K.M. Sullivan, R. Alliston-Greiner, F.I.A. Archampong, R. Piercy, G. Tully, P. Gill, and C. Lloyd-Davies
The Forensic Science Service, Priory House, Gooch Street North, Birmingham, B5 6QQ, U.K.

× Ø × Ø × Ø × Ø × Ø × Ø × Ø × Ø × Ø × Ø × Ø × Ø × Ø × Ø × Ø

Sequencing of the mtDNA control region is an effective forensic method for the comparison of samples containing insufficient nuclear DNA for analysis, such as hair shaft, bone, and old and degraded samples (1 - 6). The Forensic Science Service has offered mtDNA sequence analysis as a service since November 1994. Hair shaft has been the main evidence type in 52% of the 23 cases undertaken to date. Work to optimise extraction of mtDNA from hair yielded unexpected results. Multiple hair shafts from an individual were analysed and appeared to differ from each other and a reference blood extract at a single position. This is in direct contrast to a previous report in which mtDNA control region sequence analysis yielded no differences between multiple hair shaft samples and reference bloods (6). Variation in levels of mtDNA heteroplasmy has been observed previously between multiple hair follicles from single donors (7, 8). However these were disease-related mutations within the mtDNA coding region.

To confirm the observation of mtDNA control region mutation differences between hair shaft and reference saliva samples, 12 hair shafts from a single donor were analysed. Differences between single hair shafts and the reference sample were authenticated by repeat extraction of the hair shaft. We report that five out of the 12 hair shafts analysed differed at one nucleotide position from the reference saliva mtDNA control region sequence.

To minimise the possibility of contamination, hair shaft DNA extraction and amplification set-up were performed in a dedicated laboratory. Reference sample extraction and amplification set-up were carried out in a separate laboratory, generally after hair shaft extractions had been processed. Amplified DNA was handled in a third laboratory. If contamination was detected in a negative control, the experimental batch was rejected.

200 µl aliquots of liquid saliva were centrifuged at 11,000 x g for 3 min, the supernatants discarded, and DNA extracted from the pellets using DNA DIRECT™ (Dynal, Oslo, Norway) according to the manufacturer’s instructions.

Hair shaft samples were extracted as follows: handling the hair with clean fine forceps and using a sterile scalpel blade, the root was removed and ~ 2 cm of hair shaft (or one third of the total hair) proximal to the root was placed in a sterile microcentrifuge tube. Following three washes (vortex mixing for 30s in 1ml sterile distilled water), the hair shaft was removed to a sterile 0.6ml microcentrifuge tube containing 250 µl extraction mix (20% Chelex 100Ò Resin, 250 mg proteinase K, 40 mM DTT) using sterilized fine forceps. The hair sample was submerged in the extraction mix, incubated until no hair fragments remained visible to the naked eye (either 56°C for up to 4 hr or at 37°C overnight) and then boiled for 8 min. The extraction mix was centrifuged at 11,000 x g for 3 min to pellet any undigested hair and the Chelex 100Ò Resin, and the supernatant removed to a clean sterile tube. To avoid repeated freeze-thawing, supernatants were aliquoted prior to storing at -20°C.

Amplification and subsequent solid-phase sequencing of hypervariable (HV) regions 1 and 2 of the mtDNA control region utilised two rounds of nested amplification as described (9) except that: 0.1 to 10 µl of DNA extracts were amplified in the first round reactions; HV 2 was amplified in the second round using primer pair L00029 and H00408 (10); and sequencing products were loaded on a 4.25% acrylamide, 6M urea gel in a Perkin Elmer 377 Sequencer (Perkin-Elmer Corp., Applied Biosystems Division, Warrington, U.K.) and electrophoresed at 1680 V for 7 hours.

Where only a single round of PCR was used, amplification was through 35 cycles as described previously (11), using primer pair (-21M13)H16239 and (Biotin)L15997.

For each HV region, two sets of PCR products were generated with alternate strands biotinylated to enable analysis and comparison of both strands from reference samples to verify the mtDNA sequence. For amplification products from hair shaft extracts, one strand was initially analysed. If differences to reference samples were identified, the complementary strand was analysed and compared. Where a sample was refractory to sequencing of both strands, duplicate analyses of the same strand were used to verify the mtDNA sequence. All sequence electropherograms were independently and critically examined by two analysts.

781 bp of mtDNA control region sequence was determined for DNA extracts from 12 hair shaft samples from a single donor. Seven out of 12 gave identical sequence to the reference saliva sample (Table 1). The remaining five samples displayed a single difference at nt 16093 (numbering according to reference 12); three hair shafts were heteroplasmic C/T (Figure 1B) and two were homoplasmic T. The reference saliva sample was homoplasmic C at this position (Figure 1A). These nucleotide assignments were confirmed by analysing both strands of the mtDNA, and by repeat extractions and analyses of further portions of the same hair shaft samples (Table 1 and Figure 1C).

Contamination was excluded as an explanation for these results as: extraction and amplification negatives were clean; careful examination of the sequence data revealed no other evidence of a mixture; and consistent results were obtained between repeat extractions. A further possible explanation is that the difference at nt 16093 was a technical artefact due to the large number of amplification cycles (50 rounds) in the analysis. To examine this explanation, one hair shaft sample was re-analysed with amplification through only 35 cycles in a single round using the conditions normally employed for amplification of degraded samples. For both the 35 cycle and 50 cycle amplifications, the mtDNA extracted from the hair shaft displayed C/T heteroplasmy at nt 16093 (Figure 2).

We have demonstrated for the first time the presence of sequence variation within the mtDNA control region between hair samples taken from the same individual. The cause of this phenomenon has not yet been determined. However, similar results have been reported for two mtDNA disease-related coding region heteroplasmic mutations (7, 8). All of these observations may be directly related to the unusual biology of the hair growth cycle. These preliminary results have important implications for the reporting of hair shaft mtDNA results in forensic casework, and indicate that extreme caution should be exercised in interpreting the significance of only one or two nucleotide differences between hair shaft and reference results.

A population survey is currently underway to determine the extent of this phenomenon.

REFERENCES

Table 1. Identity of nt 16093 in the mtDNA control region sequence from 12 single hair shafts of a single donor.
The saliva reference mtDNA sequence was C at nt 16093. All other nucleotides determined in the
mtDNA sequence from the hair shafts matched those of the donor reference sample.

Figure 1. Comparison of mtDNA sequence (nt 16084 - 16100) from the saliva and a single hair shaft of the same individual. Sequence from the saliva (A) is homoplasmic C at nt 16093 (arrowed). Sequences from two separate extractions of the single hair shaft (B and C) are heteroplasmic C/T at nt 16093 (arrowed).

Figure 2. MtDNA sequence (nt 16078 - 16105) from a single hair shaft heteroplasmic C/T at nt 16093 (arrowed). Sequence data generated from mtDNA amplified through two rounds of 25 cycles each (A), and through one round of 35 cycles (B) are identical.