Mitochondrial DNA Polymorphisms in Chilean Aboriginal Populations: Implications for the Peopling of the Southern Cone of the Continent

AMERICAN JOURNAL OF PHYSICAL ANTHROPOLOGY 113:19–29 (2000)

MAURICIO L. MORAGA,1 PAOLA ROCCO,1 JUAN F. MIQUEL,2 FLAVIO NERVI,2 ELENA LLOP,1 RANAJIT CHAKRABORTY,3 FRANCISCO ROTHHAMMER,1 AND PILAR CARVALLO1* 1Programa de Gene´tica Humana, Instituto de Ciencias Biome´dicas, Facultad de Medicina, Universidad de Chile, Santiago 7, Chile 2Departamento de Gastroenterologı´a, Facultad de Medicina, Pontiﬁcia Universidad Cato´lica de Chile, Santiago, Chile 3Human Genetic Center, University of Texas School of Public Health, Houston, Texas 77225

KEY WORDS Amerindian mtDNA haplogroups; sequence analysis; peopling of the southern cone

ABSTRACT The mitochondrial DNAs (mtDNAs) from individuals belonging to three Chilean tribes, the Mapuche, the Pehuenche, and the Yaghan, were studied both by RFLP analysis and D-loop (control region) sequencing. RFLP analysis showed that 3 individuals (1.3%) belonged to haplogroup A, 19 (8%) to haplogroup B, 102 (43%) to haplogroup C, and 113 (47.7%) to haplogroup D. Among the 73 individuals analyzed by D-loop sequencing, we observed 37 different haplotypes deﬁned by 52 polymorphic sites. Joint analysis of data obtained by RFLP and sequencing methods demonstrated that, regardless of the method of analysis, the mtDNA haplotypes of these three contemporary South American aborigine groups clus- tered into four main haplogroups, in a way similar to those previously described for other Amerindians. These results further revealed the absence of haplogroup A in both the Mapuche and Yaghan as well as the absence of haplogroup B in the Yaghan. These results suggest that the people of Tierra del Fuego are related to tribes from south-central South America. Am J Phys Anthropol 113:19–29, 2000. © 2000 Wiley-Liss, Inc.

The analysis of mitochondrial DNA vari- and Central America. Of those studies con- ation has been extensively used in the re- ducted with South American groups, almost cent past, for genetic characterization of all have involved populations from the contemporary aboriginal populations of the northern region of the continent (Torroni et Americas. The primary goals of this analy- al., 1993; Easton et al., 1996; Santos et al., sis have been to determine the origins, re- 1996). Most of the data generated from lationships, and migrational patterns of these studies were obtained through RFLP New World populations. In this respect, the study of the frequency distribution of the four founding Amerindian haplogroups, Grant sponsor: FONDECYT; Grant number: 198-1111; Grant sponsor: Ministero degli Affari Esteri d’Italia; Grant number: suggested by Schurr et al. (1990) and de- 1091-G224/ICU/CILE; Grant sponsor: NIH; Grant number: GM fined by Torroni et al. (1992), has proven 41399. *Correspondence to: Pilar Carvallo, Departamento de Biologıá useful. Celular y Molecular, Facultad de Ciencias Biolo´gicas, Pontificia Universidad Cato´lica de Chile, Casilla 114-D, Santiago, Chile. The majority of these studies have fo- E-mail: [email protected] cused on aboriginal populations from North Received 23 November 1998; accepted 28 March 2000.

