National and International Reference Laboratory for Triatomine Taxonomy, Oswaldo Cruz Institute/Fiocruz
Members of the Triatominae subfamily (Hemiptera, Reduviidae) (Figure 1) are very relevant because they are hematophagous and transmitters of theTrypanosoma cruzi (Chagas 1909), the etiologic agent of Chagas disease. Only a few other members of the order Hemiptera, belonging to the families Cimicidae (bedbugs), Polyctenidae (bat ectoparasites), and some members of the Cleradini tribe also feed on blood. As important vectors, triatomine bugs have always been worthy of permanent attention. As a result, various aspects of their biology, ecology, biogeography and evolution have been studied for decades. From their first description in the 18th century to this day, the classification of these insects has been done predominantly through the traditional morphological concept, and the subfamily was considered a monophyletic group. In the 80’s, a new proposal was put forth, based on ecological and biogeographical aspects that suggested the polyphyletic origin of the subfamily. Recent cladistic analysis using DNA have brought back the hypothesis of the monophyletic origin, breaking the paradigm of the polyphyletic origin. In this text we summarize the current knowledge of the systematics of these vectors, including a brief history of the most relevant studies of subfamily systematics.
The first great subdivision of animals is the phylum, which is subdivided in classes. Classes are divided in orders, which are divided in families; families are divided into genera, and these are divided in species. There are also many intermediary divisions, such as subfamilies and tribes. Phylum Arthropoda includes insects, among other animals. The main characteristics of arthropods are articulated legs and the presence of an exoskeleton (a hard and resistant external skeleton, forming a sort of shell that protects and supports the internal organs).
Within the Phylum Arthropoda, it is in Class Insecta that all insects are included, and they are classified in different Orders, among which is Hemiptera. This order includes insects more commonly known as true bugs. One of the main characteristics of insects of this order is the structure of their forewings: in most hemiptera, the basal part of the wing is thickened and leathery, while the apical part is membranous. This type of wing is called hemielytron, hence the name of the order (hemi = half; ptera = wing). The mouthpieces of hemiptera insects are of the piercing and sucking type. The mouthpiece is called a labium and originates in the front part of the head. While most species are terrestrian, some are aquatic. Many suck from plants, while others predate other insects (and are therefore called entomophagous), while others still are hematophagous.
Of the more than 40 hemiptera families, only two are relevant for human medicine, due to the hematophagic habit of their species: Cimicidae (bedbugs) and Reduviidae. The number of subfamilies accepted within Reduviidae varies from 21 to 29, according to different specialists. Only one family (Triatominae) feeds on blood; the other subfamilies consist of predator insects.The first report on the physical appearance and the habits of a triatomine dates back to 1590 and was made by Reginaldo de Lizárraga, who, during inspection trips to convents in Peru and Chile, noticed the presence of large hematophagous insects that attacked at night. Many other travellers and naturalists also mentioned the presence of these insects in South America. One of the most famous of these reports was provided by Charles Darwin, who, during his journey to South America on the H.M.S, Beagle, in 1835, recorded the following in his journal:
“When placed on a table, and though surrounded by people, if a finger was presented, the bold insect would immediately protrude its sucker, make a charge, and if allowed, draw blood, No pain was caused by the wound, It was curious to watch its body during the act of sucking, as in less than ten minutes it changed from being as flat as a wafer to a globular form…”
The publication of Systema Naturae by Carolus Linnaeus in 1758 already included the creation of order Hemiptera. However, at the time the grouping of these insects into a subfamily had not yet been proposed. Taxonomic studies of triatomines began with these publications: Mémoire pour servir à l´histoire des insectes por (De Geer, 1773), in which the first species was described as Cimex rubrofasciatus (Figure 2). Sixty years later, Laporte (1833) designated it as a type species of genusTriatoma and it then changed name to Triatoma rubrofasciata. In 1811, Latreille published Insectes de l´Amérique équinoxiale describing Reduviusdimidiatus and Reduviusgeniculatus, and Stål published, in 1859, his Monographie der Gattung Conorhinus und Verwandten, in Berliner Entomologisches Zeitschrift. In 1873, Walker published Catalogue of the Species of Hemiptera Heteroptera in the Collection of the British Museum. Later on, Berg, Uhler, Champion, Breddin and Distant described new species, considerably expanding their number. In 1907 there were 57 species, of which 32 are considered valid to this day.
