John A. Angelos


Moraxella (phylum: Proteobacteria; class: Gammaproteobacteria; order: Pseudomonadales; family: Moraxellaceae) is a genus of Gram‐negative bacteria for which 20 species are currently recognized; seven are derived from human and 13 from animal sources (Parte et al. 2020; Schoch et al. 2020). A summary of Moraxella species isolated from animals and their characteristic morphology is presented in Table 15.1. Five species formerly designated in the genus Moraxella have been reclassified as other genera; one of these arose from an animal source: “Moraxella anatipestifer”, now classified as Riemerella anatipestifer, was isolated from septicemic ducks.

Of the Moraxella species isolated from animals exhibiting disease, our most complete understanding of pathogenesis exists for Moraxella bovis, an organism that has long been associated with infectious bovine keratoconjunctivitis (IBK), commonly referred to as pinkeye of cattle. Another less‐well characterized Moraxella that has also been associated with IBK is Moraxella bovoculi (Angelos et al. 2007c). It is likely that M. bovoculi has existed in cattle populations for a long time and that Gram‐negative cocci previously isolated from cattle with IBK were, in fact, M. bovoculi that were reported variably as “M. ovis”, “M. ovis‐like”, “Branhamella ovis”, or “Branhamella ovis‐like” species. The primary focus of this chapter is M. bovis and our understanding of its role in the pathogenesis of IBK.

Infectious Bovine Keratoconjunctivitis

IBK (pinkeye) is the most common eye disease of cattle. All breeds of cattle are susceptible to infection, although lower incidences have been reported in Brahman cattle and in cattle with increased pigmentation at the ocular margins. Cattle with IBK exhibit characteristic central corneal ulceration, corneal edema, photophobia, blepharospasm, and lacrimation. Most corneal ulcerations associated with M. bovis heal with varying degrees of corneal scarring; severe scarring may result in reduced vision due to corneal scarring and formation of synechiae (adhesions between the iris and cornea), or even, in the most severe cases, corneal rupture leading to permanent blindness (Figure 15.1).

Corneal ulceration associated with IBK results in economic losses that can be attributed to reduced weight gain, costs of treatment and associated labor, and reduced marketability of affected/treated animals. Together with the economic effects of infection and treatment, the considerable negative impact on animal welfare should not be overlooked.

Table 15.1 Species of Moraxella isolated from animals and their morphology.

Species Source Morphology
Moraxella boevrei Nasal passage of healthy goats Rod, short rod
Moraxella bovis Cattle with infectious bovine keratoconjunctivitis Rod, short rod
Moraxella bovoculi Cattle with infectious bovine keratoconjunctivitis Cocci
Moraxella canis Muzzles healthy dogs and cats Cocci
Moraxella caprae Nasal passage of healthy goats Rod, short rod
Moraxella caviae Pharynx of healthy guinea pigs Cocci
Moraxella cuniculi Nasopharynx of rabbits Cocci
Moraxella equi Horses with conjunctivitis Diplococcus
Moraxella macacae Macaques with epistaxis Cocci
Moraxella oblonga Oral cavity of sheep Cocci
Moraxella ovis Sheep with keratoconjunctivitis Cocci
Moraxella pluranimalium Sheep with meningitis; pigs with pleuritis and polyserositis; healthy pigs Cocci
Moraxella porci Pigs with pleuritis and pneumonia, pericarditis, meningitis, and lymphadenopathy Cocci
Schematic illustration of clinical progression of healing in a steer with infectious bovine keratoconjunctivitis.

Figure 15.1 Clinical progression of healing in a steer with infectious bovine keratoconjunctivitis. On day 0, the ulcer was first identified. Yellow coloration shows areas of fluorescein uptake indicating areas of corneal ulceration. The steer received antibiotics on day 0 when this ulcer was first identified. Corneal ulcer healing with scar formation had occurred by day 21. Ruler markings indicate millimeters.