© 2000 WILEY-LISS, INC. 20 M.L. MORAGA ET AL. mapping, rather than by D-loop sequencing, To expand our knowledge about the ex- although some D-loop sequence data from tent of mtDNA variation in southern South these populations are available for analysis. America and to test the hypothesis of a sin- Similarly, the mtDNAs taken from southern gle wave of migration to the southern part of South American groups were primarily South America, we performed RFLP and characterized by RFLP analysis (Merri- sequence analysis on three Chilean aborig- wether et al., 1995; Lalueza et al., 1997). inal populations (the Pehuenche, the Mapu- Thus, there remains limited D-loop se- che, and the Yaghan). quence information from groups inhabiting MATERIALS AND METHODS the southern part of South America. Studies of South Amerindians using Population samples RFLP and D-loop sequence analysis have The samples include three different Chil- revealed a concordance between the distri- ean aboriginal populations: 105 Pehuenche bution of haplogroups A–D, and haplo- from the community of Trapa Trapa in the groups (lineage clusters) I–IV as defined by Province of Bio Bio, 8th region (37° 43Ј S; Horai et al. (1993), based on the nucleotide 71° 16Ј W), 111 Mapuche from the Huapi polymorphisms occurring in the D-loop of Island, Province of Valdivia, 10th region each haplogroup (Torroni et al., 1993; Eas- (40° 15Ј S; 72° 25Ј W), and the 21 last sur- ton et al., 1996; Santos et al., 1996). How- viving Yaghan from Puerto Williams, Na- ever, Santos et al. (1996) found that 7% of varino Island, Province of Antartica, 12th individuals belonging to tribes from the region (55° S; 67° 40Ј W). The three groups Brazilian Amazon Region comprise unusual analyzed correspond to Amerinds showing a haplogroups because of their lack of corre- percentage of admixture between 2–13% spondence between RFLP and sequencing (Llop et al., 1993, 1994). The geographic loca- analysis. There is also extensive informa- tions of the subjects are shown in Figure 1. tion referring to RFLP analysis in these populations. In addition, Bailliet et al. DNA extraction and PCR amplification (1994) described a subdivision of the classi- Total DNA was prepared from peripheral cal Amerindian haplogroups A, C, and D on blood lymphocytes (Gustincich et al., 1991). the basis of the presence or absence of the DNA amplification of the four polymorphic HaeIII np 16517 site. This division has led mtDNA regions, defining the four Amerin- to the creation of haplotypes A1/A2, C1/C2, dian haplogroups, was carried out by using and D1/D2. However, the HaeIII np 16517 the following specific primers: site has been shown to be highly polymorphic in African (Chen et al., 1995), Euro- Haplogroup A: pean (Torroni et al., 1996), and Asian (Ball- L604 GTAGCTTACCTCCTCAAAGCAA inger et al., 1992) populations, indicating H708 AGGGTGAACTCACTGGAACG that it may not be useful for distinguishing Haplogroup B: within Native American haplogroups A, C, L8214 CACAGTTTCATGCCCATCGT and D. Similarly, Forster et al. (1996) also H8294 ATGCTAAGTTAGCTTTACAGTGG postulated subdivisions of founding haplo- Haplogroup C: types, using D-loop sequence data. In this L13232 CGCCCTTACACAAAATGACATCAA case, they divided haplotypes A1 and A2 by H13344 GGAGCACATAAATAGTATGGC aC3T transition at np 16111 and defined Haplogroup D: D1 and D2 haplotypes by the presence or L5120 CCTAACTACTACCGAATTCCTA absence of the 16271 T3C transition, with H5255 ATTCTTCGATAATGGCCCATTTG this transition being present only in North American Indian mtDNAs. Hence, there PCR was performed in a final volume of would appear to be specific nucleotide 50 ␮l, containing 300 ng genomic DNA, 1 U changes in the D-loop, which distinguish Taq DNA polymerase (Promega), 25 pmoles between North-Central and South Ameri- of each primer, 200 nM of each deoxinucle- can populations. otide, and the appropriate buffer. Samples MTDNA POLYMORPHISMS IN ABORIGINAL CHILEANS 21 ers L29, GGTCTATCACCCTATTAACCAC; and H16401, CACCATCCTCCGTGAAATCA, labeled at the 5Ј end with ␥-32P-ATP. DNA sequence analysis was performed by electro- phoresis on high-resolution 6% polyacryl- amide gels, followed by autoradiography.

Data analysis Sequence alignment was performed by using the Clustal W computer program (Thompson et al., 1994). Substitutions and deletions were inferred from direct sequence comparisons. The phylogenetic trees were obtained through two different methods: neighbor joining (Saitou and Nei, 1987), using the Clustal W program, and parsimony analysis (Swofford and Olson, 1990). Maximum parsimony trees were generated through random addition of sequences using the tree bisection and reconnection (TBR) algorithm, with the PAUP (Swofford, 1993) computer program.

RESULTS RFLP analysis Fig. 1. Map of Chile, showing the geographic loca- tion of the three Chilean tribes included in the present All individuals analyzed showed one of study. the characteristic haplogroups, A, B, C, or D. As shown in Table 1, haplogroups C and D were the most frequent in the three pop- were processed under the following PCR ulations, appearing at very similar frequen- conditions: 1 cycle at 95°C for 5 min, fol- cies in all of them. Haplogroup A was lowed by 30 cycles at 95°C for 1 min, 55°C present in the Pehuenche at a very low fre- for 1 min, and 72°C for 1 min, and finally quency (2.8%), while both haplogroups A one cycle at 72°C for 5 min. Haplotypes A, C, and B were absent in the Yaghan group. We and D were analyzed by restriction diges- tested the gain of the HaeIII site at position tion, using HaeIII for haplotype A, HincII 16517 only in individuals belonging to hap- for haplotype C, and AluI for haplotype D. logroup B, since this is a specific marker to The resulting restriction fragments and differentiate Amerindians from non-Amer- the PCR product including region V, defin- indian B individuals (Torroni et al., 1992, ing haplotype B, were analyzed by electro- 1993). The gain of the HaeIII site was phoresis on 3% NuSieve-Agarose (2:1) (FMC present in all individuals exhibiting haplo- BioProducts) gels. PCR amplification of the group B, including 8 Mapuche and 11 Pehu- D-loop region was carried out with two enches. specific primers, spanning a region of A3ϫ 4 contingency table analysis sug- 1,021 bp: L15997, CACCATTAGCACCCA- gests that in terms of haplogroup frequen- AAGCT; and H408, CTGTTAAAAGTGCAT- cies, these three populations cannot be sta- ACCGCCA. tistically distinguished (chi-square ϭ 6.78, exact P-value ϭ 0.32 with 10,000 permuta- Direct sequencing of PCR products tions). However, in terms of haplotype di- DNA sequencing was carried out by us- versity (Nei, 1978) defined by these four ing the PCR dsDNA Cycle Sequencing haplogroups, the Yaghan have the lowest System (Life Technologies), with the prim- diversity (52.4%), and the Pehuenche the 22 M.L. MORAGA ET AL.

