lable at ScienceDirect

Anaerobe 15 (2009) 44–54

Contents lists avai

Anaerobe

journal homepage: www.elsevier .com/locate/anaerobe

Veterinary anaerobes and diseases

Ecology and pathogenicity of gastrointestinal Streptococcus bovis

Paul Herrera a,b,1, Young Min Kwon a,b, Steven C. Ricke a,b,c,*

a Department of Poultry Science, University of Arkansas, Fayetteville, AR 72701, United Statesb Center for Food Safety, IFSE, University of Arkansas, Fayetteville, AR 72704, United Statesc Department of Food Science, University of Arkansas, Fayetteville, AR 72704, United States

a r t i c l e i n f o

Article history:Received 18 October 2008Accepted 29 November 2008Available online 7 December 2008

Keywords:Streptococcus bovisRuminal acidosisGastrointestinal tractBloatColon cancer

* Corresponding author. Department of Food Scie2650 N. Young Ave, Fayetteville, AR 72704, Unitedfax: þ479 575 6936.

E-mail address: sricke@uark.edu (S.C. Ricke).1 Current address: 2736 Lake Shore Drive, Waco, T

1075-9964/$ – see front matter � 2008 Elsevier Ltd.doi:10.1016/j.anaerobe.2008.11.003

a b s t r a c t

Streptococcus bovis is an indigenous resident in the gastrointestinal tracts of both humans and animals.S. bovis is one of the major causes of bacterial endocarditis and has been implicated in the incidence ofhuman colon cancer, possibly due to chronic inflammatory response at the site of intestinal colonization.Certain feeding regimens in ruminants can lead to overgrowth of S. bovis in the rumen, resulting in theover-production of lactate and capsular polysaccharide causing acute ruminal acidosis and bloat,respectively. There are multiple strategies in controlling acute lactic acidosis and bloat. The incidence ofthe two diseases may be controlled by strict dietary management. Gradual introduction of grain-baseddiets and the feeding of coarsely chopped roughage decrease the incidence of the two disease entities.Ionophores, which have been used to enhance feed conversion and growth rate in cattle, have beenshown to inhibit the growth of lactic acid bacteria in the rumen. Other methods of controlling lactic acidbacteria in the ruminal environment (dietary supplementation of long-chain fatty acids, induction ofpassive and active immune responses to the bacteria, and the use of lytic bacteriophages) have also beeninvestigated. It is anticipated that through continued in-depth ecological analysis of S. bovis the char-acteristics responsible for human and animal pathogenesis would be sufficiently identified to a pointwhere more effective control strategies for the control of this bacteria can be developed.

� 2008 Elsevier Ltd. All rights reserved.

1. Introduction

Streptococcus is a genus of diverse spherical Gram-positive lacticacid bacteria [1]. Cellular division occurs along one axis and thusthe bacteria can be observed growing in pairs or as chains. Severalspecies of Streptococci are responsible for numerous diseases andillnesses in humans and animals including sore throat, pneumonia,meningitis, endocarditis, necrotyzing fasciitis, ruminal acidosis, andbloat. However, many strains are non-pathogenic and occur asnatural commensal flora on the skin, in the oral cavity, the naso-pharynx, the upper respiratory tract, the gastrointestinal tract, andurogenital tract of both humans and animals [1]. Those living in thegastrointestinal tract are facultatively anaerobic [1]. One of the bestcharacterized gastrointestinal Streptococci in both humans andanimals is S. bovis. In ruminants, S. bovis under optimal growthconditions can outgrow other ruminal bacterial flora, and producelarge amounts of lactate and capsular polysaccharide causing acute

nce, University of Arkansas,States. Tel.: þ479 575 4678;

X 72708, United States.

All rights reserved.

ruminal acidosis and bloat respectively [2,3]. In humans, S. bovis hasbeen implicated in colon cancer [4–6]. This review will explore therelationship of S. bovis to other Streptococci and differencesbetween S. bovis strains isolated from ruminants and humans. Therole of S. bovis in the etiology of both lactic acidosis and feedlotbloat in ruminants and colon cancer in humans will also bereviewed.

2. Immunological and biochemical characteristics of S. bovis

2.1. S. bovis’ relation to other group D Streptococci

S. bovis is classified as a member of the group D Streptococci bythe serological classification system based on polysaccharide anti-gens present in the bacterial cell walls [7]. Schlegel et al. [8]analyzed the strains making up the group D Streptococci usingquantitative DNA–DNA hybridization, 16S rDNA sequencing andphylogenetic analysis (Table 1). The group D Streptococci includeS. bovis, Streptococcus equinus, Streptococcus caprinus, Streptococcusgallolyticus, Streptococcus alactolyticus, Streptococcus infantarius,Streptococcus macedonius and Streptococcus waius. Group D Strep-tococci were divided into 4 DNA clusters on the basis of their 16SrDNA sequences [8]. Strains of S. bovis were found to belong in both

Table 1Schlegel et al. [8] characterization of Group D Streptococci by DNA–DNA hybridiza-tion using 16S rDNA probes.

Streptococcus species Source and pathogenic role

DNA cluster I1. Streptococcus bovis biotype II.1 Bovine feces and human endocarditis2. Streptococcus equinus Equine feces and reproductive tract

DNA cluster IIStreptococcus gallolyticus subspecies gallolyticus (previously classifiedas S. bovis biotype I)

1. S. gallolyticus Human endocarditis, bovine mastitis,and koala feces

2. Streptococcus caprinus Goat fecesS. gallolyticus subspecies macedonicus

1. S. macedonicus Bovine mastitis and2. S. waius Dairy products

S. gallolyticus subspecies pasteurianus1. S. bovis biotype II.2 Human endocarditis, blood, urinary tract,

and infectionsDNA cluster III

1. S. infantarius subspeciesinfantarius

Human endocarditis, blood, infections,and dairy products,

2. S. infantarius subspecies coli Human endocarditis, urine, feces, blood,and infections

DNA cluster IV1. S. alactolyticus Porcine infection

P. Herrera et al. / Anaerobe 15 (2009) 44–54 45

DNA clusters I and II. S. bovis can be further classified in twobiotypes depending on their ability (biotype I) or inability(biotype II) to ferment mannitol. Biotype I is also associated withthe hydrolysis of tannins and decarboxylation of gallic acid [9].Biotype II strains can be further subdivided into sub-types: biotypeII.1 is b-glucuronidase negative and a-galactosidase positive,whereas biotype II.2 is b-glucuronidase and b-mannosidase posi-tive. S. bovis biotype II.1 and S. equinus comprise DNA cluster I [8].The high degree of DNA–DNA hybridization and 16S rDNA sequencesimilarity (99.0%) revealed that S. bovis and S. equinus were a singlespecies. Streptococci from DNA cluster II can be isolated fromhuman infections, bovine mastitis, and cheese and include S. bovisbiotypes I and II.2, S. gallolyticus, S. macedonius and S. waius [10].Further analysis of the DNA sequences of these strains suggests thatthis cluster is one species (S. gallolyticus) with three subspecies(gallolyticus, macedonicus, and pasteurianus).

The subspecies gallolyticus is characterized by the fermentationof mannitol, and the presence of the enzymes tannase and gallatedecarboxylase [9]. This subspecies consists of S. bovis biotype Istrains which have been associated with endocarditis and coloncancer [8,11]. This subspecies also includes S. caprinus, which isreadily isolated from ruminants and is noted for its high toleranceto tannins [12]. The subspecies macedonicus consists of S. mace-donicus and S. waius which have been isolated from animal infec-tions and dairy products. The subspecies pasteurianus consists ofS. bovis biotype II.2 which is also associated with disease in humans(endocarditis and urinary tract infections). DNA cluster III consistsof S. infantarius, which is further divided into two subspecies: coliand infantarius. This species has been isolated from food items andinfected humans, where it is associated with systemic disease ininfants. DNA group IV is composed of S. alactolyticus, which origi-nate from animal infections, primarily porcine. It is the only speciesin the S. bovis/S. equinus complex that is unable to ferment lactoseto produce acid.

2.2. Differences between human versus ruminal S. bovis isolates

S. bovis is a common inhabitant of the gastrointestinal tracts ofcattle, sheep, horses, camels, deer, and pigs [13,14]. While it is

infrequently isolated from the intestinal tracts of humans, it hasbeen implicated as the causative agent in several human diseasesincluding associations with meningitis, endocarditis, and bacter-emia in AIDS patients [4–6,15]. Several studies have suggested thatthe presence of S. bovis in feces is associated with an increased riskof developing colonic cancer [6]. As some strains of S. bovis havebeen implicated in human disease, determining which strains arepotential pathogens becomes important. Biochemical tests are nothelpful as diagnostic tools because of the wide variety phenotypesseen in the S. bovis/gallolyticus complex, making it necessary to usemolecular methods to make such distinctions.

Klieve and coworkers [16] attempted to measure the geneticdiversity of 33 presumptive strains of S. bovis isolated fromAustralian cattle and goats fed on a variety of diets. All bacterialstrains exhibited similar morphology and biochemical profilesupon culturing. To confirm the identity of the bacteria, restrictionfragment length polymorphism (RFLP) analysis of the 16S rRNAgenes was performed. The RFLP patterns of the isolate’s 16S rRNAsequences were compared to those of 3 reference strains of S. bovisand a reference strain of Streptococcus intermedius, anothercommon ruminal Streptococcus isolate. Thirty of the isolatesexhibited identical banding patterns to three of the S. bovis refer-ence strains. The other three isolates shared similar bandingpatterns to the S. intermedius reference strain. These findings sug-gested that there was considerable genetic homogeneity in theS. bovis strains isolated from ruminants in Australia. Comparableresults were observed when Ghali et al. [14] compared the strainsof S. bovis isolated from four dromedary camels and two Rusa deerfed alfalfa hay to six strains isolated from cattle and sheep feddifferent diets. The strains exhibited similar morphology andbiochemical profiles during culturing. RFLP analysis of the 16S rRNAgene sequences revealed that the banding patterns of the isolatesfrom the camel and deer rumen samples were identical to thosefrom cattle and sheep samples. When the DNA was sequenced andcompared to the 16S rRNA sequences from three reference strainsfrom cattle and sheep, greater than 99% similarity of the sequenceswas observed.

