Outcomes of Trypanosoma cruzi and Trypanosoma evansi infections on health of Southern coati (Nasua nasua), crab-eating fox (Cerdocyon thous), and ocelot (Leopardus pardalis) in the Brazilian Pantanal

The occurrence of Trypanosoma spp. in wild carnivore populations has been intensively investigated during the last decades. However, the impact of these parasites on the health of free-living infected animals has been largely neglected. The Pantanal biome is the world’s largest seasonal wetland, harboring a great diversity of species and habitats. This includes 174 species of mammals, of which 20 belong to the order Carnivora. The present study aimed to investigate the effect of Trypanosoma evansi and Trypanosoma cruzi infections and coinfections on the health of the most abundant carnivores in the Pantanal: coati (Nasua nasua), crab-eating fox (Cerdocyon thous), and ocelot (Leopardus pardalis). We captured 39 coatis, 48 crab-eating foxes, and 19 ocelots. Diagnostic tests showed T. cruzi infection in 7 crab-eating foxes and 5 coatis. Additionally, 7 crab-eating foxes, 10 coatis, and 12 ocelots were positive for T. evansi. We observed coinfections in 9 crab-eating foxes, 8 coatis, and 2 ocelots. This is the first report of T. evansi and T. cruzi infection on the health of free-living ocelots and crab-eating foxes. We showed that single T. evansi or T. cruzi infection, as well as coinfection, caused some degree of anemia in all animals, as well as an indirect negative effect on body condition in coatis and crab-eating foxes via anemia indicators and immune investment, respectively. Furthermore, the vigorous immune investment observed in sampled coatis, crab-eating foxes and ocelots infected by T. evansi, T. cruzi and coinfected can be highly harmful to their health. Overall, our results indicate that single and combined infection with T. evansi and T. cruzi represent a severe risk to the health of wild carnivores in the Pantanal region.

Introduction them, three species are considered abundant in the Pantanal: the southern coati, the crab-eating fox, and the ocelot [25].

Sample collection
Carnivores were sampled from November 2015 to October 2016. We used 50 Box-traps (90 × 45 × 50; EquiposFauna 1 ) baited with bacon and tinned sardines to capture the target species. Once trapped, animals were sedated with an intramuscular injection of Zoletil 50 (containing tiletamine hydrochloride and zolazepan hydrochloride; Virbac) respecting the dosages currently recommended for each species, and marked with subcutaneous transponders (AnimalTag 1 ). Body condition (body size and mass) were recorded. Blood (~4 mL) was collected from the jugular vein, placed in tubes with and without ethylenediamine tetraacetic acid (EDTA), and stored in cool boxes until laboratory procedures. The animals were released at the capture site after recovery from anesthesia.

Ethical approval
All field procedures were conducted in accordance with a license granted by the Biodiversity Information and Authorization System of the Chico Mendes Institute for Biodiversity Conservation (license number 49662-5). The present study was approved by the Ethics Committee for Animal Use of Dom Bosco Catholic University, Campo Grande, MS (license number 19/ 2015).

Health parameters
The health of carnivores was inferred, mainly, by means of blood parameters. Packed cell volume (PCV), red blood cell counts (RBC), and white blood cell counts (WBC) were measured up to 8 h after blood collection in Neubauer chambers, as described by Voigt [26]. Mean corpuscular volume (MCV) was calculated based on the RBC and PCV values. The immunoglobulin concentration (IgG) was determined by titration with the indirect fluorescent antibody test (IFAT) [27,28] and by optical density using enzyme-linked immunosorbent assay (ELISA) [29]. Leukocyte (eosinophils, lymphocytes, monocytes, and neutrophils) counts were performed using blood smears fixed with methanol and stained with Giemsa [30].
We evaluated the health condition of sampled carnivores in terms of: (a) PCV, RBC, and MCV as anemia indicators; (b) monocyte and neutrophil counts as indicators of infection responses; and (c) lymphocyte counts and IgG concentration as indicators of immune investment [30].

