Isozyme Evidence of Lack of Speciation Between Wild and Domestic Triatoma Infestans (Heteroptera: Reduviidae) in Bolivia’

1987 Dujardin et al.: Lack of speciation in wild and domestic T. i,+sta?stn,ls J. Med. Entomol. Vol. 24. no. 1: 40-45 January 1987

5 196; b\ lhe Enramolagical Socicu of Amenca

ISOZYME EVIDENCE OF LACK OF SPECIATION BETWEEN WILD AND DOMESTIC TRIATOMA INFESTANS (HETEROPTERA: REDUVIIDAE) IN BOLIVIA’

J.P. Dujardin,2.s M. Tibayrenc?*-‘.jE. Venegas,6 L. Maldonado,6 P. Desjeux?.’ and F.J. Ayala5

Abstract. UÏld and domestic Triatoma b+stans from probably fed on wild animals (Usinger et al. the Cochabamba region of Bolivia were virtually iden- 1966) and was less numerous and less wide- tical at 19 gene loci coding for enzymes. ìio allele \,.as spread than at present. distinctive of either the wild or the domestic popu- Wild T. ii$sfesln,2s population. occur in the lations. Hence, there is no evidence that the 2 pop- Cochabamba I’alley of Bolivia, t\.here they live ulations are different species. Domestic populations separated by 20 km shoL%.edstatistically significant in rocky habitats associated with wild guinea pigs (Fig. 1). \t‘¡Id T. ii2festam are morphologi- ’ differences in allelic frequencies: this is compatible’ with the hypothesis of other authors that migrations cally indistinguishable from domestic speci- of T. iq’2stans are limited Ghen feeding conditions are mens. satisfactory. Fifteen trypanosomatid stocks isolated To clarify the taxonomic and e\olutionary from wild T. i?zJ>:taiis \,.ere shown by isozyme analysis relationships between wild and domestic T. in- to be Tr$anosonia rruri. This provides evidence that festam, we compared them genetically by means wild T. i+staizs are involved in the Chagas‘ disease of isozyme anal!sis and similarly compared the cycle. The T. crui isozymic strains from wild T. in- T~@iiosomacru3 isozymic strains isolated from fpjtans were genetically similar to those isolated from them. The results bear on the important ques- domestic T. i+vtans in the same region. Thissupports tion of whether xild ijtfestans might be in- the hypothesis that there is no speciation between wild T. and domestic T. injjtai::. and that wild and domestic volved in the domestic cycle of Chagas‘ disease. T. crud cycles may overlap in this region. MATERIALS AND METHODS Among T. crici vectors, Triatoma iiqtstaiii We studied 32 T. i+es/a,zs from a \\.¡Id pop- (Klug) is the best adapted to human habitats ulation and 86 from 3 domestic populations in (Lent & Wygodzinsky 1979). Its wide distribu- the Cochabamba region, Department of Cocha- tion in South America may be due to the recent bamba, Bolivia, where the wild populations oc- spread of human populations on that continent. cur (Table 1, Fig. 2). Some 3rd-, 4th-, and 5th- Before it adapted to human habitats, T. in,hstaizj instar nymphs were included in the present study, as were male and female adults. Indeed, the enzyme patterns studied here shoI\-ed nei- ther ontogenic nor sex specificity (Dujardin, un- 1. This study was carried out with the support of publ.). Domestic T. ii2festuizs populations sam- the French Technical Cooperation Program, of the pled from or near human habitats were F:G 1. Rocky habitat of wild Triatoma infestati:. Feces from guinea pig hosts can be seen on stones at bottom. Ministère de l’Industrie et de la Recherche (extension .Irea covered by photograph is ca. 1 m*. of grant no. PVD/81/L-1423), and of the LSDP/ designated “I,” ”2,” and “3.” The xsild pop- I\.orld Bank/\$‘HO Special Programme for Research ulation, collected from wild guinea pig (Caiw and Training in Tropical Diseases. sp.) habitats about 15 km south of the town of isomerase (EC 