To compare 27 S. bovis strains derived from animal and humansources at the genetic level, Whitehead and Cotta [13] generatedDNA probes of the V1 region of the 16S rDNA of S. bovis strains JB1(ruminal strain) and ATCC 43143 (human strain). Both DNAhybridization and PCR assays confirmed that the DNA from theruminant strains hybridized with the JB1 primer and that the DNAfrom the human clinical isolates hybridized with the ATCC 43143primer. The ruminal strains exhibited at least a 99% similarity in thegenetic sequence between strains, as did the human clinicalisolates. When the human and ruminal strains were compared toeach other, the similarity dropped to below 97% with the majorityof the differences being observed within the first 350 bases of theV1 region of the 16s rDNA. Furthermore, specific primers for 16SrRNA gene sequences could be used to distinguish between ruminaland human strains of S. bovis.

Kurtovic et al. [17] amplified repetitive DNA sequences (BOXsequences) using PCR from 36 bovine and 15 human isolates ofS. bovis. The sequences were compared by unweighed pair groupmethod (UPGAMA) analysis to produce a dendrogram of thebacterial strains representing their relationships. The humanisolates were found in 6 DNA groups and the bovine isolates in 13groups. Four of the human DNA groups formed a subgroup whichwas distinct from all other bovine isolates. UPGAMA revealed thatthe BOX sequences from these human DNA groups differed fromthe bovine groups by as much as 50%. The other two human DNAgroups were found in subgroups with closely related bovine DNAgroups. They subsequently compared the ability of human andbovine strains to grow in the presence of the antimicrobial enzyme

P. Herrera et al. / Anaerobe 15 (2009) 44–5446

lysozyme and demonstrated that the bovine strains were four toeight fold more sensitive. The growth inhibition of the bovinestrains were further enhanced by replacing glucose in the growthmedia with 2-deoxyglucose (2 mg/ml). When growth rates ofhuman and bovine strains in media containing both lysozyme and2-deoxyglucose were measured by culture turbidity, an almost tenfold growth inhibition in the bovine strains was observed after 48 hincubation when compared to the human strains. These findingssuggest that the addition of lysozyme and 2-deoxyglucose ingrowth media could be used as a culture-based diagnostic tool todistinguish human from bovine isolates of S. bovis.

3. S. bovis and human colon cancer

3.1. Association of S. bovis with colon cancer

In human medicine, there has been an increase in interest inStreptococci as it has been noted that bacteremia and endocarditisdue to these bacteria were associated with increased incidence ofcolorectal neoplasia [4–6,15]. The Streptococcal species associatedwith colonic malignancy include agalactiae, bovis, equinus, milleri,salivarius, and sanguis [6]. The genus Enterococcus, once consideredas part of the genus Streptococcus, has also been implicated incausing colon cancer [18]. The species most frequently isolatedfrom clinical cases of Streptococcal bacteremia or endocarditis isS. bovis, and is responsible for 5–14% of the cases of bacterialendocarditis [19,20].

There has been controversy over the identification of theStreptococci associated with colonic malignancy [20]. The associa-tion between S. bovis bacteremia and endocarditis with coloncancer was first reported in 1974 [21]. Ruoff and coworkers [22]characterized clinical isolates by their cellular fatty acids andbiochemical profiles. They discovered that S. bovis biotype I(mannitol positive) was more likely to be associated with bothendocarditis and malignant colon lesions than biotype II (mannitolnegative). Since that time, there has been a re-evaluation of therelationship of S. bovis with other closely related species. Numerousstudies using molecular analysis of human clinical isolates haveidentified S. gallolyticus as the species most likely to be associatedwith bacteremia, endocarditis and malignant colon lesions [23,24].Clarridge and coworkers [19] used partial 16S ribosomal DNAsequences and biochemical tests specific for Streptococci to identifythe species involved in 22 clinical isolates from men diagnosedwith Streptococcal bacteremia, endocarditis, or urinary tract infec-tions. All isolates were found to be distinctly different from S. gal-lolyticus with most of the isolates belonging to S. bovis biotype II.2(mannitol negative, b-glucuronidase positive). In a similar study,Herrero and coworkers [20] re-evaluated cases from the MayoClinic diagnosed as Streptococcal endocarditis and analyzed theisolates using full length 16S rDNA sequences to identify the speciesinvolved. In 11 of the 14 cases, the causative agent was determinedto be S. bovis biotype I. Two of the cases were diagnosed with cancerupon colonoscopy and two others exhibited possible pre-cancerouslesions. Another case, although never undergoing colonoscopy, hadphysiological changes (anemia and occult blood in the stool) whichwere consistent with colon cancer [20].

3.2. Potential etiologic role

There is further controversy whether the presence of S. bovis isan actual causal factor or an incidental association in the diagnosisof colon cancer [21]. Some researchers have theorized that systemicdisease (endocarditis, meningitis, abscesses) is secondary to thegastrointestinal involvement. It has been noted that in patientsdiagnosed with colon cancer, the chance they will develop S. bovis

endocarditis is 3–6% [21]. However, some studies have suggestedthat S. bovis is more prevalent in the feces of patients with colo-rectal cancer as compared to normal patients with no evidence ofcancer [6,15,25]. S. bovis strains NTTC 8133 and ATCC 41344 havebeen shown to adhere to the intestinal mucosa and stimulate thecytokine production causing vasodilation and increased capillarypermeability [20]. These changes in intestinal function at the site ofthe neoplasia could represent a potential means for the bacteria toenter the bloodstream, spread to other organs, and proliferate. Ithas been further suggested that the presence of antibodies toS. bovis antigens or the antigens themselves in the bloodstreammay act as markers for carcinogenesis in the colon [15,21,25].Although the findings of such studies have been mixed, profilingthe immune response to Streptococcal antigens may eventuallyserve as an early diagnostic tool for the detection of colon cancer.

The association between Streptococcal bacteremia or endo-carditis and colon malignancy is considered so strong that it hasbeen recommended that a complete gastrointestinal examinationbe performed to screen for possible colonic lesions [21,25]. It hasbeen estimated that 60–70% of patients with diagnosed S. bovisendocarditis will also exhibit malignant gastrointestinal lesionsupon colonoscopy [6,21]. Periodic colonoscopic examinations aretypically recommended for 2–4 years after the resolution of theoriginal Streptococcal infection, as the appearance of new lesionscan still occur over that time period [6]. In a prospective study, 29patients with a diagnosis of S. bovis septicemia without colonicpathology were re-evaluated 2 years later. Of the 15 that were givena full gastrointestinal examination, 13 were found to have pre-cancerous or cancerous lesions.

Ellmerich and coworkers [4] studied the colorectal pathogenesisof rats pretreated with the carcinogen azomethane (AOM). All ratsreceived a dose of 15 mg AOM/kg once a week for two weeks. Ratsin the treatment groups received either a S. bovis suspension(1010 CFU/ml) or S. bovis wall-extracted antigens (WEA, 1 mg/ml) inbrain-heart infusion broth (BHI) by gavage twice a week during the5 weeks trial. Controls received unmodified BHI. At the end of thetrial the rats were euthanized, and the tissues assayed for levels ofthe cytokine interleukin 8 (IL-8) and proliferating cell nuclearantigen (PCNA). Tissue samples were examined histologically andwere graded on the presence of aberrant crypt foci (ACF), colonicadenomas, and proliferative crypt cells. Levels of IL-8 wereobserved to be 4- and 3-fold higher in the groups receiving theS. bovis suspension and WEA, respectively, compared to the controlrats. Similarly, there was a 2-fold increase in the levels of PCNA inthe groups receiving either the whole cell or the WEA of S. boviscompared to the controls. When the tissues were examinedhistopathologically, all rats pretreated with AOM exhibitednumerous abnormal and hyperplastic colonic crypts. However, thetreatment groups exhibited a nearly 2-fold increase in the numberof ACF and also a doubling of proliferative crypt cells. Colonicadenomas were observed in 50% of the group receiving the WEA,but not in the other treatment group or the controls. This wouldsuggest that some cell wall components can induce increasedsecretion of cytokines and promote carcinogenesis in chemicallyinduced pre-cancerous colon lesions. This parallels the findings ofresearchers who have examined the role of the bacteria Heli-cobacter pylori in diseases of the stomach [26,27]. H. pylori has beenimplicated in the promotion of gastric ulcers and cancer and isknown to induce increased levels of the cytokines interleukin 1band IL-8 within gastric lesions.

Birac et al. [5] characterized the proteins of S. bovis that inducedthe inflammatory and carcinogenic processes in chemicallyinduced pre-cancerous lesions. S. bovis WEA were fractionated bygel filtration chromatography and tested for IL-8 induction activity.The active fraction consisted of approximately 9 protein bands and

P. Herrera et al. / Anaerobe 15 (2009) 44–54 47

induced the production of IL-8 in Caco-2 cells. IL-8 is known toinduce the over-expression of cyclooxygenase 2 (Cox-2) which inturn leads to increased levels of prostaglandin E2 (PGE2), both ofwhich are elevated in the tissues of colonic adenocarcinomas. Thisactive fraction also increased the expression of Cox-2 and PGE2 intreated Caco-2 cells. The mode of action was believed to be due tothe increased formation of oxygen radicals and nitric oxide whichproduced mutagenesis in the cells of the intestinal mucosa.

