In 2018 an international team of scientists — from labs in Houston, Copenhagen, Barcelona and beyond — got their hands on a remarkable biological specimen: a skin sample from a 52,000-year-old woolly mammoth that had been recovered from the permafrost in Siberia. They probed the sample with an innovative experimental technique that revealed the three-dimensional architecture of the mammoth’s genome. The resulting paper was published on Thursday in the journal Cell.
Hendrik Poinar, an evolutionary geneticist at McMaster University in Canada, was “floored” — the technique had successfully captured the original geometry of long stretches of DNA, a feat never before accomplished with an ancient DNA sample. “It’s absolutely beautiful,” said Dr. Poinar, who reviewed the paper for the journal.
The typical method for extracting ancient DNA from fossils, Dr. Poinar said, is still “kind of cave man.” It produces short fragments of code composed of a four-letter molecular alphabet: A (adenine), G (guanine), C (cytosine), T (thymine). An organism’s full genome resides in cell nuclei, in long, unfragmented DNA strands called chromosomes. And, vitally, the genome is three-dimensional; as it dynamically folds with fractal complexity, its looping points of contact help dictate gene activity.
“To have the actual architectural structure of the genome, which suggests gene expression patterns, that’s a whole other level,” Dr. Poinar said.
“It’s a new kind of fossil, a fossil chromosome,” said Erez Lieberman Aiden, a team member who is an applied mathematician, a biophysicist and a geneticist and directs the Center for Genome Architecture at Baylor College of Medicine in Houston. Technically, he noted, it is a non-mineralized fossil, or subfossil, since it has not turned to stone.
The information gleaned from such chromosome fossils will no doubt aid plans to “de-extinct” animals like the woolly mammoth. Three members of the research team are on the scientific advisory board of, and hold stock options in, Colossal Biosciences, a company that hopes to resurrect the mammoth, the Tasmanian tiger and the dodo. Colossal Biosciences did not provide funding or support for the research.
And 3-D genomic data will also be useful in efforts to save existing organisms from extinction. “Given the climate crisis that we’re currently in, there are big questions of how rapidly or not animals can adjust to warming or cooling patterns,” Dr. Poinar said. “Mammoths are a great thing to study in that sense, because they traveled huge ranges over the course of their lifetime and dealt with varying climates and environments.”
Recovering such a relic was fortuitous in the extreme. “It is so mysterious that these things exist,” Dr. Aiden said. “Why do these fossil chromosomes still persist?”
In their paper, the researchers drew upon physics and hypothesized that the mammoth had “spontaneously freeze-dried” in the cold of Siberia. The cells transitioned into a glasslike state, which prevented molecular movement and so preserved the chromosomes’ shape, or morphology, all the way down to 50-nanometer genomic loops. Ragini Mahajan, a doctoral student in the Houston lab, coined the term “chromoglass.” Dr. Aiden likes to call it “woolly mammoth jerky.”
‘Genomic origami’
The mammoth study, a decade in the making, built on pathbreaking research by Dr. Aiden and collaborators in 2009 and 2014. In a masterpiece of “genomic origami,” the earlier work provided the first high-resolution, 3-D maps of folded genomes. And it prompted the invention of a technique called “Hi-C” (unrelated to the fruit drink, other than as a lab mascot of sorts — Dr. Aiden keeps a stash of juice boxes in his office).
The technique probed the 3-D architecture of whole genomes and dealt with a vexing problem. A genome is like a book containing all of an organism’s genetic information; DNA sequencing extracts and reads individual pages of the book, but without page numbers. Hi-C puts the pages in order.
Dr. Aiden wondered whether this protocol could be applied to ancient specimens: “paleoHi-C.” He set his sights on the woolly mammoth.
Cynthia Pérez Estrada, a neuroscientist and genomicist and a team member at the Houston lab, conducted an initial phase of “crazy experiments” that accelerated at Thanksgiving in 2016. “Erez invited us for dinner, and I collected the turkey bones and started to perform experiments,” Dr. Peréz Estrada recalled. “The question was, Can we recover genome architecture from degraded samples?”
