Detection of arboviruses of public health interest in free-living New World primates ( Sapajus spp . ; Alouatta caraya ) captured in Mato Grosso do Sul , Brazil

Introduction: A sero-epidemiological survey was undertaken to detect the circulation of arboviruses in free-living non-human primates. Methods: Blood samples were obtained from 16 non-human primates ( 13 Sapajus spp. and three Alouatta caraya) that were captured using terrestrial traps and anesthetic darts in woodland regions in the municipalities of Campo Grande, Aquidauana, Jardim, Miranda and Corumbá in the State of Mato Grosso do Sul, Brazil. The samples were sent to the Instituto Evandro Chagas (IEC) in Ananindeua, Pará, Brazil, to detect antibodies against 19 species of arboviruses using a hemagglutination inhibition test (HI). Results: Of the 16 primates investigated in the present study, fi ve (31.2%) were serologically positive for an arbovirus. Of these fi ve, two (12.5%) exhibited antibodies to the Flavivirus genus, one (6.2%) exhibited a monotypic reaction to Cacipacoré virus, one (6.2%) was associated with Mayaro virus, and one (6.2%) was positive for Oropouche virus. Conclusions: Based on the positive serology observed in the present study, it was possible to conclude that arboviruses circulate among free-living primates. The viruses in the areas studied might have been introduced by infected humans or by primates from endemic or enzootic areas. Studies of this nature, as well as effi cient and continuous surveillance programs, are needed to monitor viral activities in endemic and enzootic regions.

Arboviruses are transmitted and maintained in nature through wild cycles, in which several species of bloodsucking arthropods act as vectors and wild vertebrates act as reservoir hosts 1 .The most commonly affected people are those who maintain close contact with wild environment s where ecological niches of arbovirus exist 2 .In these environments, arboviruses can cause meningitis and diseases of the central nervous system , as seen for infections with St. Louis encephalitis virus (SLEV), Rocio virus (ROCV), eastern equine encephalitis virus (EEEV) and western equine encephalitis virus (WEEV) 1 .
Certain arboviruses appear regularly in urban areas, such as dengue virus (DENV) and Oropouche virus (OROV), or in peri-urban areas, such as Mayaro virus (MAYV) and yellow fever virus (YFV), causing epidemic febrile illnesses characterized by exanthematous and/or hemorrhagic fever 1 .
Arboviruses contain genomes made up of ribonucleic acid (RNA) that is either segmented or non-segmented 3 .These viruses are classifi ed into fi ve principal families according to their antigenic properties and physicochemical characteristics: Bunyaviridae, Flaviviridae, Reoviridae, Togaviridae and Rhabdoviridae 1 .
Arboviruses are considered to be emerging disease agents when they initially appear in a population or when their incidence or geographic distribution increases considerably 3 , as observed in a recent serological study of non-human primates (NHPs) in the municipality of Bonito, Mato Grosso do Sul, which detected antibodies to MAYV and OROV 4 .A number of factors can precipitate emerging disease, such as ecological changes due to economic development, agricultural or climatic abnormalities, demographic and behavioral changes, international traffi c and wildlife trade, microbial adaptation or a collapse in public health control programs 3 .
Due to their arboreal and diurnal habits, non-human primates are more frequently infected by arboviruses than other terrestrial animals.They are infected when they feed in Batista PM et al -Arbovirus infections in Mato Grosso do Sul METHODS treetops, using the same feeding schedule as vectors 5 .Thus, they can act as important reservoirs in the cycle of several zoonoses.Furthermore, because they are part of a habitat with high biological diversity, they could be used for natural sentinel surveillance of emerging arboviruses, even when endangered by these diseases 6 .
Non-human primates can be used in biomedical research because they have anatomical, physiological, biochemical and behavioral similarities with humans 7 .Serological studies in wild animals have been limited.They have been performed through cross-sectional surveys, using several animals at a time , to detect the distribution of seropositivity and antibody titers to determine whether the host has been exposed to an antigen 5 .
Because they have genetic and physiological characteristics similar to those of humans, primates are susceptible to several common pathogens that can cross the boundaries of species through various transmission routes, particularly through vectors.Interaction between humans and wild primates living in tropical forests has increased due to ecotourism and the invasion of the forests, thereby increasing the likelihood of the transmission of pathogens 6 .
Arboviruses could potentially become an important public health problem in Brazil in the coming years .Monitoring programs should be able to detect these emerging viruses before they cause major outbreaks.Laboratory tests to confi rm diagnosis are of great importance because these viruses cannot be easily distinguished clinically from other viral diseases, such as dengue 8 .
The aim of the present study was to conduct serological and virological tests to detect circulating arboviruses in non-human primates in the regions of Serra da Bodoquena and the Pantanal in Mato Grosso do Sul (Brazil ) to provide information that could be used to defi ne and improve prevention and control strategies for these important zoonoses.