4. S. bovis in cattle

4.1. Rumen ecology

Ruminant gastrointestinal tracts are well designed for thedigestion of fibrous plant material [2,28]. Chewing triggers the flowof bicarbonate-rich saliva which acts to both lubricate the food andto buffer the rumen pH [2]. The rumen contains a highly diversepopulation of bacteria, fungi, and protozoa capable of digestingfiber and subsequently via anaerobic fermentation produceproteins, volatile fatty acids, and vitamins needed by the ruminant[29]. During rumination, partially digested material is periodicallytransported from the rumen to the mouth where it is re-chewedand mixed with saliva before being returned to the rumen. Gases(e.g. CO2, CH4) are produced as a by-product of fermentation ata rate of 0.2–2.0 L/min. The buildup of gas in the dorsal sac of therumen induces a series of coordinated muscular contractions whichpulls contents away from the esophageal/ruminal junction andforces the gas through the cardia of the rumen and out theesophagus [3]. Normal coordinated muscular contraction of therumen is contingent on the presence of fibrous plant material [2].A fiber deficient diet can lead to reduced frequency and intensity ofruminal muscular contractions altering ruminal function andfermentation. Nutritional alterations can create an imbalancedrumen microbial ecology leading to severe disease state in theruminant animal.

5. Development of ruminal acidosis

Ruminal acidosis occurs during abrupt switching of cattle froma forage-based diet to a low-fiber high-carbohydrate diet [2,30].The introduction of excess easily digestible carbohydrates can leadto the overgrowth of starch-fermenting, acid-resistant, lactate-producing bacteria, such as S. bovis and Lactobacillus spp. [31–33].S. bovis is normally present in the rumen at populations rangingfrom 104 to 107 colony forming units (CFU) per gram of ruminalcontents but it can proliferate to 1011 CFU/g under conditions ofabnormal ruminal fermentation [34]. When supplied with excesscereal grain starch, S. bovis can outgrow other ruminal bacteria,with a doubling time as low as 12 min. Under normal conditions, S.bovis produces acetate, formate, and ethanol from fermentablesubstrates [34,35]. However, when exposed to excess carbohy-drates, fermentation becomes uncoupled from growth and bacteriashift to homolactic fermentation. With a lower pKa than otherfermentation acids, lactate accumulates overcoming the rumenfluid’s buffering capacity, and ultimately leads to rumen contentacidification [2,32,36]. Consequently, ruminal pH can drop belownormal range (less than pH 5.6) within 1–3 h after a dietary changewith symptoms of lactic acidosis arising shortly afterward [30].

The resulting acidosis can inhibit the growth and functioning ofother ruminal microorganisms able to digest cellulose or fermentthe lactate into other end-products [32,36]. Ruminal bacteria able toferment lactate to volatile fatty acids include: Megasphaera elsdenii,Selenomonas ruminantium, and Fusobacterium necrophorum [34].M. elsdenii is estimated to be responsible for up to 75% of lactatecatabolism in the rumen [37]. However, its growth is sensitive to

acidic conditions. As the pH drops further, Gram-positive bacteriabecome the predominant bacterial population [32]. Lactobacilli arehighly acid-resistant, becoming more numerous in the acidic rumen,and further contributing to the acidosis [34].

As the rumen epithelium is not protected by a coating ofmucous, the increased acidity causes inflammation and ulceration[2,32]. This may permit the entrance of opportunistic bacteria orendotoxins from lysed bacterial cells into the bloodstream [38].Liver abscesses, due to bacteria such as F. necrophorum, are commonsequelae to ruminal acidosis and leads to the condemnation of theorgan [34,39,40]. Endotoxins, histamines, and metalloproteinasescan affect the fine capillary beds in the hooves causing laminitis andlameness [2,38,41]. Ruminal acidosis also suppresses salivation,inhibiting buffering of the ruminal contents by the bicarbonate inthe saliva [42].

As lactic acid concentration increases, the osmolarity of theruminal contents increases, drawing water into the lumen andcausing dehydration, secondary systemic acidosis, and hypo-volemic shock [2,42]. Normal ruminal osmalality can range from240 to 265 mOsm/L with roughage diets and 280–300 mOsm/L incattle acclimated to grain-based diets [43]. When lactic acidosiswas experimentally induced in cattle, ruminal osmalality rangedfrom 339 to 420 mOsm/L with the increased lactate concentrationcontributing approximately 61% of the increased osmolality [44].The influx of water can damage the epithelium of the rumenalready weakened by the increased acidity and ulceration [43].Gozho and coworkers [38] measured the ruminal concentration oflipopolysaccharide (LPS) and serum concentrations of the systemicmarkers of acute inflammatory response, haptoglobin (Hp) andserum amyloid-A (SAA), in rumen-fistulated Jersey steers withexperimentally induced subacute ruminal acidosis. The steers wereswitched from a diet consisting primarily of chopped alfalfa hay topelleted concentrate over a five day feeding trial. Two days after theintroduction of concentrate, the levels of LPS and SAA increasedsignificantly; a day later the levels of HP also increased. LPS insystemic circulation induced the production of cytokines, reactiveoxygen and nitrogen compounds, and bioactive lipids which cancause profound metabolic imbalances in the host. Symptomscaused by rumen acidosis can range anywhere from decreased feedintake, reduced weight gain, to ruminal ulceration, hoof problems,and even death [30,36].

6. S. bovis and grain bloat in feedlot cattle

6.1. Etiology and microbial ecology of grain bloat

Bloat occurs when excessive gas builds up in the rumen [3].Grain bloat can occur if the ruminal conditions prevent normalmuscular contractions or if the movement of gas through the cardiais blocked. As the gas accumulates, the rumen expands and appliespressure on the diaphragm and lungs, impairing respiration andultimately death [2,32,45].

Grain bloat is often associated with acute ruminal acidosis.Increases of facultatively anaerobic Streptococci and Lactobacillipopulations in animals with bloat have been observed [3,45].Several studies have noted that specific ruminal bacteria, such as S.bovis and M. elsdenii, are greater in number in cattle with bloat [48].The cause of bloat is the formation of a stable froth which interfereswith the expulsion of excess gas [32,48]. The formation of this frothis the result of the over-production of capsular polysaccharide byS. bovis and other ruminal bacteria [3,32,49]. Normally the ruminalcontents would exhibit low viscosity [3]. Bubbles of gas rise to thetop of the rumen fluid and coalesce into free gas which is easilyexpelled. During bloat, the viscosity of the rumen fluid increasesand gas bubbles do not burst but form stable foam. The foam

P. Herrera et al. / Anaerobe 15 (2009) 44–5448

prevents the expulsion of gas by eructation [2,48]. In cases withconcurrent acidosis, as the acidity in the rumen increases, thenormal muscular contractions of the rumen are inhibited includingthose involved with eructation. Furthermore as the rumen contentsbecome more acidic, pH-sensitive bacteria die and lyse, releasingtheir intracellular contents. Many ruminal bacteria store carbohy-drates intracellularly when energy is plentiful. When released,these cellular components add to the stability of the froth [3].M. elsdenii contributes to the formation of bloat by producing gasfrom the fermentation of lactate and the formation of a mat whichstabilizes the froth [45,48].

The most basic requirements for polysaccharide production invitro are the availability of CO2 and sucrose [50]. The presence of Bvitamins such as biotin, thiamine and calcium pantothenateenhance the growth and dextran production of S. bovis. Thechemical structure of the extracellular polysaccharide is complexand consists of different monosaccharides, non-carbohydrate sub-stitutents, and a combination of a- and b-linkages [49]. Thecomposition is species specific. Chemical analysis of the poly-saccharides from S. gallolyticus and S. bovis revealed that the formerhad a lower proportion of mannose than the latter. In vitro exper-iments have shown that all strains of S. bovis produce extracellularpolysaccharide when the basic nutritional requirements are met[50]. The production of the mucopolysaccharide is closely related tothe availability of energy [3,51]. The over-production of extracel-lular capsular polysaccharide and increased ruminal contentviscosity frequently occur in ruminants fed diets consisting ofgreater than 50% grain or abruptly switched from a forage-baseddiet [3]. Interfering with the production of extracellular poly-saccharide or inhibiting the growth of ruminal bacteria thatcontribute to the formation of bloat could be a direct means forreducing the incidence of this gastrointestinal disorder.

6.2. Prevention and clinical treatment of bloat

The best strategy for controlling feedlot bloat is prevention andmost of these approaches are the same as those used in preventingruminal acidosis [46]. Gradual introduction of grain into the dietreduces the incidence of both bloat and acute lactic acidosis [2]. Theuse of coarsely chopped forage or coarsely ground grains slows therate of ruminal fermentation and prevents the runaway growth bythe lactic acid bacteria [46]. To prevent feedlot bloat, it has beensuggested that the diet include 10–15% roughage (dry weight).

Table 2Strategies for the control of S. bovis overgrowth and the prevention of ruminal acidosis.

