Hybridisation, Southern, Northern, Western Blots and Microarrays

Two single-stranded nucleic acid molecules with complementary base sequences will form a double-stranded hybrid under conditions that allow hydrogen bonds to form between bases (A-T, G-C for DNA; A-U, G-C for RNA). The stability of the hybrid depends on four variables: (1) the number of complementary base pairs (longer hybrids are more stable); (2) G+C content (G-C pairs form three hydrogen bonds versus two for A-T, increasing melting temperature by approximately 0.41°C per G-C pair at standard conditions); (3) ionic strength (Na+ ions shield the negatively charged phosphate backbone, reducing electrostatic repulsion between strands); and (4) temperature (higher temperatures increase thermal energy, competing with hydrogen-bond formation).

The melting temperature (Tm) of a DNA duplex is the temperature at which 50% of the molecules are double-stranded. For oligonucleotide probes, a useful approximation is the Wallace rule: Tm (°C) = 2 × (A+T) + 4 × (G+C). For longer probes (>50 bp), the more accurate formula includes a salt-correction term. In hybridisation experiments, a washing step at 5-12°C below the Tm (the "stringency wash") removes partially matched or non-specifically bound probe molecules, while perfectly matched hybrids survive.

The term "stringency" is central to all hybridisation assays. High stringency (high temperature + low salt in the wash) retains only perfect or near-perfect complements. Low stringency tolerates more mismatches, broadening the specificity to related sequences. Forensic probe-based assays (Jeffreys' original probes, commercial dot-blot kits for species identification) are all high-stringency assays: a probe designed against a human-specific Alu sequence will not hybridise to dog or cat DNA under high-stringency conditions, which is how it identifies human stains in mixed crime-scene samples.

E.M. Southern published "Detection of specific sequences among DNA fragments separated by gel electrophoresis" in Journal of Molecular Biology in 1975. The method's conceptual contribution was that electrophoretic size separation and sequence-specific detection could be combined: separate fragments by size on a gel, then transfer the pattern to a membrane that can be probed without disturbing the spatial arrangement.

The original transfer method was capillary action: the agarose gel (containing separated, denatured single-stranded DNA fragments) was placed face-down on a wick submerged in 20× SSC (saline-sodium citrate) buffer; a nitrocellulose membrane was laid on top of the gel; absorbent paper towels were stacked above; the upward capillary flow of buffer carried DNA out of the gel and deposited it on the membrane in a spatial distribution matching the gel pattern. Nitrocellulose was later replaced by positively charged nylon membranes (Hybond-N+, GeneScreen Plus), which bind DNA more strongly and can be reprobed multiple times.

Modern Southern blotting can also use vacuum transfer (much faster, 30-60 minutes versus overnight capillary transfer) or electroblotting. After transfer, the DNA is cross-linked to the membrane by UV (photoactivated cross-linking of thymine to membrane amines) or heat (80°C for 2 hours in an oven for nitrocellulose). The membrane is then prehybridised in a blocking buffer (Denhardt's solution, salmon sperm DNA, or formamide-SSC mixtures) to saturate non-specific binding sites, followed by hybridisation with the labelled probe in the same buffer at 60-68°C overnight.

Detection in the original Jeffreys work used a 32P-radiolabelled probe; the membrane was exposed to X-ray film (autoradiography) for 1-5 days, developing bands at positions where the probe hybridised. Modern Southern blotting uses digoxigenin (DIG) or biotin-labelled probes detected with enzyme-conjugated antibodies (alkaline phosphatase or horseradish peroxidase) and chemiluminescent substrates, which gives equivalent sensitivity without radioactivity. This is the format used in the Quantiblot human DNA dot-blot kit (Applied Biosystems) for forensic sample quantification, which applies Southern-blot-style probe hybridisation to membrane-bound samples before STR kit setup.

In the US, RFLP Southern blots on five to seven loci were the basis for criminal convictions from the late 1980s through the early 2000s. The FBI's RFLP typing protocol used HaeIII digestion, agarose gel separation, capillary transfer to Hybond-N+ membranes, and hybridisation with locus-specific probes (D1S7, D2S44, D4S139, D10S28, D17S79 and others). Each blot required 50-150 ng of high-molecular-weight genomic DNA, a requirement that excluded degraded samples and led to the eventual adoption of PCR-based STR typing for casework. Legacy RFLP evidence in US federal convictions is still occasionally reviewed in post-conviction DNA proceedings, requiring forensic scientists to explain the Southern blot methodology to courts.

In the UK, the Forensic Science Service used a modified Jeffreys multi-locus probe system on HinfI-digested DNA before transitioning to single-locus probes and eventually to STR typing. In India, RFLP-based evidence was admitted in civil paternity disputes and a small number of criminal cases in the early 1990s before the adoption of STR typing at CFSL Hyderabad and state FSLs.

