Introduction
Mannheimia haemolytica is a bovine pathogen of considerable economic importance to the global cattle industry and in particular to the North American feedlot industry. Although the organism naturally exists as a commensal of the upper respiratory tract and nasopharynx of healthy ruminants (Frank, Reference Frank, Adlam and Rutter1989; Carter et al., Reference Carter, Chengappa and Roberts1995), it is also associated with the diseased state, pneumonic pasteurellosis, and is considered the major bacterial agent of bovine respiratory disease complex. The organism may cause disease in young calves as a component of enzootic pneumonia of beef and dairy calves (Kiorpes et al., Reference Kiorpes, Butler, Dubielzig and Beck1988; Van Donkersgoed et al., Reference Van Donkersgoed, Ribble, Boyer and Townsend1993a; Ames, Reference Ames1997); however, its greatest impact occurs in recently weaned beef calves shortly after entry to feedlots (Jubb and Kennedy, Reference Jubb and Kennedy1970; Mosier et al., Reference Mosier, Confer and Panciera1989; Wilson, Reference Wilson1989). Economic losses to the North American feedlot industry due to respiratory disease have been estimated to be as high as 1 billion dollars annually (Whiteley et al., Reference Whiteley, Maheswaran, Weiss, Ames and Kannan1992). Despite improved management practices and extensive use of vaccination programs, bovine respiratory disease continues to be a major cause of losses in feedlot cattle.
Clinically, cattle suffering from M. haemolytica respiratory infections may have fever, nasal discharge, cough and respiratory distress along with inappetance and weight loss (Friend et al., Reference Friend, Wilkie, Thomson and Barnum1977). The major cause of death is acute fibrinous pleuropneumonia. Necropsy findings include obstruction of bronchioles with fibrinous exudate, accumulation of neutrophils, macrophages and fibrin in the alveoli, and thrombosis and distention of lymphatic vessels.
The organism
M. haemolytica is a Gram-negative, non-motile, non-spore-forming, fermentative, oxidase-positive, facultative anaerobic coccobacillus (Quinn, Reference Quinn1994; Hirsh and Zee, Reference Hirsh and Zee1999). The organism is a member of the family Pasteurellaceae, genus Mannheimia. First named Bacterium bipolare multocidum by Theodore Kitt (Kitt, Reference Kitt1885), it was renamed Pasteurella haemolytica in 1932, to reflect its weakly hemolytic phenotype on sheep's blood agar plates (Newsome and Cross, Reference Newsome and Cross1932) and was historically classified into 16 serotypes, based on an indirect hemagglutination test using extractable capsular surface antigens (Biberstein, Reference Biberstein, Bergan and Norris1978). The genus was further divided into distinct biotypes (A and T) based on the ability to ferment arabinose or trehalose, respectively (Smith, Reference Smith1961; Lo and Shewen, Reference Lo, Shewen, Balows, Truper, Dworkin, Harder and Schliefer1991). Twelve A serotypes and four T serotypes were identified. Later, Younan and Fodor (Reference Younan and Fodor1995) characterized a new serotype of M. haemolytica, A17, isolated from sheep in Syria. Through DNA–DNA hybridization studies and 16S RNA sequencing, all but one of the A biotypes were assigned the species designation M. haemolytica (Angen et al., Reference Angen, Mutters, Caugant, Olsen and Bisgaard1999a). P. haemolytica T biotypes were renamed Pasteurella trehalosi (Bingham et al., Reference Bingham, Moore and Richards1990; Sneath and Stevens, Reference Sneath and Stevens1990), then Bibersteinia trehalosi (Blackall et al., Reference Blackall, Bojesen, Christensen and Bisgaard2007). The remaining A11 serotype was renamed Mannheimia glucosida (Angen et al., Reference Angen, Quirie, Donachie and Bisgaard1999b).
M. haemolytica comprises 12 capsular serotypes based on those originally assigned to P. haemolytica (A1, A2, A5, A6, A7, A8, A9, A12, A13, A14, A16 and A17) (Angen et al., Reference Angen, Quirie, Donachie and Bisgaard1999b). Both serotypes A1 and A2 colonize the upper respiratory tract of cattle and sheep. Pneumonia in cattle is mainly associated with isolation of serotype A1 from lungs at necropsy (Frank and Smith, Reference Frank and Smith1983; Allan et al., Reference Allan, Wiseman, Gibbs and Selman1985), even though healthy cattle frequently carry both serotypes A1 and A2 in the nasopharynx. Davies et al. (Reference Davies, Whittam and Selander2001) state that `serotype A1 and A6 strains account for almost all cases of bovine pneumonic pasteurellosis', and recent surveys in Germany (Ewers et al., Reference Ewers, Lubke-Becker and Wieler2004) and the USA (Purdy et al., Reference Purdy, Raleigh, Collins, Watts and Straus1997; Al-Ghamdi et al., Reference Al-Ghamdi, Ames, Baker, Walker, Chase, Frank and Maheswaran2000) document that serotype A6 constitutes 30% of the total number of serotyped isolates. Interestingly, apart from the capsule structure, serotype A1 and A6 are extremely similar structurally (Davies and Donachie, Reference Davies and Donachie1996; Morton et al., Reference Morton, Simons and Confer1996). Serotype A2 is a common cause of pneumonic pasteurellosis in sheep (Shewen and Conlon, Reference Shewen, Conlon, Gyles and Thoen1993) but there is also an increased prevalence of serotypes A5, A6 and A7 (Ewers et al., Reference Ewers, Lubke-Becker and Wieler2004).
