Isolation, selection, and characterization of lactic acid bacteria for a competitive exclusion product to reduce shedding of escherichia coli o157:h7 in cattle

Journal of Food Protection, Vol. 66, No. 3, 2003, Pages 355–363Copyright q, International Association for Food Protection Isolation, Selection, and Characterization of Lactic Acid Bacteria
for a Competitive Exclusion Product To Reduce Shedding of
Escherichia coli O157:H7 in Cattle
M. M. BRASHEARS,1* D. JARONI,2 AND J. TRIMBLE2
1Department of Animal and Food Sciences, Texas Tech University, P.O. Box 42141, Lubbock, Texas 79409; and 2Department of Veterinary and Biomedical Sciences, University of Nebraska–Lincoln, 124 VDC, East, Lincoln, Nebraska 68583-0907, USA MS 02-206: Received 20 June 2002/Accepted 16 September 2002 ABSTRACT
Lactic acid bacteria (LAB) were selected on the basis of characteristics indicating that they would be good candidates for a competitive exclusion product (CEP) that would inhibit Escherichia coli O157:H7 in the intestinal tract of live cattle.
Fecal samples from cattle that were culture negative for E. coli O157:H7 were collected. LAB were isolated from cattle fecesby repeated plating on deMan Rogosa Sharpe agar and lactobacillus selection agar. Six hundred eighty-six pure colonies wereisolated, and an agar spot test was used to test each isolate for its inhibition of a four-strain mixture of E. coli O157:H7. Threehundred fty- ve isolates (52%) showed signi cant inhibition. Seventy- ve isolates showing maximum inhibition werescreened for acid and bile tolerance. Most isolates were tolerant of acid at pH levels of 2, 4, 5, and 7 and at bile levels of0.05, 0.15, and 0.3% (oxgall) and were subsequently identi ed with the API system. Lactobacillus acidophilus, Lactobacillusfermentum, Lactobacillus delbreukii, Lactobacillus salivarius, Lactobacillus brevis, Lactobacillus cellobiosus, Leuconostocspp., and Pediococcus acidilactici were the most commonly identi ed LAB. Nineteen strains were further tested for antibioticresistance and inhibition of E. coli O157:H7 in manure and rumen uid. Four of these 19 strains showed susceptibility to allof the antibiotics, 13 signi cantly reduced E. coli counts in manure, and 15 signi cantly reduced E. coli counts in rumen uid(P , 0.05) during at least one of the sampling periods. One of the strains, M35, was selected as the best candidate for a CEP.
A 16S rRNA sequence analysis of M35 revealed its close homology to Lactobacillus crispatus. The CEP developed will beused in cattle-feeding trials.
Escherichia coli O157:H7 belongs to the enterohemor- Because E. coli O157:H7 is shed in the feces, concerns rhagic group of E. coli and is an important human patho- about the safety of products of bovine origin have been gen, causing bloody diarrhea (hemorrhagic colitis) and re- raised, since the intestinal tract can be a source of contam- nal failure (hemolytic uremic syndrome). It was rst iden- ination of the nal product. The carcass can be contami- ti ed as a human pathogen in 1982 following two food- nated with the pathogen during slaughter through the punc- related outbreaks of hemorrhagic colitis (25). Both turing of the intestinal tract or by fecal material on the hide.
outbreaks were linked to ground beef. Since then, numerous Considering the high prevalence of the pathogen in the fe- outbreaks of diseases caused by E. coli O157:H7 have been ces and the potential contamination of carcasses during pro- reported around the world (4, 6, 15–18, 20, 24, 25, 27–29). cessing, appropriate measures to reduce or eliminate the Recent estimates by the Centers for Disease Control suggest carriage of E. coli O157:H7 in cattle should begin at the that about 74,000 illnesses and 250 deaths due to E. coli farm. Competitive exclusion involves the use of live mi- O157:H7 infections occur each year in the United States crobial cultures that exhibit antagonistic effects against spe- alone (15, 20). Although several sources of infection have ci c groups of organisms, resulting in a decrease in their been identi ed, the principal vehicle implicated in out- numbers in the intestinal tract. The U.S. Food and Drug breaks has been ground beef. Additionally, some of the out- Administration (FDA) recently approved the rst competi- breaks associated with nonbovine products have been tive exclusion product (CEP) for use in commercial poultry linked to cross-contamination of the implicated food with production to reduce salmonella in chicks. Recent studies indicate that competitive exclusion may be an effective in- Cattle have been implicated as the principal reservoirs tervention strategy for pathogen control for cattle (31). of E. coli O157:H7 (3, 5, 10, 19, 21, 26, 30, 32), with the It has long been known that lactic acid bacteria (LAB) primary sites of localization being the rumen and the colon.
inhibit pathogens in laboratory media and in foods. Several Recent studies indicate that up to 30% of the animals in a LAB, most commonly strains from the genera Lactobacil- feedlot may shed the pathogen and that up to 70% of feedlots lus, Enterococcus, and Streptococcus, have been tested as contain at least one E. coli O157:H7–positive animal (9). probiotic agents or CEPs for livestock. Additionally, pro-biotic LAB are fed to cattle to improve animal perfor- * Author for correspondence. Tel: 806-742-2469; Fax: 806-742-2427; mance, so a probiotic agent could be easily integrated into In order to produce the desired effects, probiotic or 48 h in plastic bags ushed with CO2 for 30 s. Approximately 10 CEP bacteria must be carefully selected and screened to to 15 well-isolated colonies were picked from each plate and maximize their inhibitory activity in the animals. In addi- transferred to individual tubes containing 10 ml of MRS broth, tion to being able to inhibit E. coli O157:H7 in laboratory which were further incubated at 378C for 18 to 72 h to obtain media, the cultures should be isolated from the host animal maximum growth of the cultures. The isolated cultures were re-streaked onto lactobacillus selection agar and MRS agar plates and should be able to adapt to the intestinal environment.