Northern blotting (the name coined as a deliberate play on Southern's surname, with no eponym intended) was developed by Alwine, Kemp, and Stark at Stanford in 1977. The method transfers RNA from a denaturing agarose gel (run in formaldehyde or glyoxal to prevent secondary-structure formation) to a nylon membrane, then hybridises with a DNA or RNA probe to detect a specific transcript. Band position reports transcript size in kilobases; band intensity estimates relative abundance.

In forensic contexts, Northern blotting is not used in routine casework; the technique requires fresh or well-preserved RNA, which degrades rapidly at room temperature, making it unsuitable for typical crime-scene samples. However, it underpins the conceptual and methodological basis for mRNA-based body-fluid identification, a technique validated at the Netherlands Forensic Institute (NFI) and at the Armed Forces DNA Identification Laboratory (AFDIL) in the US between 2005 and 2012.

The mRNA profiling approach (Haas, Kiesler, Stangier, and colleagues, published in Forensic Science International Genetics 2009-2012) uses RT-PCR (reverse transcription PCR) rather than Northern blotting to detect tissue-specific transcripts: blood-specific mRNAs (HBB, ALAS2), saliva-specific (STATH, PRB3, PRH1), semen-specific (SEMG1, KLK3), vaginal secretion-specific (MUC4, CYP2B7P1), and skin epithelial (LCE1C, LCE1D) markers. Presumptive and confirmatory tests for these same fluids using antibody and colour-reaction methods are described in body-fluid identification: presumptive and confirmatory tests. The assay is multiplexed and run on the same capillary electrophoresis platform as STR profiling. This is the descendant of the Northern blot concept: both ask the same question (which mRNA is present?) by different mechanisms.

The RSID (Rapid Stain Identification) line of immunochromatographic strips and the ABAcard (abacus diagnostics) lateral-flow assay for PSA/p30 (semen identification) are protein-target analogues: they detect specific proteins by antibody binding rather than nucleic acid hybridisation, which connects to the Western blot principle. In the UK, the Forensic Science Service incorporated mRNA profiling into its body-fluid confirmation workflow before closure; the Netherlands Forensic Institute and the German BKA continue to use RT-PCR-based mRNA profiling as a confirmatory body-fluid identification step.

Western blotting (also called protein immunoblotting) was developed by Towbin, Staehelin, and Gordon in 1979 at the Friedrich Miescher Institute, Basel, Switzerland. Proteins are denatured and given a uniform negative charge by heating in sodium dodecyl sulphate (SDS) sample buffer with 2-mercaptoethanol; separated by molecular weight on an SDS-polyacrylamide gel (SDS-PAGE); transferred to a nitrocellulose or PVDF membrane by electrophoresis; blocked to prevent non-specific antibody binding; probed with a primary antibody specific to the target protein; detected with a secondary antibody conjugated to a reporter enzyme (HRP or alkaline phosphatase) and a chemiluminescent or colorimetric substrate.

The result is a band at a position corresponding to the molecular weight of the target protein, visible because the antibody binds specifically and the enzyme produces a detectable signal.

In forensic applications, Western blotting is used in several contexts. Confirmatory identification of protein markers in body fluids is one: PSA (prostate-specific antigen, also called p30) detection in semen stains can be confirmed by Western blot when immunochromatographic strip results are ambiguous; the band at approximately 33 kDa confirms PSA identity. Post-mortem toxicology and protein biomarkers represent another application: Western blot has been used in research settings to detect post-translational modifications associated with drug toxicity and to characterise changes in cardiac proteins in sudden-death cases, though this remains a research tool rather than a routine casework method. Species identification in food fraud and wildlife forensics is a third: Western blot with species-specific antibodies can distinguish beef, horse, pig, and poultry proteins in processed products, complementing PCR-based DNA approaches; the UK's Food Standards Agency and the EU's food safety networks used Western blot alongside PCR during the 2013 horse-meat adulteration scandal that implicated supply chains across Ireland, the UK, France, Sweden, and Romania.

A specialised application is the identification of prion proteins in animals and humans. Western blot is the confirmatory diagnostic test for bovine spongiform encephalopathy (BSE) in cattle and Creutzfeldt-Jakob disease (CJD) in humans, where the abnormal PrP(Sc) isoform migrates differently from normal PrP(C) on SDS-PAGE after proteinase K treatment. This has legal implications in contaminated-blood-product litigation and in food-safety prosecutions in the UK, Ireland, France, and Canada.

A DNA microarray (gene chip) immobilises thousands to millions of short oligonucleotide probes at defined positions on a glass or silicon substrate. A labelled target (cDNA synthesised from RNA, genomic DNA, or PCR products) is hybridised across the array surface under controlled stringency. Where the target sequence is complementary to a probe, a fluorescent signal is detected at that position. The intensity of the signal reflects the abundance of the complementary sequence in the target. Because each probe position is spatially registered, the identity of the complementary sequence is determined by geography: position (x, y) = known probe sequence.

