A review of antibiotic resistance among Campylobacter species in human, animal, and water sources in South Africa: a One Health Approach

Campylobacter species are among the aetiological agents for gastroenteritis in humans and are a leading cause of gastroenteritis in developed and developing countries. Globally, pathogenic Campylobacter species are responsible for 400–500 million diarrhoea cases per annum and have been classified as emerging infections (Igwaran & Okoh 2020b). It is estimated that Campylobacter jejuni and Campylobacter coli are responsible for about 98% of all human Campylobacter gastroenteritis cases (Gahamanyi et al. 2020). Campylobacter infections are generally self-limiting, but in cases where symptoms persist and in immunocompromised individuals, antibiotics such as macrolides (erythromycin), fluoroquinolones (ciprofloxacin), and tetracycline are administered. Macrolides and fluoroquinolones are the first line of treatment when therapeutic intervention is required, while tetracycline serves as a choice in treating clinical Campylobacter infection (Shobo et al. 2016). The development of antibiotic resistance is observed in Campylobacter species, and this has serious implications for treating Campylobacter infections in humans (Sithole et al. 2021). Antibiotic resistance is suspected in cases of treatment failure evidenced by relapse.

The increasing number of immunocompromised individuals and bacterial infections have stirred up a pattern of largely empirical antibiotic prescriptions in humans in South Africa. This is coupled with the common use of antibiotics in animal feeds and veterinary medicine (Manderson 2020). However, the burden and incidence of antibiotic resistance for Campylobacter species in South Africa are uncertain and may be underestimated due to poor diagnosis and surveillance (Thobela 2017). Locally, Campylobacter has been isolated from numerous sources such as water (Kalule et al. 2019), pigs (Leblanc-Maridor et al. 2011), cattle (Karama et al. 2020), poultry (Bester & Essack 2012; Roess et al. 2015; Pillay et al. 2020), companion animals (Karama et al. 2019), and wild animals (Moré et al. 2017). The presence of Campylobacter species in different reservoirs suggests that they can be transmitted through these reservoirs. South Africa is a water-scarce country and therefore relies on its surface water for domestic, agricultural, and industrial applications. Besides, in some communities where best grazing management is not practiced, humans and animals share the same water source. Also, the use of antibiotics in animal feeds is a common practice in South Africa, and this practice has been linked to the development of resistance to animal pathogens. Therefore, there is potential for sharing of antibiotic resistance bacteria and genes between animals, humans, and the environment (Bengtsson-Palme 2017). The risk of dissemination of antibiotic-resistant (AR) Campylobacter species between humans, animals, and the environment is anticipated. Contamination of the environment with antibiotics, resistant bacterial strains, and antibiotic-resistant genes (ARGs) boosts the dissemination and spread of drug-resistant pathogens. Antibiotic resistance genes can potentially be incorporated into human pathogens and passer-by pathogens with time through horizontal gene transfer. Thus, having emerged at the human–animal–environment interface, Campylobacter presents a One Health challenge. One Health recognises the link between the health of animals, humans, and the environment. Thus, applying approaches that consider the inclusion of environmental and animal health as an essential component of disease surveillance, control, prevention, and mitigation is critical (Mackenzie & Jeggo 2019).

It is against this background that this review was conducted to analyse available information on the prevalence of antibiotic resistance among clinical Campylobacter species isolated from humans, animals (avian, cattle, swine), and environment (water) in South Africa. A better understanding of patterns and determinants of Campylobacter antibiotic resistance will enable more appropriately targeted interventions. Furthermore, with the current increasing travel and trade among intercontinental cities, there is an increasing risk of the spread of diseases from any part of the globe through a highly mobile populace. Therefore, beyond South Africa, sharing such information with the global health community is one way of alerting responsible authorities to enable timely prevention of transnational impact. This is in line with the Global Health Security Agenda (GHSA). The objective of this paper is to present a situation analysis that reports the current patterns and determinants of Campylobacter antibiotic resistance in South Africa. Local prevalence data are needed to generate evidence and information which can be used to develop effective and targeted interventions and influence policies, on issues such as antimicrobial use. This review highlights the prevalence of ARGs, AR Campylobacter, and virulence genes (VGs), and the methods that have been employed in these studies. Firstly, the method of data collection for this systematic review is described. This is followed by a section describing the methods that have been employed in the studies on AR Campylobacter in humans, water, and animals (pigs, chickens, and cattle samples) in South Africa. The methods employed in isolation/identification, DNA extraction, antibiotic sensitivity testing, detection of virulence and ARGs, as well as relatedness among species are described. Thereafter, the prevalence of AR Campylobacter, VGs, and ARGs, in relation to the regulations for antibiotic use in South Africa are discussed. The last two sections discuss the findings of this review and give a conclusion.