5.3.1.9, GPI); aglycerophos- 1.1.I .44, 6PGD). Conditions for sample prep- 2. Institut de Yédecine Tropicale “Prince Leo- Cochabamba, was designated population “S.” phate dehydrogenase (EC 1.1.1.8, aGPD): iso- -aration, electrophoresis, and enzyme staining pold,” 155 Kationalestrzat. B-2000 Anr\verpen. Bel- Domestic population 3 and the wild population citrate dehydrogenase (EC 1.1.1.42, IDH): leu- have been previously described (Dujardin & Ti- gium. Send reprint requests to this address. S are separated by ca. 1 km; population 2 is the cine aminopeptidase (cytosol aminopeptidase, bayrenc 1985). 3. Instituto Boliviano de Biologia de Altura. Em- farthest from them, ca. 20 km away. EC 3.4.1 1.1, LAP); malate dehydrogenase IEC Fifteen trypanosomatid stocks were isolated bajada de Francia. Casilla S24, La Paz, Bolivia. Genetic analysis was performed b! electro- 1.1.1.37, JIDH); malate dehydrogenase (osalo- from wild T. in$estam using the method de- 4. Entomologie medicale. SSC ORSTOSI. 70-74 phoretic separation of isozymes on cellulose acetate decarboxylating) (SADP+) or malic en- scribed by Tibayrenc et al. (1982). These were route d‘A~il~t~.93140 Bondy, France. acetate plates (Helena Laboratories, Beaumont, z) me (EC 1.1.1.40, ME): peptidase 1 (ficin. EC also studied by enzyme electrophoresis on cel- 5. Department of Genetics. University of Califar- 3.4.22.3, formerly EC 3.4.4.12, substrate: leu- :lia. Davis, California 956ld. USA. TX). The following enzyme systems were as- lulose acetate plates. The 12 enzyme systems 6. Universidad San Simon. Cochabamba, Bol¡\ ia. sayed: aconitase (aconitate hydratase, EC 4.2.1.3, cyl-leuc\I-leucine): phosphoglucomutase (EC assayed were glucose-6-phosphate dehydroge- 7. Institut Pasteur, 28 rue du Doctcur Rous. 75015 .4COX); glucose-6-phosphate dehydrogenase 5.4.2.2, formerly EC 2.7.5.1, PGlI): and nase; glucose-&phosphate isomerase: glutamate Paris, 1:rancc. ... 6-phosphogluconate dehydrogenase (EC (EC 1.1.1.49, GGPD): glucose-6-phosphate dehydrogenase NAD+ (EC 1 .4.1 .2, GDH p \1’;%B fi ',p

42 J. Med. Entomol. Vol. 24, no. 1 1987 Dujardin et al.: Lack of speciation in wild and domestic T. hzjkk71u 43

TABLE1. Genotype numbers and allele frequencies NAD+); glutamate dehydrogenase NADP+ (EC at 2 loci in 4 populations of Triofoina iifezfaiu. 1.4.1.4, GDH NADP+); isocitrate dehydroge- ochsbamba nase; leucine aminopeptidase; malate dehydro- Populations (sample date and size) genase; malic enzyme; peptidase 1 (substrate: s: leucyl-leucyl-leucine); peptidase 2 (bromelain, 1: 3: Laguna EC 3.4.22.4, formerly EC 3.4.4.24, substrate: Cerro 2: Laguna Angos- San Pueblo Angos- tura leucyl-L-alanine); phosphoglucomutase; and San Miguel Miguel Maica tura strands 6-phosphogluconate dehydrogenase. The tech- tl, (Dec. (Feb. (Mar. (Mar. Genotypes 198% '1984; 1984; 1984; nical conditions were as described previously by and alleles 31 = 12) n,= 44*) 11 = 30*) 11 = 32) Tibayrenc & Le Ray (1984). Two reference strains were added on each electrophoresis plate: Genotype numbers Tehuantepec (isozyme strain 12) and Tulahuen Pgn1/1 1 , O 4 3 (isozyme strain 43); the classification of T. cruri Pgin 1/2 2 6 9 8 isozyme strains is according to Tibayrenc et al. Pgrn2/2 9 38 17 21 (1986b). x*** 0.01*** 0.01 0.33 0.29 P 0.9 0.9 0.5 0.5 RESULTS - -_ 6Pgd 1/1 O 'O 4 2 Wild and domestic TriaYdÏna infestans 6Pgd 1/2 8 17 13 16 No 6Pgd2/2 4 23 9 14 diagnostic loci between the wild and domestic populations were found among the 19 X' 0.6 0.33 0.03 0.15 gene loci (11 enzyme systems) studied. The P 0.3 I0.5 0.5 0.5 nearest domestic population was viftually iden- Allele frequencies tical to the wild population genetically, and the Pgm 1 0.167 0.068 0.283 0.219 other 2 domestic populations were only slightly Pgtn 2 0.833 0.932 0.717 0.781 different from them and about equally different 6Pgd 1 0.333 0.213 0.404 0.313 from both. 6P~d2 0.667 0.787 0.596 0.687 Sixteen gene loci (8 enzyme systems: ACON, / , * At the 6Pgd locus,. only 40 individuals were GGPD, GPI, IDH, LAP, MDH, ME, and PEP I. analyzed in population 2 and 26 in population 3. 