For over a century, since the first description by De Geer, triatomines were studied only from the descriptive standpoint. From 1909, withCarlos Chagas‘s discovery, and due to the obvious medical-sanitary relevance acquired, studies began on the clinical form of the disease, on the protozoa and its vertebrate hosts, and on the biology and mechanism of transmission of vectors (Figure 3). As for the taxonomy of vectors, one of the main people responsible for breakthroughs in this new phase wasArthur Neiva, who, beginning in 1911, described several species, peaking with the publication of his thesis: Revision of genus Triatoma Lap. Important monographs were later published by Pinto (1925) and Del Ponte (1930), in addition to other papers published by Neiva & Lent (1936, 1941), by Usinger (1944), by Abalos & Wygodzinsky (1951) and by Ryckman (1962), culminating with the review by Lent e Wygodzinsky (1979).
The use of morphometry on Triatominae began in the 1990’s, as an attempt to refine conventional taxonomic study. Later, a new branch of this method, geometric morphometry, also began to be used in triatomine studies, advancing beyond the reach of traditional morphometry. The first attempt to use non-morphological characteristics to solve taxonomic issues was attributed to Actis and collaborators (1964, 1965), who used electrophoresis of hemolymph proteins to compare species of the T. sordidacomplex. Similar but more encompassing studies were published three years later by Brodie and Ryckman. Since then, many studies using genetic analysis for specific characterization or as taxonomic tool have been published. More recently, Carcavallo et al. (2000) and Galvão et al (2003) summarized the taxonomic papers subsequent to the revision by Lent & Wygodzinsky (1979), while Schofield & Galvão (2009) reorganized genus Triatoma into three groups and eight complexes. The subfamily is currently divided in six tribes, 18 genuses and 152 species.
The hypothesis that triatomines have evolved from predator ancestors (non-triatomines) is currently accepted. However, it is unknown which is/are the predator group(s) that gave origin to triatomines. This issue is made even more difficult to solve due to the scarcity of fossil records, restricted to only two species described by Ponair (2005, 2013). Getting to know the evolutionary pathways followed during the process of adaptation to blood sucking is no easy task: the different proposals that exist at the moment are the reason for much polemics and debates. Consensus on the origin of triatomines is far from being reached. New cladistic analysis with more significant samples and combining morphological and molecular features are necessary to reach a firm conclusion on the issue.The taxonomic review of Lent and Wygodzinsky (1979) was a game changer as it made the work of taxonomists easier. The current classification of the subfamily is still based on this publication. Although the theory proposed by Willi Hennig was not formally applied, it was applied to the first postulates on the phylogenetic relation of the subfamily and on its relation with possible “sister” groups, clearly defending the monophyletic origin initially proposed by Usinger (1944). According to Lent and Wygodzinsky (1979), the monophyletic origin of triatomines could be supported by three possible autapomorphies: a) the hematophagic habit, b) the presence of a flexible membrane connecting the second and third rostral segments, facilitation suction, and c) loss of abdominal scent glands in nymphs. Later on, the monophyletic hypothesis was supported by studies by Clayton (1990), Schuh & Slater (1995), Guant & Miles (2000) and Hypsa et al. (2002). Although it is acknowledged by those authors as a monophyletic subfamily of Reduviidae adapted to hematophagy, several questions on the origin and evolution of these vectors remain unknown. In 1988, Schofield proposed a concept of polyphyletic origin, in which current Asian fauna would be derived from at least two independent lineages of ancestral groups of Reduviidae. The first lineage would consist of species ofTriatoma supposedly descending of T. rubrofasciata after its introduction in the Old World. The other lineage, represented by genus Linshcosteus, would be supposedly autochtonous in India. This hypothesis was supported by Gorla and collaborators (1997) by means of morphometric analysis, in which Linshcosteus species would be dissociated from Triatomaspecies of the Old World. On the other hand, molecular cladistic analysis of 57 species by Hypsa and collaborators (2002) showed genus Linshcosteus and species Triatoma rubrofasciata as sister groups, with clade Linshcosteus-T. rubrofasciatarooted into Triatomini, supporting the subfamily’s monophyletic origin. According to Schaefer, the most relevant issue on the phylogenetic relation of triatomines can be divided in two parts: 1) Is Triatominae truly an independent subfamily? 2) If Triatominae is independent, which other Reduviidae subfamilies is it related to? In attempt to solve this issue, several scientists have been using molecular data to infer phylogenetic relations. Sadly, most of these papers were based on a limited number of taxons, making it impossible to answer these questions with conviction. Just like we should not use a single morphological characteristic to define or identify a species, it would be prudent to not do the same with a single DNA sequence. It is also important to highlight that cladistic analysis always need a hypothesis a priori (to be tested), and this hypothesis is usually provided by conventional taxonomy. Another interesting hypothesis of conventional taxonomy is its utilitarian aspect, which makes its results more easily comprehensible to a non-academic audience (such as vector control agents, for instance). Conventional taxonomy is also “protected” by ICZN, but Phylogenetic Systematics already has its own nomenclature code, PhyloCode, an alternative to the nomenclature system based on conventional code categories. Although it is not yet formally in operation, PhyloCode is available on the Internet, with increasing interest among scientists. The progress of Phylogenetic Systematics appears to be irreversible.
