Paleoproteomics and Ancient Protein Sequencing

Paleoproteomics is the biomolecular discipline that extracts, authenticates, and sequences degraded endogenous proteins from archaeological and paleontological remains—primarily mineralized bone, tooth dentin, dental enamel, and avian eggshell—using liquid chromatography coupled to high-resolution tandem mass-spectrometry (LC-MS/MS). Because the phosphodiester backbone of ancient-dna (aDNA) undergoes rapid spontaneous hydrolytic depurination after cell death, recoverable DNA fragments (>30–35 base pairs) rarely survive beyond 400,000 to 500,000 years in temperate latitudes or 1.0 to 2.0 million years even in permanently frozen Arctic permafrost. By contrast, peptide bonds linking amino acids are kinetically more stable against hydrolysis, and extracellular matrix proteins tightly bound to biomineral surfaces can persist for 2 to over 6.5 million years, extending molecular phylogenetics deep into the Early Pleistocene, Pliocene, and late Miocene epochs.

Through both rapid peptide mass fingerprinting (ZooMS, Zooarchaeology by Mass Spectrometry) of type I collagen and shotgun de novo sequencing of the dental enamelome, paleoproteomics has resolved long-standing evolutionary questions across vertebrate paleontology and paleoanthropology. Landmark achievements include identifying the 160,000-year-old Xiahe mandible from the Tibetan Plateau as Denisovan in the complete absence of preserved DNA, recovering biological sex and phylogenetic trees from 1.77-million-year-old Homo erectus and Stephanorhinus fossils at Dmanisi, Georgia, and sequencing 3.5- to 6.5-million-year-old camelid and ostrich eggshell proteomes from the Canadian High Arctic and East African Rift Valley.

Mineral-Binding Thermodynamics and Enamel Closed Systems

The extraordinary geological longevity of ancient proteins is governed by their electrostatic binding affinity to biogenic hydroxyapatite (Ca₁₀(PO₄)₆(OH)₂) in vertebrate bone and teeth or calcite (CaCO₃) in eggshell. In cortical bone and dentin, fibrillar Type I Collagen (a heterotrimer composed of two α1 chains encoded by COL1A1 and one α2 chain encoded by COL1A2, forming a tight right-handed Gly-X-Y triple helix stabilized by hydroxyproline) accounts for ~90% of the organic matrix. Intercalated inside mineral nanopores (<5 nm diameter), collagen fibrils are sterically shielded from microbial collagenases and bulk groundwater flow, surviving up to ~1.5–3.5 Ma in cold caves or permafrost, alongside non-collagenous bone proteins such as osteocalcin (BGLAP), α2-HS-glycoprotein (fetuin-A, AHSG), and biglycan (BGN).

Even greater temporal survival is achieved within dental enamel, the hardest and most heavily mineralized tissue in vertebrates (96% to 98% crystalline hydroxyapatite by weight, with virtually zero open porosity after crown maturation). During amelogenesis, secretory ameloblasts deposit a specialized suite of extracellular matrix proteins that guide crystal elongation before being enzymatically cleaved by MMP20 and KLK4:

• Amelogenin (AMELX on the X chromosome; AMELY on the Y chromosome): Comprises ~90% of developing enamel matrix; because human and mammalian AMELY carries diagnostic sequence deletions and amino acid substitutions relative to AMELX, detecting AMELY-specific peptides provides an unambiguous molecular assay for biological sex (XY male vs. XX female) in fossil hominins.
• Enamelin (ENAM), Ameloblastin (AMBN), and Enamel-Specific Proteases (MMP20, KLK4): Evolve under weaker purifying selection than highly conserved collagen, accumulating ~3- to 5-fold higher densities of phylogenetically informative Single Amino Acid Polymorphisms (SAAPs) per 100 residues.