The study site
The wild animals used in the present study were captured from forests in the municipalities of Campo Grande (8), Aquidauana (1), Jardim (4), Miranda (2) and Corumbá (1) in the State of Mato Grosso do Sul (Figure 1).In these municipalities, there is a predominance of the cerrado ecosystem, which is characterized by short, sloping, twisted trees with thin trunks.The cerrado has well-defi ned climatic seasons : a rainy season in summer, from October to April, and a dry season in winter, which extends from June to August.
The City of Campo Grande, which is the capital of the State of Mato Grosso do Sul, is situated in the central region of the state, at approximately 532m above sea level (20°26'34" S and 54°38'47" W).It has an estimated area of 8,096.05km 2 , and its population is approximately 765,000 inhabitants.Th e municipality of Aquidauana is located in the south of the Central West region of Brazil, in the Pantanal of Mato Grosso do Sul (mi cro-region -Aquidauana) in the region of the Serra da Piraputanga and Maracajú (2 0º28'15" S and 55°47 '13" W).Th e municipality of Jardim is located in southwest Mato Grosso do Sul (21º28'49" S and 56º08'17" W).It h as an area of 2,207.6km²,with a humid sub-tropical climate and temperatures ranging from 15ºC to 39ºC.
In t he municipality of Miranda, animals were captured at a park hotel on P antanal Park Road, Mato Grosso do Sul (20 º30' S and 56º15' W).In t he City of Corumbá, non-human primates were captured in the Nhecolândia sub-region (18º20'-19º40' S and 57º57'-55º00' W).
The Pant anal is a seasonal tropical wetland with an area of approximately 140,000km², and it i s considered one of the largest freshwater ecosystems in the world 9 .The Br azilian Pantanal, which represents 85% of the total area of the Pantanal, is locat ed in the States of Mato Grosso and Mato Grosso do Sul in the Central-West region of Brazil, and th e parts of the Pantanal in these two states are known as the Pantanal North and South, respectively 10 .This reg ion is ecologically classifi ed into subregions according to vegetation, fl ooding and physiography 11 .The Nhec olândia sub-region, which is situated in the S outh Pantanal, comprise s approximately one-fifth of the total area of the ecosystem, and it is characterized by hundreds of shallow lakes that exhibit different degrees of salinity, as well as coalescence of the system during fl oods 12 .

Primates
Blood sa mples from 16 free-living non-human primates were analyzed.Thirteen Sapajus spp.were captured using humane live terrestrial traps (Tomahawk) set in previously determined areas for easy viewing 13 .Three Alouatta caraya were ca ptured with the aid of an anesthetic dart rifl e 14 .Of the a nimals captured, four (25 %) were female and 12 (75%) were male, and 11 ( 68.8%) were adults and fi ve (31.2%) were juveniles.The anim als were anesthetized using a protocol based on the association of tiletamine hydrochloride and zolazepam hydrochloride 15 .The dosa ge, adjusted to the weight of the animal, was injected intramuscularly.All of t he capture procedures were performed by a group of biologists and veterinarians, with the authorization of the Brazilian Institute of Environment and Renewable Natural Resources (Instituto Brasileiro do Meio Ambiente e dos Recursos Naturais Renováveis -IBAMA) under authorization number 21808-1.
Blood sa mples were collected, and aliquots of serum and whole blood were initially frozen in liquid nitrogen and then stored in a freezer at -70°C until processing 16 .Biometri c and clinical data were collected (data not shown), and microchi p identifi cation transpo nders were attached to the animals.After re covery from anesthesia, the animals were released.

