Touch DNA, Hair, Bone, Teeth and Challenging Substrates

Hair is one of the most commonly submitted forensic exhibits, and one of the most misunderstood in terms of DNA yield. The DNA source within a hair is almost entirely restricted to the nucleated cells of the root sheath, not the shaft.

Rooted (anagen-phase) hair. A hair pulled forcibly from the scalp retains a visible root tag, a translucent gelatinous mass of root sheath cells at the proximal end. These cells carry nuclear DNA in sufficient quantity for autosomal STR profiling on a standard amplification. Published data from the FBI Hair Evidence Analysis project (2013) and from the SWGDAM mtDNA guidelines show that rooted hairs with an intact root tag regularly yield profileable nuclear DNA. The Nuclear DNA Advisory Board (NDAB) standards for hair root analysis require documentation of root morphology and cellular content before extraction.

Shed (telogen or anagen-shed) hair. Naturally shed hairs have no root tag or a dry, club-shaped anagen root with little attached cellular material. Nuclear DNA yield from the shaft itself is insufficient for autosomal STR profiling in the vast majority of cases. The shaft contains mitochondrial DNA (mtDNA) in the cells of the medulla and in the residual cytoplasm of the cortical cells, but not in quantities that reliably support nuclear typing. mtDNA profiling from the shaft is well-validated and routinely used, it is the analytical pathway for shed hairs recovered from crime scenes. The SWGMAT mtDNA guidelines and the FBI Forensic Biology manual document the protocol. Casework examples include the serial-killer cases prosecuted in the 1990s using mtDNA from hair, a technique that was validated in R v. Deen (UK, 1994) and People v. Soto (California, 1999).

Practical workflow. Before any hair extraction, the examiner documents the root tag status and morphology under a stereomicroscope. A tag-bearing hair proceeds to nuclear DNA extraction (typically organic or column-based). A tag-absent hair proceeds to mtDNA extraction unless contextual information (e.g. the hair is from a body and has a fresh root) warrants a nuclear attempt. Hair shaft washes (a detergent wash before extraction) remove surface contamination, a standard step per UK LGC Forensics and US FBI protocols.

Bone is the substrate of choice when soft tissue is unavailable: it resists decomposition and environmental insult far better than any cellular tissue. But recovering DNA from bone requires disaggregating a mineralised matrix in which DNA is distributed through the hydroxyapatite crystal lattice of the cortex and through the cells (osteocytes) within that lattice.

Sampling site selection. Not all bone is equally useful. The dense cortical bone of the mid-shaft femur is the standard reference substrate for skeletal DNA casework, used in INTERPOL DVI guidelines and in the FBI mass-fatality protocols applied at the 9/11 World Trade Center identification effort and the MH17 victim identification (Netherlands, 2014-ongoing). Trabecular (spongy) bone degrades faster because its larger surface area exposes more DNA to environmental DNases and acidic groundwater. In practice, INTERPOL DVI casework teams and the International Commission on Missing Persons (ICMP) preferentially sample the petrous portion of the temporal bone of the skull, which is the densest bone in the human body. A 2016 study by Pilli and colleagues (University of Rome, published in Forensic Science International: Genetics) quantified the petrous bone as yielding 10-200 times more human DNA per gram than femoral cortex from the same individual after burial, confirming field observations from tsunami and conflict victim identifications.

Decontamination. The outer surface of bone recovered from decomposed or buried remains is heavily contaminated with environmental DNA (plant, bacterial, fungal). Surface decontamination is achieved by sequential treatment: abrasion with a Dremel tool or grinding wheel under laminar-flow conditions removes the outer 1-2 mm, followed by UV irradiation of all cut surfaces for at least 30 minutes. ICMP Sarajevo protocols, the German Bundeskriminalamt mass-fatality SOPs, and US military AFMES (Armed Forces Medical Examiner System) DNA laboratory procedures all include this decontamination step.

Pulverisation. Decontaminated bone is reduced to a fine powder using a cryogenic mill (SPEX SamplePrep 6870D Freezer/Mill is the instrument referenced in most published protocols, including NIST OSAC and FBI procedures) or by a liquid-nitrogen grinding method. Cryogenic milling reduces the sample to 100-200 micron particles, maximising surface area for demineralisation while preventing heat-induced DNA denaturation.

Demineralisation. Two main approaches exist. (a) Organic acid decalcification: 0.5 M EDTA (pH 8.0) chelates calcium ions from hydroxyapatite, leaving the organic matrix including DNA. This step takes 24-72 hours at 56°C with agitation. (b) HCl demineralisation: a brief wash with dilute hydrochloric acid (0.1-0.5 M) dissolves the mineral phase quickly, but risks DNA damage if the pH is not controlled carefully. EDTA decalcification is the validated standard in most operational labs. After demineralisation, the organic matrix is digested with proteinase K and SDS, then the extract proceeds to DNA extraction (silica column or organic method).

