Am. J. Trop. Med. Hyg., 60(3), 1999, pp. 377-386 Copyright 1999 by The American Society of Tropical Medicine and Hygiene

Abstract. Kissing bugs or triatomines (Reduviidae: Triatominae) are vectors of the Chagas' disease agent Trypanosoma cruzi. There is a current need for more sensitive tools for use in discrimination of different bug populations and species, thus allowing a better understanding of these insects as it relates to disease transmission and control. In a preliminary analysis of the mitochondrial large subunit ribosomal RNA (mtlsurRNA) and cytochrome B (mtCytB) genes, we used DNA sequencing to study species identification and phylogeny. In both examined gene regions, about 46% of nucleotide positions exhibited polymorphism. The examined region of mtCytB appears to have evolved more rapidly than the examined region of mtlsurRNA. Phylogenetic analysis of both gene fragments in the examined species produced similar results that were generally consistent with the accepted taxonomy of the subfamily. The two major tribes, Rhodniini and Triatomini, were supported, along with additional clades that corresponded to accepted species complexes within the Rhodnius and Triatoma genera. The one chief exception was that Psammolestes coreodes sorted into the Rhodnius prolixus-robustus-neglectus clade, with bootsrap values of 99% and 81%, respectively, for the mtlsurRNA and mtCytB fragments. All of the individual species examined could be distinguished at both genetic loci.

    The Triatominae (Hemiptera: Reduviidae) comprise a subfamily of hematophagous insects that are vectors for the parasitic agent of Chagas' disease, Trypanosoma cruzi¹. This protozoan is usually transmitted during nocturnal feeding by domestic insects. Chagas' disease is a substantial public health problem in the Americas, affecting approximately 1618 million people throughout this region, with an additional 100 million estimated to be at risk.²

    The Triatominae are widely distributed primarily in the New World, their center of diversity and likely site of origin.³ One hundred twenty-five species are currently recognized. In the Americas, these bugs occupy diverse habitats from southern Argentina to the Great Lakes of North America. Primary habitats for these bugs are nests of birds and mammals, which provide a stable source of food and shelter.

    Triatomine bugs are distinguished from other Reduviidae by features related to obligate hematophagy, a primary factor in their biology, distribution, and evolution.^1,4 Traditionally, the Triatominae have been classified as a monophyletic taxon derived from a bloodfeeding ancestor.¹ However, Schofield has proposed that the Triatominae are polyphyletic and that hematophagy arose from diverse predatory ancestors.³ In this view, morphologic similarities reflect convergent apomorphic characters related to hematophagy and shared ecology.

    Genetic analysis of triatomine phylogeny has mainly involved examination of isozyme variability, cytogenetic features, or random amplified polymorphic DNA profiles, which all suggest that the Triatominae are polyphyletic.^1-15 However, more sensitive tools that are able to distinguish both different populations of bugs as well as morphologically similar species are needed for a better understanding of triatomine population structure and biology. Greater definition of phylogenetic relations, species status, and rates of gene flow could be obtained from DNA sequence analysis of appropriately selected genes. A recent study has reported preliminary examination of triatomine phylogeny based on mitochondrial DNA sequence analysis. ¹⁶

    Here we report a preliminary study of mitochondrial DNA (mtDNA) sequence determination for species identities and phylogenetic relationships among the Triatominae. Molecular analysis of mtDNA is widely used for phylogenetic and population studies and mitochondrial gene phylogenies are representative of species relationships. 17, " Also, universal primers have been designed for polymerase chain reaction (PCR) amplification of the insect mitochondrial genome, allowing rapid molecular analysis.19 We have examined regions of the cytochrome B (mtCytB) and large subunit ribosomal RNA (mtlsurRNA) genes from individuals representing five genera of the Triatominae. Sequence comparisons of these gene fragments were useful for distinguishing all of the examined triatomine species and for a preliminary evaluation of phylogenetic relationships within the subfamily Triatominae.

