Detection and Profiling of Antibiotic Resistance among Culturable Bacterial Isolates in Vended Food and Soil Samples

The emergence and persistence of antibiotic resistance remain formidable health challenges. This study aimed at detecting and profiling antibiotic resistance of bacterial contaminants in vended food and the environment. Seventy antibiotic-resistant bacterial isolates were isolated from fried fish, African sausages, roasted meat, smokies, samosa, chips (potato fries), vegetable salads, and soil samples collected from Embu Town and Kangaru Market in Embu County, Kenya. The antibiotic susceptibility test, morphological and biochemical characterization, antibiosis assay, polymerase chain reaction-based detection of antibiotic resistance genes, and sequencing of the 16S rRNA gene were done. Analysis of variance on all measured data was done, and Tukey's honest test was used to compare and separate mean diameters of zones inhibition. Resistance of bacterial isolates to antibiotics was chloramphenicol (90%), cefotaxime (84.29%), nalidixic acid (81.43%), tetracycline (77.14%), amoxicillin (72.86%), gentamycin (48.57%), streptomycin (32.86%), and trimethoprim + sulphamethoxazole (30%). Isolate KMP337, Salmonella spp., exhibited highly significant antibiosis against S. aureus recording a mean inhibition diameter and standard error (SE) of 16.33 ± 0.88 mm, respectively, at P=0.001. The 70 bacterial isolates belonged to Bacillus, Paraclostridium, Lysinibacillus, Virgibacillus, and Serratia genera. The study isolated Bacillus wiedmannii (KC75) which is a risk group 2 as well as Serratia marcescens (KMP95) and Bacillus anthracis (KS606) which are risk group 3 organisms. The presence of antibiotic resistance genes Tet A, BlaTEM, StrB, Dfr A, Amp, and FloR genes was confirmed by a polymerase chain reaction. Samples from Kangaru Market recorded a higher (88.57%) proportion of resistant isolates as compared to isolates from Embu Town (11.43%). The study confirmed the presence of antibiotic-resistant bacteria in vended fast food and the soil in Embu Town and Kangaru Market. This study calls for continuous monitoring of bacterial status and hygienic handling of vended food.


Introduction
Urban centers in developing countries face socioeconomic challenges that can adversely affect health standards. Foodborne diseases, particularly those of microbial origin, are a major health problem and contribute to reduced economic growth [1]. Street foods are popular in urban areas because of their availability, convenience, and affordability. However, the hygiene aspects of sales operations are a major source of concern for the health of both handlers and consumers [2]. Running surface water, inadequate sanitation, and the disposal of untreated wastewater provide an environment conducive to the growth of pathogenic microbes. Besides, street foods are prone to cross-contamination due to lack of personal hygiene, sharing of contaminated utensils, and free movement of flies that sporadically land on the food.
Food and environmental samples are reservoirs of foodborne pathogens such as Listeria monocytogenes, Vibrio parahaemolyticus, Escherichia coli, Salmonella spp., Vibrio cholerae, and Staphylococcus aureus [1,3]. In addition to antibiotic resistance in bacterial pathogens, commensal bacteria and the environment are also reservoirs of antibiotic resistance. The release of animal and human waste into the environment results in the subsequent release of antibiotics into water and soil, which encourages further spread of antibiotic resistance [4]. Bacterial resistance to commonly used antibiotics has been associated with enteric pathogens such as E. coli, Salmonella spp., Shigella spp., Klebsiella spp., and V. cholerae [5].
A cross-sectional study conducted in 2014-2016 in 22 of Kenya's 47 counties diagnosed 11,033 people with cholera [6]. An outbreak of cholera was reported in Embu County in 2019 [7]. Waterborne diseases in Embu County have been attributed to poor sanitation and contamination of water sources [8]. Despite efforts to report the prevalence of foodand waterborne pathogens in Embu County, data on their susceptibility to antibiotics are scarce, so this study focused on the isolation and characterization of antibiotic-resistant bacteria in food sold in Embu Town and the nearby Kangaru Market.