ZooMS Peptide Mass Fingerprinting vs. LC-MS/MS

Paleoproteomic investigations employ two complementary proteomics platforms depending on sample abundance and taxonomic goals:

1. ZooMS (Zooarchaeology by Mass Spectrometry): A high-throughput, minimally destructive screening method (requiring 5–20 mg of bone or non-destructive extraction on PVC eraser crumbs via triboelectric friction) that digests extracted collagen with trypsin and measures the mass-to-charge (m/z) profile of the resulting peptides using MALDI-TOF (Matrix-Assisted Laser Desorption/Ionization Time-of-Flight) mass spectrometry. By comparing 7 to 12 diagnostic collagen peptide marker masses against reference libraries, ZooMS rapidly classifies thousands of morphologically unidentifiable Pleistocene bone splinters to family or genus level, enabling targeted discovery of hominin fragments (such as Denisova 11, the Neanderthal–Denisovan F1 hybrid "Denny").
2. Shotgun Nano-LC-MS/MS: For deep phylogenetic reconstruction and ancient enamel analysis, demineralized peptides (often extracted without enzymatic digestion because endogenous diagenetic hydrolysis has already cleaved the backbone) are separated by reverse-phase nano-liquid chromatography and ionized via electrospray into an Orbitrap or TIMS-TOF tandem mass spectrometer. Precursor peptide ions (MS1) are isolated and fragmented by higher-energy collisional dissociation (HCD) to generate b- and y-ion fragment ladders (MS2) at <5 ppm mass accuracy, allowing exact amino acid sequence reconstruction and discovery of novel evolutionary variants.

Authentication via Deamidation Kinetics

Because modern human skin keratins, laboratory reagents (porcine trypsin, bovine serum albumin), and soil fungi can contaminate fossil extracts, ancient peptides must be authenticated by quantifying time-dependent diagenetic post-translational modifications (PTMs). The primary chemical clock in paleoproteomics is the non-enzymatic hydrolytic deamidation of the neutral amide side chains of asparagine (Asn, N) and glutamine (Gln, Q) into negatively charged aspartic acid (Asp, D) and glutamic acid (Glu, E) (plus isoaspartate and γ-glutamate isomers), which adds a precise mass shift of +0.9840 Da (-NH₂ → -OH) per modified residue.

In aqueous or mineral environments, Asn deamidates rapidly via a five-membered succinimide ring intermediate (half-life t_1/2 of days to centuries), whereas Gln deamidates 10 to 100 times more slowly via a less favorable six-membered glutarimide intermediate (t_1/2 of 10³ to 10⁶ years, modulated by temperature and primary sequence neighbor effects). Authentic Early and Middle Pleistocene fossil proteins exhibit near-complete Asn deamidation (80%–100%) and advanced Gln deamidation (40%–95%), partnered with progressively shorter average peptide chain lengths (8–15 amino acids) driven by terminal hydrolysis and oxidative conversion of methionine and tryptophan, cleanly separating endogenous fossil proteomes from intact modern contaminants (<5% Gln deamidation).

Denisovan Mandibles and Early Pleistocene Hominins

Paleoproteomics has transformed Middle and Early Pleistocene paleoanthropology in warm and subtropical regions where ancient-dna fails to survive. In 2019, analysis of degraded dentine collagen from the Xiahe mandible—a 160,000-year-old hominin jaw discovered in Baishiya Karst Cave on the Tibetan Plateau at 3,280 meters elevation—identified a derived single amino acid substitution (COL1A2 R996K) shared uniquely with the Altai Denisovan genome and absent in modern humans and Neanderthal lineages, proving that Denisovans occupied high-altitude Tibet long before Homo sapiens. Similarly, sequencing of the dental enamelome from Dmanisi (Georgia, 1.77 Ma) and South African Paranthropus robustus (~2.0 Ma) and Asian Gigantopithecus blacki (~1.9 Ma in subtropical Guangxi) resolved the sister-group relationship between Gigantopithecus and modern orangutans (Pongo) and revealed intra-taxon genetic variation within early African hominins.

