Mitochondrial PCR-RFLP Assay to Distinguish Triatoma brasiliensis macromelasoma from Triatoma brasiliensis brasiliensis Subspecies (Hemiptera: Reduviidae)

Triatoma brasiliensis sensu lato (s.l.), the main vector of Chagas disease in northeastern Brazil, is a species complex comprising four species, one with two subspecies (T. brasiliensis brasiliensis, T. brasiliensis macromelasoma, T. juazeirensis, T. sherlocki, and T. melanica), and each taxon displaying distinct ecological requirements. In order to evaluate the genetic relationships among nine T. brasiliensis s.l. populations from northeastern Brazil, we analyzed their mitochondrial cytochrome c oxidase subunit 1 sequences and suggested a PCR-RFLP assay to distinguish between T. b. macromelasoma and T. b. brasiliensis subspecies. All the specimens were morphologically identified as T. b. brasiliensis. The resulting phylogenies identified two major clades that are congruent with the geographical populations studied. Based on collection sites and in accordance with type-location, one clade was identified as the subspecies T. b. macromelasoma. The second clade grouped T. b. brasiliensis populations. Restriction endonuclease sites were observed in the sequences and used in PCR-RFLP assays, producing distinct fingerprints for T. b. macromelasoma and T. b. brasiliensis populations. The results suggest that these are different species and that gene flow occurs only among T. b. brasiliensis populations, possibly associated with human activity in the area.


Introduction
About 28 million people live in areas at risk of Chagas disease, 11-14.5 million of whom are affected worldwide. Trypanosoma cruzi, the pathogen that causes Chagas disease, is found in most South American countries, representing an important cause of heart damage among the economically active population [1]. After a successful chemical control of Triatoma infestans (Klugi, 1834), the other main vectors of Chagas causing agent, Panstrongylus megistus Burmeister, 1835, Rhodnius prolixus Stal, 1859, and Triatoma brasiliensis sensu lato Neiva 1911. T. brasiliensis kept attracting considerable attention from local entomological surveillance.
The systematic of Triatominae species is based on morphological characters of the adult exoskeleton and male phallic structures [5]. However, insects captured during vector monitoring and control or received for identification and notification are often immature. Although their characteristics are similar to those of adult individuals, they are difficult to distinguish. The morphology of Triatominae species is not well described; with studies on the immature forms performed for only 40 species, eggs and nymphs described for only 20 species, a key to identify nymphs to the species level has yet to be developed. Available keys are useful and partially applicable to other stages, but specific identification of all live forms remains unresolved [6].
Members of the T. brasiliensis complex have been distinguished by analyzing isoenzymes [7], mitochondrial DNA sequences [8], and random amplification of polymorphic DNA-RAPD [9]. In the present study we analyzed the barcoding CO1 sequences of nine T. brasiliensis s.l. populations from northeastern Brazil (States of Pernambuco, Paraíba, and Rio Grande do Norte) in order to identify their genetic relationships. We also conducted a PCR-RFLP assay to distinguish between T. b. macromelasoma and T. b. brasiliensis subspecies.

Sample Collection.
Triatominae were collected by surveillance technicians during active inspections in anthropogenic environments (domestic and peridomestic habitats). Live specimens were collected in nine localities in Northeast Brazil (Table 1, Figure 1) using tweezers, flashlights, and PIRISA. Housed in plastic boxes (7 cm diameter × 8 cm high) lined with folded filter paper, the bugs were transported in coolers to the Culicid and Triatomine Laboratory of the Department of Epidemiology, Faculty of Public Health/USP. Adults were identified as T. b. brasiliensis according to the key by Lent and Wygodzinsky [5]. Nymphs were assumed to be T. brasiliensis s.l. because, in addition to the difficulty in distinguishing immature Triatominae based on chromatic and morphological characters, comparative material or a key for nymph identification to species level are not available. After morphological identification, DNA was extracted from individuals from the nine localities and sequenced the 520 bp barcode portion of the CO1 gene (Table 1). Specimens collected in Pernambuco State (even nymphs) were assumed to be T. b. macromelasoma, according to Costa et al. [10] and Costa et al. [11].