    Triatomine specimens. The DNA sequence analysis of regions of the mtCytB and mtlsurRNA genes was performed on the 17 triatomine species and one outgroup reduviid (nontriatomine) species listed in Table 1. The traditional taxonomic classification and source material of each species examined in this study are also indicated, along with the GenBank accession numbers of each sequence reported in this study.

    Isolation and purification of DNA. The DNA from individual bugs was purified using the Wizard Genomic Purification Kit (Promega Corporation, Madison, WI), following the manufacturer's protocol as recommended for DNA isolation from animal tissue. The DNA was extracted either from eggs, if available, or from dissected leg muscle tissue.

    Oligonucleotide Primers for PCR amplification and direct DNA sequencing. Forward and reverse oligonucleotide primers, used for PCR amplification and DNA sequencing of the mtlsurRNA region, were designed based on the consensus sequences of two highly conserved locations within the mtlsurRNA gene of six invertebrate species deposited in GenBank. The species chosen for primer design and the corresponding GenBank accession number of their mtlsurRNA genes were: Anopheles gambiae (L20934), An. quadrimaculatus (LO4272), Apis mellifera (LO6178), Artemia franciscana (X69067), Drosophila melanogaster (U37541), and Locustus migratoria (X80245). Alignment and comparison of these sequences revealed two sites each approximately 20 basepairs (bp) long, lying nearly 400 bp apart, that exhibited substantial sequence similarities among the selected invertebrate species. The primer sequences used are the following, with each number in brackets signifying the nucleotide position in the An. gambiae mitochondrial genome that corresponds to the 5'-most nucleotide position of the primer sequence:²⁰

Forward:      LRN 13393 [13303]

5'-C(G/A)C CTG TTT AAC AAA AAC AT,

or

LRN 13393A [13303]

5'-CAT CTG TTT A(A/T)C AAA (A/G)AC AT;

Reverse:       LRJ 12966 [12882]

5'-AAA AAA ATT ACG CTG TTA TCC CTA AAG TAA,

or

LRJ 12966A [12882]

5'-AAA AAA ATT ACG CTG TTA TCC CTA A(A/G)G TAA,

or

LRJ 12966B [12882]

5'-AAA AAA ATT ACG CTG TTA TCC CTA A.

    By the same strategy, oligonucleotide primers were designed for PCR amplification and DNA sequencing of a region of the mtCytB gene. Alignment and comparison of this gene region from the same invertebrate species as above revealed two sites (each at least 20 hp in length), lying approximately 500 hp apart, which exhibited substantial sequence similarities. The derived primer sequences are the following, with each number in brackets signifying the nucleotide position in the An. gambiae mitochondrial genome that corresponds to the 5'-most nucleotide position of the primer sequence:²⁰

Forward:      CYT BF [10821]

5'-CGA CAA ATA TCA TTT TGA GGA GCA ACA G,

or

CYT BF1 [10821]

5'-GGT CAA ATA TCA TTT TGA G(T/G)A GC(T/A) AC(T/A) G,

or

CYT BF2 [10813]

5'-TAC CAT GAG C(A/T)C AAA TAT CAT (T/A)TT GAG;

Reverse:       CYT BR [11290]

5'-ATT ACT CCT CCT AGC TTA TTA GGA ATT G,

or

CYT BR1 [11277]

5'-ATT TAT TAG GAA T(A/T)G ATC GTA AAA T(T/A)G,

or

CYT BR2 [11309]

5'-ATT TGA TAT AAC TAA (T/A)GC AAT (A/T)AC TCC TCC.

    All primers used in this study were synthesized by the Centers for Disease Control and Prevention/National Center for Infectious Diseases (CDC/NCID) Biotechnology Core Facility. Primers LRJ 12966, 12966A, and 12966B are sequence modifications of the oligonucleotide primer LR-J13017.21 Primers LRN 13393 and 13393A are sequence modifications of the oligonucleotide primer LR-N-13398 and primer CYT BR is a sequence modification of the primer CB-N-11367.¹⁹

