The study was carried out at primary schools located at Kibera informal settlement, Nairobi County, Kenya in July 2015. Kibera is located at an altitude of 1670 m above sea level, at latitude 36°50’ east and longitude 1°17’ south, about 140 km south of the equator. Kibera is located 5 km South of Nairobi Central Business District (CBD), the Capital of Kenya. Kibera is divided into 9 official villages. The average living place is 3 m ², with an average of 5 persons per place. The study site presents a population with diverse enteric infections ²⁸ . In total, 11 primary schools with pupil population ranging from 120 to 189 were randomly sampled and, 40 to 80 stool samples collected from pupils in each school, depending on the school population, making a total of 580 stool samples. With a known prevalence of 40.7% of soil transmitted helminths in school going children in urban Kenya, the formula by Martin et al. (1998) was used to determine the desired minimum sample size. The schools were distributed in five administrative villages, namely Lindi, Silanga, Laini Saba, Gatwekera and Mashimoni. Participants’ parents provided written consent through the care givers. This was done during parents’ school meetings, where parents were informed of the intended study and its benefits, those who agreed their children to participate were issued with consent forms for them to sign and return to their class teacher. Only those who their parents consented participated in the study.

Research clearances were given by National Commission for Science, Technology and Innovation (research clearance permit No. 3756) and ethical clearance (PKU/278/1274) was granted by Kenyatta University Ethical Review Committees.

Campylobacter spp. culture. In the laboratory, 5 g of freshly collected faecal sample was pre-enriched by suspending the faeces in 45 ml buffered peptone water (BPW) (Oxoid, Hampshire, England) and incubating the suspension at 42°C for 18 hours in a 50-ml closed culture tube. The pre-enrichment was inoculated onto modified campylobacter charcoal-cefoperazone deoxycholate (mCCDA) agar plates with supplement (polymyxin B 2500IU, rifampincin 5 mg, trimethoprim 5 mg and cycloheximide 50 mg) using a sterile swab and the plates incubated at 45°C for up to 48 hours under anaerobic conditions. The mCCDA culture media (Oxoid, Hampshire, England) was prepared according to manufacturer’s instructions and stored at 4°C until use. Anaerobic conditions were achieved by adding a 21.3-g sachet of CampyGen ^TM 3.5 L (Oxoid, Hampshire, England) in an anaerobic jar with the cultures resulting to a maximum of 13.2% O ₂ within 24 hours and 9.5% CO ₂ in 1 hours. After 24 hours of incubation, the plates were checked for characteristic growth and plates without growth were re-incubated for an additional 24 hours. Characteristic colonies (grey/white or creamy grey in colour with moist appearance) were examined and counted. Distinct colonies were harvested and tested for oxidase and peroxidase breakdown, by picking a portion of distinct colony with a sterile wire loop and placing it on a drop of 30% hydrogen peroxide on a clean microscope slide. Production of effervescent air bubbles was recorded as peroxidases positive. The same colonies were tested for cytochrome oxidase enzyme production by placing a portion of the test colony onto oxidase paper impregnated with NNN’N’ tetramethyl-p-phenylene-diamine dihydrochloride (Oxoid, Basingstoke, UK). Purple colour change was recorded as positive reaction. Reactive colonies were processed for DNA and a portion stored in skimmed milk at -80°C for further characterization.

DNA preparation from bacteria colonies and multiplex PCR. For DNA preparation, three distinct colonies from pure bacteria cultures were lifted with a sterile wire loop and suspended in 0.5 ml sterile, distilled water. The suspension was boiled for 30 minutes in a water bath. After cooling to room temperature, the preparation was centrifuged at 2000 x g and the supernatant harvested and stored at -20°C until analysis by polymerase chain reaction (PCR). PCR was first undertaken to confirm Campylobacter genus for the isolates after which three specific species were also identified: C. coli, C. jejuni and C. lari. The Campylobacter DNA preparation (2 µl) was amplified in a 25 µl reaction mix by mixing 2.5 µl 10X PCR buffer (Coraload), 0.5 µl dNTPs, 0.125 µl Taq DNA polymerase (Inqaba biotec, Pretoria, South Africa) and 0.1 µl of each specific primer to 10 pmole (Inqaba Biotec, Pretoria, South Africa), 2 µl DNA template and 18.657 µl DNAse/RNAse-free distilled water. The DNA was amplified using a program of initial heating at 94°C for 5 min followed by 30 cycles of denaturation at 94°C for 1 minutes, annealing at 56°C for 1 min, extension at 72°C for 1 min with a final extension of 72°C for 10 min using a Veriti 96 wells thermocycler, (Applied Biosystems, model 9902, Singapore) in 0.2-ml PCR tubes. The PCR products were kept at -20°C until gel electrophoresis was done.

The Campylobacter genus-specific primers, C412F and C1228 R, described by Linton et al. ²⁹ were used to amplify a 812 bp fragment within the 16S rRNA gene of Campylobacter species using forward primer C412F 5’-GGATGACACTTTTCGGAGC-3’ and reverse primer; C1228R 5’-R-CATTGTAGCACGTGTGTC-3’. Multiplex PCR was carried out for C. jejuni and C. coli with specific primers CjejlpxAF, CjejipxAR (shared by both species) and CcollpxAF, described by Klena et al. ¹⁵ to amplify 331 bp and 391 bp fragment flanking the lpxA gene. The primer sequences were; CjejlpxAF (forward) 5’-ACAACTTGGTGACGATGTTGTA-3’, CjejipxAR (reverse, shared by CjejlpxA and CcollpxA) 5’-CAATCATGDGCDATATGASAATAHGCCAT-3’ for C. jejuni and for C. coli CcollpxAF (forward) 5’-AGACAAATAAGAGAGAATCAG -3’. The C. lari specific primers were forward primer lpxAC , 5’-AGACAATAAGAGAGAATCAG-3’ and reverse primer lpxARKK2M, 5’CAATCATGDGCDATATGASAATAHGCCAT-3’.

