This study was performed in Maranhão state, a transition area two Brazilian biomes, Amazon and Cerrado. During 2011–2013, 1,560 domestic dogs were sampled for collection of serum blood samples and ticks in eight counties (3 within the Amazon and 5 within the Cerrado). A total of 959 ticks were collected on 150 dogs (9.6%). Rhipicephalus sanguineus sensu lato (s.l.) was the most abundant tick (68% of all collected specimens), followed by Amblyomma cajennense sensu lato (s.l.) (12.9%), Amblyomma parvum (9.2%), and Amblyomma ovale (5.2%). Other less abundant species (<1%) were Amblyomma oblongoguttatum, Rhipicephalus microplus, Haemaphysalis juxtakochi, and Amblyomma rotundatum. Females of A. cajennense s.l. ticks were morphologically identified as A. cajennense sensu stricto (s.s.) or A. sculptum. Molecular analyses of 779 canine ticks revealed three Rickettsia species: Rickettsia amblyommatis in 1% (1/100) A. cajennense s.l., ‘Candidatus Rickettsia andeanae’ in 20.7% (12/58) A. parvum, Rickettsia bellii in 6.8% (3/44) A. ovale and 100% (1/1) A. rotundatum ticks. An additional collection of A. sculptum from horses in a Cerrado area, and A. cajennense s.s. from pigs in an Amazon area revealed R. amblyommatis infecting only the A. cajennense s.s. ticks. Serological analysis of the 1,560 canine blood samples revealed 12.6% canine seroreactivity to Rickettsia spp., with the highest specific seroreactivity rate (10.2%) for R. amblyommatis. Endpoint titers to R. amblyommatis were significantly higher than those for the other Rickettsia antigens, suggesting that most of the seroreactive dogs were exposed to R. amblyommatis-infected ticks. Highest canine seroreactivity rates per locality (13.1–30.8%) were found in Amazon biome, where A. cajennense s.s. predominated. Lowest seroreactivity rates (1.9–6.5%) were found in Cerrado localities that were further from the Amazon, where A. sculptum predominated. Multivariate analyses revealed that canine seroreactivity to Rickettsia spp. or R. amblyommatis was statistically associated with rural dogs, exposed to Amblyomma ticks.
Citation: Costa FB, da Costa AP, Moraes-Filho J, Martins TF, Soares HS, Ramirez DG, et al. (2017) Rickettsia amblyommatis infecting ticks and exposure of domestic dogs to Rickettsia spp. in an Amazon-Cerrado transition region of northeastern Brazil. PLoS ONE 12(6): e0179163. https://doi.org/10.1371/journal.pone.0179163
Editor: Ulrike Gertrud Munderloh, University of Minnesota, UNITED STATES
Received: February 25, 2017; Accepted: May 24, 2017; Published: June 8, 2017
Copyright: © 2017 Costa et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are within the paper and its Supporting Information files.
Funding: This work received financial support by the São Paulo Research Foundation (FAPESP grant # 2010/52395-0). The funder had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Amazon and Cerrado are the two largest biomes representing the greatest biodiversity of Brazil . A few studies have reported the diversity of ticks infesting domestic dogs in areas within the Amazon [2–4] and Cerrado [5–7]. Overall, these studies indicate that dogs living in urban areas of counties within these two biomes are infested mainly by Rhipicephalus sanguineus sensu lato (s.l.), whereas dogs from rural areas are infested mainly by ticks of the genus Amblyomma, and also by R. sanguineus s.l. in some areas. While a number of Amblyomma species were reported on these dogs, Amblyomma cajennense sensu lato (s.l.) prevailed in the Cerrado biome, and only rarely reported on dogs from the Amazon biome.
The taxon A. cajennense s.l. was recently reassessed and demonstrated to represent a complex of six species, namely Amblyomma cajennense sensu stricto (s.s.) (restricted to the Amazonian region), Amblyomma mixtum (from Texas to western Ecuador), Amblyomma sculptum (northern Argentina, Bolivia, Paraguay, Brazil), Amblyomma interandinum (inter-Andean valley of Peru), Amblyomma tonelliae (dry areas of northern Argentina, Bolivia and Paraguay), and Amblyomma patinoi (Eastern Andes of Colombia) . A subsequent extensive study demonstrated that only two of these species, A. cajennense s.s. and A. sculptum occur in Brazil . In this case, the distribution of A. cajennense s.s. was primarily associated with the Amazon biome, whereas A. sculptum was primarily associated with the Cerrado and Atlantic rainforest biomes . Interestingly, some transition areas were found with the two tick species in sympatry. One of these transition areas is the state of Maranhão, composed by the Amazon biome in the western part (where A. cajennense s.s. predominates), and by the Cerrado biome in the eastern and southern parts (where A. sculptum seems to predominate .
The A. cajennense species complex is of great medical importance in Brazil because A. sculptum is the main vector of Rickettsia rickettsii, the etiological agent of Brazilian spotted fever (BSF), the deadliest spotted fever of the world . In southeastern Brazil, where BSF is endemic, R. rickettsii is the sole Rickettsia species that has been reported to infect A. sculptum; however, infection rates are extremely low among A. sculptum populations, usually with less than 1% of the ticks harboring R. rickettsii [11,12], albeit most of the A. sculptum populations are not found infected by any Rickettsia species [13–15]. On the other hand, the species A. cajennense s.s. is commonly found infected by another spotted fever group (SFG) agent, the species Rickettsia amblyommatis (formerly ‘Candidatus R. amblyommii’). Studies from the Brazilian Amazon have reported 26.8 to 69.4% R. amblyommatis-infection rates among A. cajennense s.s. populations [16,17]. While A. cajennense s.s. is an important human-biting tick in the Amazon region , this rickettsial agent was never confirmed as a human pathogen ), although there has been serological evidence that R. amblyommatis could cause spotted fever illness in the United States [19,20]. Importantly to note, there has been laboratory evidence that R. amblyommatis- naturally infected ticks are competent to transmit this rickettsial agent to rabbits .
Considering that the domestic dog is an important sentinel for tick-borne spotted fever [13,22,23], herein we investigated ticks infesting dogs and the rickettsial infection in ticks, and canine exposure to Rickettsia spp. in the state of Maranhão, where A. sculptum and A. cajennense s.s. have been reported to occur in sympatry in many areas.