A search was conducted on PubMed and Science Direct for relevant publications that were published between 2012 and 2021 in English. The search was conducted with a predefined search string and adapted keywords. The key terms used were Campylobacter AND Prevalence AND South Africa. The search on PubMed yielded 874 results but after restricting the search to research articles only, 213 results were obtained. On the other hand, the search on PubMed yielded 52 results. A total of 265 articles were screened using the criteria described below.

The selection criteria for the inclusion of studies were:

• The studies which were conducted in South Africa.

• Studies on AR Campylobacter in humans, animals, food, and water.

• Studies on the prevalence of Campylobacter in South Africa.

• Studies that reported Campylobacter VGs and ARGS specifically from humans, animals, food, and water.

• Studies in which the methods employed in isolation/identification, DNA extraction, antibiotic sensitivity testing, detection of virulence genes/ARGs, relatedness among species, and statistical analysis conducted are described.

Exclusion criteria were

• Studies on antibiotic-sensitive Campylobacter species.

• Studies that were not conducted in South Africa.

• Duplicate articles.

• Studies conducted on other bacteria (other than Campylobacter).

Titles and abstracts were screened for location within South Africa and the type of samples (humans, animals, or water). After screening, a total of 16 articles were eligible for review. Full articles of these eligible articles were obtained and after reviewing the full articles, four articles were excluded on the basis that they did not report on antibiotic resistance in Campylobacter species. The following articles were considered for review; two (2) studies on human samples, seven (7) studies on humans, and three (3) on water samples. Data were collected independently from each publication and tabularised in Microsoft word.

The following data were extracted from included studies:

• Study characteristics: period, location, type of samples (human, animal, water).

• Data on AR Campylobacter: prevalence, types of antibiotics, virulence genes, ARGs.

• Methods of detection: microbiological methods, molecular techniques.

Regardless of the sample type investigated, only culture-based methods were employed in the studies recorded. Water samples were firstly filtered through a 0.45 μm Chukwu et al. (2019); Igwaran & Okoh (2020a) or 0.65 μm Otigbu et al. (2018) sterile membrane on a vacuum filter. After that, the concentrated cells on the membrane filters are inoculated on Campylobacter selective enrichment broth and incubated at 37–42 °C in a microaerophilic environment (MAE) for 24–72 h.

For animal studies, samples were either directly inoculated into enrichment media or processed before enrichment. Samples that were directly inoculated into enrichment media included milk samples Igwaran & Okoh (2020a), pig faecal samples Sithole et al. (2021), chicken faecal samples Bester & Essack (2012); Reddy & Zishiri (2017); Mileng et al. (2021), and cattle faecal samples (Karama et al. 2020). Igwaran & Okoh (2020a) homogenised the meat samples in buffered peptone water and collected a portion for enrichment. The enrichment media used included Bolton selective enrichment broth mixed with Bolton broth selective supplement with 5% (v/v) defibrinated horse blood Igwaran & Okoh (2020a) and enrichment charcoal broth (Sithole et al. 2021). The enrichment mixture was incubated at 37–42 °C for 24–48 h under microaerophilic conditions (Igwaran & Okoh 2020a; Sithole et al. 2021).

Notably, the investigation of AR Campylobacter species in pigs by Sithole et al. (2021) was conducted using a ‘farm-to-fork approach’. This approach involved collecting samples at five stages in pork processing, i.e. production, animal handlers, transportation, post-slaughter samples, and meat products.

