Antimicrobial resistance screening and profiles: a glimpse from the South African perspective

The WHO predicts that globally deaths from antibiotic-resistant infections could increase to 10 million deaths annually by 2050 and that antimicrobial resistance (AMR) could force up to 24 million people into extreme poverty due to the combined effects of AMR on human health as well as food systems by 2030 (WHO 2019). The development of resistance in microorganisms against commonly used antibiotics can be exacerbated through the use of antibiotics in livestock farming, unregulated administration of antibiotics and patient default of antibiotic treatment. South Africa represents a paradox of antibiotic management similar to other developing countries, with both overuse and underuse (resulting from lack of access) of antibiotics (Duse 2011; Laxminarayan et al. 2013). A further concern is that wastewater reuse must be encouraged as a measure to provide alternative water sources to provide resilience to climate change (Adewumi et al. 2010) and can contribute towards antibiotic resistance through selective pressure that increases antibiotic resistance in native bacteria (Kraemer et al. 2019). There is also growing concern that environmental concentrations of antibiotics exert selective pressure on clinically relevant bacteria. In some African countries, inadequate wastewater treatment facilities may also contribute to the increased resistance profiles of the common pathogens (Momba et al. 2006). Irrigation with water containing antibiotic-resistant bacteria may lead to the uptake of these bacteria into the crops, leading to an additional source of exposure to antibiotic-resistant pathogens (Ruimy et al. 2010; Hirneisen et al. 2012; Drissner & Zürcher 2014; Nüesch-Inderbinen et al. 2015; Thanner et al. 2016). It is estimated that between 2000 and 2010, the antibiotic consumption in the world increased 35%, with Brazil, Russia, India, China and South Africa accounting for 76% of this increase, although their contribution in population increase was only 25% (Van Boeckel et al. 2014). This increase in global consumption was attributed both to more affordable antibiotics in low-income countries, as well as inadequate prescription or over-the-counter purchases (Van Boeckel et al. 2014). Amoxicillin, sulphamethoxazole/trimethoprim and ciprofloxacin contributed to 75% of the total oral antibiotic consumption in the African region, whilst beta-lactams were 70% of the consumed antibiotics in the Western Pacific region; amoxicillin, macrolides and quinolones were also the majorly consumed antibiotics in the Eastern Mediterranean as well as the European region. The trend observed was similar for the most consumed classes of antibiotics across the global region, with beta-lactams, macrolides and fluoroquinolones reflected, in the African region; however, sulphonamides were consumed more than macrolides (WHO 2018).

Escherishia coli is sometimes used as a sentinel for monitoring antimicrobial drug resistance in faecal bacteria because it is found more frequently in a wide range of hosts, acquires resistance easily and is a reliable indicator of resistance in other pathogenic bacteria (Tadesse et al. 2012). Resistance to one of the most widely used antibacterial drugs for the oral treatment of urinary tract infections caused by E. coli, fluoroquinolones, is very widespread (WHO 2014). The incorrect use of antimicrobial drugs, for instance in animal husbandry, supports the development and selection of resistant bacteria. Antibiotics in sewage, treated effluent, sewage sludge and water sediments allow for the selection of antibiotic resistance (Fouz et al. 2020). In addition, exposure to antibiotic-resistant bacteria may be through the irrigation of crops. Selection for antibiotic resistance is not confined to the human body or to hospitals, clinics and farms. Selection takes place anywhere an antibiotic is present, especially in natural environments, most notably sewage and surface water sediments, where antibiotics are likely to be coupled with high densities of various microorganisms. Large amounts of antibiotics and biocides end up in sewage sludge, making it a source for the development of antibiotic resistance (Fouz et al. 2009). When dewatered sludge is applied as fertiliser to agricultural land, there is a renewed risk of introducing both antibiotics and resistant strains into the food supply.

E. coli is normally effectively treated by the antibiotics ampicillin, cloxacillin, colistin sulphate, doxycycline, enrofloxacin, florfenicol, fosfomycin, gentamicin, kanamycin, nalidixic acid, penicillin, streptomycin, sulphamethoxazole, trimethoprim, sulphonamides and tetracycline (Hertz et al. 2014; Tekin et al. 2018; Madappa 2019).

Antibiotics are not only given to humans and animals for the treatment of infections. Certain antibiotics, when given in low, sub-therapeutic doses, are known to improve feed conversion efficiency (more output, such as muscle or milk, for a given amount of feed) and/or may promote greater growth, most likely by affecting gut flora.