and incubated at 378C for 48 h until isolated colonies of one form Most bacteria do not survive well at low pH values. The were obtained. Pure colonies were Gram stained for preliminary severe acidic conditions of the digestive tract could have identi cation. The isolated cultures were maintained as frozen an adverse effect on the bacteria. It is therefore important (2708C) stocks in MRS broth supplemented with 10% (vol/vol) that microorganisms to be used as probiotic agents or CEPs sterile glycerol. Isolates were subcultured in MRS broth at 378C be tested for acid tolerance (7). Bile secreted in the duo- for 24 to 48 h before they were used for further studies.
denum (small intestine) also reduces the survival of bacteriaby destroying their cell membranes, whose major compo- Preliminary screening of isolated colonies for inhibition
nents are lipids and fatty acids (13, 14). Hence, the success of E. coli O157:H7. Isolated strains were screened for their in-
hibition of E. coli O157:H7 with an agar spot test. The agar spot
of a probiotic agent or CEP also depends on the bile-resis- test was conducted with fresh cultures of individual strains that tant qualities of the selected strain. The cultures should also were spot inoculated onto the surfaces of MRS agar plates and grow rapidly and survive well during storage so that viable incubated for 24 to 48 h at 378C in plastic bags ushed with CO2 cultures are delivered to the animals (1, 11). for 30 s. Four strains of E. coli O157:H7, obtained from the food A recently introduced criterion for new probiotic or microbiology laboratory at UNL (strains 920, 922, 944, and 966, competitive exclusion strains is their susceptibility to anti- all cattle isolates), were individually prepared by inoculating 10 biotics used for humans and food-producing animals. It is ml of tryptic soy broth (Difco Becton Dickinson) with frozen now well known that bacteria can acquire or develop resis- (2708C) stocks of each E. coli strain and incubating it at 378C tance to antimicrobial agents, which may pose a threat to for 18 to 24 h. A cocktail of the four E. coli O157:H7 strains was human health (10). The emergence of resistant bacteria can prepared and inoculated into 10 ml of tempered soft tryptic soyagar (TSA; Difco) to obtain a nal concentration of approximately lead either to widespread antibiotic resistance of opportu- 5 3 107 CFU/ml. This mixture was then poured onto the surface nistic pathogens, which enhances the potential threat posed of spot-inoculated MRS agar plates containing the LAB. After 24 by these organisms, or to the transfer of resistance to human h of incubation at 378C, the sizes of inhibition zones around col- pathogens through the gastrointestinal tract or the environ- onies were measured with calipers to evaluate inhibitory activity.
ment (10). It is therefore considered important that bacteria Previously isolated and screened control strains (positive and neg- selected for probiotic use do not exhibit resistance to stan- ative) were used to validate the effectiveness of the test. Inhibitory dard antibiotics used for humans and food-producing ani- activity was considered signi cant when the width of the clear zone around the colonies of the isolated strains was The overall objective of this study was therefore to The test for each isolate against the pathogen cocktail was repli- develop a LAB CEP for cattle that would potentially reduce the intestinal colonization and fecal shedding of E. coli Tests for acid and bile tolerance. The 75 isolates that most
O157:H7. The speci c objectives of this study were (i) to extensively inhibited E. coli O157:H7 in the agar spot test were isolate strains of LAB from healthy cattle; (ii) to screen screened for acid and bile tolerance. The method of Conway et isolated LAB strains for their inhibition of E. coli O157: al. (7) was used to study the acid tolerance of cattle isolates. Brief- H7; (iii) to screen LAB for acid and bile tolerance, storage ly, a fresh culture of the LAB was harvested from MRS broth by stability, and antibiotic resistance; (iv) to identify the LAB centrifugation, and pellets were washed and resuspended in sterile strains selected; and (v) to test these LAB strains for an- phosphate-buffered saline (PBS; Fisher Scienti c, Fair Lawn, tagonistic action against E. coli O157:H7 in rumen uid N.J.). Each strain was individually added to 4 ml of sterile PBS adjusted to pH values of 2, 4, 5, and 7 (control) and incubatedfor 2, 4, and 24 h in a water bath at 378C. After each incubation MATERIALS AND METHODS
period, the growth of the strain was determined by measuringabsorbance at 620 nm. The acid tolerance of cattle isolates was Isolation of LAB strains from healthy cattle. Cattle that
categorized as excellent, very good, good, fair, and poor on the were fecal culture negative for E. coli O157:H7 were selected to basis of their growth at different pH values. The acid tolerance of supply fecal samples for the isolation of LAB that were to be an isolate showing increased growth at all pH values over a 24-h screened for their inhibitory activity against E. coli O157:H7. Ma- period was considered excellent; the acid tolerance of an isolate nure samples from cattle housed in the Animal Science Depart- showing increased growth at three or more pH levels over a 24- ment of the University of Nebraska–Lincoln (UNL) were obtained h period was considered very good; the acid tolerance of an isolate at 3-week intervals over a period of 12 weeks (four sampling was considered good if that isolate exhibited (i) increased growth at two or more pH levels for all incubation periods, (ii) increased One gram of each fecal sample was added to 10 ml of sterile growth for all incubation periods at one pH level and over a period deMan Rogosa Sharpe (MRS) broth (Difco Becton Dickinson Mi- of 4 h at two or more pH levels, or (iii) increased growth at all crobiology Systems, Sparks, Md.) and mixed thoroughly. The fe- pH levels for 4 h but not for 24 h; the acid tolerance of an isolate cal material was then streaked onto lactobacillus selection agar was considered fair if it exhibited an increase in growth at only (BBL Becton Dickinson Microbiology Systems, Cockeysville, one pH level over a 24-h period or at two pH levels over a 4-h Md.) and MRS agar plates. The plates were incubated at 378C for period; the acid tolerance of an isolate showing no increase in COMPETITIVE EXCLUSION PRODUCT TO REDUCE E. COLI O157:H7 growth at any pH level for 4 or 24 h was considered poor. Ex- Cow manure samples were obtained from the Animal Science De- periments were repeated three times.