Materials and methods
Study areas and canine sampling
This study was carried out in eight counties of the Maranhão state, northeastern Brazil. Three of the counties are located in the Amazon biome, whereas the remaining 5 counties belong to the Cerrado biome. Overall, the climate in the state of Maranhão is characterized by a rainy season extending from December to June, and the dry season from July to November. The annual average temperature, relative humidity and rainfall are 26.2°C, 70–85% and 1,000 to 2,500 mm, respectively .
From 2011 to 2013, a total of 1,560 domestic dogs were sampled. This sample included 160 to 260 dogs from each county, where most the times a similar number of urban or rural dogs was sampled (Table 1). Dogs from rural areas were sampled distantly at least 10 km from the urban area of each county. The minimum number of sampled dogs per county was previously calculated to be representative of each county, considering 18.9% estimate prevalence for rickettsial seropositivity , 6% error probability, and 95% confidence level, using Epitools epidemiological calculators (http://epitools.ausvet.com.au/). For this purpose, the canine population was estimated as 10% of the human population to each county. Every sampled dog was owned and was apparently health (with no external signs of clinical diseases) at the moment of sample collections. Canine health assessment was carried out by at least two veterinarians (FBC and APdaC) during sampling of every dog. The two veterinarians certified that dogs had normal behavior (no prostration, no locomotor or nervous disorder) and did not present any external sign of clinical disease (e.g., cutaneous diseases, fracture). If a dog had any sign of disease, it was not included in the present study.
Ticks and hemolymph test
Each dog had its entire body examined for the presence of ticks, which were collected and placed into plastic vials containing a moistened piece of cotton to keep ticks alive until arriving at the laboratory. Ticks (adults and nymphs) were identified to genus and species according to current literature [9,26,27]. Larvae of the genus Amblyomma were not identified to species level due to lack of literature support.
Adult ticks that arrived alive in the laboratory were submitted to the hemolymph test for screening of rickettsial infection, as described . For this purpose, a tick leg was cut and a drop of hemolymph was spread on a glass slide, which was stained by Gimenez staining. The tick was immediately frozen at -80°C for further processing. The slide was examined under an optical microscope with 1000x magnification with oil immersion, which identifies organisms morphologically compatible with bacteria of the genus Rickettsia.
Isolation of Rickettsia in cell culture
Some of the ticks tested by the hemolymph test were randomly selected to be processed by isolation of rickettsiae in cell culture by the shell vial technique, as previously described . Briefly, ticks were surface disinfected for 10 min in iodine alcohol, followed by several washes in sterile water. Then each tick was macerated by using a sterile mortar and pestle in 600μL of brain heart infusion broth. Two shell vials containing confluent monolayers of Vero cells were each inoculated with 300μL of the tick homogenate, and then centrifuged for 1h at 700 g. Thereafter, monolayers were washed and fed with 1 mL of minimal essential medium supplemented 10% with bovine calf serum and antibiotics [1% (each) penicillin, and streptomycin, and 1.5 g/ml amphotericin B), and incubated at 28°C. Every 3 days, the medium was switched to new medium without antibiotics, and the aspirated medium was checked by Giménez staining for the presence of Rickettsia-like organisms. If there were Rickettsia-like organisms, the monolayer of the shell vial was harvested and inoculated into a 25 cm2 flask containing a monolayer of confluent uninfected Vero cells. The level of infection of cells was monitored by Giménez staining of scraped cells from the inoculated monolayer. When ricketsial infection reached >90% of the cells, they were harvested and inoculated into 75-cm 2 flasks of Vero cells. The rickettsial isolate was considered to be established in the laboratory after at least three passages through 75 cm2 flasks, each achieving a proportion >90% of infected cells .
Rickettsia-infected cells of the third passage were submitted to DNA extraction by using the DNeasy Blood and Tissue kit (Qiagen, Chatsworth, CA). The extracted DNA was tested by polymerase chain reaction (PCR) protocols to amplify fragments of the rickettsial genes citrate synthase (gltA; primers CS-78, CS-323), the outer membrane proteins ompA (primers Rr190.70p, Rr190.701) and ompB (primers 190.59F, 190.807R), and the 17-kDa protein (htrA, primers 17k-5 and 17k-3), as described [29–31]. PCR products were purified using ExoSAP-IT (USB Corp., Cleveland, OH) and underwent DNA sequencing in an ABI automated sequencer (Applied Biosystems/Perkin Elmer, model ABI Prism 3500 Genetic, Foster City, CA), and the resultant sequences were compared with GenBank data by BLAST analysis (http://blast.ncbi.nlm.nih.gov/Blast.cgi).
Molecular analysis of ticks
Ticks were individually submitted to DNA extraction by the guanidine isothiocyanate technique, as described elsewhere , and tested by PCR targeting a 401-fragment (primers CS-78, CS-323) of the rickettsial gltA gene . Ticks positive by this PCR assay were tested by a second PCR assay (primers Rr190.70p, Rr190.701) targeting a ≈632 fragment of the SFG rickettsial ompA gene . For each PCR run, a negative control (water) and positive control (Rickettsia parkeri strain NOD) were included. PCR products were DNA sequenced and submitted to BLAST analysis as described above. Body remnants of the shell vial-processed tick specimens were also included in these molecular analyses.
Serological analyses of dogs
From each sampled dog, a blood sample was collected by venipuncture; the serum was separated by centrifugation, and kept frozen at -20°C until tested. Canine sera were tested by immunofluorescence assay (IFA) as described elsewhere  using Vero cells infected with each of the following five Rickettsia species known to infect ticks in Brazil: R. rickettsii strain Taiaçu , R. parkeri strain At24  R. amblyommatis strain Ac37, R. rhipicephali strain HJ5 , and R. bellii strain Mogi  as crude antigens. In each slide, a serum previously shown to be non-reactive (negative control) and a known reactive serum (positive control) from a previous study  were included. Slides were incubated with fluorescein isothiocyanate-labelled rabbit anti-dog IgG (Sigma, St Louis, MO, USA). Sera showing antibodies titers to a Rickettsia species at least fourfold higher than those observed for the other Rickettsia species were supposed to be homologous to the first Rickettsia species or to a very closely related genotype [36,37].