TABLE 1. MTDNA haplogroup distribution in three Chilean aboriginal populations Haplogroups (%) Populations NA B C D Mapuche 111 0.0 7.2 44.1 48.7 Pehuenche (Trapa-Trapa) 105 2.8 10.5 41.0 45.7 Yaghan 21 0.0 0.0 47.6 52.4 Total 237 1.3 8.0 43.0 47.7 highest (61.7%), with the Mapuche being quence diversity. The Yaghan, likewise, intermediate (56.8%), the differences being showed the smallest sequence diversity of statistically insignificant (P ϭ 0.11). 91.4%, although this was not significantly smaller than the others. Sequence analysis Disregarding the phylogenetic relation- The two hypervariable regions of the D- ships of the observed sequences, at the se- loop were analyzed by direct sequencing of quence level, the Yaghan appear to be al- the DNA from 73 individuals from the three most 3–5 times as distant from the native populations, which were representa- Pehuenche and the Mapuche, (the standard tive of the four haplogroups described pre- distance of Nei (1978) being 0.31 between viously. As shown in Table 2, 37 different the Pehuenche and the Mapuche, 1.55 be- haplotypes were observed, defined by 52 tween the Pehuenche and the Yaghan, and polymorphic sites. It is important to note 1.09 between the Mapuche and the Yaghan). that two haplotypes shared by individuals The contingency table analysis of sequence- from different populations are shown sepa- lineage frequency differences suggests that rately. This gives a total of 40 lines in Table the genetic distances among these three 2, corresponding to only 37 haplotypes. In populations are statistically significant (chi- haplotype C, lineages PE3C, YA4C, and square ϭ 15.12, P Ͻ 10Ϫ5, with 10,000 per- HU4C are the same, as well as YA2C and mutations). HU1C. Haplogroup D was the most diverse at the Only nucleotide changes at np 73 and np sequence level, revealing 30 polymorphic 263 are shared by all haplogroups. One of sites, in contrast to haplogroups A with 13 these, transition A3G at position 73, has changes, B with 14 changes, and C with 16 also been described in all individuals from changes. We also found that, from all indian Argentinean Mapuche group (Ginther et viduals classified in the four haplogroups al., 1993), as well as in half of the individu- A–D (Torroni et al., 1992), only haplogroups als from a Hue´tar sample of Costa Rica A–C corresponded perfectly to 3 of the 4 (Santos et al., 1994). Two types of adenine clusters described by Horai et al. (1993). deletions, an A at np 248 and an AA at The T nucleotide at position 16290 and A 286–291, were found at hypervariable re- 16319 define haplogroup A/cluster III. Hap- gion II (HVS-II). These A deletions have logroup B/cluster I is defined by C 16189 been found in all haplogroup C individuals and C 16217. Haplogroup C/cluster IV is from the three Chilean aboriginal popula- defined by a C 16298 and T 16327. tions, as well as the Argentinean Mapuche Regarding haplogroup D/cluster II, C nu- sample (Ginther et al., 1993). Since these A cleotides at position 16325 and 16362 have deletions have not been described for other been described by Horai et al. (1993) as populations (Handt et al., 1998), they may characteristics of cluster II, since they were be characteristic nucleotide changes at least shared by 72 native Americans, including for South American aboriginal populations. 45 Chilean aborigines. From our study, one In these three Chilean groups, as with Mapuche individual lacked the 16362 T3C haplogroup diversity, we found the highest transition, and the other 13 individuals be- sequence diversity in the Pehuenche sam- longing to the three Chilean populations did ple, showing 38 out of 52 total polymorphic not show the characteristic 16325 T3C sites at the D-loop, amounting to 93.2% se- transition. Hence, our results do not corrob- TABLE 2. Polymorphic mtDNA D-loop sites in a sample of Chilean aborigines1 Anderson 16086 16092 16111 16129 16153 16187 16189 16207 16209 16213 16217 16218 16223 16234 16241 16249 16270 16286 16290 16291 16298 16301 16304 16311 16319 16325 16327 16342 16343 16362 53 55 56 64 73 77 94 129 143 146 150 152 153 195 207 215 234 248 258 263 286–291 Individuals Lineages

