Zoonotic diseases are a major concern in today's globalized world. Sri Lanka is a biological hotspot (Myers et al. 2000) with moderate to high risk of transmission of zoonotic disease from wildlife to humans (Jones et al. 2008). All primate species in Sri Lanka have decreasing population trends due to anthropogenic disturbance and increasing human-primate conflicts. However, at temples and other holy places, primates can roam relatively undisturbed where they may consume human waste and food offerings, potentially exposing themselves to foodborne pathogens (Nahallage and Huffman 2013).

Campylobacter spp. and Salmonella spp. can colonize or infect many species, including primates that may become asymptomatic carriers or develop clinical disease (Nizeyi et al. 2001; Ngotho et al. 2006). The bacteria can be spread by direct contact between animals and humans or indirectly via the environment. Most infected humans recover without specific treatment. Antimicrobials are warranted for patients with severe, prolonged enteritis, or life-threatening conditions.

The presence of these bacteria and antibiotic resistance has not been reported in Sri Lankan nonhuman primates. The aim of our study was to investigate the occurrence of Campylobacter spp. and Salmonella spp. in toque macaques (Macaca sinica) and tufted gray langurs (Semnopithecus priam) in Sri Lanka, to assess the antibiotic susceptibility of detected isolates, and to genetically subtype the isolates.

During March–May 2015, fresh fecal pat samples were collected from endemic, endangered toque macaques and near-threatened tufted gray langurs in five locations in Sri Lanka (Fig. 1 and Table 1). The primates were resident at temples or archeological sites frequented by local and international visitors. To minimize risk of double sampling, no specimens within 2 m of each other were sampled, unless they were directly observed defecations from different individuals. The samples were collected in duplicates using Copan FecalSwab^™ (Copan Diagnostics Inc., Murrieta, California, USA) and immediately stored in a cooler box with ice packs (5±3 C), followed by refrigeration for up to 2 days, until analyses at the University of Kelaniya (Ragama, Sri Lanka; Fig. 1).

Figure1

Five sampling locations (circles) in Sri Lanka where fecal pat samples were collected in March–May 2015 for isolation of Salmonella spp. and Campylobacter spp. in free-ranging tufted gray langurs (Semnopithecus priam) and toque macaques (Macaca sinica). Laboratory analyses were conducted at the University of Kelaniya in Ragama, near the capital Colombo (squares).

Table 1

Distribution and number (n) of fecal pat samples collected in March–May 2015 from tufted gray langurs (Semnopithecus priam) and toque macaques (Macaca sinica) at five locations in Sri Lanka. Salmonella enterica subsp. enterica serovar Virchow, Campylobacter jejuni, and Campylobacter coli were found in fecal pat samples from toque macaques but not from tufted gray langurs.

Culturing of Campylobacter spp. and Salmonella spp. was performed according to the International Organization for Standardization (ISO 10272, 2017) and Nordic Committee on Food Analysis (NMKL no. 71, 1999), respectively. Suspected isolates were stored in serum broth with 15% glycerol at –20 C for later transport on dry ice to the National Veterinary Institute (SVA, Uppsala, Sweden). Species identification of Campylobacter spp. was performed with matrix-associated laser desorption/ionization-time of flight mass spectrometry. Suspected Salmonella isolates were identified by serotyping with antisera to determine species and serovar according to the Kauffman-White classification system (Popoff and Minor 1997). Susceptibility to selected antimicrobial substances was assessed with VetMIC^™ panel analysis systems (SVA), determining the antimicrobial minimum inhibitory concentration. Multidrug resistance was defined as resistance to three or more antibiotic classes. For example, resistance to ciprofloxacin, enrofloxacin, and nalidixic acid was considered resistance to one antibiotic class (fluoroquinolones). Genetic subtyping was performed by pulsed field gel electrophoresis (PFGE). Computer-assisted identification (BioNumerics version 7.5, Applied Maths, Sint-Martens-Latem, Belgium) was used to construct a dendrogram of Campylobacter isolates (Fig. 2). A dendrogram was constructed based on PFGE band pattern similarities according to the standardized Campynet protocol (On et al. 2000). Sequence types (STs) and clonal complexes are based on multilocus sequence typing (MLST) using the standardized PubMLST Campylobacter method (Dingle et al. 2001).

Figure2

Dendrogram of Campylobacter spp. isolates originating from fecal pat samples collected in March–May 2015 from toque macaques (Macaca sinica) in Sri Lanka. PFGE=pulsed field gel electrophoresis.

Out of 58 toque macaque samples, Salmonella enterica subsp. enterica serovar Virchow (Salmonella Virchow) was isolated in two (3%), C. jejuni in 10 (17%), and C. coli in four (7%). The C. jejuni isolates were classified into four clonal complexes: ST-21, ST-52, ST-362, and ST-404, and divided in six unique PFGE profiles (Fig. 2). All four C. coli isolates and both Salmonella Virchow isolates originated from the same location (Table 1), and the isolates were indistinguishable from each other within each species, based on PFGE band patterns (Fig. 2). Two C. jejuni isolates (ST-21, PFGE profile 2) were resistant to ciprofloxacin, tetracycline, and nalidixic acid. The third (ST-52) was resistant to ampicillin, ciprofloxacin, tetracycline, and nalidixic acid and was therefore considered multidrug resistant. The four C. coli isolates (ST 828) were resistant only to ampicillin, and the two Salmonella Virchow isolates were susceptible to all antibiotics tested (Table 2). Neither Campylobacter spp. nor Salmonella spp. were detected in the 40 samples from tufted gray langurs (Table 1).