Laboratory for Parasitic Diseases, Oswaldo Cruz Institute/Fiocruz
Molecular analysis have been revolutionizing biology. Among their countless applications are the development of molecular markers for the differentiation of biological entities (populations, species etc) and the evaluation of their systematic and evolutionary relationship.
Vector control programs already benefit from the knowledge generated by molecular studies. Some of the most relevant contributions include the identification and characterization of introduced vector populations, such as R. prolixus in Central America or T. infestans in the Southern Cone of South America; the precise definition of targets for control interventions, including R. prolixus as a valid species, separate from R. robustus, and the chromatic forms of T. brasiliensis making up distinct species, as described by Monteiro and collaborators. Anwers to the most crucial questions are frequently in the context of “fine” systematics, such as the kind that deals with intraspecies variation. However, the evaluation of interspecies variability and phylogenetic relations are relevant in many cases, as they allow for wider extrapolations on epidemiologically relevant groups, as revised by Schaefer.We present here a review of the use of molecular markers in triatomines. We also comment on the applicability, advantages and disadvantages of the main molecular techniques (Tables 1 and 2), provide suggestions for future studies, and highlight important issues that should be taken into account during the planning of future projects and their experimental design (Table 3).
It is important to emphasize the idea that the proper identification of insects is a crucial step for the reliability of comprehension of vector biology. Ecological, behavioral, physiological, epidemiological or evolutional investigations, as well as the planning and monitoring of control interventions, are necessarily subject to accurate taxonomic judgments. At the same time, the analysis of molecular data sheds light on more basic questions regarding relations, evolution and adaptative trends of vectors. We discussed the use of different techniques for the analysis of genetic material (cytogenetics, methods based on PCR and genic products (aloenzymes) in the study of populations, species, genuses and tribes of Triatominae. This essay focuses on the methods that have generated more wealth of information and, in particular, on those that have produced key pieces of knowledge for the designing of rational strategies to control Chagas disease.
Different molecular approaches supported by good sampling strategies and adequate choice of markers can be employed in the elucidation of issues regarding biogeography, behavior, taxonomy, evolution, and population structure of triatomines. We present here a general appreciation of the results obtained in these fields so far.We also attempt to stimulate investigation in the area, pointing to new pathways and all the while warning of the requisites and cautionary steps that should be taken to elaborate the projects. With high-quality results, molecular genetic studies will continue to contribute to the elaboration of more effective surveillance and control programs for Chagas disease vectors.
Triatomines have holocentric chromosomes, usually with a diploid complement of 2n=20A+XX/XY (20 autosomes plus XX♀ and XY♂); karyotypes with 18A and 22A are rare. Males of Triatoma species from Central and North America usually present a X1X2Y sexual mechanism, while South American species typically present XY. Panstrongylus and Eratyrus also have X1X2Y. All Rhodniini have a diploid complement of 2n=20A+XX/XY. Special cytogenetic techniques, such as C-banding and detailed analysis of heterochromatic regions, or the study of the meiotic behavior of male chromosomes, have been used to complement the morphological characterization of Triatominae. Applications of these methods range from the identification of similar species and detection of intraspecies variability to the investigation of evolutionary relations in different levels of divergence, as reviewed by Dujardin.
The use of DNA techniques for the study of triatomines is fairly recent and yet has a very dynamic history. Molecular studies have resulted in new perspectives within the evolutionary trends of the subfamily, with projects in levels that range from intertribes to intraspecies. Polymorphisms both in nuclear and mitochondrial genic sequences have been studied. Many results have been used to refine and improve control strategies in Latin America, and they will certainly prove crucial for the study of secondary vectors as well. Many DNA-based techniques have been applied to triatomines. These include random amplification of polymorphic DNA (RAPD), the analysis of single-strand conformation polymorphism (SSCP), sequencing of DNA from fragments of nuclear and mitochondrial genes, and characterization of microsatellite loci.
The amplification of polymorphic DNA with random oligonucleotides (Figure 1) has been used to investigate genetic variability in various insects with medical relevance. In triatomines, diversity among populations of T. infestans and T. sordida was evaluated by Carlier and collaborators. Band patterns clearly separated the species and discriminated wild and domestic species of T. infestans. In another study by Noireau and collaborators regarding Bolivian T. infestans, RAPD profiles separated two main groups (Andes and Chaco). Within each clade, wild and domestic populations, García and collaborators developed RAPD diagnostic profiles for the pairs R. prolixus-robustus and R. neglectus-nasutus. Honduras and Colombia populations of R. prolixus presented patterns of diagnostic RAPD. The reduced variability within these insects in Central America suggests a recent anthropic dispersion of a population that originated in South America, encouraging the thought that its erradication is feasible in the region.