Additional collection of ticks
With the purpose to obtain additional adult ticks to be processed for isolation of rickettsiae in cell culture, Amblyomma cajennense s.l. ticks were collect from horses in Balsas county (July 2011) and from pigs in Viana county (August 2013). Ticks that arrived alive in the laboratory were submitted to the shell vial technique for isolation of rickettsiae in Vero cells, as described above.
The proportions of tick-infested dogs or Rickettsia-seroreactive dogs between urban and rural areas were compared by the chi-square test or Fisher exact test. Canine endpoint titers to the five Rickettsia species were compared by the non-parametric Mann-Whitney test. For each sampled dog, a questionnaire was given to the dog owner with the purpose of gaining information about independent variables that could be associated with seroreactivity to Rickettsia spp. The following independent variables related to dogs were subjected to univariate analysis: living area (urban or rural), age (≤ 1 or > 1 year old), sex, breed (pure or mongrel), hunting activity (yes or no), living place near to forest (yes or no), tick species (only R. sanguineus s.l. or other species). Frequencies of independent variables were compared according to four dependent variables: seroreactivity (titer ≥64) to any of the five Rickettsia antigens; seroreactivity (titer ≥64) to R. amblyommatis; seroreactivity (titer ≥512) to R. amblyommatis; or seroreactivity to R. amblyommatis with endpoint titers at least four-fold higher than the titers to the other four Rickettsia species. Variables with statistical association (P<0.20, chi-square test or Fisher exact test) were tested in the multivariate model by the stepwise forward method. Four models, one for each dependent variable, were created. The variables were included in the multivariate model if they displayed statistical significance of P < 0.05. All analyses were performed using SPSS for Windows version 17.
Characterization of the sampled dogs
A total of 1,560 dogs (737 from the urban areas and 823 from rural areas) were evaluated in this survey. Overall, 919 (58.9%) dogs were males with a proportion of 394 (42.9%) from urban areas and 525 (57.1%) from rural areas. A total of 641 (36.2%) dogs were females, 343 (53.5%) from urban areas and 298 (46.5%) for rural areas. Only 5.7% of the sampled dogs were ≤ 1 year old, and almost all sampled dogs (99.6%) were mongrel. Hunting activity was reported for 213 (13.7%) dogs, from which 32 (15.0%) lived in urban areas and 181 (85.0%) in rural areas.
Ticks on dogs
A total of 959 ticks were collected on 150 dogs (9.6%), being 6.1% (45/737) and 12.8% (105/823) of the urban and rural dogs, respectively. Overall, significantly more rural dogs were infested by ticks than urban dogs (P< 0.05). However, considering each of the eight counties separately, only Açailândia and Caxias had significantly more infested dogs in the rural area; the remaining 6 counties had statistically similar (P> 0.05) proportions of infested dogs among urban and rural areas.
Rhipicephalus sanguineus s.l. was the most abundant tick, with 652 (68%) specimens (307 males, 252 females, 88 nymphs, 5 larvae), followed by A. cajennense s.l. [124 (12.9%) specimens (13 males, 12 females, 99 nymphs)], Amblyomma parvum [88 (9.2%) specimens (17 males, 68 females, 3 nymphs)], and Amblyomma ovale [50 (5.2%) specimens (14 males, 36 females)]. Other less abundant species were Amblyomma oblongoguttatum [9 (0.9%) specimens (2 males, 7 females)], Rhipicephalus microplus [5 (0.5%) females], and Haemaphysalis juxtakochi [1 (0.1%) nymph]. A total of 29 (3%) specimens were retained as Amblyomma spp. One (0.1%) specimen of Amblyomma rotundatum was non-attached to the dog body.
Regarding the A. cajennense s.l. ticks, currently, only the female stage can be morphologically identified with certainty to species level (i.e., A. cajennense s.s. or A. sculptum). Among the 124 specimens of A. cajennense s.l. collected from dogs, only 12 specimens were females. Eight of these specimens were identified as A. cajennense s.s. (1 from Açailândia, 3 from Barreirinhas, and 4 from São Bento), and 4 specimens were identified as A. sculptum (2 from Açailândia and 2 from Balsas).
Considering each tick species separately, the proportions of infested dogs in rural areas were significantly higher (P<0.05) than in urban areas for R. sanguineus s.l., A. cajennense s.l., and A. parvum. The remaining tick species were found in much smaller numbers, and did not show significantly different proportions between urban and rural dogs (Table 1). Among the 150 infested dogs, most of them (120 dogs) were found infested by a single tick species. Twenty-seven dogs had dual infestations (simultaneous infestations by 2 tick species), usually R. sanguineus s.l. with a second tick species; 24 of these dogs were rural, and only 3 were urban. Triple infestations were found on only 2 rural dogs, which were simultaneously infested by R. sanguineus s.l. and 2 Amblyomma species.
Infection and isolation of Rickettsia from ticks
A total of 99 tick specimens belonging to the species A. cajennense s.l., A. ovale, or A. parvum were screened by the hemolymph test, which indicated Rickettsia-like structures in 17 (17.1%) ticks (3 A. ovale and 14 A. parvum). Four of these ticks (2 A. ovale, 2 A. parvum) were tested by the shell vial technique. In addition, we also tested 9 negative-hemolymph ticks (5 A. ovale, 3 A. parvum, 1 A. cajennense s.s) by the shell vial technique, considering that the hemolymph test is not 100% sensitive for detection of rickettsiae in ticks . None rickettsial isolation was established in Vero cells from these ticks.
A total of 779 ticks collected from dogs, including remnants of the 13 specimens processed by the shell vial technique, were tested by PCR targeting a fragment of the rickettsial gltA gene. Rickettsial DNA was detected in 12 A. parvum adults, 3 A. ovale adults, 1 A. rotundatum adult, and 1 A. cajennense s.l. nymph. The 12 A. parvum and the single A. cajennense s.l. nymph also yielded rickettsial DNA by the ompA PCR. Analyses of the DNA sequences generated from these PCR products indicated infection with ‘Candidatus Rickettsia andeanae’ in A. parvum, Rickettsia bellii in A. ovale and A. rotundatum, and R. amblyommatis in A. cajennense s.l. (Table 2). From the body remnants of the 13 tick specimens that were processed by shell vial, only 2 specimens (1 A. ovale, 1 A. parvum, both positive by the hemolymph test) were shown by PCR to contain rickettsiae.