Early reports into the etiology of IBK suggested associations with infectious bovine rhinotracheitis virus and Mycoplasma. Infection with these agents may promote ocular injury (Pugh et al. 1970, 1976; George et al. 1988) as well as causing an increase in ocular and nasal secretions which may facilitate the transmission of M. bovis between animals. In an experimental challenge study involving infections of calves with M. bovis and Mycoplasma bovoculi, extended colonization by M. bovis and keratitis appeared to be associated with Mycoplasma infection (Rosenbusch 1983). More recent work found that herds with higher Mycoplasma presence have a greater predisposition to IBK outbreaks (Schnee et al. 2015). Outbreaks of IBK have also been reported where Mycoplasma species were the only identified agents (Levisohn et al. 2004). Taken together, these observations suggest that Mycoplasma can be a risk factor for IBK in some cattle populations. Debate also exists as to the role that M. bovoculi plays in IBK. Controlled studies using a scarification model of exposure to a type strain of M. bovoculi did not demonstrate that this isolate could cause corneal ulceration (Gould et al. 2013).

Flies, ultraviolet irradiation, direct mechanical trauma from seed awns such as foxtails, and trace mineral deficiencies are other important risk factors for IBK. M. bovis can survive on the external surface and in the alimentary tract of face flies. Exposure of cattle to face flies that had contacted an M. bovis infected bacterial culture plate developed IBK (Arends et al. 1984). Fly control using insecticide impregnated ear tags (Allan and Van Winden 2020) or back/face rubbers can be effective in reducing IBK. An association between IBK and ultraviolet (UV) irradiation has also been documented and is the basis for experimental challenge models that incorporate exposure of calves to suntanning lamps prior to experimental challenge. The corneas of UV‐exposed calves develop corneal epithelial cell degeneration (Vogelweid et al. 1986) that is believed to predispose to the establishment of an M. bovis infection and IBK. Seed awns can also mechanically scrape the corneal surface and predispose cattle to IBK.

Trace minerals such as copper and selenium are generally accepted to play an important role in cattle health and, although this has not been proven, supplementation with copper and selenium may be important in supporting immune responsivity in cattle with IBK. An in vitro study reported that neutrophils from copper‐deficient cattle had reduced superoxide dismutase and hydrogen peroxide generation compared with copper‐replete animals, although phagocytic and bactericidal activities of neutrophils from copper replete and deficient animals were not different (Cintia et al. 2016). During IBK, increased tear film levels of endogenous anti‐inflammatory lipids and hydroperoxyl glycerophospholipids have been reported (Wood et al. 2018). This work suggests that M. bovis infections can cause lipid peroxidation, and that trace minerals that are involved in protection against oxidant injury should be important in the host response during IBK.

Pathogenesis of Moraxellabovis Infection

Experimental M. bovis infection of cattle results in corneal ulcers, conjunctival erosions, and the accumulation of fibrin, neutrophils, and bacteria within the corneal stroma (Rogers et al. 1987a, 1987b). Most research on M. bovis pathogenesis has focused on proteins that allow this organism firstly to attach to the cornea and, secondly, to damage the corneal epithelium. The M. bovis adhesins that have been best characterized are pili, while damage to the corneal epithelium following attachment has largely been attributed to a repeats in the structural toxin (RTX) cytotoxin (also called cytolysin or hemolysin).

In addition to pili and cytotoxin, M. bovis produces a variety of other enzymes that, while not directly proven to be involved in pathogenesis and not studied in as much detail as cytotoxin, may also be involved in pathogenesis. These include hydrolytic enzymes such as C4 esterase, C8 esterase‐lipase, C14 lipase, phosphoamidase, phosphatase, leucine and valine aminopeptidases and gelatinase (Frank and Gerber 1981), fibrinolysins (Nakazawa and Nemoto 1979), and cell detachment proteins (Marrion and Riley 2000).