5000111, 3&S 1) have been reported as monomorphic in T. : Laguna ** Because numbers were low, the 2 less numer- iilfstam (Dujardin & Tibayrenc 1985). The Angostura ous genotypes were combined in the calculations of present study confirms this and reveals exactly all xp's. FIG.2. Vicinity of Cochabamba, Bolivia, showing the collection sites. the same enzyme patterns in both the wild and *** Yates' corrected x*. domestic specimens. In other domestic populations from Bolivia, aGPD exhibits variability pattern in the heterozygotes) and 6PGD is di- (Dujardin, unpubl. data), but it is monomorphic meric (3-banded pattern in the heterozygotes) in the present samples, which show the same (Tibayrenc et al. 1981; Dujardin & Tibayrenc TABLE2. Standard deviate statistics for painvise comparison between allelic frequencies using individual or phenotype in wild and domestic bugs. This low 1985). grouped populations. Levels of significance (P) are in parentheses. The results are for Pgm (above) and 6Pgd level of polymorphism cannot be due to our Chi-square tests showed no significant depar- (below). electrophoretic procedures, because similar tures from the expectations of Hardy-Weinberg methods have been used to study the Tvpaizo- equilibrium for any of the 4 populations (see soina cruzi stocks, which exhibit high allozyme Table 1). If the data for the 4 populations are * variability. pooled, the combined genotype numbers are 2 Table 1 gives results for the 2 polymorphic also consistent with a Hardy-Weinberg equilib- 1.21 (NS)** loci, Pgm and 6Pgd. For each locality we list the rium (x' = 0.54 and 3.14, P > 0.3 and > 0.2, 3 1.11 (NS) 3.55 (0.001) number of individuals exhibiting each genotype forPpn and 6Pgd, respectively). ForPgin we had 0.59 (NS) 2.37 (0.02) * and the allele frequencies. These 2 loci are poly- to pool the 2 less numerous genotypes to obtain S 2.72 (0.01) 0.82 (NS) morphic in other domestic populations (Ti- sufficiently large expected numbers, and so there 0.18 (NS) 1.36 (NS) 1.02 (NS) bayrenc et al. 1981; Dujardin & Tibayrenc is only 1 degree of freedom: for 6Pgd there are 1+3 - 3.28 (0.001) L 0.44 (NS) 1985), and only 2 alleles that are the same have 2. In Table 2 standard deviate statistics are pre- 2.31 (0.03) L ' 0.85 (NS) been observed in the 4 populations now sur- sented for comparing the allelic frequencies in 1+3+S - 3.29 (0.001) - - veyed. The heterozygous patterns observed in the various populations: e = (p, - pB)/(pq/NA+ - 2.13 (0.04) L - these populations are consistent with the qua- pq/-YB)%, where p, and pBare the frequencies * Numbers are too small to make calculation meaningful. ternary structures previously inferred for do- of a given allele in populations A and B, p is the ** h'S = not significant,i.e., P > 0.05. mestic samples: PGM is monomeric (2-banded frequency of this allele in the 2 populations con- J. 1987 Dujardin al.: wild and domestic T. II Med. Entomol. Vol. 24, no. 1 et Lack of speciation in infestam 45 I 44