Additional collection of ticks
A total of 41 A. cajennense s.l. adults (5 males, 36 females) were collected from horses in Balsas County, and 93 A. cajennense s.l. adults (68 males, 25 females) were collected from pigs in Viana County (in the southern border of São Bento County). All females from Balsas were identified as A. sculptum, whereas all females from Viana were identified as A. cajennense s.s. A total of 5 A. sculptum and 14 A. cajennense s.s. were processed for isolation of rickettsiae by the shell vial technique. Rickettsial isolates were obtained in Vero cell cultures from 2 A. cajennense s.s. ticks. DNA fragments of the gltA, htrA, ompA and ompB genes of the two isolates were 100% (350/350-bp), 100% (419/419-bp), 99.8% (586/587-bp) and 99.9% (816/817-bp), respectively, identical to corresponding fragments in the genome of R. amblyommatis from Brazil (CP012420). In addition, the ompA fragment was 100% identical to a R. amblyommatis ompA fragment from the western Brazilian Amazon (KM042860), similarly to the ompA fragments amplified from canine ticks in this study (Table 2). However, only one Rickettsial isolate was established in Vero cells for at least 3 passages, since the second isolate was lost due to contamination by extracellular bacteria. Body remnants of the 19 shell vial-inoculated ticks were submitted to DNA extraction and PCR targeting rickettsiae; 5 A. cajennense s.s. ticks (including the 2 specimens that yielded rickettsial isolation) yielded gltA and ompA amplicons, which generated DNA sequences 100% identical to the established isolate of R. amblyommatis.
DNA partial sequences generated in this study have been submitted to GenBank and received the following accession numbers: KY628365-KY628368 for R. amblyommatis gltA, htrA, ompA and ompB; KY628369-KY628370 for ‘Ca. R. andeanae’ gltA and ompA; and KY628371 for R. bellii gltA.
Canine seroprevalence for Rickettsia spp.
A total of 1,560 canine serum samples were tested by IFA employing antigens of five Rickettsia species. Overall, 196 (12.6%) dogs were seroreactive to at least one Rickettsia species, with endpoint titers ranging from 64 to 16,384. The number of sera reacting at the screening dilution (1:64) to each Rickettsia species were 160 (10.2%) for R. amblyommatis, 151 (9.7%) for R. rhipicephali, 66 (4.2%) for R. parkeri, 64 (4.1%) for R. rickettsii, and 57 (3.6%) for R. bellii (Table 3). Endpoint titers to R. amblyommatis (median: 512) was significantly higher (P<0.05) than the endpoint titers to the other four Rickettsia species, for which the median values varied from 128 to 256 (Fig 1).
Different lower case letters mean statistically different (P<0.05) endpoint titers between Rickettsia species.
A total of 73 (4.7%) canine sera from the eight counties showed endpoint titers to R. amblyommatis at least 4-fold higher than those to any of the other four antigens. The antibody titers in these 73 dogs were considered to have been stimulated by R. amblyommatis or a very closely related species (Table 3). Based on this criterion, 12 sera were considered to have been stimulated by R. bellii, 6 by R. rhipicephali, and 1 by R. parkeri. For the remaining 104 seroreactive dogs, it was not possible to discriminate the infection agent, because they displayed similar titers (< 4-fold difference) for two or more Rickettsia species or had a single titer of 64 for a single Rickettsia species.
In three counties (São Bento, Caxias, Balsas), rickettsial seroprevalence values were significantly higher (P<0.05) for rural than for urban dogs; in the remaining counties, seroprevalence values were similar (P>0.05) between urban and rural dogs. Considering the overall seroprevalence in each county, the highest seroprevalence value (30.8%, P<0.05) was found in São Bento in the Amazon biome, followed by Açailândia (18.8%) also in Amazon, Grajaú (14.5%) in Cerrado, and Cururupu (13.1%) in Amazon. The remaining four counties, all located in the Cerrado, had significantly lower (P<0.05) seroprevalence values, ranging from 1.9 to 6.5% (Fig 2).
Pie charts represent the proportions of seropositive and seronegative dogs in urban and rural areas in each locality (Urban dogs +: seropositive dogs in the urban area; Urban dogs -: seronegative dogs in the urban area; Rural dogs +: seropositive dogs in the rural area; Rural dogs -: seronegative dogs in the rural area). Percent numbers above each pie chart represent canine seroreactivity values in urban and rural areas (lower numbers), and total canine seropositivity (top number). Different lower case letters inside parenthesis mean statistically different (P<0.05) total seroreactivity values between the 8 localities. The distribution of Amblyomma cajennense s.s. and Amblyomma sculptum in the map was retrieved from the study of Martins et al..
By the univariate analysis, 4 canine independent variables (living area; age; hunting activity; proximity to forest; or tick species) showed statistical association (P<0.20) with each of the four dependent variables (S1 Table). These independent variables were selected for the multivariate analyses, which revealed that all four dependent variables related to canine seroreactivity to Rickettsia spp. or R. amblyommatis were statistically associated (P<0.05) with dogs living in the rural area, dogs with hunting activity, and dogs that were infested with ticks different from R. sanguineus s.l. (Table 4).
A total of 8 tick species were collected from dogs in the state of Maranhão. This number excludes A. rotundatum because the collected specimen was not attached to dog, and includes two species of A. cajennense s.l. (A. cajennense s.s. and A. sculptum). Tick infestations were more frequent among dogs from rural areas, where there was also a greater diversity of tick species, when compared to dogs from urban areas. These findings are similar to previous studies that compared infestations between urban and rural dogs in Brazil [5–7]. On the other hand, our variety of 8 tick species infesting dogs is one of the greatest diversities found in Brazil, since previous studies encompassing rural dogs from all five major Brazilian biomes have reported 5 to 11 tick species for the Amazon biome [2–4], 3 to 7 tick species for the Cerrado biome [5–7,38,39], 2 to 6 tick species for the Atlantic rainforest biome [40–43], 2 tick species for the Pampa biome , and 1 tick species for the Caatinga biome .
