Post-mortem hemoparasite detection in free-living Brazilian brown brocket deer ( Mazama gouazoubira , Fischer 1814 ) Detecção post-mortem de hemoparasitos em veados catingueiro ( Mazama gouazoubira , Fischer 1814 ) de vida livre no Brasil

Tick-borne infections can result in serious health problems for wild ruminants, and some of these infectious agents can be considered zoonosis. The aim of the present study was the post-mortem detection of hemoparasites in free-living Mazama gouazoubira from Minas Gerais state, Brazil. The deer samples consisted of free-living M. gouazoubira (n = 9) individuals that died after capture. Necropsy examinations of the carcasses were performed to search for macroscopic alterations. Organ samples were collected for subsequent imprint slides, and nested PCR assays were performed to detect hemoparasite species. Imprint slide assays from four deer showed erythrocytes infected with Piroplasmida small trophozoites, and A. marginale corpuscles were observed in erythrocytes from two animals. A. marginale and trophozoite co-infections occurred in two deer. A nested PCR analysis of the organs showed that six of the nine samples were positive for Theileria sp., five were positive for A. phagocytophilum and three were positive for A. marginale, with co-infection occurring in four deer. The results of the present study demonstrate that post-mortem diagnostics using imprint slides and molecular assays are an effective method for detecting hemoparasites in organs.


Introduction
Tick-borne infections caused by protozoa and rickettsial organisms can result in serious problems for wild and domestic hosts, particularly when the animals have been weakened by other parasites or are undernourished or stressed (FOWLER, 1993).The brown brocket deer (Mazama gouazoubira) is a small forest cervid with solitary habits that can live in anthropogenic areas and in close proximity to farmers in neotropical forest areas.This behavior favors the interchange of infectious microorganisms between wild and domestic populations (DUARTE, 1997).
Babesia and Theileria species have been detected by molecular means in Brazilian cervids from Minas Gerais and Mato Grosso do Sul states (SILVEIRA et al., 2011(SILVEIRA et al., , 2013)).These species are known to cause clinical signs and hematological alterations in stressed animals (KREIER et al., 1962;BARKER et al., 1973).On the other hand members of the Anaplasmataceae family can infect brown brocket deer, but the consequences of the infection on deer health are not well established.
Anaplasma marginale strains from deer exhibit low virulence in bovine hosts, and they have been studied as alternatives for controlling A. marginale infection in cattle (KUTTLER;ZAUG, 1988;KOCAN et al., 2003;SILVEIRA, 2012).Anaplasma phagocytophilum infects humans and a wide range of other animal hosts including rodents, ruminants, felids, horses, donkeys, dogs and birds (MADIGAN;PUSTERLA, 2000;STUEN, 2007).Capturing free-range deer is a fundamental concern for studies that require close access to the animals for sample collection and other ecological studies.However, cervid immobilization is a difficult process because these animals are easily frightened and very sensitive to capture myopathy (DUARTE, 2006).In some cases, a restrained deer can die a few days after being captured.Therefore, an alternative way to study hemoparasites from these hosts is to harvest organs from free-living deer when these animals are found injured or bowled over.The aim of the present study was the post-mortem detection of hemoparasites in organs from free-living Mazama gouazoubira in Minas Gerais state.

Material and Methods
This study was approved by the Ethical Committee on Animal Experimentation (CETEA/UFMG, Belo Horizonte, MG, Brazil) under protocol no.142/08 and by the Brazilian Institute for Environment and Natural Renewable Resources (IBAMA, Belo Horizonte, MG, Brazil) under license no.16064-1.

Animal samples
This study was performed from June 2007 to April 2013.The deer samples came from free-living M. gouazoubira (n = 24) that were captured by the Forestry Police of Minas Gerais state and transported to IBAMA.Blood from these animals was collected by puncturing the jugular vein, and the samples were transferred immediately to vials containing EDTA.Blood smears were prepared, subjected to quick Romanowsky staining and examined under an optical microscope at 100X magnification.At least 40 microscopic fields were observed for each deer blood sample.Blood sample aliquots were frozen and stored for subsequent DNA extraction.
Nine (9/24) of these animals eventually died, and necropsy examinations were performed.The animals most likely died from capture myopathy or because of erroneous immobilization.Dog attacks were confirmed in three cases (MGM specimens).The deer were identified as follows: MGI8, MGI11, MGI12 and MGM1 (adult females), MGM2 (pregnant female) MGI13 and MGI19 (young females), MGM3 (adult male) and MGI14 (young male).Their ages were determined according to the following parameters: young (1-2 months), for animals presenting white hair spots; or adult, for animals presenting mature features.