    Amplification of DNA by the PCR. The PCR amplification was performed in 50-m L reactions containing 1 m L (2.5 units) of Amplitaq DNA polymerase (PE Applied Biosystems, Foster City, CA) and 1 m L of DNA template, using either a GeneAmp 2400 or 9600 PCR Instrument System (PE Applied Biosystems), according to supplier's recommendations. The following reaction conditions were used: denaturation at 94° C for 5 min; then 35 cycles of denaturation at 94° C for 30 sec, annealing at 45-47° C for 30 seconds, and extension at 72° C for 1-2 min; followed by a 72° C extension for 7 min, and 4° C indefinitely. In some samples these reaction conditions failed to produce a product. For those cases, the following conditions were used: denaturation at 94° C for 5 min; then 16 cycles of denaturation at 94° C for 30 sec, annealing at 52° C for 30 sec (reduced 1° C/cycle), and extension at 72° C for 1-2 min. The reactions then continued for 21 cycles of denaturation at 94° C for 30 sec, annealing at 55° C for 30 sec, and synthesis at 72° C for 1-2 minutes; then extension at 72° C for 7 min and 4° C indefinitely. Successful amplification was confirmed by examining a 5-m L aliquot of the amplification product. Amplified products were then purified for DNA sequence analysis with the Wizard PCR DNA Purification System (Promega Corporation), according to supplier's specifications, and stored at -20'C.

    Sequence analysis of DNA. Purified PCR amplification products were examined further by direct, automated fluorescence sequencing using the ABI Prism Dye Terminator Cycle Sequencing Kit (PE Applied Biosystems) and following the manufacturer's recommendations. The same oligonucleotide primers used for PCR amplification were employed for the sequencing reactions. Products of the sequencing reactions were then purified with Centri-Sep columns (Princeton Separations, Adelphia, NJ), according to the supplier's recommendations, vacuum dried, and stored at -20'C in the dark until used. For analysis, samples were resuspended in 6 m l of formamide:EDTA/blue dextran (5:1) and subjected to electrophoresis a 4% acrylamide gel using an ABI Prism 377 Automated DNA Sequencing apparatus (PE Applied Biosystems). Downstream analysis of the sequence data was performed using Sequence Navigator version 1.0.1 (PE Applied Biosystems). Further sequence analyses and comparisons were performed using various programs in the GCG Wisconsin Package, Version 9.1.²²

    Phylogenetic analysis of the samples was performed by the neighbor joining (NJ) method, with Galtier and Gouy distances.23,24 This model does not assume that sequence base composition (A, T, G, and C) is at equilibrium, and so corrects for unequal base composition. It is, therefore, applicable to analysis of insect mitochondrial DNA, which has a high A + T content.14,21 Tree reliability was assessed by the bootstrap method with 500 replications.26 Only bootstrap values above 70% are reported.27 The nodes with bootstrap support between 70 and 95% should be considered tentative, and require additional data for analysis.28 All NJ analyses were performed using the TreeconW program.29

RESULTS

    Analysis of the mtlsurRNA gene region. The examined region of the mtlsurRNA gene from 18 reduviid species (Table 1),corresponds to nucleotides 12,882 to 13,303 inclusively of the An. gambiae mitochondrial genome.20 Using different combination sets of the designed forward and reverse primers (see Materials and Methods) allowed amplification of this gene product from all examined specimens, a selection of which are shown in Figure 1 A. A single product was generated by PCR amplification and detected by agarose gel electrophoresis for each examined species. For each species, a reaction product of approximately 400 hp was obtained (Figure 1A), in good agreement with the predicted product size of 421 bp (see Materials and Methods).

    A 383-bp nucleotide sequence for the forward and reverse strands was obtained for each examined species (Figure 2).The putative mtlsurRNA DNA sequences of Triatoma infestans and Rhodnius prolixus were examined for homology to sequences in the Genbank database and found to have highest identity (> 75%) to other animal mtlsurRNA sequences in a minimum 368-bp overlap. The overall base composition of the mtlsurRNA sequences was approximately 43% A, 18% C, 9% G, and 30% T, exhibiting a high A: T rich composition (73%) as expected for insect mitochondrial DNA. ²⁵ A comparison alignment of the mtlsurRNA sequences of all examined species is shown in Figure 2. Of the 383 nucleotide positions of examined mtlsurRNA gene sequence in these species, 164 nucleotide positions (approximately 43%) exhibited polymorphism. Changes in 25 nucleotide positions involved insertions or deletions. The remaining sequence modifications comprised simple base substitutions.

