Resistance profile of Salmonella isolated from food sold in the streets of N'Djamena, Chad

The preparation and sale of street foods is booming in African cities in general and especially in N’Djamena. However, the hygienic failures observed during this activity constitute a source of contamination of these foods. The objective of this work is to determine the different serotypes of Salmonella isolated from food sold in the streets of N’Djamena in Chad and to test their sensitivity antibiotics. In total, 447 samples from 12 different types of food were collected and analyzed using standard food microbiology methods. The disk diffusion method was used to test the antibiotic susceptibility of Salmonella strains detected. The serotyping of the 5 strains of Salmonella allowed us to identify 3 serotypes namely Salmonella Mbandaka, Salmonella Idikan and Salmonella Anatum. The susceptibility profiles of the strains to antibiotics were varied. Resistance were observed with the antibiotics Amoxicillin + Clavulanic acid, Cefotaxime and Nalidixic acid. The most active antibiotics were Cefoxitne, Ciprofloxacin, Aztreonam, Imipenem and Choramphenicol with a rate of 100% sensitivity. However, the resistance of these strains to certain antibiotics is a real public health problem that calls out to food safety.


INTRODUCTION
In Africa, street food vending and consumption have proliferated in the last three and a half decades (FAO, 2016). Street food can be defined as any ready to eat food or beverage sold and sometimes prepared in outdoor public spaces by vendors or cooked stationary outlet with or without indoor space to accommodate *Corresponding author. E-mail: djibrine_ateba@yahoo.fr.
Author(s) agree that this article remain permanently open access under the terms of the Creative Commons Attribution License 4.0 International License consumers. This kind of foods concerns all genders, all age groups and all socio-professional categories (FAO, 2016). Street food is a regular source of income for millions of unskilled men and women in developing countries (FAO, 2010).
Antibiotics have since their discovery played a very important role in reducing the cases of disease and death related to microorganisms. These antibiotics have revolutionized the treatment of infectious diseases. However, the excessive use of these antimicrobials would be the reason for these multidrug resistance observed in certain strains (Mwambete and Peter, 2011;Makut et al., 2013;Rashed et al., 2013).
In a study conducted on foods in Morocco by Bouchrif et al. (2009), out of 104 isolates of Salmonella enterica including the serovars Infantis, Bredeney, Blokley, Typhimurium, Mbadaka, Branderup II and Kiambu, approximately 29% of the isolates were resistant to at least one antimicrobial. Resistance to tetracycline was most common (21%), followed by resistance to ampicillin (13%), amoxicillin + clavulanic acid (9%), streptomycin (7%), chloramphenicol (4%) and nalidixic acid (3.8%). Isolates of S. Enteritidis and S. Typhimurium obtained from chicken meat samples in Malaysia showed resistance to several antibiotics including erythromycin, penicillin and vancomycin (Thung et al., 2016). Antibiotic resistance in Salmonella has been associated with higher frequency and duration of hospital stays, longer disease, higher risk of invasive infection, and double increased risk of death within two years following infection (WHO, 2011).
The main aim of the present study is to determine the different serotypes of Salmonella isolated from food sold in the streets of N'Djaména in Chad and to test their sensitivity to antibiotics commonly used.

Site and period of the study
The study took place in N'Djamena, the political capital of Chad between October 2014 and January 2018. Of the 10 districts in N'Djaména, eight were selected for this study.

Sampling and sample processing
Approximately 447 samples were collected from the sites chosen for the conduct of the study (Figure 1). Random sampling was carried out and consisted of collecting samples of minced beef sandwich, the "ball" with okra sauce, rice with sorrel sauce, rice with tomato sauce, meat grilled mutton, fried fish, banana juice, avocado juice, raw mutton, raw beef, raw fish and grilled mutton seasoning. Quantities of 500 g (for solid products) were collected in sterile sachets and 500 ml (for liquid products) were collected in sterile bottles, placed in a cooler containing and sent to the laboratory, then analyzed within hours followed.