Virulence Factors and Pathogenomics

Adhesins (Known and Postulated)


Pilin expression is an important feature of M. bovis pathogenesis. During passage in the laboratory, piliated M. bovis often stop expressing pili resulting in a smooth phenotype. The pili of M. bovis are peritrichously distributed 6.5–8.5 nm diameter elongated unbranched filaments on the surface of rough, but not smooth colonies of M. bovis (Simpson et al. 1976). They are of the N‐methyl phenylalanine type 4 (Patel et al. 1991). Piliated strains of M. bovis adhered better than non‐piliated strains to various bovine cell surfaces, and detachment of pili from M. bovis by magnesium chloride treatment greatly reduced the ability of M. bovis to adhere to these surfaces (Annuar and Wilcox 1985). The importance of pili in M. bovis pathogenesis was subsequently demonstrated in vivo. Piliated M. bovis strain Epp63 induced IBK while non‐piliated strains did not, and vaccines derived from piliated cultures of M. bovis protected against IBK in an experimental challenge (Jayappa and Lehr 1986). Six serogroups of M. bovis pili, designated I through VI, have been described based on enzyme‐linked immunosorbent assay (ELISA) testing methods (Lepper and Hermans 1986). A later and larger study with more M. bovis strains that used slide agglutination, ELISA, and tandem‐crossed immunoelectrophoresis reported seven different pilus serogroups, designated A through G (Moore and Lepper 1991).

It is likely that some of the observed differences in M. bovis vaccine efficacy is related to pilin serogroup differences between herd strains and vaccinal strains. In addition to serogroup differences, an added layer of diversity between strains exists because of two distinctly different phases of pilin that M. bovis strain Epp63 can express: I (also referred to as the α form) and Q (also referred to as the β form); these two distinct forms result from the inversion of a 2 kb region of DNA (Marrs et al. 1988). When phase I‐ and Q‐ pili expressing strains of M. bovis strain Epp63 were examined to determine differences in infectivity, Q pili strains were found to be more infectious than phase I pili strains (Ruehl et al. 1993a, 1993b). In vivo switching between the different pilus forms has been observed and led to the conclusion that Q pili were important for colonization of the bovine cornea, while I pili were involved in maintaining an established infection (Ruehl et al. 1993b). Such switching may be important in M. bovis evasion of host immune responses. Methods for assaying and improving piliation of M. bovis grown in large bioreactors for vaccine production have been described (Prieto et al. 2003, 2008).

If pilus type switching occurs within other serogroups of M. bovis, the antigenic diversity of M. bovis pilus types is expected to be high. Cross‐protection might, however, be possible between pilus‐based vaccines that incorporate antigens comprising the amino end of pilin, which is highly conserved between different serogroups (Atwell et al. 1994). Cyanogen bromide cleavage of pilin protein preparations was shown to expose shared conserved antigenic determinants between heterologous M. bovis strains (Greene et al. 2001a, 2001b); such treatment could be employed in developing pilus‐based vaccines that cross protect between different M. bovis serogroups. An experimental recombinant subunit vaccine that incorporated the conserved amino terminus of one M. bovis strain coupled with the carboxy terminus of the M. bovis cytotoxin has been tested in the field against naturally occurring IBK (Angelos et al. 2007b), but it is not known whether this vaccine antigen imparts an advantage over M. bovis expressing heterologous pilin serogroups. In that study, rates of IBK were lowest in calves that received the pilin cytotoxin vaccine, however, the differences from adjuvant control and cytotoxin vaccinated groups were not statistically significant.

A high degree of similarity exists in the structural pilin gene, pilA, among geographically diverse isolates of M. bovoculi (Angelos et al. 2021). Among 94 different M. bovoculi isolates derived from cattle with IBK throughout California and other western US states, 10 (designated groups A through J) unique PilA sequences were identified that exhibited around 96–99% identity. These 10 pilin sequences were only around 74–76% identical to deduced sequences of putative pilin proteins that were reported in a study of M. bovoculi that were derived from deep nasopharyngeal swabs of IBK, asymptomatic cattle (genotype 2; see below; Dickey et al. 2016). To date, there have been no published studies that have defined whether or not pilin is important in the association of M. bovoculi to the bovine ocular surface.

