Forensic DNA Analysis: Strengths and Limitations

Mitochondrial DNA (mtDNA) analysis is becoming more common in criminal investigations to characterize forensic biological specimen. This paper will examine mtDNA analysis in the forensic field, the expertise and training required and its strengths and limitations. The strengths of mtDNA analysis are the following: mtDNA has a high copy number, it provides an alternative option when nuclear DNA (nucDNA) is not viable, better recovery from degraded samples, and confirms maternal relatedness and some discriminatory power using hypervariable regions. The limitations of mtDNA analysis are that it does not provide definitive identification, heteroplasmy, risk of contamination, paternal leakage and is time-consuming and expensive. Lastly, this paper will discuss the development of next generation sequencing (NGS) technologies and its implications on future cases.

Forensic mtDNA Analysis

Forensic mtDNA analysis allows analysts to develop a DNA profile from samples that cannot yield results of nucDNA. One mitochondrion contains 2-10 copies of mtDNA and there are thousands of mitochondria in each cell (Hameed, Jebor & Kareem, 2015). Mitochondria are energy-producing organelles and are often referred as the powerhouse of the cell (van Oven & Kayser, 2008). mtDNA is a circular, double-stranded molecule and has 16,569 base pairs, which encode 37 genes for 22 tRNAs, two rRNAs and 13 mRNAs (Anderson, 2017; Hameed et al., 2015). Also, mtDNA haplotypes are genetic information that is uniparentally inherited from the mother to offspring through the egg cell; accordingly, individuals from the same maternal lineage have identical mtDNA (Parson, 2014). For this reason, the assumption is that all mtDNA types can be traced back to a common maternal ancestor; mtDNA sequence variation evolved as a result of mutations (van Oven & Kayser, 2008). There is more sequence divergence in mtDNA than nucDNA (Hameed et al., 2015). Mutations can occur during mtDNA replication; consequently, rate of change of mtDNA is five times faster than nucDNA (Hameed et al., 2015).

Expertise and Training

There are several steps necessary to become a forensic biologist. Although the minimum requirement for entry level forensic analyst is a three-year diploma, most analysts have Bachelor of Science (BSc) degrees and some also have Master’s degrees (Anderson, 2007). Forensic lab personnel have a strong science background; analysts can specialize in any biological science, such as Biology, Biochemistry, Chemistry, Molecular Biology or Forensic Biology (Anderson, 2007). Once hired, applicants undergo an understudy period of 9-18 months and receive on-going training (Anderson, 2007). Additionally, forensic analysts must be competent in their written and oral communication skills, as a large proportion of their duties require them to provide expert testimony in court (Anderson, 2007). The analysts must be able to effectively communicate relevant scientific knowledge in lay terms for those without a scientific background, such as the jury (Anderson, 2007).

Strengths

One significant strength of mtDNA is that it has a high copy number in each cell. Compared to only two copies of nucDNA in a cell, thousands of mtDNA copies are present in each cell, increasing the likelihood of obtaining viable DNA (Alzarez-Cubero et al., 2012). The mtDNA copies are distributed throughout the cytoplasm of cells, which increases its sensitivity and chances of detection (Sampath & Jagannathan, 2014). Also, mtDNA is present in red blood cells, whereas nucDNA is not due to absence of nuclei (Sampath & Jagannathan, 2014). In hair samples, mtDNA can be detected anywhere along the hair follicle, including the hair shaft; in comparison, nucDNA is only present in the hair root (Budowle, Allard, Wilson & Chakraborty, 2003). As expressed, mtDNA is present in more areas of the body and in higher quantities.

Related to a high copy number, mtDNA analyses can be used when only small quantities of DNA are available. When obtaining nucDNA is not possible, sequencing of mtDNA is an alternative option (Nilsson, Andréasson-Jansson, Ingram & Allen, 2008). Samples, such as hairs, teeth or bone often contain low amounts of DNA (Nilsson et al., 2008). Collecting mtDNA from dental tissue is relatively easy due to large pulp chambers and resistance to decomposition (Sampath & Jagannathan, 2014). Hair samples as small as 0.2 cm have a high probability (86%) of obtaining partial or full profiles; however, thick and pigmented hairs produce the best results (Melton, Dimick, Higgins, Lindstrom & Nelson, 2005). As a result, mtDNA analysis is preferred when forensic analysts only have access to small quantities of DNA.