Isolation and identification of Salmonella
For the detection of Salmonella a pre-enrichment in buffered peptone water for 24 h at 37°C was carried out, then two broths, Rappaport-Vassiliadis (41.5°C) and Muller-Kauffmann with Tetra Thionate-novobiocin (37°C) were used for enrichment. For isolation, Xylose Lysine Desoxycholate (XLD) and Hektoen agar media were used. The colonies presenting the typical appearance of Salmonella were subjected to biochemical tests namely Kligler-Hajna agar which consisted of looking for a slope (red), a pellet (yellow), with formation of gas and hydrogen sulphide (blackened agar). The second test was to look for the production or not of urease and indole. The presumed Salmonella colonies were confirmed by the API 20E gallery (BioMérieux, France).

Serotyping of Salmonella
The serotyping of the Salmonella strains was carried out by the technique of direct agglutination on slide with the association of antigens (Ag) O (of the wall) and H (of the flagellum) according to the Kauffmann-White diagram (Kauffmann, 1966). The autoagglutinating property or not of the Salmonella strains was detected using physiological water. Serotypes were determined by combining the "O" and "H" antigens obtained by agglutination test on a pure, non-self-agglutinating culture, isolated for 24 h on nonselective agar. For the strains on which auto-agglutination was not detected, we tested them successively with the O-polyvalent (OMA, OMB and OMC), O-monovalent and H (HMA, HMB, HMC and H1) according to the Kauffmann-White diagram (Kauffmann, 1966 The OMB + strains were tested with a monovalent O agglutinating serum O: 6, 7, 8. All strains agglutinating with antisera from groups C and D were tested with Vi antiserum. From the Svend Gard (SG) soft agar, we used H: a flagellar antisera, for the strains agglutinated with the antisera of groups A, the antisera H: b, H: i and H: g, m for those of group B and H: d for those of group D / Vi positive, then H: g, m and H: p for the other strains of group D / Vinegative. The antigenic formula and the reading of the serotyping results were carried out using the Kauffman-White minor table (Grimont and Weill, 2007;Guibourdenche et al., 2010).

Preparation of the inoculum
For performing the antibiogram, the agar diffusion method was used (Bauer et al., 1966). Thus, the strains of Salmonella sp isolated previously from food samples were subcultured on Mueller Hinton agar (Liofilchem, Italy) then incubated at 37°C. After incubation for 18-24 h, a few bacterial colonies were suspended in physiological water. The resulting suspension was calibrated to the MacFarland 0.5 standard (~ 108 CFU / ml) using a densimat.

Inoculation of Salmonella sp.
Agar medium (Mueller Hinton) was as described by Kirby-Bauer. Flood seeding was carried out using 3 to 5 ml of inoculum. The excess bacterial suspension was removed with a micropipette. The Petri dish was then kept for 15 min at room temperature for drying under a microbiological safety post (PSM). Antibiotic discs were then placed at an equal distance on the inoculated agar medium using sterile forceps. A light pressure was exerted on the discs in order to consolidate their adhesion to the agar medium. The Petri dishes containing the antibiotic discs were finally kept at room temperature before being incubated for 24 h at 37°C.

Reading
After incubation, sterile circular culture areas were observed around some of the antibiotic discs. The diameter of these zones or diameter of inhibition is proportional to the antibacterial activity of the antibiotic. Thus, these inhibition diameters were measured using a caliper. The results were interpreted according to the recommendations of AC-FSM/EUCAST (2016).

RESULTS
The different sampling sites are made up of markets, primary schools, avenues, the administrative area, high schools, stations and the industrial area (Figure 1). The different foods analyzed are based on cereals, meat and fish (Table 1). The different strains isolated showed varying levels of sensitivity and resistance to the antibiotics tested (Table 3).

Prevalence and serotypes of Salmonella in street foods in N'Djaména
Out of the 447 samples from 12 different types of food analyzed, we noted the presence of Salmonella in 5

N°
Type of food Description of the food Cereal-based foods 01 Minced beef sandwich Bread loaded with minced meat cooked in a sauce with tomato, potato, onions, mayonnaise, aroma of oil and salt 02 "Boule" with okra sauce Rice flour dough prepared and eaten with okra meat sauce (beef or mutton) 03 Rice with sorrel sauce White rice boiled and served with sorrel sauce with meat (beef or mutton) 04 Rice with tomato sauce White rice boiled and served with a tomato sauce made with meat (beef or mutton) or fish  (Table 2). It should be remembered that the 5 contaminated samples were those of rice with sorrel sauce. Serotyping made it possible to identify 3 Salmonella serotypes as shown in Table  2.

