Running title: E. COLI O157 IN ORGANICALLY- OR NATURALLY-RAISED CATTLE Prevalence of Escherichia coli O157:H7 in Organically- and Naturally-Raised Beef Cattle†

1 We determined the prevalence of Escherichia coli O157:H7 in organicallyand 2 naturally-raised beef cattle at slaughter and compared antibiotic susceptibility profiles of 3 the isolates to those from conventionally-raised beef cattle. Prevalence of E. coli O157:H7 4 were 14.8 and 14.2% for organicallyand naturally-raised cattle, respectively. No major 5 difference in antibiotic susceptibility patterns was observed between the isolates. 6 7 Many cattle producers have adopted production methods, termed niche marketing, to meet 8 consumer demand for safe and healthy beef. The two main niches for beef cattle producers are 9 organic and natural production (4). Organic beef cattle production, regulated by the United 10 States Department of Agriculture, requires feeding of certified organic feed (17) and raising 11 cattle without the use of antibiotics, hormones and other veterinary products (4). Guidelines for 12 producers to label the product as “natural” differ between natural beef programs, and such 13 programs are administered and regulated by the company or organization that owns the brand 14 name rather than the USDA (12). Natural production guidelines often include a complete 15 restriction on the use of antibiotics and growth-promoting hormones, but unlike organic 16 production, feed from non-organic sources is allowed (12). Escherichia coli O157:H7 is a major 17 food-borne pathogen that causes outbreaks of hemorrhagic enteritis, which often leads to 18 hemolytic uremic syndrome in children and the elderly (11). Cattle are major reservoirs of E. 19 coli O157:H7 where it colonizes the hindgut, specifically the rectoanal mucosal region. Cattle 20 feces are the major source of food and water contamination (11). Organic production methods 21 impact on prevalence of food-borne pathogens, including E. coli O157:H7 and Campylobacter in 22 dairy cattle (8, 15) and Campylobacter and Salmonella in chickens (7, 19), have been studied. 23 on N ovem er 2, 2017 by gest ht://aem .sm .rg/ D ow nladed fom 2 However, there is no published study on the prevalence of E. coli O157:H7 in organicallyand 24 naturally-raised beef cattle. Additionally, nothing is known regarding the effects of these 25 production methods on antibiotic susceptibilities of E. coli O157:H7 in beef cattle. Our 26 objectives were to determine the fecal prevalence of E. coli O157:H7 in organically and naturally 27 raised beef cattle at slaughter and compare antibiotic susceptibilities of isolates from organically28 , naturally-, and conventionally-raised beef cattle production systems. 29 Cattle sampled in this study were from three types of production systems, organic, natural, or 30 coventional. Organically-raised beef cattle were from farms that were certified by the National 31 Organic Program (1). The naturally-raised beef cattle were from farms that were certified by the 32 All Natural Source Verified Beef Program (1). Sample collection occurred in an abattoir. 33 Conventionally-raised cattle from two feedlots were sampled in a different abattoir so antibiotic 34 susceptibilities of their isolates could be compared with those from organically and naturally 35 raised cattle. Fecal samples were accessed by spooning contents after cutting open the rectum. 36 The mucosa of the rectum was then rinsed with water until free of visible fecal material and 37 swabbed with a sterile foam-tipped applicator (5). Isolation and identification of E. coli O157 38 and PCR detection of major virulence genes (eae, stx1, stx2, hlyA, and fliC) were as described by 39 Reinstein et al. (14). A subset of 60 isolates (20 [10 fecal and 10 RAMS] from each production 40 system) was randomly chosen to determine the antibiotic susceptibility patterns by broth 41 microdilution method (10). Antibiotics (all from Sigma-Aldrich) tested were: amikacin, 42 amoxicillin, ampicillin, apramycin, bacitracin, cefoxitin, ceftazidime, ceftriaxone, cephalothin, 43 chloramphenicol, chlortetracycline, ciprofloxacin, enrofloxacin, erythromycin, florfenicol, 44 gentamicin, kanamycin, lincomycin, monensin, naladixic acid, neomycin, norfloxacin, 45 novobiocin, oxytetracycline, penicillin, rifampicin, spectinomycin, and