' sidered as one, q = 1 - p, and NA and NBare 39). Hence, the 2 isozyme strains isolated from seems most probable, because the triatomine Nei, M. 1972. Genetic distances between popula- . the numbers of alleles sampled in each popu- wild T. infestans are strains already recorded in bugs that we have found in these wild guinea tions. Am. h'at. 106 283-92. lation (see Schwartz 1963, p. 58). Populations domestic bugs and the numbers of stocks of the pig habitats were naturally infected. In addi- Schofield, C.J. 1985. Population dynamics and con- trol of Tri~fomainfesfans. Ann. Soc. Belge Med. Trop. 2 and 3, which are farthest apart geographi- tion, we found eggs, exuviae, and larvae of 2 isozyme strains of wild and domestic origin Tri- 65(S~ppl.1): 149-64. in wild guinea pig nests, indicat- cally, have significantly different allele frequen- are not significantly different (x' = 0.61, P = atoma infestam Schwartz, D. 1963. Méfhodes sfafisfiquesà l'usage des cies at both loci Pgm and 6Pgd. Population 2 is 0.3). ing that the triatomine bugs breed in these nests. médecins et des biologiites. Flammarion Ed., Paris. significantly different from the wild population Given that no other warm-blooded animals are 318 p. I at the Pgm locus but not at 6Pgd. Population 1 DISCUSSION available, the bugs probably also feed on wild Tibayrenc, M., L. Echalar & Y. Carlier. 198 1. Don- is not significantly different from any of the The comparison between wild and domestic guinea pigs. In this area, therefore, the wild I nées de génétique formelle pour 6 loci isoenzy- other 3 populations at either locus. Population Triatoma infstans suggests that they are mem- population of T.infestans could act as a reservoir I matiques chez Triafomainfestans (Hemiptera, Re- 3, which is geographically closest to the wild bers of the same species. No distinctive allele to recolonize houses following attempts to con- duviidae). Cah. ORSTOM Ser. Entomol. Med. population (ca. 1 km), does not differ from it at exists in either group: all 17 monomorphic loci trol domestic populations of this species. Parasitol. 19 121-23. either locus. are identical and the 2 polymorphic loci exhibit We have found significant differences in al- Tibayrenc, M., L. Echalar & P. Desjeux. 1982. Une méthode simple pour obtenir directement des When the populations were grouped the re- the same alleles with similar frequencies. The lelic frequencies between domestic populations ' isolats de TgFanosoma cruri à partir du tube di- sults were as follows: Population 2, geographi- of T. separated by relatively short dis- genetic distances between the wild population infstans gestif de l'insecte vecteur. Coh. ORSTOM Ser. cally most distant from the others, was signifi- and the domestic ones are extremely small. The tances (20 km). This is consistent with the hy- Enfomol. Med. Paraatol. 20: 187-88. cantly different from thelother 3 populations- I pothesis of Schofield (19.85) that has - lack of speciation between wild and domestic T. T. infestans Tibayrenc, M., A. Hoffman, O. Poch, L. Echalar, F. - I. combined and also from the pooled data for infestans is corroborated by the fact that the 2 little tendency to migrate when it enjoys satis- Le Pont, J.L. Lemesre, P. Desjeux & F.J. Ayala. populations 1 and 3, but these 2 combined were Trypanosoma mziisozyme strains found in wild ; factory feeding and habitat conditions. Never- 1986a. Additional data on T~ypanosomacruii iso- not significantly different from the wild popu- T. infestuns are the same ones found in domestic theless, we were unable to find any deviation zymic strains encountered in Bolivian domestic lation. T. infestuns from the same region, in comparable from Hardy-Weinberg equilibria, either within transmission cycles. Trans. R. Soc. Trop. Mcd. HJg. We calculated the genetic distance, D (which frequencies. populations or between them. This suggests that 8 80:442-47. estimates the average number of codon differ- Wild and domestic T. infestuns might differ at the populations, at the level studied, tend to- Tibayrenc, M. & D. Le