Filamentous Hemagglutinin

Research has demonstrated that non‐piliated strains of M. bovis adhere to different cell types (Annuar and Wilcox 1985) suggesting that additional proteins besides pili may be important in adherence and pathogenesis. Two large open‐reading frames, designated flpA and flpB, were reported on a large 44 kb plasmid, pMBO‐1, of M. bovis strain Epp63, which encodes proteins that have homology to the Bordetella pertussis filamentous hemagglutinin (FHA), an important virulence factor that allows B. pertussis to adhere to mucosal surfaces (Kakuda et al. 2006). Both FlpA and FlpB have integrin‐binding sequences including LDV and IDS (involved in α4β1 integrin binding), RLD (αVβ3 and αMβ2 binding), and IET (αLβ2 binding) sequences (Kakuda et al. 2006). Along with these similarities, FlpA and FlpB share similarity to the B. pertussis FHA precursor FhaB in that both have a putative approximately 77 amino acid long signal sequence possibly involved in Sec‐dependent transport. Kakuda et al. (2006) also reported another orf on pMBO‐1 encoding a putative Flp accessory protein (Fap), which has homology to the accessory protein FhaC that is necessary to interact with B. pertussis FHA as it traverses the periplasmic space on its way to the cell surface. Reverse transcription polymerase chain reaction analysis also demonstrated the presence of flpA, flpB, and fap gene transcripts in RNA extracted from M. bovis Epp63 grown in vitro (Kakuda et al. 2006). These genes were identified in all but one isolate among a group of geographically diverse M. bovis. In Moraxella catarrhalis, a human pathogen of the middle ear and respiratory tract, filamentous hemagglutinin appears to have an important role in adherence (Balder et al. 2007). Numerous accessions exist in the GenBank database of filamentous hemagglutinin related‐proteins in a number of other Moraxella species including M. bovoculi, and a similar role for this protein in adherence of M. bovis and M. bovoculi is very likely. However, further studies are necessary to definitively confirm a role for FHA in pathogenesis of IBK related to Moraxella spp.

Phospholipase B

In 2001, an approximately 66 kDa secreted protein that has phospholipase B activity was described in M. bovis (Farn et al. 2001). This protein, designated PLB and encoded by plb, belongs to the GDSL (Gly‐Asp‐Ser‐Leu) family of lipolytic proteins (Shiell et al. 2007). Antibodies to a protein (McaP) with phospholipase B activity produced by M. catarrhalis, a human respiratory tract pathogen, caused reduced adherence of M. catarrhalis to human lung cells (Lipski et al. 2007). While PLB has not been conclusively shown to be a key feature of M. bovis pathogenesis, it is probable that it is also involved in cell adhesion and could be important in immunoprophylaxis against IBK. Sequence alignments of available data in GenBank indicates that approximately 40% pairwise identity exists between the amino acid sequence of McaP (GenBank: ABM05621.1) and PLB (GenBank: AAK53448.1), as well as between McaP and autotransporter domain‐containing proteins reported for M. bovoculi (National Center for Biotechnology Information reference sequence: WP_173944456.1).


Early work demonstrated that M. bovis could be identified on the basis of its lipopolysaccharide (LPS) composition (Wannemuehler et al. 1989), and the chemical composition of LPS was determined (Araujo et al. 1989). Later efforts to characterize M. bovis isolates on the basis of outer‐membrane proteins, LPS profiles, and DNA fingerprinting have shown that increased diversity among isolates can be found by combining these three characterization methods (Prieto et al. 1999). A capsular polysaccharide of M. bovis Mb25 was subsequently identified as an unmodified chondroitin disaccharide repeat unit (Wilson et al. 2005). Further work on M. bovis lipooligosaccharides (LOS) have characterized the oligosaccharide structure (De Castro et al. 2014) and its biosynthetic locus (Faglin et al. 2016). Compared with wild‐type strains of M. bovis, experimental mutant strains with truncated LOS have slower growth rates in vitro, increased susceptibility to some antibiotics, reduced survival in serum, and reduced adherence to certain human cell lines (Singh et al. 2018).


M. bovis (Prieto et al. 2013), M. bovoculi and M. ovis (Ely et al. 2019) all possess the ability to form biofilms, and it appears that different isolates of M. bovis, M. bovoculi, and M. ovis vary in biofilm forming capacity (Ely et al. 2019). Exposure to magnesium chloride resulted in M. bovis type IV pili to be removed from bacterial cells and not only prevented biofilm formation, but also caused disassembly of preformed biofilms (Prieto et al. 2013). Biofilm formation in M. bovis imparts protection against antibiotic exposure (Prieto et al. 2013), and lysozyme was reported to have a negative effect on biofilm formation (Ely et al. 2019).