Although the resolution of RAPD analysis is much higher than that of aloenzyme electrophoresis in terms of detecting intraspecies variability, the use of random oligos requires extreme care to minimize contamination and reproductibility problems (Figure 2). Another serious issue with RAPDs is that there is little guarantee of homology between same-sized bands, because oligos aneel randomly. Even the specific conditions employed in the preparation of the samples and of the PCR can affect results, reducing reliability and reproductibility. Finally, as RAPD markers are dominant (Figure 3), it is not possible to determine whether a single band in the gel corresponds to a homozygote or heterozygote individual. This rules out the possibility of using one of these markers in genetic studies of populations, because conformity to the expected Hardy-Weinberg equilibrium cannot be tested.
The sequencing of gene fragments (Figure 4) allows for the direct evaluation of DNA polymorphisms, providing the best type of data for phylogenetic inferences and for the determination of kinship relations between individuals and populations. Different segments of mitochondrial (mt) and nuclear DNA evolve at different rates. Rapid-evolution regions are suitable for the study of closely related organisms, while more preserved regions are more suitable for comparisons between more divergent taxons.
Mitochondria are cytoplasmatic organelles of eukaryotic cells. They are involved in key physiological and pathological processes. In metazoa, each mitochondria possesses a single molecule of circular DNA, 15-20 kb in size, containing genes that code for 2 rRNAs, 22 tRNAs and 13 mRNAs responsible for the synthesis of cell respiration proteins. This same genic composition is found in the only triatomine mitochondrial genome already sequenced (that of T. dimidiata). When compared with that of other insects, the mitochondrial genome of 17,019 pairs of bases of T. dimidata (Figure 5) has a lower composition of adenine + thymine (69.5% A+T), as revised by Dotson and Beard.
The mtDNA molecule, present in multiple copies (102-140) per cell, is inherited from the mother, replicates without recombination, and evolves faster than nuclear genome. High replacement rates are partly due to poor repair mechanisms during replication.
Mitochondrial genes are widely used in populational genetics and phylogenetic studies. Variabilities present in rapidly evolving mitochondrial genes are useful to investigate taxons that have diverged in more recent geological times. In older divergence levels, variable sites are subject to the accumulation of multiple substitutions (especially of transition-type mutations in the third position of codons), therefore leading to saturation. Saturation wipes out the phylogenetic signal and is one of the main causes of homoplasia. Other precautions that should be adopted when interpreting mtDNA results in the study of closely related taxons include the possibility of introgression occurring in hybrid zones (usually detectable with the use of additional data of nuclear loci) and the retention of ancestral polymorphisms. On the other hand, the direct sequencing of mitochondrial genes is easier than that of nuclear genes (which may require cloning), and the alignment of sequences does not pose any problem.
In 1999, Monteiro and collaborators reported the first study of polymorphisms in mtDNA sequences (gene cytochrome b [cyt b]) among populations of the vector triatomine, T. infestans. The results showed that T. melanosoma and the wild melanic forms of T. infestans are simply phenotypical variations of the same species. Small yet consistent differences were found between Bolivian populations and those of Argentina and Brazil. In Bolivia, all Andean insects had the same haplotype, while the melanic insects from the Chaco flatlands region had slightly divergent haplotypes. These melanic forms differ phenotypically and ecologically (they are dark and occupy tree hollows in Chaco) from AndeanT. infestans. Aloenzymes, cytogenetics, inter-cross-breeding, RAPD and mtDNA all corroborate the idea of a single species. A recent study based on mitochondrial gene cyt b which included all chromatic variations of T. brasiliensis (Figure 6) showed that the juazeiro and melanica forms deserve their own specific status. These two chromatic forms were recently elevated to the specific category by Costa and collaborators. T. b. brasiliensis and T. b. macromelasoma (very different from juazeiro and melanica) also seem to represent distinct evolutionary lineages. However, as the distance between them is not as big as that between other populations, a sturdy taxonomic decision should await future analysis.
