Environmental infestation and rickettsial infection in ticks in an area endemic for Brazilian spotted fever

Brazilian spotted fever (BSF), caused by Rickettsia rickettsii, is endemic in the municipality of Americana, southeastern Brazil, where the disease is transmitted by the tick Amblyomma cajennense. This study evaluated the tick fauna and rickettsial infection in free-living ticks that were captured monthly using dry ice traps in areas endemic for BSF in Americana, from July 2009 to June 2010. Two tick species were captured: A. cajennense (6,122 larvae; 4,265 nymphs; 2,355 adults) and Amblyomma dubitatum (7,814 larvae; 3,364 nymphs; 1,193 adults). The immature stages of A. cajennense and A. dubitatum had similar distribution through the 12-month period, with larvae of both species collected in highest numbers between April and July, and nymphs between June and October. The highest numbers of A. cajennense adults were collected between October and December, whereas A. dubitatum adults were collected in relatively similar numbers throughout the 12-month period. Rickettsial infection was evaluated by means of PCR in 1,157 A. cajennense and 1,040 A. dubitatum ticks; only 41 (3.9%) A. dubitatum were found to be infected by Rickettsia bellii. The present study showed that the areas of Americana that are endemic for BSF are characterized by high environmental burdens of A. cajennense and A. dubitatum.


Introduction
The bacterium Rickettsia rickettsii is the etiological agent for the deadliest form of rickettsiosis in the world, namely Brazilian spotted fever (BSF) (LABRUNA, 2009).This disease is endemic in southeastern Brazil, especially in the state of São Paulo, where 555 laboratory-confirmed cases occurred from 1985 to 2012, with a 40% case-fatality rate (official data from the São Paulo State Health Office available at http://www.cve.saude.sp.gov.br/).BSF is transmitted by ticks.In Brazil, Amblyomma cajennense is the most important vector, since it is incriminated in transmitting R. rickettsii to humans in most of the endemic areas, including rural areas in the interior of the state of São Paulo (LABRUNA, 2009;PINTER et al., 2011).In these endemic areas, capybaras (Hydrochoerus hydrochaeris) and domestic horses are the main hosts for all parasitic stages of A. cajennense.In addition, capybaras are also considered to be amplifier hosts for R. rickettsii, which means that they are responsible for generating new lineages of infected ticks in endemic areas (SOUZA et al., 2009;LABRUNA, 2009;SOARES et al., 2012).Besides R. rickettsii, a number of other Rickettsia species have been reported in Brazil, mostly infecting only ticks, as is the case of Rickettsia bellii, the most common species of the genus in Brazilian ticks.Until now, R. bellii has been reported infecting 11 different tick species in Brazil; however, this Rickettsia species is considered to be non-pathogenic (LABRUNA et al., 2011).
Americana is a municipality located in the eastern part of the state of São Paulo.It has a population of 212,791 inhabitants, mostly living in a 92 km 2 urban area surrounded and crossed by water courses, namely the Salto Grande reservoir (9.3 km 2 ) and four rivers: Piracicaba, Jaguari, Atibaia and Ribeirão Quilombo (FELICIANO, 2012).From 2004 to 2012, there were ten confirmed cases of BSF with a 60% fatality rate in Americana, thus indicating endemicity for BSF.All these cases were related to tick bites acquired along the water courses of the municipality, where there are established populations of capybaras (BRITES-NETO, 2011).Therefore, the present study aimed to evaluate the tick fauna of these risk areas, and to investigate rickettsial infection in these ticks.