Study Site and Sampling.
Samples consisting of fried fish, African sausages, roasted meat, smokies, samosa, chips (potato fries), vegetable salads, and soil were purposively collected from 176 out of the 324 licensed food vendors in the densely populated Central Business District of Embu Town and the nearby Kangaru Market (latitude: 0°31′58.80″N and longitude: 37°27′0.00″) (Figure 1). Samples were collected aseptically and transported to the Microbiology Laboratory at the University of Embu where they were stored in a refrigerator at 4°C. Treated water from Embu Water and Sanitation Company (EWASCO) was also sampled.

Preparation of Inoculum and
Subculturing. The samples were incubated in eight selective media, namely, MacConkey agar, Salmonella Shigella agar, brain-heart infusion agar, phenylalanine agar, Pseudomonas fluorescens agar, brilliant green agar, eosin ethidium bromide agar, and deoxycholate agar. The standard cultivation method for bacterial food contaminants isolation was carried out as recommended by the International Organization for Standardization (ISO) 6579 [9]. The plates were incubated at 37°C for 24 hours, in an inverted position, after which single colonies were subcultured on fresh media.

Characterization and Identification of Bacterial Isolates.
The isolates were characterized as described in [10]. The ability of the isolates to ferment sugars, utilize citrate, and produce catalase enzyme and phenyl-pyruvic acid was determined as previously described [11,12]. The isolates were screened for their ability to produce amylase, protease, cellulase, and esterase using the methods described in [13,14]. Probable identification of the isolates was done using Bergey's manual of determinative bacteriology and confirmed by partial sequencing of the 16S rRNA gene of some isolates.

DNA Extraction from Antibiotic-Resistant Bacteria.
The bacterial isolates were grown overnight in Luria Broth for 24 hours, and DNA was extracted using the phenolchloroform method with some modifications [19]. To isolate plasmid DNA, bacterial isolates were grown overnight in Luria Broth. Plasmids were extracted using a protocol modified from [20].

Partial
Sequencing of Bacterial 16S rRNA Gene. The 16S rRNA gene sequence was amplified by PCR using bacterial universal primers 8F (5′-AGCTTTGATCCTGGCTCAG-3′) and 1492R (5′-CGGCTACCTTGTTACGACTT-3′) for all selected bacterial isolates according to [21] using Phusion-High fidelity PCR Kit (New England and Biolabs Inc). The PCR amplicons were checked by gel electrophoresis, labeled, and shipped to Inqaba Biotech African's Genomics Company in South Africa for sequencing. The sequence outputs were edited, analyzed, and deposited in GenBank for assignment of accession numbers.

PCR Amplification of Antibiotic Resistance Genes.
Genes encoding resistance to tetracycline, streptomycin, and beta-lactam antibiotics such as ampicillin, chloramphenicol, and trimethoprim were detected using the primer specific PCR approach (Table 1) as described in [22]. The presence of a target gene was confirmed by the presence of a band on the gel.

Data
Analysis. Data on antibiosis, extracellular enzymatic activity, and antibiotic sensitivity were recorded in excel sheets. The general linear model (PROC GLM) procedure of SAS software version 9.1 (SAS Institute, Cary, NC) was used to perform analysis of variance (ANOVA). Tukey's honest significant difference (HSD) test was used to compare and separate the means of the diameter of zones of inhibition. Correlation profiles of zones of clearance on different substrates and zones of inhibition were visualized as a heatmap generated by a hierarchical clustering R script using R version 3.3.1 software. The genetic affiliation of the sequenced genes of the isolates was deduced from a phylogenetic tree generated using MEGA X. The analysis involved 20 nucleotide sequences. The evolutionary distances were computed using the maximum composite likelihood method and were recorded as units of the number of base substitutions.