For human studies, faecal specimens were collected from diarrhoeal and non-diarrhoeal patients (Chukwu et al. 2019) using a modification of the method reported by Besse et al. (2011), which involves processing samples by filtration or direct inoculation on a Karmali agar plate. Karmali agar medium is based on the formula described by Karmali, and it is recommended for the isolation of C. jejuni and C. coli from clinical specimens and foods. The filtration method, on the other hand, involves preparing a suspension of stool suspension in a Brucella broth and then inoculating a drop of the suspension on a 0.65 μm Millipore filter. In other studies, Shobo et al. (2016), and Reddy & Zishiri (2017) analysed preserved human Campylobacter isolates collected before these studies. The isolates were cultured on Campylobacter blood-free Selective Agar Base, supplemented with CCDA Selective Supplement, and incubated at 37 °C in a microaerobic environment for 48 h.

To obtain pure cultures, colonies are generally subcultured by streaking on modified cefoperazone deoxycholate agar (mCCDA) supplemented with antibiotic selective supplement CCDA selective supplement (cefoperazone and amphotericin). The media selective for Campylobacter contain antibiotics cefoperazone and amphotericin inhibit the normal faecal flora making it easy to detect Campylobacter (Igwaran & Okoh 2020a, 2020b; Pillay et al. 2020; Mileng et al. 2021). Tryptose Blood Agar (TBA) supplemented with 5% defibrinated sheep blood is also used to culture the enriched cells (Chukwu et al. 2019) and to sub-culture colonies (Shobo et al. 2016; Pillay et al. 2020). This medium is very nutritious, and the blood supplement promotes the growth of fastidious organisms such as Campylobacter.

This is followed by identification using colony morphology, motility, Gram-stain, no-growth in aerobic conditions, biochemical tests (hippurate hydrolysis and oxidase tests) (Shobo et al. 2016; Otigbu et al. 2018; Chukwu et al. 2019; Sithole et al. 2021). Campylobacter spp. are gram-negative, microaerophilic, motile, and oxidase-positive. The colonies are smooth, colourless translucent to grey in appearance (Sithole et al. 2021). The hippurate test is used to differentiate C. jejuni from C. coli and other Campylobacter species (Bester & Essack 2012; Shobo et al. 2016; Chukwu et al. 2019). C. jejuni can enzymatically hydrolyse hippurate while others from other Campylobacter species are not (Miljković-Selimović et al. 2014). C. jejuni ATCC 33560 and C. coli ATCC 33559 are commonly used control strains in the bacterial identification process (Shobo et al. 2016; Igwaran & Okoh 2020a; Sithole et al. 2021).

Additionally, some of the studies confirmed the presence of Campylobacter in the respective samples using molecular methods, such as conventional or quantitative polymerase chain reaction (PCR or qPCR), targeting the genus-specific 16sRNA gene, or species-specific hipO and asp genes (Shobo et al. 2016; Reddy & Zishiri 2017; Chukwu et al. 2019). The hippuricase (hipO) gene is for C. jejuni, while the aspartokinase (asp) gene is for C. coli. The positive controls commonly used are C. jejuni ATCC 33560, C. coli ATCC 33559, and a reaction mixture without template DNA is used as the negative control (Sithole et al. 2021). Extraction of DNA from the isolates is a prerequisite for PCR-based analyses. To extract DNA from the isolates from the different samples, the different studies applied different methods, such as the heat lysis method (Chukwu et al. 2019), conventional boiling method (Reddy & Zishiri 2017; Platts-Mills et al. 2018; Igwaran & Okoh 2020b; Pillay et al. 2020; Sithole et al. 2021), and a modified heat lysis method (Shobo et al. 2016). In all the studies, the purity and concentration of the extracted DNA were assessed using Nanodrop Spectrophotometry. Samples with a 260/280 ratio ranging from 1.7 to 2.1 (Shobo et al. 2016; Reddy & Zishiri 2017; Chukwu et al. 2019), (Chukwu et al. 2019; Igwaran & Okoh 2020b), 1.8–1.9 (Reddy & Zishiri 2017; Pillay et al. 2020; Mileng et al. 2021) are considered as good quality DNA, which is usable for PCR.

In the recorded studies where antibiotic susceptibility tests have been conducted the antibiotics tested are listed in Table 1.

