Y-chromosome-specific microsatellite variation in Australian aboriginals

Human Biology, Dec 1999 by Vandenberg, N, Van Oorschot, R A H, Tyler-Smith, C, Mitchell, R J

KEY WORDS: Y CHROMOSOME, MICROSATELLITE, HAPLOTYPE, AUSTRALIAN AB

ORIGINAL, GENE DIVERSITY, NETWORK ANALYSIS, DYS19, DYS390, DYS391, DYS392

Abstract The frequency distributions of 4 highly polymorphic Y-chromosome-specific microsatellites (DYS19, DYS390, DYS391, and DYS392) were determined in 79 unrelated Australian Aboriginal males from the Northern Territory. These results are compared with those observed in worldwide populations at both the locus and the haplotype level. Common alleles in Aboriginals are DYS19*15 (49%), DYS19*14 (28%), DYS390*19 (39%), DYS390*24 (20%), DYS391*10 (72%), DYS392*11 (63%), and DYS392*13 (28%). No evidence of reduced gene diversity was observed for these Y-chromosome alleles. DYS390 exhibits the most complex arrangement, displaying a bimodal distribution composed of common alleles (*22-*26), and rare short alleles (*18-*20), with an intermediate allele (*21) being absent. DYS390*20, previously reported only in Papuans and Samoans, is observed for the first time in Aboriginals. Compared with a recent study of Aboriginals, our sample exhibits considerable diversity in the haplotypes associated with the rare DYS390*19 allele, indicating that this allele is of considerable antiquity, if it arose as a single deletion event. Combining all 4 Y-chromosomelinked microsatellites produced 41 unique haplotypes, which were linked using a median-joining network. This network shows that most (78%) of our Aboriginal haplotypes fall into 2 distinct clusters, which likely represent 2 separate lineages. Seven haplotypes are shared with haplotypes found in a recent study of Aboriginals, and 7 are shared with a Spanish population. The cluster of Aboriginal haplotypes associated with the short DYS390 alleles does not share any haplotypes with the Spanish, indicating that this cluster of haplotypes is unique to Australian Aboriginals. Limited data from 4 worldwide populations used to construct haplotypes based on 3 loci (DYS19, DYS390, DYS392) show that only 4 of these haplotypes are seen in Australian Aboriginals. Shared haplotypes may be the result of admixture and/or recurrent mutation at these loci. Expanding the haplotype analysis to include biallelic markers on the Y chromosome will resolve this issue.

The nonrecombining portion of the human Y chromosome is inherited as a single unit, or haplotype, following a paternal line of transmission. Its genetic sequence therefore remains unchanged, except when altered by mutation, which then cannot be lost by recombination. These characteristics make Y-chromosome polymorphisms suitable for studying human phylogeny at the molecular level in addition to their application in forensic casework and paternity disputes (Trabetti et al. 1994; Kayser et al. 1997; de Knijff et al. 1997; Prinz et al. 1997).

Until relatively recently, there was a paucity of reported polymorphisms on the Y chromosome (Malaspina et al. 1990; Dorit et al. 1995). However, with the advent of new methods of detecting variation, an increasing number of sequence variants have been identified (Underhill et al. 1997). It has also been shown that the nonrecombining portion of the Y chromosome contains I minisatellite (Jobling et al. 1998) and several microsatellite [or short tandem repeat (STR)] polymorphisms (Kayser et al. 1997). These microsatellites are just as polymorphic as their autosomal equivalents (Roewer et al. 1992) and, as a result of their strict linkage disequilibrium, a group of microsatellites can generate Y-chromosome haplotypes directly. Haplotype diversity can be of immense value in investigating population affinities.

However, there has been considerable debate concerning the usefulness of microsatellites in detecting evolutionary relationships because these loci are subject to recurrent mutation (Ciminelli et al. 1995; Santos et al. 1996; Jorde et al. 1997). Recently, Heyer et al. (1997) used deep rooting pedigrees to estimate Y-chromosome-specific microsatellite mutation frequencies and obtained an average mutation frequency of 0.21% per locus per generation. Thus, assuming a stepwise mutation model (Kimura and Ohta 1978) and an average mutation frequency of 0.21%, de Knijff et al. (1997) estimated that Y-chromosome microsatellites may be reliable in elucidating population histories over approximately the last 50,000 years.

Although some large-scale surveys of human Y-chromosome polymorphisms, including microsatellites, have been undertaken (Jobling and TylerSmith 1995; de Knijff et al. 1997; Hammer et al. 1997), 1 of the least studied groups is the indigenous population of Australia. This region of the globe is of considerable importance to our understanding of human origins and their subsequent spread across the world. Controversy exists over the date of first arrival of the ancestors of the present indigenous inhabitants of Australia and also the source of these ancestors. Previous studies have looked at the genetic relationships between indigenous Australians and other Oceanic populations, especially those of Papua New Guinea and Southeast Asia, and results have suggested a shared ancestry between Papua New Guinea Highlanders and Australian Aboriginals (Gao and Sejentson 1991; Lin et al. 1994). This therefore indicates that the first human migration from the north into Australia may have taken place during the late Pleistocene (about 50,000 years B.P.) when sea levels were lower and Australia was still connected to Papua New Guinea (White and O'Connell 1982). However, there is debate over the number of waves of migration and the degree of diversity attributable to different founding populations or to microevolution within Australia (Tsintof et al. 1990; Roberts-Thomson et al. 1996; van Holst Pellekaan et al. 1998).