TTCGGCTATGTCCCATCCCCTCTTGTCTATGTACAAGTGTCTATGAA A CA A A

⅐⅐TA⅐⅐⅐⅐⅐⅐⅐⅐T ⅐⅐⅐⅐⅐T ⅐⅐⅐⅐⅐A ⅐⅐⅐⅐C ⅐⅐⅐TG⅐⅐⅐⅐C ⅐⅐GC⅐⅐G ⅐⅐G ⅐⅐3 PE1A ⅐⅐⅐⅐⅐⅐C ⅐⅐⅐C ⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐G ⅐⅐⅐⅐⅐⅐⅐⅐⅐A ⅐⅐⅐ ⅐G ⅐⅐2 PE6B ⅐⅐⅐⅐⅐⅐C ⅐⅐⅐C ⅐⅐⅐⅐⅐⅐⅐⅐T ⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐G ⅐⅐⅐⅐⅐⅐⅐⅐⅐A ⅐⅐⅐ ⅐G ⅐⅐1 PE5B ⅐⅐⅐⅐⅐⅐C ⅐⅐⅐CT⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐G ⅐⅐⅐⅐⅐⅐C ⅐⅐A ⅐⅐⅐ ⅐G ⅐⅐1 PE4B ⅐⅐⅐⅐⅐⅐CG⅐⅐C ⅐⅐⅐⅐⅐⅐⅐⅐T ⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐G ⅐⅐⅐⅐⅐⅐⅐⅐C ⅐⅐⅐⅐ ⅐G ⅐⅐1 PE1B ⅐⅐⅐⅐⅐⅐CG⅐⅐C ⅐⅐⅐⅐⅐⅐⅐⅐T ⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐G ⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐ ⅐G ⅐⅐4 HU1B ⅐⅐⅐⅐⅐⅐CG⅐⅐C ⅐⅐⅐⅐C ⅐⅐⅐T ⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐G ⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐ ⅐G ⅐⅐3 HU3B ⅐⅐⅐⅐⅐⅐CG⅐⅐C ⅐⅐⅐⅐⅐⅐⅐⅐T ⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐G ⅐⅐G ⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐ ⅐G ⅐⅐1 HU2B ⅐⅐⅐⅐⅐⅐C ⅐⅐AC⅐⅐⅐⅐C ⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐G ⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐ ⅐G ⅐⅐1 PE2B ⅐⅐⅐⅐⅐⅐C ⅐⅐AC⅐⅐T ⅐ C ⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐G ⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐ ⅐G ⅐⅐1 PE3B ⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐T ⅐⅐⅐⅐⅐⅐⅐C ⅐⅐C ⅐ CT⅐⅐⅐⅐⅐⅐⅐G ⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐—TG——1 PE2C ⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐T ⅐⅐⅐⅐⅐⅐⅐C ⅐⅐C ⅐ CT⅐ G ⅐⅐⅐⅐⅐G ⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐—TG——1 YA3C ⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐T ⅐⅐⅐⅐⅐⅐⅐C ⅐⅐⅐⅐CT⅐⅐⅐⅐⅐⅐⅐G ⅐⅐⅐⅐⅐⅐⅐⅐C ⅐⅐⅐—TG——1 PE1C ⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐T ⅐⅐⅐⅐⅐⅐⅐C ⅐⅐⅐⅐CT⅐⅐⅐⅐⅐⅐⅐G ⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐—TG——6 PE3C ⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐T ⅐⅐⅐⅐⅐⅐⅐C ⅐⅐⅐⅐CT⅐⅐⅐⅐⅐⅐⅐G ⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐—TG——1 YA4C ⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐T ⅐⅐⅐⅐⅐⅐⅐C ⅐⅐⅐⅐CT⅐⅐⅐⅐⅐⅐⅐G ⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐—TG——7 HU4C ⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐T ⅐⅐⅐⅐⅐⅐⅐C ⅐⅐⅐⅐CT⅐⅐⅐⅐⅐⅐⅐G ⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐— ⅐ G——2 YA2C ⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐T ⅐⅐⅐⅐⅐⅐⅐C ⅐⅐⅐⅐CT⅐⅐⅐⅐⅐⅐⅐G ⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐— ⅐ G——3 HU1C ⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐T ⅐⅐⅐⅐⅐⅐⅐C ⅐⅐⅐⅐CT⅐⅐⅐⅐⅐⅐⅐G ⅐⅐⅐⅐⅐T ⅐⅐⅐⅐⅐⅐— ⅐ G——1 HU2C ⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐T ⅐⅐⅐⅐⅐⅐⅐C ⅐⅐⅐⅐CT⅐⅐⅐⅐⅐⅐⅐G ⅐ T ⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐—TG——1 HU3C ⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐T ⅐⅐⅐⅐⅐⅐TC⅐⅐⅐⅐CT⅐⅐⅐⅐⅐⅐⅐G ⅐⅐⅐⅐⅐⅐⅐⅐C ⅐⅐⅐— ⅐ G——2 YA1C ⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐T ⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐C ⅐⅐⅐C ⅐⅐⅐⅐G ⅐⅐⅐⅐⅐⅐C ⅐ C ⅐⅐⅐⅐ ⅐G ⅐⅐2 YA4D ⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐TTG⅐⅐⅐⅐⅐⅐⅐⅐C ⅐⅐⅐C ⅐ C ⅐⅐⅐⅐G ⅐⅐⅐⅐⅐⅐C ⅐⅐⅐⅐⅐⅐ ⅐G ⅐⅐2 YA3D ⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐T ⅐ G ⅐⅐⅐⅐⅐⅐T ⅐⅐⅐⅐⅐C ⅐ C ⅐⅐⅐⅐G ⅐⅐⅐⅐⅐⅐C ⅐⅐⅐⅐⅐⅐ ⅐G ⅐⅐2 HU10D ⅐⅐⅐⅐⅐T ⅐⅐⅐⅐⅐⅐T ⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐C ⅐⅐⅐⅐⅐⅐⅐⅐G ⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐ ⅐G ⅐⅐1 HU9D ⅐⅐⅐⅐⅐T ⅐⅐⅐⅐⅐⅐T ⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐C ⅐⅐⅐C ⅐⅐⅐⅐G ⅐⅐⅐⅐C ⅐ C ⅐⅐⅐G ⅐⅐ ⅐G ⅐⅐1 HU1D ⅐⅐⅐⅐⅐TC⅐ C ⅐⅐⅐T ⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐C ⅐⅐⅐C ⅐ CGTG⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐ ⅐G ⅐⅐1 PE3D ⅐⅐⅐⅐⅐TC⅐ C ⅐⅐⅐T ⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐C ⅐⅐⅐C ⅐ CGTG⅐⅐⅐⅐⅐⅐⅐⅐C ⅐⅐⅐⅐ ⅐G ⅐⅐1 PE4D ⅐⅐⅐⅐⅐T ⅐⅐⅐⅐⅐⅐T ⅐⅐⅐⅐⅐⅐⅐⅐⅐C ⅐⅐C ⅐⅐⅐C ⅐⅐⅐⅐G ⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐ ⅐G ⅐⅐2 PE5D ⅐ C ⅐⅐⅐TC⅐⅐⅐⅐⅐T ⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐C ⅐⅐⅐⅐G ⅐⅐⅐A ⅐⅐⅐⅐⅐⅐⅐⅐⅐ ⅐G ⅐⅐1 PE2D ⅐⅐⅐⅐⅐TC⅐⅐⅐⅐⅐T ⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐C ⅐⅐⅐⅐G ⅐⅐⅐A ⅐⅐⅐⅐⅐⅐⅐⅐⅐ ⅐G ⅐⅐1 PE1D ⅐⅐⅐⅐⅐TC⅐⅐⅐⅐⅐T ⅐⅐⅐T ⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐C ⅐⅐⅐⅐GT⅐⅐A ⅐⅐⅐⅐⅐⅐⅐⅐⅐ ⅐G ⅐⅐1 HU7D ⅐ C ⅐⅐⅐TC⅐⅐⅐⅐⅐T ⅐⅐⅐T ⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐CT⅐⅐⅐G ⅐⅐⅐A ⅐⅐⅐⅐⅐⅐⅐⅐⅐ ⅐G ⅐⅐1 HU4D ⅐⅐⅐⅐⅐T ⅐⅐⅐⅐⅐⅐T ⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐C ⅐⅐⅐CT⅐⅐⅐G ⅐⅐⅐A ⅐⅐⅐⅐⅐⅐⅐⅐⅐ ⅐G ⅐⅐1 HU2D ⅐ C ⅐⅐⅐TC⅐⅐⅐⅐⅐T ⅐⅐⅐T ⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐C ⅐⅐⅐⅐G ⅐⅐⅐A ⅐⅐⅐⅐⅐⅐⅐⅐⅐ ⅐G ⅐⅐3 HU5D ⅐⅐⅐⅐⅐TC⅐⅐⅐⅐⅐T ⅐⅐⅐T ⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐C ⅐⅐⅐⅐G ⅐⅐⅐A ⅐⅐⅐⅐⅐⅐⅐⅐⅐ ⅐G ⅐⅐1 HU6D ⅐⅐⅐⅐A ⅐⅐⅐⅐⅐⅐⅐T ⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐C ⅐⅐⅐C ⅐⅐⅐⅐G ⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐ ⅐G ⅐⅐2 HU8D ⅐⅐⅐⅐⅐TC⅐⅐⅐⅐⅐T ⅐⅐⅐T ⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐C ⅐⅐⅐⅐G ⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐ ⅐G ⅐⅐1 HU3D C ⅐⅐⅐⅐TC⅐⅐⅐⅐⅐T ⅐⅐⅐⅐T ⅐⅐⅐⅐⅐⅐⅐C ⅐⅐⅐C ⅐⅐⅐⅐G ⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐ ⅐G ⅐⅐4 YA2D C ⅐⅐⅐⅐TC⅐⅐⅐C ⅐ T ⅐⅐⅐⅐T ⅐⅐⅐⅐⅐⅐⅐C ⅐⅐⅐C ⅐⅐⅐⅐G ⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐⅐ ⅐G ⅐⅐1 YA1D