Cytotoxin (Hemolysin/Cytolysin)

In addition to pili, M. bovis pathogenesis also involves an RTX cytotoxin (hemolysin, cytolysin) that has calcium‐dependent hemolytic, corneotoxic, and leukotoxic properties. The significance of M. bovis cytotoxin as a critical virulence determinant was first demonstrated by intracorneal injection of a hemolytic fraction from pathogenic M. bovis; the lesions that were induced were similar to those in naturally occurring IBK, however, equivalent fractions extracted from nonhemolytic M. bovis, did not cause ocular lesions (Beard and Moore 1994), and strains of M. bovis that are nonhemolytic are generally considered to be non‐pathogenic for cattle.

Colonies of hemolytic M. bovis growing on blood agar exhibit beta hemolysis. On a cellular level, when erythrocytes are exposed to the M bovis cytotoxin, potassium efflux, colloid‐osmotic cell swelling, and lysis occurs, presumably due to formation of transmembrane pores in target cell membranes (Clinkenbeard and Thiessen 1991). Cattle that recover from IBK have been shown to develop a hemolysin antibody response that can neutralize the hemolysin from different M. bovis (Ostle and Rosenbusch 1985). Calves vaccinated with M. bovis hemolysin are protected against heterologous M. bovis challenge (Billson et al. 1994; George et al. 2005). Although molecular Koch’s postulates have not been attempted, these findings have helped to establish M. bovis cytotoxin as a key feature of pathogenesis, as well as a worthwhile vaccine candidate. A method for partially purifying and stabilizing the M. bovis cytotoxin from culture supernatants has been described that employs diafiltration of culture supernatant (George et al. 2004). This method results in a cytotoxin‐enriched product that remains stable for at least four months at minus 80°C. The native cytotoxin prepared using this method was reported to have efficacy at preventing IBK (George et al. 2005).

The M. bovis cytotoxin gene designated mbxA encodes a circa 98.8 kDa protein, MbxA, which is related to RTX toxins of other bacteria such as Actinobacillus species (Chapter 12) and Mannheimia haemolytica (Chapter 11). RTX are known to be important in virulence in these and other Gram‐negative pathogens (Ristow and Welch 2019). Amino acid sequence motifs in MbxA are characteristic of RTX including glycine‐rich repeats in the carboxy terminus (Angelos et al. 2001). Four of six such repeats in MbxA match the predicted consensus sequence (Leu/Val‐Xaa‐Gly‐Gly‐Xaa‐Gly‐Asn/Asp‐Asp‐Xaa [L/V‐X‐G‐G‐X‐G‐N/D‐D‐X]) for glycine repeats in RTX toxins. In addition, key lysine residues that are likely targets for activation of the toxin by fatty acylation are present in MbxA. Both hemolytic and cytotoxic activity of a preparation of native M. bovis cytotoxin were neutralized by rabbit antisera against the carboxy terminus of MbxA (Angelos et al. 2001).

Classical RTX operons are composed of four genes arranged 5′‐C‐A‐B‐D‐3′. One of the best characterized RTX operons is hly from uropathogenic E. coli. HlyC activates the structural RTX proper (HlyA) by fatty acylation of conserved lysine residues. HlyB and HlyD are important for the extracellular transport of HlyA, and a secretion accessory protein, TolC, is necessary for the transport of HlyA out of the cell; tolC is unlinked to the Hly operon in E. coli. Following the discovery of MbxA, sequencing of the DNA flanking mbxA revealed the presence of a classical RTX operon (Angelos et al. 2003). In M. bovis, the mbx operon genes are designated mbxC (toxin activating protein); mbxA (structural cytotoxin proper); mbxB (transport); and mbxD (transport) genes. In addition, a gene encoding a protein related to TolC flanks mbxD and is predicted to be necessary for cytotoxin secretion.

Only gold members can continue reading. Log In or Register to continue

Stay updated, free articles. Join our Telegram channel

Nov 13, 2022 | Posted by in GENERAL | Comments Off on Moraxella

Full access? Get Clinical Tree

Get Clinical Tree app for offline access