partment at UNL on the day of the experiment, and about 20 g The method used for testing bile tolerance was similar to the of the manure was reserved for use as an uninoculated control. A one described by Gilliland et al. (14). A fresh culture of the se- 400-g manure sample was inoculated with the E. coli O157:H7 lected isolates was inoculated into tubes containing 10 ml of MRS cocktail to obtain a population of 1 3 105 CFU/ml, and this sam- broth with 0 (control), 0.05, 0.15, and 0.3% oxgall. The inoculated ple was divided into 20 sterile Whirl-Pak bags (20 g each). One tubes were incubated at 378C in a water bath. The growth of the of these manure samples was used as an inoculated control, and isolates was monitored at 2, 4, 6, and 24 h by measuring absor- each of the other samples was inoculated with 2 ml (1 3 107 bance at 660 nm. The test for each strain was repeated three times.
CFU/ml) of the fresh MRS broth cultures of LAB (one isolate permanure sample). Each of the samples was incubated at 378C and Storage stability. Nineteen of the 75 isolates selected for
plated in duplicate on MSA supplemented with nalidixic acid (50 further studies were tested for storage stability in a commercial mg/ml) for 0, 24, and 48 h to monitor E. coli O157:H7 growth.
freezer (08C). Following isolation, an overnight culture of each The pH of each manure sample was recorded after 0, 24, and 48 isolate was prepared in MRS broth and supplemented with 10% h. Samples containing no viable colonies on MSA were subjected (vol/vol) sterile glycerol. About 1 ml of each culture was then to further enrichment to determine whether the cultures were pos- dispensed into sterile 1.5-ml cryogenic vials and frozen at 2708C.
itive or negative. The entire experiment was repeated three times.
These frozen stocks were transferred the next day to a freezer set To simulate rumen conditions by the method of de Vaux et at 08C. The viability of the isolates was tested on day 0 and after al. (8), fresh bovine rumen uid obtained from the Animal Science 3 months of storage in the commercial freezer by plating them on Department at UNL was mixed with 1 g of corn starch per liter, MRS agar and incubating them at 378C for 18 to 24 h.
2 g of maltose per liter, and equal parts of phosphate buffer so- When all the screening procedures had been completed (24 months), the isolates nally selected as potential CEPs were test- ed for viability by plating concentrated cultures on MRS agar at 2PO4 [Fisher Scienti c] mixed together). The mix- ture was sterilized and put in an anaerobic jar to facilitate thor- appropriate dilutions and incubating the plates at 378C for 18 to ough reduction for 1 to 4 days. After reduction, about 10 ml of rumen uid mixture was reserved for use as uninoculated control Preliminary identi cation of isolates with the API. On the
uid and about 250 ml of the mixture was inoculated with 0.25 basis of their inhibitory activity against E. coli O157:H7 and their ml of E. coli O157:H7 cocktail as prepared in the manure exper- tolerance to acid and bile, 75 isolates were chosen for identi ca- iment. The inoculated mixture was divided into 20 sterile tubes tion by fermentation patterns with standard Analytical Pro le In- (10 ml each). One of these samples was kept for use as an inoc- dex (API; bioMe´rieux, Inc., Hazelwood, Mo.) tests. API 50CHL, ulated control sample, and the other samples were inoculated with API 20Strep, and API 20E were used to identify gram-positive 1 ml of freshly prepared cultures of selected LAB (one isolate per rods, gram-positive cocci, and gram-negative rods, respectively.
sample). Each of the samples were anaerobically incubated at378C and plated in duplicate on MSA supplemented with nalidixic Antibiotic resistance. Isolates were selected after further
acid for 0, 24, and 48 h to monitor E. coli O157:H7 growth.