Strategy Rationale

Dietary management Gradual introduction of concentrate to allowbalanced growth of lactate-producing and lacbacteria in response to the increasedcarbohydrate load

Antibiotics and ionophores Inhibition of the growth of Gram-positive rum(including S. bovis) while not affecting the grof Gram-negative and lactate-utilizing bacter

Long-chain fatty acids Inhibition of the growth of Gram-positive rumbacteria (including S. bovis) while not affectinthe growth of Gram-negative and lactate-util

Probiotics Use of live cultures of ruminal bacteria and yto compete with S. bovis for fermentable carbor to utilize the excess lactate

Immunological approaches Introduction of S. bovis-specific antibodies byor dietary supplementation of pre-formed aninhibit the growth and metabolism of S. bovis

Bacteriophage Use of stocks of S. bovis-specific viruses to inhof the bacteria

Tallow, vegetable oils (soybean, peanut, corn) and mineral oils(paraffin) have been added to feed as anti-foaming agents at levels of60–120 mL/head/day [46]. These fats and oils act to reduce thesurface tension of the ruminal contents and inhibit the formation ofbubbles. Synthetic nonionic surfactants have also been added intofood, water, or supplements (licks, molasses) as anti-foaming agentsat a dose of 10–20 g/head/day. Some of the anti-foaming agents thathave been examined include dioctyl sodium sulfosuccinate (doc-usate), poloxalene, and alcohol ethoxylate detergents. Anti-foamingagents can be delivered directly to the rumen via a rubber hose toalleviate mild to moderate cases of bloat [47]. In severe cases of bloat,where the life of the animal is at stake, it may necessary to performan emergency rumenotomy or use a trocar to physically relieve thepressure [46]. More specific strategies focused on controlling S. bovisovergrowth will be discussed in the following sections.

7. Strategies for controlling S. bovis overgrowth in the rumen

7.1. Dietary management

Given the potential causative role of S. bovis, numerous strate-gies to specifically prevent its overgrowth, acidosis, and bloat infeedlot cattle have been devised (Table 2) and included dietmanagement, feed supplementation with cultures of lactate-utilizing bacteria, vaccination, use of ionophores, antibiotics, orbacteriophages [3,30,32]. Of these approaches, careful dietarymanagement has been the most extensively characterized. Intro-ducing grain-based diets gradually allows the ruminal microor-ganisms to adapt the increased carbohydrate load [2]. This allowsthe balanced growth of lactate-producing bacteria and lactate-utilizing microorganisms [52]. However, staggered introduction ofgrain-based diets is both labor-intensive and expensive [30].

Cereal grains differ in how rapidly they are fermented [39].Wheat, barley, high-moisture or steam-flaked corn can cause rapidreductions in ruminal pH. The difference is due to the nature of theprotein matrix surrounding the starch granules [53]. While judi-cious choice of grains fed to cattle could reduce the incidence ofbloat, the choice of grain fed is usually dictated by economicalconcerns and seasonal availability.

Grain is normally cracked, cut or rolled to break the pericarp,allowing ruminal microorganisms greater access to the nutrients,thus increasing the grain’s digestibility [54]. However, the smallerthe particle size of the processed grain becomes, the greater the

Comments

for thetate-utilizing

Gradual dietary adaptation is both timeconsuming and costly

inal bacteriaowthia

There are concerns regarding the targetedorganisms gaining resistance to these antimicrobialcompounds

inalgizing bacteria

Inconsistent results

eastsohydrates

Inconsistent results

vaccinationtibodies to

Expensive and/or time consuming

ibit the growth Effects transient, bacteria become resistant

P. Herrera et al. / Anaerobe 15 (2009) 44–54 49

incidence of acute rumen acidosis, bloat, laminitis, and liverabscesses. It has been suggested that the optimal size that couldprovide adequate grain digestibility without a concurrent increasein ruminal pathology would be between 0.25 and 1.0 mm [53].Beauchemin et al. [54] characterized the effects of temper-rollingbarley on feed intake, digestibility, ruminal passage rate and pH insteers fed a 90% barley diet. Barley was passed through roller millsand rolled into 4 different thicknesses. The amount of processingwas represented by a processing index (PI). Rolling grain into finerconsistencies reduced the value of the PI. The four barley dietsprepared had PIs ranging from 85% (course) to 65% (flat). Total drymatter intake was not affected by the amount of processing thegrain received. Modest increases in the digestibility of starch, fiber,and nitrogen were observed as the amount of grain processingincreased. There were no observed differences in the passage rateof ruminal contents in the four experimental groups. However, theeffect of grain processing on ruminal pH was significant. RuminalpH for all groups decreased from an average value of 6.6 from thetime of feeding to at 15 h ranging from 5.7 (course; PI¼ 85%) to 5.2(medium-flat; PI¼ 70%). Essentially the time period during whichruminal pH was greater than 6.2 decreased quadratically withincreased grain processing and similar patterns were observed forfecal pH.

Treating cereal grain with heat alters its digestibility [39]. Steamflaking of sorghum gelatinizes the starch granules, making them upto 3-fold more accessible to ruminal fermentation. In contrast,treatment with dry heat, 90–100 �C for 1 min results in complexesformed between the starch and protein in the grain and slowsruminal fermentation without negatively affecting feed efficiencyor average daily weight gain.

The amount and type of roughage in the diet also affects theincidence of lactic acidosis, bloat, and liver abscesses [39,47]. Ingeneral, the lower the amount of roughage in the finishing diet,the greater the chance for fluctuations in ruminal fermentationand pH. Increasing the amount of forage in a diet stimulates salivaproduction, which is rich in bicarbonate and increases the pH ofthe rumen [55,56]. Increased amounts of forage in the dietincrease the time a cow will chew and ruminate [57]. Salivaryflow rates during the act of rumination can be nearly twice theresting flow rate, which can range from 0.10 to 0.15 L/min [56].Cattle consuming high-grain diets produce only 60–70% of thevolume of saliva of cattle consuming equivalent amounts offorage. Increased forage also slows the rate of ruminal fermenta-tion and reduces the incidence of ruminal acidosis and bloat. Thusthe ratio of forage to grain in a diet has to be balanced betweenincreased feed efficiency and weight gain and the incidence ofruminal pathology.

The feeding of pelleted diets in cattle feedlots is not recom-mended as it has been correlated with increased incidences ofboth ruminal acidosis and bloat [46]. Nocek and Kesler [57a]characterized the changes in rumen development in steers feda pelleted diet versus those fed a hay-based diet. Rumen pH insteers in the experimental group (pH 5.65) was lower than in thecontrol steers (pH 6.40). In addition, there was a higher incidenceof bloat in the steers consuming the pelleted diet. However, at theend of the 32 weeks trial, both groups of cattle exhibited similargains in weight, height, and chest circumference. Examination ofthe ruminal mucosa of the steers consuming the pelleted dietsrevealed fewer papillae with many of the papillae being shorter,thickened, or malformed and in some cases, entire areas ofmucosa denuded of papillae. The changes in ruminal papillae inthe experimental steers are believed to be due to the increasedconcentrations of lactic acid which altered papillae metabolism,development and function, reducing the absorptive capability ofthe rumen.

7.2. Antibiotics

Antibiotics are administered to cattle as a feed supplement inorder to increase feed conversion, improve carcass quality, andreduce disease [2,30]. A variety of antibiotics have historically beenfed to cattle including tylosin, bacitracin methylene disalicyclate,chlortetracycline, oxytetracycline, neomycin, and virginimycin [2].Bryant and coworkers [45] attempted to enumerate the variousclasses of ruminal bacteria and protozoa in cattle fed a diet high inconcentrate. They compared the populations of ruminal microor-ganisms before and after treatment with penicillin with a primaryfocus on the facultatively anaerobic Streptococci and cellulolyticbacteria. They found no correlation between changes in bacterialpopulations and the biological manifestations of bloat [45]. It wasnoted that the initial administration of penicillin caused decreasesin the general bacterial population in cattle with bloat. However,this change was temporary as the bacteria quickly adapted to theantibiotic treatment and returned to pre-treatment levels.However, the rise of antibiotic resistance in livestock and concernthat these resistant microorganisms may cross-over into humanpopulations have led to reduction of the indiscriminate use of theseantimicrobial agents [2,30].

7.3. Ionophores

7.3.1. Use in beef industryIonophores reduce the incidence of feedlot bloat and lactic

acidosis [58,59]. Some of the most commercially and commonlyused ionophores include monensin, lasalocid, laidlomycin, andsalinomycin [58]. A dose of 1.32 mg/kg body weight of lasalocid andmonensin reduced bloat in cattle consuming high-grain diets by92% and 64%, respectively [59]. A dose of lasalocid of 0.66 mg/kgprevented bloat formation when administered before the intro-duction of the grain-based diet.

Historically ionophores have benefited the beef cattle industrywhen used as feed supplements to increase feed conversion andgrowth rates in feedlot cattle. Treatment with monensin decreasesfeed intake in a dose related manner, but does not negatively affectdaily weight gain resulting in the potential for a 10–17% increase infeed efficiency [60]. Ionophores are known to enhance the energymetabolism by increasing the production of propionate by ruminalfermentation while concurrently reducing methane production[60]. Ionophores also decrease the production of ammonia duringrumen fermentation [58,60,61]. This is believed to stem from theinhibition of proteins and amino acids degradation by ruminalmicroorganisms allowing more intact dietary protein to reach thesmall intestine and be absorbed by the ruminant. Consequently,less protein is converted to ammonia by ruminal microflora and lostthrough the urine. Likewise, decreased methane productionrepresents energy no longer lost to the animal.