The disc diffusion method was used to conduct antibiotic susceptibility tests for confirmed isolates from clinical samples (Shobo et al. 2016; Chukwu et al. 2019) and water samples (Otigbu et al. 2018; Chukwu et al. 2019; Igwaran & Okoh 2020b), pig samples (Sithole et al. 2021), chicken samples (Bester & Essack 2012; Pillay et al. 2020), and cattle samples (Igwaran & Okoh 2020b; Karama et al. 2020). This disc diffusion method involves inoculating a direct suspension of colonies from fresh cultures onto the surface of Mueller Hinton agar and placing filter paper disks saturated with a standardised concentration of an antimicrobial agent on the surface. After incubation, the presence/absence and the size of the zone of inhibition around the disk are measured. The measured inhibition zones are interpreted according to reference values based on the Clinical and Laboratory Standards Institute (CLSI) and EUCAST recommended guidelines (CLSI 2019). Guidelines for Enterobacteriaceae are used for the interpretation of results against ampicillin, azithromycin, gentamicin, clindamycin, chloramphenicol, levofloxacin, ceftriaxone and imipenem as there are no guidelines for Campylobacter for these antibiotics (Igwaran & Okoh 2020b).

Antibiotic resistance can also be determined using the minimal inhibitory concentration (MIC). The MIC (mg/L or μg/mL) is the lowest concentration of an antibiotic which can completely hinder the visible growth of bacteria after incubation under controlled in vitro conditions (Kowalska-Krochmal & Dudek-Wicher 2021). Different antibiotics are prepared (mainly at two-fold dilution series) and added to molten agar medium plates in different concentrations. The bacteria being tested are diluted to around log 7.0 CFU/mL and 1–2 μL are added to the plates in different spots (4.0 CFU per spot). One agar plate was used as a control seeded without an antibacterial agent. The plates are incubated under anaerobic conditions at 35 °C for 18–24 h Campylobacter because they are fastidious organisms. Also, Campylobacter grows under microaerophilic conditions. After incubation, the growth on the agar plates is observed, and the MICs are determined as the lowest concentration of drug tested that stopped the growth of bacteria (Benkova et al. 2020). In the study by Bester & Essack (2012), MICs were determined by agar dilution using a modification of Columbian agar supplemented with 7% lysed horse blood and Campylobacter growth supplements.

Chukwu et al. (2019) determined the MIC of the antibiotics using the E-test procedure. An impervious inert strip with a marked, continuous concentration gradient of a predefined antibiotic consisting of more than 15 twofold dilutions is incubated on the agar. After incubation, the lowest concentration of an antibiotic that hindered visible growth is read at the point where the elliptical zone of inhibition intersected against the MIC scale on the strip is considered the MIC (Kowalska-Krochmal & Dudek-Wicher 2021).

Multiple antibiotic resistance (MAR) characteristics of a bacterial isolate are also sought to determine resistance to three or more antibiotics. The MAR index of each of the Campylobacter isolates is determined by dividing the number of antibiotics to which the test isolate showed resistance to (a) by the total number of antibiotics to which the test isolate has been evaluated for susceptibility (b) (expressed as MAR = a/b) (Igwaran & Okoh 2020b).

The studies by Shobo et al. (2016), Reddy & Zishiri (2017), and Chukwu et al. (2019) combined antimicrobial susceptibility testing with the molecular screening of isolates for the presence of genes responsible for antimicrobial resistance. The isolates showing phenotypic resistance to the test antibiotics were subjected to molecular screening to detect genotypic resistance genes employing the conventional PCR method. The presence of quinolone resistance is determined by amplifying the Thr-86-lle mutations in the Quinolone Resistance-Determining Region (QRDR) of the gyrA gene in Campylobacter (Shobo et al. 2016; Reddy & Zishiri 2017; Chukwu et al. 2019). Tetracycline resistance is detected by amplifying the tetO gene while point mutations at position 2075 and 2074 in the 23S rRNA gene determine macrolides (erythromycin) resistance (Reddy & Zishiri 2017; Chukwu et al. 2019). Other ARGs that have also been screened in Campylobacter species include gyrA (235 bp), gyrA (270 bp), bla_OXA-61 (Reddy & Zishiri 2017), and Campylobacter multidrug efflux (cme) gene B (gene Cj366c) (Shobo et al. 2016), tetA, tetB, tetC, tetD, tetK, tetM genes, gyrA gene, ermB gene, catI, and catII genes and the aac(3)-IIa-(aacC2) and VIM, KPC, Ges, bla_OXA-48-like and IMI genes (Igwaran & Okoh 2020b). Following conventional PCR, PCR products (amplicons) were visualised by gel electrophoresis. Selected amplicons were sequenced to identify known/novel mutations conferring resistance on a Genetic Analyser using the Sanger method of DNA sequencing. A Basic Local Alignment Search Tool (BLAST) was then conducted and results were compared to known Campylobacter gene sequences in GENBANK (Sithole et al. 2021). Multiplex PCR assays have also been used to screen the following ARGs, tetA, tetB, tetC, tetD, tetK, tetM, gyrA, ermB gene, catI,catII, aac(3)-IIa (aacC2)^a IMI, KPC, VIM, and bla_OXA-48-like genes in cattle (Igwaran & Okoh 2020b).