Antibiotic resistance has shown up soon after their development resistance for some antibiotics such as linizolid and methicillin developed only 1–2 years after its development. An extensive evaluation of antibiotic resistance on a global scale was conducted by Laxminarayan et al. (2013). The complexities in low- to middle-income countries stem from a combination of overuse and underuse as previously mentioned of certain antibiotics. In addition, the unregulated and/or misinformed use of antibiotics results in continued mortalities in adolescents from treatable infections and increased resistance to certain classes of antibiotics that are more easily accessible. This, coupled with inadequate wastewater treatment and poor data collation contribute to greater challenge of tackling antibiotic resistance. Within these developing countries, these are the combined considerations in understanding AMR. Compounding this are the limited comparable methods applicable in low and highly resource countries, with the disc diffusion method being the most commonly applied and standard method in research for screening (Syal et al. 2017). The AMR in different water sources in this study were assessed within this context.

Crop irrigation is thought to be a potential source of exposure to antibiotic-resistant bacteria through the transfer from the water or sludge into crops (Thanner et al. 2016). The objectives of this study were to examine the antibiotic-resistance profiles of E. coli from rural water supplies used for agriculture, including passive wastewater treatment systems, river waters and agricultural farm dams from three agricultural locations in the Western Cape, South Africa. Surface water sediments and wastewater sludge provide a place where antibiotic-resistance genes are transferred to other bacteria.

Using a classical microbiology approach, the resistance profiles of E. coli species detected from water and underlying sediments were assessed for resistance to 14 common antibiotics. These antibiotics included ampicillin, colistin sulphate, cloxacillin, doxycycline, enrofloxacin, florfenicol, fosfomycin, gentamicin, kanamycin, nalidixic acid, penicillin, streptomycin, tetracycline and trimethoprim–sulphamethoxazole. Randomly selected E. coli isolates from the sediment and surface water of wastewater and surface water were tested for resistance.

The Western Cape has a warm temperate Mediterranean climate, with rainfall predominating during austral winter and early spring. There is a history of concern for water shortages and compromised water quality, and hence, the interest in reuse of wastewater with E. coli levels and nutrient concentrations being the major considerations. As part of independent research being carried out in projects looking at water quality issues in the Western Cape, an opportunity arose to investigate three agricultural sites in the Western Cape, South Africa to provide an indication of the occurrence of antibiotic-resistant bacteria in the water, used for the irrigation of crops. The sites included the Dwars river site in the upper reaches of the Berg River catchment, the Touws River site in the Gouritz catchment and the Brandwacht passive wastewater treatment works (WWTWs) also in the Gouritz Water Management Area (WMA) (Figure 1(1-1)–1(1-3)). The three sites offer a selection where agricultural activity takes place and could contribute to the exposure to antibiotic-resistant bacteria, through the irrigation of crops with water containing antibiotic-resistant bacteria.

The Berg River has been a concern for agricultural use regarding the water quality over many years. The upper Berg River catchment is a mountainous sub-catchment of the Berg River catchment in the Western Cape and is bound by the Franschhoek and Drakenstein mountains to the south and south-west of the catchment. The main tributaries of the Berg River in this area are the Franschhoek, Wemmershhoek and Dwars Rivers and are used to irrigate crops. Samples were collected from the Dwaars tributary.

The Touws River is part of the Coastal Belt in the Gouritz catchment which includes the Gouritz/Goukou/Duiwenhoks catchment. Land use is a mix of residential, natural vegetation and intense dairy farming with some fresh produce agriculture.

The Brandwacht topography comprises land that drains into the Brandwacht WWTWs which serves a small population of 1,470 individuals or 398 households. The wastewater treatment system is a small one making use of seven earth ponds designed to allow gravity to flow from one pond to the next without electricity. Depending on the need, effluent may be made available for agricultural activities.

Surface water (1 l in sterile bottles) and sediment samples (100 g) were collected from different sites of the Touws river and farm dams, the Upper Berg river (surface water only), and the seven ponds of the Brandwacht WWTWs. Sediment samples from the centre of farm dams were collected in the benthic zone using a canoe or rowing boat (depending on the availability) and a Van Veen grab sampler. Samples were also collected from three sites in the littoral zone evenly distributed around the dam using a metal scoop, before combining the sediment samples into a composite sample in zip-lock plastic bags. In Brandwacht WWTW, 1 l grab samples were collected from the outlets of Pond 1–7 and from the sediment at these sites. Nitrile gloves were used for sample handling and to prevent contamination. All samples were collected using bailers, transferred to sterile water bottles and kept on ice, in the dark during transportation to the laboratory before analysis.

To determine the AMR profiles of E. coli isolates and to understand the role of WWTWs, surface water sources including farm dams and rivers in the development of AMR, samples from three different sites were collected in the Western Cape in South Africa. The antimicrobial susceptibility of E. coli isolates from surface waters, wastewater effluents and sediments were determined, using antibiotics used in the control of E. coli infections.