1 The number of individuals for each lineage is shown. In the right column each lineage is named by its corresponding population (PE, Pehuenche; HU, Mapuche (Huapi Island); YA, Yaghan) the sample number, and the RFLP haplogroup. Individuals highlighted in bold share the same sequence and belong to different groups (PE3C, YA4C, HU4C, and YA2C, HU1C). The sequence published by Anderson et al. (1981) was used as a reference. 24 M.L. MORAGA ET AL. orate the classification of haplogroup analysis. D-loop sequences from the 3 Pehu- D/cluster II on the basis of these mutations, enche individuals leading to haplogroup A but are in agreement with the results of correspond to one lineage, clustering obvi- Ginther et al. (1993), who also observed the ously in one branch of 100% consensus. absence of the 16325 T3C mutation in hap- DISCUSSION logroup D mtDNAs from the Argentinean Mapuche group. In the present paper we report on the It is interesting to note that one of these results of a mtDNA RFLP and sequence nucleotide changes, C 16325, is present in polymorphism analysis of three Chilean ab- all individuals of haplogroup C/cluster IV original populations. The RFLP analysis from our study, in the Argentinean Mapu- showed that all individuals belong to one of che, and in Amazonian individuals (Santos the previously defined Amerindian haplo- et al., 1996). Also, in our study, C in position groups, A, B, C, or D. In terms of haplogroup 16362 is present in individuals from cluster frequencies, the three Chilean aboriginal III/haplogroup A. Therefore it seems that groups studied here cannot be genetically there are no characteristic and exclusive nu- distinguished. However, as shown in Table cleotide changes in either hypervariable re- 3, combining the present data with other gion defining haplogroup D, at least for published haplogroup frequencies of South these aboriginal populations from South American aborigines, it is noteworthy that America. However, we found a C3T change the frequencies of haplogroups A and B de- at nucleotide 16187 which is present in the crease from north to south in meridian majority of South Amerindians of haplo- South America, whereas haplogroups C and group D, which is absent from haplogroups D tend to increase. This trend is most likely A–C. This change is also present in 6/10 the result of founder effect, which occurred Argentinean Mapuche (Ginther et al., during the initial peopling of the southern 1993), and in 14/18 native Americans (Horai cone of the continent. As Paleoindians et al., 1993). From this analysis, we suggest moved south, new territory was probably that this change is the most characteristic colonized by small bands that were carriers one for South Amerindian populations. of a subsample of genes from the ancestral populations they left behind. Haplogroups A Phylogenetic analysis and B probably got lost as a consequence of The D-loop DNA sequences, spanning hy- this process. An alternative hypothesis, e.g., pervariable regions I and II from the 37 gene flow from migrant populations carry- Chilean lineages, were aligned and submit- ing A and B into more ancient groups pos- ted to phylogenetic analysis through neigh- sessing C and D, would imply that more bor-joining (NJ) (Fig. 2) and parsimony than one migration event occurred during methods (data not shown). In the NJ tree, the peopling of the southern cone of South we can see that all lineages fall into four America. This hypothesis is not supported differentiated branches, including haplo- by recently published gene geography maps groups A–D. Nevertheless, it is noteworthy of South America (Rothhammer and Silva, that three D lineages, HU8D, HU9D, and 1992; Rothhammer et al., 1997). YA4D, cluster close to the C branch, while Turning now to the sequence data, we note HU10D and YA3D cluster close to branch A. that the three Chilean aboriginal groups, can All the other D lineages fall into one clade. be genetically distinguished at the D-loop se- The majority rule consensus (data not quence level, making the Yaghan most dis- shown) shows that two branches B and C tant (as well as least diverse) from the Pehu- cluster independently in 100% of maximum enche and the Mapuche. parsimony trees. Seven out of 19 lineages Further, the nucleotide changes described corresponding to haplogroup D cluster in by Forster et al. (1996), including one base one branch, with 74 % consensus from a substitution at nucleotide 16111 leading to total of 500 maximum parsimony trees. The subhaplogroups A1/A2, and another substi- remaining lineages, also corresponding to tution at 16271 defining subhaplogroups haplogroup D, appeared unresolved in this D1/D2, have been tested in the present MTDNA POLYMORPHISMS IN ABORIGINAL CHILEANS 25