The PCR products were visualized by electrophoresis in a 1.5% agarose (Genetics analysis grade, Fisher Scientific, New Jersey) gel stained with 0.02% ethidium bromide and amplicons identified against molecular marker (50 bp DNA ladder, England Biolab) run alongside the samples.

For confirmation, the positively identified PCR products were submitted for sequencing. The PCR products were fist purified using exonuclease1, shrimp alkaline phosphatase mixture (ExoSAP mix) according to the manufacturer’s instructions. Briefly, this was done by adding 2.5 µl of ExoSAP mix to 10 µl PCR product. The mixture was then incubated at 37°C for 30 minutes and reaction stopped by heating at 95°C for 5 minutes. The clean PCR product was then quantified using a fluorimeter (Qubit 2.0, Invitrogen, USA). The clean DNA was first labelled with BigDye terminator v3.1kit (Applied Biosystem, CA, USA) according to the manufacturer’s instructions and loaded into Genetic Analyzer (ABI 3730 capillary analyser; Applied Biosystems, Foster City, CA, USA) for sequencing. Sequences were obtained in ABI files that were opened and edited to remove unspecific ends using BioEdit version 7.0.4 (Hall, CA, USA) software. Clean sequences were then submitted to NCBI GenBank database and BLASTn program used to test for homology and genetic identity of bacteria isolates.

Antimicrobial sensitivity test (AST) for PCR-confirmed Campylobacter spp.. Campylobacter spp. isolates were phenotypically tested for resistance using selected antimicrobial agents according to European committee on antimicrobial susceptibility testing (EUCAST) ³⁰ . Only those antibiotics with EUCAST established breakpoints were tested, namely tetracyclines (tetracycline 30 mg), quinolones (ciprofloxacin 5 mg, naladixic acid 30 mg) and macrolides (erythromycin 15mg). Mueller-Hinton agar plates plus 5% de-fibrinated horse blood with 20 mg/L β-nicotinamide adenine dinucleotide Mueller-Hinton fastidious (β-NAD (MH-F)); (Oxoid, Basingstoke, UK) were prepared and dried at 35°C, with the lid removed, for 15 min prior to inoculation to reduce swarming. Inoculum turbidity was adjusted to McFarland 0.5 prior to inoculation. The antibiotic discs were placed on the inoculated plates using a sterile multi-disc dispenser and incubated in a microaerobic environment at 41±1°C for 24 hours. Isolates with insufficient growth after 24 hours of incubation were re-incubated immediately and inhibition zones read after a total of 40–48 hours incubation. The inhibition zones were defined by the point showing no growth when viewed from the front of the plate with the lid removed and with reflected light.

Genotypic characterization of Campylobacter spp. isolates for antimicrobial resistance. The 90 antibiotic resistant, PCR-characterized Campylobacter spp. isolates including; 11 C. lari, 30 C. coli and 49 C. jejuni were selected for demonstration of genes encoding for resistance to tetracyclines. Presence of four tetracycline resistance genes, tet(A), tet(B), tet(C) and tet(O), were tested. Multiplex PCR was carried out as described above. Primers used for amplification of products encoding for the resistant genes to tetracyclines are shown in Table 1.

The EUCAST disk diffusion method was used to determine the resistance patterns of only the identified isolates, 68 ( C. jejuni, C. coli and C. lari) confirmed by PCR ( Table 2). Fifteen isolates were not recovered from storage culture after identification and thus not tested . All of the antibiotics studied had isolates showing resistance towards them, with 96% of isolates resistant to tetracycline (30 mg), 93% to naladixic acid (30 mg) and all the isolates tested resistant to erythromycin (15 mg). The antibiotic that most isolates were sensitive to was ciprofloxacin (5 mg) which still had 84% of the isolates showing resistance ( Table 3). Of the four tet genes tested, tet(A) was most frequently identified in 20 (29.1%) of the isolates followed by tet(O) in 8 (11.7%) isolates and tet(C) in only 2 (2.9%) isolates. None of the isolates had more than one tet gene demonstrated. ( Table 3).

Four MDR profiles were observed. All of the tested isolates were resistant to two or more antimicrobial agents, but the majority of isolates (84%) were resistant to all the antibiotics studied (profile 3 and 4). Campylobacter jejuni had the highest number of isolates that were MDR with 25 (37%) isolates being resistant to all four antibiotics tested (profile 1). C. coli had 23 (34%) isolates resistant to all the four antibiotics while C. lari had 9 (13%) isolates resistant to the four antibiotics. One (1.5%) C. jejuni and C. lari isolates was resistant to drugs in profile 2, while three (3%) C. coli isolates were in this profile. However, profile 4 had only one (1%) C. coli isolate while profile 4 had 2 (3%) C. lari, 4 (6%) C. coli and 5 (7%) C. jejuni MDR isolates ( Table 4).