White-Nosed Coatis (Nasua narica) Are a Potential Reservoir of Trypanosoma cruzi and Other Potentially Zoonotic Pathogens in Monteverde, Costa Rica

Trypanosoma cruzi, a vector-borne protozoan parasite, infects more than 180 mammal species and can be a significant cause of morbidity and mortality for humans, domestic dogs, and some exotic animals (Guedes et al., 2007; Noireau et al., 2009). Although T. cruzi is estimated to infect nearly 10 million people worldwide and can result in chronic and fatal myocardial degeneration, Chagas disease is widely considered to be a neglected tropical disease (Hotez et al., 2008).

In Monteverde, Costa Rica, the reduviid bug Triatoma dimidiata can transmit T. cruzi with prevalence rates of 42% and 83% reported in domestic and sylvatic habitats, respectively (Zuriaga et al., 2012). This popular ecotourism destination receives more than 200,000 predominantly international tourists annually (Ferguson, 2010). Because the chronic form of Chagas disease does not develop for 10–20 yr, visitors could become infected and, even in the absence of an appropriate vector, potentially transmit the parasite via unscreened blood donations.

The white-nosed coati (Nasua narica) is a peridomestic omnivore with a wide range in Central and South America (Glatston, 1994). The gregarious and inquisitive nature of white-nosed coatis brings them into contact with humans and domestic animals in semiurban areas. Other procyonids, such as ring-tailed coatis (Nasua nasua) and raccoons (Procyon lotor), and some rodents are reservoirs of T. cruzi (Charles et al., 2013; Herrera et al., 2008). Trypanosoma cruzi has also been cultured from the blood and anal glands of the opossum (Didelphis marsupialis) (UrdanetaMorales and Nironi, 1996). The reservoir status of white-nosed coatis (hereafter, coatis) is unknown. To elucidate the role of coatis in the epidemiology of Chagas disease in Costa Rica, we conducted a field investigation to determine T. cruzi prevalence in coatis. Additionally, we tested coatis for a suite of selected pathogens, including Babesia, Mycoplasma, and Baylisascaris.

Between 10 July and 8 August, 2011, 20 free-ranging coatis were live-trapped with the use of Tomahawk™ (Tomahawk Live Trap Co., Tomahawk, Wisconsin, USA) and Havahart™ traps (Havahart Products, Litiz, Pennsylvania, USA) at three sites in Monteverde (10°18′36.039″N, −84°49′11.5968″W; 10°18′17.6472″N, −84°48′54.0822″W; and 10°18′10.2162″N, −84°48′42.8394″W). Sites were nonrandomly selected near homes, in semiurban parks, and at tourist destinations. Traps were baited with local bananas, mangos, and eggs and checked every 3 hr. Coatis were weighed and chemically immobilized with an intramuscular injection of acepromazine (Alfasan International, Woerden, Holland; 0.1 mg/kg) and ketamine (Bremer-Pharma, Warburg, Germany; 20 mg/kg).

Blood samples (between 0.4 and 0.8% body weight) were collected from the jugular vein, and fecal samples were collected rectally. Physiologic saline solution (Pisa Agropecuaria, Guadalajara, Mexico; 15 mL/kg) was administered intravenously or subcutaneously as fluid replacement therapy during anesthesia. Respiratory rate, heart rate, and body temperature were monitored throughout anesthesia. Animals were allowed to recover in traps until sternal and responsive (3–5 hr) before release at their original capture site.

Thin blood smears for visualizing T. cruzi and Babesia by light microscopy were fixed in methanol and stained with Wright-Giemsa stain (Blevins et al., 2008). Remaining whole blood was stored at −18 C until testing. Buffy coats and anal gland secretions were inoculated into liver-infusion tryptose culture media and maintained at room temperature (Brown et al., 2010). Plasma was tested for anti–T. cruzi antibodies with the Chagas-Stat-Pak™ (Chembio Diagnostics Inc., Medford, New York, USA) according to manufacturer's instructions. DNA (extracted from blood samples) and culture isolates were tested for T. cruzi with the use of nested polymerase chain reaction (PCR) with primers specific to T. cruzi (D75/D76, in the primary reaction, and D71/ D72 in a secondary reaction), and sequenced. Blood samples were also screened for Babesia, Mycoplasma, Anaplasma, Candidatus Neoehrlichia lotoris, Ehrlichia, Bartonella, and apicomplexan parasites (Sarcocystis, Toxoplasma, and Eimeria) by PCR with the use of published primers: 1) Babesia, primers 3.1/5.1 and RLBH-F/RLBH-R; 2) Mycoplasma spp., HaembartF and HaembartR; 3) Anaplasma and Candidatus Neoehrlichia lotoris, ECC/ECB and GE9f/GAUR1; 4) Ehrlichia, Dsb-330/Dsb-728; 5) Bartonella, 325s/100as; and 6) Apicomplexan parasites, 5.1/B and 18S1H/19S9L (Jensen et al., 2001; Yabsley et al., 2005; Labruna et al., 2007; Cadenas et al., 2008; Yabsley et al., 2008, 2009b). Sheather's sugar flotation was performed on all feces to detect Baylisascaris.

