Antimicrobial resistant and enteropathogenic bacteria in ‘filth flies’: a cross-sectional study from Nigeria

‘Filth flies’ facilitate the dispersal of pathogens between animals and humans. The objective was to study the intestinal colonization with antimicrobial resistant and enteropathogenic bacteria in ‘filth flies’ from Nigeria. Flies from Southern Nigeria were screened for extended-spectrum β-lactamase producing Enterobacterales (ESBL-E), Staphylococcus aureus, Salmonella sp., Shigella sp., Campylobacter sp. and Yersinia enterocolitica by culture. ESBL-E were tested for blaSHV, blaCTX-M and blaTEM; S. aureus was screened for enterotoxins. Spa typing and multilocus sequence typing (MLST) was done for S. aureus and MLST for Escherichia coli. Of 2,000 flies, 400 were randomly collected for species identification. The most common species were Musca domestica (44.8%, 179/400), Chrysomya putoria (21.6%, 85/400) and Musca sorbens (18.8%, 75/400). Flies were colonized with S. aureus (13.8%, 275/2,000) and ESBL-E (0.8%, 16/2,000). No other enteropathogenic bacteria were detected. The enterotoxin sei was most common (26%, 70/275) in S. aureus, followed by sea (12%, n = 32/275). Four S. aureus isolates were methicillin resistant (mecA positive, t674 and t5305, ST15). The blaCTX-M (n = 16) was the most prevalent ESBL subtype, followed by blaTEM (n = 8). ‘Filth flies’ can carry antimicrobial resistant bacteria in Nigeria. Enterotoxin-positive S. aureus might be the main reason for food poisoning by ‘filth flies’ in the study area.

Diptera (true flies) are the most abundant and diversified endopterygota (or holometabola) of the insect order, with more than 180 families of about 158,000 described species 1,2 . 'Filth flies' (diptera) are universal, ubiquitous, coprophagic and synanthropic (living in close association with humans) insects that breed in garbage, animal and human faeces 3 . Known to serve as vectors of many pathogens, house flies can disperse pathogens through a flight distance of about 7 km between animals and humans 4 .
Flies transmit pathogens through three routes: mechanical translocation from the exoskeleton, regurgitation and defecation 3 . During feeding, flies can either pick up pathogens on its exoskeletal surfaces or ingest fluids contaminated with pathogens. Ingested pathogens can multiply in the crop (a blind sac of the digestive tract in higher flies) and after regurgitation which coined the term of "bioenhanced transmission" 5 .
Antimicrobial resistance (AMR) affects both humans and animals and antimicrobial resistant bacteria can be transmitted between animals and humans in both direction. This challenge is considered in the "one Health" approach. The most widespread resistance mechanisms in Enterobacterales is based on plasmid-mediated production of extended spectrum β-lactamases (ESBL) which hydrolyse β-lactam rings, thereby reducing the efficacy of cephalosporins and monobactams 6 . Flies are important reservoirs and vectors of antimicrobial resistant bacteria (such as methicillin resistant Staphylococcus aureus, ESBL-producing Enterobacterales [ESBL-E]) 3,7 . One short report suggests that antimicrobial resistant bacteria can also be detected in flies (n = 25) in sub-Saharan Africa 8 . However, the true burden of AMR in 'filth flies' in Africa is unknown and it is currently unclear which fly species are the main vectors of ESBL-E and enteropathogenic bacteria. Therefore, the objective of this study was to analyse the colonization rates of 'filth flies' from Southern Nigeria with ESBL-E, S. aureus and other enteropathogenic bacteria and to identify those fly species which are mainly colonized with these target organisms.  Figure S1).

