Study population

A cohort of 30 post-mortem cases admitted for an autopsy to Salt River Mortuary in Cape Town, South Africa between 1 May 2017 and 30 August 2017 were recruited in 2017. Written informed consent for the collection of blood samples from the deceased individuals was obtained, which adhered to published ethical considerations [21]. The final cohort consisted of 22 males between the age of 22 years and 66 years, and 8 females between the age of 21 years and 57 years. Cases were included in the study if drug toxicity was a suspected primary or contributory cause of death, or if the case was a sudden or unexpected death in adults (SUDA), where the decedent had a reported history of drug use. Cases were excluded from the study if the decedent was below the age of 18 or if the body was severely burnt or decomposed.

Sample collection

Femoral blood (20 mL) was collected from each of the 30 cases at the time of autopsy. Samples were aliquoted in 4 mL volumes into one purple-top vial containing EDTA (BD Vacutainer, New Jersey, USA), two gray-top tubes with sodium fluoride and potassium oxalate (SG Vac, Johannesburg, SA), and two red-top tubes (no additive) (SG Vac, Johannesburg, SA) for each of the cases. Each case and their samples were anonymized through unique case identification numbers to ensure confidentiality.

DNA extraction (silica column)

Within 72 h after sample collection, DNA was extracted from the purple-top vial, one gray-top vial and one red-top vial for each case (n = 90). These samples did not undergo toxicological analysis and were used as controls (termed herein as P-No-Tox, G-No-Tox, and R-No-Tox, respectively). The remaining gray-top and red-top tubes for each case (termed herein as G-Tox and R-Tox, respectively) were prepared for analysis in the UCT Forensic Toxicology Unit laboratory and submitted for toxicological screening at the UCT Division of Pharmacology prior to DNA analysis (n = 60). DNA was extracted from the 150 stored blood samples using the QIAamp® DNA Investigator kit (Qiagen, Hilden, Germany) following the manufacturer’s protocol, with slight adjustments as follows: A total of 100 µL of each blood sample was utilized, and the purified DNA was eluted into 50 µL of Qiagen® ATE Elution Buffer and centrifuged at full speed (20 238 rcf) for an extended period of 90 s. This elution step was repeated, resulting in two eluates of extracted DNA for each sample. Following DNA extraction, blood samples were stored at 4 °C until the one-year and five-year time points, after which the DNA extraction process was repeated using the same protocol. At the five-year time point, 13 of the blood samples had dried, and were reconstituted in 2 mL 1X phosphate-buffered saline (Thermo Fisher Scientific, Massachusetts, USA) according to the method described by Tagliaferro et al. [22]. Molecular biology grade water (Thermo Fisher Scientific, Massachusetts, USA) stored in identical purple, gray, and red-top tubes underwent the same DNA extraction process to serve as negative controls.

DNA extraction (salting-out)

After five years, the procedure described above was no longer sufficient to recover usable DNA for Sanger sequencing. Therefore, the salting-out DNA extraction procedure developed by Miller et al. [23] was assessed, optimized, and applied to the 150 samples. A total of 2 mL of whole blood from each sample was digested with 12 mL of red blood cell lysis buffer (8.28 g/L > 99% ammonium chloride (Kimix, Durban, South Africa), 0.79 g/L > 99% ammonium bicarbonate (Sigma-Aldrich, Missouri, USA), 0.5 M EDTA at pH 7.4 (Thermo Fisher Scientific, Massachusetts, USA)) and centrifuged at 376 rcf for 20 min, after which the supernatant was discarded. The red blood cell digestion step was repeated a total of four times. The resulting cell pellet was then resuspended in 1 mL white blood cell lysis buffer (1 M Tris-HCl at pH 7.5 (Lonza, Basel, Switzerland), 0.5 M EDTA at pH 7.4 (Thermo Fisher Scientific, Massachusetts, USA), 3 M sodium chloride (Merck, New Jersey, USA)). The lysate was incubated with 40 µL of 20% SDS (Merck, New Jersey, USA) and 40 µL of 20 mg/mL Proteinase K (Qiagen, Hilden, Germany) for 24 h at 56 °C with shaking at 211 rcf.

Following overnight incubation, an equivalent amount of 6 M sodium chloride (Merck, New Jersey, USA) was added to the lysate and the samples were incubated at -20 °C for 30 min. The samples were then centrifuged at maximum speed (20 238 rcf) for 15 min, after which the resulting pellet was discarded, and the supernatant was transferred to a clean tube containing an equivalent amount of > 99.9% HPLC grade isopropanol (Sigma-Aldrich, Missouri, USA). For samples where the supernatant was not clear following centrifugation, the supernatant was transferred to a clean tube and centrifuged for an additional 10 min at 6010 rcf before being added to the absolute isopropanol. The samples were then incubated at -80 °C for 30 min and centrifuged at maximum speed for 15 min. The resulting supernatant was decanted, and the DNA pellet was washed twice with 70% ethanol (Merck, New Jersey, USA), with a 5 min centrifugation step at 15 871 rcf separating the washes. Finally, the DNA pellet was dried at 37 °C and resuspended in 50 µL of 1X Tris-EDTA buffer (Merck, New Jersey, USA).

An additional purification step using sodium acetate was tested on a subset of DNA extracts, but this did not improve results and was not tested further.

A summary of the sample collection and laboratory procedures followed throughout the longitudinal study is depicted in Fig. 1.