DNA Extraction and CO1
Amplification. Genomic DNA was extracted from the legs of 10 individual samples of each population using the Qiagen DNeasy Blood and Tissue Kit (Qiagen, Crawley, United Kingdom) following the manufacturer's protocol. The CO1 barcode region was amplified from whole genomic DNA using primers LCOI 1490 (5 -GGT CAA CAA ATC ATA AAG ATA TTG G-3 ) and HCOI 2198 (5 -TAA ACT TCA GGG TGA CCA AAA AAT-3 ) [12].
PCR amplification was carried out in a final volume of 50 L containing PCR buffer, 0.2 mM of each dNTP, 2.5 mM MgCl 2 , and 1.25 units of Taq polymerase. Initial PCR denaturation was at 94 ∘ C for 5 min, followed by 40 cycles of denaturation (1 min) at 94 ∘ C, annealing (2 min) at 50 ∘ C, and extension (2 min) at 72 ∘ C. A final extension step at 72 ∘ C was performed for 10 min. The amplified DNA was loaded onto 1.5% agarose gel and stained with ethidium bromide. Amplicons were sequenced in both forward and reverse directions using ABI PRISM BigDye Terminator Cycle Sequencing Ready Reaction Kits (Perkin Elmer, Foster City, CA) on an ABI PRISM 3100 Genetic Analyzer/HITACHI.

Analysis of DNA Sequence.
Sequences were aligned using Clustal W [13]. One or two representative haplotypes for each population was chosen because of the low variation within populations. Phylogenetic reconstructions were performed by Neighbor Joining and Maximum Likelihood methods (both using the Kimura-3-parameter distance model K81) in MEGA 5.0 [14], and a divergence matrix was constructed under Kimura two parameters (K2P) ( Table 2). Maximum Parsimony was carried out using the branch-and-bound search option. Phylogenetic analyses included 1000 bootstrap replicates and a Triatoma sordida CO1 sequence (Genbank acc. no. AF021213) as outgroup. Triatoma brasiliensis s.l. CO1 sequences were also analyzed using NEBcutter version 2.0 [15] to select appropriate endonuclease enzymes.

Enzyme Restriction Analysis.
Individual CO1 sequences were PCR-amplified using the above parameters and digested in a 10 L reaction with StyI (Promega) and HincII (New England Biolabs, Ipswich, MA) enzymes. The reaction contained 1 L of 10x buffer, 4 L deionized water, 4 L of amplification product, and 1 unit of restriction enzyme. The digestion mixture was incubated at 37 ∘ C for 2 h and then resolved on 2.0% agarose gel.

Results
Phylogenetic trees derived from the Neighbor Joining, Maximum Likelihood, and Parsimony methods showed similar topologies (data shown for ML tree in Figure 2). Since most of CO1 sequences were identical into and among the populations, the ML tree was constructed using only two samples of each one. This ML tree and the Nucleotide Distance Matrix ( Conversely, the HincII enzyme yielded two fragments (297 bp and 240 bp) in all population samples ( = 10 samples of each population) except that from Salgueiro (Figure 3). StyI and HincII enzymes therefore produced distinct fingerprints for T. b. macromelasoma and T. b. brasiliensis, suggesting that they are different subspecies. The molecular protocols described above are a useful tool in the study of populations and cryptic species, contributing to the identification of insect vectors.