Analysis of mtDNA sequences revealed that populations of T. dimidiata of the Yucatan peninsula are very divergent (possibly cryptic species) and that the natural dispersion of the species probably followed a north-south route from southern Mexico through Central America and all the way to Colombia. Ecuadorean populations, however, have probably derived recently from populations from Honduras/Guatemala, supporting the hypothesis of an artificial introduction of the species in Ecuador and Peru suggested by the groups of Marcilla and Abad-Franch.Similarly, many populations of R. prolixus (Venezuela, Colombia, Honduras, Guatemala and Brazil) are virtually identical, based on cyt b haplotypes, indicating a recent artificial dispersion of synanthropic forms. This conclusion is also supported by metric analysis and by RAPD. On the other hand, as described by Monteiro and collaborators, high levels of structuration were found between various populations of R. robustus, suggesting that this taxon contains a complex of at least four cryptic species (Figure 7).
In a study of sequences of gene mt cyt b with data from five populations of R. ecuadoriensis (Figure 8) (four from Ecuador and one from Peru), Abad-Franch and collaborators discovered that insects from Peru had a markedly different haplotype, while all sequences from Ecuador (nine haplotypes) were similar. Results suggest that the two clades represent discrete phylogropus (or even incipient species), and indicate that control programs can attack them independently.
Data from mitochondrial sequences have been used to solve taxonomic issues. The specific status of many problematic taxons was confirmed, such as with R. prolixus, robustus, nasutus and neglectus, while acknowledged species have proven to be simply chromatic variants, such as T. infestans and T. melanosoma. In many cases, analysis revealed that a single species actually consisted of more than one cryptic taxon, such as T. dimidiata, T. brasiliensis, R. robustus, or R. ecuadoriensis. In a recent analysis of sequences of mt cyt b, it was shown that R. pallescens is a very variable species, with four moderately divergent haplotypes detected in a small sample. The close species R. colombiensis presented a haplotype that was within the intraspecies variation of pallescens, suggesting either co-specificity (colombiensis possibly representing a breed or subspecies) or a very recent divergence. An interesting case of sequencing of a single specimen (identified as R. robustus and belonging to a population that is frequently collected from palm trees in the Ecuadorean Amazon) revealed a haplotype of cyt bthat is quite singular and had never been found in Rhodnius. Through phylogenetic analysis, Abad-Franch showed that this specimen represents the most basal taxon of the lineage that includes R. prolixus, neglectus, the robustus complex and Psammolestes. If future studies confirm these preliminary results, a new species of triatomine will have been discovered through DNA analysis.
The use of mtDNA in the investigation of systematic and phylogenetic issues in triatomines began in the 1990’s. Stothard and collaborators showed that the three main genuses (Rhodnius, Triatoma and Panstrongylus) are well separated from each other. Two main clades (Rhodniini and Triatomini) were evident, but genetic distances were found to be greater between R. pictipes and R. prolixus than between Panstrongylus and Triatoma.
The phylogeny of 11 species of the infestans complex and four others was explored based on mitochondrial genes 12S, 16S and cytochrome oxidase I. García and collaborators confirmed the evolutionary proximity of T. infestans and T. platensis, and evidence was also found of introgression of mtDNA among them, T. circummaculata and T. rubrovaria remained within theinfestans complex, while P. megistus was grouped with North American species, T. sordida was near T. matogrossensis and distant from T. guasayana.
Mitochondrial genes 16S rRNA and cyt b were used by Lyman and collaborators. The Rhodniini tribe proved to be very different from the Triatomini tribe, with a profound separation between the branches that connect them in the phylogenetic tree. Triatomine species could also be separated into a group from Central America and a group from North America, while another group was for South American species. The positioning of P. megistus and Dipetalogaster maxima was uncertain, but both remained within Triatoma, which is an indication of the paraphyletic nature of this genus.Monteiro and collaborators studied the relation between various Rhodniini species, including the two genuses of this tribe: Rhodnius and Psammolestes. Results confirmed the paraphylia of Rhodnius(Figure 9), with Psammolestes appearing closer to the prolixus clade than to the pictipes clade. Parsimony analysis of 1429bp (including fragments of mitochondrial genes 16S rRNA and cyt b, in addition to variable region D2 of nuclear ribosolam gene 28S) generated two main clades: ((brethesi-pictipes) (ecuadoriensis-pallescens) and (Ps, tertius (neivai (domesticus (nasutus (neglectus-prolixus) (robustus-prolixus).
Hypša and collaborators combined different fragments of DNA (predominantly mitochondrial 16S) to build a phylogeny of 57 species of Triatominae, representing nine genuses and including Linshcosteus from the Old World. The sharp separation between Triatomini and Rhodniini was once again confirmed. Most of the generic subdivisions currently acknowledged did not appear as monophyletic groups. Panstrongylus, Dipetalogaster, Mepraiae Linshcosteus, all appeared as sister groups of different species of “Triatoma”. The paraphylia of Rhodnius regarding Psammolestes was also confirmed. Various taxonomic and systematic rearrangements were proposed, but some significant questions (such as the non-monophylia of Panstrongyluls) were not discussed.
