Polymorphic Alu insertions and the Asian origin of Native American population

GABRIEL E. NOVICK,1 CORINA C. NOVICK, 1JUAN YUNIS,2 EMILIO YUNIS,2 PAMELA ANTUNEZ DE MAYOLO, 1W. DOUGLAS SCHEER3 PRESCOTT L. DEININGER,4 MARK STONEKING, DANIEL 5. YORK,6 MARK A. BATZER,3 AND RENE J. HERRERA1

Abstract A rapid PCR-based assay was used to study the distribution of 5 polymorphic Alu insertions in 895 unrelated individuals from 30 populations, 24 from North, Central, and South America. Although a significant level of interpopulation variability was detected, the variability was less than that observed in a worldwide population survey. This is consistent with the bottleneck effect and genetic drift forces that may have acted on the migrating founder groups. The results corroborate the Asian origin of native American populations but do not support the multiple-wave migration hypothesis supposedly responsible for the tripartite Eskaleut, Nadene, and Amerind linguistic groups. Instead, these populations exhibit three major identifiable clusters reflecting geographic distribution. Close similarity between the Chinese and native Americans suggests recent gene flow from Asia.

A wealth of anthropological, dental, linguistic, and genetic data have been accumulated in the area of human evolution in the Americas. However, the issues of origin, point of entry, number of migrations, timing, routes of expansion, and survivorship of native Americans remain controversial. Traditional anthropological analyses support the hypothesis that native Americans are derived from northern Asian ancestors who reached the New World by walking over the Bering land bridge that was exposed during the last glaciation, some 20,000 years ago (Fladmark 1983).

Several hypotheses have been formulated to explain the origins of humans in the New World. The tripartite hypothesis proposes three waves of migration that originated in northern Asia and gave rise to the three main genetic and linguistic clusters: Eskaleut, Nadene, and Amerind (Greenberg 1987). Under this hypothesis there were three distinct migrations. The first one, sometime before 15,000 years ago, originated the Amerind cluster, which spread over most of the New World. The second migration, 15,000-10,000 years ago, gave rise to the Nadene group. The third wave, 10,000 years ago, founded the Eskaleut cluster. This hypothesis has been corroborated by dental data (Greenberg et al. 1986) and by nuclear and mitochondrial DNA genetic data (Wallace and Torroni 1992; Cavalli-Sforza et al. 1994).

Multiple migrations also have been proposed (Cavalli-Sforza et al. 1994), as have single-migration models (Rogers et al. 1991). One possibility, according to the single-migration model, portrays a mostly ice-covered North America during the last glacial maximum with some ice areas suitable for habitation (Rogers et al. 1991). The isolation that may have occurred in this ice-free refugium may have encouraged differentiation, genetic drift, and independent differentiation (Rogers et al. 1991). This differentiation may have yielded the profile of multiple migrations postulated by other researchers.

Several polymorphic genetic systems have been used to study native American phylogeny. These include mitochondrial DNA, variable number of tandem repeats (VNTRs), restriction fragment length polymorphisms (RFLPs), and point mutations in Alu sequences (O'Rourke et al. 1992; Schanfield 1992; Kidd, Pakstis et al. 1993; Torroni et al. 1994; Kidd and Kidd 1996; Knight et al. 1996). Alu sequences are the largest family of short interspersed repetitive elements (SINEs) in humans, with an excess of 500,000 copies per haploid genome [for reviews see Deininger and Batzer (1993) and Novick et al. (1996)]. Alu elements are ancestrally derived from the endoplasmic reticulum signal recognition particle 7SL RNA gene (Ullu and Tschudi 1984), with which they share about 90% sequence similarity throughout most of their sequences (Ullu and Tschudi 1984). Alu elements are distributed throughout the genomes of primates. Recently, one Alu subfamily was found to be largely human specific (HS) (Batzer 1990; Batzer and Deininger 1991). Members of this subfamily have been inserted recently into the human lineage genome, within the last 200,000 to 6 million years (Batzer 1991; Batzer and Deininger 1991). A limited number of Alu elements are transcriptionally active (Matera et al. 1990) and undergo amplification into other genomic locations (Wallace et al. 1991) in a process termed retroposition.

In addition to the polymorphic nature of many Alu insertions, several features make Alu elements exceptional genetic markers: the stability of the Alu insertion event, the lack of a known mechanism for the precise removal of Alu elements from their specific chromosomal site of insertion, and the low rate of de novo insertions that reach polymorphic levels (Batzer and Deininger 1991). These characteristics make it highly unlikely that the same Alu insertion could occur more than once independently at the same locus or that once inserted, an element could be removed without leaving vestiges of its existence behind. Furthermore, the ancestral state of an Alu insertion invariably is the absence (complete and exact) of the element at a particular locus and the presence of an insertion at that site, the forward mutational change. This is an invaluable attribute not present in other polymorphic systems (e.g., RFLPs) where the ancestral state is ambiguous.