Discussion
The identification of adult Triatominae based on morphological and chromatic pattern is considered relatively easy for most species; however, this is commonly misguided owing to the wide phenotypic variability within this subfamily. For instance, Triatoma maculata Erichson, 1848 and Triatoma pseudomaculata Corrêa & Espínola, 1964 which were first treated as members of a same species complex due to morphological similarities [5] thereafter proved to be genetically distant [17][18][19]. Panstrongylus herreri Wygodzinsky, 1948 and Panstrongylus lignarius Walker, 1873 in turn, were considered to be separate species until Marcilla et al. [20] and Crossa et al. [21] demonstrated that they are the same species, cytogenetically identical with regard to the second internal transcribed spacer. Another difficulty in identifying genera and species of Triatominae is their extensive chromatic variability. The color of some species such as Rhodnius sp. (light brown tones) and Rhodnius nasutus Stål, 1859 (pinkish tones) seems to be associated with the color of the palm trees they colonize [22], but others such as Triatoma rubrovaria Blanchard, 1843, exhibit well-known 4 chromatic morphotypes [23]. Other studies report the occurrence of natural homoploid hybrids between T. infestans and Triatoma platensis Neiva, 1913, T. infestans and Triatoma rubrovaria, and sympatric species of Phyllosoma complex and species of the T. brasiliensis  complex. In addition, several hybrid species have been obtained experimentally [24]. This interspecific crossing can be decisive in originating and diversifying wild species, resulting in important epidemiological consequences due to differential competence and the capacity of hybrid vectors [3,10,11]. Therefore, the characterization (or identification) of Triatominae specimens based only on morphological and chromatic patterns, the most common identification method, is more complex than previously believed. Studies on immature stages are crucial for group systematics. However, literature reports on immature forms of certain groups are scarce, difficult to use, or nonexistent. Many species undergo changes in color, structure, and morphology during their development, hindering their identification [25,26]. Triatominae nymphs at this development stage are difficult to identify. To that end, molecular analyses are successfully used to characterize morphotypes of species complexes such as T. brasiliensis sp., which exhibits wide chromatic and morphological variation [7,8].
Marked differences in color pattern and ecological features among species from the Triatoma brasiliensis complex were detected by microsatellites, mitochondrial 12S, and cytochrome b genes, reinforcing species diagnosis [7,16]. However, individuals from subspecies T. b. brasiliensis and T. b. macromelasoma might be clustered within the same CO1 clade, since earlier studies have shown that some cytochrome b haplotypes of T. b. macromelasoma are similar to those of T. b. brasiliensis [16]. Moreover, these subspecies produce fertile hybrids when crossed in laboratory [3].
The basal clade of the Maximum Likelihood tree (Figure 1) was identified as subspecies T. b. macromelasoma because it is highly divergent (6%, as shown in the Nucleotide Distance Matrix, Table 2) from the other populations and was collected in its type locality [16]. On the other hand, all populations in the second clade were identified as T. b. brasiliensis. Their genetic similarity may be related to geographic proximity and similar habitat conditions. However, interpopulation divergence values (<1%) suggest that T. b. brasiliensis is still diversifying and/or exhibiting ongoing gene flow, probably due to human-assisted dispersal. Based on wing morphometry, Costa et al. [11] recently formulated a hypothesis that T. b. macromelasoma is the result of homoploidal hybridization between T. b. brasiliensis and T. juazeirensis in the state of Pernambuco, and that this is a form of speciation in sympatric populations.  In northeastern Brazil, the epidemiological importance of Triatominae bugs is mainly defined by their high rate of natural T. cruzi infection and ability to adapt to multiple ecotopes. Control measures therefore require a precise identification of which species of the T. brasiliensis complex is being targeted. Moreover, it is important to understand the ecoepidemiology of Triatominae since these vectors are found in large numbers in their natural habitat [27]. In this respect, the PCR-RFLP protocol described here is suggested as rapid, relatively simple, and economical assay to distinguish Triatoma b. macromelasoma from Triatoma b. brasiliensis subspecies. Even at small geographic scales, domestic populations are genetically structured by ecological parameters, thereby exhibiting small differences from the wild counterparts from which they are derived [8]. The present study highlights the effectiveness of the CO1 gene in identifying subspecies of the T. brasiliensis complex and its contribution to classic taxonomy.

Disclosure
No competing financial interests exist in the paper.