Furthermore, real-time PCR, also referred to as qPCR, has also been used to detect the presence of genes and chromosomal mutations conferring resistance to antibiotics in some of the recorded studies. The presence of erythromycin resistance was determined by detecting point mutations at positions 2074 and 2075 in domain V of the 23S rRNA indicating the presence of erythromycin resistance while Thr-86-Ile mutations in the QRDR of the gyrA gene in Campylobacter were used to detect fluoroquinolone resistance in chicken isolates (Pillay et al. 2020). The presence of resistance genes such as gyrA, tetO, 23SrRNA, bla_OXA-61, and cmeB were screened in samples from pigs using qPCR (Sithole et al. 2021). Similar to the conventional PCR, in qPCR, also referred to as Real-Time PCR, DNA is amplified by three repeating steps (denaturation, annealing, and elongation). However, with qPCR, the presence of target genes is noted/seen as PCR progresses (i.e. the DNA is quantified in ‘real-time’). The advantage is that both the presence and quantity data are achieved using the qPCR method (Karlen et al. 2007). PCR primers specific to the target genes are designed based on the gene sequence information in the GenBank database and previously published studies (Reddy & Zishiri 2017). C. jejuni 33560, C. jejuni ATCC 33560, and C. coli ATCC 33559 are generally used as positive controls for the PCR assay of Campylobacter, while the reaction mixture without template DNA is used as a negative control (Pillay et al. 2020).

Out of the recorded studies, nine (9) studies screened for the presence of virulence genes while investigating antibiotic resistance in Campylobacter isolates from the sources. Their virulence factors are critical for their colonisation and pathogenicity. They enable the bacteria to survive within the host. The virulence mechanisms of Campylobacter species are well understood, and a sequence of virulence factors has been described, such as adhesion, chemotaxis, motility, response to oxidative stress, cytolethal distensor toxin production, regulation of iron uptake and invasion, lipopolysaccharide and capsular polysaccharide, and multidrug and bile resistance (Otigbu et al. 2018). Virulence genes responsible for invasion (iam), invasion protein (ciaB) colonisation (flaA), adherence (cadF), toxin production (cdtB), flagella synthesis and modification (flgR), Invasion (cstII), Carbon starvation regulation (csrA), response to environmental stress (htrB), stress tolerance (clpP), and multidrug efflux system (cmeA, cmeB and cmeC) have been assessed. High antimicrobial resistance to antibiotics such as quinolones and tetracycline of C. jejuni isolated from humans and chicken has been closely linked with the presence of several virulence genes (Wieczorek et al. 2018). Molecular methods such as PCR can be used to investigate virulence genes involved in adhesion (cadF, pldA), thermotolerance (dnaJ), invasion (ciaB), and toxin production (cdtA, cdtB, and cdtC). Molecular methods have been used to test for the presence of virulence genes; ciaB, cadF in humans, Chukwu et al. (2019), cdtA, dnaJ, ciaB, cdtC, cdtB, pldA, cadF in chickensPillay et al. (2020), iam, cadF, cdtB, flgR, and flaA in assorted meat samplesIgwaran & Okoh (2020b),cdtB, iam, cadF, flgR in cow milk Igwaran & Okoh (2020b), ciaB, dnaJ, pldA, cdtA, cdtB, cdtC and cadF in pigs (Sithole et al. 2021), cadF, ciaB, dnaJ, pldA, cdtA, cdtB, cdtC in chickens (Reddy & Zishiri 2018), cadF, ciaB in water (Chukwu et al. 2019) cadF, flgR, cdtB, iam in water (Igwaran & Okoh 2020b) and cdtB, cadF, cstII, csrA, htrB, clpP in water as well (Otigbu et al. 2018).