Antibiotic sensitivity of Salmonella sp.
Table 3 the antibiotic sensitivity profile of the Salmonella strains isolated in this study. The different strains isolated showed varying levels of sensitivity and resistance to the antibiotics tested.
On the one hand, the isolates were all susceptible to Ciprofloxacin, Cefoxitin, Chloramphenicol, Aztreonam and Imipenem. On the other hand, the phenotypic resistance profile of the Salmonella reveals that the only strain of S. Idikan showed resistance to both Amoxicillin + Clavulanic Acid and Nalidixic Acid. Finally, one in two strains of S. Anatum exhibited resistance to Cefotaximé.

DISCUSSION
Analysis of the different foods showed that the majority were not contaminated with Salmonella, this result is consistent with those of many studies conducted on street foods (El Marnissi et al., 2012;Diane et al., 2017;Boko et al., 2017;Dossou et al., 2018;Mayoré et al., 2018;Doutoum et al., 2019). In rice with sorrel sauce, there is a strong presence of Salmonella as in several other studies (Bereda et al., 2016;Justin et al., 2018). The presence of Salmonella in the rice sorrel sauce could come from the sorrel leaves which are used for the preparation of the sauce if the cooking is not done well or from cross contamination following poorly washed utensils  used for the service (Mayoré, 2019). In food, the presence of Salmonella is worrying for the health of the consumer. According to the World Health Organization, Salmonella (non-tyhoid) is responsible for tens of millions of cases of water-borne and food-borne illnesses in humans each year around the world (WHO, 2013). The serotype Anatum identified in rice with sorrel sauce were also found in Carcass sponge, Bone meal, Raw chicken and Process chicken (Shafini et al., 2017;Da Cunha-Neto et al., 2017). S. Anatum is frequently detected in cattle and in faeces, skin, lymph nodes, meat fluids, and carcasses of dairy cattle and beef in the southern United States (Kunze et al., 2008), such as in the case of beef from Mexico (Varela-Guerrero et al., 2013), Namibia (Shilangale et al., 2015) and South Africa (Madoroba et al., 2016). In France for example, between 2007 and 2012, the data collected on the food samples analyzed underline an almost constant evolution in the number of strains of the S. Mbandaka serovar (Renaud et al., 2015).
Foodborne illness caused by non-typhoid salmonellosis is a major public health problem around the world. However, examination of the resistance profile to different classes of antibiotics performed on isolates of the different Salmonella serotypes isolated from food samples shows some resistance to one class of antibiotics. Since these pathogens could be transmitted through the environment or contaminated water, their presence could have an impact on public health. Our results show resistance of Salmonella to Amoxicillin, Cefotaxime and Nalidixic Acid. An earlier study carried out in Chad revealed resistance to at least one of these antibiotics (Djim-Adjim et al., 2013) and also studies carried out in several other countries have made similar findings (Thong and Modarressi, 2011;Bagré et al. 2014;Bawa et al. 2015). Quinolones are generally the drugs of choice for the treatment of invasive Salmonella infections in adults, and resistance to nalidixic acid may invalidate treatment with quinolone. The total sensitivity of our isolates to Ciprofloxacin, Cefoxitin, Chloramphenicol, Aztreonam and Imipenem has also been variously reported by Bagré et al. (2014); Bawa et al. (2015) and Bsadjo (2015). The resistance of Salmonella strains to different antibiotics could be explained by the fact that these antibiotics are often used inappropriately in unregulated doses, thus becoming a factor favoring the adaptation of strains to antibiotics (Iroha et al., 2011;Carvalho et al., 2013;Li et al., 2013).

Conclusion
The study showed that the strains of Salmonella sp., Isolated from street foods, are of different serotypes and resistant to the antibiotics commonly used in human medicine in Chad. This particularly pronounced resistance to antibiotics of the quinolone family is a real concern for public health. The prescription and use of antibiotics, whether for the treatment of animals or humans, must be the subject of mature studies in order to better control the proliferation of antibiotics.
Institute of Livestock Development ("IRED") who offered us a technical platform to a higher level for our research.