However, the most important finding of this work, although not that well supported, was the indication that the Triatominae subfamily is monophyletic. The observation that the Old World genus Linshcosteus is grouped with Triatomini (with T. rubrofasciata as its sister group) clearly rules out the possibility of a separate and independent origin for Linshcosteus. An elucidative discussion on the issue of a mono- or polyphyletic origin for Triatominae was elaborated by Schaefer.
In a recent work, Gaunt and Miles used sequences of the COI mitochondrial gene and amino acid data to calibrate a mitochondrial molecular clock for various orders of insects, including Hemiptera. Their results suggest that the evolutionary separation between Triatomini and Rhodniini took place 93 million years ago, coinciding with the beginning of the separation of South America from Africa in the Cretaceous. The paraphyletic nature of Triatoma became apparent, as both Eratyrus and Panstrongylus were grouped with representatives of that genus.
Nuclear genes are usually more preserved than mitochondrial ones. They are, therefore, more indicated for the analysis of diversity and relations above species level.
However, ribosomal RNA spacers such as internal transcribed spacers (ITSs) can be informative for populational analysis. These markers have also been used in taxonomic and evolutionary studies of Triatominae. The biggest problems regarding the use of ITSs in systematics are related to the alignment of sequences and to intragenomic variability.
In addition to the combined analysis of mitochondrial fragments with the D2 variable region of 28S rRNA mentioned above, the second internal transcribed spacer (ITS-2) of the nuclear rRNA was recently tested as a molecular marker for populations, species, and phylogenetic relations in Triatominae. These comparisons included predominantly Meso American species of Triatoma belonging to the phyllosoma complex, other Triatoma from South America and various species of Panstrongylus; many populations of T. infestans and some Rhodniini were also analyzed by Marcilla and collaborators.The phylogenetic analysis of sequences of ITS-2 (Figure 10) carried out by Marcilla and collaborators corroborated significant findings based on mitochondrial DNA as they revealed two major clades within Triatomines, a compound of species from Central and North Americas (including T. dimidiata, three species of the phyllosoma complex, T. barberi and D. maxima) and the second containing South American species (T. infestans, T. sordida, T. brasiliensis and P. megistus). They also confirmed the paraphylia of Rhodnius, showing that P. tertius is more closely related to R. prolixus than R. prolixus is to R. stali. This analysis was later expanded and inncluded various species of Panstrongylus. These studies showed that some taxonomic rearrangements may be necessary in the phyllosoma complex and in genus Panstrongylus. For instance, T. picturata and T. longipennis presented identical sequences, and the degree of variability within T. dimidiata was higher than what was found for the remaining species of phyllosoma. In addition, P. lignarius (considered wild) and P. herreri (an important vector in northern Peru) presented identical ITS-2 sequences, even though the geographical sampling was wide. These species were recently synonymized by Galvão and collaborators, a procedure than could also be appropriate for P. chinai and P. howardi, which are probably chromatic variations of the same species. The unexpected position of P. rufotuberculatus as a sister species of the dimidiata-phyllosoma group, with T. barberi – D. maxima occupying the immediate external branch, was interpreted as a strong indication of non-monophylia of Panstrongylus.
The full identity of the sequences of T. dimidiata specimens from Ecuador and Honduras suggests that the first derived recently from a Meso American population, probably introduced into the west of Ecuador in a recent past. Similar conclusions regarding human intervention in the passive transportation of populations of synanthrope insects were drawn from analysis of variation in ITS-2 sequences among geographical populations of T. infestans carried out by Marcilla and collaborators.
The analysis of the 18S subunit of nuclear rDNA carried out by Bargues and collaborators proved this unit to be more preserved than ITS-2 or even mitochondrial genes. In triatomines, the substitution rate can be as high as 1.8% divergence per 100 million years, which is 55 times slower than ITS-2. These substitution rates were used to calculate the divergence time between the various species of triatomines; according to this molecular clock of 18S rDNA, the divergence between Triatomini and Rhodniini would yield an estimate of 48.9-64.4 million years ago (Paleocene-Eocene). Estimates of divergence among Meso- and South American species of Triatomini range from 22.8 to 31.9 million years ago (for 18S) and from 19.5 to 34.1 million years ago (for ITS-2), much before the Isthmus of Panama connected North and South America in the Pliocene (3 million years).