M. haemolytica resides in the nasopharynx (Babiuk and Acres, Reference Babiuk, Acres and Loan1984) and tonsils (Frank and Briggs, Reference Frank and Briggs1992; Frank et al., Reference Frank, Briggs and Zehr1995) of healthy calves. As a commensal organism, M. haemolytica inhabits the nasopharynx and maintains a symbiotic relationship with its host; however, key inciting events including stress of weaning, adverse weather conditions, changes in feed, transportation over long distances, mixing of cattle and infection with other microorganisms (including viruses and mycoplasma sp.) cause this benign coexistence to become a fulminating disease state (Farley, Reference Farley1932; Blood et al., Reference Blood, Radostits and Henderson1983).
Tonsillar tissue has been identified as a reservoir for M. haemolytica (Frank and Briggs, Reference Frank and Briggs1992; Frank et al., Reference Frank, Briggs and Zehr1995). Calves may be negative for M. haemolytica on culture of nasal swabs, but positive on culture of the tonsils (Frank et al., Reference Frank, Briggs, Loan, Purdy and Zehr1994). Not only does the frequency of isolation of M. haemolytica A1 increase as calves move to the feedlot, but the number of bacteria also increases rapidly (Frank, Reference Frank and Loan1984). When high numbers of M. haemolytica are present on the nasal mucosa of calves, the bacteria are inhaled into the lungs (Grey and Thomson, Reference Grey and Thomson1971). In healthy calves, lung clearance of inhaled M. haemolytica is highly efficient, with elimination of 90% of an administered dose within 4 h (Lillie and Thomson, Reference Lillie and Thomson1972) (Fig. 1A).
Fig. 1. Host–bacterium–environment interactions in Mannheimia haemolytica pneumonia of cattle. (A) Homeostasis in the commensal state; (B) environmental factors tip the balance to favor lung colonization and disease; (C) intervention strategies restore homeostasis.
Several studies have attempted to understand the mechanisms associated with the switch from commensal to pathogen. Cold stress (chilling of calves with cold water) and transportation were studied for their effect on immunosuppression. Cold stress was demonstrated to transiently increase plasma cortisol levels, but had no effect on histamine or bradykinin levels (Slocombe et al., Reference Slocombe, Derksen, Robinson, Trapp, Gupta and Newman1984). Transportation was found to cause a transient elevation of plasma cortisol levels, suppressing in vitro lymphocyte blastogenesis (Filion et al., Reference Filion, Willson, Bielfeldt-Ohmamn, Babiuk and Thomson1984). Serum complement activity has also been noted to be decreased in calves that are purchased at auction and moved into the feedlot (Purdy et al., Reference Purdy, Richards and Foster1991). These alterations in immune function could impact the ability of the calf to maintain normal homeostasis with the commensal organism.
The effects of mixing of calves (during transportation, in auction barns or after sorting at the feedlot) on the prevalence of respiratory disease may result from the stresses of interacting with strange calves or may reflect increased opportunities to contact infectious agents (Jericho, Reference Jericho1979). Viral and bacterial agents break down the antimicrobial barrier of β-defensins, anionic peptides and serous and mucous secretions of the respiratory tract, allowing M. haemolytica to be released from its commensal status (Brogden et al., Reference Brogden, Lehmkuhl and Cutlip1998). Infection with bovine herpes virus 1 (BHV-1) parainfluenza virus 3 (PI-3) and bovine viral diarrhea virus (BVDV) leads to proliferation of the bacteria in the nasopharynx, interferes with normal clearance from the lungs (Lopez et al., Reference Lopez, Thomson and Savan1976) and impairs the ciliary activity of epithelial cells in the trachea (Rossi and Kiesel, Reference Rossi and Kiesel1977). BVDV infection also leads to impaired neutrophil and lymphocyte functions (Brown et al., Reference Brown, Bolin, Frank and Roth1991) predisposing to bacterial pneumonia. Immunosuppression in the case of bovine respiratory syncytial (BRS) virus (Woldehiwet and Sharma, Reference Woldehiwet and Sharma1992) and BHV-1 diminishes the activity of T lymphocytes, B lymphocytes, monocytes and macrophages (Brown and Ananaba, Reference Brown and Ananaba1988). All of these factors tend to impair innate immune defenses and can provide an opportunity for the organism to gain access to deeper structures of the respiratory tract (Ribble et al., Reference Ribble, Meek, Shewen, Guichon and Jim1995a) (Fig. 1B).