Serological tests
The seru m samples from the 16 animals were analyzed using the hemagglutination inhibition test (HI) and a p anel of 19 different types of arbovirus distrib uted among the following genera: Alphavir us (WEEV, EEEV, MAYV and Mucambo virus), Flavivirus (YFV, Ilheus virus, SLEV, Cacipacoré virus, ROCV and Bussuquara virus), Orthobunyavirus (Guaroa, Maguari, Tacaiuma, Utinga, Belém, Caraparu, Catu and ORO virus) and Phlebovirus (Icoaraci virus).The HI t est used in the present study was standardized by the Section of Arbovirology and Hemorrhagic Fevers (SAARB) of the I nstituto Evandro Chagas (IEC) followin g the protocol described by Clarke and Casals 17 , as adapt ed for microplates by Shope 18 .
The anti gens used were pre pared from the brains and/or serum of newborn infected mice using t he sucrose-acetone and sorovirus extraction methods, respectively 19 .The seru m samples were pre-treated with acetone to remove natural inhibitors, and they were adsorbed on goose erythrocytes (Anser cinereus) 20 to remove non-specific agglutinins that mig ht inhibit the agglutination of red blood cells, thereby avoiding false-positive results.

Virological tests
For viru s isolation, newborn Swiss albino mice (Mus musculus) were ino culated intracerebrally with 0.02mL of the serum and/or blood of the captured primates.The serum or blood samples were diluted 1:10 in buffered saline (PBS) containing antibiotics (100IU/ m L penicillin and 100µg/mL streptomycin) and 0.4% bovine a lbumin.The anim als were observed daily for 21 days, and an y change was recorded on their identifi cation cards.Concomitantly, a continuous culture of gut cells from Aedes albopictus Clone C 6/36) was inoculated with the primate samples.The cultures were maintained in Leibovitz's modifi ed culture medium with L-glutamine (L-15), suppleme nted with tryptose, non-essential amino acids, penicillin (100IU/mL) and streptomycin (100µg/mL), using 5% fetal bovine serum (FBS) for the growth medium and 2% for the maintenance medium.The cult ures were observed daily for 10 days with the aid of an invert ed optical microscope to assess cytopathic effects (CPEs).Confi rmation of viral replication in the cells was performed using the indirect fl uorescent antibody test (IFAT) 21 with po lyclonal antibodies to Alphavirus and Flavivirus genera 1 .

Statistical analysis
Associations between variables of interest and arbovirus infection were quantifi ed using prevalence ratios and their 95% confi dence intervals, and the signifi cance level was 5%.
There were no differences in the prevalence of infection by arboviruses according to sex, age or species.