trimethoprim. The MIC 46 on N ovem er 2, 2017 by gest ht://aem .sm .rg/ D ow nladed fom 3 was defined as the lowest concentration of an antibiotic that prevented visible growth of the 47 organism. Each concentration of the antibiotic compound was duplicated in the microtiter plate 48 and MIC determination was repeated with a different inoculum preparation. Logistic regression 49 was performed using PROC GENMOD of SAS (SAS Institute, Cary, NC) to compare prevalence 50 (binomial distribution of outcomes) of E. coli O157:H7 detected in fecal samples, RAMS 51 samples, and fecal or RAMS samples (overall animal-level prevalence). The MIC of antibiotics 52 for E. coli O157:H7 isolates were analyzed using a non-parametric survival test in PROC 53 LIFETEST of SAS to determine the effects of the production system (natural, organic, or 54 conventional). Data were right-censored when necessary (resistant at the highest concentration 55 evaluated). The Wilcoxon test was utilized to determine the effect of production system on MIC 56 values. 57 A total of 553, 506, and 322 organically-, naturally-, and conventionally-raised cattle were 58 sampled, respectively. In organically-raised cattle, the prevalence of E. coli O157:H7 in fecal 59 samples across sampling days ranged from 0 to 24.4% with an average of 9.3%, and RAMS 60 prevalence ranged from 0 to 30.9% with an average of 8.7% (Figure 1). In naturally-raised 61 cattle, the prevalence of E. coli O157:H7 in fecal samples ranged from 0 to 20.3% with an 62 average of 7.2%, and RAMS prevalence ranged from 0 to 23.8%, with an average of 8.9% 63 (Figure 1). In both organicallyand naturally-raised cattle, prevalence detected by both sampling 64 methods was greater (P < 0.05) than the prevalence (total) detected by either method (Figure 1). 65 Thirty-six of 322 (11.2%) of conventionally-raised feedlot cattle were culture positive for E. coli 66 O157:H7 (either feces or RAMS). Fecal prevalence of E. coli O157:H7 was 6.5% and the 67 prevalence by the RAMS sampling method of sampling was 7.1%. Most isolates (66.7% from 68 organically-raised beef cattle and 77.8% from naturally-raised beef cattle) were positive for eae, 69 on N ovem er 2, 2017 by gest ht://aem .sm .rg/ D ow nladed fom 4 stx2, hlyA, and fliC but negative for stx1. The stx2 gene was present in 100% and 95% of isolates 70 from organically-raised cattle, respectively. Prevalence of E. coli O157:H7 that we observed in 71 organically-and naturally-raised beef cattle were similar to prevalence reported in 72 conventionally-raised cattle (2). Our study did not include statistical comparison of the 73 prevalence data because of a number of differences in the production systems, particularly diets, 74 between organically-, naturallyand conventionally-raised beef cattle. Organicallyand 75 naturally-raised cattle are either required to graze pasture or fed a forage-based diet. Although 76 conflicting data exist (2), studies have shown that cattle fed a forage diet have both higher levels 77 and longer durations of fecal shedding of E. coli O157:H7 compared with cattle on a grain diet 78 (18). 79 None of the isolates tested from the three production systems was susceptible (MIC > 80 50μg/ml) to bacitracin, lincomycin, monensin, novobiocin, tilmicosin, tylosin, and vancomycin. 81 The MIC values of isolates collected from the production systems were significantly different (P 82 < 0.05) for 12 antibiotics (amikacin, apramycin, cefoxitin, ceftriaxone, gentamycin, kanamycin, 83 nalidixic acid, neomycin, rifampicin, streptomycin, and tetracycline). Escherichia coli O157:H7 84 isolates from conventionally-raised cattle had higher (P < 0.05) MIC values compared to isolates 85 from naturallyand/or organically-raised cattle for gentamicin and neomycin (Table 1). 86 However, isolates from conventionally-fed cattle had lower (P < 0.05) MIC values for amikacin, 87 apramycin, cefoxitin, ceftriaxone, kanamycin, nalidixic acid, penicillin, rifampicin, and 88 tetracycline compared to isolates from naturallyand/or organically-raised cattle (Table 1). 