Bacterial virulence factors
Multiple products and components of M. haemolytica A1 have been proposed as virulence factors, including an adhesin (Jaramillo et al., Reference Jaramillo, Diaz, Hernandez, Debray, Trigo, Mendoza and Zenteno2000), capsular polysaccharide (Confer et al., Reference Confer, Simons, Panciera, Mort and Mosier1989; Conlon and Shewen, Reference Conlon and Shewen1993; Brogden et al., Reference Brogden, Debey, Audibert, Lehmkuhl and Chedid1995), fimbriae (Morck et al., Reference Morck, Raybould, Acres, Babiuk, Nelligan and Costerton1987, Reference Morck, Olson, Acres, Daoust and Costerton1989), iron-regulated outer membrane proteins (OMPs) (Squire et al., Reference Squire, Smiley and Croskell1984; Morck et al., Reference Morck, Ellis, Domingue, Olson and Costerton1991; Gatewood et al., Reference Gatewood, Fenwick and Chengappa1994), leukotoxin (Lkt) (Gentry et al., Reference Gentry, Confer and Panciera1985; Shewen and Wilkie, Reference Shewen and Wilkie1985, Reference Shewen and Wilkie1988), lipopolysaccharide (LPS) (Confer and Simons, Reference Confer and Simons1986), lipoproteins (Cooney and Lo, Reference Cooney and Lo1993; Nardini et al., Reference Nardini, Mellors and Lo1998), neuraminidase (Frank and Tabatabai, Reference Frank and Tabatabai1981; Straus and Purdy, Reference Straus and Purdy1994; Straus et al., Reference Straus, Purdy, Loan, Briggs and Frank1998), a serotype-specific antigen (Gonzalez-Rayos et al., Reference Gonzalez-Rayos, Lo, Shewen and Beveridge1986; Lo et al, Reference Lo, Strathdee, Shewen and Cooney1991), sialoglycoprotease (Abdullah et al., Reference Abdullah, Udoh, Shewen and Mellors1992; Lee et al., Reference Lee, Shewen, Cladman, Conlon, Mellors and Lo1994) and transferrin-binding proteins (Ogunnariwo and Schryvers, Reference Ogunnariwo and Schryvers1990; Potter et al., Reference Potter, Schryvers, Ogunnariwo, Hutchins, Lo and Watts1999). The adhesin protein, capsular polysaccharide, fimbriae, sialoglycoprotease and neuraminidase may have roles in the attachment of M. haemolytica and its colonization of cells of the respiratory tract of calves (Babiuk and Acres, Reference Babiuk, Acres and Loan1984; Morck et al., Reference Morck, Watts, Acres and Costerton1988, Reference Morck, Olson, Acres, Daoust and Costerton1989; Whiteley et al., Reference Whiteley, Maheswaran, Weiss, Ames and Kannan1992). The capsular polysaccharide also has antiphagocytic properties (Chae et al., Reference Chae, Gentry, Confer and Anderson1990) and increases directed migration of neutrophils (Czuprynski et al., Reference Czuprynski, Noel and Adlam1989). The sialoglycoprotease, by cleaving bovine IgG1, may reduce the effectiveness of opsonizing antibodies (Lee and Shewen, Reference Lee and Shewen1996). Transferrin-binding proteins and other iron-regulated proteins enable M. haemolytica to proliferate in vivo in spite of the low iron environment normally maintained by the host. LPS as an inducer of inflammation has a central role in the development of vascular lesions in lung tissue (Whiteley et al., Reference Whiteley, Maheswaran, Weiss, Ames and Kannan1992).
The virulence factor believed to be pivotal in the pathogenesis of M. haemolytica is the ruminant-specific Lkt. The effects of Lkt range from impairment of function to lysis of ruminant leukocytes (Baluyut et al., Reference Baluyut, Simonson, Bemrick and Maheswaran1981; Berggren et al., Reference Berggren, Baluyut, Simonson, Bemrick and Maheswaran1981; Shewen and Wilkie, Reference Shewen and Wilkie1982; Gentry et al., Reference Gentry, Confer and Panciera1985; Clinkenbeard et al., Reference Clinkenbeard, Mosier and Confer1989; DeBey et al., Reference DeBey, Roth, Brogden, Cutlip, Stevens, Briggs and Kluge1996). Lkt is a heat labile protein, actively secreted by all serotypes of the bacterium during the logarithmic phase of in vitro growth (Shewen and Wilkie, Reference Shewen and Wilkie1985). It is a member of the RTX (repeats in toxin) family of multidomain exotoxins (Lo, Reference Lo1990) and contains six highly conserved regions, glycine-rich nonapeptide repeats, near the C-terminal end of its structure (Lo, Reference Lo1990; Coote, Reference Coote1992; Jeyaseelan et al., Reference Jeyaseelan, Sreevatsan and Maheswaran2002). Strathdee and Lo (Reference Strathdee and Lo1987) found Lkt to be very similar genetically to Escherichia coli α-hemolysin, and Lkt has been shown to lyse sheep erythrocytes (Murphy et al., Reference Murphy, Whitworth, Clinkenbeard and Clinkenbeard1995). Like the α-hemolysin, the Lkt is encoded by four genes in the RTX toxin operon, designated C, A, B and D (Lo et al., Reference Lo, Shewen, Strathdee and Greer1985, Reference Lo, Strathdee and Shewen1987; Strathdee and Lo, Reference Strathdee and Lo1987). The A gene (lktA) codes for the structural toxin. The product of the C gene is involved in toxin activation in conjunction with an acyl carrier protein. The product of the lktA gene is biologically inactive until modified post-translationally by fatty acid acylation (Issartel et al., Reference Issartel, Koronaskis and Hughes1991). The products of B and D genes are involved in toxin secretion (Chang et al., Reference Chang, Young, Post and Struck1989; Highlander et al., Reference Highlander, Chidambaram, Engler and Weinstock1989; Strathdee and Lo, Reference Strathdee and Lo1989).
