Genetic variation in microsatellites is rarely examined in the field of ancient DNA (aDNA) due to the low quantity of nuclear DNA in the fossil record together with the lack of characterized nuclear markers in extinct species. 454 sequencing platforms provide a new high-throughput technology capable of generating up to 1 gigabases per run as short (200–400-bp) read lengths. 454 data were generated from the fossil bone of an extinct New Zealand moa (Aves: Dinornithiformes). We identified numerous short tandem repeat (STR) motifs, and here present the successful isolation and characterization of one polymorphic microsatellite (Moa_MS2). Primers designed to flank this locus amplified all three moa species tested here. The presented method proved to be a fast and efficient way of identifying microsatellite markers in ancient DNA templates and, depending on biomolecule preservation, has the potential of enabling high-resolution population genetic studies of extinct taxa. As sequence read lengths of the 454 platforms and its competitors (e.g., the SOLEXA and SOLiD platforms) increase, this approach will become increasingly powerful in identifying microsatellites in extinct (and extant) organisms, and will afford new opportunities to study past biodiversity and extinction processes.

Introduction

With the introduction of new high-throughput DNA sequencing techniques capable of generating millions/billions of sequence reads per run, genomic research is advancing faster than ever (1,2,3,4). In the field of paleogenetics, the first complete nuclear genome has yet to be recovered, but major sequencing projects of woolly mammoth (5) and Neanderthal (6) are heading what has been called the “third wave” of progress in ancient DNA (aDNA) studies (7). These new sequencing platforms generate large quantities of protein coding data, which will undoubtedly assist in the study of molecular evolution, functional genomics, and adaptation. In addition to coding data, the huge number of randomly amplified sequences provides the opportunity to search for microsatellites or short tandem repeats (STRs). These non-coding sequences with a high rate of mutation are applied as markers in a wide array of genetic research, especially in relation to forensics, modern population biology, and parentage analyses.

A limited number of successful STR amplifications have been reported from ancient substrates using “modern” STRs as templates (8,9,10) but to our knowledge, no microsatellite primers developed directly from aDNA templates (let alone extinct species) have been published prior to this study. This is likely due to complications in traditional microsatellite library construction as the result of the degraded and cross-linked nature of ancient DNA (11). Greenwood et al. (10) managed to amplify a single microsatellite locus in woolly mammoth using primers developed for modern elephants. However, to rely solely on primers developed for related modern taxa, when targeting microsatellites in extinct ones, is problematic because of possible low cross-species amplification rates and chance of monomorphism in the target species (12,13). This is especially pertinent in taxa such as moa, where the likelihood of cross-species amplification is limited as a result of the >80 million years that separate moa from their closest living relatives among the ratite birds (14). Consequently, the chance of identifying polymorphic microsatellite markers in ancient DNA templates of an extinct taxon seems greatly enhanced when the potential markers have been identified directly on the target species.

The Roche GS-FLX (454 Life Sciences, Branford, CT, USA) sequencing technology is currently capable of producing 0.1 gigabases per run with a read length averaging 200–300 nucleotides—a sequence size that allows for the presence of an STR and sufficient flanking regions to design primers. A GS-FLX run was conducted on a Pachyornis elephantopus (heavy-footed moa) bone extract to identify a series of microsatellite loci. To illustrate the viability of this technique, we identified an (AC)₁₂ microsatellite (directly from raw GS-FLX data) and demonstrated cross-species amplification in three species of moa. All eleven New Zealand moa species were driven to extinction in the early 15th century, following the arrival of Polynesians. Identification of new STR markers, such as described here, will enable detailed DNA profiling of extinction processes and past population dynamics of these ancient ratites.

Materials and methods

Sampling of moa fossils and DNA isolation

Sampling of moa fossils was conducted by drilling out cylindrical elements (diameter of ∼1 cm) of moa tibiotarsal bones using a power drill and diamond-dust coated drill bits. Each sample was then ground into bone powder using a Dremel tool (Part no. 114; Racine, WI, USA). To minimize the incorporation of any possible DNA contamination present on the bones, the bone surfaces and the inner porous parts were excluded, and only solid cortical bone was processed. Contamination from external sources—as well as cross-contamination between samples—was minimized by thoroughly cleaning equipment and sampling environment (with 10% bleach and 100% alcohol) between the processing of each individual. To minimize the risk of a ubiquitous DNA contaminant being present on all the bones, fossils representing three different moa species—from two different sites, and from several different museum/university collections—were included (Supplementary Table 1). Pyramid Valley and Bell Hill Vineyard Swamp both represent late Holocene deposits in North Canterbury, New Zealand, with a known fossil record spanning app. 3700–700 bp (15, R.N.H. unpublished data).