89 Among the 60 isolates tested for antibiotic susceptibilities, six isolates (10%) were susceptible to 90 all antibiotics included in the study, excluding the seven antibiotics that were resistant to all 91 isolates. Forty-two isolates (70%) were resistant (MIC > 50 μg or IU/ml) to one antibiotic, nine 92 on N ovem er 2, 2017 by gest ht://aem .sm .rg/ D ow nladed fom 5 isolates (15%) were resistant to two antibiotics, and two isolates (3%) were resistant to five 93 antibiotics. One isolate from the organically-raised cattle group was resistant to 10 (amoxicillin, 94 ampicillin, cefoxitin, cephalothin, chloramphenicol, florfenicol, oxytetracycline, penicillin, 95 streptomycin, and tetracycline) of the 26 antibiotics that were inhibitory to other isolates. We 96 have presented data as median MIC values for each production system. In some instances, the 97 median values were same, but the actual MIC data for the type of production system differed. 98 This was because the data were right censored if isolates were not susceptible at 50 μg or IU/ml. 99 If more isolates are censored for a particular production system compared to another, it may lead 100 to statistical differences. This justifies the use of survival analysis for this type of data. There 101 were differences between MIC values of isolates from organically-raised cattle and 102 conventionally-raised cattle for many antibiotics (cefoxitin, ceftriaxone, gentamicin, nalidixic 103 acid, neomycin, penicillin, rifampicin, and tetracycline). Similarly, there were differences 104 between MIC values of isolates from naturally-raised cattle and conventionally-raised cattle for 105 many antibiotics (amikacin, apramycin, ceftriaxone, kanamycin, nalidixic acid, and rifampicin. 106 For many of these antibiotics, MIC values for isolates from organically-or naturally-raised cattle 107 were greater than for isolates from conventionally-raised cattle. Resistance genes can be 108 transferred from food animals to humans among the enteric pathogen population (9) and it is 109 possible that resistance genes from other bacteria in the gastrointestinal system of cattle could be 110 acquired by E. coli O157:H7. In cattle, heavy metals like copper or zinc, which are also 111 antimicrobial are included in diets at concentrations in excess of the nutritional requirements, 112 often to replace conventional antibiotics, to achieve growth promotion (6). Feeding of metals 113 also results in the emergence of bacterial populations resistant to metals (6), which in some 114 instances could lead to resistance to antibiotics. Copper resistance mechanisms to concentrations 115 on N ovem er 2, 2017 by gest ht://aem .sm .rg/ D ow nladed fom 6 above those normally tolerated by normal cellular processes have been found on plasmids linked 116 to resistance to antibiotics in some bacteria (6). Therefore, it is possible that isolates from 117 organicallyor naturally-raised cattle that are not exposed to antibiotics still could become 118 resistant to antibiotics. 119 Information on prevalence and antibiotic susceptibilities of foodborne pathogens in organic 120 or natural livestock production systems is limited and variable. In a study of organic and 121 conventional dairy cattle farms, conventional farms were more likely to have at least one 122 Salmonella isolate resistant to antibiotics than were organic farms (13). Kuhnert et. al (8) 123 showed no difference between prevalence of E. coli O157:H7 isolated from organic and 124 conventional dairy farms. Escherichia coli isolates from conventional dairies had significantly 125 higher rates of resistance to certain antibiotics than isolates from organic dairies (16). Cho et al. 126 (3) compared antibiotic susceptibilities of Shiga toxin-producing, O157 and non-O157 isolates 127 from organic and conventional dairy farms and concluded that that there was no overall 128 significant difference in resistance between the two production systems. 129 Although organic and natural beef production systems are becoming popular, little is known 130 about the effects these production systems on foodborne pathogens. Because safety of the food 131 supply is crucial, further investigation into these production systems and their potential for 132 altering risk of human illness is warranted. Our study found that organically and naturally raised 133 beef cattle have a similar fecal prevalence of E. coli O157:H7 and our prevalence estimates for 134 these types of production systems are similar to previous reports for conventionally raised 135 feedlot cattle. 136