Fig. 2. Tree obtained by the neighbor-joining method for 37 Chilean haplogroups. Individuals highlighted in bold share the same sequence and belong to different groups (PE3C, YA4C, HU4C, and YA2C, HU1C). An African sequence was used as an outgroup in order to root the tree. The alignment was performed over 549 bp, including hypervariable regions I and II. Letters on the side indicate respective RFLP haplogroups. study (Table 2). All individuals from haplo- Forster et al. (1996). Indeed, all haplotype D group A did have the C3T transition at individuals lacked the T3C transition and, 16111, and therefore are subhaplotype A2 of therefore, are haplotype D1. 26 M.L. MORAGA ET AL.

TABLE 3. Frequency distribution of founding haplogroups in aboriginal populations of Chile and Argentina1 Haplogroups Populations N A B C D Others References Aymara 172 0.06 0.68 0.12 0.14 Merriwether et al., 1995 Atacamen˜ o 50 0.12 0.72 0.10 0.06 Merriwether et al., 1995 Whichi 72 0.06 0.63 0.03 0.26 0.03 Bravi et al., 1995 Chorote 20 0.15 0.40 0.30 0.15 Bravi et al., 1995 Mapuche 1 58 0.05 0.31 0.21 0.30 0.14 Ginther et al., 1993 Mapuche 2 111 0.07 0.44 0.49 This study Pehuenche 1 100 0.02 0.09 0.37 0.52 Merriwether et al., 1995 Pehuenche 2 105 0.03 0.10 0.41 0.46 This study Huilliche 80 0.04 0.28 0.19 0.49 Merriwether et al., 1995 Tehuelche 29 0.21 0.24 0.55 Bravi (pers. Comm.) Fueguian 45 0.42 0.56 0.02 Lalueza et al., 1997 Yaghan 21 0.48 0.52 This study 1 Populations are listed from top to bottom in order of distribution along Chile and Argentina, from northernmost to southernmost.