    Pairwise comparisons of all the examined mtlsurRNA sequences demonstrated a greater average percent similarity among the Rhodnius sequences (approximately 90%) and among the Triatoma sequences (approximately 89%) than the average similarity of the Rhodnius sequences compared with the Triatoma sequences (approximately 80%). The Rhodnius sequences also exhibited a high average percent similarity with Psammolestes coreodes mtlsurRNA sequence (approximately 88%). The Triatoma sequences had significant percent similarity with the Dipetalogaster maxima mtlsurRNA sequence (approximately 89%) and the Panstrongylus megistus (approximately 88%) mtlsurRNA sequence.

    Analysis of the mtCytB gene region. The portion of the examined mtCytB gene for 18 species of reduviid bugs corresponds to sequences within the cytochrome B gene of An. gambiae mitochondrial DNA (nucleotides 10821-11309, inclusive).20 Using different combination sets of the designed forward and reverse primers (see Materials and Methods) allowed amplification of the gene product from all examined specimens, a selection of which are shown in Figure 1B. A single product was generated by PCR amplification and detected by agarose gel electrophoresis for each examined species. For each species, a reaction product of approximately 500 hp was obtained (Figure 1B), in good agreement with the predicted product size of 456-496 bp, depending on the primer pair used (see Materials and Methods).

    A 399-bp nucleotide sequence of the forward and reverse strands was determined for each examined species (Figure 3A). The putative mtCytB nucleotide sequences of T. infestans and R. prolixus were analyzed for homology to sequences in the Genbank database and found to have highest similarity (> 70%) with other animal mtCytB sequences in a minimum 398-bp overlap. Overall, the base composition of the mtCytB gene sequences was approximately 31% A, 23% C, 12% G, and 34% T, corresponding to a high A:T rich composition (65%) as expected for insect mitochondrial DNA. ²⁵

    Of the 399 nucleotide positions of examined mtCytB gene DNA sequence, 192 nucleotide positions (48%) exhibited polymorphism (Figure 3A). All modifications involved base substitutions. Approximately 70% of the base changes occuffed in the third codon position, 25% involved the first position, and only 5% occurred in the second codon position. Codon frequency analysis of all mtCytB gene sequences demonstrated a preference for A or T bases in the third codon position, reflecting the A:T rich composition of insect mitochondrial DNA. Of 133 characters at the amino acid level, 37 involved an amino acid change (approximately 28%). Comparison alignments of the mtCytB nucleotide and amino acid sequences are shown in Figure 3A and B, respectively.

    Pairwise identity comparison at the amino acid level of mtCytB sequences indicated an average identity between the Rhodnius sequences (approximately 96%) and between the Triatoma sequences (approximately 93%) that exceeded the average identity of the Rhodnius sequences compared with the Triatoma sequences (approximately 91%). The Rhodnius mtCytB sequences exhibited approximately 94% identity with Psammolestes coreodes sequence, while the Triatoma mtCytB sequences exhibited greater than 93% identity with the Dipetalogaster maxima sequence and the Panstrongylus megistus sequence.

    Inferred phylogenetic relations based on mtlsurRNA and mtCytB gene sequences. Results of phylogenetic analysis of the two examined mitochondrial loci are shown in Figure 4 (mtlsurRNA) and Figure 5 (mtCytB). In general, both genes yielded similar results. The examined region of mtCytB appears to have evolved faster than the examined region of mtlsurRNA. Using the wheel bug Arilus cristatus (Reduviidae, Harpactorinae) as an outgroup, phylogenetic analysis of both gene fragments suggested that all of the examined species sorted into one of two main groups, which corresponded to the accepted Triato@ni and Rhodniini tribes. The bootstrap support for this distinction was quite high, 100% and 93%, respectively, for the mtlsurRNA and mtCytB loci.