Materials and Methods
From July 2009 to June 2010, free-living ticks were collected by means of dry ice traps, as previously described (WILSON et al., 1972;OLIVEIRA et al., 2000), in the following areas within the municipality of Americana, state of São Paulo, Brazil: 1-Sobrado Velho (22°41'16"S and 47° 15' 09" W,516 m);496 m);534 m);558 m);508 m);528 m).All of these six areas (1 to 6) were inhabited by capybaras, whereas horses were present only in areas 1, 2, and 4. Human parasitism by ticks had frequently been reported in these six areas.Confirmed cases of BSF have been reported in area 1 (5 cases; 4 deaths) and area 2 (1 case) during the last nine years.
In each area, 10 to 13 traps were mounted every month for 12 consecutive months, totaling 830 traps for the whole study.The ticks collected were counted and taken to the laboratory, where they were kept frozen at -20 °C until further testing.The adult ticks were taxonomically identified in accordance with Onofrio et al. (2006), whereas nymphs were identified as described by Martins et al. (2010).Larvae were identified to species level by means of morphological comparisons with laboratory-reared larvae.
The frozen adult ticks were thawed and their salivary glands were separated through dissection under a stereomicroscope, as previously described (EDWARDS et al., 2009).The salivary glands of each tick were subjected to DNA extraction using the guanidine isothiocyanate-phenol technique, as previously described (SANGIONI et al., 2005).For every 10 individual ticks, a blank tube was included in the DNA extraction.Samples were tested individually by means of Rickettsia genus-specific PCR, targeting a 401-bp fragment of the rickettsial gene gltA, as previously described (LABRUNA et al., 2004).
Samples that yielded visible amplicons of the expected size from this gltA-PCR were further tested by two other PCR assays: (i) one assay targeting spotted fever group Rickettsia, using the primers Rr190.70p and Rr190.602n,which amplified a 532-bp fragment of the rickettsial gene ompA (REGNERY et al., 1991); and (ii) another assay specific to Rickettsia bellii, using the primers 5′-ATCCTGATTTGCTGAATTTTTT-3′ (forward) and 5′-TGCAATACCAGTACTGACG-3′) (reverse), which amplified a 338-bp fragment of the R. bellii gltA gene (SZABÓ et al., 2013).A random sample of amplicons generated by the Rickettsia genusspecific PCR protocol was subjected to direct DNA sequencing using an automated ABI Prism 310 Genetic Analyzer (Applied Biosystems, Foster City, CA, USA).The BLAST software (National Center for Biotechnology Information, Bethesda, MD, USA) was used to determine the similarities of the partial rickettsial sequences generated in the current study.In each PCR run, two negative control tubes containing water were included, and also a positive control tube containing R. parkeri DNA.
The proportions of the numbers of adult ticks of the species A. cajennense and A. dubitatum collected among the six areas were compared, taking the null hypothesis to be that these ticks were represented by equal proportions (50% each) in each sampled area.Analyses were performed using the Minitab software, release 16.

Results
Over the 12-month period, a total of 3,548 adult ticks (1,649 males and 1,899 females) and 21,565 immature ticks (7,629 nymphs and 13,936 larvae) were collected in the 830 dry ice traps mounted in the six sampled areas.Two tick species, A. cajennense and A. dubitatum, were collected in all sampled areas (Table 1).Adults of A. cajennense and A. dubitatum were collected in similar numbers, except in areas 2 and 4, where significantly more A. cajennense were collected.Taxonomic differentiation of unfed larvae of A. cajennense and A. dubitatum under a stereoscope microscope relied basically on the idiosome size, which was visually larger in A. dubitatum (Figure 1).The numbers of ticks collected in the six areas were pooled and are presented in Figure 2. Overall, the immature stages of A. cajennense and A. dubitatum had similar distribution throughout the 12-month period, with larvae of both species collected in highest numbers during the autumn and early winter (April to July), and nymphs during late autumn, winter and early spring (June to October).On the other hand, the highest numbers of A. cajennense adult ticks were collected during the spring and early summer months (October to December), whereas A. dubitatum adult ticks were collected in relatively similar numbers throughout the 12-month period (Figure 2).
The salivary glands of 2,197 adult ticks were subjected to DNA extraction and PCR.Among the 1,157 A. cajennense ticks, none was positive according to the initial gltA-PCR protocol.Among the 1,040 A. dubitatum ticks, 41 (3.9%) were positive according to the initial gltA-PCR, negative according to the ompA-PCR and positive according to the R. bellii-specific PCR protocol.These PCR positive samples were found in 5 out of the 6 sampled areas, with infection rates varying from 1.3 to 8.5% (Table 2).PCR products were randomly selected from 10 of these ticks, and were subjected to DNA sequencing.The 10 ticks generated sequences that were 100% identical to the corresponding sequence of R. bellii in GenBank (accession number CP000087).