Pliocene Mammals and Darwin's South American Ungulates

Beyond hominins, shotgun collagen sequencing resolved one of Charles Darwin's oldest evolutionary puzzles: the phylogenetic placement of South America's extinct native ungulates (Meridiungulata). Sequencing Late Pleistocene collagen from Toxodon (Notoungulata) and Macrauchenia (Litopterna) demonstrated that rather than being related to African elephants/manatees (Afrotheria) or rodents, both orders form a monophyletic sister clade (Panperissodactyla) to Perissodactyla (horses, tapirs, and rhinoceroses). Pushing deeper into the Neogene, enamel proteomes recovered from 1.77 Ma Stephanorhinus teeth at Dmanisi and 3.5 Ma Paracamelus fossils on Ellesmere Island demonstrated that the extinct Eurasian rhinoceros Stephanorhinus forms a clade with the woolly rhinoceros (Coelodonta), invalidating earlier morphological hypotheses and confirming that paleoproteomics can reconstruct molecular phylogenies across the Pliocene–Pleistocene transition.

Why do ancient proteins survive millions of years longer than ancient DNA?

Ancient DNA degrades rapidly because water molecules attack purine bases (adenine and guanine), causing depurination and strand breakage within 100,000 to 1,000,000 years. By contrast, peptide bonds linking amino acids have a higher activation energy barrier against hydrolysis, and extracellular proteins like collagen, amelogenin, and struthiocalcin have negatively charged acidic domains that bind tightly to calcium ions on hydroxyapatite or calcite mineral crystals, locking them inside a water-excluding mineral cage for 2 to 6.5 million years.

How much phylogenetic information does a paleoproteome contain compared to a genome?

A complete ancient enamel or bone proteome typically recovers 500 to 3,000 amino acid positions across 6 to 15 proteins—equivalent to roughly 1,500 to 9,000 base pairs of coding DNA, or less than 0.0003% of a 3-gigabase mammalian genome. Because protein sequences evolve slowly under functional constraint, paleoproteomics cannot resolve fine-scale cousin marriages or recent population splits like ancient DNA, but it preserves enough diagnostic amino acid substitutions to resolve genus- and species-level evolutionary trees across millions of years.

How can paleoproteomics determine the biological sex of a 2-million-year-old fossil tooth?

Mammalian tooth enamel contains amelogenin proteins encoded by two sex-linked genes: AMELX on the X chromosome and AMELY on the Y chromosome. Because AMELY has diverged in sequence from AMELX, the presence of AMELY-specific peptides (containing methionine at position 59 in hominins) alongside AMELX peptides in a mass-spectrometry scan identifies an XY male, whereas high-coverage detection of AMELX peptides alone indicates an XX female.

1. Cappellini, E., Prohaska, A., Racimo, F., Welker, F., Pedersen, M. W., Allentoft, M. E., et al. (2018). "Ancient biomolecules and evolutionary inference." Annual Review of Biochemistry, 87, 1029–1060. https://doi.org/10.1146/annurev-biochem-062917-012002
2. Cappellini, E., Welker, F., Pandolfi, L., Ramos-Madrigal, J., Samodova, D., Rüther, P. L., et al. (2019). "Early Pleistocene enamel proteome from Dmanisi resolves Stephanorhinus phylogeny." Nature, 574(7776), 103–107. https://doi.org/10.1038/s41586-019-1555-y
3. Chen, F., Welker, F., Shen, C. C., Bailey, S. E., Bergmann, I., Davis, S., et al. (2019). "A late Middle Pleistocene Denisovan mandible from the Tibetan Plateau." Nature, 569(7756), 409–412. https://doi.org/10.1038/s41586-019-1139-x
4. Welker, F., Collins, M. J., Thomas, J. A., Wadsley, M., Brace, S., Cappellini, E., et al. (2015). "Ancient proteins resolve the evolutionary history of Darwin's South American ungulates." Nature, 522(7554), 81–84. https://doi.org/10.1038/nature14249