________________________________
During the weeks and months after our publication of the Neanderthal mtDNA sequence, I reflected on what had led up to it. I had come a long way from my first attempts sixteen years earlier to extract DNA from a piece of dried calf’s liver from the supermarket. Now, for the first time, we had used ancient DNA to say something new and profound about human history. We had shown that the archetypical Neanderthal carried mitochondrial DNA very different from the mtDNA in people today, and that he or his relatives had not, before they became extinct, contributed their mtDNA to modern people. The achievement had required years of painstaking work to develop techniques to reliably determine DNA sequences from individuals long dead. Now that I had these techniques at my disposal, and a group of dedicated people able and willing to try new things, the biggest question was: Where should we go from here?
One task seemed of immediate importance: to determine mitochondrial DNA sequences from other Neanderthals. As long as we had studied only one individual, it remained possible that other Neanderthals carried mitochondrial genomes very different from the one from Neander Valley, perhaps even carrying mitochondrial genomes that were like those of present-day humans. Mitochondrial DNA sequences from additional Neanderthals would also reveal something about the genetic history of the Neanderthals themselves. Present-day humans, for example, have relatively little genetic mtDNA variation. If Neanderthals did, too, this would suggest that they had originated and expanded from a small population. If, on the other hand, they had as much mtDNA variation as any of the great apes have, this would suggest that over their history their numbers had never been very low. They would not have had such a dramatic history with ups and downs in population size as modern humans have. Matthias Krings, eager to follow up on his success with the iconic type specimen from Neander Valley, was keen to examine other Neanderthal specimens.
The major problem was getting access to fossils sufficiently well preserved for us to do work.
I thought a great deal about why we had been successful with the Neander Valley type specimen and came to realize that the fact that it had come from a limestone cave might be significant. Tomas Lindahl had taught me that acid conditions cause DNA strands to disintegrate, which was why the Bronze Age people found in acid bogs in northern Europe had never yielded any DNA. But when water passes over limestone, it becomes slightly alkaline. So I decided we should concentrate on Neanderthal remains unearthed in limestone caves.
Unfortunately, I had never paid much attention to the geological features of Europe in school. But I remembered the first anthropological conference I had ever attended, in Zagreb, in what was then Yugoslavia, in 1986. During the conference, we were taken on excursions to Krapina and Vindija, two sites where large amounts of Neanderthal bones had been found in caves. I made a quick search in the literature and confirmed that both Krapina and Vindija were limestone caves, which was promising. Promising as well was the presence of large numbers of animal bones, particularly of cave bears, in the caves. Cave bears, which were a large plant-eating species, became extinct shortly after 30,000 years ago, just like the Neanderthals. Their bones often abound in caves, often in circumstances suggesting that they died during hibernation. I was happy about the presence of cave-bear bones because they could possibly serve as a convenient tool to check whether DNA was preserved in the caves. If we could show that their bones contained DNA, this might be a good means to convince hesitant curators that they should allow us to try the much more valuable Neanderthal remains from the same cave. I decided to interest myself in cave-bear history, especially in the Balkans.
After a bloody war with Serbia, Zagreb had become the capital of the independent Republic of Croatia. The largest collection of Neanderthals there is from Krapina, in northern Croatia, where starting in 1899 the paleontologist Dragutin Gorjanović-Kramberger discovered more than eight hundred bones from some seventy-five Neanderthals—the richest cache of Neanderthals ever found. These bones are today housed in the Museum of Natural History in the medieval center of Zagreb. The other site, Vindija Cave (see Figure 6.1) in northwestern Croatia, was excavated by another Croatian paleontologist, Mirko Malez, in the late 1970s and early 1980s. He found bone fragments of several Neanderthals but no spectacular crania like those found in Krapina. Malez also found enormous amounts of cave-bear bones. His finds are housed in Zagreb, too, in the Institute for Quaternary Paleontology and Geology, which belongs to the Croatian Academy of Sciences and Arts. I arranged to visit both this institute and the Museum of Natural History. In August 1999, I arrived in Zagreb.
The Krapina Neanderthal collection was extremely impressive, but I was skeptical about its potential for DNA research. The bones were at least 120,000 years old and therefore older than anything we had found to yield DNA. The Vindija collection looked more promising. First of all, it was younger. Several layers in the excavation had yielded Neanderthal remains, but the uppermost and thus the youngest one to do so was between 30,000 or 40,000 years old—young, as far as Neanderthals go. I saw a second exciting feature of the Vindija collection: it was full to overflowing with ancient cave-bear bones. They were stored, according to bone type and layer, in innumerable paper sacks that were coming apart in the humidity of the Quaternary Institute’s basement. There were sacks full of ribs, others full of vertebrae, others of long bones, and yet others of foot bones. It was an ancient DNA gold mine.
