Introduction
Resistance to respiratory disease in cattle requires orchestration of host defense mechanisms that protect against viral and bacterial pathogens. There is substantial evidence that active viral infection, with any of a number of bovine respiratory viruses, predisposes cattle to secondary bacterial pneumonia with Mannheimia haemolytica (formerly Pasteurella haemolytica) and Histophilus somni (formerly Haemophilus somnus) (Hodgson et al., Reference Hodgson, Aich, Manuja, Hokamp, Roche, Brinkman, Potter, Babiuk and Griebel2005). The challenge for the animal is to initiate an innate immune response that overcomes the virulence mechanisms utilized by the viral and bacterial pathogens without eliciting extensive inflammation that can compromise lung function and lead to decreased weight gain.
Results and discussion
Our laboratory has been studying events related to involvement of leukocytes and other cells in the pathogenesis of bovine respiratory disease (BRD). We have established an in vitro model for viral–bacterial synergism in which bovine bronchial epithelial cells are infected with bovine herpesvirus type 1 (BHV-1). Viral infection results in increased release of chemokines and cytokines, which in turn attract, activate and increase the adhesion of neutrophils (Rivera et al., Reference Rivera, Kisiela and Czuprynski2009). These events lead to inflammation, as seen in BRD. BHV-1 infection also results in greater adherence of M. haemolytica to bronchial epithelial cells. M. haemolytica adhesion appears to be dependent largely on two outer membrane proteins, OmpA and Lpp1 (Kisiela and Czuprynski, Reference Kisiela and Czuprynski2009). This finding is interesting in light of previous evidence that a serological response to M. haemolytica outer membrane protein correlated with resistance following immunization or natural infection (Ayalew et al., Reference Ayalew, Confer and Blackwood2004). The net result of BHV-1 infection of epithelial cells is increased adhesion of M. haemolytica under circumstances in which a vigorous inflammatory response is also initiated. Although this scenario could be beneficial in host defense against M. haemolytica, it also establishes conditions that predispose to poorly regulated inflammation in the lung. The sum of these events could lead to the intense inflammation that characterizes fibrinous pleuropneumonia, the most acute form of BRD (Hodgson et al., Reference Hodgson, Aich, Manuja, Hokamp, Roche, Brinkman, Potter, Babiuk and Griebel2005).
Once M. haemolytica is established and begins to multiply in the lung, the bacterial cells release a protein exotoxin (leukotoxin or LKT) that has potent effects on bovine leukocytes. Epithelial cells are resistant to the LKT, although they can be affected by the lipopolysaccharide in the outer envelope of M. haemolytica (McClenahan and Czuprynski, Reference McClenahan, Hellenbrand, Atapattu, Aulik, Carlton, Kapur and Czuprynski2008). Small amounts of LKT can activate Neutrophils (PMNs) and mononuclear phagocytes to release cytokines and reactive oxygen intermediates that might further exacerbate the inflammatory response in the respiratory tract (Czuprynski et al., Reference Czuprynski, Noel, Ortiz-Carranza and Srikumaran1991; Stevens and Czuprynski, Reference Stevens and Czuprynski1996). Greater amounts, or sustained exposure to LKT, impairs leukocyte function and leads to leukocyte death. These events, together with those initiated by viral infection, set the stage for a vigorous inflammatory response that might not be effective in eliminating the bacterial cells from the respiratory tract.
Formerly it was thought that the LKT exerted its effects solely at the surface of leukocytes, causing membrane lesions (pores) that led to calcium influx, efflux of macromolecules and cell death (Jeyaseelan et al., Reference Jeyaseelan, Sreevatsan and Maheswaran2002). Other evidence demonstrates that LKT binds to the CD18 chain on adhesion molecules (β2 integrins) on the surface of bovine leukocytes (Ambagala et al., Reference Ambagala, Ambagala and Srikumaran1999; Dassanayake et al., Reference Dassanayake, Maheswaran and Srikumaran2007; Shanthalingam and Srikumaran, Reference Shanthalingam and Srikumaran2009). This in turn initiates a series of intracellular events that compromise mitochondrial function and lead to release of cytochrome c. The latter in turn causes leukocytes to die by a programmed cell death pathway (apoptosis or pyroptosis) (Atapattu and Czuprynski, Reference Atapattu and Czuprynski2005). Our laboratory has shown that after binding to CD18, the LKT is internalized into bovine leukocytes and is then transported via the cytoskeleton to the mitochondria. The LKT damages the outer mitochondrial membrane, which impairs mitochondrial membrane potential and leads to cell death (Atapattu et al., Reference Atapattu, Albrecht, McClenahan and Czuprynski2008).