Diagnosis of T. evansi and T. cruzi infection
Infections with T. evansi and T. cruzi were assessed by parasitological, molecular and serological tests. The parasitological test for T. evansi used the microhematocrit centrifuge technique (MHCT) according to Woo [31]. The absence of kinetoplast in buffy coat smears confirms T. evansi. For T. cruzi, the test was based on hemoculture by inoculating 300 μL of blood in Novy McNeal Nicole (NNN) medium with liver infusion tryptose (LIT), in duplicate. Hemoculture tubes were incubated at 27˚C for 30 days and monitored once a week.
Molecular detection of Trypanosoma spp. infection was performed by nested polymerase chain reaction (nPCR). Genomic DNA was extracted from 200 μL of blood with EDTA using the QIAamp Blood DNA Mini Kit (Qiagen) according to the manufacturer's instructions. Total DNA was diluted with 50 μL elution buffer and stored at -20˚C until molecular diagnosis. We used as a target a variable region of the trypanosome 18S rRNA gene (600 bp), with external primers TRY927F and TRY927R, and internal primers SSU561F and SSU561R, according to Smith et al. [32]. TBR1 and TBR2 primers were applied to positive 18S rRNA samples to amplify a sequence of mini-chromosome satellite DNA for T. evansi, according to Masiga et al. [33]. Furthermore, D71 and D72 primers were used to amplify a conserved sequence of the large subunit of the ribosomal DNA gene (24Sα rDNA) in T. cruzi, according to Souto and Zingales [34]. Each reaction included sterile distilled water instead of DNA as negative control, and positive control samples from T. cruzi and T. evansi strains. PCR products were visualized in 2% agarose gel after ethidium bromide staining under ultraviolet light.
Serological tests were used to detect anti-T. evansi IgG antibodies in crab-eating foxes and ocelots by IFAT using a commercial fluorescein-conjugated antibody against dogs and cats IgGs, respectively. The cut-off value for IFAT was 1:40 [27]. There is presently no fluoresceinconjugated antibody against coatis' IgGs. To detect anti-T. cruzi IgG antibodies, we used IFAT (fluorescein-conjugated antibody against dogs and cats IgGs) and ELISA (fluorescein-conjugated antibody against raccoon's IgGs), as described by Rocha et al. [28] and Alves et al. [29], respectively. The cut-off value for ELISA was defined as the mean optical absorbance of the negative controls +20%. We added two positive and two negative control sera to each reaction plate, as described by Alves et al. [29].
We considered an animal to be positive to Trypanosoma infection, when any of the four diagnostic tests used (hemoculture, MHCT, PCR or/and serological tests) was positive.

Data analysis
Descriptive statistic (mean ± standard deviation) was applied to obtain the mean health parameters of the specimens. The Shapiro-Wilk test served to establish whether the distribution was normal. Finally, a Kruskal-Wallis test was applied to determine the differences between: no infection, T. evansi infection, T. cruzi infection, and coinfection. Post hoc Mann-Whitney tests were used to assess pair-wise results of the Kruskal-Wallis test.
To determine the direct and indirect influences of infections and coinfections in relation to anemia, infection responses, immune investment and body condition, we carried out a path analysis. We assessed variation in body condition based on the standardized residuals from an ordinary linear regression between body mass (g) and head-body length (mm) of individuals, while accounting for age and sex effects (13). This should circumvent the effects of animal growth on the condition index. Therefore, the residuals were calculated for males and females separately. To perform dimensionality reduction of anemia, infection responses, and immune investment values, we used the principal coordinate analysis, a geometric technique that converts a matrix of distances between points in multivariate space into a projection that maximizes the amount of variation along a series of orthogonal axes. We used an r value ! 0.60 to interpret the results (positive or negative effect) of the path analysis.
Path analysis describes two types of effects: direct and indirect. When the exogenous variable has an arrow directed towards the dependent variable, the effect is direct. When the effect is indirect, the arrow crosses one or more than one dependent variable until the final effect. The variables were considered to be statistically significant for p values 0.05. All data were analyzed using R (version 3.4.2) [35].
Path analysis showed a negative direct effect of T. evansi infection (path coefficient = -0.30, p < 0.05) on anemia indicators, resulting in lower PCV (r = 0.84) and MCV (r = 0.65). Although we did not observe a direct effect on anemia indicators of coatis infected with T. cruzi, we found an increased negative direct effect on these values in coinfected animals (path coefficient = -0.47, p < 0.05). Additionally, our results showed that T. evansi infection had a negative influence on body condition via anemia indicators (path coefficient = 0.37, p < 0.05). Moreover, this effect was potentiated in coinfected animals (S1 Fig).
Path analysis revealed that infections with T. cruzi and T. evansi had no effect on anemia indicators of crab-eating foxes. However, we found a negative effect on the infection responses following

Discussion
Our results reveal that T. evansi infection in coatis, crab-eating foxes and ocelots causes some degree of anemia. Anemia has been recorded previously in coatis infected with T. evansi Table 4 [13,[17][18][19], but the present study is the first report of T. evansi infection resulting in anemia also in free-living ocelots and crab-eating foxes. Anemia is characteristic of T. evansi infections [17,[36][37][38] and can represent a threat to the health of carnivores in the Pantanal wetland, as suggested by infection rates of 89% (17/19) in ocelots, 46% (18/39) in coatis, and 33% (16/48) in crab-eating foxes. Even though T. cruzi infection could not induce anemia in coatis, coinfection with T. evansi caused the degree of anemia to become more severe, a finding previously observed by Olifiers et al. 2015 [13]. The microcytic hypochromic anemia, characterized by the low MCV values in T. evansi-infected coatis and the even lower values in coinfected animals, may correlate to deficient hemoglobin synthesis due to iron deficiency [39][40][41], as observed in T. evansi infections [42,43]. The low MCV values could also result from the influx of iron into the cell, which is necessary for the multiplication of intracellular amastigote forms of T. cruzi [44,45].