Through sequence analysis of the lkt genes and studies of polymorphism, it has been shown that different serotypes produce different Lkt types and that LktA may vary between bovine and ovine isolates of the same serotype (Davies et al., Reference Davies, Campbell and Whittam2002). Serotypes A1, A5, A6, A8, A9 and A12 have very similar Lkts (LktA1.1, LktA1.2 and LktA1.3), whereas serotype A2 isolates may express any of four Lkt types (LktA2, LktA3, LktA8 and LktA10) (Davies et al., Reference Davies, Whittam and Selander2001; Davis and Baillie, 2003). Even though these differences exist, polyclonal antibodies raised to one Lkt cross-neutralize Lkt produced by other serotypes, although antisera neutralize homologous Lkt more efficiently (Shewen and Wilkie, Reference Shewen and Wilkie1983; Lainson et al., Reference Lainson, Murray, Davis and Donachie1996). The epitope associated with neutralizing activity has been identified for M. haemolytica serotype 1 and is located at the C-terminal end of Lkt A (Lainson et al., Reference Lainson, Murray, Davis and Donachie1996).
Once the bacterium gains entry into deeper respiratory structures, Lkt plays a major role in lung injury and also in allowing bacteria to survive by evading phagocytic cell destruction. Tatum et al. (Reference Tatum, Briggs, Sreevatsan, Zehr, Hsuan, Whiteley, Ames and Maheswaran1998) generated M. haemolytica deficient in Lkt through gene knock-out. Although the mutant retained the ability to colonize the upper respiratory tract, it could not induce lung lesions. Fedorova and Highlander (Reference Fedorova and Highlander1997) created a mutant strain that secreted an antigenic proLkt that was not leukotoxic or hemolytic, confirming that the lktC gene was required for activation of proLkt to mature Lkt. Later, Highlander et al. (Reference Highlander, Fedorova, Dusek, Panciera, Alvarez and Rinehart2000) demonstrated that a strain that secretes inactive Lkt had attenuation of virulence in a calf model.
RTX toxins attach to cells through passive adsorption, which does not always lead to cell lysis, and through specific cell surface receptors. The receptor for Lkt has been identified as the transmembrane receptor CD18, the constant β-subunit of the β2-integrin family. CD18 complexed with CD11a forms the lymphocyte-function-associated antigen 1 (LFA-1), which is responsible for the high affinity adhesion of Lkt to ruminant leukocytes and platelets (Bailly et al., Reference Bailly, Tontti, Hermand, Cartron and Gahmberg1995; Gahmberg, Reference Gahmberg1997; Lally et al., Reference Lally, Kieba, Sato, Green, Rosenbloom, Korostoff, Wang, Shenker, Otrlepp, Robinson and Billings1997; Wang (JF), Reference Wang, Kieba, Korostoff, Guo, Yamaguchi, Rozmiarek, Billings, Shenker and Lally1998a; Ambagala et al., Reference Ambagala, Ambagala and Srikumaran1999; Li et al., Reference Li, Clinkenbeard and Ritchey1999; Deshpande et al., Reference Deshpande, Ambagala, Ambagala, Kehrli and Srikumaran2002; Berman et al., Reference Berman, Kozlova and Morozevich2003; Dassanayake et al., Reference Dassanayake, Maheswaran and Srikumaran2007). The repeating nanopeptide of the LktA molecule is the ligand for LFA-1 on host cells (Lally et al., Reference Lally, Hill, Kieba and Korostoff1999).