In the present study, Rhipicephalus sanguineus s.l. was the most frequent tick among dogs from either urban or rural areas. Indeed, such finding is expected for urban areas, where human dwellings provide optimal conditions for the off-host development of R. sanguineus s.l. . The predominance of R sanguineus s.l. among rural dogs also agrees with a number of previous studies from different parts of Brazil [5–7,40], although other studies reported higher prevalence of Amblyomma spp. ticks than R. sanguineus s.l. among rural dogs [39,42,43]. Such differences shall be related to canine habits; i.e., as more time dogs spend on human dwellings, higher the chances for establishment of R. sanguineus s.l. populations in rural areas. On the other hand, absence of R. sanguineus s.l. on rural dogs from some Amazonian sites [2,46] could be related to the yet non-introduction of this exotic tick into more remote areas, as discussed elsewhere . Indeed, this is not the case of the state of Maranhão, where R. sanguineus seems to be widespread in rural areas.
Three rickettsial agents were detected in canine ticks: R. amblyommatis in an A. cajennense s.l. nymph, ‘Ca. R. andeanae’ in A. parvum, and R. bellii in A. ovale and A. rotundatum. The R. amblyommatis-infected nymph is likely to belong to the species A. cajennense s.s., since this nymph was collected in Açailândia county (Amazon biome), where A. cajennense s.s. in known to occur. This statement is based on previous studies that reported that R. amblyommatis usually infects 26.8–69.4% of A. cajennense s.s. ticks [16,17], and at the same time, is usually not found in A. sculptum ticks [13–15]. In fact, our attempts to detect or isolate rickettsiae from A. cajennense s.s. and A. sculptum ticks collected from other domestic animals (pigs and horses) were successful only for A. cajennense s.s., from which R. amblyommatis was detected/isolated in 35.7% (5/14) of the ticks. Our findings of ‘Ca. R. andeanae’ infecting 20.7% of the A. parvum ticks is concordant with previous studies from Brazil and Argentina, where A. parvum ticks were commonly found infected by this rickettsial agent [47,48]. Finally, the presence of R. bellii in A. ovale or A. rotundatum ticks has also been previously reported in Brazil [16,25,42,49]. No rickettsial infections were detected in any of 541 R. sanguineus s.l. attached to dogs. These findings contrast to studies in southeastern where a few R. sanguineus s.l. ticks collected from dogs were infected by the human pathogens R. rickettsii or Rickettsia sp. strain Atlantic rainforest [42,43,50]. Noteworthy, these other studies were from spotted fever-endemic areas where R. rickettsii or Rickettsia sp. strain Atlantic rainforest were primarily associated with canine ticks of the genus Amblyomma, what could have facilitated horizontal transmission between different tick species via dogs.
Overall, 196 (12.6%) dogs were seroreactive to at least one Rickettsia species, with the highest specific seroreactivity rate (10.2%) for R. amblyommatis (Table 3). Even though serological cross-reaction is very common between SFG Rickettsia species in dogs [36,51], 73 dogs (37.2% of all 196 seroreactive dogs) presented titers to R. amblyommatis at least four-fold higher than the titers to the other five Rickettsia antigens that were tested. Indeed, the overall endpoint titers to R. amblyommatis were significantly higher than those for the other Rickettsia species (Fig 1). These results suggest that most of seroreactive dogs of the present study have been exposed to R. amblyommatis-infected ticks. In fact, the R. amblyommatis-seroreactive dogs were statistically associated with infestation by ticks different from R. sanguineus (Table 4); e.g., A. cajennense s.l. ticks, the most frequent Amblyomma ticks, which were found infected by R. amblyommatis. In addition, considering the 8 sites of the present study, the highest canine seroreactivity rates (13.1 to 30.8%) were found in the western part of Maranhão state, composed basically by the Amazon biome, and consequently, with the presence/predominance of A. cajennense s.s. (Fig 2), which is naturally found infected by R. amblyommatis ([16,17], present study). On the other hand, the lowest seroreactivity rates (1.9 to 6.5%) were found in the eastern part of Maranhão state (Cerrado biome), where A. sculptum predominates (Fig 2). All these above statements provide strong epidemiological evidence for the natural infection of dogs, via tick transmission, by R. amblyommatis in the state of Maranhão. This evidence is also corroborated by a laboratory study showing that R. amblyommatis is tick transmitted . In addition, one study in the United States showed canine seroconversion to R. amblyommatis after natural exposure to Amblyomma americanum ticks, which are frequently found infected by this rickettsial agent under high infection rates; these seroconverted dogs also showed much higher endpoint titers to R. amblyommatis than to other two SFG agents .
The third most abundant tick species in the present study, A. parvum, is a relatively common canine tick in the Cerrado biome [38,52], including in the present study (Table 1, Fig 2), where ‘Ca. R. andeanae’ was detected in 20.7% of the A. parvum ticks. While there is a possibility that canine exposure to ‘Ca. R. andeanae’-infected ticks could have interfered in our serological results, this possibility is unlikely considering that one study demonstrated that ‘Ca. R. andeanae’ is not efficiently transmitted to the host skin during tick feeding . Finally, our serological results also suggest that a few dogs could have been infected by R. bellii, R. rhipicephali, or R. parkeri (Table 3). These three rickettsial agents have been isolated from some of the tick species that were collected on dogs in the present study; e.g., R. bellii from A. ovale ([42,43], present study), R. rhipicephali from H. juxtakochi , and R. parkeri-like agent (strain Atlantic rainforest) from A. ovale [42,43]. Due to low number of dogs with highest endpoint titers to these 3 rickettsial agents in the present study, more epidemiological background is required to support canine exposure by these agents in Maranhão. Finally, while A. sculptum is a natural vector of R. rickettsii in southeastern Brazil [12,11], our results showed no evidence for the circulation of this pathogen among domestic dogs, even though they have been considered efficient sentinels for R. rickettsii [13,22,36].
This study evaluated ticks, rickettsial infection in ticks, and exposure of dogs to Rickettsia spp. in an Amazon/Cerrado transition area of Brazil. Our results showed that the dogs, especially from the rural areas, are exposed to a great diversity of tick species, some of them carrying Rickettsia organisms. Dogs living in the Amazon biome of Maranhão (western part of the state) are much more likely to be seropositive to R. amblyommatis than dogs living in the Cerrado biome (eastern part of the state). This scenario is indeed related to the distribution of A. cajennense s.l. ticks in Maranhão, represented by A. cajennense s.s. ticks (usually infected by R. amblyommatis) mostly in the Amazon, and A. sculptum (usually not infected by R. amblyommatis) mostly in the Cerrado. Since R amblyommatis is suspected to be a human pathogen, surveillance of human infection by this agent should be directed to Amazon areas. On the other hand, if BSF (caused by R. rickettsii) is to be detected in Maranhão state someday, it is more likely to occur in the eastern part (Cerrado), where the natural vector (A. sculptum) predominates and there would be less interference by another SFG agent such as R. amblyommatis. This statement is based on previous studies that have shown that ticks cannot simultaneously maintain separate species of Rickettsia by vertical transmission, as recently discussed .