Discussion
A total of 2,000 flies from Southern Nigeria were screened for the intestinal colonization with antimicrobial resistant and enteropathogenic bacteria. Our main findings are a high proportion of S. aureus (13.8%, 275/2,000) and low occurrence of ESBL-E (0.8%, 16/2,000).
The analysis of the intestinal culturome was done to assess the occurrence of bacterial species independent of the resistance phenotype. In general, Enterobacterales were only rarely reported (e.g. K. pneumoniae, n = 1/82, www.nature.com/scientificreports/ Figure S1) which is in line with an overall low colonization rate with ESBL-E in flies (0.8%). The vast majority of isolates belonged to the genera Bacillus and Clostridioides suggesting a selection of spore-forming bacteria by the treatment with ethanol. However, we confirmed the observation that ethanol sanitation of the exoskeleton does not alter the intestinal colonization 10 . Although the culturome was assessed under aerobic conditions, some Clostridioides species were detected. This applies for those species, that are known to be aerotolerant (i.e. C. histolyticum, C. terium, Fig. S1) 11 . The S. aureus intestinal colonization rate in our study (13.8%) is in stark contrast to a 0.4% colonization rate in flies in a comparable approach (i.e. same trap and bait) from Germany 12 . The differing proportions are most likely due to the different settings (tropical vs. temperate regions). All S. aureus from our study shared a very similar genetic background based on the spa repeat pattern, ST and cgMLST (Fig. 2). This is suggestive either for a common source for all isolates from flies or a cross-contamination between flies during sampling or an adaptation (due to e.g. fitness factors) of this ST15 lineage to the intestinal tract of the flies. An artificial contamination of flies would challenge the scientific significance of our work. However, the widespread detection of S. aureus from various sampling site (Fig. 1), both the absence and presence of mecA in isolates belonging to t674, different antimicrobial resistance rates, the high diversity of ESBL-E sampled simultaneously and the known effective sanitation of the exoskeleton by ethanol argues against a cross-contamination during sampling 13 . All isolates belonged to ST15, which is in line with a predominance of the clonal complex CC15 in isolates from community-acquired infection in sub-Saharan Africa 14 . In addition, CC15 is known to be well adapted to the human host 15 . However, the low antimicrobial resistance rates, the absence of lukF-PV/lukS-PV and sak might also suggest that S. aureus from flies are rather of animal (e.g. livestock or wildlife) than of human origin [16][17][18] . However, other factors not investigated in our study (e.g. adaptation of S. aureus to the gut of 'filth flies' , fitness factors) could explain the predominance of ST15/t674-S. aureus in flies.
Among all S. aureus from 'filth flies' , we only detected sea and sei. The superantigenic activity of both is superior to other staphylococcal enterotoxins (e.g. members of the SEB group) due to an additional high-affinity MHC II binding site 19 .
Although our study has strengths (e.g. large sample size, broad bacterial culture), some limitations need to be addressed: first, culturome results revealed the abundance of Bacillus and Clostridioides suggesting a selection of spore-forming genera by ethanol. This might point towards a methodological bias of our work. However, others have also reported a higher prevalence of Bacillus cereus than Enterobacteriaceae 23 . Second, despite poor sanitation systems and access of flies to human and animal faeces, we did not detect any other enteropathogens (e.g. www.nature.com/scientificreports/ Salmonella sp., Shigella sp.). Since we were unable to culture the fly samples immediately, some isolates might have not stayed viable during storage and transport.
In conclusion, diptera-borne S. aureus food poisoning might be or become a health issue in the study region due to the high prevalence of enterotoxins (sea, sei) in S. aureus from 'filth flies' . In contrast, a transmission of ESBL-E through flies by defecation and regurgitation does not seem to play a major role.