7.3.2. Mechanism of antimicrobial activity of ionophoresIonophores have both polar and non-polar domains which allow

them to capture cations, interact with cellular membranes, trans-locate the cations across the bacterial membrane, and disrupt iongradients [36,60]. Ionophores inhibit the growth of most Gram-positive rumen bacteria, including S. bovis, but have no effect onGram-negative bacteria or on lactate-utilizing and propionate-producing bacteria [2]. The difference in activity is related to theconstruction of the cell membrane [29]. Ionophores can more easilytraverse the membranes of Gram-positive bacteria [58]. Thisbecomes critical because bacterial membranes are normally rela-tively impermeable to ions allowing them to establish and maintainion gradients to energize nutrient uptake and other metabolicprocesses [62]. This relationship is important in ruminal bacteria as

P. Herrera et al. / Anaerobe 15 (2009) 44–5450

well. The intracellular spaces of ruminal bacteria have high Kþ andlow Naþ concentrations compared to rumen fluid which has low Kþ

and high Naþ concentrations. In addition ruminal pH is slightlyacidic due to the presence of VFA [58]. In contrast, ruminal bacteriapossess an intracellular pH that is neutral or slightly alkaline due tothe active transport of protons outward through the cellmembrane. When an ionophore such as monensin which isa metal/proton antiporter inserts into the cytoplasmic membrane,Hþ is exchanged for either Kþ or Naþ [29]. As the Kþ gradient isgreater than the Naþ gradient, there is a net movement of Kþ out ofthe cell and Hþ into the cell [36]. The bacteria respond to theacidification of the cytoplasm by hydrolyzing ATP via F1F0 ATPase totransport Hþ out of the cell. The bacteria also activate other ATP-powered pumps for Kþ uptake and Naþ excretion to reestablish theion gradient. The ionophore interferes with bacterial growth bydepleting ATP stores and can lead to cellular death.

Newbold and Wallace [37] characterized the effects of theionophore tetronasin on a co-culture of 4 ruminal bacteria (M.elsdenii, Se. ruminantium, S. bovis, and Lactobacillus) in continuousculture. Under continuous culture conditions with limited glucose,the four bacteria formed a stable co-culture with a distribution ofapproximately 40% Se. ruminantium, 20–30% of both S. bovis and M.elsdenii, and 2–5% Lactobacillus. Lactate was not detected asa fermentation product. The addition of excess glucose to theculture without the ionophore caused an initial increase in thepopulations of all the bacteria. However, the culture pH decreasedand reached a minimum value of 4.5 at approximately 5–6 h afterthe introduction of glucose. This corresponded with an increase inthe lactate concentration (approximately 23.8–24.2 mmol/L) in theco-culture. During this phase S. bovis outgrew the other bacteria,with both M. elsdenii and Se. ruminantium populations simulta-neously declining. Approximately 12 h after glucose addition,S. bovis also declined and Lactobacillus dominated after 48 h.

When the ionophore tetronasin (0.5 mg/ml) was introduced byNewbold and Wallace [37] into the co-culture at the same time asexcess glucose, there was a decrease in pH values although both therate and full extent (5.1) were not as severe. Lactate productionincreased rapidly for the first 3 h after the introduction of glucoseand reached a peak of 16.1 mmol/L, after which the concentrationsdecreased for the duration of the trial. Increases in the volatile fattyacids propionate and butyrate were observed and reached peakconcentrations at 9.7 and 26.2 mmol/L, respectively. S. bovisincreased with the introduction of glucose; however, there was noinhibition in the growth of the other bacteria. When the co-culturereached steady state the bacterial population consisted of approx-imately 60% Se. ruminantium, 23% M. elsdenii, 15% S. bovis, and lessthan 1% Lactobacillus. When tetronasin was administered 24 h afterglucose addition to the co-culture lactate concentration reached27.0 mmol/L, the pH value was 4.4, and lactic bacteria dominated.After ionophore addition, co-culture pH rose to 5.0 within 24 h.Lactate decreased to below detectable limits, butyrate increased,and lactic acid bacterial growth was inhibited while Se. ruminan-tium and M. elsdenii increased.

Studies have shown that ionophore-sensitive ruminal bacteriacan quickly become resistant in vitro [63]. In the majority of cases itis believed that the bacteria alter the characteristics of the outermembrane to hinder the ionophores’ access to the cytoplasmicmembrane. Prevotella ruminicola that are adapted to monensin bindless of the ionophore and it is hypothesized that they exclude theionophore by reducing porin size [64]. The change in cellmembrane characteristics also results in other metabolic conse-quences. For example, sensitive strains of P. ruminicola are able toferment both di- and tri-peptides; whereas the adapted strains losethe ability to ferment tri-peptides [58]. Clostridium aminophilum,when serially transferred 20 times into media containing sub-lethal

doses of monensin (0.3 mM), increased its resistance to the iono-phore 8-fold. In contrast to P. ruminicola, C. aminophilum acquiredresistance by increased extracellular polysaccharide production[58]. S. bovis has also been shown to be initially very sensitive to theeffects of monensin, but repeated passage in media with sub-lethaldoses of the ionophore (0.25 mM) results in a 7-fold increase inmonensin resistance [63].

7.4. Long-chain fatty acids

Attempts to use long-chain fatty acids to alter the ruminalmicroflora have met with mixed results [51,53]. Lauric acid, foundin high concentrations in coconut and palm kernel oil, has anti-bacterial activity against Gram-positive bacteria [51]. It inhibitsthe growth of Gram-positive ruminal bacteria such as Butyrivibrioand Ruminococcus, but not Gram-negative ruminal bacteria suchas Se. ruminantium, P. ruminicola, Anaerovibrio lipolytica, and M.elsdenii. Other studies suggest that lauric acid can also reduce thenumber of fibrolytic bacteria and protozoa, which could nega-tively influence ruminal fermentation. Lauric acid has been shownto affect ruminal fermentation in in vitro experiments bydecreasing the production of methane and ammonia, butincreasing propionate production. Yabuuchi and coworkers [51]fed Holstein steers diets high in grain content supplemented withlauric acid and reported changes in rumen pH, viscosity, VFAconcentrations, ammonia as well as ruminal protozoa and bacte-rial counts. The supplementation did not affect feed intake or thedigestibility of the diet or any of the measured parameters ofruminal fermentation. When selected species of ruminal protozoaand bacteria were measured by real-time PCR assays, ruminalmicroflora between the controls and treatment groups weresimilar. In in vitro trials, S. bovis growth was initially inhibitedwhen grown in media supplemented with lauric acid. However,the bacteria quickly adapted after serial passage with increasingconcentrations of the fatty acid. It appears that although dietarysupplementation with short chained fatty acids may alter ruminalfermentation, the effect is transient.

7.5. Probiotics

Probiotics are cultures of potentially beneficial bacteria used asdietary supplements to favorably alter gastrointestinal function.Bacteria such as M. elsdenii, Se. ruminantium, Veillonella parvula andpropionibacteria utilize lactic acid to produce propionate [32,52]. Ithas been suggested that cultures of these bacteria could be used asprobiotics to prevent the accumulation of lactic acid in susceptibleanimals. M. elsdenii may be an ideal candidate as it can use lactate,glucose, and maltose, thus directly competing with S. bovis for thesemetabolites [52]. Kung and Hession [52] measured the changes inpH and lactate production in batch fermentation of ruminal fluidinoculated with M. elsdenii. A mixture of filtered ruminal fluid fromhay-fed cattle and a ruminal microflora suspension were inoculatedwith low or high doses (105–106 CFU/ml) of M. elsdenii. In controlcultures, pH decreased to 4.8 after 10 h of fermentation andremained stable for the remainder of the 24 h test period. In thecultures inoculated with M. elsdenii, pH decreased to approximately5.5 after 10 h. After 10 h of fermentation, 17.5 mM of lactate waspresent in the control cultures while lactate concentrations werebelow the detection limit for M. elsdenii cultures. It is conceivablethat dietary supplementation with lactate-utilizing bacteria duringthe introduction of a grain-based diet may decrease the incidenceof ruminal acidosis by stabilizing non-related rumen bacterialspecies. The lactate-utilizing bacteria enhance the growth of themost active cellulolytic ruminal bacteria such as Ruminococcusalbus, Ruminococcus flavefaciens, and Fibrobacter succinogenes [2].

P. Herrera et al. / Anaerobe 15 (2009) 44–54 51

Shunting ruminal fermentation from lactate to propionateproduction also reduces methanogenesis [32].

Lactic acid consumption by lactate-utilizing bacteria has beenshown to be enhanced by co-culturing with yeasts such asSaccharomyces cerevisiae and Aspergillus oryzae or the addition ofdicarboxylic organic acids, such as malate or fumarate [2,32]. Activedry yeast products, some of which contain no less than severalbillion Sa. cerevisiae viable cells per gram, have been used as dietarysupplements in dairy cattle [65]. The addition of yeast to the diet ata rate of 0.5 g/head/day enhanced the efficiency of ruminal cellu-lolytic activity, reduced dietary nitrogen wastage and ruminalammonium production, and increased total milk production andmilk protein. Diet supplementation with yeast has the addedbenefit of reducing the incidence of ruminal acidosis in cattle feddiets high in concentrates.

7.6. Immunological approaches

7.6.1. Passive immunizationAttempts to immunize livestock against S. bovis and other lactic

acid bacteria in order to prevent the overgrowth of the bacteriumhave been examined [30]. The resulting antibodies to lactic acidbacteria have been shown to inhibit growth in these bacteria invitro [66]. The mechanism by which the antibodies inhibit lacticacidosis is believed to be due to binding of the antibodies to thelactic acid bacteria and interference with either their proliferationor metabolism [30].

The induction of passive immunization by the feeding ofcolostrum or egg-derived antibodies has been attempted to controlbovine coronovirus, F. necrophorum, and S. bovis in cattle [66].DiLorenzo and coworkers [66] treated ruminally cannulated steersfed a diet consisting of 83% dry-rolled corn with polyclonal egg-derived antibodies against S. bovis. A daily dose of 2.5 ml of theantibodies reduced S. bovis from 2.85�108 CFU/ml on day 0 ofthe trial to 4.76�107 CFU/ml on day 14. In contrast, the levels of thebacteria in the control cattle climbed from 3.63�108 CFU/ml onday 0 of the trial to 9.77�108 CFU/ml on day 14 while totalanaerobic bacterial populations did not differ between the treat-ment and control groups. It was also noted that the cattle receivingthe antibodies had higher ruminal pH values than the control cattle,6.08 versus 5.67, respectively.

