Synanthropic triatomines as potential vectors of Trypanosoma cruzi in Central Brazil

Introduction: Chagas disease surveillance requires current knowledge on synanthropic triatomines. We analyzed the occurrence and Trypanosoma cruzi infection rates of triatomine bugs in central Brazil, during 2012-2014. Methods: Triatomines were collected inside or around houses, and T. cruzi infection was determined by optical microscopy and conventional/quantitative polymerase chain reaction. Results: Of the 2706 triatomines collected, Triatoma sordida was the most frequent species in Goiás State, whereas Panstrongylus megistus predominated in the Federal District. Parasites identifi ed were T. cruzi, T. rangeli, and Blastocrithidia sp. Conclusions: P. megistus and T. sordida sustained the risk of T. cruzi transmission to humans in central Brazil.

More than 150 triatomine species have so far been described; 68 of them occur in Brazil 1 , and at least 27 triatomine species have been recorded in the Central-West region of Brazil 2 .The Brazilian Ministry of Health reported about 3,500 deaths caused by Chagas disease in Goiás between 2007 and 2011; most involved adults over 39 years (chronic cases).This Chagas disease-specifi c mortality rate (~11 deaths per 100,000 inhabitants per year) is much higher than the Brazilian average (~3.4) 3 .Furthermore, nine confi rmed cases of acute Chagas disease were reported from 2006-2012, showing that transmission persists in Goiás 4 .Fourteen triatomine species occur in the state, among which Triatoma sordida is particularly common 2 .
Vector-borne Trypanosoma cruzi transmission to humans has never been documented in the Federal District of Brazil, although infected triatomines invade, and at times colonize, houses and peridomestic structures.Panstrongylus megistus, Triatoma pseudomaculata, Triatoma sordida, Rhodnius neglectus, Psammolestes tertius, Panstrongylus geniculatus, and Panstrongylus diasi have all been recorded in the Federal District 5,6 .
Chagas disease surveillance requires current knowledge on the distribution and infection of synanthropic triatomines.These data are also crucial for the planning and evaluation of vector control interventions.Our aims were I) to analyze the occurrence of triatomine bugs in and around houses in Goiás State and the Federal District, Brazil, from 2012-2014, and II) to investigate T. cruzi infection rates in those bugs.
Triatomines were collected inside houses or in peridomestic structures during routine entomological surveillance between August 2012 and December 2014.They were sent to the University of Brasília [Universidade de Brasília (UnB)] from the Central Laboratory (LACEN) of Goiás State Health Department or from the Environmental Surveillance Agency (DIVAL) of the Federal District.
In Goiás, municipality technicians identifi ed the bugs and checked them for T. cruzi infection using optical microscopy.All bugs and their Giemsa-stained microscope slides were submitted to local reference laboratories and state-level LACEN for a fi rst and second quality checks, respectively.All bugs and slides received at the Goiás LACEN between 2013 and 2014 were sent to the Medical Parasitology and Vector Biology Laboratory (MPVBL) at the UnB.
In the Federal District, triatomines were collected through either community-based surveillance, whereby residents can notify about the presence of suspected insects in their homes, or through active searches by DIVAL staff in dwellings where householders had found triatomines.Between 2012 and 2014, all the bugs arriving at the DIVAL were submitted directly to the MPVBL at UnB. Triatomine species were identifi ed at the MPBVL using dichotomic keys 7 and categorized according to developmental stages (fi rst to fi fth instar nymphs and adults) and sex.The date, locality, and site of capture (indoors or in peridomestic structures) were also recorded.
Live triatomines submitted by the DIVAL were examined for infection at the MPVBL.Trypanosoma cruzi infection was investigated by optical microscopy in live bugs at the laboratory as described by Minuzzi-Souza et al. 8 .Bugs from LACEN in Goiás were brought dead to MPVBL, precluding fresh-slide examination.We re-examined (at 1,000× magnifi cation) all the Giemsa-stained slides of the bugs.A specifi c parasitological diagnosis based on morphological criteria to differentiate T. cruzi, Trypanosoma rangeli, and Blastocrithidia sp. was performed 9 .