API screening for resistance to a panel of six antibiotics repre- Samples with counts below the detection limit on MSA were sub- senting categories of antimicrobial agents that are commonly used jected to further enrichment to determine if E. coli O157 was for animals and humans (10). The antibiotics used in the present present. The pHs of all rumen uid samples were recorded after study included erythromycin, ampicillin, tetracycline, trimetho- 0, 24, and 48 h. The entire experiment was repeated three times.
prim-sulfamethoxazole, grepa oxacin (quinolone), and levo ox-acin ( uoroquinolone). A commercial antimicrobial susceptibility Identi cation of strains by ribotyping and 16S rRNA se-
test kit, Etest (AB Biodisk, North America Inc., Piscataway, N.J.), quence analysis. A genotypic ngerprint analysis of 13 isolates
was used to determine the MICs of the antibiotics for the isolates was carried out at the Laboratory for Molecular Typing at Cornell selected. Etest strips containing a prede ned exponential gradient University (Ithaca, N.Y.). Additionally, one of the cultures was of a particular antibiotic were applied to agar plates containing identi ed by 16S rRNA sequence analysis. In this analysis, a por- Mueller-Hinton agar (Difco) preinoculated with an overnight cul- tion of the 16S rRNA gene of a bacterium was ampli ed and ture of the given bacteria (an inoculum of 0.5 McFarland turbidity sequenced and was then compared with sequences contained in in 0.85% NaCl). The agar plates with the strips were immediately the GenBank public database to obtain matches to the isolate.
incubated at 378C for 16 to 20 h, when bacterial growth becamedistinctly visible. After incubation, MIC values (mg/ml) were read Statistical analysis. Analysis of variance was used to per-
at the point at which the inhibition ellipse (centered along the form interaction studies for manure and rumen uid with a ran- strip) intersected the scale at the edge of the strip.
domized complete block design using the single-factor repeated-measures analysis of the Statistical Analysis System software Inhibition of E. coli O157:H7 in rumen uid and manure.
(SAS Institute Inc., Cary, N.C.). The means of duplicate plate Nineteen isolates, selected by the previously described screening counts were converted to a log10 scale before the analysis was methods, were further tested for their antagonistic action against carried out. A factorial design (20 3 3, with each plate count as strains of E. coli O157:H7 in rumen uid and manure. To facilitate a sampling unit and duplicate plate counts as an experimentalunit) their recovery in the presence of background ora, the E. coli was used. Differences among treatment means were determined O157:H7 strains used in the initial inhibition study were selected by the least square difference method.
for nalidixic acid resistance as described by Zhao et al. (31). Iso-lates were tested in the presence of increasing concentrations of RESULTS AND DISCUSSION
nalidixic acid, and cells that were resistant to 50 mg of nalidixicacid per ml were selected.
Isolation of LAB. Of 686 isolates, 15% were gram
A cocktail of the four E. coli strains was prepared. This mix- negative and 85% were gram positive. Gram staining in- ture was serially diluted, pour plated onto TSA, and incubated at dicated that 72% of the gram-positive isolates were gram- 378C for 48 h to determine nal concentrations of the pathogen.
positive rods and 27% were gram-positive cocci. Gram re- TABLE 1. Cattle isolates chosen for further analysis on the basis 0.5 mm), and 275 isolates (40%) produced inhibition of E. coli O157:H7 inhibition, Gram reaction, bile tolerance,a and zones of 1 mm. The sizes of inhibition zones ranged from Analytical Pro le Index (API) pro le 0.5 mm to as large as 12 mm. Inhibition zones for the most inhibitory isolates are illustrated in Table 1.
The cause of inhibition was not identi ed, but since most of these isolates were LAB, the inhibition observed was probably due to one of the mechanisms employed by LAB during competitive exclusion. Because of the design of the agar spot test, it is likely that the LAB that inhibited E. coli O157 were producing substances that were inhibi- tory to the pathogen. The observed degree of inhibition of E. coli O157:H7 by the isolated LAB was not unexpected.
Several studies have reported the antagonistic effects of LAB on various other pathogenic microorganisms in broth, Lactobacillus delbreukii delbreukii Acid and bile tolerance. Seventy- ve of the most in-
hibitory isolates were further screened for acid and bile tolerance. These 75 isolates were selected because they in- hibited E. coli O157, were clearly identi ed as pure iso- lates, and were from distinct separate sources (i.e., they were from different animal sources and/or were collected at different sampling times). Most of the isolates showed acid resistance: 23% of them showed excellent to very good a Acid and bile tolerance data are not shown. All isolates were acid tolerance; 34% of them showed good acid tolerance; resistant to all acid and bile conditions.
and 30% of them showed fair acid tolerance. Only 13% of b Mean for three separate trials. Inhibition was measured as the the isolates showed decreased growth at all pH values over size of the clear zone around the colony in an agar spot test.
a period of 24 h. The isolates not showing good acid re-sistance were not considered good candidates for a CEP.
Bile tolerance studies indicated that all isolates except actions of the most inhibitory isolates, selected for further one showed increased growth under all bile conditions (0, investigation, are illustrated in Table 1. Several criteria were 0.05, 0.15, and 0.30%). The isolate that did not show in- used to select isolates for further screening. The type of creased growth was not considered a good candidate for a organism, the host animal, and the source of the organism Acid and bile tolerance were examined to predict the Although several genera and species of microorgan- survival of the CEP microorganisms in the intestinal tracts isms have been used as live microbial cultures in animals, of live animals. Bacteria used as CEPs are generally deliv- the most commonly used bacteria are intestinal strains of ered to the animals through their food and thus should have LAB. LAB have a history of safe use and are found uni- the ability to resist the stressful conditions in the intestinal versally. Additionally, the focus of our selection criteria tract, including acid and bile secretion. The CEP microor- was on obtaining LAB isolates that were generally recog- ganisms can be subjected to cellular stress arising from low nized as safe. Numbers of lactobacilli in the large intestine pH values in the rumen and bile secretion in the intestinal range from 108 to 109 per gram (dry weight) of gut con- tract. Rumen pH, which, like other factors, depends on the tents, so the isolation of this group of organisms from the nature of the diet, can range from 4.3 to 6.5 (7), with diets fecal material of cattle was not dif cult.