She tested everything: roadkill, the leather trim on her knapsack, friends’ leftovers that she had let rot outside in the scorching Houston heat. Once Dr. Pérez Estrada felt somewhat optimistic about the prospect, she emailed Love Dalén, a geneticist at Stockholm University. “About mammoths, Love is the guy to talk to,” she said. He soon joined the team.
Dr. Dalén introduced the researchers to Thomas Gilbert, director of the Center for Evolutionary Hologenomics at the University of Copenhagen. Dr. Gilbert has long investigated the paleogenomics of many species. “When I heard how Hi-C worked, it clicked in my head that in theory it should work on ancient DNA,” he said.
Aiming at the same target — the mammoth — the two labs joined forces.
Marcela Sandoval-Velasco, a paleogenomicist who was then a member of the team at the Copenhagen lab, spent hours “cracking protocols,” modifying the experiments in an attempt to make museum samples cooperate. Dr. Sandoval-Velasco and Dr. Pérez Estrada visited back and forth. They tested bees, ants, wild asses, fish, pieces of polar bear skull, pickled remnants of the last great auk. Almost all failed.
“But despite these failures, and some successes, we learned something very important,” Dr. Pérez Estrada said. “For these experiments to work, the sample would have to be preserved in a very specific manner.”
Dr. Sandoval-Velasco, now at the University of Mexico, noted that “with ancient DNA, 1 percent of your sequencing output is going to be data from the organism that you’re interested in. That’s normal. Ninety-nine percent is going to be a pool of environmental, microbial, human contamination.”
In 2019, Dr. Dalén shared a specimen from a recently excavated woolly mammoth nicknamed Chris Waddle, after a British soccer player known for his mullet (although the mammoth was female). When Dr. Sandoval-Velasco tried this mammoth sample, the experiment worked. But she didn’t know it until the data had been analyzed.
Just as the Covid pandemic hit, she visited Marc Marti-Renom, a structural genomicist and a member of the team who runs the National Center for Genomic Analysis in Barcelona. His lab performed some “computational massaging” on the experimental data, Dr. Marti-Renom said.
“No one was believing that the structure could be there,” recalled Juan Antonio Rodríguez, a researcher and a team member in Barcelona, now at the University of Copenhagen. “The first 3-D genome map we generated was not good, but it was very promising,” he said. “So we just tweaked the method here and there, and then eureka, we got an awesome 3-D map.”
The 4.1 percent difference
Once the researchers had the 3-D signal, they could examine how the 52,000-year-old mammoth’s genome was folded. One remarkable finding was that classic features of modern chromosomes, at many scales, were preserved.
Olga Dudchenko, a researcher in the Houston lab whose expertise lies at the intersection of applied physics, mathematics and genomics, worked algorithmic wonders over the course of the study. And for an ancient-versus-modern comparison, she put together the genome of an Asian elephant, an endangered species and the mammoth’s closest living relative. (The necessary samples were provided by Methai, the 55-year-old matriarch of the herd at the Houston Zoo, and a deceased Asian elephant at the San Antonio Zoo.)
The elephant and mammoth genomes were strikingly similar: Each had 28 chromosomes — not surprising, yet a valuable confirmation. “It makes it really, really obvious that we must have done this right,” Dr. Aiden said. It’s a proof of principle for the technology.”
But Dr. Dudchenko noted that a comparison of the two 3-D genomes showed that about 800 genes in skin — 4.1 percent of the total — exhibited differing activity profiles. One class of differing genes pertained to hair follicle development and hair growth regulation. “These genes give us a different kind of clue as to why the woolly mammoth was woolly,” Dr. Aiden said.
That these clues exist at all seemed “ridiculous,” he added. To better understand how the sample could have been so well preserved, the researchers looked to both physics and mathematics.
Dr. Dudchenko translated the folding patterns of the mammoth’s genome into space-filling curves and random walks. She proposed a simple model and a proved a theorem describing how chromosomes would degrade over millenniums — a rare glimpse of certainty in the messy world of biology.
All the same, Dr. Aiden admitted, there is “some woolliness around this theory.”
“There’s a few different ways in which this could actually correspond to the exact arrangement of atoms in the universe,” he said, “and we don’t 100 percent know what the answer is.”