Ray. 1984. General classification of the isoenzymic strains of TTpanosoma ences per locus; Nei 1972), between the various some loci not surveyed in the present study, but wards panmixia. (Schizof~panutn)CWLI and comparison with T. (S.) populations. For the comparison between the the similarity of allelic and genotypic fiequen- I LITERATURE CITED c. marenkelki and T. (Herpefasoma)rangeli. Ann. wild population and population 3, which is geo- cies at the loci sampled make it highly unlikely Soc. Belge Med. Trop. 64: 39-248. graphically closest, D = 0.001; for the compar- that these populations have been breeding in- ' Ayala, F.J. 1982. Biologie moléculaire et holution. Mas- Tibayrenc, M., P. Ward, A. Moya & F.J. Ayala. isons between the wild population and popu- dependently for very long. Indeed, the genetic 1 son Ed., Paris. 136 p. 1986b. Natural populations of Tvpanosoina cruri, lations 1 and 2, D = 0.001 and 0.002, distadces observed are very small, even for local Dujarpin, J.P. & M. Tibayrenc. 1985. Etude de 11 ' the agent of Chagas' disease, havea complex mul- respectively. For the comparisons between pop- populations of the same species (Ayala 1982). enzymes et données de génétique formelle pour ticlonal structure. Proc. Natl. Acad. Sci. USA 83 19 Triafoinainfesstans. Ann. ulation 1 and either 2 or 3, D = 0.001; and for We then accept as a working hypothesis that loci enzymatiques chez 115-19. Soc. Belge Med. Trop. 65: 271-80. the comparison between populations 2 and 3, the wild and domestic are members Usinger, R.L., P. Wygodzinsky& R. Ryckman. 1966. T. infstans Lent, H. & P. Wygodzinsky. 1979. Revision of the the 2 domestic populations farthest apart, D = of the same species. The biosystematics of Triatominae. Annu. Rev. Triatominae (Hemiptera, Reduviidae), and their Enfomol. 11: 309-30. 0.004. These results confirm the extreme ge- There are 2 obvious alternatives as to the significance as vectors of Chagas' disease. Bull. netic similarity among the 4 populations. origin of the wild population. First, it might be Am. iMus. Ara¿.Hisf. 163: 125-520. a relic of the ancestral population from which Analysis of the trypanosomatid stocks the domestic T. infestans derived. The great ge- isolated from wild T. infestans netic similarity between the wild and domestic The 15 trypanosomatid stocks isolated from populations could be a consequence of occa- wild T. infestans were assayed for 13 genetic loci sional gene migration (interbreeding) between (YE exhibiting the activity of 2 loci). According them. The 2nd alternative is that the wild pop to their. isozyme patterns, all 15 stocks were dation has derived from domestic T. infestans determined as Tvpanosoma cruzi. This is evi- that would have relatively recently colonized dence that wild T. infestans are indeed vectors wild guinea pig habitats. Only the 1st possibility in the Chagas' disease cycle. has been previously considered by other au- Using the classification presented by Tibay- thors (Usinger et al. 1966); yet the isozyme data renc et al. (1986b), the 15 stocks are as follows: are equally consistent with both alternatives. 11 are identical to isozyme strain 20 (IS 20), 3 An important consideration epidemiological- are identical to IS 39, and 1 is a mixed stock of ly is that the wild T. infestans is infected with these 2 strains. We have isolated from domestic Tvpanosoma cruzi and does not appear to be T. infestans in the same region 14 T. cruzi stocks reproductively isolated from the domestic T. (Tibayrenc et al. 1986a). Eight belong to IS 20, infestans. No data are available on the possible 3 are IS 39, and 1 is IS 32 (closely related to IS infection of wild guinea pigs by T. cruzi, but this