S1 Table. Results of univariate analysis (Fisher's exact test) for the association between independent variables with the serological results of domestic dogs, analyzed through four serologic profiles (four dependent variables) determined by the immunofluorescence assay: (i) canine seroreactivity to Rickettsia spp. (titer ≥64); (ii) canine seroreactivity to R. amblyommatis (titer ≥64); (iii) canine seroreactivity to R. amblyommatis (titer ≥512); (iv) canine seroreactivity to R. amblyommatis with titers at least 4-fold higher than the titers for other Rickettsia species.
We are grateful to Tatiana Ueno and Amália Barbieri for their help in performing tick hemolymph tests. This work received financial support by the São Paulo Research Foundation (FAPESP grant # 2010/52395-0).
- Conceptualization: FBC APdaC RAD MBL.
- Data curation: FBC APdaC JM-F TFM HSS DGR RAD MBL.
- Formal analysis: FBC RAD MBL.
- Funding acquisition: RAD MBL.
- Investigation: FBC APdaC JM-F TFM HSS DGR MBL.
- Methodology: FBC MBL.
- Project administration: FBC APdaC MBL.
- Resources: RAD MBL.
- Supervision: MBL.
- Validation: FBC MBL.
- Visualization: FBC APdaC RAD MBL.
- Writing – original draft: FBC MBL.
- Writing – review & editing: FBC APdaC JM-F TFM HSS DGR RAD MBL.
- 1. Myers N, Mittermeier RA, Mittermeier CG, da Fonseca GA, Kent J, Fonseca GAB. Biodiversity hotspots for conservation priorities. Nature. 2000;403: 853–8. pmid:10706275
- 2. Labruna MB, Homem VS, Heinemann MB, Ferreira Neto JS. Ticks (Acari: Ixodidae) associated with rural dogs in Uruará, eastern Amazon, Brazil. J Med Entomol. 2000;37: 774–6. Available: http://www.ncbi.nlm.nih.gov/pubmed/11004794 pmid:11004794
- 3. Labruna MB, Camargo LMA, Terrassini FA, Ferreira F, Schumaker TTS, Camargo EP. Ticks (Acari: Ixodidae) from the state of Rondonia, western Amazon, Brazil. Syst Appl Acarol. 2005;10: 17–32.
- 4. Minervino AHH, Lima JTR, Soares HS, Malheiros AF, Marcili A, Krawczak F da S, et al. Seroprevalence of Tick-Borne Pathogens and Tick Infestation in Dogs from Tapirapé and Karajá Indigenous Communities, Brazil. Vector Borne Zoonotic Dis. 2015;15: 412–8. pmid:26186512
- 5. Szabó MPJ, Marquez Cunha T, Pinter A, Vicentini F. Ticks (Acari: Ixodidae) Associated with Domestic Dogs in Franca Region, São Paulo, Brazil. Exp Appl Acarol. Kluwer Academic Publishers; 2001;25: 909–916. pmid:12455880
- 6. Szabó MPJ, de Souza LGA, Olegário MMM, Ferreira FA, de Albuquerque Pajuaba Neto A. Ticks (Acari: Ixodidae) on dogs from Uberlândia, Minas Gerais, Brazil. Transbound Emerg Dis. 2010;57: 72–74. pmid:20537111
- 7. Costa A, Silva A, Costa F, Xavier G, Martins T, Labruna M, et al. A survey of ectoparasites infesting urban and rural dogs of Maranhão state, Brazil. J Med Entomol. 2013;50: 674–678. pmid:23802466
- 8. Nava S, Beati L, Labruna MB, Cáceres AG, Mangold AJ, Guglielmone AA. Reassessment of the taxonomic status of Amblyomma cajennense (Fabricius, 1787) with the description of three new species, Amblyomma tonelliae n. sp., Amblyomma interandinum n. sp. and Amblyomma patinoi n. sp., and reinstatement of Amblyomma mixtum Koch, 1. Ticks Tick Borne Dis. Elsevier GmbH.; 2014;5: 252–276. pmid:24556273
- 9. Martins TF, Barbieri ARM, Costa FB, Terassini FA, Camargo LMA, Peterka CRL, et al. Geographical distribution of Amblyomma cajennense (sensu lato) ticks (Parasitiformes: Ixodidae) in Brazil, with description of the nymph of A. cajennense (sensu stricto). Parasit Vectors. Parasites & Vectors; 2016;9: 186. pmid:27036324
- 10. Labruna MB, Santos FCP, Ogrzewalska M, Nascimento EMM, Colombo S, Marcili A, et al. Genetic identification of Rickettsial isolates from fatal cases of Brazilian spotted fever and comparison with Rickettsia rickettsii isolates from the American continents. J Clin Microbiol. 2014;52: 3788–3791. pmid:25078908
- 11. Krawczak FS, Nieri-Bastos FA, Nunes FP, Soares JF, Moraes-Filho J, Labruna MB. Rickettsial infection in Amblyomma cajennense ticks and capybaras (Hydrochoerus hydrochaeris) in a Brazilian spotted fever-endemic area. Parasit Vectors. 2014;7: 7. pmid:24387674
- 12. Labruna MB, Krawczak FS, Gerardi M, Binder LC, Barbieri ARM, Paz GF, et al. Isolation of Rickettsia rickettsii from the tick Amblyomma sculptum from a Brazilian spotted fever-endemic area in the Pampulha Lake region, southeastern Brazil. Vet Parasitol Reg. 2017;Stud Rep (.