Necropsy
Necropsy examinations were performed on nine dead deer, and the carcasses were evaluated for macroscopic alterations.Imprint slides of lymph node, spleen, liver and brain samples were prepared, and samples of these organs were collected and frozen for subsequent DNA extraction.The imprint slides were subjected to quick Romanowsky staining and examined under the optical microscope at 100× magnification.

DNA extraction and PCR amplification
DNA was extracted from the cervid organ and blood samples using a Wizard Genomic DNA Purification Kit (Promega, Madison, WI, USA), which was employed according to the manufacturer's instructions for animal tissue and whole blood, respectively.
The first reaction mixture contained 1.2 µl of dNTPs (2.5 mM), 0.15 µl of Taq polymerase (Phoneutria, Belo Horizonte, MG, Brazil) (0.05 U), 1.5 µl of reaction buffer IB (Phoneutria, Belo Horizonte, MG, Brazil), 0.6 µl of a solution containing the mixed primers (10 µM) and 10.05 µl of sterile ultra-pure water.A 1.5 µl aliquot of the DNA template was added to the reaction mixture to achieve a final volume of 15 µl.
The second reaction mixture contained 2.0 µl of dNTPs (2.5 mM), 0.25 µl of Taq polymerase, 2.5 µl of reaction buffer IB, 1.0 µl of a solution containing the mixed primers (10 µM) and 16.75 µl of sterile ultra-pure water.An aliquot (2.5 µl) of amplicon that was obtained after the first reaction was added to the reaction mixture to achieve a final volume of 25 µl.
Amplification was performed with an Eppendorf Mastercycler (Eppendorf, São Paulo, SP, Brazil) thermocycler.The amplification program used in all nPCR assays was as follows: 94 °C for 5 min (initial denaturation step); 30 cycles of 92 °C for 1 min (denaturation), 54 °C for 1 min (annealing) and 72 °C for 2 min (extension), followed by 72 °C for 8 min (final extension step).After the amplification, the reaction mixtures were maintained at 12 °C.The PCR amplicons were separated by electrophoresis on 1% agarose gel (40 min, 100 V), which was stained with GelRed (Biotium, Hayward, CA, USA) and visualized under ultraviolet light.
The positive products of the second-round PCRs were purified using a QIAquick PCR Purification Kit (Qiagen Biotecnologia Brazil, São Paulo, Brazil), according to the manufacturer's instructions.The purified amplicons were sequenced by using ABI3130 with POP7 polymer and BigDye v3.1 (Myleus Biotechnology, Belo Horizonte, MG, Brazil) with the second reaction primers.The sequences were aligned, edited and analyzed at the URL http:// asparagin.cenargen.embrapa.br/phph/and using the MEGA 6.0 software (TAMURA et al., 2013).The identity of each sequence was confirmed through a comparison with the sequences available at GenBank using the BLAST software (ALTSCHUL et al., 1990).Phylogenetic analyses were conducted using the MEGA 6.0 software, and the phylogenetic position of the pathogens was inferred using the neighbor-joining method.The combination of phylogenetic clusters was assessed by bootstrap tests with 1000 replicates, and a Kimura 2-parameter model was used to test the stability of the trees.

Results
No visible macroscopic lesions were observed during the necropsy examinations of nine brown brocket deer, which led to the conclusion that the animal deaths were most likely caused by stress from animal handling.Imprint slides from deer MGI19 could not be analyzed because of the low quality of the smears; thus, slides from eight of the nine deer were analyzed.Protozoa in the form of Piroplasmida small trophozoites (<2 µm) were  (2009) found in the erythrocytes of 50% (4/8) of the analyzed animals, and A. marginale corpuscles were observed in the erythrocytes of 25% (2/8) of the deer.Intraerythrocytic Piroplasmida trophozoites were detected in three spleen samples, three liver samples and one brain sample.A co-infection with A. marginale corpuscles and piroplasmid trophozoites was observed in two animals using spleen and liver slides (Table 2).The liver and spleen were the most parasitized organs, and no infected erythrocytes were observed in lymph nodes.
A molecular examination of whole blood that was published in Silveira et al. (2011Silveira et al. ( , 2012) ) revealed that 55.5% (5/9) of the samples were infected with hemoparasites; five were infected with A. marginale, four with T. cervi and one with Anaplasma sp.(which is closely related to A. platys/A.phagocytophilum).Co-infection was present in 44.4% (4/9) of the samples; three had T. cervi and A. marginale, and one had T. cervi, Anaplasma sp. and A. marginale (Table 2).
A phylogenetic analysis of the partial 16S rRNA and the msp4 gene showed that the phylogenetic topology generally placed samples from this study in close proximity with other A. phagocytophilum isolates, including isolates from humans and wild ruminants, reinforcing that these isolates are from A. phagocytophilum (Figures 1  and 2).The 18S rRNA from the Piroplasmida phylogenetic tree showed a cluster containing T. cervi isolates from other deer, including one sequence (gb HM466930.1,Silveira et al. ( 2012)) from M. gouazoubira in Minas Gerais, Brazil (Figure 3).
No hemoparasites were detected in organ samples from MGI12, although the infection was detected in whole blood.However, MGI13 was not positive for any parasites in the whole blood, but it was positive for Babesia/Theileria species in the lymph node sample.Half of the lymph nodes tested (3/6) were positive for hemoparasites, and the liver samples from five animals were positive.Co-infection was most common in whole blood (4/9) compared with organ samples; however, one animal (MGI14) presented co-infection only in the organ samples (Table 2).None of the young animals were positive for hemoparasites according to the imprint slides analysis, but nPCR assays revealed that coinfection occurred in two of these animals (2/3).