    Within the Rhodniini clade, three species groups were suggested from analysis of both loci. Rhodnius pallescens and R. ecuadoriensis grouped together with bootstrap values of 94% (mtlsurRNA) and 99% (mtCytB), while R. brethesi and R. pictipes clustered together with bootstrap values of 100% (mtlsurRNA) and 99% (mtCytB). Similarly, R. prolixus, R. robustus, and R. neglectus grouped together with bootstrap values of 100% for both genes. Interestingly, P. coreodes grouped with the R. prolixus cluster. The bootstrap support for this grouping was quite high, 99% and 81% respectively for mtlsurRNA and mtCytB.

    Within the Triatomini branch, two clades were distinguishable with relatively high probabilities. The bootstrap values were generally lower for this region of the tree; however, T. infestans and T. sordida were grouped with a bootstrap value of 97% at the mtCytB fragment. Triatoma dimidiata, T. pallidipennis, and T. sanguisuga sorted into another group with a 92% probability.

DISCUSSION

    Direct DNA sequence analysis of both gene fragments distinguished all of the closely related triatomine species evaluated in this study. The two major tribes were readily distinguished as well as subgroups within these tribes. Within the Rhodniini tribe, three clades appeared to be recognized from analysis of mtlsurRNA and mtCytB (Figures 4 and 5), generally agreeing with previous categorization of these species. Interestingly, for both mitochondrial loci, P. coreodes appeared quite similar to the clade (species group) containing Rhodnius prolixus, R. robustus, and R. neglectus; more so than these species were to species in the other Rhodnius clades. These results were statistically significant, supported by strong bootstrap values for the two loci. Members of the genus Psammolestes are morphologically similar to Rhodnius, to which they have previously been proposed to be closely related.' The analysis also supported the phylogenetic relatedness of R. pallescens to R. ecuadoriensis and of R. pictipes to R. brethesi. Both of these results are consistent with the currently accepted taxonomy of the Rhodniini.

    While R. prolixus, R. robustus, and R. neglectus are morphologically very similar, sequence analysis of the two gene fragments indicates that these species are distinguishable by DNA comparison. In side-by-side sequence comparisons between individuals of each species, significant nucleotide divergence could be detected (Figures 2 and 3A). In comparisons of R. prolixus and R. robustus, a difference of approximately 3.9% (15 of 383 bp) was observed in the mtlsurRNA gene fragment and a 6.8% difference (27 of 399 bp) was observed in mtCytB. Phylogenetic analysis of sequence data from both genes in the specimens examined herein (Figures 4 and 5) also supports the view that R. prolixus and R. robustus are distinct species. Also, in pairwise sequence comparisons between R. prolixus and R. neglectus, a 3.9% difference (15 of 383) bp) was observed in the mtlsurRNA locus, while a 5.5% difference (22 of 399 bp) was detected in the mtCytB gene fragment. Comparisons between R. robustus and R. neglectus revealed a 2.9% nucleotide difference (I 1 of 383 bp) at mtslurRNA and a 7.8% difference (31 of 399 bp) at mtCtyB. Consequently, data from these two loci appear to distinguish this group of similar species; however, this hypothesis awaits confirmation judged on the basis of consistent results obtained among replicates of each species collected from different geographic regions.

    Genetic variation based on isozyme analysis of several loci has been reported between R. robustus and R. neglectus, as well as between R. prolixus and R. neglectus. However, it can be quite difficult to distinguish R. prolixus and R. robustus using this method, prompting the suggestion that R. prolixus and R. robustus may not constitute separate species.',' In these comparisons, we were able to distinguish all three of these species, suggesting that mitochondrial DNA sequence variation at the selected loci appears to be highly sensitive and useful in resolving species identities among closely related members of the Triatominae.