Forensic mtDNA analysis is optimal for old or degraded samples, as mtDNA is resistant to harsh environmental conditions. Mitochondria’s have a strong protein coat that protects mtDNA from bacterial enzymes (Hameed et al., 2015). Conditions where mtDNA analysis may be ideal are for fragmented skeletal remains, bones or teeth that have been exposed to long durations of acidity, high temperature or humidity, fingernails, shed hairs and samples that were unsuccessful for nucDNA detection (Alaeddini, Walsh, & Abbas, 2010; Kumar, Kumar, Honnugar, Kumar & Hallikeri, 2012). Also, improperly stored biological materials may not yield a nucDNA profile but may yield results for mtDNA (U.S. Department of Justice, 2002).

Remains from the World Trade Center attack, Tsar Nicholas II (remains from 1918) and Neanderthal bones have been characterized using mtDNA analysis (Budowle et al., 2003). Since these remains are heavily degraded and the latter two are significantly older, it is unlikely that obtaining nucDNA is possible. For Tsar Nicholas II, mtDNA was extracted from the putative bones and was matched with living maternal relatives (Budowle et al., 2003). In another case, mtDNA samples were extracted from the mouth, small fragments of the brain, feet and hand bones of bodies that were severely burned; this demonstrates that mtDNA can survive even in severely degraded and charred remains (Ricci, 2015).

Matrilineal inheritance of mtDNA allows living maternal relatives to be a reference sample to compare with unidentified remains (e.g. missing persons and mass casualties). There are large databases with DNA profiles of convicted offenders, human remains, crime scene samples and maternal relatives of missing persons (Alvarez-Cubero et al., 2012). In missing persons case where nucDNA samples cannot be obtained of the victim, living maternal relatives are able to submit their mtDNA to determine if the samples match (Alvarez-Cubero et al., 2012). Also, since there is a lack of recombination, maternal relatives several generations down can also serve as reference samples (Alvarez-Cubero et al., 2012). Many missing persons database of unidentified remains and maternal relatives automatically compare genetic data to find similar matches; this is a useful tool and has been successful in identifying remains (Alzarez-Cubero et al., 2012). Lastly, mtDNA discovered at crime scenes can potentially be an exculpatory tool to reduce wrongful convictions and potential candidates, consequently, redirecting resources and increasing efficiency of investigations (Alzarez-Cubero et al., 2012).

Although mtDNA does not provide positive identification, there are distinctive differences in the hypervariable regions of mtDNA that have considerable discriminatory power. Individuals have unique differences in two hypervariable regions of mtDNA: hypervariable region 1 (HV1) and hypervariable 2 (HV2) (Anderson, 2017). HV1 and HV2 contain approximately 610 bases of information; these control regions are highly variable due to a high evolutionary rate of mtDNA (Parson & Coble, 2001).

Gonçalves, Fridman & Krieger (2011) evaluated the discriminatory power of HV1/HV2 variables using 290 unrelated participants. The results showed that 77% of the participants could be discriminated using the HV1/HV2 types (Gonçalves et al., 2011). Although 23% of participants could not be distinguished using HV1/HV2 types, the study showed a large number of unique haplotypes. Parson & Coble (2001) found majority of the mtDNA types vary and 982 of 1175 mtDNA types are unique. It is likely that newer technology will be able to detect more distinguishing features (Gonçalves et al., 2011).

Limitations

One significant limitation of mtDNA is that it does not provide definitive identification. Although mtDNA is a class characteristic, it is a single linked molecule and the probabilities of individual polymorphism cannot be multiplied (Parson & Coble, 2001). mtDNA only reveals that the unknown sample shares the same maternal lineage as the reference sample (Parson & Coble, 2001). Parson & Coble (2001) claim that mtDNA should be treated as supplementary circumstantial evidence to assist in criminal proceedings. Additionally, because all maternal relatives share the same mtDNA, they are indistinguishable from one another on this basis. In cases where multiple family members are missing or died in a mass casualty, mtDNA analysis will have low probative value and will not help identify the unknown sample. Furthermore, studies have found some common HV1/HV2 types in the population. Parson & Coble (2001) found that about 7% of U.S. Caucasians share the most common type of HV1/HV2 type and Americans with African and Hispanic descent share similar mtDNA sequences as Caucasians. As a result, there is a chance for a random match to occur between known and unknown samples (Parson & Coble, 2001).