including large amounts of starch and soluble carbohydrates The proper selection of the strains on the basis of sev- resulting in lower pH values than diets that are high in eral criteria is the most important factor in the ef cacy of cellulose and other complex carbohydrates (7, 13). a CEP. We isolated our organisms from cattle because host The LAB isolates were also tolerant to various bile speci city appears to play a role in the ef cacy of a CEP.
conditions. Because the isolates were obtained from cattle Studies involving pigs (2), chickens (12), and calves (22) and had already been subjected to bile conditions within have demonstrated that isolates from host animals have the animals, they were intrinsically bile tolerant. On the been effective in colonizing the host, but isolates from other basis of the results of acid and bile tolerance tests, these hosts have not. Therefore, we selected E. coli O157–inhib- isolates show promise with regard to their use as CEPs in iting LAB from cattle for a product that would be fed to live animals because of their ability to survive intestinal cattle to reduce shedding of the pathogen.
Inhibition of E. coli O157:H7. Agar spot tests were
API. Some isolates were eliminated on the basis of the
used to test all 686 isolates for their inhibition of E. coli API either because their identi cation was not clear (un- O157:H7. Three hundred fty- ve isolates (52%) met or acceptable or doubtful pro le) or because they were not exceeded the inhibition criteria (with inhibition zones of LAB. On the basis of the API, the most commonly iden- COMPETITIVE EXCLUSION PRODUCT TO REDUCE E. COLI O157:H7 ti ed strains of LAB were Lactobacillus acidophilus, Lac- TABLE 2. MICs of antibiotics for cattle isolatesa tobacillus fermentum, Lactobacillus delbreukii, Lactobacil- lus salivarius, Lactobacillus brevis, Lactobacillus cellobio- sus, Leuconostoc spp., and Pediococcus acidilactici (Table1). Of the 75 isolates identi ed and screened for acid and bile tolerance, 19 were selected for further interaction stud- ies with manure and rumen uid. A list of the selected isolates along with their API pro les is given in Table 1.
Storage stability. To ensure that cultures could survive
shipment and storage and thus would be viable when de- livered to the animals, all cultures were evaluated for stor- age stability. Because they would be stored in a commercial freezer, commercial freezer conditions were mimicked in the laboratory after cultures were frozen at 2708C. All iso- lates except two maintained their viability over a period of 3 months of storage at 08C. Isolates (M35 and L411) that were selected as CEPs showed good survival after 24 months of storage at 2708C. A population of about 1 3 109 CFU/ml was obtained for both of these isolates after Test for antibiotic resistance. Of the 19 strains tested,
21% (4 of 19; M35, A43, C312, and I212) showed suscep- a EM, erythromycin; TS, trimethoprim-sulfamethoxazole; GP, gre- tibility to all of the antibiotics. Most isolates were resistant pa oxacin; AM, ampicillin; LE, levo oxacin; TC, tetracycline; to trimethoprim-sulfamethoxazole and grepa oxacin (63 H, highest MIC value on the scale (256 mg/ml for EM, AM, and and 58%, respectively). Thirteen of 19 isolates (68%) TC and 32 mg/ml for TS, GP, and LE); L, lowest MIC value on showed resistance to more than one antibiotic; these isolates were most commonly resistant to trimethoprim-sulfameth-oxazole, grepa oxacin, and tetracycline. Only one isolate portant in the treatment of a serious disease for which an (B41) was resistant to levo oxacin, and two isolates (T28 alternative therapy is available. Category III includes anti- and C314) were resistant to ampicillin. The results of the microbial agents that have little or no use in human med- antibiotic resistance tests are shown in Table 2. The two icine (10). We selected antibiotics from categories I and II selected isolates, M35 and L411, were not resistant to van- (vancomycin, grepa oxacin, and levo oxacin [both quino- comycin or cephalosporin. Both of these isolates exhibited lones] belong to category I, and the rest of the antibiotics zones larger than 15 and 18 mm for vancomycin and ceph- examined belong to category II). In addition to being used in humans, these antibiotics (with the exception of grepa- Antibiotic resistance of microorganisms used as pro- oxacin and levo axacin) are also used in food-producing biotic agents or CEPs is an area of growing concern. It is animals. Although only four isolates (21%) showed suscep- believed, but not scienti cally documented, that antimicro- tibility to all of the antibiotics, about 53% of them showed bial agents used for food-producing animals can promote susceptibility to at least four of the six antibiotics tested, the emergence of resistance in these bacteria that can lead and only four isolates were resistant to three or more an- to the transfer of resistance to other pathogenic bacteria tibiotics. Isolates that had high resistance pro les were through the exchange of genetic material or can increase eliminated from consideration as CEP microorganisms.
the potential threat posed by these organisms as opportu- Since isolate M35 was not resistant to any of the an- nistic pathogens. Selected isolates were tested for their re- tibiotics, it is safer for use in live animals and eventually sistance to six antibiotics (erythromycin, tetracycline, am- in humans, and its lack of resistance also indicates that it picillin, grepa oxacin, levo oxacin, and trimethoprim-sul- may not contribute to the transfer of resistance to other famethoxazole), and the isolates selected for CEP were also microorganisms (especially foodborne pathogens).