Discussion
Free-ranging wild animals may be considered reservoirs of an increasing number of pathogens with zoonotic potential.When some parasites infect both susceptible wild and domestic species that share the same habitats, both present enzootic importance and must be monitored (KUIKEN et al., 2005).Furthermore, some hemoparasites are a serious problem for domestic ruminants and humans, and both can be in close contact with wild animals, particularly deer, in South America (DUARTE et al., 2001).
According to Warns-Petit et al. ( 2010), post-mortem findings are the primary data collected for general disease surveillance.No macroscopic lesions were observed during the necropsy examinations of the nine brown brocket deer that were evaluated in the present study.Death caused by stressful handling can be acute, and post-mortem macroscopic alterations may not be evident, as found in this study.However, deer are often capable of harboring some hemoparasite infections in a latent asymptomatic stage for a long time (BARKER et al., 1973;KUTTLER;ZAUG, 1988;DAWSON et al., 1994;CANTU et al., 2007;DAVIDSON et al., 2001).
The blood smear is not reliable for distinguishing between some hemoparasites, such as Babesia and Theileria species, especially when both species occur together in endemic areas (SILVEIRA et al., 2011).Although the diagnosis of a hemoparasite infection is normally achieved through the examination of blood smears, this method exhibits poor sensitivity when there is a low level of parasitemia.
Nested PCR assays have been shown to provide more consistent data for mapping the distribution of hemoparasites that affect the wild fauna of Brazil.Whole blood analysis is a useful way to detect hemoparasites by using molecular techniques, and it is more amenable to DNA extraction compared with using organ samples.However, organ samples were found to be a suitable alternative to post-mortem detection.The present study reinforces the importance and utility of molecular techniques for post-mortem diagnostics and organ samples were found to be a suitable alternative to detection in necropsy.Erythrocytes in lymph node imprint slides have not shown infected, but half of the lymph node samples tested by nPCR were positive for hemoparasites.
No hemoparasites were detected in the whole blood of MGI13 and MGM3 deer; however, lymph nodes from MGI13 and the liver and brain of MGM3 were positive for Theileria species.Furthermore, none of the young were positive for hemoparasites in their imprint slides, but nPCR assays of blood or organ samples revealed that co-infection occurred in these animals.These facts Table 2. Hemoparasite detection in organs and whole blood from nine free-living brown brocket deer in the state of Minas Gerais, Brazil, as determined by imprint slides and nested PCR.In relation to hemoparasites finding in this study, deer are important natural reservoir hosts of Anaplasmataceae, and they are capable of harboring latent A. marginale and A. phagocytophilum infections (CHRISTENSEN et al., 1958).Natural infection with A. phagocytophilum has been reported in wild deer, and these animals are considered reservoirs in Europe, Asia and North America (revised by Stuen et al. (2013)).Pathogenic and apathogenic genetic variants have been found in these animals (MASSUNG et al., 2005;SILAGHI et al., 2011;OVERZIER et al., 2013).