- 13. Sangioni LA, Horta MC, Vianna MCB, Gennari SM, Soares RM, Galvão MAM, et al. Rickettsial infection in animals and Brazilian spotted fever endemicity. Emerg Infect Dis. 2005;11: 265–270. pmid:15752445
- 14. Vianna MCB, Horta MC, Sangioni LA, Cortez A, Soares RM, Mafra CL, et al. Rickettsial spotted fever in Capoeirão village, Itabira, Minas Gerais, Brazil. Rev Inst Med Trop Sao Paulo. 2008;50: 297–301. pmid:18949349
- 15. De Campos Pacheco R, Horta MC, Pinter A, Moraes-Filho J, Martins TF, Nardi MS, et al. Pesquisa de Rickettsia spp em carrapatos Amblyomma cajennense e Amblyomma dubitatum no Estado de São Paulo. Rev Soc Bras Med Trop. 2009;42: 351–353. pmid:19684990
- 16. Labruna MB, Whitworth T, Bouyer DH, McBride J, Camargo LMA, Camargo EP, et al. Rickettsia bellii and Rickettsia amblyommii in Amblyomma ticks from the State of Rondônia, Western Amazon, Brazil. J Med Entomol. 2004;41: 1073–1081. pmid:15605647
- 17. Soares HS, Barbieri ARM, Martins TF, Minervino AHH, de Lima JTR, Marcili A, et al. Ticks and rickettsial infection in the wildlife of two regions of the Brazilian Amazon. Exp Appl Acarol. 2014;65: 125–140. pmid:25273064
- 18. Karpathy Sandor E., Slater Kimetha S., Goldsmith CS, Nicholson WL, Paddock CD. Rickettsia amblyommatis sp. nov.,a spotted fever group rickettsia associated with multiple species of Amblyomma ticks in North and South America. Int J Syst Evol Microbiol. 2016;66: 5236–5243. pmid:27638476
- 19. Apperson CS, Engber B, Nicholson WL, Mead DG, Engel J, Yabsley MJ, et al. Tick-Borne Diseases in North Carolina: Is “Rickettsia amblyommii” a Possible Cause of Rickettsiosis Reported as Rocky Mountain Spotted Fever? Vector-Borne Zoonotic Dis. 2008;8: 597–606. pmid:18447622
- 20. Delisle J, Mendell NL, Stull-Lane A, Bloch KC, Bouyer DH, Moncayo AC. Human Infections by Multiple Spotted Fever Group Rickettsiae in Tennessee. Am J Trop Med Hyg. 2016;94: 1212–1217. pmid:27022147
- 21. Saraiva DG, Nieri-Bastos FA, Horta MC, Soares HS, Nicola PA, Pereira LCM, et al. Rickettsia amblyommii Infecting Amblyomma auricularium Ticks in Pernambuco, Northeastern Brazil: Isolation, Transovarial Transmission, and Transstadial Perpetuation. Vector-Borne Zoonotic Dis. 2013;13: 615–618. pmid:23705586
- 22. Horta MC, Labruna MB, Pinter A, Linardi PM, Schumaker TTS. Rickettsia infection in five areas of the state of São Paulo, Brazil. Mem Inst Oswaldo Cruz. 2007;102: 793–801. pmid:18094887
- 23. Pinter A, Horta MC, Pacheco RC, Moraes-Filho J, Labruna MB. Serosurvey of Rickettsia spp. in dogs and humans from an endemic area for Brazilian spotted fever in the State of São Paulo, Brazil. Cad Saude Publica. 2008;24: 247–252. pmid:18278271
- 24. GEPLAN/ UEMA. Atlas do Maranhão/Gerência de Estado de Planejamento e Desenvolvimento Econômico. Laboratório de Geoprocessamento-UEMA, editor. São Luís, Maranhão; 2002.
- 25. Costa AP da, Costa FB, Labruna MB, Silveira I, Moraes-Filho J, Soares JF, et al. A serological and molecular survey of Babesia vogeli, Ehrlichia canis and Rickettsia spp. among dogs in the state of Maranhão, northeastern Brazil. Brazilian J Vet Parasitol. 2015;24: 28–35. pmid:25909250
- 26. Barros-Batesti DM, Arzua M, Bechara GH. Carrapatos de importância médico-veterinária da região neotropical : um guia ilustrado para idenfiticação de espécies [Internet]. São Paulo: Vox/ICTTD-3/Instituto Butantan; 2006. Available: https://books.google.com.br/books?id=XfBAAAAACAAJ
- 27. Martins TF, Onofrio VC, Barros-Battesti DM, Labruna MB. Nymphs of the genus Amblyomma (Acari: Ixodidae) of Brazil: descriptions, redescriptions, and identification key. Ticks Tick Borne Dis. 2010;1: 75–99. pmid:21771514
- 28. Burgdorfer W. Hemolymph test a technique for detection of Rickettsiae. 1970;19: 1010–1014. pmid:4992724
- 29. Labruna MB, Whitworth T, Horta MC, Bouyer DH, McBride JW, Pinter A, et al. Rickettsia species infecting Amblyomma cooperi ticks from an area in the state of São Paulo, Brazil, where Brazilian spotted fever is endemic. J Clin Microbiol. 2004;42: 90–98. pmid:14715737
- 30. Roux V, Fournier P, Raoult D. Differentiation of spotted fever group rickettsiae by sequencing and analysis of restriction fragment length polymorphism of PCR-amplified DNA of the gene encoding. J Clin Microbiol. 1996;34: 2058–2065. Available: http://jcm.asm.org/content/34/9/2058.short pmid:8862558
- 31. Roux V, Raoult D. Phylogenetic analysis of members of the genus Rickettsia using the gene encoding the outer-membrane protein rOmp (ompB). Int J Syst Evol Microbiol. 2000;50: 1449–1455. pmid:10939649
- 32. Horta MC, Labruna MB, Sangioni LA, Vianna MCB, Gennari SM, Galvão MAM, et al. Prevalence of antibodies to spotted fever group rickettsiae in humans and domestic animals in a Brazilian spotted fever-endemic area in the state of São Paulo, Brazil: Serologic evidence for infection by Rickettsia rickettsii and another spotted fever gr. Am J Trop Med Hyg. 2004;71: 93–97. 71/1/93 [pii] pmid:15238696
- 33. Pinter A, Labruna MB. Isolation of Rickettsia rickettsii and Rickettsia bellii in cell culture from the tick Amblyomma aureolatum in Brazil. Ann N Y Acad Sci. 2006;1078: 523–529. pmid:17114770
- 34. Silveira I, Pacheco RC, Szabó MPJ, Ramos HGC, Labruna MB. Rickettsia parkeri in Brazil. Emerg Infect Dis. 2007;13: 1111–1113. pmid:18214195
- 35. Labruna MB, Pacheco RC, Richtzenhain LJ, Szabó MPJ, Szabo MP. Isolation of Rickettsia rhipicephali and Rickettsia bellii from Haemaphysalis juxtakochi ticks in the state of São Paulo, Brazil. Appl Environ Microbiol. 2007;73: 869–73. pmid:17142361
- 36. Piranda EM, Faccini JLH, Pinter A, Saito TB, Pacheco RC, Hagiwara MK, et al. Experimental infection of dogs with a Brazilian strain of Rickettsia rickettsii: clinical and laboratory findings. Mem Inst Oswaldo Cruz. 2008;103: 696–701. pmid:19057821
- 37. Saito TB, Cunha-Filho NA, Pacheco RC, Ferreira F, Pappen FG, Farias NAR, et al. Canine infection by rickettsiae and ehrlichiae in southern Brazil. Am J Trop Med Hyg. 2008;79: 102–108. 79/1/102 [pii] pmid:18606772
- 38. Martins TF, Spolidorio MG, Batista TCA, Oliveira IAS, Yoshinari NH, Labruna MB. Occurrence of ticks (Acari: Ixodidae) in the municipality of Goiatins, Tocantins / [Ocorrência de carrapatos (Acari: Ixodidae) no município de Goiatins,Tocantins]. Rev Bras Parasitol Veterinária Brazilian J Vet Parasitol. 2009;2961: 50–52.