7.6.2. Active immunizationThere is wide antigenic variation within the numerous strains of

S. bovis and Lactobacillus. In order to be effective, a vaccine wouldhave to react with numerous strains. Shu and coworkers [33]characterized the cross-reactivity of antiserum to S. bovis strain Sb-5 to 9 strains of S. bovis (3 encapsulated and 6 un-encapsulated).Mixtures of bacterial suspensions and antisera were incubated for1 h at 37 �C. The bacteria were subsequently removed and theantibody concentration (units/ml) in supernatant was determinedby ELISA. Cross-reactivity index (CRI) was calculated by dividing theantibody concentration in the antiserum after absorption by theantibody concentration in the antiserum prior to absorption. A lowCRI indicated high cross-reactivity, whereas a high CRI meant lowcross-reactivity. When the Sb-5 antiserum was incubated with theencapsulated strains CRI ranging from 7.3% to 12.4% were observed.The six un-encapsulated strains of S. bovis had higher CRI rangingfrom 28.9% to 56.1%.

How S. bovis cells are prepared prior to immunization may bea factor as well. When laying hens were inoculated with a killedS. bovis suspension, there was a moderate increase in specific eggyolk antibodies from three to seven weeks after the inoculation[67]. When comparative ELISA tests were performed with threenon-related ruminal bacteria, the S. bovis-specific antibodies

exhibited significant cross-reactivity with the other bacteria [67].Inconsistent immune responses may be due to the masking ofantigens by the extracellular polysaccharide capsule. Consequently,producing a vaccine specific to S. bovis may require more extensiveprocessing of the bacterial cells, as well as isolation and identifi-cation of appropriate antigens to optimize the immune responseand subsequent antibody production.

The profile and location of the antibody response should also beconsidered. The primary antibody type present in the mucosalsecretions of ruminants is IgG1 [68]. When radiolabeled IgG1 wasintravenously injected into pregnant heifers, plasma IgG1 concen-trations decreased sharply over a 48 h period. Concurrently, theconcentration of the antibody in the mucosal secretions peaked atapproximately 48 h. However, when the amount of radiolabeledIgG1 was compared to the total amount of IgG1 in the mucosalsecretion and tissue, it was discovered that the radiolabeled IgG1was only a minor component in the mucosa. The majority of theantibody was locally produced by plasma cells within the mucosa.In contrast, the rumen is a poor site for inducing a local immuneresponse due to its non-glandular and keratinized tissue charac-teristics [69]. For the control of acute lactic acidosis in the rumen,immunoglobulins in saliva affect S. bovis growth and ruminalfermentation [68]. Local immune response in the intestinal tractmucosa may inhibit the proliferation of S. bovis in the hindgut,reduce the amount of lactic acid production, and prevent diarrheain immunized animals [30].

7.6.3. Vaccination in sheep and cattleGill et al. [30] characterized the effects of vaccination with

S. bovis suspensions on ruminal fermentation and the symptoms ofacute lactic acidosis in Merino wethers. The two treatment groupswere vaccinated intramuscularly with either suspensions of livebacteria (SB) or formalin-killed bacteria (kSB). The feed consump-tion rates and diarrhea scores were monitored and anti-SB anti-body titers in the body fluids, ruminal pH and lactateconcentrations were periodically measured. Prior to the firstvaccination, low levels of anti-S. bovis antibodies were detected inthe saliva and serum of all test animals [30]. After the third booster,significantly higher titers of the antibodies were detected in thebodily fluids of the immunized animals compared to the controlsand antibody levels in all animals inoculated with SB were at leastdouble compared to those receiving kSB. After the introduction ofthe grain-based diet, a decrease in feed consumption and anincrease in diarrhea scores were observed in all groups [30].However, significantly higher feed consumption was observed inthe SB group. All sheep vaccinated with the live SB survived thefeeding trials but one sheep from the control and kSB groups diedfrom acute acidosis. When the grain-based diet was initiated therumen fluid pH dropped in all groups but the rumen pH of the SBgroup was significantly higher than the kSB and control groupswhile ruminal L-lactate concentration was 7-fold higher in thecontrol group versus the SB group [30]. In general, the animalsimmunized with live S. bovis had greater antibody response, higherfeed consumption, higher rumen pH, and lower diarrhea scoresthan the controls or the animals immunized with killed bacteria.Although it appears that immunization with S. bovis suspensionscan reduce the severity of lactic acidosis in animals fed a grain-based diet, optimizing the immune response may require morerefinement of the SB antigens used to vaccinate.

Shu et al. [69] performed a similar study with cattle. FiveHereford steers were given four vaccinations with live suspensionsof S. bovis strain Sb-5. Samples of serum, rumen fluid, and salivawere collected periodically throughout the vaccination stage andthe feeding trial. By the third booster, concentrations of the anti-bodies to Sb-5 peaked at 200 units/ml. Titers slowly decreased over

P. Herrera et al. / Anaerobe 15 (2009) 44–5452

the course of the trial, yet remained significantly higher than thecontrol levels by the end of the trial at day 121. There was a stronglinear correlation between the antibody concentrations found inthe serum and those in the saliva. Two days after the cattle wereplaced on the wheat diet, the feed intakes of the immunizedanimals were significantly higher than the controls [69]. The lactateconcentration was 7–10-fold higher in the control steers comparedto the immunized steers 24 h after the introduction of the wheatdiet. Consequently, the rumen pH was lower in the control steers.Prior to the start of the wheat diet, there was a significant differencein the number of S. bovis between the vaccinated (3.69 log10 CFU/ml) and the control (5.68 log10 CFU/ml) steers. The direct reductionin S. bovis and the increase in antibody concentrations in blood andsaliva, and the resulting changes in fermentation and animalresponse suggest that intramuscular vaccination with live S. boviseffectively reduces acute lactic acidosis in animals fed grain-baseddiets [30,69]. It would be interesting to follow vaccinated animalsover several cycles of changes in feeding regimens to determinewhether resistance to S. bovis or reduced lactic acidosis is retained.

7.7. Bacteriophages

7.7.1. Isolation of S. bovis-specific phageBacteriophages have multiple advantages as antibacterial

agents. They can be easily isolated from the environment andcultivated, and unlike pharmaceutical agents which are metabo-lized and excreted from the body, they replicate in their hosts,growing exponentially in number and potency [70–72]. Phages areubiquitous and are found wherever bacteria are found includingthe rumen where they can occur at concentrations as high as 109

plaque forming units (PFU) per ml of fluid [73]. Klieve et al. [74]isolated 36 strains of S. bovis from rumen fluid samples from cattleand goats fed varying diets. They characterized the genetic diversityof the bacterial strains by amplifying the 16S rDNA sequences,subjecting them to digestion by restriction enzymes, and visual-izing the DNA profiles by gel electrophoresis. When the restrictionenzyme profiles of the bacterial isolates were compared to 3representative strains of S. bovis, nearly identical profiles wereobserved. They also isolated 3 bacteriophages from abattoir waste-water which infected the S. bovis strain 2B and tested the sensitivityof the bacterial isolates to the phages using the soft agar overlaymethod. Plaque formation was only observed with S. bovis 2B andtwo other strains. In all cases, the plaques were small (less than0.5 mm diameter) and turbid, suggesting that lysogeny wasoccurring. In a lysogenic cycle, the phage’s genetic material isincorporated into the host’s genome [71]. Consequently, phageswhich possess the capability of lysogeny are reduced in theireffectiveness as candidates for bacteriophage therapy. Numerousresearchers have reported high rates of lysogeny in S. bovis phagesranging anywhere from 23.7% to 55.5% [74,75].

7.7.2. Phage therapy and potential for resistanceAnother process which may limit the therapeutic use of phages

is the rise of resistance [71,76]. Klieve and coworkers [74] isolatedthe bacteriophage phi Sb01 which was lytic to the S. bovis strain 2B.Repeated culturing of phages and bacteria together rapidly led tothe development of resistance in the bacteria. Normally thebacteria would grow as single cells or as diplococci. However, theresistant bacteria initially grew in long chains and eventuallyformed clumps of cells, suggesting that the continued exposure tothe phage induced changes in the bacterial phenotype which mayhave been protective.

The rapid rise of bacteriophage resistance in ruminal bacteriasuggests that there is a high rate of adaptation between phages andtheir hosts in the dynamic environment of the rumen. Iverson and

Millis [77] reported rapid and spontaneous development ofa secession of S. bovis strains with differing sensitivities to bacte-riophages within the rumens of sheep. They isolated lytic phagesfrom slaughterhouse wastes and grew high titer stocks that wereadministered in in vivo trials. While the phage stocks were beingprepared, Iverson and Millis [77] monitored the populations ofphage-sensitive and phage-resistant strains in the rumens of thetwo test sheep. The total concentration of S. bovis remained rela-tively consistent over this time period and no phage-resistantbacteria were initially detected. However, in spite of the fact thatthere were no interactions between the isolated phages and therumens of the 2 sheep by the 36th day of the trial, numbers of thesensitive bacteria began to drop and bacteria resistant to the phageappeared. By day 43 of the trial, the phage-resistant straincompletely replaced the phage-sensitive strain. No phages weredetected in the rumens of the 2 sheep. Undeterred, they twicerepeated the process isolating a new lytic bacteriophage fromslaughterhouse waste and growing new phage stocks. Each time,bacteria resistant to the new phage would spontaneously appear inthe rumen of the two sheep.

It appears the mutations that confer resistance to lytic bacte-riophage occur frequently and rapidly in the dynamic environmentof the rumen. The fact that no phages were detected during thechange in bacteriophage sensitivities strongly suggests that phageresistance arises fairly easily and externally administered phage israpidly cleared from gastrointestinal systems such as the rumen. Todevelop a more effective phage therapy will require a betterunderstanding of the biology and ecology of phages during their lifecycles in the rumen. In particular understanding how they areintroduced and sustained may offer clues for developing better andmore long term phage treatment approaches. Strategies involvingmanipulation of lysogeny and lytic phases may be necessary toretain phage in the rumen.