Comparison of the two hypervariable D- tinean Mapuche (Fig. 3). These findings are loop regions reveals a high variability of noteworthy in relation to unconventional haplogroup D, in comparison to the other speculations concerning the origin of the ab- three haplogroups, and especially with hap- origines of Tierra del Fuego. More than half logroup C, including a similar number of a century ago, it was proposed that people of individuals, as shown in Table 2. This vari- Tierra del Fuego had direct ancestral links ability is expressed by 19 different lineages to Australoid populations (Frenguelli, 1963; defined by 30 polymorphic sites. We were Imbelloni, 1938). This hypothesis was based unable to find any nucleotide change shared on the unusual morphological characteris- by 100% of D individuals, as described pre- tics, particularly the cranial robustness viously by Horai et al. (1993) in a similar (supposedly a sign of antiquity), of the ab- sample. Nevertheless, our results indicate a origines of Patagonia and Tierra del Fuego, C3T 16187 change which is present in the that is to say the Ona (Selk’nam), Yaghan majority of D individuals, and is absent in (Yamana), and Alacaluf (Kaweskar). haplogroups A–C (Table 2), concordant with In the interim, a consensus among an- lineages found by Ginther et al. (1993) in an thropologists was reached in the sense that Argentinean Mapuche sample, as well as Amerinds were derived from Mongoloids Chilean lineages described by Horai et al. who, moving from Asia, crossed the Bering (1993). However, this change seems to be Land Bridge as suggested by Hrdlicka exclusive to Chilean and Argentinean (Frenguelli, 1963). There are, however, still southern aborigines, since it is not present anthropologists with different views. Re- in other populations from South America. cently, on the basis of the statistical manip- As shown in Figure 3, it is noteworthy ulation of craniometric means, Neves and that 26/26 individuals from this study be- Pucciarelli (1991) suggested the existence of longing to haplogroup C, two lineages from a pre-Mongoloid migration to South Amer- Chilean aborigines described by Horai et al. ica, coming close to reviving the old ideas of (1993), and 7/8 Argentinean aborigine indi- Imbelloni (1938) in a more current context. viduals (Ginther et al. 1993), are grouped in In contrast, our results indicate that the one cluster. This extreme low diversity Yaghan do not exhibit divergent D-Loop se- within haplogroup C for southern Chilean quences, clustering together with the corre- and Argentinean groups could also be ex- sponding haplogroup sequences from other plained by a founder effect. South Amerindian populations (Fig. 3), and Comparing our findings with other Amer- presenting similar haplogroup frequencies indian mtDNA sequences from South Amer- compared to the Mapuche and Pehuenche. ica, we found that the retrieved sequences of In other words, our data do not support the the three Chilean groups cluster into the hypothesis that the Yaghan descend from a four basic mtDNA haplogroups, together different Paleoindian migration lacking with the Amazonian tribes and the Argen- haplogroups A and B, as suggested by Lalu- MTDNA POLYMORPHISMS IN ABORIGINAL CHILEANS 27