Using E. coli as an indicator organism, enumeration was conducted using the Colilert 18 method (IDEXX, South Africa) according to the manufacturer's specifications. Briefly, Colilert nutrient powder capsules were dissolved in 100 ml volumes of surface water samples. For sediment samples, 0.1 g samples of sediment were dissolved in volumes of 100 ml sterile tap water, sealed in 49 well Quanti-Trays and incubated at 35 °C, over an 18-h period. After incubation, samples tray wells were analysed based on colour changes. A change in medium from colourless to yellow indicated the presence of coliforms, whilst fluorescence of the yellow wells under UV light indicated the presence of E. coli.

E. coli bacteria isolated from sediment and surface water samples were tested for resistance against various antibiotics. The E. coli originally cultured using the Colilert 18 method were extracted by removing 0.1 ml of medium from fluorescent wells (positive wells) using a needle and syringe. Spread plate antibiotic disc diffusion assays were conducted with the cell suspensions and monitored over a 48-h incubation period at 35 °C on nutrient agar (Merck, South Africa) plates. At least ten samples containing E. coli were tested in all three sites. Overall, 14 antibiotics were tested, namely, ampicillin, streptomycin, florfenicol, trimethoprim, colistin, enroflaxin, doxycycline, fosfomycin, nalidixic acid, tetramycin, sulphamethoxazole, gentamicin and kanamycin (Oxoid, South Africa). Zones of inhibition were recorded as either present (susceptible) or absent (resistant), according to the EUCAST method, with zones of inhibition typically indicated by no growth, when held up about 30 cm from the naked eye (Matuschek et al. 2014). The observed zones were generally greater than 5 mm, although only presence/absence criteria were used in this study.

Microsoft Excel was used to create figures and conduct trend analysis.

The majority of E. coli isolates were resistant to ampicillin, penicillin, cloxacillin, sulphamethoxazole and trimethoprim (Table 1; Figure 2(a)–2(c)), whereas streptomycin, kanamycin, enrofloxacin, doxycycline and gentamycin remained effective against E. coli in the majority of isolates tested (Figure 2(a)–2(c)), ranging between 70 and 100% sensitivity.

The resistance profile of E. coli isolated from sediments differed from those isolated from surface water, with 90% of the surface water isolates being resistant to ampicillin. Isolates from the sediment were less resistant (40%) to ampicillin, whereas nearly all the isolates from the pond and sediment were resistant to sulphamethoxazole. In farm dams, only the E. coli isolates found in sediment were found to be resistant to fluorquinolone or fluorifenicol.

There is an indication of antibiotic resistance retained within the sediments of the Touws river, whilst this is only observed for sulphamethoxazole in the Brandwacht WWTWs.

In the wastewater ponds, the number of antibiotics that the E. coli were resistant to, increased in successive maturation ponds (Figure 3(b)), whereas in the E. coli isolated from sediments, the number of antibiotic they were resistant to started at 9 in pond 1 and only 5 in pond 7 (Figure 3(c)).

Penicillin and sulphonamides are examples of antibiotics which are known for overuse and persistence in the environment (Lobanovska & Pilla 2017). In the water and sediment samples, the greatest resistance across the sites is towards sulphamethoxazole, trimethoprim and penicillin. Ampicillin, which is a later generation of the penicillin classes and initially caused microbial sensitivity (Lobanovska & Pilla 2017), is now one of the antibiotics that E. coli isolates from the different sites were widely resistant to.

Sulphonamides such as sulphamethoxazole and/or the sulphamethoxazole–trimethoprim combination have been recorded as the more persistent antibiotics in the environment (Grenni et al. 2019), with an approximate 60-day degradation, with synergistic actions of the metabolite with other antibiotics resulting in a longer degradation time. This has led to the development of resistance genes against this antibiotic. In South Africa, HIV-positive patients are treated with low doses of this combination as preventative to opportunistic infections (Kaplan et al. 2009); however, 100% of E. coli isolates were resistant to sulphamethoxazole, highlighting its relevance in the South African context. This finding is supported by earlier work by Nyamukamba et al. (2019) in the Vaal triangle area, where sulphamethoxazole was detected in higher quantities than other antibiotics tested, which were interestingly below the limit of detection. This is interesting as this is one of the more prevalent antibiotics consumed in the African region in comparison to other global regions.

In surface water isolates of E. coli resistance to all of the antibiotics tested was detected, whereas in the maturation pond isolates, the E. coli isolates remained sensitive to streptomycin, enroflaxin and the majority of isolates remained sensitive to gentamycin and kanamycin (Figure 2(a)). In the agricultural dam water and sediment, more resistance was found in sediment isolates (Figure 2(b)).