Another limitation for forensic mtDNA analysis is heteroplasmy. Heteroplasmy is the presence of more than one mtDNA type in an individual (Budowle et al., 2003). Heteroplasmy can manifest in three ways: multiple mtDNA type in a single tissue, heteroplasmic in one tissue and homoplasmic in another, or one mtDNA type in one tissue and a different type in another tissue (Budowle et al., 2003). Although rare, when heteroplasmy is observed, it is usually found in a single base of the mtDNA (Budowle et al., 2003). Budowle et al. (2003) report that one individual had some hairs that were homoplasmic and others were heteroplasmic; these results challenge the value of mtDNA analysis in forensic cases. mtDNA analysis is premised on the assumption that mtDNA is identical across maternal lines; the findings suggest that DNA databases and forensic analysts are potentially administrating false negatives.

Heteroplasmy is frequently observed in hair samples because of mitochondrial bottleneck (Budowle et al., 2003). Bottlenecking is a developmental process of cell-to-cell variability from a mutation; cell-level selection eliminates cells with high heteroplasmy and cells with low heteroplasmy are retained (Johnson et al., 2015). Samples with low heteroplasmy create an obstacle for mtDNA analysis, as they are difficult to detect. In cases where heteroplasmy is observed in a known sample and not in the recovered sample (or vice versa), additional samples should be collected to determine if they are a match (Bär et al., 2000).

There is an increased risk of contamination in forensic analysis of mtDNA. In highly degraded remains, there is often minute amounts of mtDNA for extraction (Wurmb-Shwark, Heinrich, Freudenberg, Gebür & Schwark, 2007). Particularly in cases of historical remains, it is difficult to be confident that the mtDNA extracted is from the victim, rather than modern DNA, such as from those that handled the sample (Wurmb-Shwark et al., 2007). When two objects are in contact with one another, there is an inevitable exchange of material, called Locard’s exchange principle (Anderson, 2017). Contamination can occur during evidence collection at crime scenes, evidence recovery or during forensic analysis (Wurmb-Shwark et al., 2007). Also, contamination can occur by inexperienced civilians that initially found and reported the remains (Wurmb-Shwark et al., 2007). Contamination concerns can be addressed by collecting known samples of those that came in contact with the DNA for elimination purposes; however, not all people that came in contact with the specimen will be known (Wurmb-Shwark et al., 2007). Also, contamination prevention measures should be taken, such as wearing a laboratory coat, disposable gear (e.g. gloves, caps, sleeves) and face mask, maintaining a clean and controlled workspace and monitoring a negative control (Bär et al., 2000).

Interestingly, recent research suggests that paternal mtDNA inheritance may coexist with maternal inheritance. Luo et al. (2018) found three unrelated multigenerational families, including 17 individuals, had high levels of mtDNA heteroplasmy (24-76%) and showed evidence of paternal leakage. Patients with diseases caused by mtDNA mutations were tested and revealed copies of paternal mtDNA, in addition to maternal mtDNA (Luo et al., 2018). From the 17 individuals, 13 members directly inherited paternal mtDNA and four inherited paternal mtDNA from previous generations (Luo et al., 2018). Also, mtDNA analyses were administered independently in at least two different laboratories with new technicians and new blood samples to increase the validity of the study; the results were replicated and showed no sign of contamination (Luo et al., 2018). The findings challenge the assumption that mtDNA is exclusively inherited by the mother. In regard to forensic cases, this raises concern as missing persons database only contain mtDNA from maternal relatives. However, evidence of biparental mtDNA inheritance is unusual, suggesting that majority of humans inherit mtDNA through maternal lineage.

Lastly, an important limitation of mtDNA analysis is that it is very time-consuming and costly. Many countries do not have mtDNA analysis laboratories and depend on private service providers (Melton, Holland & Holland, 2012). Forensic mtDNA analysis is expensive and tedious, as analysts need to work meticulously to avoid contamination of degraded samples (Melton et al., 2012). The cost of private testing of mtDNA is a significant barrier and law enforcement agencies infrequently use mtDNA laboratories (Melton et al., 2012). Also, many cases in mtDNA laboratories require analysis of a large number of samples; on average, each case requires about four samples but can require as many as 200 samples for a single case (Melton et al., 2012). Therefore, it is possible that mtDNA analysis laboratories experience back-logs, thus, decreasing its efficacy. Melton et al. (2012) report that majority of cases in their laboratory involve mtDNA analysis of samples from old or cold cases, crime scene hairs less than 10 mm and nonhuman samples; this demonstrates the irregularity of mtDNA analyses in criminal investigations. Furthermore, mtDNA testing laboratories have large costs due to “training, quality control, quality assurance, accreditation, proficiency testing and casework costs” (Melton et al., 2012, p.8). For these reasons, mtDNA analysis is conducted only when nucDNA is not viable and DNA is essential to the case.