tested for their resistance to vancomycin and cephalosporin(cephalothin). The basis for the selection of these antibi- Antagonistic action of cattle isolates against E. coli
otics was that the FDA is considering dividing antimicro- O157:H7 in manure and rumen uid. Selected isolates
bial drugs into three categories (I, II, and III) on the basis were further screened for their survival in the rumen and of their importance with regard to human medicine (10). colon, as well as for their antagonistic action against E. coli Category I includes antibiotics that are considered essential O157:H7. The interaction of E. coli O157:H7 with the se- for the treatment of a serious or life-threatening human dis- lected LAB was studied in manure and rumen uid after 0, ease that does not have an alternative treatment, for the 24, and 48 h of incubation. As expected, none of the strains treatment of a foodborne disease whose treatment has been showed differences from the control with regard to inhibi- limited by its resistance to alternative therapy, or for long- tion of E. coli O157:H7 at 0 h in both rumen uid and term therapy. Category II includes antibiotics that are im- TABLE 3. Growth of Escherichia coli O157:H7 in association with cattle isolates in manure at 0, 24, and 48 h E. coli O157:H7 count (log10 CFU/ml) Lactobacillus delbreukii delbreukii a Values with different letters are signi cantly different (P , 0.05).
Results of the interaction studies for manure indicate at 24 and 48 h in rumen uid is probably that some strains that after 24 h of incubation, the control (no LAB added) had faster growth rates than others and therefore were able showed an increase in E. coli O157:H7 counts of ca. 2.0 to affect E. coli growth faster. The simulated rumen envi- log10 cycles (Table 3). All strains signi cantly (P , 0.05) ronment (anaerobic) may have been more conducive to the reduced E. coli counts (by ca. 1 log10 cycle) compared with growth of the LAB than the manure, enabling these LAB the control count at 24 h. At 48 h, all strains except S310 to exhibit stronger antagonistic effects on E. coli O157:H7.
and R44 were able to maintain their reductions in E. coli Validation studies by Brashears et al. (3) have indicated that O157:H7 counts, and these reductions were statistically sig- the nalidixic acid–resistant E. coli O157:H7 strains are ni cant (P , 0.05) compared with the control count. Ad- more tolerant to stress conditions, including acid and bac- ditionally, samples inoculated with strains M35, L411, P31, teriocin stress, than are the nonresistant E. coli strains.
C312, T28, I33, and P32 had signi cantly lower (P , 0.05) However, the fact that selected LAB were still able to in- E. coli O157:H7 counts at 48 h than they did initially. No hibit E. coli in manure suggests that these bacteria have changes in the pH values of the treated samples (which ranged from 6.1 to 6.8 for all incubation periods) were ob- To select nal strains for the CEP, isolates were com- pared with regard to their inhibition of E. coli O157:H7 Interaction studies for rumen uid revealed that after and with regard to various other characteristics. The only 24 h of incubation, strains M38, C312, T21, H27, T28, two strains that showed signi cant E. coli O157:H7 inhi- C314, M35, A43, A46, L411, and R44 signi cantly (P , bition in both rumen uid and manure at 24 and 48 h of 0.05) reduced E. coli counts (by ca. 1 to 4 log10 cycles) incubation relative to the control count and relative to the compared with the control count at 24 h (Table 4). Strain initial count were strains M35 and L411. Strain M35 was M35 reduced the E. coli O157:H7 count from 5.6 log10 also susceptible to all of the antibiotics tested and exhibited CFU/ml to an undetectable level (,1.0 log10 CFU/ml) at excellent storage stability. Ultimately, this isolate was con- 24 h. With further enrichment, we were unable to detect sidered the ‘‘ideal’’ candidate for a CEP. L411 may also be the pathogen. At 48 h, all of the strains except I212 and a good candidate, but it exhibited some resistance to anti- C315 signi cantly inhibited E. coli O157:H7 compared biotics, so the introduction of this culture into a cattle-feed- with the control (P , 0.05). Strains H27, C314, M35, M38, ing program may not be desirable. The selection process is A43, A46, L411, and R44 reduced E. coli O157:H7 counts in rumen uid to undetectable levels at 48 h. With furtherenrichment, no viable cells were detected. No change in pH Ribotyping and 16S rRNA sequence analysis of se-
was observed for any of the treatments, with pH values lected isolates. The genotypic identi cation of the selected
ranging from 6.0 to 6.3 throughout the study.
cattle isolates was accomplished through ribotyping. Isolate The reason for the differences in inhibition observed L411 was identi ed as L. acidophilus. The pattern for iso- COMPETITIVE EXCLUSION PRODUCT TO REDUCE E. COLI O157:H7 TABLE 4. Growth of Escherichia coli O157:H7 in association with cattle isolates in rumen uid at 0, 24, and 48 h E. coli O157:H7 count (log10 CFU/ml) Lactobacillus delbreukii delbreukii a Values with different letters are signi cantly different (P , 0.05).
late M35 did not closely match any of the patterns in eitherthe Dupont or the LMT database. Isolate M35 was furtheridenti ed by 16S rRNA sequence analysis. Isolate M35,which was identi ed as L. acidophilus by the API fermen-tation pattern, was identi ed as Lactobacillus crispatus bycomparison of the 16S rRNA sequence of the strain with TABLE 5. Selection of isolates for competitive exclusion productto inhibit Escherichia coli O157:H7 in cattle the sequences in the GenBank database. M35 had signi -cant homology to L. crispatus (98%).