    Although supported by generally weaker bootstrap values than those seen in the Rhodniini, analysis of both loci suggested some clades within the Triatomini that seem to correlate with species complexes of the genus Triatoma (Figures 4 and 5). Phylogenetic analysis of both genes, for example, suggested that T infestans and T sordida (Infestans complex) were related inter se, as were the Central and North American species T dimidiata, T. pallidipennis (Phyllosoma), and T sanguisuga (Lecticularia). Interestingly, D. maxima and P. megistus could not be distinguished from the other Triatomini with any degree of certainty based on bootstrap values.

    While this study demonstrates the utility of mtDNA sequence analysis for taxonomic examination of triatomine vectors of Chagas' disease, we consider the present results preliminary for two reasons. First, we examined only a single specimen for each of the species in the study. Consequently, we have not yet assessed intraspecific sequence variation for comparison with our present results. Second, we analyzed a relatively small amount of sequence data, approximately 782 bp from two different loci. Analysis of more individuals, species, and genetic loci, including both mitochondrial and nuclear, will likely add further definition to the observations reported.

Acknowledgments: We are greatly appreciative of Drs. Chris J. Schofield and Michael A. Miles (London School of Hygiene and Tropical Medicine) for critical reading of the manuscript. We thank Drs. Jose Jurberg (Instituto Oswaldo Cruz, Rio de Janeiro, Brazil) and Jose Ribeiro (National Institutes of Health, Bethesda, MD) for supplying specimens that were used in this studv. We also thank Brian Holloway and the staff of the NCID Biotechnology Core Facility for synthesis of the oligonucleotide primers.

Financial support: This work was supported in part by a grant from WHO/TDR, ID no. 960320 T80/181/148.

Disclaimer: The use of trade names does not constitute endorsement by the U.S. Public Health Service or the Centers for Disease Control and Prevention.

Authors' addresses: Daniel F Lyman, Fernando A. Monteiro, Ananias A. Escalante, and Charles B. Beard, Division of Parasitic Diseases, National Center for Infectious Diseases, Centers for Disease Control and Prevention, 4770 Buford Highway, NE, Mailstop F-22, Atlanta, GA 30341-3724. Celia Cordon-Rosales, Medical Entomology Research and Training Unit - Guatemala, Guatemala City, Guatemala. Dawn M. Wesson, Department of Tropical Medicine, School of Public Health and Tropical Medicine, Tulane University, 1501 Canal Street, New Orleans, LA 70112. Jean-Pierre Dujardin, ORSTOM, CP 9214, La Paz, Bolivia.

REFERENCES

1. Lent H, Wygodzinski P, 1979. Revision of the Triatominae (Hemiptera: Reduviidae) and their significance as vectors of Chagas disease. Bull Am Museum  Natural His 163: 123-520.

2. WHO, 1990. Tropical Diseases 1990. TDR/CTD/HH20.1 Geneva: WHO.

3. Schofield CJ, 1988. Biosystematics of the Triatominae. Service MW, ed. Biosystematics of Haematophagous Insects. Oxford: Clarendon Press, 285-312.

4. Schofield CJ, Dolling WR, 1993. Bedbugs and kissing-bugs (bloodsucking Hemiptera). Lane RC, Crosskey RW, eds. Medical Insects and Arachnids. London: Chapman and Hall, 483516.

5. Harry M, Galindez I, Cariou ML, 1992. Isozyme variability and differentiation between Rhodnius prolixus, R. robustus and R. pictipes, vectors of Chagas disease in Venezuela. Med Vet Entomol 6: 37 43.

6. Harry M, 1993. Isozymic data question the specific status of some blood-sucking bugs of the genus Rhodnius, vectors of Chagas disease. Trans R Soc Trop Med Hyg 87: 492-493.

7. Garcia BA, Canale DM, Blanco A, 1995. Genetic structure of four species of Triatoma (Hemiptera: Reduviidae) from Argentina. J Med Entomol 32: 134-137.

8. Garcia BA, Soares-Barata JM, Blanco A, 1995. Enzyme polymorphism among Triatoma infestans (Hemiptera: Reduviidae) colonies. J Med Entomol 32: 126-133.