We used quantitative real time polymerase chain reaction (qPCR) to detect T. cruzi deoxyribonucleic acid (DNA) in bugs for which no Giemsa-stained slides were available and to confi rm parasitological diagnosis based on morphological criteria.DNA was extracted using Illustra Tissue and Cells Genomic (GE), the QIAamp DNA Mini Kit (Qiagen), or the Biopur Kit Extraction Mini Spin Plus following the manufacturers' instructions.To confi rm the integrity of DNA in each sample, we PCR-amplifi ed a 414-bp fragment of the mitochondrial cytochrome b gene 10 .The qPCR was performed with primers TCZ3 and TCZ4, which amplifi ed a 168-bp region of the T. cruzi nuclear microsatellite repetitive region 11 .Reactions were run in a fi nal volume of 20µl, with 1X SYBR® Green PCR Master Mix (Applied Biosystems), 0.2µM of each primer, and 2µl of DNA (10ng).All qPCRs were run in duplicate in a StepOnePlus Real Time PCR System (Applied Biosystems) as follows: 50°C for 2 min, 95°C for 10 min, plus 40 cycles of 95°C for 15s, 60°C for 45s, and 72°C for 10s.The results were analyzed with the StepOne 2.3 software (Applied Biosystems).Each run included positive and negative controls, as well as blank controls with no template DNA.
To identify the DNA of T. rangeli, we performed a qPCR with primers S36 and S67 antisense, modifi ed after Sturm et al. 12 which amplify approximately 330-pb region of T. rangeli kinetoplast DNA minicircle (kDNA).Reactions were run in a fi nal volume of 20µl, using SYBR® Green PCR Master Mix (Applied Biosystems, CA, USA) under the same described conditions of qPCR TCZ.The following amplifi cation conditions were used 50°C for 2 min, 95°C for 10 min and 40 cycles at 95ºC for 15s, 57°C for 1 min, and 72°C for 10s.
A conventional PCR was used to identify Blastocrithidia by a touch-down PCR nuclear ribosomal deoxyribonucleic acid (rDNA-24sα) with primers D75 and D76 which amplifi ed a 250 region of Blastocrithidia nuclear 24sα ribosomal DNA as described by Schijman et al. 13 .
In total, 1,812 triatomines of eight species were collected from 95 of the 246 municipalities, in Goiás State.Most of these vectors (89.9%) were found around houses (Table 1).Triatoma sordida was the most frequently collected species, with records from 63 municipalities.
Of the 1,602 T. sordida collected, 74 (4.6%) were recorded as T. cruzi infected by Goiás State surveillance staff.However, re-examination of stained slides at the UnB and qPCR confi rmed T. cruzi infection in only 48 T. sordida (3%) (Table 1).Trypanosomatids other than T. cruzi were detected in two of the slides; one was positive for Blastocrithidia sp. and another for T. rangeli -also confi rmed by conventional PCR and qPCR respectively (Table 1 and Figure 1).Eight T. sordida nymphs infected with T. cruzi, as confi rmed by microscopy and qPCR, were caught indoors.
The second most frequently collected species in Goiás was R. neglectus (7% of all bugs), which was recorded in 36 municipalities.Twenty-nine (24.2%) of the 120 R. neglectus caught inside or around houses in Goiás were scored as infected with T. cruzi by surveillance staff.Most of the infected bugs were found indoors and 18 (15%) were confi rmed as infected by microscopy and qPCR (Table 1).Some positive slides by surveillance staff were found to be negative at re-examination by UnB researchers and by qPCR.Panstrongylus geniculatus, Panstrongylus diasi, Triatoma pseudomaculata, Triatoma williami, and Triatoma costalimai were rarely recorded in Goiás (Table 1).
Five triatomine species (894 bugs in total) were recorded in 12 of the 30 administrative regions in the Federal District, most frequently in Vicente Pires and Park Way.The majority of these records corresponded to P. megistus, with 828 specimens overall, of those 53 were caught indoors (Table 2).Thirteen P. megistus (1.6%) were confi rmed as infected with T. cruzi in stained slides and qPCR; four infected P. megistus were found inside houses.Moreover, 41 peridomiciliary specimens were infected by Blastocrithidia sp.(Table 2) in the same locality.The second most common species was T. pseudomaculata, but none of the 51 specimens collected (one found indoors) were infected.One P. geniculatus caught inside a house was T. cruzi-positive by both microscopy and qPCR.P. diasi and R. neglectus were also recorded in the Federal District (Table 2).