- 39. Costa-Junior LM, Rembeck K, Mendonça FLDM, Azevedo SC, Passos LMF, Ribeiro MFB. Occurrence of ectoparasites on dogs in rural regions of the state of Minas Gerais, Brazil. Brazilian J Vet Parasitol. 2012;21: 237–242.
- 40. Labruna M, Pereira M. Carrapato em Cães no Brasil. Clínica Veterinária. 2001. pp. 24–32.
- 41. Dantas-Torres F, Melo MF, Figueredo LA, Brandão-Filho SP. Ectoparasite infestation on rural dogs in the municipality of São Vicente Férrer, Pernambuco, Northeastern Brazil. Rev Bras Parasitol Vet. 2009;18: 75–7. pmid:19772781
- 42. Sabatini GS, Pinter A, Nieri-Bastos FA, Marcili A, Labruna MB. Survey of ticks (Acari: Ixodidae) and their rickettsia in an Atlantic rain forest reserve in the State of São Paulo, Brazil. J Med Entomol. 2010;47: 913–916. pmid:20939390
- 43. Szabó MPJ, Nieri-Bastos FA, Spolidorio MG, Martins TF, Barbieri AM, Labruna MB. In vitro isolation from Amblyomma ovale (Acari: Ixodidae) and ecological aspects of the Atlantic rainforest Rickettsia, the causative agent of a novel spotted fever rickettsiosis in Brazil. Parasitology. 2013;140: 719–28. pmid:23363571
- 44. Weck B, Dall’ Agnol B, Souza U, Webster A, Stenzel B, Klafke G, et al. Spotted fever group Rickettsia in the Pampa Biome, Brazil, 2015???2016. Emerg Infect Dis. 2016;22: 2014–2016. pmid:27767913
- 45. Araes-Santos AI, Moraes-Filho J, Peixoto RM, Spolidorio MG, Azevedo SS, Costa MM, et al. Ectoparasite Infestations and Canine Infection by Rickettsiae and Ehrlichiae in a Semi-Arid Region of Northeastern Brazil. Vector Borne Zoonotic Dis. 2015;15: 645–51. pmid:26565771
- 46. Soares HS, Camargo LMA, Gennari SM, Labruna MB. Survey of canine tick-borne diseases in Lábrea, Brazilian Amazon: “accidental” findings of Dirofilaria immitis infection. Rev Bras Parasitol Veterinária. 2014;23: 473–480. pmid:25517525
- 47. Pacheco RC, Moraes-Filho J, Nava S, Brandão PE, Richtzenhain LJ, Labruna MB. Detection of a novel spotted fever group rickettsia in Amblyomma parvum ticks (Acari: Ixodidae) from Argentina. Exp Appl Acarol. 2007;43: 63–71. pmid:17768597
- 48. Nieri-Bastos FA, Lopes MG, Cançado PHD, Rossa GAR, Faccini JLH, Gennari SM, et al. Candidatus Rickettsia andeanae, a spotted fever group agent infecting Amblyomma parvum ticks in two Brazilian biomes. Mem Inst Oswaldo Cruz. 2014;109: 259–261. pmid:24714968
- 49. Horta MC, Saraiva DG, Oliveira GMB, Martins TF, Labruna MB. Rickettsia bellii in Amblyomma rotundatum ticks parasitizing Rhinella jimi from northeastern Brazil. Microbes Infect. 2015;17: 856–858. pmid:26344601
- 50. Ogrzewalska M, Saraiva DG, Moraes-Filho J, Martins TF, Costa FB, Pinter A, et al. Epidemiology of Brazilian spotted fever in the Atlantic Forest, state of São Paulo, Brazil. Parasitology. 2012;139:1283–300. pmid:22716923
- 51. Barrett A, Little SE, Shaw E. “Rickettsia amblyommii” and R. montanensis Infection in Dogs Following Natural Exposure to Ticks. Vector-Borne Zoonotic Dis. 2014;14: 20–25. pmid:24359419
- 52. Szabó MPJ, Olegário MMM, Santos ALQ. Tick fauna from two locations in the Brazilian savannah. Exp Appl Acarol. 2007;43: 73–84. pmid:17828441
- 53. Grasperge BJ, Morgan TW, Paddock CD, Peterson KE, Macaluso KR. Feeding by Amblyomma maculatum (Acari: Ixodidae) enhances Rickettsia parkeri (Rickettsiales: Rickettsiaceae) infection in the skin. J Med Entomol. 2014;51: 855–863. pmid:25118419
- 54. Paddock CD, Allerdice ME, Karpathy SE, Nicholson WL, Levin ML, Smith TC, et al. Isolation and characterization of a unique strain of Rickettsia parkeri associated with the hard tick Dermacentor parumapertus Neumann in the western United States. Appl Environ Microbiol. 2017; pii: AEM.03463-16.