The effect of Lkt on bovine cells is dose-dependent. At very low concentrations, the toxin activates target cells triggering respiratory burst and degranulation. As the concentration of Lkt is increased, target cells are stimulated to undergo apoptosis (Lally et al., Reference Lally, Hill, Kieba and Korostoff1999). At high Lkt concentrations, necrosis of target cells occurs as a consequence of the formation of pore-like structures in the plasma membrane (Clarke et al., Reference Clarke, Confer and Mosier1998), which leads to K+ efflux and Ca2+ influx, colloidal osmotic swelling and eventual cell lysis (Orrenius et al., Reference Orrenius, Zhivotovsky and Nicotera2003). The size of the pore varies among RTX toxins; in the case of M. haemolytica Lkt, the transmembrane pore is 0.6–1.0 nm in diameter (Clinkenbeard et al., Reference Clinkenbeard, Mosier and Confer1989). Extensive in vitro studies have been successful in reproducing neutrophil necrosis utilizing purified Lkt (Wang (Z), Reference Wang, Clarke and Clinkenbeard1998b; Ambagala et al., Reference Ambagala, Ambagala and Srikumaran1999; Sun et al., Reference Sun, Clinkenbeard, Cudd, Clarke and Clinkenbeard1999, Reference Sun, Clinkenbeard, Ownby, Cudd, Clarke and Highlander2000; Jeyaseelan et al., Reference Jeyaseelan, Hsuan, Kannan, Walcheck, Wang, Kehrli, Lally, Sieck and Maheswaran2000, Reference Jeyaseelan, Kannan, Briggs, Thumbikat and Maheswaran2001; Cudd et al., Reference Cudd, Ownby, Clarke, Sun and Clinkenbeard2001; Davies and Baillie, Reference Davies and Baillie2003). At subcytolytic concentrations, Lkt enhances the inflammatory response by activating cells to produce mediators and release reactive oxygen metabolites and proteases. The lesions seen in infected lungs, including fibrinous exudate and thrombosis of lymphatic vessels, result, in part, from effects of the toxin on neutrophils (Slocombe et al., Reference Slocombe, Malark, Ingersoll, Derksen and Robinson1985; Breider et al., Reference Breider, Walker, Hopkins, Schultz and Bowersock1988) and from lysis of platelets (Clinkenbeard and Upton, Reference Clinkenbeard and Upton1991).
Another critical virulence factor is the lipid A component of the LPS of the cell wall of the organism. The lipid A fraction is responsible for endotoxic effects, such as pyrexia, macrophage activation, release of tumor necrosis factor and induction of hypotensive shock (Keiss et al., Reference Keiss, Will and Collier1964), and plays a role in the vascular lesions seen in diseased lung tissue (Whiteley et al., Reference Whiteley, Maheswaran, Weiss, Ames and Kannan1992). In addition, LPS forms complexes with Lkt, which may enhance cytotoxicity (Li and Clinkenbeard, Reference Li and Clinkenbeard1999). Although LPS has been reported to be a major antigenic determinant (Confer et al., Reference Confer, Panciera, Gentry and Fulton1986), antibody titers to LPS do not correlate with resistance to experimental pneumonia (Confer et al., Reference Confer, Simons, Panciera, Mort and Mosier1989; Mosier et al., Reference Mosier, Simons and Vestweber1995).
Lipoproteins have also been identified (Cooney and Lo, Reference Cooney and Lo1993; Nardini et al., Reference Nardini, Mellors and Lo1998) and are present in most serotypes. A surface-exposed 45-kDa OMP, designated PlpE, was sequenced and cloned by Pandher et al. (Reference Pandher, Murphy and Confer1999) . Mosier et al. (Reference Mosier, Confer and Panciera1989) reported this protein to be immunogenic in cattle and Pandher et al. (Reference Pandher, Confer and Murphy1998) found that antibodies to PlpE were associated with complement-mediated killing of M. haemolytica. A recombinant PlpE was highly immunogenic when injected subcutaneously in vaccination studies (Confer et al., Reference Confer, Ayalew, Panciera, Montelongo, Whitworth and Hammer2003). The same group demonstrated that the surface-exposed immunodominant epitope between amino acids 26 and 76 conferred protection from challenge (Ayalew et al., Reference Ayalew, Confer and Blackwood2004). A recent report from these researchers suggests that addition of recombinant PlpE to existing commercial vaccines enhances protection against experimental challenge (Confer et al., Reference Confer, Ayalew, Panciera, Montelongo and Wray2006).
Acute pulmonary infection in feedlot cattle is characterized by a fibrinosuppurative and necrotizing inflammatory response. Parenchymal necrosis is most likely caused by Lkt and LPS, as well as inflammatory factors released by neutrophils and other cells of the acute inflammatory process. Neutrophil infiltration during M. haemolytica pneumonia is associated with alveolar epithelial cell damage and necrosis. Slocombe et al. (Reference Slocombe, Malark, Ingersoll, Derksen and Robinson1985) demonstrated the contribution of neutrophils to parenchymal damage of the lung in pasteurellosis. Depletion of neutrophils prior to inoculation with the bacteria protects calves from the gross fibrinopurulent pneumonic and pleuritic lesions (Weiss et al., Reference Weiss, Bauer, Whiteley, Maheswaran and Ames1991; Ulevitch and Tobias, Reference Ulevitch and Tobias1995) but less severe changes still occur (Breider et al., Reference Breider, Walker, Hopkins, Schultz and Bowersock1988). Pathognomonic for bovine pneumonic pasteurellosis is necrosis of the alveolar epithelium due to the strong influx of neutrophils and accumulation of fibrin in the lungs. Depending on the size and distribution of the fibronecrotizing lesions, this pneumonia may result in death.