Evidence from other laboratories shows that viral infection can also enhance the severity of H. somni infection in cattle (Gershwin et al., Reference Gershwin, Berghaus, Arnold, Anderson and Corbeil2005). One common characteristic of H. somni infection is formation of thrombi in blood vessels (Gogolewski et al., Reference Gogolewski, Leathers, Liggitt and Corbeil1987). Our laboratory has shown that H. somni adheres to, but does not invade, bovine endothelial cells in vitro (Behling-Kelly et al., Reference Behling-Kelly, Vonderheid, Kim, Corbeil and Czuprynski2006). This attachment makes the endothelial cell surface procoagulant, which promotes thrombus formation (Behling-Kelly et al., Reference Behling-Kelly, Kim and Czuprynski2007a). H. somni also triggers cytoskeletal alterations that increase the permeability of the endothelium (Behling-Kelly et al., Reference Behling-Kelly, McClenahan, Kim and Czuprynski2007b). The effects of H. somni on endothelial cells are amplified by its interactions with blood platelets (Kuckleburg et al., Reference Kuckleburg, Sylte, Inzana, Corbeil, Darien and Czuprynski2005). Phosphorylcholine on H. somni cells is required for platelet aggregation, a process that can be blocked by inhibitors of the platelet activating factor receptor (Kuckleburg et al., Reference Kuckleburg, Elswaifi, Inzana and Czuprynski2007). Activated platelets stimulate endothelial cell expression of adhesion molecules, inflammatory cytokines and tissue factor, and cause some endothelial cells to undergo apoptosis (Kuckleburg et al., Reference Kuckleburg, McClenahan and Czuprynski2008).
The ability of bovine leukocytes to ingest and kill M. haemolytica and H. somni is limited due to properties of these bacterial pathogens (Chiang et al., Reference Chiang, Kaeberle and Roth1986; Czuprynski et al., Reference Czuprynski, Hamilton and Noel1987; Gomis et al., Reference Gomis, Godson, Beskorwayne, Wobeser and Potter1997). We have been investigating further the interaction of bovine neutrophils with M. haemolytica and H. somni. Confocal microscopy reveals that exposure to H. somni leads to redistribution of PECAM−1 on the surface of bronchial endothelial cells, and that neutrophil transmigration across H. somni-treated monolayers is reduced by anti-platelet endothelial cell adhesion molecule-1 (PECAM-1) antibodies (Tiwari et al., Reference Tiwari, Sullivan and Czuprynski2009). Recently, we investigated the role of neutrophil ‘nets’, extracellular structures composed of DNA and protein extruded from neutrophils (Wartha et al., Reference Wartha, Beiter, Normark and Henriques-Normark2007), in BRD. We find that both the M. haemolytica LKT and intact H. somni cells stimulate net formation by bovine neutrophils. We are in the process of investigating the antimicrobial activity of these nets for M. haemolytica and H. somni.
Conclusion
Collectively, the above events result in an inflammatory response to bovine BRD pathogens that is pro-inflammatory. The ability of M. haemolytica to circumvent leukocyte antibacterial activity via its LKT, and of H. somni to resist leukocyte killing while creating a pro-inflammatory and pro-coagulative environment on the endothelial cell surface, likely contribute to the intense inflammation that characterizes BRD. Efforts to reduce the severity of BRD through vaccination must take these events into account and strive to eliminate the bacterial cells without further exacerbating pulmonary inflammation.
Acknowledgments
The work done in my laboratory was totally dependent on the outstanding contributions of Dhammika Atapattu, Nicole Aulik, Erica Behling-Kelly, Dagmara Kisiela, Chris Kuckleburg, Dave McClenahan, Jose Rivera and Raksha Tiwari. Funding was provided by grants from the USDA NRI (25-6239-0117-002, 2005-01694 and 2006-01615) and the Walter and Martha Renk Endowed Laboratory for Food Safety.
