DNA from keratinous tissue. Part I: Hair and nail
Introduction
DNA derived from the source organism (endogenous DNA) is present in most, if not all, naturally occurring biological tissues and other biomaterials, including solid and liquid waste products. However, its quality and quantity vary considerable due to a complicated array of factors that relate to their biological role, and their biogenesis. For example, at the simplest level, a tissue's cell density will provide an obvious contribution to the total amount of extractable DNA per unit mass sampled, with cell-dense tissues (e.g. muscle, liver and blood) yielding considerably more DNA than those that contain many fewer cells (e.g. bone, hair, nail, scales, horn and feather). However, this relationship itself is not straightforward, with cell densities of particular tissues varying not only between taxa, but between individuals within a species, and even within single individuals, varying either spatially across the organism, or temporally throughout the organism's life. In addition to the total amount of DNA present, the relative contribution of different endogenous DNA sources to the total will also vary considerably. For example, metabolically active tissues such as vertebrate red muscle or liver contain significantly higher levels of mitochondria per cell, thus mitochondrial DNA per cell, than other, less active tissues (e.g. bone). In addition to living tissues, many organisms also contain or produce tissues that contain much lower levels of DNA as a result of their formation. Examples include tissues that are secreted from living cells, and that might contain shed, or entrapped remnants of dead progenitor cells (e.g. trapped ameloblasts or osteocytes in tooth enamel or bone), waste products that might contain shed cells from the waste producing organs (e.g. gut epithelial cells shed into feces) or tissues that derive specifically from living progenitor cells that undergo cell-death during their biogenesis. Notably this includes the so-called ‘keratinous’ tissues, that in vertebrates include hair, nail, horn, feather and scales. As a result of this cell death, the DNA present is not just at low level, but heavily fragmented (e.g. Forslind and Swanbeck, 1966, Linch et al., 2001, Olsen et al., this issue).
Given this, it is not surprising that the earliest studies investigating the DNA content of hair and nail (and horn, which has similar properties to both) did not herald good news for future genetic analyses. In 1988, Higuchi et al. (Higuchi et al., 1988) reported that a single, freshly shed human hair contains on average no more than 200 ng of extractable DNA. A subsequent study investigating the DNA content of naturally shed hair indicated that as little as 10 ng DNA could be extracted per hair, falling after several months to less than 1 ng (Allen et al., 1998). Nevertheless, despite these drawbacks, over the past 2 decades scientists have demonstrated that, although there is a clear difference in quality of DNA recoverable from the root (high quality) versus non-root components (poor quality) of hair and nail, within limits, PCR amplifiable mitochondrial (mtDNA) and nuclear DNA (nuDNA) can be retrieved from almost all sources of ‘fresh’ (i.e. not historic/ancient) hair, including from the human head, eyebrow, pubic and torso hair (e.g. Baker et al., 2001), and nail (e.g. Tahir and Watson, 1995, Anderson et al., 1999, Cline et al., 2003). As such, having been validated for forensic use in 1995 (Wilson et al., 1995a, Wilson et al., 1995b), DNA extracted from hair was first used as forensic evidence 1996 (cf. Allen et al., 1998). To date, numerous biological (e.g. Vigilant et al., 1989, Taberlet et al., 1993, Morin et al., 1994, Garner and Ryder, 1996, Murata and Masuda, 1996, Gagneux et al., 1997, Takami et al., 1998, Hsieh et al., 2003, Alberts et al., 2010), and several thousand forensic cases have relied on hair, horn and nail as a source of DNA (cf. Allen et al., 1998).
Furthermore, over the past decade, studies have begun to demonstrate that under the appropriate conditions, DNA may also survive for considerable time periods within hair and nail. Initial PCR-based studies demonstrated the survival of mtDNA in museum and archaeological collections that were several hundred years old (e.g. Baker, 2001, Thangaraj et al., 2003, Gilbert et al., 2004b, Ricaut et al., 2006, Amory et al., 2007, Gilbert et al., 2007a, Wilson et al., 2007a, Wilson et al., 2007b, Matheson et al., 2009, Melchior et al., 2010), and in much older archaeologically preserved samples including approximately 2500 year-old horses, 9400 year-old bighorn sheep, and >64,800 year-old bison (Bonnichsen et al., 2001, Gilbert et al., 2004b). A number of studies also demonstrated that, occasionally, short fragments of nuDNA could be recovered (e.g. Amory et al., 2007), although this was challenged by the extremely short nature of the nuDNA in the samples. However breakthroughs in sequencing technologies – in particular initially the release of the first generation of Roche/454 ‘FLX’ platforms, and subsequently the Illumina ‘Genome Analyzer’ platforms – changed this. These so-called ‘second-generation’ sequencers rely on conversion of extracted DNA into ‘libraries’ through the ligation of universal adaptor sequences to the ends of the DNA templates. These adaptors subsequently provide PCR and sequencing primer ligation sites, leading ultimately to the sequencing of hundreds of thousands, to millions, of template molecules in parallel per sequencing reaction. In addition to the extremely large increase in sequence that can be generated through massive parallel sequencing of the libraries, a key benefit of this technique with regard to degraded DNA, is that extremely short DNA fragments can be sequenced – well below the 50–60 bp minima that conventional studies require. With this in mind, beginning in 2007, a series of studies were published that demonstrated how such techniques could be applied to DNA extracted from historic and ancient hair and nail material, in order to generate large amounts of mitochondrial data, including complete mtDNA genomes of taxa including mammoths, humans, thylacines and extinct and extant rhinos (e.g. Gilbert et al., 2007b, Gilbert et al., 2008a, Gilbert et al., 2008b, Willerslev et al., 2009, Miller et al., 2009), and nuclear DNA (Miller et al., 2008, Gilbert et al., 2008b) – culminating in 2010 with the publication of the first high-quality ancient nuclear genome, that of a ca. 4500 year old extinct ‘Saqqaq’ Greenlander sequenced at 20× coverage (Rasmussen et al., 2010).