Materials and methods
Ethical statement. An ethical approval is not required for the analysis of invertebrates. All methods were carried out in accordance with relevant guidelines and regulations. The local Ph.D. committee, Medical Faculty, Westfälische Wilhelms-Universität Münster, approved all experimental protocols. Study area/mapping. 'Filth flies' were collected in Southern Nigeria between June and July 2017. Sampling sites were classified into "urban", "semi-urban" and "rural" according to the European Union Methodological manual on territorial typologies 24 . GPS coordinates were taken for each sampling site (eTrex 10, Garmin, Olathe, Kansas, USA). For every sampling spot, key environmental conditions were documented (i.e. livestock/human faeces within 10 m radius, presence of refuse dump, presence of decomposing organic matters and setting [urban, semi-urban, rural]). Atmospheric conditions (i.e. humidity, temperature, air pressure, wind force, sunshine hours) were recorded as reported by the Nigerian Meteorological Agency (NiMet, https ://nimet .gov.ng/).
The maps were downloaded as jpg-files from openstreetmap (https ://www.opens treet map.org) with bounding boxes (latitude, longitude) related to the region of interest and used as a background for the plots. All plots were produced with "R" (Version 3.6.2) using the package "ggplot2" (grammar of graphics, version 3.2.0) 25 . The data used for the plots were taken from the original files and, where necessary, transformed by methods from the package "dplyr" (grammar of data manipulation, version 0.8.3).
Diptera. Flies (n = 2,000) were collected using the Gaze trap method 12 . Insect bait, conventionally made from animal proteins, carbohydrates and sugar (Feldner, Waldsee, Germany) in a container covered with gaze and placed under the gaze trap was used to lure the flies. Trapped flies (approximately 20/sampling site) were collected and killed in 70% ethanol and dried in silica gel (2-4 mm, Carl Roth, Karlsruhe, Germany). Ethanol sanitizes the outer surfaces of the flies, avoiding cross-contamination during sampling without altering the intestinal microbiome of the flies 10 .
Each fly was sent for further analysis to Germany in 1 ml sterile sodium chloride (0.45%) at -18 °C (Peli Biothermal Credo Cube, UK).
Culturome of random flies. We analysed the culturome of 82 randomly selected flies in order to describe the overall colonization pattern of flies with bacteria independent of the AMR phenotype. Since our target bacterial species (e.g. Escherichia coli, Klebsiella sp.) are aerobic bacteria, we did not use specific anaerobic conditions for culture. After removal of the legs and wings for molecular species identification, the remaining body (head, thorax and abdomen) was mechanically homogenized and cultured in 1 ml BHI broth overnight (37 °C, ambient air). A total of 10 µl of overnight culture was sub-cultured on MacConkey Agar (Oxoid GmbH, Wesel, Deutschland), Columbia Blood Agar (Oxoid), Trypticase Soy Agar (BD, Heidelberg, Germany), Kimmig Agar (Oxoid), Chocolate Agar (BD) and Colistin-Aztreonam (CAP) Agar (Oxoid). All phenotypically different colonies were selected for species identification using Matrix-Assisted Laser Desorption/Ionization-Time-Of-Flight (MALDI-TOF, microflex LT, Bruker Daltonics, Bremen, Germany).

Identification and characterization.
Species of suspected S. aureus was identified using MALDI-TOF and confirmed by the detection of a species specific polymorphism of the non-ribosomal peptide synthetase (NRPS) 26 and the S. aureus specific thermostable nuclease (nuc) 27 . All S. aureus were screened for the immune evasion cluster (IEC) 28 , the Panton-Valentine leucocidin gene (lukS-PV/lukF-PV) 29 and the enterotoxin genes sea, seb, sec, sed, see, sef, seg, seh and sei 30,31 .
Enterobacterales were identified with VITEK2 automated systems (bioMérieux) due to ambiguous delineation of E. coli and Shigella sp. using MALDI-TOF.
Antimicrobial resistance. The antimicrobial susceptibility testing was done with VITEK2 automated systems (bioMérieux) using EUCAST clinical breakpoints (Version 9.0). ESBL-E were confirmed using the double disc diffusion test (Mast diagnostics, Bootle, UK) and were screened for the presence of bla SHV , bla CTX-M , bla TEM and bla CMY-2 32,33 . Subtypes of the detected beta-lactamase genes were determined by Sanger sequencing. All S. aureus isolates were screened for the staphylococcal beta-lactamase blaZ 34 .

Scientific Reports
| (2020) 10:16990 | https://doi.org/10.1038/s41598-020-74112-x www.nature.com/scientificreports/ Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creat iveco mmons .org/licen ses/by/4.0/.