The advent of passive wastewater treatment in developing countries offers a low cost alternative to the challenge of failing WWTWs system, which however consume large amounts of power, proving to be a challenge in developed countries as well (Hossain et al. 2010). Waste stabilisation ponds are a technology used prolifically by South African municipalities due to their simplicity, economy and reliability (Mambo et al. 2014a). In only very limited locations in South Africa are alternative wastewater treatment processes being trialled, such as constructed wetlands (Mthembu et al. 2013) and Algal Integrated Wastewater Pond Systems (AIWPS) or Integrated Algae Pond Systems (IAPS) as a municipal sewage treatment technology (Mambo et al. 2014b). When assessing compliance of this treatment, the effluent produced required additional tertiary treatment to meet the required standards in coliform and total suspended solids (Mambo et al. 2014a). In fact, it is recommended to decision-makers to engage in mitigating risks posed by poorly performing WWTWs by investing in in-stream biotechnologies concurrent to investing in the refurbishment of WWTWs (Mitchell et al. 2014).

The comparison of antibiotic-resistance profiles from samples of conventional and this treatment alternative provide useful information on the treatment efficiency and how AMR profiles may change in a passive treatment system. The use of algae to utilise pollutants for nutritional benefit in wastewater provides an environmentally friendly alternative to the conventional wastewater treatments, with higher retention times (Wang et al. 2010), which may explain the final reduced resistance profile in the sediment of pond 7 in Brandwacht WWTW, which uses gravitational energy for effluent flow into treatment ponds (Figure 2(c)). In South Africa, a majority of the conventional WWTWs are dysfunctional and not treating water to the required standards (Momba et al. 2006). Coupled with the complexities of increased populations, climate change concerns, overuse and underuse of antibiotics; the challenge in screening and effective control of AMR in South Africa is critical.

The use of passive treatment in wastewater treatment, where residence times are significantly longer than in conventional wastewater treatment systems, may allow the reduction of antibiotic resistance, as shown by the lower numbers of antibiotics that the isolated E. coli were resistant to compared with surface water isolates (3 versus 6 antibiotics, respectively; Table 1). What is not known is whether the passive systems perform better than conventional WWTWs in reducing the presence of antibiotic-resistant bacteria. In the passive wastewater treatment system, the number of antibiotics that the E. coli isolated from the water samples were resistant to, increased in successive maturation ponds, whereas the reverse was seen in the E. coli isolated from sediment samples (Figure 3(b) and 3(c)).

These results illustrate the importance of developing a better understanding of antibiotic resistance in agriculture and wastewater scenarios to ensure remedial measures take place where the greatest benefit can be realised in countries with limited financial resources. Future research is needed to assess the contribution of conventional WWTWs to the growing antibiotic resistance of bacterial isolates and to establish the potential of passive treatment systems.

In South Africa (and most developing countries), pathology laboratories and water facility laboratories predominantly make use of the culture method to screen for AMR. This method, among many more sophisticated techniques, remains the gold standard method from a cost perspective. Among the methods to test for antibiotic resistance, there is still no standardised cohesive guide that is widely practised for reporting and comparable analysis in countries with limited resources, thereby making the classical microbiological culture technique the more feasible and standardised approach (Khan et al. 2019). The WHO has implemented an AMR surveillance project which recommends cultivation methods in support of the concept that a simplified, integrated, trans-sectoral surveillance system of bacterial resistance to antibiotics could be implemented on a global basis, the so-called Tricycle project (GLASS 2020).

The findings in this study indicate that the theory of wastewater treatment systems being a hub for horizontal antibiotic-resistance acquisition in pathogens is possible based on the increased resistance profiles in the sediment samples of conventional wastewater plants, where increased resistance has been linked to wastewater treatment (Manaia et al. 2018). This is observed in the surface water assessments of all the sites and in the sediment of the passive wastewater treatment plant. The use of passive wastewater treatment to manage persistence of sulphamethoxazole and other persistent antibiotics cannot be concluded within the limits of this research, especially considering that a 100% of isolates from each of the seven ponds were resistant to sulphamethoxazole. The sites indicated resistance to the largely used and persistent antibiotics, with increased resistance to fluoroquinolones in the animal influenced samples, this has been reported in other parts of the world where there is intense animal husbandry (Tang et al. 2017; Schulz et al. 2019), which may be due to the unaltered excretion of these antibiotics by the animals, into waters where inadequate treatment exists to effectively degrade or manage the impacts of genetic mutations in exposed bacteria. The comparison of the three sites indicates a global trend in resistance profile influence. Overall, the South African perspective is fairly bleak and complexed with less than 50% of the wastewater treatment systems in South Africa meeting national and international water quality standards for wastewater treatment (Mthembu et al. 2013; Mitchell et al. 2014). These findings are proof that South Africa's wastewater treatment systems are inadequate to meet the effluent required standards. This has resulted in the urgent need for the development and implementation of innovative systems to resolve the wastewater treatment constraints. This research, among the existing body of the literature, strengthens the call for a more stringent management approach of this challenge, with passive treatment showing a better sensitivity profile.

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