Out of the reviewed studies, three (3) determined the relatedness among the isolates (Chukwu et al. 2019; Pillay et al. 2020; Sithole et al. 2021). The genetic relatedness among isolates helps understand the molecular epidemiology and transmission of the pathogen. This is very important in the implementation of integrated disease surveillance, especially for zoonotic pathogens such as Campylobacter (Du et al. 2018). Chukwu et al. (2019) determined the relatedness (phylogeny) among the isolates by amplifying the conserved region of the 16S rRNA gene for the Campylobacter genus by PCR. The obtained PCR products were purified and subjected to Sanger sequencing. The authors then conducted a search by BLAST to compare the sequences of the PCR product to known Campylobacter sequences in the GenBank. After analysis, a phylogenetic tree based on 16S rRNA is created using MEGA7 (Chukwu et al. 2019).

In another recorded study, Enterobacterial Repetitive Intergenic Consensus PCR (ERIC-PCR) was used to determine the genetic relatedness among Campylobacter isolates. Hence, the clonality of Campylobacter isolates was clonality determined to determine if the isolates originated from common sources (Sithole et al. 2021). The ERICs are 127-bp imperfect palindromes that occur in multiple copies in the genomes of enteric bacteria. ERIC-PCR is used for discriminating different types of strains (Ghorbanalizadgan et al. 2014; Ranjbar et al. 2017). In the study by Sithole et al. (2021) isolates with antibiograms showing similar resistance patterns from different sources along the farm-to-fork continuum were investigated for clonal relatedness. PCR was conducted. The amplicons were electrophoresed, and gel images were captured. Fingerprint patterns were analysed, and dendograms were created using cluster generation (Pearson correlation with a 1% optimisation and an unweighted pair group with arithmetic averages (UPGMA)). Isolates were grouped and could be divided into clusters based on a 75% fingerprint similarity (Sithole et al. 2021).

In the study by Pillay et al. (2020), the genetic relationship among Campylobacter isolates from pigs was determined by pulsed-field gel electrophoresis (PFGE). PFGE involves subjecting isolated genomic DNA to restriction enzyme analysis and analysing digestion products on an agarose gel by applying an electric field that periodically pulses the electric field in different directions, hence allowing for separation of the larger DNA fragments. A portion of isolates representative of sources and antibiograms were selected. Using Electrophoresis macro-restriction fragments and images were analysed. Isolates were grouped into pulse types based on an 80% similarity cut-off (Pillay et al. 2020).

The following section reports the prevalence of antibiotic resistance among Campylobacter isolated from humans, animals, and water environments in South Africa.

In humans, a Campylobacter multidrug resistance rate of 76% is observed. On the other hand, in terms of mono-resistance, the highest level of resistance is to azithromycin (69.7%), whereas the lowest level of resistance of Campylobacter is to gatifloxacin (8.3%) (Table 2). The high resistance observed may be a result of the overuse of azithromycin. Azithromycin is a broad-spectrum antibiotic that is commonly used in both humans and animals. Owing to the high incidence of resistance to fluoroquinolones among human isolates, macrolides such as erythromycin and azithromycin have become the drugs of choice for human campylobacteriosis. High resistance to clinically relevant antibiotics, as observed in these studies, raises serious public health concerns (Shobo et al. 2016; Bolinger & Kathariou 2017). Between 2016 and 2019, the level of resistance to tetracycline in humans remains constant, at 33.30 and 32.00%, respectively. Alternatively, a drop in the level of resistance Ciprofloxacin and Erythromycin from 23.60 to 18.00%, and 33.30 to 26.70%, respectively.