Discussion
This study was conducted in six areas of the municipality of Americana, which has been considered to be an area endemic for BSF since 2004 (PINTER et al., 2011).It was found that all the active stages of A. cajennense and A. dubitatum were abundant in the six sampled areas, which were also all inhabited by freeranging capybaras.Indeed, the environmental tick burdens found in the present study are directly related to capybaras, which are primary hosts for all the parasitic stages of both A. cajennense and A. dubitatum (PEREZ et al., 2008;NAVA et al., 2010;LABRUNA, 2013).Horses, another primary host species for all parasitic stages of A. cajennense (LABRUNA et al., 2002), were also present in areas 1, 2 and 4. Interestingly, areas 2 and 4 were the only ones where significantly greater numbers of A. cajennense than of A. dubitatum were collected.It is possible that this higher A. cajennense burden was related to higher host availability, namely horses and capybaras.Unfortunately, we could not quantify the populations of capybaras or horses in the present study, thus   precluding any comparative analysis of host density and tick burdens.Interestingly, the six sampled areas were interconnected by water courses, in which capybaras were present and to which they had free access (data not shown).Therefore, it is possible that constant gene (and pathogen) exchange exists between ticks in the six sampled areas.Nonetheless, the present study corroborates a number of previous studies that have reported the presence of capybaras associated with the ticks A. cajennense and A. dubitatum in southeastern Brazil, in both BSF-endemic and non-endemic areas (GUEDES et al., 2005;PEREZ et al., 2008;TOLEDO et al., 2008;PACHECO et al., 2009;QUEIROGAS et al., 2012).The seasonal dynamics of A. cajennense has been well studied in southeastern Brazil, where this tick completes one generation per year, with larvae predominating in autumn, nymphs in winter and adults during spring and summer (OLIVEIRA et al., 2000(OLIVEIRA et al., , 2003;;LABRUNA et al., 2002).This seasonal pattern is determined by the behavioral diapause of unfed larvae, as regulated by photoperiod and ground temperature (LABRUNA et al., 2003;CABRERA;LABRUNA, 2009).In the present study, albeit encompassing only a 12 month period, the highest peaks of larvae, nymphs and adults of A. cajennense followed the well known seasonal dynamics of this tick in southeastern Brazil.
Three previous studies evaluated the seasonal dynamics of free-living A. cajennense and A. dubitatum for two consecutive years in areas ecologically similar to the present study.i.e. with capybaras sustaining simultaneous populations of A. cajennense and A. dubitatum.In Jaguariúna, state of São Paulo, Souza et al. (2006) observed A. dubitatum and A. cajennense adults peaking during spring and summer; immature stages were not identified to species level.On the border between the states of São Paulo and Mato Grosso do Sul, Szabó et al. (2007) observed A. dubitatum and A. cajennense nymphs during winter and spring, and adults peaking during spring to autumn.Only three larval clusters were identified, thus precluding any seasonal inferences for this stage.In the state of Minas Gerais, Guedes and Leite (2008) observed A. dubitatum and A. cajennense adults peaking during spring and summer; again, immature stages were not identified to species level.On the other hand, although the present study only encompassed a 12-month period, it quantified not only adults but also larvae and nymphs of both A. dubitatum and A. cajennense ticks.In this, we observed that the larval and nymphal peaks of A. dubitatum were congruent with those of immature states of A. cajennense, i.e. larvae peaking in autumn and nymphs in winter.Therefore, our results, together with those previous studies, suggest that the seasonal dynamics of A. dubitatum are similar to those of A. cajennense in southeastern Brazil.
