Ancient DNA

Ancient DNA (aDNA) refers to trace, post-mortem degraded DNA molecules extracted from subfossil bone, dental cementum, hair shafts, coprolites, permafrost cores, and stratified archaeological cave sediments (sedimentary ancient DNA, or sedaDNA). Pioneered in 1984 when Russell Higuchi and Allan Wilson sequenced a 229-base-pair mitochondrial fragment from a museum skin of the extinct quagga (Equus quagga quagga), and revolutionized over the subsequent three decades by Svante Pääbo (recipient of the 2022 Nobel Prize in Physiology or Medicine), paleogenomics has progressed from PCR amplification of short mitochondrial loci to high-coverage whole-genome sequencing of archaic hominins, Pleistocene megafauna, and historical pathogens.

Once an organism dies, endogenous nucleases and invading soil microbes cleave chromosomal DNA until desiccation, freezing, high salinity, or adsorption onto mineral matrices arrests enzymatic activity. Thereafter, spontaneous chemical reactions—chiefly hydrolytic depurination and hydrolytic cytosine deamination—progressively fragment the remaining strands to median lengths of 30 to 70 base pairs (bp) and introduce characteristic post-mortem nucleotide misincorporations at fragment termini. While these chemical lesions limit the survival horizon of authentic ancient DNA to roughly 1.0 to 2.0 million years under optimal permafrost conditions (beyond which paleoproteomics of tooth-enamel proteins is required), the resulting terminal damage signatures serve as the primary bioinformatic criterion for authenticating ancient reads against modern human or microbial contamination.

Hydrolytic Depurination, Strand Cleavage, and Cytosine Deamination (C → T)

After post-mortem autolysis ceases, non-enzymatic chemical decay proceeds at rates governed by an Arrhenius exponential relationship k = A exp(−E_a / RT) dependent on ambient temperature T, water activity, pH, and redox state:

1. Hydrolytic Depurination and β-Elimination (Strand Fragmentation): Water molecules attack the N-glycosidic bond linking purine bases (adenine and guanine, which are ~20-fold more labile than pyrimidines) to the deoxyribose sugar ring, releasing the free purine base and creating an abasic (apurinic, AP) site. The aldehyde-open form of the abasic deoxyribose rapidly undergoes β-elimination cleavage of the 3′ phosphodiester backbone, generating single-strand nicks with a high enrichment of purines (A and G) immediately upstream (5′) of the strand break. Accumulation of single-strand nicks on opposite strands produces ultra-short double-stranded fragments with jagged single-stranded 5′ and 3′ overhangs (typically 1–6 nucleotides long).
2. Hydrolytic Cytosine Deamination (Terminal C → T Misincorporations): Within these exposed single-stranded 5′ and 3′ overhangs (where hydrogen-bonded base pairing no longer shields the pyrimidine ring from water), cytosine (C) undergoes hydrolytic loss of its exocyclic 4-amino group (–NH₂) to form uracil (U), while 5-methylcytosine (5mC) deaminates to thymine (T) at rates roughly two orders of magnitude faster than in intact double-stranded DNA. When a standard DNA polymerase copies uracil during library amplification, it pairs U with adenine (A), yielding an apparent cytosine-to-thymine (C → T) transition at the 5′ end of the sequenced read (and a complementary guanine-to-adenine (G → A) transition at the 3′ end in double-stranded libraries). Plotting the position-specific C → T substitution frequency along mapped reads produces a diagnostic exponential decay curve (often reaching 15%–45% at the terminal nucleotide and dropping to baseline by positions 10–15) monitored via tools such as mapDamage2 and PMDtools.

Mineral Adsorption in Petrous Bone and Cave Sediments (sedaDNA)

The preservation of endogenous DNA across tens of thousands of years depends on micro-anatomical and geochemical shielding from circulating groundwater and microbial colonization:

• Osteological Preservation in the Petrous Pyramid: In vertebrate skeletons, the cochlear region of the petrous part of the temporal bone (the densest bone in the mammalian body, with mineral density ~2.1 g/cm³ and minimal vascular remodeling) preserves up to 4-fold to 16-fold higher endogenous DNA fractions than cortical limb bones or ribs (Pinhasi et al., 2015), alongside the acellular cementum layer of tooth roots. Negatively charged phosphate groups on the DNA backbone bind electrostatically to positively charged calcium (Ca²⁺) ions on the crystal faces of bioapatite (carbonated hydroxyapatite, Ca₁₀(PO₄)₆(OH)₂), stabilizing the helix against hydrolysis.
• Sedimentary Ancient DNA (sedaDNA) Adsorption: Even in archaeological cave strata completely devoid of macroscopic skeletal fossils, shed skin cells, hair, feces, and urine release extracellular DNA that adsorbs onto phyllosilicate clay minerals (such as smectite, montmorillonite, and illite), quartz, and humic colloids via divalent cation bridging (Ca²⁺ and Mg²⁺). Mineral-bound sedaDNA experiences an activation energy barrier against nuclease digestion, allowing stratified sedimentary rock and unconsolidated cave silts to preserve time-resolved genetic archives of Pleistocene fauna and hominins.

Single-Stranded Library Preparation, UDG Treatment, and Hybridization Capture

Because environmental microbes (bacteria, fungi, and soil archaea) colonize buried remains after death, endogenous host DNA often constitutes less than 0.1% to 5% of the total extracted DNA, while even trace skin-cell or aerosol contamination from excavation or laboratory personnel can overwhelm degraded ancient templates. Modern paleogenomic facilities operate under positive-pressure cleanroom protocols (bleach decontamination, UV irradiation, and negative extraction blanks) paired with three specialized molecular innovations:

1. Single-Stranded Library Preparation: Conventional double-stranded library prep repairs overhangs with T4 DNA polymerase and ligates adapters to blunt double-stranded ends, losing any molecule that carries single-strand nicks on both strands or is shorter than ~40 bp. Developed by Matthias Meyer and Marie-Theres Gansauge (2012, 2013), single-stranded library preparation heat-denatures the extract into individual single strands and uses CircLigase II (or splint ligation in ssDNA2.0) to attach a biotinylated adapter directly to the 3′ end of every single strand, increasing template recovery 2-fold to 10-fold in heavily degraded Middle Pleistocene specimens (such as the ~430,000-year-old Sima de los Huesos hominins).
2. Uracil-DNA Glycosylase (UDG) Repair: While terminal C → T damage is essential for authentication, high internal uracil rates degrade genotype calling accuracy and sequence alignment. Treating extracts with E. coli Uracil-DNA Glycosylase (UDG) and Endonuclease VIII (USER enzyme) excises uracil bases and cleaves the resulting abasic sites; partial UDG ("UDG-half") protocols retain terminal deamination on the outermost 1–2 nucleotides for authentication while scrubbing uracils from the interior of each fragment.
3. In-Solution Hybridization Capture: When shotgun sequencing is uneconomical due to < 1% endogenous content, libraries are hybridized against synthetic biotinylated oligonucleotide baits—such as mitochondrial genome baits or the 1240k SNP capture array targeting ~1.24 million informative nuclear human polymorphisms—and pulled down on streptavidin-coated magnetic beads, enriching target sequences by orders of magnitude.

Archaic Hominin Admixture and Cave Sediment Paleogenomics

Ancient DNA sequencing has reshaped the evolutionary model of the genus Homo and Late Quaternary ecosystems:

• Neanderthals, Denisovans, and Reticulate Admixture: Sequencing the draft and high-coverage nuclear genomes of Neanderthals (Homo neanderthalensis, Green et al., 2010; Prüfer et al., 2014) demonstrated that non-African modern human populations carry ~1.5% to 2.1% Neanderthal-derived nuclear ancestry resulting from interbreeding ~50,000–60,000 years ago. In 2010, sequencing a single juvenile distal manual phalanx from Denisova Cave in the Altai Mountains revealed a previously unknown archaic sister group to Neanderthals—the Denisovans (Reich et al., 2010; Meyer et al., 2012)—who contributed ~3% to 6% nuclear ancestry to Indigenous Papuan, Australian, and Melanesian populations and transferred the high-altitude hypoxia-adaptive EPAS1 haplotype to modern Tibetans. Direct proof of contemporaneous contact emerged from Denisova 11 ("Denny"), a ~90,000-year-old bone fragment identified via paleogenomics as the first-generation (F₁) female offspring of a Neanderthal mother and a Denisovan father (Slon et al., 2018).
• Cave Sediment Ancient DNA (Denisova, Baishiya Karst, Satsurblia, and El Mirón): By applying mitochondrial and nuclear hybridization capture to micro-stratigraphic cave sediments (Slon et al., 2017; Vernot et al., 2021), researchers have recovered Neanderthal and Denisovan nuclear genomes from strata at Denisova Cave, Chagyrskaya Cave, Baishiya Karst Cave (Tibetan Plateau), and Galería de las Estatuas (Spain) without requiring physical hominin bones. Similarly, Upper Paleolithic sediments from Satsurblia Cave (Georgia, ~25,000 cal BP) yielded environmental genomes of a basal West Eurasian human lineage Alongside extinct Eurasian wolf and bison lineages, while Magdalenian human remains from El Mirón Cave (Cantabria, Spain, ~18,700 cal BP) anchored the post-Last Glacial Maximum demographic re-expansion across western Europe. At the extreme temporal frontier, permafrost-preserved mammoth molars from northeastern Siberia (~1.2–1.65 Ma, van der Valk et al., 2021) and Kap København marine-delta sediments in northern Greenland (~2.0 Ma, Kjær et al., 2022) have extended environmental DNA recovery into the Early Pleistocene.

