Emerging Markers: DNA Methylation, miRNA, Cell-Free DNA and HIrisPlex-S Phenotyping

MicroRNAs (miRNAs) are small non-coding RNA molecules, typically 21-23 nucleotides in length, that regulate gene expression post-transcriptionally. Their tissue-specific expression patterns, blood expresses a different set of miRNAs from semen, saliva, vaginal fluid, or menstrual blood, have been proposed as a body-fluid identification system since approximately 2011.

The forensic appeal is straightforward: existing confirmatory body-fluid tests (the ABAcard p30 immunochromatographic strip for semen, the RSID-saliva for salivary amylase, phadebas for amylase non-specifically) are antibody-based or enzymatic, and each tests for a single fluid. A miRNA panel that simultaneously classifies multiple body fluids from a single extract would be a significant operational improvement. Several candidate panels have been published:

• Blood-specific miRNAs: miR-451a and miR-16-5p (both highly expressed in erythrocytes, low in other tissues)
• Semen-specific: miR-891a-5p, miR-888-5p, miR-10a-5p (expressed in seminal fluid components)
• Saliva-specific: miR-200c-3p, miR-141-3p, miR-203a-3p (salivary gland expression)
• Vaginal fluid: miR-124-3p, miR-372-3p (cervicovaginal epithelial expression)
• Menstrual blood: distinguishable from peripheral blood by a combination of miR-451a (lower than peripheral blood) and endometrial-specific markers

Quantification of miRNA panels is performed by reverse-transcription quantitative PCR (RT-qPCR) using a TaqMan miRNA assay plate or a Luminex xMAP multiplexed bead array. The key technical challenge is RNA stability: RNA degrades faster than DNA in ambient conditions, and a bloodstain left outdoors for several days may have lost sufficient miRNA for reliable profiling even when DNA extraction yields a full STR profile. Studies from the laboratory of Margreet Kloosterman at NFI (Netherlands) and from the group of Roland Wehrens at Radboud University have characterised miRNA decay kinetics at different temperatures and humidities, showing that refrigerated or dry-stored stains retain miRNA reliably for months while wet stains at high temperature show significant degradation within 48 hours.

The ENFSI DNA WG has published a provisional miRNA profiling guide (2022 update) recommending that any lab implementing miRNA body-fluid identification validate the selected panel against at least 100 reference samples per body fluid, at least three degradation conditions, and at least two technical replicates per sample. No kit has yet received ENFSI endorsement for accredited operational use; validation studies are ongoing in Germany, Sweden, the Netherlands, and Australia.

Cell-free DNA (cfDNA) refers to DNA fragments circulating extracellularly in body fluids, shed primarily from apoptotic cells and to a lesser extent from necrotic cells. In healthy individuals, plasma cfDNA is predominantly of hematopoietic origin (blood cell turnover) and consists of fragments averaging 166 base pairs, corresponding to mononucleosomal DNA. Because cfDNA fragments are short, highly fragmented, and in solution (not protected by an intact cell nucleus), they degrade faster than cellular DNA from a blood stain, but they also distribute to body surfaces and compartments that cellular DNA cannot reach.

Forensic trace cfDNA: The forensic relevance of cfDNA is primarily in two areas. First, cfDNA provides a mechanism to explain positive DNA signals on surfaces where no visible stain or cellular material is apparent: handled surfaces, doorknobs, environmental swabs, and swimming pools. The Locard exchange principle predicts DNA transfer; cfDNA shed from skin surface is the molecular substrate of that prediction for many non-touch-DNA scenarios. Second, cfDNA in plasma can be typed at STR loci if sufficient copies are present, which matters in blood-transfusion recipients whose plasma may carry donor cfDNA as a complicating factor in forensic blood typing.