The two major array platforms relevant to forensic biotechnology are the Affymetrix GeneChip (now Thermo Fisher Scientific) and the Illumina BeadChip (BeadArray). Affymetrix chips use photolithographic synthesis of oligonucleotides directly on the surface; the 500K human SNP genotyping array contains 500,000 SNP probes in a 1.28 cm × 1.28 cm area. Illumina BeadChip arrays (including the CytoSNP-850K and the Global Screening Array used in forensic genetic genealogy research) use silica beads (3 µm diameter) deposited into etched wells on a glass substrate; each bead type carries a specific 50-mer probe, and self-assembly of the beads at random positions is decoded by reading barcode hybridisation patterns.

HLA typing was the first forensic/medical application of microarray hybridisation to reach widespread adoption. Human leukocyte antigen (HLA) typing identifies the polymorphic MHC class I and class II alleles that govern immune compatibility. Traditional serological typing was largely replaced from the mid-1990s by PCR-SSP (sequence-specific priming) and PCR-SSOP (sequence-specific oligonucleotide probe hybridisation on membrane strips, a direct descendant of the Southern blot), and subsequently by microarray-based methods (LABType SSO using Luminex bead arrays, now widely used in US and European tissue-typing labs). HLA typing has forensic relevance in parentage testing (HLA provides approximately 8 bits of exclusion power as a non-autosomal marker), in disaster victim identification where DNA is not available or insufficient, and in transplant-related homicide investigations. The Luminex bead-array system used for HLA typing is in routine clinical use in transplant centres in the US (UNOS network), the UK (NHS Blood and Transplant), India (NOTTO), and Australia (AOTP).

SNP genotyping chips in a forensic context were developed primarily as research tools before NGS made whole-genome SNP typing practical. The Affymetrix 500K SNP array was used in the initial human HapMap Project phase II (2005-2007) to characterise linkage disequilibrium patterns and identify ancestry-informative marker (AIM) panels. These AIM panels are the intellectual predecessors of the forensic ancestry-estimation SNP panels (the FORENSeq ANCestry QS panel, the SNPforID 34-plex), which were designed by selecting a subset of the array-characterised SNPs that most efficiently distinguish global ancestral populations.

For forensic phenotyping, the Parabon Snapshot pipeline that generated predicted physical-trait predictions from crime-scene DNA in US and UK cases (2014 onwards) used a BeadChip-derived SNP training set to build the predictive model for eye colour (OCA2, HERC2, SLC24A4), hair colour (MC1R, TPCN2, SLC24A4), and skin colour (SLC45A2, TYR, OCA2) predictions. The Parabon Nanopore DMP pipeline used in the Golden State Killer investigation (People v. DeAngelo, Sacramento Superior Court, 2020) also relied on genome-wide SNP data from Illumina-format BeadChip arrays uploaded to GEDmatch and FamilyTreeDNA for kinship triangulation.

Southern blotting is no longer used in new forensic casework in any major jurisdiction (US, UK, EU, Australia, India). The marker classes it characterised and their inheritance rules are described in polymorphism, heredity and forensic markers. However, it remains forensically relevant because:

• Post-conviction review of RFLP evidence (common in US federal habeas corpus proceedings and UK Criminal Cases Review Commission referrals) requires experts who can explain the Southern blot methodology and assess whether the original analysis was conducted correctly.
• The Quantiblot membrane dot-blot quantification assay (a miniaturised Southern blot variant) was widely used in US forensic labs from 1994 to approximately 2005 as the standard pre-PCR DNA quantification method; legacy cases from that window may involve interpretation of Quantiblot results.
• Southern blotting remains the gold-standard method for confirming restriction fragment polymorphisms in research settings, including in the forensic genetics research literature.

Northern blotting's forensic descendant is RT-PCR-based mRNA body-fluid identification, now validated at the NFI (Netherlands), BKA (Germany), the FBI Laboratory's Chemistry Unit (US), and assessed by ENFSI working groups for adoption across EU member-state forensic laboratories. The ENFSI DNA Working Group published evaluation data on mRNA profiling in 2012.

Western blotting is in active forensic use in species identification, food-fraud investigations, prion confirmation, and some body-fluid confirmation workflows. Its limitation for crime-scene trace analysis is protein degradation; proteins survive poorly on dried stains exposed to humidity and UV, unlike DNA.

Microarrays were the dominant SNP-genotyping technology from approximately 2001 to 2015, when Illumina short-read sequencing became price-competitive. Their legacy is felt in every ancestry-informative and phenotyping SNP panel used in forensic genetics today: those panels were discovered, validated, and characterised using microarray data. The current forensic-grade equivalent is the Verogen MiSeq FGx running the FORENSeq DNA Signature Prep assay, an NGS panel that simultaneously types 27 autosomal STRs, 7 X-STRs, 23 Y-STRs, and 94 iSNPs, all of which were defined using microarray-era SNP discovery data, the full bioinformatics pipeline from FASTQ to reportable allele is described in NGS data analysis: allele calling and variant callers.