9. Lopez G, Moreno J. 1995. Genetic variability and differentiation between populations of Rhodnius prolixus and R. pallescens, vectors of Chagas disease in Colombia. Mem Inst Oswaldo Cruz 90: 353-357.

11. Panzera F, Perez R, Panzera Y, Alvarez F, Scvortzoff E, Salvatela R, 1995. Karyotype evolution in holocentric chromosomes of three related species of triatomines (Hemiptera: Reduviidae). Chromosome Res 3: 143-150.

12. Perez R, Panzera F, Page J, Suja JA, Rufas JS, 1997. Meiotic behavior of holocentric chromosomes: orientation and segregation of autosomes in Triatoma infestans (Heteroptera). Chromosome Res 5: 47-56.

13. Dujardin JP, Schofield, CJ, Tibayrenc M, 1998. Population structure of Andean Triatoma infestans: allozyme frequencies and their epidemiological relevance. Med Vet Entomol 12: 20-29.

14. Solano P, Dujardin JP Schofield CJ, Romana C, Tibayrenc M, 1996. Isoenzymes as a tool for identification of Rhodnius species. Res Rev Parasitol 56: 41-47.

15. Garcia AL, Carrasco HJ, Schofield CJ, Stothard JR, Frame [A, Valente SAS, Miles MA, 1998. Random amplification of polymorphic DNA as a tool for taxonomic studies of triatomine bugs (Hemiptera: Reduviidae). J Med Entomol 35: 38-45.

16. Stothard JR, Yamamoto Y, Cherchi A, Garcia AL, Valente SAS, Schofield CJ, Miles MA, 1998. A preliminary survey of mitochondrial sequence variation within triatomine bugs (Hemiptera: Reduviidae) using polymerase chain reaction-based single strand conformational polymorphism (SSCP) analysis and direct sequencing. Bull Entomol Res 88: 553-560.

17. Avise JC, 1994. Molecular Markers, Natural History and Evolution. New York: Chapman and Hall.

18. Moore WS, 1995. Inferring phylogenies from mtDNA variation: mitochondrial gene trees versus nuclear gene trees. Evolution 49:

Simon C, Frati F Beckenbach A, Crespi B, Liu H, Flook P, 1994. Evolution, weighting, and phylogenetic utility of mitochondrial gene sequences and a compilation of conserved polymerase chain reaction primers. Ann Entomol Soc Am 87.651-701.

20. Beard CB, Hamm DM, Collins FH, 1993. The mitochondrial genome of the mosquito Anopheles gambiae: DNA sequence, genome organization, and comparisons with mitochondrial sequences of other insects. Insect Mol Biol 2: 103-124.

21. Kambhampati S, Smith PT, 1995. PCR primers for the amplification of four insect mitochondrial gene fragments. Insect Mol Biol 4: 233-236.

22. Wisconsin Package, 1997. Version 9. 1. Madison, WI: Genetics Computer Group.

23. Saitou N, Nei M, 1987. The neighbor-joining method: a new method for reconstructing phylogenetic trees. Mol Biol Evol 4: 406-425.

24. Galtier N, Gouy M, 1995. Inferring phylogenies from DNA sequences of unequal base composition. Proc Nat Acad Sci USA 92: 11317-11321.

25. DeSalle R, Freeman T, Prager EM, Wilson AC, 1987. Tempo and mode of sequence variation in mitochondrial DNA of Hawaiian Drosophila. Evolution 42: 1076-1084.

26. Felsenstein J, 1985. Confidence limits on phylogenies: an approach using bootstrap. Evolution 39: 783-791.

27. Hillis DM, Bull JJ, 1993. An empirical test of bootstrapping as a method for assessing confidence in phylogenetic analysis. Syst Biol 8: 189-191.

28. Efron B, Halloran E, Holmes S, 1996. Bootstrap confidence levels for phylogenetic trees. Proc Natl Acad Sci USA 93: 13429-13434.

29. Van de Peer Y, De Watcher R, 1994. TREECON for Windows: a software package for the construction and drawing of evolutionary trees for the Microsoft Windows environment. Comput Applic Biosci 10: 569 570.