8. Conclusions

S. bovis belongs to a diverse genus of bacteria which can befound in the gastrointestinal and respiratory tracts of animals andhumans. S. bovis has been implicated as a cause of colorectal cancerin humans. It is also the causative agent for ruminal acidosis andbloat in ruminants. S. bovis causes disease in ruminants when theyare introduced to a diet high in readily digestible carbohydrates,such as cereal grains because it can readily ferment starches andproduce copious amounts of lactic acid. The increase in lactic acidconcentration reduces the ruminal pH, thus inhibiting the growthof cellolytic and lactate-utilizing microorganisms. The lactic acidalso affects ruminal fermentation and absorption by inhibitingruminal wall contractions and disrupting the architecture of theruminal mucosa. Over-production of extracellular polysaccharideacts to trap gas formed during fermentation in a stable form pre-venting its expulsion by eructation. The combined effects of S. bovison ruminal fermentation can be as mild as a transitory reduction infeed intake and as severe as laminitis, liver abscesses, and death.

Numerous strategies have been proposed for the control ofruminal acidosis in dairy and feedlot cattle. Careful dietarymanagement has been the most explored strategy to date. Antibi-otics and ionophores have been used to increase weight gain andfeed conversion in feedlot cattle for several decades. Althoughthese antimicrobial agents can also alter the metabolic function inGram-positive bacteria such as S. bovis and reduce both the inci-dence and severity of ruminal acidosis, the exact mechanism ofaction remains unknown. Attempts have been made to control thegrowth of S. bovis in the rumen by vaccinating cattle in order toproduce antibodies specific to the bacteria albeit with some successbut the process remains time consuming. Controlling S. bovis

P. Herrera et al. / Anaerobe 15 (2009) 44–54 53

biologically using lytic bacteriophages is enticing as phages havemultiple advantages: they act specifically on the host bacteria, theyare easy to isolate and grow, and their numbers increase during thetreatment period. However, resistance to phages can occur rapidlyin S. bovis and may limit their usefulness. Clearly further more in-depth studies of S. bovis ecology and metabolism are required. Thisin turn should lead to new strategies devised in limiting theuncontrolled growth, and these approaches could avoid thesubsequent dominance of S. bovis that occurs during unbalancedrumen fermentation. Ultimately combinations of preventativemeasures such as vaccination in conjunction with immediateclearance approaches such as phages may prove to be the mostconsistent and effective means to limit S. bovis over extendedperiods of time.

Acknowledgements

This review was supported by an Arkansas ABI grant (ArkansasBiosciences Institute, 4301 W. Markham St., Little Rock, AR), anda USDA Food Consortium grant. PH was supported by a postdoctoralfellowship (USDA-NRI Grant No. 2004-04751).

References

[1] Holt JG. Genus Streptococcus. In: Bergy DH, Holt JG, Krieg NR, Sneath PH,editors. Bergey’s manual of determinative bacteriology. Baltimore, MD: Lip-pincott Williams & Wilkins; 1994. p. 523.

[2] Russell JB, Rychlik JL. Factors that alter rumen microbial ecology. Science2001;292:1119–22.

[3] Cheng K-J, McAllister TA, Popp JD, Hristov AN, Mir Z, Shin HT. A review of bloatin feedlot cattle. J Anim Sci 1998;76:299–308.

[4] Ellmerich S, Scholler M, Duranton B, Gosse F, Galluser M, Klein J-P, et al.Promotion of intestinal carcinogenesis by Streptococcus bovis. Carcinogenesis2000;21:753–6.

[5] Biarc J, Nguyen IS, Pini A, Gosse F, Richert S, Thierse D, et al. Carcinogenicproperties of proteins with pro-inflammatory activity from Streptococcusinfantarius (formerly S. bovis). Carcinogenesis 2004;25:1477–84.

[6] Wentling GK, Metzger PP, Dozois EJ, Chua HK, Krishna M. Unusual bacterialinfections and colorectal carcinoma – Streptococcus bovis and Clostridiumsepticum: report of three cases. Dis Colon Rectum 2006;49:1223–7.

[7] Lancefield RC. A serological differentiation of human and other groups ofhemolytic Streptococci. J Exp Med 1933;57:571–95.

[8] Schlegel L, Grimont F, Ageron E, Grimont PAD, Bouvet A. Reappraisal of thetaxonomy of the Streptococcus bovis/Streptococcus equinus complex andrelated species: description of Streptococcus gallolyticus subsp. gallolyticussubsp. nov., S. gallolyticus subsp. macedonicus subsp. nov. and S. gallolyticussubsp. pasteurianus subsp. nov. Int J Syst Evol Microbiol 2003;53:631–45.

[9] Sly LI, Cahill MM, Osawa R, Fujisawa T. The tannin-degrading species Strep-tococcus gallolyticus and Streptococcus caprinus are subjective synonyms. IntJ Syst Bacteriol 1997;47:893–4.

[10] Osawa R, Sasaki E. Novel observations of genotypic and metabolic character-istics of three subspecies of Streptococcus gallolyticus. J Clin Microbiol2004;42:4912–3.

[11] van’t Wout JW, Bijlmer HA. Bacteremia due to Streptococcus gallolyticus, or theperils of revised nomenclature in bacteriology. Clin Infect Dis 2005;40:1070–1.

[12] Brooker JD, Lum DK, Thomson AM, Ward HM. A gene-targeting suicide vectorfor Streptococcus bovis. Lett Appl Microbiol 1995;21:292–7.

[13] Whitehead TR, Cotta MA. Development of molecular methods for identifi-cation of Streptococcus bovis from human and ruminal origins. FEMS Micro-biol Lett 2000;182:237–40.

[14] Ghali MB, Scott PT, Al Jassim RAM. Characterization of Streptococcus bovisfrom the rumen of the dromedary camel and Rusa deer. Lett Appl Microbiol2004;39:341–6.

[15] Darjee R, Gibb AP. Serological investigation into the association betweenStreptococcus bovis and colonic cancer. J Clin Pathol 1993;46:1116–9.

[16] Klieve AV, Heck GL, Prance MA, Shu Q. Genetic homogeneity and phagesusceptibility of ruminal strains of Streptococcus bovis isolated in Australia.Lett Appl Microbiol 1999;29:108–12.

[17] Kurtovic A, Jarvis GN, Mantovani HC, Russell JB. Ability of lysozyme and2-deoxyglucose to differentiate human and bovine Streptococcus bovis strains.J Clin Microbiol 2003;41:3951–4.

[18] Jett BD, Huycke MM, Gilmore MS. Virulence of enterococci. Clin Microbiol Rev1994;7:462–78.

[19] Clarridge III JE, Attorri SM, Zhang Q, Bartell J. 16S Ribosomal DNA sequenceanalysis distinguishes biotypes of Streptococcus bovis: Streptococcus bovis

biotype II/2 is a separate genospecies and the predominant clinical isolate inadult males. J Clin Microbiol 2001;39:1549–52.

[20] Herrero IA, Rouse MS, Piper KE, Alyaseen SA, Steckelberg JM, Patel R.Reevaluation of Streptococcus bovis endocarditis cases from 1975 to 1985by 16S ribosomal DNA sequence analysis. J Clin Microbiol 2002;40:3848–50.

[21] zur Hausen H. Streptococcus bovis: causal or incidental involvement in cancerof the colon? Int J Cancer 2006;119:xi–xii.

[22] Ruoff KL, Miller SI, Garner CV, Ferraro MJ, Calderwood SB. Bacteremia withStreptococcus bovis and Streptococcus salivarius: clinical correlates of moreaccurate identification of isolates. J Clin Microbiol 1989;27:305–8.

[23] Devriese LA, Vandamme P, Pot B, Vanrobaeys M, Kersters K, Haesebrouck F.Differentiation between Streptococcus gallolyticus strains of human clinicaland veterinary origins and Streptococcus bovis strains from the intestinaltracts of ruminants. J Clin Microbiol 1998;36:3520–3.

[24] Schlegel L, Grimont F, Collins MD, Regnault B, Grimont PAD, Bouvet A.Streptococcus infantarius sp. nov., Streptococcus infantarius subsp. infantariussubsp. nov. and Streptococcus infantarius subsp. coli subsp. nov., isolated fromhumans and food. Int J Syst Evol Microbiol 2000;50:1425–34.

[25] Potter MA, Cunliffe NA, Smith M, Miles RS, Flapan AD, Dunlop MG.A prospective controlled study of the association of Streptococcus bovis withcolorectal carcinoma. J Clin Pathol 1998;51:473–4.

[26] Yamaoka Y, Kodama T, Kita M, Imanishi J, Kashima K, Graham DY. Relationbetween clinical presentation, Helicobacter pylori density, interleukin 1b and8 production, and cagA status. Gut 1999;45:804–11.

[27] El-Omar EM. The importance of interleukin 1b in Helicobacter pylori associ-ated disease. Gut 2001;48:743–7.

[28] Hungate RE. The rumen microbial ecosystem. Annu Rev Ecol Syst 1975;6:39–66.

[29] Russell JB. A proposed mechanism of monensin action in inhibiting ruminalbacterial growth: effects on ion flux and protonmotive force. J Anim Sci1987;64:1519–25.

[30] Gill HS, Shu Q, Leng RA. Immunization with Streptococcus bovis protectsagainst lactic acidosis in sheep. Vaccine 2000;18:2541–8.

[31] Freer SN. Purification and characterization of the extracellular a-amylasefrom Streptococcus bovis JB1. Appl Environ Microbiol 1993;59:1398–402.