Feedlot management practices and therapeutic intervention
It has been questioned whether pneumonic pasteurellosis in feedlot calves should be considered a highly contagious disease (Thomson, Reference Thomson and Loan1984). Although morbidity rates as high as 69% have been reported in the first weeks after feedlot arrival (Kelly and Janzen, Reference Kelly and Janzen1986), it has been observed that ‘fibrinous pneumonia does not sweep through the feedlot like an epizootic but rather it centres on certain pens’ (Thomson, Reference Thomson and Loan1984). This would suggest that characteristics of the calves in the various pens rather than mere exposure to M. haemolytica are critical in determining disease outcomes. In an experimental study, 50% of non-challenged control calves (naive to M. haemolytica) in contact with animals challenged with M. haemolytica developed clinical signs of respiratory disease and responded serologically (Gibbs et al., Reference Gibbs, Allan, Wiseman and Selman1984), but it is unclear how well this reflects transmission under field conditions.
In reality, pneumonic pasteurellosis is a management disease resulting from an incompatibility between the biology of calves (and their pathogens) and the managing and/or marketing systems devised by humans. Calves that move directly from a ranch into feedlots without moving through saleyards, and without mixing with calves from other sources, have an expected morbidity of less than 5% (Radostits et al., Reference Radostits, Blood and Gay1994). While vaccines and antibiotics can be useful in controlling pneumonia, basic changes in when and how calves are weaned, sold and transported to feedlots would have a profound impact on the prevalence of disease.
Antibiotics
Injectable antibiotics are employed extensively in the feedlot industry in North America in attempts to prevent and treat bovine respiratory disease. Antibiotics are rarely selected on the basis of in vitro sensitivity of isolates from nasal or pharyngeal swabs. Isolates from these sources do not accurately reflect organisms present in the lower respiratory tract of the same animal (Allen et al., Reference Allen, Viel, Bateman, Rosendal, Shewen and Physick-Sheard1991). Necropsy specimens from antibiotic-treated calves also do not provide reliable information as to the antibiotic sensitivities of the organisms initiating the pneumonia. Most published studies examining the efficacy of antibiotics against disease in feedlots make no attempt to distinguish pneumonic pasteurellosis from respiratory conditions caused by other bacteria or by viruses (Mechor et al., Reference Mechor, Jim and Janzen1988; Libersa et al., Reference Libersa, Van Huffel and Madelenat1995; Vogel et al., Reference Vogel, Laudert, Zimmermann, Guthrie, Mechor and Moore1998). The term ‘undifferentiated fever’ has been considered synonymous with ‘bovine respiratory disease complex’ by some workers (Jim et al., Reference Jim, Booker, Guichon, Schunicht, Wildman, Johnson and Lockwood1999; Booker et al., Reference Booker, Abutarbush, Schunicht, Jim, Perrett, Wildman, Guichon, Pittman, Jones and Pollock2007; Schunicht et al., Reference Schunicht, Booker, Guichon, Jim, Wildman, Pitman and Perrett2007). Thus a succession of antimicrobial drugs have been examined for efficacy against undifferentiated respiratory disease under field conditions, including penicillin (Mechor et al., Reference Mechor, Jim and Janzen1988; Bateman et al., Reference Bateman, Martin, Shewen and Menzies1990), oxytetracycline (Mechor et al., Reference Mechor, Jim and Janzen1988; Bateman et al., Reference Bateman, Martin, Shewen and Menzies1990; Harland et al., Reference Harland, Jim, Guichon, Townsend and Janzen1991), trimethoprim/sulfadoxine (Mechor et al., Reference Mechor, Jim and Janzen1988; Bateman et al., Reference Bateman, Martin, Shewen and Menzies1990; Harland et al., Reference Harland, Jim, Guichon, Townsend and Janzen1991), ampicillin (Bentley and Cummins, Reference Bentley and Cummins1987; Libersa et al., Reference Libersa, Van Huffel and Madelenat1995), tilmicosin (Gorham et al., Reference Gorham, Carroll, McAskill, Watkins, Ose, Tonkinson and Merrill1990; Schumann et al., Reference Schumann, Janzen and McKinnon1990; Hoar et al., Reference Hoar, Jelinski, Ribble, Janzen and Johnson1998), florfenicol (Libersa et al., Reference Libersa, Van Huffel and Madelenat1995; Hoar et al., Reference Hoar, Jelinski, Ribble, Janzen and Johnson1998; Jim et al., Reference Jim, Booker, Guichon, Schunicht, Wildman, Johnson and Lockwood1999) and tulathromycin (Booker et al., Reference Booker, Abutarbush, Schunicht, Jim, Perrett, Wildman, Guichon, Pittman, Jones and Pollock2007; Schunicht et al., Reference Schunicht, Booker, Guichon, Jim, Wildman, Pitman and Perrett2007; Wellman and O'Connor, Reference Wellman and O'Connor2007). Although all of these antimicrobials have been efficacious for treatment of bovine respiratory disease, many isolates of M. haemolytica are now resistant to penicillin, ampicillin, tetracycline, sulfonamides and tilmicosin (Bateman, Reference Bateman, Prescott and Baggot1993; Watts et