The studies involving Campylobacter isolated from animals and animal products in South Africa show high resistance to commonly used antibiotics among Campylobacter species (Table 2). Campylobacter isolates from commercial free-range and industrially raised chickens demonstrate high resistance to antibiotics (tetracycline 100%, Ciprofloxacin 98%) while the lowest resistance is in rural-raised chickens (0% resistant to Erythromycin and Gentamicin) (Bester & Essack 2012). Campylobacter isolates from milk samples and pig samples show very high resistance (over 50%) to common antibiotics, with the highest observed resistance to erythromycin (95.06%), clindamycin (95.68%), and doxycycline (87.65%). The highest resistance recorded for isolates from pigs is erythromycin (90%), tetracycline (84.3%), streptomycin (88%), and ampicillin (73%), while the lowest is gentamycin (11.6%), and nalidixic acid (27.2%) (Sithole et al. 2021). Clindamycin (95.68%) and erythromycin (95.06%) are the highest reported resistance among the isolated milk samples from cattle. (Igwaran & Okoh 2020b). High resistance among isolates from animal and animal products can be attributed to the high usage of antibiotics in livestock production, evidenced by low antibiotic resistance in rural-raised chickens where usage is minimal (Bester & Essack 2012). In general, between the years 2012 and 2021, the level of resistance to Ciprofloxacin in poultry has dropped to 21%. In contrast, within that same time frame, the level of resistance to tetracycline in poultry has shown no significant reduction.

Overall, 100% resistance of Campylobacter isolates from water samples to tetracycline, imipenem is recorded. Also, resistance to clarithromycin (95%), azithromycin 92%, clindamycin 84.2%, clarithromycin 80%, doxycycline 80%, and ciprofloxacin 77.8% (Otigbu et al. 2018; Chukwu et al. 2019). Water samples recorded the highest antibiotic resistance among Campylobacter species. The aquatic environment is a recipient and reservoir of AR bacteria and antibiotics from animals, run-off, humans, and other anthropogenic activities. This may be why Campylobacter isolates from water samples demonstrate high resistance to clinically relevant pathogens, as shown in Table 2.

The clonality and relatedness among isolates have been analysed in some of the studies. In the study by Sithole et al. (2021), a high diversity among Campylobacter isolates (collected from Farm to Fork) is reported. However, Campylobacter isolates from the paediatric stool and household drinking water samples in the Northwest Province of South Africa have demonstrated genetic similarity, suggesting the movement of AR bacteria between humans and the environment. The study also reported a close relationship among some Campylobacter isolates from both water and stool samples, indicating the likelihood of transmission of pathogenic Campylobacter from water to humans (Chukwu et al. 2019).

This study reviews the status of resistance of Campylobacter spp. to commonly used antibiotics in South Africa. In general, the level of resistance of Campylobacter isolated from humans, animals, and environment to tetracycline, imipenem, cefuroxime, clarithromycin, azithromycin, clindamycin, clarithromycin, doxycycline, ciprofloxacin, amoxicillin/clavulanic acid, ampicillin, cephalozolin, erythromycin, and doxycycline is high for normal practical treatment schemes.

This raises serious concerns about the use of these antibiotics as a drug of choice to achieve the best treatment of Campylobacter infections in South Africa. Important information regarding second-line antibiotic options is also presented. The low to moderate resistance levels (15, 11, and 45%) observed for Campylobacter isolated from humans, animals, and water samples, respectively, could be because these antibiotics are not the preferred choice of medicine used, compared to tetracycline, imipenem, cefuroxime, clarithromycin, azithromycin, clindamycin, doxycycline, ciprofloxacin, amoxicillin/clavulanic acid, ampicillin, cephalosporin and erythromycin.

Therefore, the choice of antibiotics to be prescribed to a patient suffering from Campylobacter infection must be guided by a pattern of antibiotics susceptibility test results of patients in South Africa. Furthermore, there is evidence of a high and widespread multidrug-resistant Campylobacter spp. in South Africa. Antibiotic resistance is suspected in cases of treatment failure of Campylobacter infections evidenced by relapse. Generally, the increasing number of immunocompromised individuals, coupled with increasing bacterial infections, has stirred up a pattern of largely empirical antibiotic prescription in humans in South Africa. This necessitates routine verification of the susceptibility of Campylobacter spp. to common antibiotics used for the treatment of the infection as a strategy to combat crisis due to AR Campylobacter infections.