Antenatal forensic contexts: Non-invasive prenatal testing (NIPT) is the major clinical application of cfDNA: fetal cfDNA constitutes approximately 10-15% of maternal plasma cfDNA from roughly 10 weeks of gestation and can be typed at fetal-specific STR loci or subjected to chromosomal aneuploidy analysis (Down syndrome, trisomies 18/13) without invasive amniocentesis. The forensic relevance is in disputed paternity cases during pregnancy: a forensic laboratory can determine whether a fetus's paternal half of the genome matches a specific alleged father using cfDNA isolated from a maternal blood draw, without amniocentesis or chorionic villus sampling. Commercial NIPT platforms available in this context include Illumina's VeriSeq NIPT Solution v2 (CE-marked in the EU), Roche's Harmony test, and several Indian platforms (LifeCell and Narayana Health have offered NIPT since approximately 2016). In the US, the American College of Obstetricians and Gynecologists (ACOG) position on forensic NIPT is that it should follow the same ethical and consent framework as any other forensic DNA test.

cfDNA degradation kinetics: Because cfDNA fragments average 166 bp and lack chromatin protection, they are more sensitive to nuclease activity, UV exposure, and humidity than cellular-source DNA. In a haemolysed blood sample or in plasma left at room temperature for more than 24 hours, cfDNA yield drops substantially. EDTA tubes (preferred for NIPT) inhibit nuclease activity and improve cfDNA yield relative to serum separator tubes; for forensic plasma samples, prompt refrigeration and processing within 4 hours is the standard recommendation in the PREANALYTIX cfDNA preservation guidelines, followed by storage at -80°C for extended intervals.

The HIrisPlex-S system is a 41-SNP DNA panel developed by Erasmus MC (Rotterdam, Netherlands) under the leadership of Manfred Kayser for simultaneous prediction of eye colour, hair colour, and skin colour from a DNA sample. HIrisPlex-S is the third generation of EVC panels from the Erasmus group: IrisPlex (2011, eye colour only, 6 SNPs) was followed by HIrisPlex (2013, eye + hair, 24 SNPs) and then HIrisPlex-S (2018, eye + hair + skin, 41 SNPs, published in Forensic Science International: Genetics Supplement).

The panel types SNPs in genes including HERC2 (the master regulator of OCA2 expression, critical for the blue/brown eye-colour switch), OCA2, MC1R (melanocortin 1 receptor, hair redness), ASIP, TYR, SLC45A2, SLC24A5, and KITLG, among others. For each of the 36 predicted phenotype categories (8 eye-colour categories, 17 hair-colour categories, 11 skin-colour categories on a von Luschan's scale basis), the model outputs a posterior probability. A prediction is reported if the highest-probability category exceeds a specified threshold, typically 0.7, at which level independent validation studies show accuracy above 70% for most categories and above 90% for the discriminative extremes (blue eyes, dark brown eyes, blonde hair, black hair, very light skin, very dark skin).

The recommended wet-laboratory protocol uses the ForenSeq DNA Signature Prep Kit (Verogen MiSeq FGx), which includes the 41 HIrisPlex-S SNPs alongside the STR panel, making simultaneous STR typing and EVC prediction technically possible from a single library preparation. Alternatively, the Precision ID Ancestry Panel (Thermo Fisher Scientific, Ion S5) includes many of the same SNPs and can be used for phenotyping analysis. The Erasmus MC team maintains a free web-based prediction tool at erasmusmc.nl/forensic that accepts typed genotypes and returns probability distributions for all 36 trait categories.

Regulatory status by jurisdiction: As noted above, Germany's 2017 amendment to the Code of Criminal Procedure (StPO § 81e, 81f) explicitly permits EVC analysis (eye colour, hair colour, skin colour, age estimation) and biogeographic ancestry from DNA for use in criminal investigations. Austria and Switzerland have followed with similar statutory provisions. The Netherlands, despite hosting the technology's development, does not yet have an explicit statutory basis for EVC use in criminal investigations; the NFI provides it as an investigative service under an administrative framework. The UK does not have explicit statutory authorisation; the Forensic Science Regulator's office has indicated that validation evidence must be submitted before EVC prediction could be considered admissible. Australia's Model Criminal Law Officers' Committee has recommended caution and noted that EVC prediction evidence is not currently admissible in most state jurisdictions. In the US, phenotype prediction is used as an investigative lead only and is not presented as direct evidence in court; the DOJ 2019 IGG policy does not explicitly cover EVC analysis independent of IGG but the permissive approach to investigative tools generally applies.