Future of mtDNA Analysis

NGS technologies are being adapted to increase accuracy, reproducibility, speed and cost efficiency for mtDNA analysis. NGS technology can detect heteroplasmy at the whole mitochondrial genome level (Yang, Xie & Yan, 2014). As mentioned earlier, heteroplasmy is problematic for mtDNA analysis; more than one haplotype is present within a single person or tissue and these mutations occur with relative frequency. Current NGS technology has largely focused on nucDNA because mtDNA mutations are often present at very low heteroplasmy levels, making them difficult to detect (Bosworth, Grandhi, Gould & LaFramboise, 2017). However, newer NGS technologies can detect mtDNA of very low frequency molecules (less than 1%) (Just, Irwin & Parson, 2015). NGS technologies simultaneously conduct analyses of large number of samples and determine the base composition of single DNA molecules (Yang et al., 2014). Also, NGS technologies can recover very short mtDNA fragments, old, poor quality or low quantity specimen (Just et al., 2015).

The NGS technologies have led to three major improvements in DNA analysis. Firstly, the new technology does not require bacterial cloning of DNA fragments; they depend on the formation of NGS libraries in a cell-free system (Yang et al., 2014). Secondly, NGS technologies can parallelize thousands to millions of sequencing reaction, compared to hundreds seen in previous technologies (Yang et al., 2014). Thirdly, sequencing outputs are detected without the use of electrophoresis (technique to separate DNA molecules) (Yang et al., 2014). For these reasons, it is likely that NGS technologies will replace older mtDNA analysis techniques.

Bosworth et al. (2017) use MitoDel (computation procedure) to detect mtDNA deletions using NGS data, which was generated from human brain tissue of a previous study. Since MitoDel focuses exclusively on mtDNA, which is significantly smaller than nucDNA, it reduces the burden of mapping (Bosworth et al., 2017). The results of MitoDel were efficacious in detecting deletions present in mtDNA below 1% heteroplasmy levels, with a low false positive rate (Bosworth et al., 2017). NGS methods are relevant to forensic cases because they produce larger volumes of mtDNA sequence data at a lower cost (Just et al., 2015).

One concern for NGS technologies is that the increased sensitivity for highly degraded specimen can mistakenly detect nuclear mitochondrial DNA segment (NUMT), transposition of mtDNA into the nucDNA and confound heteroplasmy detection (Just et al., 2015). Although NUMT detection does not pose as a significant challenge, laboratory-based and bioinformatic modifications need to be made for optimal efficiency (Just et al., 2015). Also, NGS technologies need to develop mechanisms to authenticate heteroplasmy to reduce risk of contamination and false positives (Just et al., 2015). Furthermore, current forensic mtDNA analysis primarily detect polymorphisms within a hypervariable region; NGS technology has the potential to increase the discriminatory power of identification by incorporating additional polymorphic loci (Yang et al., 2014).

NGS technologies have implications on the field of forensic science and future criminal cases. Firstly, a high resolution of the full mitochondrial genome will help detect heteroplasmy. Secondly, NGS technologies will contribute to larger mtDNA databases; these databases will improve haplotype frequency estimates and provide a greater reference sample (Butler, 2015). Thirdly, and perhaps most importantly, NGS technologies significantly reduce the time and cost for mtDNA analysis. It is likely that NGS technologies will be available more widespread and more frequently used by law enforcement. Familial DNA can significantly aid investigations of missing persons, identify victims, eliminate potential candidates and redirect resources accordingly.

Overall, mtDNA analysis is a useful tool in forensic science. mtDNA has several advantages in comparison to nucDNA; mtDNA analysis is optimal for small, degraded or old samples, an alternative when nucDNA does not yield results and for reference samples of maternal relatives. Although mtDNA does not provide positive identification and has issues with heteroplasmy, paternal leakage and contamination, it is beneficial when nucDNA is not possible. The development of NGS technologies is significant for forensic mtDNA analysis; it is likely that mtDNA analysis will more commonly used in the future by law enforcement agencies.