Back in the Jurassic Park summer of 1993, paleontologist Jack Horner and his graduate student Mary Schweitzer announced that they had isolated DNA from a 65-million-year-old Tyrannosaurus rex bone. Horner and Schweitzer were cautious about their find. The DNA, they said, might have come from a fungus or plant that had contaminated the bone--either in the ground or in the lab--rather than from the dinosaur itself. When Scott Woodward reported his own discovery of dinosaur DNA last November, though, he seemed more sure of himself. It was quite a rush to see it on the gel and realize this was something that had been missing for 80 million years, recalls Woodward, a microbiologist at Brigham Young University in Utah. And here it was, right before our eyes.
Woodward’s dinosaur DNA came from a coal mine, which was actually the first place he thought to look for it. For years he and his colleagues had been isolating DNA from Egyptian mummies, Utah mammoths, and other ancient specimens, and they’d come to realize there were a number of situations in which the genetic molecule could survive much longer than people used to think. Woodward had also read about the well-preserved human corpses found in peat bogs in Florida and the United Kingdom. That got him thinking. I had grown up in coal mining country, so I knew that coal mines are ancient bogs, he says. I also knew that dinosaur tracks were often found in coal mines, and sometimes dinosaur bones. So in December 1992 he contacted some old friends in the coal mining industry and asked them to keep their eyes peeled for big bones.
In June 1993, about when Horner’s results became public, one of Woodward’s friends came through--he’d found some large bone fragments in a sandstone deposit in the roof of a Utah coal mine. Although the fragments themselves could not be reliably linked to a particular type of animal, the sandstone had previously been dated to 80 million years ago--a time when dinosaurs ruled and large mammals were yet to come. Moreover, dinosaur tracks had been found in the same mine, and both tracks and bones had been found in other coal mines nearby. All those tracks and fossils, along with the size of the bone fragments, made it seem likely to Woodward that the fragments had come from a dinosaur.
When he and his colleagues extracted a couple of the fragments from the rocks, they noticed something promising about them. As sometimes happens in coal formations, the bone hadn’t fossilized--the organic material hadn’t been replaced with rock. The bone was very fragile and kind of waxy, like a hard soap, Woodward recalls. Examining sections of the bone under a microscope, the researchers found that cellular structures were preserved, indicating that some DNA might be, too. They set out right away to isolate it.
Woodward’s team never expected to find more than small pieces of ancient DNA, because the long molecule tends to fall apart over time. They looked instead for ten small segments of several genes that most animals have today, including six segments of the gene for cytochrome b, a protein involved in the creation of the energy carrier ATP. The genes had all been isolated and sequenced in a wide variety of animals. That meant Woodward’s team could use pieces of the modern genes as primers to seek out, by means of the polymerase chain reaction (PCR), similar DNA in a solution prepared from the ancient bone fragments. It also meant the researchers would be able to compare whatever ancient DNA they found with DNA from living organisms.
They found only one of the gene segments they looked for; it came from the cytochrome b gene. The segment was just 174 base pairs long, whereas the complete gene would have included more than 1,000 base pairs. But when Woodward and his colleagues quickly compared the segment’s sequence with that of cytochrome b genes from hundreds of animals, including birds, reptiles, amphibians, mammals, and insects, they noticed something striking. Although all cytochrome b genes bear a family resemblance--otherwise Woodward’s search wouldn’t have worked at all--the resemblance of the bone-fragment DNA to anything living wasn’t especially close. This was our first, partial suggestion that we had something ancient, Woodward says. That by itself was not proof, but it was certainly a strong indication that our sample wasn’t contaminated by any DNA we had in the laboratory.
Next Woodward’s team tried to replicate the result--to see if they could generate the same section of the cytochrome b gene again, with the same base-pair sequence, from the original bones. But that second sequence didn’t match the first. Neither did the third, the fourth, or any of the sequences the researchers obtained on subsequent attempts, nine in all. Woodward doesn’t see this failure to replicate as undermining his results, however--on the contrary. I think the reason the sequences are so different is that we are dealing with ancient DNA, he says. We know the DNA is going to be damaged--it’s 80 million years old.
The way the PCR technique picks ancient DNA out of a solution, Woodward explains, is by making many copies of the particular segment the researcher is looking for. Thus if the original DNA is damaged, the PCR enzymes will make flawed copies, inserting bases at random at the damaged points, and thereby producing a different base sequence each time. In any case, Woodward argues, the nine sequences were not all that different--each sequence had at least 92 percent of its bases in common with any one of the others, and at 112 positions along the 174-base-pair chain, all nine were identical.
To get as close as possible to what the ancient gene would have looked like, Woodward derived a consensus sequence--a sequence consisting of the 112 bases that all nine fragments shared, and at each of the other positions, of whatever base was most common among the nine. Then he and his colleagues compared that sequence with the cytochrome b sequences from living animals. While the ancient DNA was unlike that of anything alive, it turned out to be more similar to the DNA of mammals--in particular, whales- -than it was to birds, reptiles, or anything else. That result is surprising indeed: dinosaurs were reptiles, after all, and most paleontologists think they were the ancestors of birds.
And in fact, many paleontologists are skeptical that Woodward has isolated dinosaur DNA at all. Some flat out don’t believe that it’s possible to recover 80-million-year-old DNA. Others, like Rob DeSalle of the American Museum of Natural History, who studies DNA from insects preserved in amber, grant that Woodward’s DNA might be that old but doubt it’s from a dinosaur. I am willing to believe they have gotten ancient DNA out of bone because the way they’ve described their experiment seems adequate for obtaining DNA, says DeSalle. But they have simply not shown that they have dinosaur DNA. There is no way. I don’t think there is enough information in the small sequence they have to do this kind of analysis. If they got a lot more sequence and showed that the DNA came out as either the sister group to reptiles or the sister group to birds, then that would convince me.
Says Woodward: Some people have said that if this sequence is not more closely related to birds, then it isn’t a dinosaur sequence. Well, I don’t think that is a good reason to throw out the sequence. I don’t believe that all dinosaurs necessarily had to be related to birds. The dinosaurs were on the Earth for 200 million years, and some certainly had enough time to diversify as far away from birds as they moved from reptiles and mammals.
Only more data will convince the skeptics, of course--including Jack Horner, who helped launch this whole line of research. Getting DNA out of bone is easy, says Horner. We have the same thing Woodward has--we have DNA, but we can’t prove that it’s from a dinosaur. I don’t think anyone right now can prove that they have DNA from a dinosaur. There’s just too much chance of contamination. Both Horner and Woodward are now looking for other organic molecules in their dinosaur bones--proteins such as hemoglobin and collagen, which are much more stable than DNA and much less likely to be contaminants, because plants and fungi don’t have them. If we find these proteins, says Horner, it will be much more convincing that we have dinosaur DNA.