Effect of Antibiotics on Bacteria Isolated from Food and
Soil. Antibiotic sensitivity test assays revealed that 70 out of 345 bacterial isolates were resistant to at least three (multiple resistant) antibiotics that were tested. Isolate KMP253 affiliated to Salmonella spp., TF152 affiliated to E. coli, and KS120 affiliated to Bacillus subtilis were resistant to all the antibiotics that were tested and were all isolated from Kangaru Market ( Table 2). Hierarchical clustering showed that trimethoprim + sulphamethoxazole was the most effective antibiotic, while chloramphenicol was the least effective antibiotic. Isolate KVS120 was resistant to the highest number of antibiotics ( Figure 2). The occurrence of Bacillus wiedmannii  Figure 3. No bacteria were recovered from treated water provided by EWASCO.

Detection of Antibiotic-Resistant
Genes. Antibiotic resistance genes were detected in 36 out of the 70 antibioticresistant bacterial isolates (  Color key Figure 2: Activity of eight antibiotics against antibiotic-resistant bacteria isolated from vended food and soil in Embu Town and Kangaru Market. Strep, streptomycin; Genta, gentamycin; Amox, amoxicillin; tetra, tetracycline; Chlo, chloramphenicol; Tri-Sul, trimethoprim + sulphamethoxazole; Cefotax, cefotaxime; Nalid, nalidixic acid. 6 International Journal of Microbiology largest zone of clearance on cellulose, while KC430 affiliated to Pseudomonas spp. had the largest zone of precipitation on the tween-20 substrate.

Amplification of 16S rRNA Genes and Phylogenetic
Analysis. The 16S rRNA gene was successfully amplified in 47 isolates out of the 70 antibiotic-resistant bacterial isolates ( Figure 5).
Phylogenetic analysis of successfully sequenced bacterial isolates revealed that the isolates belonged to five genera, namely, Bacillus, Paraclostridium, Lysinibacillus, Virgibacillus, and Serratia with the similarity of above 90%. The Bacillus clade had the highest number of successfully amplified bacterial isolates distributed in nine subclades with bootstrap values of 98 ( Figure 6).