al., Reference Watts, Yancey, Salmon and Case1994; Apley, Reference Apley1997; Welsh et al., Reference Welsh, Dye, Payton and Confer2004). Penicillin and oxytetracycline have been used at higher dosage rates than recommended by the manufacturers in attempts to improve efficacy without resorting to the use of more expensive antibiotics (Bateman, Reference Bateman, Prescott and Baggot1993), prompting concerns about antibiotic residues in meat (Mechor et al., Reference Mechor, Jim and Janzen1988). The use of antibiotics at dosage rates, or by routes, or in species not approved by governmental agencies, has led to legislation in the USA to curtail ‘extra-label’ use of antibiotics in food-producing animals (Apley, Reference Apley1997). The availability of ceftiofur (1988), tilmicosin (1992), florfenicol (1996) and most recently tulathromycin (2005) for use in cattle has provided several drugs efficacious at label dosage rates against bovine respiratory pathogens (Bateman, Reference Bateman, Prescott and Baggot1993; Apley, Reference Apley1997; Hoar et al., Reference Hoar, Jelinski, Ribble, Janzen and Johnson1998; Jim et al., Reference Jim, Booker, Guichon, Schunicht, Wildman, Johnson and Lockwood1999; Schunicht et al., Reference Schunicht, Booker, Guichon, Jim, Wildman, Pitman and Perrett2007), at least for the time being. Current use of antibiotics in treatment of bovine respiratory disease has been reviewed (Apley, Reference Apley, Howard and Smith1999). Product formulations providing sustained blood levels of antibiotics for 48–72 h or more [oxytetracyclines (Bateman, Reference Bateman, Prescott and Baggot1993), experimental sustained release ceftiofur (Kesler and Bechtol, Reference Kesler and Bechtol1999) and high-dose florfenicol (Varma et al., Reference Varma, Lockwood, Cosgrove and Rogers1998)] are especially valued for use in feedlots as a means of reducing handling and treatment stresses of sick animals and of minimizing labor costs.
In a meta-analysis of 107 field trials of prophylactic mass medication of feedlot cattle, it was concluded that the parenteral administration of tilmicosin or long-acting oxytetracycline preparations on arrival was associated with significant reductions in morbidity (Van Donkersgoed, Reference Van Donkersgoed1992). Available data were considered inadequate to judge the efficacy of mass medication administered in water or feed. Another meta-analysis of 14 field trials found that treatment of bovine respiratory disease with tulathromycin was associated with approximately 50% reduction in the risk of re-treatment compared to treatment with tilomicosin (Wellman and O'Connor, Reference Wellman and O'Connor2007). Several investigators have promoted the concept of ‘metaphylactic’ (Young, Reference Young1995; Jim et al., Reference Jim, Booker, Guichon, Schunicht, Wildman, Johnson and Lockwood1999) use of parenteral antibiotics, meaning the use of mass medication at therapeutic doses before overt signs of disease are evident (Vogel et al., Reference Vogel, Laudert, Zimmermann, Guthrie, Mechor and Moore1998, Booker et al., Reference Booker, Abutarbush, Schunicht, Jim, Perrett, Wildman, Guichon, Pittman, Jones and Pollock2007). A single dose (most commonly oxytetracycline, tilmicosin or tulathromycin) is administered to all calves on arrival, or at some later time point chosen on the basis of observed signs of disease in a minority of animals or on the basis of historical patterns of disease. This practice is considered standard procedure in some parts of North America (Jim et al., Reference Jim, Booker, Guichon, Schunicht, Wildman, Johnson and Lockwood1999). With current concerns about the development of antibiotic-resistant bacteria through excessive agricultural use of antibiotics, antibiotic residues in food products, and animal welfare, it is imperative that efficacious means that are not antibiotic-dependent be developed to prevent pneumonic pasteurellosis.
Vaccines
Vaccines intended for prevention of respiratory disease in feedlot cattle have been manufactured for over a century (Mosier et al., Reference Mosier, Confer and Panciera1989). Commercial vaccines available in Canada (Bowland and Shewen, Reference Bowland and Shewen2000) and the USA (Hjerpe, Reference Hjerpe1990) for prevention of bovine respiratory disease have been reviewed. Evaluation of vaccine efficacy in beef calves involves considerable difficulties (Martin, Reference Martin1989; Ribble, Reference Ribble1989), in part because of difficulties in defining objective outcome measures (Ribble, Reference Ribble1989), in part because of difficulties in achieving adequate group sizes to obtain acceptable statistical power (Wilson, Reference Wilson1989), and in part because of the unpredictability of morbidity and mortality rates from year to year (Van Donkersgoed et al., Reference Van Donkersgoed, Potter, Mollinson and Harland1994; Ribble et al., Reference Ribble, Meek, Jim and Guichon1995b). It is considered that a single field trial is inadequate to assess the economic impact of a particular vaccine on respiratory disease in feedlots (Wilson, Reference Wilson1989).