Table 2

Distribution of susceptibility of isolates of Salmonella enterica subsp. enterica serovar Virchow from fecal pat samples from toque macaques (Macaca sinica) in Sri Lanka collected in March–May 2015. Isolates of Salmonella Virchow, Campylobacter jejuni, and Campylobacter coli were tested for susceptibility of selected antimicrobial substances. Values for susceptibility were obtained from the European Committee on Antimicrobial Susceptibility Testing (2015).

Nonhuman primates can be asymptomatic carriers of these bacteria, but fatal illness may develop (Nizeyi et al. 2001; Ngotho et al. 2006). This is a reason for concern regarding the conservation of free-ranging primate populations. In 58 fecal samples from toque macaques, C. jejuni, C. coli, and Salmonella Virchow were detected in 17, 3, and 3% of the samples, respectively, whereas none of the pathogens were detected in 40 samples from tufted gray langurs. Varying prevalence of C. jejuni has been documented in Peru (21%) and China (2%) in wild monkeys and in Uganda (19%) in mountain gorillas (Gorilla gorilla beringei; Tresierra-Ayala and Fernandez 1997; Nizeyi et al. 2001; Zeng et al. 2016).

Possible transmission routes of the bacteria to the primates in our study could have been human food waste, contaminated water, soil, feces, or direct contact with other species or conspecifics. Poultry meat is an important vehicle for both Salmonella spp. and Campylobacter spp. in Sri Lanka (Coorey and Perera 2007; Jayatilleke et al. 2015), where an incidence of Campylobacter greater than 70% in broiler flocks has been reported (Kottawatta et al. 2007). In Brazil, four subtypes of C. jejuni that were found among humans were also found in captive marmosets (Callitrix spp.) that were given chicken (Gallus domesticus) as part of their diet (Scarcelli et al. 2005). Contaminated water is another possible route, because Campylobacter and Salmonella have been isolated from both surface and bottled water (Mannapperuma et al. 2013; Kuhn et al. 2017). Salmonella form a large group of bacteria with varying pathogenicity in different animal species depending on serovar (Ohl and Miller 2001). Salmonella Virchow has been isolated from both humans and other animals (Salisbury et al. 2011). For determining transmission routes and identifying sources of infection, bacterial characterization below the species level, so-called subtyping, such as PFGE and MLST, is useful (Fig. 2). The sequence types identified in this study have been isolated from humans with gastrointestinal disease and chickens in different parts of the world (Dingle et al. 2001; Scarcelli et al. 2005; Islam et al. 2009).

The study could have been strengthened by sampling of more individuals and additional sampling sites, which possibly could have detected the bacteria in both species of nonhuman primates. However, langurs are frugivorous foregut fermenters, with high levels of bacteriolytic lysozymes in the stomach (Stewart et al. 1987), whereas macaques are hindgut fermenters. The differences between the two species' digestive systems may influence bacterial colonization. Toque macaques, being less selective in their foraging, may be subjected to a higher risk of food borne transmission.

The emergence of resistant bacteria has been strongly linked with the use of antibiotics in animal production (Aarestrup and Engberg 2001). Despite national guidelines on the use of antimicrobials in Sri Lanka (Sri Lanka College of Microbiologists 2014), the use of antibiotics as growth promoters in animal production and aquaculture is still practiced, leading to increasing problems with antibiotic resistance (Patabendige et al. 2011). To the best of our knowledge the primates in this study had never been treated with antibiotics. Despite this, 30% (3/10) C. jejuni isolates showed resistance to both quinolones (ciprofloxacin and nalidixic acid) and tetracycline. In comparison, in human Campylobacter isolates in Sri Lanka, 38% were resistant to ciprofloxacin and 69% to nalidixic acid (Cooray and Perera 2007). Animals living in proximity to human settlements tend to harbor a higher number of resistant bacteria (Allen et al. 2010). In conclusion, the presence of zoonotic, multidrug-resistant bacteria in toque macaques may affect conservation of endangered primates and public health in Sri Lanka.

We thank the staff at the Department of Medical Microbiology, Faculty of Medicine, University of Kelaniya, Sri Lanka, and at the Department of Bacteriology at the National Veterinary Institute in Uppsala, Sweden, for support. We also thank Prof. Chamalie A. D. Nahallage, University of Sri Jayawardenapura, Sri Lanka, for sharing knowledge about Sri Lanka's primates as well as Shantha Waidyarathna for field assistance. We thank our generous funders: the Swedish Research Council (2017-05479) and its Swedish Research Links program (2013-6722), Michael Forsgren's Foundation, the Faculty of Veterinary Medicine and Animal Sciences' Scholarship Fund at the Swedish University of Agricultural Sciences, the Swedish International Development Cooperation Agency Minor Field Studies, the Zebra Foundation for Veterinary Zoological Education, UK, and Royal Canin, Sweden. We also thank Kruuse AB, Nordic Biolabs, and ANL Products in Sweden for providing supplies.