The route from a published emerging marker to an accredited forensic method is well-charted but time-consuming. The key steps, as defined by SWGDAM's Validation Guidelines (US) and the ENFSI DNA WG Best Practice Manual (EU), are:

1. Developmental validation: characterise the method's performance (sensitivity, specificity, reproducibility, mixture behaviour, degradation tolerance) in a controlled research setting. For methylation age estimation, this is the equivalent of the Zbiec-Piekarska 2015 ELOVL2 paper and its replication studies.
2. Internal validation: implement the method in a specific laboratory, demonstrate that the laboratory's version of the method performs within the developmental validation parameters, and generate a set of validation samples that will serve as reference cases for future QC.
3. Proficiency testing: demonstrate that examiner performance on unknown samples meets the expected error rate across multiple rounds.
4. Technical note publication and peer review: publish the internal validation results so other laboratories can assess the method.
5. Accreditation body review: the laboratory's accreditation body (UKAS in the UK, NATA in Australia, NABL in India, DAkkS in Germany) conducts a documented technical review of the validation evidence before adding the method to the laboratory's scope of accreditation.

For ELOVL2 methylation age estimation, the Netherlands Forensic Institute and the LKA Berlin (state police forensic lab) have both reached steps 3-4 as of 2023-2024. ENFSI DNA WG's proficiency exercise on methylation age estimation is planned for the 2025 cycle. For HIrisPlex-S, Erasmus MC and NFI are the lead validators; several German LKA labs (Baden-Württemberg, Bavaria) operate HIrisPlex-S under their StPO § 81e statutory framework. For miRNA body-fluid profiling, no laboratory has yet reached accreditation-scope status.

In India, CFSL Hyderabad and CFSL Chandigarh are the most technically advanced state-and-central FSL labs in terms of MPS implementation. NABL accreditation for MPS-based STR typing (ForenSeq) has been in progress at CFSL Chandigarh since approximately 2022. No Indian FSL has published internal validation data for methylation age estimation or HIrisPlex-S phenotyping, though research papers from AIIMS Delhi and NFSU Gandhinagar groups have examined ELOVL2 methylation in Indian population samples, establishing, critically, that the ELOVL2 clock's performance parameters in South Asian populations are within range of the European training cohort (MAD approximately 4-6 years vs the European 3.6 years), which supports future forensic implementation.

1. Sample and protocol selection
   Determine which emerging marker is appropriate for the forensic question: methylation clock for age, miRNA for body-fluid, HIrisPlex-S for EVC. Select the extraction and library-preparation protocol validated for that marker in that sample type (blood, saliva, semen, bone) and confirm that sufficient DNA quantity and quality exist.
2. Bisulphite conversion (for methylation)
   Convert DNA using an EZ DNA Methylation Kit (Zymo Research) or equivalent. Bisulphite conversion deaminates unmethylated cytosines to uracil, allowing differentiation from methylated cytosines by sequencing or pyrosequencing. Note that this step is destructive; reserve an aliquot of unconverted DNA for STR typing.
3. Quantification and amplification
   Quantify bisulphite-converted DNA by a conversion-specific qPCR assay; standard Quantifiler assays underestimate converted DNA. Amplify the target CpG region (e.g., ELOVL2 amplicon) using bisulphite-specific primers. For miRNA: reverse-transcribe total RNA extract and amplify miRNA targets using TaqMan miRNA assays.
4. Measurement
   For ELOVL2: pyrosequence the amplicon on a PyroMark Q24 or equivalent; calculate per-site methylation percentage. For multi-CpG clocks: hybridise to Illumina EPIC array and export beta-value matrix. For HIrisPlex-S: type 41 SNPs by ForenSeq MPS or Ion S5 Precision ID; export genotype calls.
5. Prediction and interpretation
   Apply the validated regression model: Zbiec-Piekarska ELOVL2 regression for age; HIrisPlex-S tool at Erasmus MC for EVC. Report the point estimate (age) or probability distribution (EVC) with the validated confidence interval or threshold. State clearly that the result is a prediction, not an identification.
6. Expert report
   State the method used, its validation basis, the laboratory's internal validation results, the population reference data, the result, and the limitations. For age estimation, state the MAD for the method in the relevant tissue type. For EVC, state the accuracy rates for each predicted trait and the threshold applied. Avoid absolute language.