Leopardus pardalis
Anemia was observed also in ocelots, as suggested by small differences in anemia indices in animals infected with T. evansi and coinfected with T. cruzi, as well as through direct effect of T. evansi infection and coinfection on PCV values tested by path analysis. Anemia has been recorded previously in domestic cats experimentally infected with T. evansi [46][47][48].
Moreover, lower RBC and higher MCV values indicated a megaloblastic anemia, which negatively influenced ocelots' body condition, irrespective of Trypanosoma spp. infection. Although we have not investigated other pathogens or other causes, in natural environments animals are constantly and concomitantly exposed to a variety of parasites, particularly Anaplasma spp., Mycoplasma spp., and piroplasmids, which cause lysis in parasited red blood cells and the consequent drop in RBC values. The same parasites have been described to infect ocelots in the studied area [49][50][51]. Additionally, the observed increase in MCV values may have metabolic origin and be associated with deficiency of vitamin B12, which is found mainly in protein diets, or/and in hepatic dysfunction.
Regarding crab-eating foxes, we observed a slight decrease in indicators of anemia only in T. evansi-infected animals. Importantly, domestic dogs that have been experimentally or naturally infected with T. evansi display evident signs of anemia and the course of infection is fatal if not treated [52,53]. Therefore, free-living crab-eating foxes parasitized with T. evansi may become sick and prostrate, consequently they may die or are not collected.
We observed discrete leucopenia due to fewer lymphocytes and eosinophils in coatis parasited with T. evansi. Immunosuppression in coatis infected with T. evansi has been described previously in natural and experimental studies [13,18,19,54]. This phenomenon varies in nature due to different communities of parasites in their hosts, as well as the influence of marked seasonality of resources, which is characteristic of the Pantanal region [12,27].
The leukocytosis observed here in T. cruzi-infected and coinfected coatis is typical of the acute phase of T. cruzi infection [55]. The increase in leucocytes during T. cruzi infection in wild mammals has been reported in Thrychomis pachyurus and coatis under experimental and natural conditions, respectively [12,56].
We observed a notable infection response in coatis infected with T. cruzi and in coinfected animals. Monocytosis is a sign of immune response during the acute phase of T. cruzi infection [57,58]. Throughout the chronic phase of T. cruzi infection, neutrophils act together with monocytes and lymphocytes to repair the tissue damage caused by T. cruzi amostigote [59]. An increase in monocyte and neutrophil values is an important hallmark of infection by T. cruzi in naturally infected coatis, as already reported by Martínez-Hernández et al. 2016 [12].
We observed a decrease in lymphocytes in crab-eating foxes infected with T. evansi, confirming the findings of Da Silva et al. 2011 [60] in the chronic phase of T. evansi infection in laboratory rodents. Indeed, domestic dogs naturally and experimentally infected with T. evansi displayed fewer WBCs and neutrophils [52,53]. Additionally, the decrease in monocytes and neutrophils observed in crab-eating foxes infected with T. cruzi or in coinfected animals, was similar to that reported in dogs during the early stages of T. cruzi experimental infection [61]. Such immunosuppression, even if transient, can impair the health of the animal.
An increase in immune investment in coatis, ocelots, and crab-eating foxes infected with T. cruzi, T. evansi, or in coinfected animals recorded in the present study may be associated with a potent stimulation of cellular and humoral immune response, characteristic of trypanosome infection [62][63][64]. The strong production of immunoglobulins results in an autoimmune hypersensitivity with consequent production of antigen-antibody molecules [65]. These immune complexes accumulate on the vascular wall, especially in the microcirculation, causing damage to their thin layer of vascular endothelial cells, and resulting in widespread micro bleeding, a phenomenon known as disseminated intravascular coagulation (DIC). DIC has been associated with trypanosome infections in various host species [66][67][68][69] and has been observed in coatis infected with trypanosomes in the Pantanal region. Indeed, as observed here by path analysis, an increase in immune investment resulted in a worse body condition in ocelots and crab-eating foxes. DIC, together with the hypoferremic response discussed above, are the main causes of anemia observed in trypanosome-infected animals. Furthermore, oxidative stress due to oxidative damage in erythrocyte membranes are related to experimental and natural infection by T. evansi [70,71].

Conclusions
As T. cruzi is restricted to the New World, it had been interacting with its hosts over millions of years. On the contrary, T. evansi originates from the African continent, and has become a parasite of South American wild mammals only recently. In the Pantanal region, T. evansi was probably introduced together with horses and dogs in the late Eighteenth century when the first cattle farms were established. According to this scenario, while the course of T. cruzi infection is known to be predominantly chronic probably due to ancient association with its hosts, T. evansi infection of endemic Neotropical fauna may cause great damage to the health of its hosts, particularly due to increased virulence and pathogenicity of present interactions.
The anemia and immunosuppression evidenced by the present study, are associated with increasing habitat fragmentation and poaching [72], which poses a threat to wild coatis, ocelots and crab-eating foxes in the Pantanal wetland. Furthermore, due to epidemiological implications and conservation importance, studies of T. cruzi and T. evansi infections in free-living mammals should be a priority for health surveillance organizations, research promotion agencies, and postgraduate programs.