Non-Typhoidal Salmonella and Lactose Fermenting Enterobacteriaceae in Human-Dairy Interface and Contamination Reduction Pilot Trials in Northwest and Central Ethiopia

Abstract

Background: Pathogens from food, animals and the environment are causing substantial social and economic impacts in the world. The dairy farms, the food from them (like milk), animals and their environment can act as sources for pathogens like non-typhoidal Salmonella (NTS) and lactose fermenting Enterobacteriaceae (LFE). The occurrence of these pathogens might be high in Ethiopia due to the risky production, handling, and consumption behaviours. Isolation and characterization of NTS and LFE in the human-dairy interface and assessing contamination reduction techniques are essential to design methods for prevention and control of outbreaks of foodborne, zoonotic, and other diseases occurring in the interface in particular and in the community in general. Despite some studies conducted to determine the presence of these bacteria; the data were scarce in the human-dairy interface in the Northwest part of Ethiopia and limited intervention studies were conducted in the country. Objectives: This study was conducted to isolate and characterize NTS and LFE, determine the effect of training on knowledge, attitude, and practice (KAP) of dairy farmers and bacterial contamination of milk. Methods: The study was conducted from January 2021 to June 2024. For isolation and characterization of NTS and LFE, a total of 362 samples consisting of pooled raw milk (58), milk container swabs (58), milker’s hand swabs (58), farm sewage (57), milker’s stool (47), and cow’s faeces (84) were collected and analyzed using standard bacteriological methods. The presumed NTS-positive colonies were confirmed by Matrix-assisted laser desorption ionisation time of flight (MALDI-TOF). Slide agglutination test according to the White-Kauffmann-Le Minor scheme was employed to identify the serovars of the NTS isolates. The genomes of NTS isolates were sequenced for further characterization. The information regarding multi-locus sequence typing (MLST), genotypic relationship, pathogenicity, pathogenicity islands, virulence genes, antimicrobial resistance genes, antimicrobial class and mechanisms of resistance were obtained by blasting the sequence of the isolates using online software and databases. The Shigatoxin-producing E.

coli (STEC) was confirmed by multiplex PCR targeting stx1 and eae genes. The antimicrobial susceptibility patterns and extended-spectrum beta-lactamase (ESBL) production ability of the LFE isolates were screened using the Kirby-Bauer disk diffusion method, and candidate isolates passing the screening criteria were phenotypically confirmed by XII |P a g e

, stx2 using cefotaxime (30 μg) and cefotaxime-clavulanic acid (30 μg/10 μg) combined-disk diffusion test. The isolates were further characterized genotypically using multiplex PCR targeting the three ESBL-encoding genes namely blaTEM XIII |P a g e

, blaSHV , and blaCTX-M. The PCR-amplified products of 26 blaCTX-M genes were sequenced for further characterization. For the training intervention study, the KAP of 120 smallholder women dairy farms and milk samples were collected from 120 dairy farmers one week before and from 115 dairy farms four to six weeks after the training in four sites in Central Ethiopia and analyzed for total coliforms, thermo-tolerant coliforms, E. coli, STEC and NTS. The LAB were isolated from 61 fermented milk samples and isolates were screened based on their inhibitory effect on E. coli. Results: Of the processed samples for NTS and LFE (n=362), 28 (7.7%) (95% CI 5.4% - 11.0%) NTS isolates were detected which belong to six serovars. Among the serovars S. Uganda (39.3%) was the dominant, followed by S. enterica subsp. diarizonae (25.0%) and S. Typhimurium (21.4%). The AMR profile showed that 89.3% of the NTS isolates were resistant to at least one antimicrobial agent and 46.4% were multidrug-resistant. Among antimicrobials, NTS isolates were highly resistant to ampicillin (57.1%), followed by tetracycline (42.9%) and chloramphenicol (35.7%). On the other hand, the NTS isolates were 100%, 96.4%, and 96.4%, susceptible to ceftriaxone, norfloxacin and azithromycin, respectively. The whole genome sequencing of 28 NTS isolates indicated that the same serovars showed similar or closer MLST, phylogenetic relatedness, pathogenicity, pathogenicity islands, and drug-resistant and virulence genes than different serovars. Among samples analyzed (n=362), 91.4% were positive for LFE, whereas 71.3%, 15.2%, 10.2% and 6.9% were positive for E. coli, Citrobacter spp, Enterobacter spp, and Klebsiella spp, respectively. E. coli was the dominant isolate in all samples. The LFE isolates were resistant to most of the commonly used antimicrobials which ranged from 0.0 to 88.9%. Among 375 LFE isolates, 70.7% and 21.3% were MDR and ESBL producers, respectively. The MDR index varied from 0.0 to 0.81 with an average of 0.30. The ESBL production was detected in all types of samples. Genotypically, the majority of the isolates (97.5%), which were positive on the phenotypic test, were carrying one or more of the three genes. The blaCTX-M gene was dominant, and the sequenced result showed that the blaCTX-M-15

subtype was the most frequently (61.5%) detected. Among E. coli isolates (n=258), 15 (5.8%) were positive for STEC. The probable transmission among sample types in the interface was observed. In the training intervention study, the KAP of farmers was substantially increased after the training. In the 33.9% and 47.8% of dairy farms, the total and thermotolerant coliform counts after the training showed a reduction, respectively. However, the milk samples contained a high proportion of both indicatory and pathogenic bacteria even after the training. Conclusions: NTS and LFE were detected in humans, dairy cows, dairy utensils, and the environment, showing the potential of the human-dairy farm-environment nexus in the circulation of these pathogens. A high proportion of the LFE isolates were MDR; showed high MDR-I and were positive for ESBL production. The findings provided evidence that the humandairy interface is one of the important reservoirs of AMR traits. Therefore, these further highlight

that the interface is a good point of intervention in the prevention and control of NTS and LFE.

The implementation of AMR mitigation strategies is highly needed in the interface. An increased

KAP

level and reduction in the detection of some milk safety bacterial hazards were observed after the training. However, considerable contamination proportions were also observed after the training. To reduce bacterial hazards from raw milk, it is important to support the training with the implementation of additional risk mitigation strategies