Campylobacter species isolated from animals have also shown high resistance to clindamycin, erythromycin, doxycycline, tetracycline, streptomycin, and ampicillin. This observation concurs with the observation of high mono-antibiotic resistance rates (close to 86%) (Founou et al. 2018). These antibiotics are commonly used for the strategic treatment of infections in humans and animals and, therefore, must be preserved. There is a need to regulate the use of clinically relevant antibiotics in animal production. For example, the guidelines for using antimicrobials in South Africa permit streptomycin, erythromycin, ampicillin, and tetracycline in pigs. These antibiotics are frontline drugs for the treatment of Campylobacter infections in humans. The most realistic explanation for the observed higher diversity of antibiotic resistance genes in animals compared to humans is the misuse, overuse, and inappropriate use of antimicrobials in livestock production as prophylaxis/therapy and growth promoters and having non-prescription antibiotics that are registered under the Farm Feeds, in South Africa (Henton et al. 2011; van den Honert et al. 2018). Therefore, the animals are exposed to a greater variety of antibiotics compared to humans. Antibiotics act as a selection pressure, which increases and accelerates the likelihood that the bacterial pathogens in animals will adapt and multiply to produce a more resistant population (Founou et al. 2016).

This study also reveals the potential risk of waterborne AR Campylobacter infections, especially for communities that rely on the rivers for domestic and agricultural (irrigation) purposes. Unfortunately, environmental surveillance of antimicrobial-resistant Campylobacter infections is limited in South Africa (van den Honert et al. 2018). Therefore, environmental isolates are rarely tested for antimicrobial susceptibility. AR Campylobacter in the environment can be easily transferred to humans upon exposure (Larsson & Flach 2022). Possible means of river water contamination include broken sewer lines, inadequately treated wastewater effluent discharged into the rivers, and illegal disposal of faecal and household waste into rivers. All of these bear bacterial pathogens, including Campylobacter and antibiotics that end up in the rivers. In local rivers, antibiotic concentrations, which are sufficient to act as a selective pressure on the pathogens, have been observed (Vumazonke et al. 2020).

The higher prevalence of virulence genes noticed among the Campylobacter water isolates shows that the Campylobacter species isolated from water samples may be more virulent than capable of causing illness in humans. The genes cdtB and cadF are responsible for pathogenicity, while htrB and clpP genes play a role in Campylobacter survival in environmental waters with high oxygen concentrations (Otigbu et al. 2018). The high prevalence of virulence genes in chickens and pigs is also a serious public health concern. These isolates expressing virulence genes can be easily transmitted to humans through food and introduced into the water.

This review highlights the likelihood of transmission of pathogenic Campylobacter from water to humans and animals (Chukwu et al. 2019). Therefore, the One Health Approach for surveillance of antibiotic resistance in Campylobacter species, which includes assaying Campylobacter resistance in environmental water samples, is necessary for early detection of antibiotic resistance developments. South Africa established a National Action Plan to fight against Antimicrobial Resistance (DAFF and DOH 2018) as a component of the global One Health strategic plan against public threats and enhance the Global Antimicrobial Resistance Surveillance System. One of the six objectives is to strengthen knowledge and evidence through monitoring and research (2015). Effective implementation of this objective, among others, will assist in reducing the burden of AR infections, including Campylobacter, nationwide.

Few studies have been conducted on AR Campylobacter in South Africa, and this may be due to the fastidious nature of this bacteria. The other reason could be poor surveillance, which has resulted in underestimating Campylobacter infections in South Africa. There is a need to conduct more studies on AR Campylobacter, especially under the One Health Approach that considers the link between human, animal, and environmental health as an enhanced action against antimicrobial resistance.

This study highlights the high level of antibiotic resistance among Campylobacter spp. in water, animals, and humans; and a high level of multidrug resistance of Campylobacter isolates in human samples. In a country with an increasing number of immunocompromised individuals, a substantial number of people rely on rivers for domestic and agricultural purposes. However, there is insufficient data on environmental surveillance to detect early antibiotic resistance development, and there is cause for alarm. Research on antibiotic resistance among Campylobacter from other sources, outside clinical isolates, including animals and environmental reservoirs, should be prioritised. Also, the use of antibiotics in animals mainly requires excessive regulation. Furthermore, long-lasting measures to prevent contamination of water resources include treating wastewater effluent to the recommended standard before discharging into nearby rivers, controlled access of grazing animals to source waters, and stricter policies to prevent dumping of domestic waste in rivers will reduce antibiotic resistance infections in South Africa.

The publication received financial support from Water Research Commission (WRC), South Africa (Project Number: K5/2886).

This manuscript does not contain experiments using animals and does not contain human studies.

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.