The results indicate that T. sordida, P. megistus, and R. neglectus play key roles in sustaining the risk of T. cruzi transmission to humans in Goiás State and the Federal District of Brazil.Over 1,800 triatomines were caught inside or around houses in 95 municipalities of Goiás State during the two years of surveillance.About ten years earlier, between 2000 and 2003, intensive entomological surveys recorded 51,570 bugs in Goiás 14 .Although these data might be suggestive of a reduction in triatomine bug presence, the two studies differed in the methods used to ascertain infestation and in dwelling coverage.Oliveira and Silva 14 used data from active bug searches in nearly 250,000 dwellings of 201 municipalities.Our data, in contrast, were generated by a 'passive' surveillance system based on the notifi cation of suspected insects by residents, so that municipality health agents only searched for bugs in houses whose owners report triatomines.These data suggest that synanthropic triatomines are probably much more common in Goiás than the results we report here indicate.Triatoma sordida was the species most frequently caught in Goiás, with most records corresponding to peridomestic environments as   Giemsa-stained slides examined by UnB researchers; some positives were confi rmed by qPCR.c Two slides were scored as negative when re-examined at the UnB, and two slides were found positive for, respectively, Blastocrithidia sp. and T. rangeli (qPCR-confi rmed).d Twenty bugs were positive by qPCR, and in 17, re-examinations of slides at the UnB revealed infection.e The slides were scored as negative when re-examined at the UnB; qPCR was also negative.f Two slides were scored as negative when re-examined at the UnB; both qPCRs were also negative.g Six slides were scored as negative when re-examined at the UnB; all six qPCRs were also negative.h Two bugs were found positive by qPCR; six slides re-examined at UnB were negative, and the corresponding qPCRs were also negative.i Three bugs were found positive by qPCR; one slide re-examined at UnB was negative, and the corresponding qPCR was also negative.
previously described 15 .Our data show that 6.4% of T. sordida caught indoors were infected -about 10 times larger than the infection index reported by Oliveira and Silva 14 or Rossi et al. 15 .
Over two years, 894 triatomines, 14 of which were infected with T. cruzi, were captured in or around houses of 12 administrative regions of the Federal District.In 2002-2010, and using methods comparable to ours, 754 bugs were caught in 20 administrative regions 6 .This suggests that sylvatic foci of synanthropic triatomines persisted in the Federal District, with the bugs (especially P. megistus) regularly invading and at times colonizing human habitats.Continuous entomological surveillance should therefore be maintained to detect and quickly eliminate domestic and peridomestic infestation foci.
In our dataset, the two administrative regions with the most reported triatomines in the Federal District were Vicente Pires and Park Way, whereas Maeda et al. 6 reported most bugs from the more rural Planaltina and Sobradinho.This suggests that the number of triatomines, particularly P. megistus, found in urbanized environments may have increased over the last few years.In Vicente Pires and Park Way, middle-upper class condominiums are often built along gallery forests likely inhabited by wild P. megistus.House invasion and colonization by P. megistus must hence be closely monitored in areas with gallery forest remnants -and particularly over the early rainy season, when the records of the species in human environments peak 6 .Autochthonous T. cruzi transmission mediated by P. megistus was recently reported in primate units at the Brasília zoo 8 .In spite of the presence of T. cruzi-infected triatomines inside houses, no autochthonous cases of human Chagas disease have so far been reported from the Federal District.
The detection of T. cruzi infection in triatomine bugs has traditionally relied on the examination of vector feces through optical microscopy 9 .This inexpensive technique has, however, several important drawbacks.For example, detection may depend on whether the bug is alive or dead, technician microscopy skills and, especially, the quality of slide staining may affect the specifi city of the test -possibly leading to confusion between T. cruzi and the morphologically similar T. rangeli or Blastocrithidia triatomae 4   slides scored as positive by surveillance staff were specifi cally reported as 'T.cruzi-positive'.When re-examined at the UnB a few of those slides were scored as positive for T. rangeli or Blastocrithidia sp., but negative for T. cruzi -or simply as negative for any trypanosomatid.This was further confi rmed by T. cruzi-negative qPCR assays.
Our results show that synanthropic triatomines maintain the risk of T. cruzi transmission to humans in Goiás and the Federal District of Brazil.Furthermore, our parasitological examination data highlight the need for continuous training of surveillance staff in charge of T. cruzi detection in vectors.We therefore recommend that Chagas disease vector surveillance be strengthened across the Central-West region of Brazil.
The development of health education strategies for Chagas disease prevention should emphasize communitybased surveillance to enhance the system's capacity to detect infestation foci.Moreover, triatomine bug surveillance should become an integral part of epidemiological surveillance.Thus, for example, disease surveillance staff could visit dwellings where entomological surveillance detected T. cruzi-positive vectors to investigate infection in dwellers and possibly their domestic mammals.Such an integrated strategy would provide public health offi cials and researchers with a more accurate understanding of how T. cruzi circulates in human environments.

TABLE 1 :
Triatomine bugs collected by routine surveillance in Goiás State, Brazil, 2013-2014: bug characteristics, collection sites, and natural infection with Trypanosoma cruzi.
ND: sex not determined; II: infection index (number and % of bugs scored as positive for T. cruzi); UnB: Universidade de Brasília; qPCR: quantitative polymerase chain reaction; T. cruzi: Trypanosoma cruzi; T. rangeli: Trypanosoma rangeli.aGiemsa-stained slides examined by Goiás State surveillance staff.b . In this study, all stained