Why can ancient DNA survive for 1–2 million years in permafrost but never in Mesozoic dinosaur bones?

Hydrolytic depurination cleaves DNA backbones at a rate controlled by the Arrhenius thermal activation relation. Even at −5 °C to −15 °C in continuous polar permafrost, kinetic measurements on dated subfossils show an average DNA half-life that reduces all templates to single-nucleotide monomers (< 20–25 bp, below unique mappability thresholds) within ~1.5 to 2.5 million years. Consequently, 66-million-year-old Mesozoic Dinosauria fossils sit more than 30-fold beyond the thermodynamic survival limit of nucleic acids; molecular systematics on Mesozoic fossils relies instead on paleoproteomics or morphological matrices.

How do bioinformaticians prove that sequenced ancient human reads are not modern lab contamination?

Modern contaminating DNA consists of long fragments (> 150 bp) with near-zero terminal cytosine deamination (< 0.5% C → T). Authentic ancient human reads are verified via three orthogonal tests: (1) an unimodal short fragment length distribution peaking between 35 and 65 bp; (2) a terminal C → T substitution rate ≥ 10% at the 5′ read end; and (3) quantitative contamination estimates (< 1%–2%) calculated from polymorphic sites on the maternally inherited mitochondrial genome (schmutzi / contamMix) or the hemizygous X chromosome in male individuals (ANGSD).

Why does 5-methylcytosine (5mC) deamination allow epigenetic reconstruction in ancient genomes?

Unmethylated cytosine (C) deaminates to uracil (U), which is excised and removed when libraries are treated with Uracil-DNA Glycosylase (UDG). By contrast, methylated 5-methylcytosine (5mC) at vertebrate CpG dinucleotides deaminates directly into thymine (T), a canonical DNA base that is untouched by UDG. Consequently, residual CpG → TpG transitions retained in UDG-treated high-coverage ancient genomes (such as the Altai Neanderthal and Denisovan genomes) map the in vivo CpG methylation landscape and tissue-specific gene regulation of extinct organisms.

1. Briggs, A. W., Stenzel, U., Johnson, P. L. F., Green, R. E., Kelso, J., Prüfer, K., Meyer, M., Krause, J., Ronan, M. T., Lachmann, M., & Pääbo, S. (2007). "Patterns of damage in genomic DNA sequences from a Neandertal." Proceedings of the National Academy of Sciences, 104(37), 14616–14621. https://doi.org/10.1073/pnas.0704665104
2. Meyer, M., Kircher, M., Gansauge, M.-T., Li, H., Racimo, F., Mallick, S., Schraiber, J. G., Jay, F., Prüfer, K., de Filippo, C., et al. (2012). "A high-coverage genome sequence from an archaic Denisovan individual." Science, 338(6104), 222–226. https://doi.org/10.1126/science.1224344
3. Pinhasi, R., Fernandes, D., Sirak, K., Novak, M., Connell, S., Alpaslan-Roodenberg, S., Gerritsen, F., Moiseyev, V., Gromov, A., Raczky, P., et al. (2015). "Optimal ancient DNA yields from the inner ear part of the human petrous bone." PLOS ONE, 10(6), e0129102. https://doi.org/10.1371/journal.pone.0129102
4. Slon, V., Hopfe, C., Weiß, C. L., Mafessoni, F., de la Rasilla, M., Lalueza-Fox, C., Rosas, A., Soressi, M., Knul, M. V., Miller, R., et al. (2017). "Neandertal and Denisovan DNA from Pleistocene sediments." Science, 356(6338), 605–608. https://doi.org/10.1126/science.aam9695