Why Can't We Find Dinosaur DNA? Ancient DNA Technology and the Limits of Time

Jurassic Park imagined that humans could recover dinosaur blood from a mosquito sealed in amber and reopen the door to a lost dinosaur world. In reality, that door cannot be opened in that way. Dinosaur bones can lie in rock strata for hundreds of millions of years; teeth can preserve traces of biting and wear; footprints can record the brief moment when an animal crossed a muddy surface. Double-stranded DNA, however, is a polymer made of two long chains of four nucleotides paired together, and once an organism dies, that molecule is no longer maintained. From that moment onward, time, temperature, moisture, microorganisms, and chemical reactions begin to dismantle it. The longer time passes, the more severely DNA degrades. This is why we may still recover DNA from mammoths (Mammuthus), cave bears (Ursus spelaeus), ancient horses, and archaic humans, but not from non-avian dinosaurs. DNA itself has physical and chemical limits on how long it can persist.

DNA（Image source：Forluvoft，CC0 1.0 ）

Even in living cells, DNA is constantly exposed to damage caused by reactive oxygen species, free radicals, hydrolysis, replication errors, ultraviolet light, chemical carcinogens, inflammation, and many other factors. Organisms therefore evolved a wide range of DNA repair proteins that continuously correct DNA damage. In vertebrates, these repair pathways include base excision repair, nucleotide excision repair, mismatch repair, homologous recombination, and non-homologous end joining. Once cells die, repair stops. Enzymes inside and outside the cell, together with microbial activity, begin breaking down tissues, and DNA begins to fragment. The earliest damage often comes from the activity of various nucleases and microorganisms, which cleave DNA molecules and gradually reduce chromosomes once composed of millions to hundreds of millions of nucleotides into many short fragments.

What ultimately makes DNA sequences unreadable is the chemical damage that accumulates after death. One important process is hydrolytic depurination, in which adenine (A) or guanine (G) is lost from the nucleotide unit, leaving behind an abasic site. These sites are prone to further cleavage, making the DNA increasingly fragmented. Another common form of damage is hydrolytic deamination. Cytosine (C), in particular, can deaminate into uracil (U), which is often read as thymine (T) in sequencing data and can therefore cause errors in sequence interpretation. This change occurs especially often near the ends of DNA fragments, where single-stranded overhangs are more vulnerable to chemical modification than double-stranded regions.

Depurination of adenine（Image source：NEUROtiker，CC0 1.0 ）

Deamination of cytosine（Image source：Yikrazuul，CC0 1.0 ）

Non-avian dinosaurs lived during the Mesozoic, more than about 66 million years ago. By contrast, the timescale in which ancient DNA research can be carried out reliably is still mostly limited to the last tens of thousands to several hundreds of thousands of years. Technical advances have certainly pushed this boundary deeper into the past. Researchers have reconstructed paleogenomes from permafrost-preserved mammoths more than 1 million years old, and the oldest isolated DNA currently comes from roughly 2-million-year-old sediments in northern Greenland, where DNA from multiple organisms was preserved together. Yet the gap between 2 million years and 66 million years remains enormous. Even if possible molecules were recovered from such deep time, they would likely be fragmented beyond recognition. Finding dinosaur bones, therefore, does not mean finding dinosaur DNA inside them.

Temporal distribution of ancient DNA studies and paleoclimatic change. The horizontal axis indicates age in million years ago (Ma). Points above the curve show published ancient DNA records, including non-human animal paleogenomes (orange), hominin paleogenomes (blue), and sedimentary ancient DNA (brown). The lower curve is a paleoclimate record based on benthic δ¹⁸O, reflecting cycles of glacial (G) and interglacial (IG) periods. LP: Late Pleistocene（Image source：Dalén L et al. (2023)， CC BY 4.0 ）

The preservation environment also has a major influence on how long DNA can persist. Cold, dry, stable settings with low microbial activity are the most favorable for DNA preservation. Many successful ancient DNA studies have come from high-latitude permafrost, cold caves, or special sedimentary environments, because these conditions slow chemical reactions and restrict microbial decomposition.

Contamination is another practical problem. Ancient DNA samples usually contain only tiny amounts of authentic endogenous DNA. More often, they are filled with DNA from modern environmental microorganisms, along with human DNA introduced during collection, curation, or laboratory handling. When polymerase chain reaction (PCR) is used to amplify very small amounts of ancient DNA, contaminating modern DNA can easily be amplified as well. The resulting sequences may be difficult to assign with confidence, which can seriously interfere with analysis. Ancient DNA research regained reliability only after high-throughput sequencing, clean laboratory procedures, damage-pattern authentication, and bioinformatic analysis became increasingly mature.

The breakthrough of modern paleogenomics lies in learning how to extract information from extremely short, damaged, and mixed molecular fragments. High-throughput sequencing can read enormous numbers of short fragments at once, giving analytical value to ultrashort DNA fragments, about 30 to 50 bp long, that traditional PCR could not amplify effectively. Researchers first extract DNA from bones, teeth, or sediments, then ligate short artificial DNA sequences, known as adapters, to ancient DNA fragments in order to construct a sequencing-ready DNA library. These methods have made paleogenomic studies of Denisovans, Neanderthals (Homo neanderthalensis), ancient horses, mammoths, and other organisms possible, and they have greatly extended the time window of ancient DNA research.

Researchers may also treat samples with enzymes such as uracil DNA glycosylase and endonuclease VIII. These enzymes remove uracils produced by cytosine deamination, reducing analytical errors caused by post-mortem damage. This approach, however, comes with a cost: the treatment can cut the DNA molecule and further fragment DNA that is already short. For this reason, it is mainly suitable for more recent or moderately old samples.

In addition to direct sequencing, researchers have developed enrichment approaches to separate authentic ancient DNA from the vast background of contaminating DNA. A common strategy uses short artificial DNA or RNA molecules as probes to capture specific target sequences, such as mitochondrial genomes, particular nuclear regions, single nucleotide polymorphism sites, exons, or even, with large probe sets, whole nuclear genomes. These methods can save substantial time and reduce the analytical burden caused by contaminating DNA.

The progress of ancient DNA technology is remarkable, but even the most refined techniques are unlikely to reach back to dinosaur DNA. Across tens of millions of years, DNA effectively disappears into the environment. Even so, dinosaur bones, eggshells, footprints, feather impressions, protein residues, and comparative genomic evidence still allow us to approach, step by step, the world in which these animals once lived.

Author: Shui-Ye You

References:

1. Dalén L et al. (2023). Deep-time paleogenomics and the limits of DNA survival. Science.

2. Orlando L et al. (2015). Reconstructing ancient genomes and epigenomes. Nature Reviews Genetics.