Fig. 1

Illustration of the procedures followed during sample collection and subsequent laboratory processing between 2017 and 2022

Assessment of DNA quantity and qualityReal-time PCR (qPCR)

DNA quantity and quality were further assessed by real-time PCR using the Quantifiler® Trio DNA Quantification kit (Thermo Fisher Scientific, Massachusetts, USA) according to the manufacturer’s protocol. This was preceded by NanoDrop™ spectrophotometry to determine whether samples required dilution to a concentration below 50 ng/µL to not exceed the upper limit of the standard curve. Amplification was performed on the 7500 Real-Time PCR system (Applied Biosystems, California, USA). The concentration of the large autosomal target was used as the final DNA concentration. An identical reaction mixture lacking template DNA functioned as a negative control. As part of the kit, a synthetic DNA template amplified alongside each sample as an internal positive control. Real-time PCR was also used to determine the ‘degradation index’ (DI). The DI values were interpreted according to the method described by Vernarecci et al. [24].

Forensic DNA profiling

In 2017, DNA profiling was performed to assess if samples that had undergone sampling for toxicological analysis in a DNA-uncontrolled laboratory had been contaminated with extraneous DNA. The PowerPlex® ESI 16 System (Promega, Wisconsin, USA) was used to prepare quarter-volume multiplex PCR reactions with 0.5 ng/µl template DNA according to the manufacturer’s recommendations. Identical reaction mixtures lacking template DNA functioned as negative controls. The multiplex system amplified the following loci: Amelogenin, D3S1358, D19S433, D2S1338, D22S1045, D16S539, D18S51, D1S1656, D10S1248, D2S441, TH01, vWA, D21S11, D12S391, D8S1179, and FGA.

Following PCR, each sample was added into an Applied Biosystems MicroAmp® optical 96-well plate containing WEN Internal Lane Standard 500 and Hi-Di™ formamide (Promega, Wisconsin, USA) according to the manufacturer’s instructions. Samples were run on an Applied Biosystems Genetic Analyser 3130 xl at 60 °C and results were analyzed using Applied Biosystems GeneMapper version 4.1 software (Applied Biosystems, California, USA).

Genetic analysis

Sanger sequencing was employed to assess whether usable sequencing data could be obtained following long-term storage in the various vial types, and not to determine the presence of variants or the phenotype of the individual. The nine coding regions of CYP2D6 (target A-I) were amplified using 2X GoTaq Green Master mix (Promega, Wisconsin, USA) and 10 µM of each primer (Table 1) in a total volume of 25 µl. Identical PCR reaction mixtures lacking template DNA to a final volume of 25 µL, functioned as negative control specimens. Agarose gel electrophoresis was utilized to determine whether amplification was successful. Prior to sequencing of the amplified products, a post-PCR clean-up step was performed with the Nucleofast® 96 PCR Clean-up kit (Macherey-Nagel, Düren, Germany) as outlined in the manufacturer’s instructions. Bi-directional sequencing was performed at the Central Analytical Facility of the University of Stellenbosch, South Africa using the BigDye® Terminator v3.0 cycle kit (Applied Biosystems, California, USA) according to their in-house protocol. The 15 resulting sequences were then compared to the CYP2D6 reference sequence available on the Ensembl genome browser Release 108 (http://www.ensembl.org; accessed 23 November 2022) using the BioEdit Sequence Alignment Editor version 7.2.5 [25] and ClustalW with 1000 bootstrap. Sequences were deemed unusable if nucleotide calling was incomplete or alignment to the reference sequence was not possible. Poor quality sequences had a high signal-to-noise ratio, and numerous chemical artefacts or errors in base pair calling, resulting in partial alignment to the reference sequence. Good quality sequences had a low signal-to-noise ratio, complete base pair calling and full alignment to the reference sequence.

Table 1 Primer sets and their corresponding CYP2D6 target regions

Based on the results of the toxicological analysis, Case 22 was selected for genetic analysis due to the detection of drugs that are metabolized by the CYP2D6 enzyme. This case involves the death of a 33-year-old male from a suspected accidental overdose. The decedent allegedly had a strong history of substance abuse and was discharged from a rehabilitation center four months prior to his death. An overdose was suspected as the decedent was allegedly found with a needle in his hand and other drug paraphernalia on scene. The postmortem examination indicated a puncture wound on the right arm, blood and vomitus in and around the mouth, and features of gastric aspiration and hepatosplenomegaly. Toxicological analysis detected the presence of methaqualone in the blood and amphetamine, methamphetamine, morphine-3-ß-D-glucuronide, 6-O-monoacetylmorphine, methaqualone, diphenhydramine, cocaine and benzoylecgonine in the urine.

DNA extracted samples from blood that had undergone toxicological sampling and screening for Case 22 were used (22G-Tox and 22R-Tox). DNA from the 22G-Tox and 22R-Tox samples extracted using the modified salting-out method was also analyzed at the five-year time point.

Data analysis

Statistical analysis was performed using the IBM SPSS Statistics Software version 28.0.1 (SPSS Inc., Illinois, USA) at a level of significance (α) of 0.05. The distributions of the data were assessed using the Shapiro-Wilk test. The data were analyzed using the Related-Samples Friedman’s Two-Way Analysis of Variance by Rank test. Effect sizes were determined using the Kendall’s W test. The main effects of vial type and time since sample collection were assessed individually to determine whether a significant difference exists in terms of DNA concentration, purity ratios and degradation index. There was no significant difference between Tox and No-Tox samples, therefore these results were combined when analyzing the effect of time since sample collection. Familywise errors were accounted for using the Bonferroni correction on multiple comparisons. To assess whether the modified salting-out DNA extraction method significantly increased DNA yield and sample purity overall and within the three tube types, a Related-Samples Wilcoxon Signed Rank test was utilized. Effect sizes were calculated by dividing the standardized test statistic Z with the square root of the number of comparisons.