While no Rickettsia species was found infecting A. cajennense ticks in the present study, R. bellii was found infecting nearly 4% of A. dubitatum adult ticks.At first sight, this result is congruent with an extensive study (PACHECO et al., 2009) that evaluated 3,545 A. cajennense and 2,666 A. dubitatum ticks from 16 municipalities (some endemic for BSF) in the state of São Paulo, in which none of the A. cajennense specimens were found to be infected by Rickettsia, and 634 (23.8%) A. dubitatum ticks were infected by R. bellii, with infection rates per municipality ranging from 6.1 to 44.9%.However, looking at the data more closely, the present study reports a relatively low infection rate for R. bellii, compared with the results of Pacheco et al. (2009).In contrast with Pacheco et al. (2009), who extracted DNA from the whole tick body, here we performed DNA extraction directly from the tick salivary glands.This procedure has the advantage of eliminating PCR inhibitors (e.g.hemoglobin derivatives) that are potentially present in the tick gut contents (KREADER, 1996), and at the same time, it provides a more reliable result relating to rickettsial infection, since we excluded the possibility of rickettsial DNA remnants in the tick gut, derived from a previous blood meal.On the other hand, we would have missed any Rickettsia species that did not infect the salivary glands and hemolymph, as is the case of infection by Rickettsia peacockii in Dermacentor andersoni ticks in the United States (NIEBYLSKI et al., 1997).R. bellii is known to infect A. dubitatum hemolymph (LABRUNA et al., 2004), but it is not known whether it infects the salivary glands.If it does not infect the salivary glands, then our findings of PCRpositive ticks would have resulted from hemolymph residues that were inevitably collected with the salivary glands.Thus, our low R. bellii-infection rate may have been related to lower tropism of R. bellii to tick salivary glands, which has yet to be demonstrated.
In two recent studies in an area endemic for BSF in the state of Minas Gerais, Guedes et al. (2005Guedes et al. ( , 2011) ) found that the proportions of A. cajennense ticks infected with R. rickettsii were 1.28% (1/78) and 0.5% (2/400), respectively.Even though Americana is an area endemic for BSF where R. rickettsii is presumably transmitted by A. cajennense (PINTER et al., 2011), we failed to encounter any R. rickettsii-infected ticks.However, this finding is not totally unexpected, since the prevalence of this pathogen among A. cajennense ticks can be very low due to the low efficiency of transovarial and transstadial transmission of R. rickettsii in A. cajennense ticks (SOARES et al., 2012).Similarly to the present study, Sangioni et al. (2005) was unsuccessful in finding any infected ticks among 810 A. cajennense adult ticks collected from areas endemic for BSF in the state of São Paulo.
In conclusion, the present study showed that the areas at risk of BSF in Americana are characterized by high environmental burdens of A. cajennense and A. dubitatum, which are primarily sustained by capybaras.It was shown for the first time that larvae and nymphs of A. dubitatum are active during the same periods as the corresponding stage of A. cajennense.While R. bellii was found infecting A. dubitatum ticks, infection by R. rickettsii among A. cajennense ticks was not found, thus indicating that it probably has a very low infection rate, as also seen in other areas endemic for BSF in the interior of the state of São Paulo.

Figure 2 .
Figure 2. Numbers of Amblyomma cajennense and Amblyomma dubitatum free-living ticks collected monthly by means of dry ice traps in the municipality of Americana, state of São Paulo, from July 2009 to June 2010.

Table 1 .
Numbers of free-living ticks collected by means of dry ice traps in six areas of the municipality of Americana from July 2009 to June 2010.
*different superscript italic letters in the same line mean significantly different proportions of A. cajennense and A. dubitatum adult ticks in the area.+ refers to the proportion (%) of each tick species according to the total number of larvae, nymphs or adults.