Discussion
The present study isolated 70 bacterial isolates that were resistant to at least three antibiotics from vended foods and the soil in Embu Town and Kangaru Market. Higher levels of antibiotic resistance were recorded from bacteria isolated from vended food and soil in Kangaru Market compared to Embu Town. The presence of antibiotic-resistant pathogens in street vended foods has been recently reported in Kisumu, Kenya [23]. An earlier study in Kibera, Kenya, reported that food and the environment are reservoirs of antibiotic resistance [24]. Food contaminated with antibiotic-resistant bacteria is a threat to public health as it encourages persistence and dissemination of resistance determinants [25]. The higher level of antibiotic resistance recorded in the Kangaru Market may be due to laxity in the enforcement of public health measures. Resistance to gentamycin and streptomycin which are administered by intramuscular injection was 48.57% and 32.86%, respectively. The intramuscular mode of administration possibly limits the use of these antibiotics and thus contributes to the preservation of the antimicrobial activity [26]. In the present study, the frequency of antibiotic resistance to oral antibiotics was high and ranged from 72.86% for amoxicillin to 81.43% for nalidixic acid. The high level of resistance to oral antibiotics is possibly due to availability, convenient mode of administration, broad-spectrum activity, and affordability.
Fish samples in the present study had the highest (18.52%) proportion of antibiotic-resistant isolates. The contamination of fish with antibiotic-resistant bacteria has been previously attributed to fecal contamination of water due to the unhygienic disposal of human waste [27]. Earlier studies reported contamination of water with bacteria that  International Journal of Microbiology   International Journal of Microbiology were multidrug resistant to sulphamethoxazole/trimethoprim, ampicillin, and tetracycline [24,28]. The present study confirmed that treated water from EWASCO that is used by most of the people in Embu Town and Kangaru Market was free from contamination by bacteria at the time of sampling. Therefore, it is unlikely that antibiotic resistance in this study originated from EWASCO water. The current study isolated antibiotic-resistant bacteria from vegetable salad and African sausages, which are mostly homemade. Contamination of smokies, which are also a kind of sausage that is factory made and packed, with antibiotic-resistant bacteria was low. Postharvest contamination of vegetables and cross-contamination of African sausages are a major risk factor for disseminating antibiotic resistance [29,30]. Soil samples from both Embu Town and Kangaru Market recorded a higher proportion of antibiotic-resistant isolates compared to food samples. In a previous study, anthropogenic activities such as long-term manure application were shown to increase antibiotic-resistant bacteria in soil [31]. Such activities were not deduced in the present study. Isolation of Pseudomonas spp. from both soil and fish in Kangaru Market indicates the possibility of food contamination originating from the environment.
The present study showed that antibiotic resistance was encoded by TetA, Bla TEM, StrB, DfrA, Amp, and FloR genes. In this study, genes encoding streptomycin resistance StrB were detected in only nine out of the 23 isolates that were resistant to streptomycin. The StrB genes are known to be encoded by both transposons and plasmids [32]. The present study did not examine transposon-encoded antibiotic resistance, which could explain the detection of StrB in fewer isolates. Five of the nine isolates positive for FloR genes were isolated from food. This could be due to the use of chloramphenicol for domestic purposes [33]. Contamination of food by antibiotic-resistant bacteria could become a major threat to public health since the antibiotic resistance determinants can be transferred to pathogenic bacteria thus complicating the treatment of bacterial infections.
Sequencing  Figure 6: Phylogenetic relationships of bacterial isolates from vended food and soil in Embu Town and Kangaru Market. The evolutionary history was inferred through the neighbor-joining method. Isolates from the present study are presented in the phylogenetic tree using the isolate number codes. 10 International Journal of Microbiology group 2 from food, and B. anthracis, a risk group 3 organism from the soil, is of public health concern. The pathogen B. anthracis can cause the death of both human beings and animals [35] (https://my.absa.org/tiki-index). Since S. marcescens occurs naturally in soil and water, it can cause urinary, respiratory, meningitis, pneumonia, septicemia, and endocarditis infections in man. Isolation of S. marcescens from the African sausage in this study could be a result of contamination during preparation, as African sausages are mostly homemade. The presence of S. marcescens a risk group 3 organism [7] in food is of grave concern. Although isolate TS572 affiliated to B. subtilis was resistant to five antibiotics, it also had antimicrobial activity against K. pneumoniae ATCC 49619. This is not surprising as B. subtilis from soil have been reported to produce antimicrobial agents [36]. KMP337 affiliated to C. freundii recorded the highest antibiosis activity against S. aureus ATCC 1026. Previous studies associated C. freundii with multiple antibiotic resistance but not the production of antimicrobial agents [37]. The present study isolated B. amyloliquefaciens from fruits in Kangaru Market. In a previous study, B. amyloliquefaciens was isolated from banana fruit surfaces and was shown to produce antimicrobial activity against crown root causing pathogens. Thus, B. amyloliquefaciens could be explored further for antimicrobial agents [38]. Isolation of bacteria with antimicrobial activity in the present study supports the need for innovative screening strategies in the search for new antimicrobial agents. Most of the bacteria isolated in the current study were able to produce extracellular enzymes. Other studies have suggested that isolates that exhibit high enzymatic activity also possess high antibiosis activity; thus, these processes could be occurring concurrently [18,39].

Conclusions
This study showed that vended food and the soil in Embu Town and Kangaru Market contain bacterial food contaminants that are resistant to commonly used antibiotics. The study further confirmed that the bacteria harbored antibiotic resistance determinants TetA, Bla TEM, StrB, DfrA, Amp, and FloR. Some of the bacterial food contaminants isolated in this study are risk group 2 and 3 organisms which is a public health concern. More resistance was recorded in the bacteria originating from Kangaru Market as compared to Embu Town. Overall, vended food in Kangaru Market may pose a health risk to the public if mitigating measures are not put in place.

Data Availability
Additional data for this manuscript have been provided as supplementary material.

Disclosure
A license to conduct the research was obtained from the National Commission for Science, Technology and Innovation (NACOSTI) in Kenya. Authorization to collect samples was granted by the Public Health Ministry of Embu County.

Conflicts of Interest
The authors declare that there are no conflicts of interest in the publication of this paper.