Figure 6.1. Vindija Cave in Croatia. Photo: J. Krause, MPI-EVA.
In charge of the Vindija collection was Maja Paunovic, a woman of a certain age who spent her days in an institute without public exhibitions and with few facilities for doing research. She was friendly enough but understandably dour—no doubt aware that science had passed her by. I spent three days with Maja, going through the bones. She gave me cave-bear bones from several layers at the Vindija site as well as small samples of fifteen of the Neanderthal bones. This seemed exactly what we needed for the next step in our exploration of the genetic variation among Neanderthals. When I flew back to Munich I felt confident that we would make quick progress.
In the meantime, Matthias Krings had extended his sequencing of the Neanderthal type specimen to a second part of the mitochondrial genome. The results confirmed that this specimen’s mitochondrial DNA shared a common ancestor with present-day human mtDNAs about half a million years ago. But this was of course what we had expected, so the news felt slightly boring after the emotional high produced by the first Neanderthal sequences. Not surprisingly, he was eager to throw himself on the fifteen Neanderthal bone samples I had gotten from Maja in Zagreb.
We first analyzed their state of amino-acid preservation. Amino acids are the building blocks of proteins and can be analyzed from much smaller samples than are needed for DNA extractions. We had shown before that if we could not find an amino-acid profile suggesting that the samples contained collagen (the main protein in bones), and if the amino acids were not present largely in the chemical form in which they are built into proteins by living cells, then our chances of finding DNA were very small and there would be no point in destroying a larger piece of the bone in an attempt to extract it. Seven of the fifteen bones looked promising, with one that particularly stood out. We sent a piece of that bone for carbon dating and the result showed that it was 42,000 years old. Matthias made five DNA extracts and amplified the two mitochondrial segments he had studied in the type specimen. It worked nicely. He sequenced hundreds of clones, taking pains to ensure that every position was observed in at least two amplifications that, at my insistence, should come from different extracts, in order to make absolutely sure that they were totally independent of each other.
In March 2000, while Matthias was working on this, a paper that appeared in Nature took us by surprise. A group based in the UK had sequenced mtDNA from another Neanderthal, unearthed at Mezmaiskaya Cave in the northern Caucasus.{36} They had not applied all the technical approaches we advocated to make sure that the sequence was correct; for example, they had not cloned the PCR products. Nevertheless, the DNA sequence they found was almost identical to our type-specimen sequence from Neander Valley. Matthias, who had his sequences almost finished, was disappointed that he’d been beaten to the publication of the world’s second Neanderthal mtDNA sequences—especially since his progress had been slow due to all the precautions and checks on which I insisted. I sympathized with him, but I was also happy that our pioneer sequence from Neander Valley had been verified by a group working independently of us. Yet I did not quite agree with the commentary Nature published along with the paper, which said that this second Neanderthal sequence was “more important” than the first because it showed that the first was correct. I wrote that off as sour grapes on Nature’s part for not getting to publish the first Neanderthal sequence.
There was a consolation prize of sorts for Matthias. The second Neanderthal DNA not only served to confirm the results in our 1997 Cell paper, but now that we knew three sequences, including the one Matthias had determined from Vindija, it became possible to say something, albeit something tentative, about genetic variation among Neanderthals. Genetic theory holds that with just three sequences there is a 50 percent chance of sampling the deepest branch of a tree relating all the mitochondrial DNAs in a population. It turned out that 3.7 percent of the nucleotides in the segment that Matthias and the British group had sequenced differed among the three Neanderthals. For perspective, we wanted to compare this degree of variation to the variation in humans and the great apes. First, we used sequence data for the same segment determined by many other groups from 5,530 humans from all over the world. In order to make a fair comparison to the three Neanderthals, we sampled three randomly chosen humans many times, so that we could calculate an average of how many differences three humans carry in the same sequence. It was 3.4 percent, very similar to that for the three Neanderthals. There were 359 chimpanzee sequences available for the same mtDNA segment. When we sampled chimpanzees in the same way, they differed by an average of 14.8 percent, and for twenty-eight gorillas the corresponding value was 18.6 percent. So Neanderthals seemed to be different from the great apes in having little mtDNA variation, just like present-day humans. Obviously, it was risky to speculate from just three individuals, and from just mtDNA, so when we published these data later in 2000, in Nature Genetics, we stressed that it would be desirable to analyze more Neanderthals; nevertheless, we suggested that Neanderthals were probably similar to modern humans in having little genetic variation and that they had therefore expanded from a small population, just like us.{37}