Micro satellites (Figure 11) are series of short repetitive sequences [such as (GT)n or (AT)n within nuclear DNA (Figure 12); they are highly polymorphic, neutral, and exhibit Mendellian inheritance and co-dominance. Micro satellites have become the tool of choice in populations genetics, when other markers do not present a suitable degree of polymorphism. Harry and collaborators studied wild populations of R. pallescens from Attalea palm trees, and found that six out of every 10 micro satellites evaluated were present in frequencies not different from those expected for the H-W balance, suggesting panmixia within the populations of palm trees in the area. Amplicons were also obtained using DNA template of R. ecuadoriensis and R. prolixus. More recently, Anderson and collaborators identified and characterized eight loci of micro satellites of populations of T. dimidiata from Mexico, Guatemala and Honduras, while García and collaborators characterized 10 loci in T. infestans.
The analysis of the electrophoretic properties of alloenzymes (Figure 13) (enzymes with the same function, but with distinct electrophoretic migration patterns, codified by alleles of the same locus (Figure 14) has been extensively used in the study of Triatominae. For decades, this methodology was the first choice to examine taxonomy and the evolutionary relations of countless organisms, including insects. Analysis at an intraspecies level make it possible to (1) evaluate the degree of genetic variability within a certain species or population (given by the expected number of heterozygote individuals in the analyzed loci – He), (2) detect deviations from the intrapopulational Hardy-Weinberg balance (by observing the excess or deficit of heterozigotes – FIS), and (3) estimate levels of genetic structuring (measured as differenciation between populations, based on the interpopulational component of total genetic variation – FST). Finally, (4) levels of gene flow (i.e. effective migration) between populations can be estimated.
Loxdale and Lushai showed that alloenzymes are suitable markers for the study of intraspecific variation in most groups of vector insects. However, triatomines, in particular, tend to present very low levels of alloenzymatic variability (in terms of heterozigotes as well as of polymorphic loci), which causes a certain limitation to the use of this marker for populational studies. Even then, some studies were able to detect sufficient variation to allow for inferences on populational structure to be made.
For instance, populations of T. infestans are panmictic within Andean villages in Bolivia and Peru, but highly structured (and following the model of isolation by distance) between villages. Dujardin’s analysis of allele frequencies made it possible to determine the probable area of origin of T. infestans. Later studies by Breniére indicated that structuring within localtions can also occur, suggesting that the basic populational unit is represented by human dwellings themselves. Using a similar methodology, Noireau and collaborators showed that the panmitic unit of Bolivian populations of T. sordida (alloenzymatic group 1) is bigger than the unit described for T. infestans, with deviations from the H-W balance detected only between populations that are located more than 20 km apart; this suggests a better capacity of dispersion of T. sordida than what was previously supposed.
In addition, the low alloenzymatic diversity in Triatominae has been seen as possible indication of higher vulnerability to chemical control, as it would be less likely that resistance to insecticides would appear in genetically impoverished populations.
In taxonomy (or alpha systematics, i.e. identification and description of species), alloenzyme electrophoresis is used to distinguish cryptic species and to determine the correct status of doubtful populations. Detecting reproductive isolation between two or more cryptic species is particularly easy when their populations are sympatric. This occurs because the premise “if the populations belong to the same gene pool, they must be exchanging genes” can be used as a null hypothesis. So, if loci present different affixed alleles in the two populations (i.e. they are diagnostic), the null hypothesis is rejected in favor of the alternative hypothesis of the existence of more than one taxon occurring in that area, as revised by Thorpe and Solé-Cava.
This reasoning cannot be applied to alopatric populations, among which genetic exchanges may not occur simply due to their geographical separation. These cases require the comparison of values of genetic distances on a par between populations, with those found in the literature. This method also works well: since a large number of individuals and loci are sampled, genetic distances are much bigger between populations of species that are closely related than among co-specific populations.
Cut-off values for genetic distance for the distinction of species have been suggested as D ≈ 0.16. For insects with medical relevance, Noireau and collaborators showed values of D > 0.1 are considered indicative of specific level. In triatomines, Dujardin and collaborators described average values of D = 0.504 ± 0.341 detected in 30 inter-specific comparisons, while co-specific populations (142 comparisons) were separated by average values of D = 0.013 ± 0.009.
In general, when clear differences (diagnostic loci in sypatry or large genetic distances in alopatry) have been detected between populations, the usual conclusion is that more than one taxon was involved, or it was due to identification mistakes, such as R. colombiensis erroneously identified as R. prolixus, or cryptic speciation (such as with T. sordida). The characterization of different chromatic populations of T. brasiliensis seemed to represent an exception when fixed alloenzymatic differences were found between co-specific populations by Costa and collaborators. The alternative view that various different taxons were involved was later supported by the analysis of DNA sequences by Monteiro and collaborators.