Deer
In Brazil, A. phagocytophilum has been molecularly detected in dogs, carnivorous birds, Rhipicephalus sanguineus and Amblyomma cajennense ticks (SANTOS et al., 2006(SANTOS et al., , 2013;;MACHADO et al., 2012).Amblyomma cajennense was found in MGI14 and MGI19 brown brocket deer, which were positive for A. phagocytophilum, by employing msp4 gene amplification (SILVEIRA, 2012).However, future studies must investigate which vectors and hosts are responsible for maintaining A. phagocytophilum in nature.
According to a BLASTn analysis of the 16S rRNA gene, the MGI8 samples shared a high percentage of their identity (100%) with an A. phagocytophilum isolate from a person from Poland (gb KF111754.1)and an isolate from white-tailed deer in the USA (gb DQ104442.1).The msp4 gene analysis showed a high shared identity (98-100%) with European samples isolated from ovine, horses and Ixodes ricinus ticks.A phylogenetic tree analysis showed that the Anaplasma agent found in the present study is closely related to A. phagocytophilum at one site that is specific to this species, namely the msp4 gene, and in another site in hypervariable regions that can provide a species-specific signature, namely the 16S rRNA gene, reinforcing that these isolates are A. phagocytophilum (Figures 1 and 2).To the best of our knowledge, this is the first molecular evidence of closely related A. phagocytophilum species in Brazilian deer, and the zoonotic role of this agent is still not understood.
Liver and brain samples from the MGI8 animal were positive for A. phagocytophilum according to a PCR of 16S rRNA and msp4 targets, reinforcing the identity of the isolate.Additionally, this agent was detected in the spleens of other cervids (MGM3, MGI14 and MGI19).Previous reports have shown that various tissues and cells are susceptible to infection by A. phagocytophilum (KLEIN et al., 1997;MUNDERLOH et al., 2004).The pathological change described during A. phagocytophilum infection in deer was an enlarged spleen with subcapsular bleeding in European roe deer (Capreolus capreolus) and reindeer (Rangifer tarandus) (STUEN, 2003).
The prevalence of A. marginale infection in Brazilian cervids has been reported at different values, specifically 6.3% and 48% in O. bezoarticus from the Pantanal region according to Picoloto et al. (2010) and Silveira et al. (2013), respectively, and 79.3% in M. gouzoubira from Minas Gerais state according to Silveira et al. (2012).However, the consequences of infection for deer health are not clearly understood.Nevertheless, anaplasmosis is known to be responsible for great economic losses among domestic ruminants in tropical and subtropical areas because of the intravascular anemia evident in infected animals (DUARTE et al., 2001).
Theileria spp.infections in Brazilian cervids were first described by our research group.The prevalence of infections in M. gouzoubira and B. dichotomus from Minas Gerais was 47.6% (SILVEIRA et al., 2012), and it was 20% in O. bezoarticus from Pantanal (SILVEIRA et al., 2013).Theileria cervi infections in MGI8 and MGI19 were confirmed by nucleotide sequencing (SILVEIRA et al., 2011), and the 18S rRNA in the Piroplasmida phylogenetic tree of the present study showed a cluster containing T. cervi isolates from other deer, including one sequence (gb HM466930.1,Silveira et al. ( 2012)) from M. gouazoubira in Minas Gerais, Brazil (Figure 3).
The pathogenesis of T. cervi infection is poorly understood, but in immunosuppressed and splenectomized deer, it can result in severe microcytic anemia (KREIER et al., 1962;FOWLER, 1993;YABSLEY et al., 2005).A high percentage of parasitized red cells in the capillaries of the brain can be found in some infections with intraerythrocytic piroplasms (RISTIC, 1988).Although the consequence may be cerebral ischemia, the magnitude of this parasitism in deer is unknown.The present study detected hemoprotozoa in the brain samples of these deer; however, whether brain parasitism can cause nervous symptoms, especially in immunosuppressed animals, remains unclear.
of examining the lymphoid organs, liver and brain by using nPCR assays.

Figure 1 .
Figure 1.The phylogenetic tree of Anaplasma phagocytophilum as isolated from M. gouazoubira MGI8 organs and based on a partial sequence of the 16S rRNA gene.The GenBank accession numbers are shown.The tree was constructed using the neighbor-joining method, and the numbers above the internal nodes indicate the percentages of 1,000 bootstrap replicates that supported the branch.Outgroup: Wolbachia spp.*samples from this study.

Figure 2 .
Figure 2. The phylogenetic tree of Anaplasma phagocytophilum as isolated from M. gouazoubira organ samples based on a partial sequence of the msp4 gene.GenBank accession numbers are shown.The tree was constructed using the neighbor-joining method, and the numbers above the internal nodes indicate the percentages of 1,000 bootstrap replicates that supported the branch.Other Anaplasmataceae were used as outgroups.*samples from this study.

Figure 3 .
Figure3.The phylogenetic tree of Theileria as isolated from M. gouazoubira MGI8 organs and based on the partial sequence of the 18S rRNA gene.The GenBank accession numbers are shown.The tree was constructed using the neighbor-joining method, and the numbers above the internal nodes indicate the percentages of 1,000 bootstrap replicates that supported the branch.Toxoplasma gondii, P. micans and S. muiris were used as outgroups.*samples from this study.

Table 1 .
Specific primers used for the detection of hemoparasites.