Abstract:
Diarrheagenic foodborne bacterial pathogens pose a significant global public health challenge, particularly in
developing countries, such as Ethiopia. Among these foodborne pathogens (FBPs), nontyphoidal Salmonella
(NTS), Escherichia coli, Shigella, and Campylobacter are major contributors to high childhood diarrhea rates.
However, there is a lack of comprehensive understanding regarding the prevalence, genomic characteristics,
and potential sources of foodborne bacterial pathogens contributing to diarrheal illnesses among children
under five in Ethiopia. This knowledge gap hinders effective preventive measures and targeted interventions
aimed at reducing the burden of diarrheal diseases associated with foodborne pathogens in vulnerable
populations. Therefore, it is important to investigate the occurrence of FBPs and their genomic antimicrobial
resistance (AMR) profiles and explore the relationships between strains isolated from children with diarrhea
and their potential exposure sources in Harar town and Kersa district, Eastern Ethiopia. A health facility-based
cross-sectional study was conducted from November 2021 to January 2023 to isolate and identify NTS,
diarrheagenic E. coli (DEC), Shigella, and Campylobacter in stool and environmental samples. Stool samples
were collected from children under five years (UFC) with diarrhea visiting healthcare facilities. Additionally,
exposure source samples, which included caregivers and/or siblings, animals, and various food and
environmental samples, were collected through case-based tracing. The bacterial pathogens were isolated and
identified using plate culture and biochemical testing procedures. Genomic DNA was extracted, libraries for
next-generation sequencing were prepared, and the samples were sequenced on a NextSeq 550 Illumina
sequencer. Data analysis was conducted using statistical software and various bioinformatics tools, including
SpeciesFinder for species determination, ResFinder and PointFinder for AMR profiling, multi- locus sequence
typing (MLST) for sequence type (ST) profiling, and CSIPhylogeny for phylogenetic analysis. The overall
prevalence of the four diarrheagenic pathogens—NTS, DEC, Shigella, and Campylobacter—was 24.8%, with
occurrence rates of 27.9% in children, 24.1% in caretakers, and 23.1% in environmental samples. The study
revealed an 8.5% coinfection rate and a 23.7% single pathogen infection rate among children with diarrhea. In
caretakers, the bacterial coinfection rate was 6%. The prevalence in children with diarrhea was 6.1% for NTS,
10.3% for DEC, 7.3% for Shigella, and 8.4% for Campylobacter, with corresponding rates of 6%, 8.4%, 9.6%,
xiii
and 6% in caretakers, respectively. Notably, 4.5% and 3.9% of diarrhea cases in children from urban and rural
settings, respectively, were attributed to bacterial coinfection. Among the 438 analyzed exposure source
samples, the overall prevalence of the three aerobic pathogens—NTS, DEC, and Shigella—was 18.3%, with
3.9% co-occurrence and 14.4% single pathogen occurrence rates. The overall prevalence of Campylobacter in
exposure sources was 5.5%, with 6% in Harar town and 5% in Kersa district. The highest prevalence of FBPs
was observed in wastewater, at 40.9% for the three aerobic FBPs (40.9%; AOR = 3.3; 95% CI: 1.1–10.1), and
9.1% for Campylobacter. The detection of Campylobacter was 1.8 times more likely in household samples (8.7%;
OR = 1.8; 95% CI: 0.7–4.5) than in samples from marketplaces, whereas the occurrence of the aerobic
pathogens in samples collected from households was 6.5% (AOR = 0.4; 95% CI: 0.2–1.0). The food
contamination rate was 17.9% (AOR = 1.2; 95% CI: 0.4–3.6) for the aerobic FBPs and 4.2% (OR = 0.4; 95%
CI: 0.1–1.4) for Campylobacter, with no significant difference between the animal- sourced and nonanimal-
sourced food categories. The occurrence rates of NTS, DEC, Shigella, and Campylobacter in the meat samples
were 13.9%, 5.4%, 6.5%, and 3.2%, respectively. Notably, DEC and Shigella were detected in 3.7 (1/27) and
7.4% (2/27) of the cooked food samples, respectively. The water contamination rate of Shigella was 7.5%
(3/40), and it was also the only FBP detected in drinking water in this study. A significantly higher prevalence
of FBPs was detected in poultry than in other animal species, with the likelihood of Campylobacter