Early bacterins were prepared from cultures of Bacillus boviseptica (which has since been divided into the species M. haemolytica and Pasteurella multocida). Following field trials, their efficacy was questioned (Miller et al., Reference Miller, Howard, Bayard, Smith, Stanard, Jones, Hilton, Killham and Truam1927) or they were found to have detrimental effects (Farley, Reference Farley1932). More recently, bacterins prepared from M. haemolytica have been reported to have no clear beneficial effect (Hamdy et al., Reference Hamdy, King and Trapp1965; Martin, Reference Martin1983) or to have detrimental effects (Schipper and Kelling, Reference Schipper and Kelling1971; Friend et al., Reference Friend, Wilkie, Thomson and Barnum1977; Wilkie et al., Reference Wilkie, Markham and Shewen1980). Live M. haemolytica vaccines have been investigated (Mosier et al., Reference Mosier, Confer and Panciera1989; Babiuk and Campos, Reference Babiuk, Campos and Peters1993) and marketed (Hjerpe, Reference Hjerpe1990), but protection studies have yielded variable results (Smith et al., Reference Smith, Davidson and Henry1985; Purdy et al., Reference Purdy, Livingston and Frank1986; Mosier et al., Reference Mosier, Panciera, Rogers, Uhlich, Butine, Confer and Basaraba1998). Use of antibiotics in cattle vaccinated with live M. haemolytica (within 7 days after or 3 days before vaccination) is not recommended, due to inhibitory effects on replication of vaccine organisms (Hjerpe, Reference Hjerpe1990). The widespread use of antibiotics in calves at feedlot arrival makes the use of live bacterial vaccines impractical.
Identification of the importance of the Lkt, produced during logarithmic phase growth, in the pathogenesis of pneumonic pasteurellosis (Shewen and Wilkie, Reference Shewen and Wilkie1982, Reference Shewen and Wilkie1985; Gentry et al., Reference Gentry, Confer and Panciera1985) led to the development of a commercial, cell-free, M. haemolytica A1 culture supernatant vaccine (Shewen and Wilkie, Reference Shewen and Wilkie1988; Shewen et al., Reference Shewen, Sharp and Wilkie1988). In experimental trials using two doses of vaccine at an interval of 21 days, efficacy against moderate to severe pneumonia ranged from 60 to 70% (Shewen et al., Reference Shewen, Sharp and Wilkie1988). Results from field trials were variable (Bateman, Reference Bateman1988; Jim et al., Reference Jim, Guichon and Shaw1988; Thorlakson et al., Reference Thorlakson, Martin and Peters1990). Numerous other antigens of M. haemolytica have been documented in logarithmic phase culture supernatant, including surface antigens involved in agglutination reactions (Shewen and Wilkie, Reference Shewen and Wilkie1988), capsular polysaccharide (Conlon and Shewen, Reference Conlon and Shewen1993), lipoproteins (Cooney and Lo, Reference Cooney and Lo1993) and sialoglycoprotease (Abdullah et al., Reference Abdullah, Udoh, Shewen and Mellors1992), and these antigens may contribute to vaccine efficacy.
Since this initial M. haemolytica subunit vaccine, various companies have introduced other subunit or subunit-enriched vaccines consisting of outer membrane extracts (Srinand et al., Reference Srinand, Hsuan, Yoo, Maheswaran, Ames and Werdin1996), or extracted antigens enriched with recombinant Lkt (Babiuk and Campos, Reference Babiuk, Campos and Peters1993; Van Donkersgoed et al., Reference Van Donkersgoed, Schumann, Harland, Potter and Janzen1993b, Reference Van Donkersgoed, Potter, Mollinson and Harland1994, Reference Van Donkersgoed, Guenther, Evans, Potter and Harland1995), or bacterins with added culture supernatant antigens (Babiuk and Campos, Reference Babiuk, Campos and Peters1993; Srinand et al., Reference Srinand, Hsuan, Yoo, Maheswaran, Ames and Werdin1996). Most recently (1999), a vaccine based on antigens expressed under iron-limiting conditions has been licensed (Gilmour et al., Reference Gilmour, Donachie, Sutherland, Gilmour, Jones and Quirie1991; Donachie, Reference Donachie1999). Serological responses to selected commercial M. haemolytica vaccines have been compared (Srinand et al., Reference Srinand, Hsuan, Yoo, Maheswaran, Ames and Werdin1996; Confer et al., Reference Confer, Fulton, Clinkenbeard and Driskell1998).
Recent studies have focused on the induction of immunity to M. haemolytica by mucosal delivery of antigen. Induction of local immunity in the nasopharynx could reduce colonization by Pasteurella spp. Viral vectors (replicating in the upper respiratory tract) expressing antigens of M. haemolytica (Babiuk and Campos, Reference Babiuk, Campos and Peters1993), oral administration of M. haemolytica antigens encapsulated in alginate microspheres (Bowersock et al., Reference Bowersock, Hogenesch, Suckow, Guimond, Martin, Borie, Torregrosa, Park and Park1999) and oral administration of modified live M. haemolytica are under consideration (Briggs and Tatum, Reference Briggs and Tatum1999). Researchers have also engineered plants for expression of M. haemolytica antigens, with the ultimate aim of producing a transgenic edible vaccine (Lee et al., Reference Lee, Strommer, Hodgins, Shewen, Niu and Lo2001)
Conclusion
Given the complex etiology of respiratory disease attributed to M. haemolytica, it is unlikely that any single strategy will be completely effective in preventing disease. A combination of more definitive diagnostic methods, more efficacious vaccines, improved therapeutic agents and more rational management practices will be needed (Fig. 1C) and the search for solutions is likely to involve veterinary researchers for some time to come.