[32] Asanuma N, Hino T. Regulation of fermentation in ruminal bacterium,Streptococcus bovis, with special reference to rumen acidosis. Anim SciJ 2002;73:313–25.

[33] Shu Q, Bird SH, Gill HS, Rowe JB. Immunological cross-reactivity between thevaccine and other isolates of Streptococcus bovis and Lactobacillus. FEMSImmunol Med Microbiol 1999a;26:153–8.

[34] Nagaraja TG, Titgemeyer EC. Ruminal acidosis in beef cattle: the currentmicrobiological and nutritional outlook. J Dairy Sci 2006;90:E17–38.

[35] Bond DR, Russell JB. A role for fructose 1,6-diphosphate in the ATPase-mediated energy spilling reaction of Streptococcus bovis. Appl EnvironMicrobiol 1996;62:2095–9.

[36] Russell JB, Strobel HJ, Driessen AJM, Konings WN. Sodium-dependenttransport of neutral amino acids by whole cells and membrane vesicles ofStreptococcus bovis, a ruminal bacterium. J Bacteriol 1988;170:3531–6.

[37] Newbold CJ, Wallace RJ. Effects of the ionophores monensin and tetronasinon simulated development of ruminal lactic acidosis in vitro. Appl EnvironMicrobiol 1988;54:2981–5.

[38] Gozho GN, Plaizier JC, Krause DO, Kennedy AD, Wittenberg KM. Subacuteruminal acidosis induces ruminal lipopolysaccharide endotoxin release andtriggers an inflammatory response. J Dairy Sci 2005;88:1399–403.

[39] Nagaraja TG, Chengappa MM. Liver abscesses in feedlot cattle: a review.J Anim Sci 1998;76:287–98.

[40] Kleen JL, Hooijer GA, Rehage J, Noordhuizen JPTM. Subacute ruminal acidosis(SARA): a review. J Vet Med A 2003;50:406–14.

[41] Thoefner MB, Pollitt CC, van Eps AW, Milinovich GJ, Trott DJ, Wattle O, et al.Acute bovine laminitis: a new induction model using alimentary oligo-fructose overload. J Dairy Sci 2004;87:2932–40.

[42] Glock RD, DeGroot BD. Sudden death of feedlot cattle. J Anim Sci 1998;76:315–9.

[43] Carter RR, Grovum WL. A review of the physiological significance of hyper-tonic body fluids on feed intake and ruminal function: salivation, motilityand microbes. J Anim Sci 1990;68:2811–32.

[44] Huber TL. Physiological effects of acidosis on feedlot cattle. J Anim Sci1976;43:902–9.

[45] Bryant MP, Robinson IM, Lindahl IL. A note on the flora and fauna in therumen of steers fed a feedlot bloat-provoking ration and the effect of peni-cillin. Appl Environ Microbiol 1961;9:511–5.

[46] Merck. In: Fraser CM, editor. Merck veterinary manual. A handbook ofdiagnosis, therapy, and disease prevention and control for the veterinarian.7th ed. Rahway, NJ: Merck Publishing Group; 1991. p. 163–6.

[47] Duren E, Miller CR. Feedlot bloat-prevention and treatment, Beef cattlehandbook–cattle producers’ library. 2nd ed. Publication #CL625. WesternBeef Resource Committee; 2002.

[48] Gutierrez J, Davis RE, Lindahl IL, Warwick EJ. Bacterial changes in the rumenduring the onset of feed-lot bloat of cattle and characteristics of Peptos-treptococcus elsdenii n. sp. Appl Microbiol 1959;7:16–22.

[49] O’Donovan L, Brooker JD. Effect of hydrolysable and condensed tannins ongrowth, morphology and metabolism of Streptococcus gallolyticus (S. capri-nus) and Streptococcus bovis. Microbiology 2001;147:1025–33.

P. Herrera et al. / Anaerobe 15 (2009) 44–5454

[50] Barnes IJ, Seeley HW, vanDemark PJ. Nutrition of Streptococcus bovis inrelation to dextran formation. J Bacteriol 1961;82:85–93.

[51] Yabuuchi Y, Tani M, Matsushita Y, Otsuka H, Kobayashi Y. Effects of lauric acidon the physical, chemical and microbial characteristics in the rumen of steerson a high grain diet. Anim Sci J 2007;78:387–94.

[52] Kung L, Hession AO. Preventing in vitro lactate accumulation in ruminalfermentations by inoculation with Megasphaera elsdenii. J Anim Sci1995;73:250–6.

[53] Owens FN, Secrist DS, Hill WJ, Gill DR. Acidosis in cattle: a review. J Anim Sci1998;76:275–86.

[54] Beauchemin KA, Yang WZ, Rode LM. Effects of barley grain processing on thesite and extent of digestion of beef feedlot finishing diets. J Anim Sci2001;79:1925–36.

[55] Schwartzkopf-Genswein KS, Beauchemin KA, Gibb DJ, Crews Jr DH,Hickman DD, Streeter M, et al. Effect of bunk management on feedingbehavior, ruminal acidosis and performance of feedlot cattle: a review. J AnimSci 2003;81:E149–58.

[56] Stone WC. Nutritional approaches to minimize subacute ruminal acidosis andlaminitis in dairy cattle. J Dairy Sci 2004;87:E13–26.

[57] Sudweeks EM, Ely LO, Mertens DR, Sisk LR. Assessing minimum amounts andform of roughages in ruminant diets: roughage value index system. J AnimSci 1981;52:1406–11.

[57a] Nocek JE, Kesler EM. Growth and rumen characteristics of holsteins fedpelleted or conventional diets. J Dairy Sci 1980;63:249–54.

[58] Callaway TR, Edrington TS, Rychlik JL, Genovese KJ, Poole TL, Jung YS, et al.Ionophores: Their use as ruminant growth promotants and impact on foodsafety. Curr Issues Intest Microbiol 2003;4:43–51.

[59] Bartley EE, Nagaraja TG, Pressman ES, Dayton AD, Katz MP, Fina LR. Effects oflasalocid or monensin on legume or grain (feedlot) bloat. J Anim Sci1983;56:1400–6.

[60] McGuffey RK, Richardson LF, Wilkinson JID. Ionophores for dairy cattle:current status and future outlook. J Dairy Sci 2001;84:E194–203.

[61] Lana RP, Russell JB. Use of potassium depletion to assess adaptation ofruminal bacteria to ionophores. Appl Environ Microbiol 1996;62:4499–503.

[62] Martin SA, Russell JB. Transport and phosphorylation of disaccharides by theruminal bacterium Streptococcus bovis. Appl Environ Microbiol 1987;53:2388–93.

[63] Callaway TR, Adams KA, Russell JB. The ability of ‘‘low G.þ.C Gram-positive’’ruminal bacteria to resist monensin and counteract potassium depletion.Curr Microbiol 1999;39:226–30.

[64] Newbold CJ, Wallace RJ, Watt ND. Properties of ionophore-resistant Bacter-oides ruminicola enriched by cultivation in the presence of tetronasin. J ApplBacteriol 1992;72:65–70.

[65] Sniffen CJ, Chaucheyras-Durand F, de Ondarza MB, Donaldson G. Pre-dicting the impact of a live yeast strain on rumen kinetics and rationformulation. In: Proceedings of the 19th southwest nutrition andmanagement conference. Arizona, USA: Phoenix; 26–27 February 2004.pp. 53–59.

[66] DiLorenzo N, Diez-Gonzalez F, DiCostanzo A. Effects of feeding polyclonalantibody preparations on ruminal bacterial populations and ruminal pH ofsteers fed high-grain diets. J Anim Sci 2006;84:2178–85.

[67] Ricke SC, Schaefer DM, Cook ME, Kang KH. Differentiation of ruminalbacterial species by enzyme-linked immunosorbent assay using egg yolkantibodies from immunized chicken hens. Appl Environ Microbiol1988;54:596–9.

[68] Curtain CC, Clark BL, Dufty JH. The origins of the immunoglobulins in themucous secretion of cattle. Clin Exp Immunol 1971;8:335–44.

[69] Shu Q, Gill HS, Hennessy DW, Leng RA, Bird SH, Rowe JB. Immunisationagainst lactic acidosis in cattle. Res Vet Sci 1999b;67:65–71.

[70] Muniesa M, Jofre J. Abundance in sewage of bacteriophages that infectEscherichia coli O157:H7 and that carry the Shiga toxin 2 gene. Appl EnvironMicrobiol 1998;64:2443–8.

[71] Sklar IB, Joerger RD. Attempts to utilize bacteriophage to combat Salmo-nella enterica serovar enteritidis infection in chickens. J Food Safety2001;21:15–29.

[72] Carlton RM. Phage therapy: past history and future prospects. Arch ImmunolTher Exp 1999;47:267–74.

[73] Klieve AV, Hegarty RS. Opportunities for biological control of ruminalmethanogenesis. Aust J Agric Res 1999;50:1315–9.

[74] Klieve AV, Hudman JF, Bauchop T. Inducible bacteriophages from ruminalbacteria. Appl Environ Microbiol 1989;55:1630–4.

[75] Styriak I, Spanova A, Zitnan R. Partial characterization of two ruminalbacteriophages with similar restriction patterns and different capsidsmorphology. Arch Tierz 2005;48:572–9.

[76] Kudva IT, Jelacic S, Tarr PI, Youderian P, Hovde CJ. Biocontrol of Escherichia coliO157 with O157-specific bacteriophages. Appl Environ Microbiol1999;65:3767–73.

[77] Iverson WG, Millis NF. Succession of Streptococcus bovis strains with differingbacteriophage sensitivities in the rumens of two fistulated sheep. ApplEnviron Microbiol 1977;33:810–3.