All coatis were negative for T. cruzi antibodies and cultures of anal gland secretions were negative. However, T. cruzi was cultured from the blood of one coati. This coati and six others (35%) were PCR positive for T. cruzi (Table 1). In addition, sequences from a blood sample and the cultured trypanosome were identical to each other and were most similar to T. cruzi IV strains (96.7%), 119 of 123 base pairs (bp) to strain 92122102r (GenBank accession AY367114) from a raccoon from the United States and 95.9% similar to strain 4166 (GenBank AF288664) from a reduviid bug, Rhodinus brethesi from Brazil. All 20 coatis were PCR positive for Babesia and Mycoplasma. Sequencing analysis from two coatis revealed identical amplicons for each respective parasite. Babesia sequence results were most similar to those from raccoons in the United States and Japan (570/573 bp, 99.5%; GenBank DQ028958 and AB251608, respectively) and from Florida, USA pumas (Puma concolor; 99.5%; GenBank DQ329138; Fig. 1). Mycoplasma sequence results were most similar to Mycoplasma haemolamae (149/169 bp, 82.8%), Mycoplasma haemominutum (82.2%) and Mycoplasma suis (80.5%). Coatis were negative for other pathogens, including Baylisascaris.

The high prevalence of T. cruzi in our study animals was similar to prevalences detected in ring-tailed coatis (Alves et al., 2011). Negative blood smear results were expected; PCR and culture are often more sensitive in chronically infected reservoirs with low parasitemias. The Chagas-Stat-Pak™ assay, which works well for testing raccoons (Yabsley et al., 2009a), failed to detect antibody in coatis. Because this assay works variably in other wildlife species (Charles et al., 2013), use of another assay (enzyme-linked immunosorbent assay, or immunoflourescent antibody test) requiring anti-coati immunoglobulin might have increased accuracy for estimating population exposure. Our failure to detect T. cruzi in anal gland secretions could be due to a true absence of infection or the difficulty of collecting samples. The morphology of coati anal glands is unusual and consists of four to five slit-like invaginations with a fold of skin acting as a diverticulum (Mivart, 1885).

We detected a 100% prevalence of Babesia infection, similar to surveys of raccoons in which prevalence can be >80% (Birkenheuer et al., 2008). Despite the high prevalence, little is known about the natural history, including the vector(s), of procyonid Babesia. We also detected a high prevalence of hemotropic Mycoplasma in coatis, but the clinical or epidemiologic relevance of our finding is unknown, as there are no published studies of hemotropic Mycoplasma in procyonids.

Our data, as well as species biology, illustrate that Monteverde coatis meet several suggested reservoir criteria for T. cruzi (Ashford, 1996). We detected a high T. cruzi prevalence in asymptomatically infected coatis. Moreover, coatis are a gregarious, long-lived species capable of surviving through nontransmission seasons. In addition to peridomestic control of the reduviid vectors, we recommend that Monteverde residents and ecotourism operators discourage coatis from foraging and resting near human dwellings, and that further research be conducted on vectors and habitats. Furthermore, the detection of Mycoplasma and Babesia in coatis suggests that coatis may harbor other pathogens that present health risks to coatis, domestic animals, or people.

This research was permitted by Sistema Nacional de Areas de Conservacion permit 044–2011–ACAT and Convention on International Trade in Endangered Species permit 11Cr000011/SJ. Funding was provided by a University of California at Davis School of Veterinary Medicine Students Training in Advanced Research grant and International Student Programs. Special thanks to Adair McNear and Bob and Marci Lawson for assistance in the field. The Monteverde Butterfly Garden and the Monteverde Conversation League made sampling sites available. Robin Houston, Joy Snipes, and Chris Steckley provided indispensable help in the laboratory, and Janine Kasper enhanced our understanding of coati anal gland anatomy. We also thank Ben Hirsch, Sonia Hernandez, and Terra Kelly for significant help with project planning.