The discrepancies between the results obtained with the API system and those obtained by genetic typing can be explained by the different the basic characterization meth- ods used by the two systems: one is genotypic and one is phenotypic. The API system constitutes a phenotypic ap- proach that uses the carbohydrate fermentation pro le for the characterization of a particular strain. On the other hand, ribotyping and 16S rRNA analysis both use a molec- ular approach. A combination of the two methods is nec- essary to obtain the nal results. Sometimes molecular typ- ing characterizes two strains as being the same, but upon phenotypic evaluation it is clear that the two strains are not the same. Careful interpretation of both types of results is needed to ensure that strains are indeed the same or differ- ent with regard to the functional characteristic and with Isolate M35 was identi ed as L. acidophilus by the API system and as L. crispatus by 16S rRNA analysis. L. cris- patus (6) belongs to the L. acidophilus complex, which comprises six distinct homology groups (A1, A2, A3, A4, B1, and B2). The identi cation table for API 50CHL re-veals that L. acidophilus and L. crispatus share very similar a Inhibition is de ned as a signi cant reduction (P , 0.05) at 24 and 48 h relative to the control count and the initial count.
fermentation patterns, with the only exception being that L. b Number of antibiotics (of the six tested) to which an isolate was crispatus also ferments amylose, arabinose, and lactose by CONCLUSIONS
4. Carter, A. O., A. A. Borczyk, J. A. Carlson, B. Harvey, J. C. Hockin, M. A. Karmali, C. Krishnan, D. A. Korn, and H. Lior. 1987. A severe Because cattle have been identi ed as a reservoir for outbreak of Escherichia coli O157:H7-associated hemorrhagic colitis E. coli O157, it is important to focus on intervention strat- egies for these animals. The approach of competitive ex- 5. Chapman, P. A., D. J. Wright, P. Norman, J. Fox, and E. Crick. 1993.
Cattle as a possible source of verocytotoxin-producing Escherichia clusion seems to have potential in this regard. Zhao et al.
(31) tested a probiotic culture consisting of several non- 6. Cieslak, P. R., T. J. Barrett, P. M. Grif n, K. F. Gensheimer, G. Beck- pathogenic strains of E. coli isolated from the intestinal ett, J. Buf ngton, and M. G. Smith. 1993. Escherichia coli O157: tracts of cattle for the prevention of E. coli O157:H7 col- H7, infection from a manured garden. Lancet 342:367.
onization in young dairy calves. Their study revealed that 7. Conway, P. L., S. L. Gorbach, and B. R. Goldin. 1987. Survival of probiotic cultures were able to reduce E. coli counts in the lactic acid bacteria in the human stomach and adhesion to intestinalcells. rumen and the colon. Although several genera and species 8. de Vaux, A., M. Morrison, and R. W. Hutkins. Displacement of of microorganisms have been used as live microbial cul- Escherichia coli O157:H7 from rumen medium by prebiotic sugars.
tures in animals, we identi ed LAB as good candidates for Appl. Environ. Microbiol., in press.
9. Elder, R. O., J. E. Keen, G. R. Siragusa, G. A. Gallagher, M. Kooh- Although the mechanism(s) of action involved in the maraie, and W. W. Laegreid. 2000. Correlation of enterohemorrhagicEscherichia coli O157 prevalence in feces, hides, and carcasses of inhibition of E. coli O157:H7 was not investigated in the beef cattle during processing. Proc. Natl. Acad. Sci. USA 97:2999– present study, it is speculated that a decrease in pH due to the extensive production of organic acids was probably not 10. Food and Drug Administration. 2001. A proposed framework for the mechanism that played a role in the reduction of E. coli evaluating and assuring the human safety of the microbial effects of O157:H7, since no reduction in pH was observed in any of antimicrobial new animal drugs intended for use in food-producinganimals.
the manure or rumen uid samples over a 48-h period dur- ing the interaction studies. It is therefore believed that some 11. Fuller, R. 1989. Probiotics in man and animals. factor(s) other than a decrease in pH due to the production of organic acids by LAB may be involved in the inhibition 12. Fuller, R., and A. Turvey. 1971. Bacteria associated with the intes- of E. coli O157:H7. Such factors may include the produc- tinal wall of the fowl (Gallus domesticus). tion of bacteriocins, hydrogen peroxide, low-molecular- 13. Gilliland, S. E. 1987. Importance of bile tolerance in Lactobacilli weight metabolites such as diacetyl and CO2, or enzymes used as dietary adjunct, p. 149–155. In T. P. Lyons (ed.), Biotech- by the selected LAB. Future studies should therefore in- nology in the feed industry. Alltech Feed Co., Lexington, Ky.
volve further investigation of mechanism(s) of action used 14. Gilliland, S. E., T. E. Stanley, and L. J. Bush. 1984. Importance of by these isolates for the inhibition of E. coli O157:H7.
bile tolerance of Lactobacillus acidophilus used as a dietary adjunct.
On the basis of the results of this study, LAB isolated 15. Grif n, P. M., S. M. Ostroff, R. V. Tauxe, K. D. Greene, J. G. Wells, from cattle show promise for use as competitive exclusion J. H. Lewis, and P. A. Blake. 1988. Illnesses associated with Esch- microorganisms in live animals. In vitro tests showed that erichia coli O157:H7 infections: a broad clinical spectrum. our isolates had good survival capabilities in the intestinal tract and over long periods of storage. These isolates were 16. Honish, A. 1986. Summary of E. coli O157:H7 at an Edmonton able to inhibit or completely eliminate E. coli O157:H7 in extended care facility. Alberta Soc. Serv. Community Health 10:363–368.
laboratory media and in ruminal and colonic contents. Their 17. Keene, W. E., J. M. McAnulty, F. C. Hosley, L. P. Williams, K.
use as CEPs in cattle holds potential for the reduction of Hedberg, G. L. Oxman, T. J. Barrett, M. A. Pfaller, and D. W. Flem- E. coli O157:H7 carriage and the shedding of the pathogen ing. 1994. A swimming-associated outbreak of hemorrhagic colitis in feces. This reduction, in turn, should reduce the contam- caused by Escherichia coli O157:H7 and Shigella sonnei. ination of meat, milk, water, fruits, and vegetables by E. coli O157:H7, thereby reducing the occurrence of human 18. Keene, W. E., J. M. McAnulty, L. P. Williams, F. C. Hosley, K.