DISCUSSION
The results found in the present study were similar to results previously reported for the municipality of Bonito (State of Mato Grosso do Sul), in which 17 (48.5%)animals exhibited antibodies to arboviruses of the genus Alphavirus and Flavivirus , as well as monotypic reactions to Mayaro and Oropouche viruses.A number of samples also exhibited reactivity to more than one type of arbovirus 4 .
When non-human primates are assessed, the use of seroepidemiological surveys as an indicator of the circulation of arboviruses, as well as sentinel animals, is a feasible method for obtaining information concerning the presence of viruses in natural environments 22 because these primates are arboreal and commonly diurnal species that are more frequently infected by arboviruses than other terrestrial animals 5 .
In the present study, the HI test was used to detect the distributions of seropositive animals and antibody titers in non-human primates.The same diagnostic technique was used to detect arbovirus antibodies in spider monkeys (Ateles geoffroyi) 23 .The HI test has often been used in serological surveys because it can detect antibodies over a long period after natural infection.It is considered a test of high sensitivity and low specifi city compared to other serological tests, such as the immunoglobulin M (IgM) capture enzyme immunoassay (MAC-ELISA) 24 .However, it is a more sensitive and accurate technique than methods that use large sample volumes, enabling savings in time and reagents.The HI test is an ideal method for detecting antibodies in wild animals captured in forests 19 .
In the present study, cross-reactivity was found between viruses belonging to the genus Flavivirus in two serum samples, precluding the identifi cation of the infecting virus and the time of infection of the animals.With the HI test, it is common to observe the occurrence of cross-reaction between viruses belonging to this genus 24 .Positive results for the presence of antibodies to Flavivirus suggest that the host was exposed at some stage to one of the arboviruses studied and produced antibodies to combat it 5 .
The negative results found for viral isolation in the present study corroborate the data found in a previous study conducted   25 .Viral isolation from blood samples of animals and humans is considered to be a sensitive and economical method, and it is the gold standard for virological diagnosis 24 .
The detection of antibodies to Cacipacoré virus in a primate from the municipality of Aquidauana, together with a previous report of virus isolation on a farm in the State of Rondônia ( northern Brazil, bordering Bolivia) 26 , suggests that the State of Mato Grosso do Sul could enable the spread of this arbovirus because the BR-262 highway connects the Southeast region of Brazil to the Bolivian border.
Similar to the results of the present study, antibody titers for MAYV were detected in 150 primate sentinels from French Guiana.A high prevalence of antibodies to arboviruses was observed in humans in the same study 27 .Antibodies to MAYV were also found in Callithrix argentata primates, using the HI test , during an investigation of outbreaks of Mayaro and yellow fever in Belterra, in the State of Pará, Brazil 28 .Similar results were also observed in a study of primates in the municipality of Bonito 4 .
MAYV can be transmitted by Aedes aegypti, which is the urban vector of dengue.Dengue is present in many Brazilian cities, including the municipalities surveyed, and there have been many cases of dengue in 2013, according to an epidemiological report of the State Department of Health 29 .MAYV might have been introduced into the urban areas near the study area by human travelers or by infected non-human primates that live in the forests around cities 30 .Mayaro virus was also detected in Venezuela, infecting the members of a family who exhibited clinical polyarthritis 31 .
The detection of antibodies to OROV in a primate sample from Passo da Lontra, in the municipality of Miranda, suggests the possible circulation of this virus in the region.Consequently, local people and tourists who maintain close contact with the natural environment could be susceptible to infection with OROV 1 .During a surveillance program of yellow fever and Oropouche in Minas Gerais, OROV was isolated in a liver sample from a primate of the genus Callithrix 32 , which is considered a new host for OROV in Brazil.
In contrast, the negative results for the primate captured in the City of Corumbá suggested that the virus did not circulate in that region or that the host was not exposed to any of the arboviruses studied.Another possible explanation for the negative result is the small number of animals captured 5 .
The positive results observed in the present study confi rm the circulation of arboviruses in the wild, non-human primate populations of the State of Mato Grosso do Sul, Brazil.The State of Mato Grosso do Sul borders Paraguay and Bolivia, where arboviruses have been detected in a variety of primate and arthropod species.These regions exhibit favorable conditions for the occurrence of outbreaks caused by arboviruses.Therefore, much larger studies of this nature, as well as effi cient and continuous epidemiological surveillance programs, are needed to monitor viral activities in endemic or enzootic areas.The results of these epidemiological studies could facilitate the discovery of diseases that affect primates and that can be transmitted to humans, thereby helping to prevent outbreaks in human populations 33 .
P O R É O U D O P E IX E

FIGURE 1 -
FIGURE 1 -Map of the State of Mato Grosso do Sul, Brazil , with the municipalities of Campo Grande, Aquidauana, Miranda, Jardim and Corumbá highlighted (study area).

TABLE 1 -
Antibodies against arboviruses in sera from 16 non-human primates (Sapajus spp.; Alouatta caraya) detected by the hemagglutination inhibition test, with the results grouped by sex, age and species.

Batista PM et al -
Arbovirus infections in Mato Grosso do Sul

TABLE 3 -
Antibody titers for arboviruses determined using the hemagglutination inhibition test, with the sex, age and the frequency of positive non-human primates (Sapajus spp.; Alouatta caraya) captured in the municipalities of Campo Grande, Jardim, Aquidauana, Miranda and Corumbá in the State of Mato Grosso do Sul, Brazil -2013.