Several examples illustrate the use of alloenzymes in the evaluaiton of triatomine taxonomy. The absence of fixed differences between R. prolixus (a very relevant vector) and R. robustus (wild and of secondary medical relevance) from Venezuela led to the suggestion that these were the same taxon. A comparison of alloenzymes of prolixus and robustus individuals, whose identity had been previously confirmed through mitochondrial DNA, still failed to detect differences. This finding was interpreted by Monteiro and collaborators as an indication of recent divergence, instead of co-specificity. In similar cases described by Flores and Noireau, alloenzymes revealed neglectable differentiation between closely related species within the phyllosoma and oliveirai complexes.
Finally, identification through alloenzymes of specimens from the field (including nymphs) has been crucial in the development of thorough eco-epidemiological investigations. After the definition of alloenzymatic markers for various species and populations, Noireau and collaborators deeply investigated the ecotopes, the behavior and the epidemiological significance of various Bolivian triatomines. An important discovery was the finding that wild spawning areas of T. infestans (Figure 15) are much larger than what was previously believed, extending from the Andean plateaus to the areas of the Bolivian Chaco.
The evolutionary interpretation of alloenzymatic data has been carried out by interpreting the direct relationship between zymography and genes. Phylogenetic relations can be investigated by comparing genetic distance measures derived from allele frequencies. Cladistic analysis are based on the identification capacity for primitive states (plesiomorphic) and states derived (apomorphic) from the characteristics of the group under study, by comparing with an outgroup. Only derived and shared characteristics (synapomorphic) are used in phylogenetic inferences, as reviewed by Avise and Thorpe and Solé-Cava. Based on these premises, the phylogenetic relationship of various groups of triatomines has been explored using alloenzymes. In general, these studies involve the evaluation of genetic distances between a few closely related species, but more complete investigations were carried out with members of the Rhodniini tribe. Using phenetic techniques, Chávez and collaborators identified three main groups within the Rhodnius genus: a basal group (pallescens (ecuadoriensis-colombiensis)) and two sister groups (prolixus (nasutusneglectus)) and (stali (pictipes-brethesi)). Most of these relations were confirmed by cladistic analysis, although the group (brethesi (pictipes-stali)) has appeared as the most basal in the cladogram. The position of Psammolestes coreodes within the tribe could not be resolved successfully, according to Dujardin and collaborators. More recently, the phylogeny of Rhodniini was studied using 12 enzymatic loci. Genetic distance analysis produced a dendrogram that showed the paraphylia of Rhodnius, with Psammolestes tertius as sister group of ‘group prolixus’ within the cluster (domesticus (Ps, Tertius (nasutus-neglectus) (prolixus-robustus))). A second group consisted of (pictipes (brethesi (pallescens-ecuadoriensis))) and was described by Monteiro and collaborators.
Biochemical systematics: Use of cuticular hydrocarbon patterns in triatomines as taxonomic markers
Catarina Macedo Lopes1, Gustavo Calderón2, Juan Girotti2 e Patrícia Juarez2
1- Department of morphology and ultrastructure of vector arthropods, Oswaldo Cruz Institute/Fiocruz
2- Instituto de Investigaciones Bioquímicas de La Plata da Facultad de Ciências Médicas de La Plata (CONICET – UNLP), Argentina.
It is increasingly necessary to obtain knowledge of triatomine species with wide and different areas in their geographical distribution, especially those with high potential to invade and/or reinfest peri- and intradomiciliary areas, even after control actions. Ávila-Pires described the need for a rational control of vector insects based on detailed knowledge of their habits, their biology and the biochemistry of their structure, as constant alterations made to the natural environment have led to an umbalance in ecological relations, fomenting the process of dispersion and domiciliation of species of insects that work as vectors of pathogens. There is consequently a need to identify sensitive techniques that can identify populational variations, with the goal of outlining the genetic variability of triatomines. In this way, various ongoing studies use morphological, molecular and biochemical markers containing the techniques that trace the profile of insects’ hydrocarbons.
Like all insects, triatomines have bodies covered with a thin layer, called epicuticle, mostly consisting of long-chain hydrocarbons, fatty alcohols and sterols of long and complex non-polar chains. Thanks to this chemical composition, this layer has different functions, ranging from thermal insulation to barrier against dehydration and chemicals and against the hazardous action of micro organisms.According to Carlson and Service, cuticular hydrocarbon profiling has been used with plenty of success in the elucidation of the specific status of cryptic species, in different groups of insects, proving to be useful to identify species, subspecies, and even castes. This same meethod has been used to study the Triatominae subfamily by Juarez and Brenner and by Lopes and collaborators, as well as for the classification of the three main genuses of triatomines involved in the transmission of the pathogenic agent of Chagas disease, by Juarez and collaborators. Recently, Calderón and collaborators carried out the analysis of the cuticular hydrocarbon profile presented by geographical populations of Triatoma dimidiata.