Hedberg, and D. W. Fleming. 1993. A two-restaurant outbreak of Escherichia coli O157:H7 enteritis associated with consumption ofmayonnaise, p. 353. In Program and abstracts of the 33rd Intersci- ACKNOWLEDGMENT
ence Conference on Antimicrobial Agents and Chemotherapy, NewOrleans.
The authors thank the State of Nebraska (LB1206) for funding this 19. Laegreid, W. W., R. O. Elder, and J. E. Keen. 1999. Prevalence of Escherichia coli O157:H7 in range beef calves at weaning. REFERENCES
20. Mead, P. S., L. Slutsker, V. Dietz, L. F. McCaig, J. S. Bresee, C.
1. Barrow, P. A., B. E. Brooker, R. Fuller, and M. J. Newport. 1980.
Shapiro, P. M. Grif n, and R. V. Tauxe. 1999. Food-related illness The attachment of bacteria to the gastric epithelium of the pig and its importance in the microecology of the intestine. 21. Orskov, F., I. Orskov, and J. A. Villar. 1987. Cattle as reservoir of verotoxin-producing Escherichia coli O157:H7. Lancet 2:276. (Let- 2. Borczyk, A. A., M. A. Karmali, H. Lior, and L. M. C. Duncan. 1987.
Bovine reservoir for verotoxin-producing Escherichia coli O157:H7.
22. Ozawa, K., K. Yabu-uchi, K. Yamanaka, Y. Yamashita, S. Nomura, and I. Oku. 1983. Effect of Streptococcus faecalis BIO-4R on intes- 3. Brashears, M. M., A. Amezquita, and J. Stratton. 2001. Validation tinal ora of weanling piglets and calves. of methods used to recover Escherichia coli O157:H7 and Salmo- nella spp. subjected to stress conditions. J. Food Prot. 64:1466– 23. Riley, L. W., R. S. Remis, S. D. Helgerson, H. B. McGee, J. G.
Wells, B. R. Davis, R. J. Hebert, E. S. Olcott, L. N. Johnson, N. T.
COMPETITIVE EXCLUSION PRODUCT TO REDUCE E. COLI O157:H7 Hargrett, P. A. Blake, and M. L. Cohen. 1983. Hemorrhagic colitis Tippen, H. D. Donnell, Jr., E. Geldreich, B. J. Payne, A. Meyer, Jr., associated with a rare Escherichia coli serotype. J. G. Wells, K. D. Greene, M. Bright, N. H. Bean, and P. A. Blake.
1992. A waterborne outbreak in Missouri of Escherichia coli O157: 24. Ryan, C. A., R. V. Tauxe, G. W. Hosek, J. G. Wells, P. A. Stoesz, H7 associated with bloody diarrhea and death. H. W. McFadden, P. Smith, G. F. Wright, and P. A. Blake. 1986.
Escherichia coli O157:H7 diarrhea in a nursing home: clinical, ep- 29. Wells, J. G., B. R. Davis, I. K. Wachsmuth, L. W. Riley, R. S. Remis, R. Sokow, and G. K. Morris. 1983. Laboratory investigation of hem- 25. Salmon, R. L., I. D. Farrell, J. G. P. Hutchison, D. J. Coleman, R.
orrhagic colitis outbreaks associated with a rare Escherichia coli J. Gross, N. K. Fry, B. Rowe, and S. R. Palmer. 1989. A christening party outbreak of hemorrhagic colitis and hemolytic uremic syn- 30. Wilson, J. B., S. A. McEwen, R. C. Clarke, K. E. Leslie, R. A.
Wilson, and C. L. Gyles. 1991. Distribution and characteristics of drome associated with Escherichia coli O157:H7. verocytotoxigenic Escherichia coli isolated from Ontario dairy cat- 26. Shere, J. A., K. J. Bartlette, and C. W. Kaspar. 1998. Longitudinal 31. Zhao, T., M. P. Doyle, B. G. Harmon, C. A. Brown, P. O. E. Mueller, study of Escherichia coli O157:H7 dissemination on four dairy farms and A. H. Parks. 1998. Reduction of carriage of Escherichia coli O157:H7 in cattle by inoculation with probiotic bacteria. 27. Stewart, P. J., W. Desormeaux, J. Chene, and H. Lior. 1983. Hem- orrhagic colitis in a home for the aged—Ontario. Can. Dis. Wkly. 32. Zhao, T., M. P. Doyle, J. Shere, and L. Garber. 1995. Prevalence of enterohemorrhagic Escherichia coli O157:H7 in a survey of dairy 28. Swerdlow, D. L., B. A. Woodruff, R. C. Brady, P. M. Grif n, S.

Source: http://www.animalnh.co.za/wp-content/uploads/2013/04/Brashears-2003-Isolation-Selection-and-Characterization-of-Lactic-Acid-Bacteria-for-a-Competitive-Exclusion-Product-To-Reduce-Shedding-of-e.-coli.pdf