Teeth are often the last tissue to yield recoverable nuclear DNA in severely compromised remains. The enamel and dentine encapsulate the pulp chamber, creating a sealed environment that protects the pulp cells from environmental insult. This protection is the reason teeth from fire victims, marine casualties and long-buried skeletal remains consistently outperform other tissue sources for nuclear DNA yield.

Pulp chamber anatomy and sampling. The pulp contains fibroblasts, odontoblasts, blood vessels and nerve fibres, all of which carry nuclear DNA. Sampling the pulp requires opening the crown: a dental bur or diamond-tipped saw separates the crown from the root along the cemento-enamel junction under sterile conditions. The pulp tissue is mechanically separated from the dentine walls, collected and transferred directly to a lysis buffer. Alternatively, the crown is ground whole after surface decontamination, though this risks diluting the pulp DNA with dentine matrix.

Root cementum and dentine. When the pulp is entirely resorbed or the tooth is heavily mineralised (common in mature adults), DNA can still be recovered from the cementum (thin cellular layer on the root surface) and from the dentine tubules, which contain cytoplasmic odontoblast processes. Dentine-based DNA extraction follows the same cryogenic milling and EDTA decalcification protocol as bone, with approximately 5-10% the yield per gram of petrous bone.

Casework application. The ICMP Srebrenica mass-grave identification programme (1995 Srebrenica massacre, identification work ongoing) relies heavily on tooth pulp for the most degraded remains. The AFMES used dental-pulp extraction as a primary nuclear-DNA strategy for WTC victims identified from fragmentary remains. In India, the Central Forensic Science Laboratory in Hyderabad applied dental-pulp DNA extraction in the Uphaar Cinema fire tragedy victim identification (1997) and in identification work following the 2004 tsunami along the Tamil Nadu and Andhra coasts.

Stochastic effects are the random sampling artefacts that arise when the number of template DNA molecules entering a PCR reaction is very small. They are the central interpretive challenge in touch DNA, shed-hair mtDNA typing, and degraded bone or dental-pulp extracts.

Allele drop-out. At very low template concentration, one allele at a heterozygous locus may fail to amplify simply because the template molecule for that allele was not present in the volume of extract added to the PCR reaction. The result is a single-peak (apparent homozygote) electropherogram where a two-peak profile would be expected. Drop-out probability increases as template quantity decreases below approximately 100 pg of amplifiable DNA per reaction, a threshold documented in the SWGDAM Interpretation Guidelines and in published studies from the Netherlands Forensic Institute.

Allele drop-in. Random low-level contributions from environmental contamination, reagent contamination, or minor contributors become visible when the signal threshold of the electropherogram is set to include very faint peaks. A single stray molecule of a technician's DNA amplified in a low-template PCR can produce an interpretable peak. Strict positive and negative controls, pre-PCR decontamination, and reagent blank monitoring are mandatory.

Stutter. STR stutter (PCR artefact producing a peak one repeat unit below the true allele) increases in apparent intensity relative to the true allele as template concentration falls. Stutter percentage thresholds validated on high-template DNA (typically 10-15% of allele peak height) may not adequately discriminate true minor-contributor alleles from stutter in a low-template mixture.

Mitigation strategies. The operational responses to stochastic effects include: (a) replication (multiple amplifications of the same extract to confirm allele calls across replicates, the LCN replication protocol first published by the UK FSS and adopted by NIST OSAC guidelines); (b) probabilistic genotyping (STRmix, TrueAllele, EuroForMix) which models stochastic effects explicitly rather than applying fixed stutter thresholds; and (c) increased cycle number PCR (up to 34 cycles) to amplify more copies from fewer template molecules, with the trade-off that artefact generation also increases.

The admissibility of low-template DNA evidence has been contested across jurisdictions since the technique entered casework.

R v. Hoey (Northern Ireland, 2007). The prosecution of Sean Hoey for the Omagh bomb (1998) relied substantially on LCN touch DNA from detonator components. Justice Weir acquitted Hoey and delivered a detailed criticism of the LCN methodology as applied in the case, finding that the process had not been validated to the standard required for the evidence to be reliable. The judgment did not rule LCN DNA inadmissible in principle; it required that the methodology meet demonstrable validation standards. The UK Forensic Science Regulator subsequently commissioned a review of LCN procedures that produced the 2007 report of the FSR DNA Working Group, which set out the standards under which LCN results could be reported. These standards were integrated into the ACPO and later FSR guidance and are now embedded in the Regulator's 2023 Codes.

People v. Holland (California, 2007). The California Court of Appeal in People v. Holland addressed touch DNA from a steering wheel in a murder case. The court found the technique admissible under the Kelly-Frye standard applied in California, provided the laboratory had validated the method and the results were interpreted within documented stochastic-effect parameters. The case established a US precedent that touch DNA admissibility depends on methodology documentation, not on categorical rejection of the technology.

Australia and New Zealand. The Victoria Police Forensic Services Centre and ESR New Zealand each publish publicly available validation documents for their low-template DNA workflows. Australian courts have admitted touch DNA evidence in multiple cases under the uniform Evidence Acts, with the Scientific Committee on DNA Analysis (SCODA) providing interpretive guidance that courts reference.