Despite these successes, and the ongoing interest in hair and nail as a source of ancient and modern DNA, large gaps exist in our knowledge of the quality of the DNA in such tissues, and in short, what one might be able to expect prior to commencing a study on such material. For example, although it is widely agreed that while the root of fresh nail or hair contains high quality DNA, and that the hair shaft/nail contains only short DNA fragments (e.g. Linch and Prahlow, 2001, McNevin et al., 2005, Opel et al., 2008), the size range of these, whether they vary between nuDNA and mtDNA, and if they vary within single samples, across a single individual, and even between individuals and species, has not been explored in detail. In this review we discuss the current state of knowledge of DNA in hair and nail in light of the biogenesis of the tissues. We furthermore discuss what this data may lead us to expect in regard to the quality of nucleic acids in such hair, and outline what future studies may offer with regard to our understanding and exploitation of, hair and nail in the genetic context.
Section snippets
The structure and biogenesis of hair and nail, and its relevance to DNA quality
Although morphologically the macrostructure of nail and hair are quite different, their underlying biogenesis, and thus microstructural components are similar, and it is these that likely play a key role in regard to the quality of DNA in such materials. As an understanding of these features is useful when considering why DNA quality may vary in such tissues, a brief description follows.
The hair (Fig. 1) is formed and hardened in the hair follicle, deriving from matrix germinal cells, via the
The implications of the hair and nail biogenesis process on genetic studies
There are a number of limitations associated with working with hair and nail in a genetic context. Some of these arise as a result of the components within the tissues, and occasionally due to artificial modification of the tissues. Examples include the aforementioned effect of melanin as a PCR inhibitor (Uchichi et al., 1992, Wilson and Budowle, 1993), and reports that some artificial hair treatments, including peroxide bleaching in humans, and the dying of sheep hair (wool) using metal-based
DNA degradation within the hair and nail
An appreciation of how DNA degrades over the long term in nail and hair may help future studies gauge a priori the chance of a successful outcome. Although there have been too few studies to allow us to draw any general conclusions about nail, DNA has been successfully recovered from a number of hair shafts, including mammoths and bison that date beyond the limit of AMS 14C dating (Gilbert et al., 2004b, Gilbert et al., 2007b, Gilbert et al., 2008b). However in other samples we have had no
Heteroplasmy and allelic dropout
Those aiming to use hair and nail as a source of modern or ancient DNA must be aware of the potentially confounding factor of mtDNA heteroplasmy (the occurrence of multiple different mitochondrial DNA sequences in individual cells or tissues). This has been recorded in several studies (Sekiguchi et al., 2004, Tully et al., 2004), including a small proportion of samples investigated as part of a collaborative study involving numerous forensic genetic research groups (Tully et al., 2004). In some
Conclusions and future directions
Over the past 20 years, hair and nail have been validated as useful sources of both mtDNA and nuDNA – with the caveat that low quantities of fragmented DNA are suitable for the study requirements. As demonstrated with the recent publication of the Saqqaq genome, a 20× covered ancient human genome sequenced to ca. 85% completeness (a value close to the theoretical limit of the technology used) (Rasmussen et al., 2010), in many situations the limits come not down to the DNA quality itself, but
Acknowledgements
MTPG acknowledges the Danish Council for Independent Research-Natural Sciences ‘Skou’ grant 272-07-0279 for funding his research.
References (80)
- et al.
STR typing of ancient DNA extracted from hair shafts of Siberian mummies
Forensic Sci. Int.
(2007) - et al.
Methods for the study of ancient hair: radiocarbon dates and gene sequences from individual hairs
J. Archaeol. Sci.
(2001) - et al.
Keratin formation in the hair follicle
Exp. Cell Res.
(1966) - et al.
Mitochondrial DNA diversity in gorillas
Mol. Phylogenet. Evol.
(1996) - et al.
Ancient mitochondrial DNA from hair
Curr. Biol.
(2004) - et al.
Histological correlates of post mortem DNA damage in degraded hair
Forensic Sci. Int.
(2006) - et al.
Species identification of rhinocerous horns using the cytochrome b gene
Forensic Sci. Int.
(2003) - et al.
Evaluation of a decontamination protocol for hair shafts before mtDNA sequencing
Forensic Sci. Int.
(1998) - et al.
Structural stability of hair over three thousand years
J. Archaeol. Sci.
(1987) - et al.
Short tandem repeat (STR) genotyping of keratinised hair. Part 1. Review of current status and knowledge gaps
Forensic Sci. Int.
(2005)