Fig. 3. Twenty-seven distinct Chilean haplogroups, puche (Huapi Island); YA, Yaghan; HOCH, Chilean ab- as well as 105 previously published Amerindian lin- origine; HOCO, Colombian aborigine; HOBR, Brazilian eages from 9 different South American populations, aborigine (Horai et al., 1993); WBR, Brazilian aborigine were analyzed using the neighbor-joining method with (Handt et al., 1998); GMA, Argentinean Mapuche a 254-bp fragment, including only the hypervariable (Ginther et al., 1993); SBR, Brazilian aborigine (Santos region I. An African sequence was used as an outgroup et al., 1996); TARG, Argentinean aborigine; TBR, Bra- in order to root the tree. Letters on the side indicate zilian aborigine (Torroni et al., 1993); EYA, Yanomama respective RFLP haplogroups. PE, Pehuenche; HU, Ma- (Easton et al., 1996). 28 M.L. MORAGA ET AL. eza et al. (1997). Our results are rather in Gustincich S, Manfioletti G, Del Sal G, Schneider C, Carninci P. 1991. A fast method for high quality agreement with the view that the Yaghan genomic DNA extraction from whole blood. Biotech- are related to tribes from south-central niques 11:298–302. Chile and Argentina such as the Huilliche, Handt O, Meyer S, von Haeseler A. 1998. Compilation of human mtDNA control region sequences. Nucleic Pehuenche, and Tehuelche, as previously Acids Res 26:126–129. suggested by Rothhammer et al. (1986) and Horai S, Kondo R, Nakagawa-Hattori Y, Hayashi S, Sonoda S, Tajima K. 1993. Peopling of the Americas, Llop (1996). Furthermore, we note that the founded by four major lineages of mitochondrial DNA. haplogroup distribution of the Yaghan is Mol Biol Evol 10:23–47. similar to the distribution obtained by Lalu- Imbelloni G. 1938. Tabla clasificatoria de los Indios: regiones biologicas y grupos raciales en America. Phy- eza Fox (1996) for extinct groups from sis 12:229–249. Tierra del Fuego-Patagonia. All samples Lalueza C, Perez-Perez A, Prats E, Cornudella L, Tur- bon D. 1997. Lack of founding Amerindian mitochon- from these geographic areas lack haplo- drial DNA lineages in extinct aborigines from Tierra groups A and B and are therefore most prob- del Fuego-Patagonia. Hum Mol Genet 6:41–46. ably descendants of one common ancestral Lalueza Fox C. 1996. Mitochondrial DNA haplogroups in four tribes from Tierra del Fuego-Patagonia: infer- Paleoindian population. In this context, it is ences about the peopling of the Americas. Hum Biol interesting to note that Bennett and Bird 68:853–871. Llop E. 1996. Genetic composition of Chilean aboriginal (1964) reported a complete archeological se- populations: HLA and other genetic marker varia- quence linking the first Paleoindian mi- tion. Am J Phys Anthropol 101:325–332. grants to contemporary Ona (Shelknam) Llop E, Harb Z, Acun˜ a M, Moreno R, Barton S, As- pillaga E, Rothhammer F. 1993. Composicioń ge- populations. ne´tica de la poblacioń chilena: los Pehuenches de Trapa-Trapa. Rev Med Chil 121:494–498. Llop E, Harb Z, Acun˜ a M, Moreno R, Aspillaga E, Van LITERATURE CITED de Maele M, Rothhammer F. 1994. Composicioń gene´tica de la poblacioń chilena: los Yamanas de Ukika. Anderson S, Bankier AT, Barrell BG, de Bruijn MHL, Rev Med Chil 122:979–985. Coulson AR, Drouin J, Eperon IC, Nierlich DP, Roe Merriwether DA, Rothhammer F, Ferrell RE. 1995. Dis- BA, Sanger F, Schreier PH, Smith AJH, Staden R, tribution of the four founding lineage haplotypes in Young IG. 1981. Sequence and organization of the Native Americans suggests a single wave of migra- human mitochondrial genome. Nature 290:457–465. tion for the New World. Am J Phys Anthropol 98:411– Bailliet G, Rothhammer F, Carnese F, Bravi C, Bianchi 430. NO. 1994. Founder mitochondrial haplogroups in Nei M. 1978. Estimation of average heterozygosity and Amerindian populations. Am J Hum Genet 55:27–33. genetic distance from a small number of individuals. Ballinger SW, Schurr TG, Torroni A, Gan YY, Hodge Genetics 89:583–590. JA, Hassan K, Chen KH, Wallace DC. 1992. Southern Neves W, Pucciarelli H. 1991. Morphological affinities Asian mitochondrial DNA analysis reveals genetic of the first Amerindian: an exploratory analysis based continuity of ancient Mongoloid migrations. Genetics on early South American human remains. J Hum 130:139–152. Evol 21:261–273. Bernnett WC, Bird JB. 1964. Andean culture history. Rothhammer F, Silva C, Cocilovo J, Quevedo S. 1986. New York: Natural History Press. Doubleday & Co., Una hipo´tesis sobre el poblamiento de Chile basada Inc. en el ana´lisis multivariado de medidas craneome´t- Chen Y-S, Torroni A, Excoffier L, Santachiara-Benere- ricas. Rev Chung 16–17:115–118. cetti AS, Wallace DC. 1995. Analysis of mtDNA vari- Rothhammer F, Silva C. 1992. Gene geography of South ation in African populations reveals the most ancient America: testing models of population displacement of all human continent-specific haplogroups. Am J based on archeological evidence. Am J Phys An- Hum Genet 57:133–149. thropol 89:441–446. Easton R, Merriwether, DA, Crews D, Ferrell R. 1996. Rothammer F, Silva C, Callegari-Jacques SM, Llop E, Salzano FM. 1997. Gradients of HLA diversity in mtDNA variation in the Yanomami: evidence for ad- South American Indians. Ann Hum Biol 24:197–208. ditional New World founding lineages. Am J Hum Saitou N, Nei M. 1987. The neighbor-joining method: a Genet 59:213–225. new method for reconstructing phylogenetic trees. Forster P, Harding R, Torroni A, Bandelt H-J. 1996. Mol Biol Evol 4:406–425. Origin and evolution of Native American mtDNA Santos M, Ward RH, Barrantes R. 1994. mtDNA vari- variation: a reappraisal. Am J Hum Genet 59:935– ation in the Chibcha Amerindian Huetar from Costa 945. Rica. Hum Biol 66:963–977. Frenguelli J. 1963. The present status of the theories Santos SE, Ribeiro-Dos-Santos AK, Meyer D, Zago MA. concerning primitive man in Argentina. In: Steward 1996. Multiple founder haplotypes of mitochondrial JH, editor. Handbook of South American Indians, Vol DNA in Amerindians revealed by RFLP and sequenc- 6. New York: Cooper Square Publisher, Inc. p 11–17. ing. Ann Hum Genet 60:305–319. Ginther C, Corach D, Penacino GA, Rey JA, Carnese Schurr TG, Ballinger SW, Gan YY, Hodge JA, Merri- FR, Hutz MH, Anderson A, Just J, Salzano FM, King wether DA, Lawrence DN, Knowler WC. 1990. Amer- MC. 1993. Genetic variation among the Mapuche In- indian mitochondrial DNAs have rare Asian muta- dians from the Patagonian region of Argentina: mito- tions at high frequencies suggesting they derived chondrial DNA sequence variation and allele frequen- from four primary maternal lineages. Am J Hum cies of several nuclear genes. EXS 67:211–219. Genet 46:613–623. MTDNA POLYMORPHISMS IN ABORIGINAL CHILEANS 29

Swofford DL. 1993. Phylogenetic analysis using parsi- Lawrence DN, Weiss KM, Wallace DC. 1992. Native mony (PAUP), version 3.1. Champaign: Illinois Nat- American mitochondrial DNA analysis indicated that ural History Survey. the Amerind and the Nadene populations were Swofford DL, Olsen GJ. 1990. Phylogeny reconstruc- founded by two independent migrations. Genetics tion. In: Hillis DM, Moritz C, editors. Molecular sys- 130:153–162. tematics. Sunderland, MA: Sinauer. p 411–500. Torroni A, Schurr TG, Cabell MF, Brown MD, Neel JV, Thompson JD, Higgins DG, Gibson TJ. 1994. Larsen M, Smith DG, Vullo CM, Wallace DC. 1993. CLUSTAL W: improving the sensitivity of progres- Asian affinities and continental radiation of the four sive multiple sequence alignment through sequence founding Native American mtDNAs. Am J Hum weighting, positions-specific gap penalties and Genet 53:563–590. weight matrix choice. Nucleic Acids Res 22:4673– Torroni A, Huoponen K, Francalacci P, Petrozzi M, Mo- 4680. relli L, Scozzari R, Obinu D. 1996. Classification of Torroni A, Schurr TG, Yang CC, Szathmary EJE, Wil- European mtDNAs from an analysis of three Euro- liams RC, Schanfield MS, Troup GA, Knowler WC, pean populations. Genetics 144:1835–1850.