contamination in the environment being 5.8 times higher in the presence of poultry (17.7%; OR = 5.8; CI: 1.1–
30.6) compared to shoats. The monthly occurrence patterns of the isolated FBPs revealed the highest frequency
during the winter (dry) season in both children with diarrhea and their exposure sources, peaking at 73.9% in
February 2022 in human stool samples. MLST analysis identified 13, 19, and 8 distinct STs for NTS, DEC, and
Campylobacter, respectively. The most common STs were ST93 for NTS, ST13857 for DEC, and ST353 for
Campylobacter. The Species determination analyses revealed low Campylobacter species diversity, with only
C. jejuni (62.2%) and C. coli (37.8%) detected. The SeqSero tool identified 17 different Salmonella serotypes,
with. Muenchen being the most frequently detected, accounting for 14% of the isolates. Among the sequenced
NTS, DEC, and Campylobacter, 100% (14/14), 60% (15/25), and 84.6% (11/13) of the child case isolates,
respectively, shared ST profiles with at least one isolate from an exposure source. Maximum likelihood
phylogenetic analysis revealed that most strains isolated from children clustered with isolates from caregivers
or environmental exposures (SNP < 30). Close genomic relatedness among case isolates and those from diverse
sources was also observed in distinct clades (SNP < 10 for Campylobacter; SNP < 5 for NTS and DEC).
PointFinder analysis revealed at least one antimicrobial resistance gene (ARG) in every NTS sample and 78%
of the DEC isolates. The most prevalent ARGs included aac(6')-Iaa and aph(3’)-Ib, which predict
xiv
aminoglycoside resistance. Other commonly identified ARGs included sul2, aph(6)-Ib, blaTEM-1B, and tet(A).
Notably, most of the identified β-lactam ARGs predicted that some NTS and DEC strains were resistant to more
than three antibiotics, with blaCTX-M-15 found in 18% of DEC and 3.5% of NTS isolates, conferring resistance
to nine antibiotics. DEC isolates from children revealed the highest ARG diversity, with genes such as aph(3'')-
Ib, aph(6)-Id, sul2, and tet(A) detected across all sample types, although some ARGs were exclusive to specific
sample types. Point mutations mediating AMR were detected in several genes, with mutations associated with
nucleotide substitution being the most frequent. In conclusion, this study revealed a high prevalence of FBPs,
with all four pathogens being detected in various exposure sources of children with diarrhea. There were also no
significant differences (P > 0.05) in the occurrence of each pathogen between settings or between animal-
sourced and nonanimal-sourced food, underscoring the importance of these pathogens in both Harar town and
Kersa district. This study revealed diverse Salmonella serotypes and DEC strains but low Campylobacter
species diversity. The majority of NTS, DEC, and Campylobacter case isolates presented similar ST profiles to
those from potential exposure sources and clustered together, even including some isolates collected at different
times within certain clades. This suggests the potential persistence and widespread distribution of specific
pathogen strains from common sources. In some phylogenetic tree clusters, isolates from various sources
clustered together and shared ST profiles, demonstrating the involvement of multiple vectors in the transmission
of each pathogen. Furthermore, genotypic AMR profiling revealed widespread drug resistant NTS and DEC
strains in Eastern Ethiopia. The study results underscore the need for enhanced food safety measures and
improved environmental and personal hygiene practices among caregivers in both urban and rural settings to
reduce the burden of FBP-associated diarrhea in children in the study area. As the first whole genome
sequencing-based source-tracking study in Ethiopia, our findings highlight the importance of employing WGS
in tracking the sources of FBPs and foodborne disease surveillance. Furthermore, one-health approach studies
employing WGS and other genome-based diagnostic tools, such as metagenomic analysis and molecular clock
analysis, are recommended to better determine the transmission dynamics of FBPs and the relative contributions
of each source. Robust and sustained AMR surveillance is also recommended for monitoring the emergence
and spread of resistant pathogens.