Neanderthal Man (8 page)

I published the quagga data in
Nature,
with Allan as my co-author.
{6}
Clearly it was now possible to study ancient DNA in a systematic and controlled way. I felt sure that extinct animals, Vikings, Romans, Pharaohs, Neanderthals, and other human ancestors would now soon be subject to the powerful methods of molecular biology, although proving that would take some time. (After all, I had to compete with my lab mates for use of our PCR machine.) One interest of Allan’s was human origins. Not much earlier, with Mark Stoneking and Rebecca Cann, he had published a controversial paper in
Nature
comparing mitochondrial DNAs from people from all over the world by means of cumbersome analyses using enzymes that cut the DNA at various places of known sequence—indicating that the mtDNAs could be traced back to a single common ancestor, who lived in Africa some 100,000 to 200,000 years ago.
{7}
Now this work could be extended by studying DNA sequences from many more individuals. A young graduate student, Linda Vigilant, who arrived at the lab on a motorcycle every morning, was doing this work. I was peripherally aware of her boyish charm but saw her mostly as a competitor for time on the coveted PCR machine. Little did I know that at a later time and in another country, we would be married and have children together.

So far, the reconstruction of human evolution from genetic data had been limited to studying differences in DNA sequences in living individuals and inferring how past migrations had resulted in the differences. These  inferences were based on models that reflected ideas about how DNA sequences accumulate nucleotide changes and how variants are transmitted from generation to generation in populations, but the models were by necessity great oversimplifications of what could have gone on in the past. They assumed, for example, that within a population every individual had an equal chance to produce children with every other individual of the opposite sex. They also assumed each generation to be a discrete entity with no intergenerational sex and no difference in survival based on the DNA sequences under study. Sometimes I felt that this amounted to little more than making up stories about the past, and very clearly all of it was indirect. To go back in time and actually see what genetic variants had existed in the past would be “catching evolution red-handed,” as I liked to say, by studying DNA sequences from many individuals in the past, adding direct historical observations to the work that Linda was doing on people living today.

These were ambitious ideas, so I decided to try them out over time periods more modest than thousands of years. The Museum of Vertebrate Zoology at UC Berkeley housed enormous collections of small mammals assembled by naturalists working in the American West over the past hundred years. With Francis Villablanca, a graduate student from the museum, and Kelley Thomas, a postdoc in Allan’s lab, I set out to study populations of kangaroo rats, small rodents named for their tendency to jump around on their inordinately large hind legs (see Figure 3.1). They are abundant in the Mojave Desert on the border between California, Nevada, Utah, and Arizona, where they are the favorite food of rattlesnakes. I extracted and sequenced mtDNA from the skins of several in the museum that had been collected at three different places in 1911, 1917, and 1937. Francis, Kelley, and I then got copies of the zoologists’ field notes and maps and took off for a series of trips to the Mojave to set traps at the same locations. We drove into the desert following the old field maps and identified the places where our zoologist predecessors had been forty to seventy years earlier. As the sun was setting, we set traps among sage brush and Joshua trees. Sleeping under the stars on clear and calm desert nights, interrupted only occasionally by the sound of our rodent traps snapping shut, was a pleasurable change from my urban, work-filled, everyday life.

Back in the lab, we extracted and sequenced mtDNA from the rodents we had collected and compared them to the sequences from animals that had lived some forty to seventy generations before. We found that these variants had not changed markedly over time, and while this observation was not entirely unexpected, it was still satisfying in that this  was the first-ever peek back in time at the genes of the ancestor populations of animals living today. We published our findings in the
Journal of Molecular Evolution
{8}
and were pleased to find a glowing comment about our work in
Nature
{9}
by up-and-coming evolutionary biologist Jared Diamond, who said that the new techniques made possible by the PCR meant that “old specimens constitute a vast, irreplaceable source of material for directly determining historical changes in gene frequencies, which are among the most important data in evolutionary biology.” He also said that “this demonstration project will make life harder for those who are too narrow-minded to understand the scientific value of museum specimens.”

Figure 3.1. A hundred-year-old kangaroo rat and a present-day one from the Museum of Vertebrate Zoology at UC Berkeley. Photo: UC Berkeley.

However, to me, human evolutionary history was the Holy Grail, and I wondered whether the PCR could open a window into our own past. In Uppsala, I had gotten a sample from some gruesome yet amazing discoveries made in Florida sinkholes. In these water-filled alkaline deposits, ancient Native American skeletons were found; inside the crania, the brains, although slightly shrunken, were preserved in surprising detail. Using old-fashioned techniques, I had shown that the sample contained preserved human DNA, and I presented these results at Cold Spring Harbor, along with my mummies. Through Allan, I now acquired a sample from a similar find in Florida of 7,000-year-old brains. I extracted DNA and retrieved short fragments that appeared to be an unusual mtDNA sequence that existed in Asia but that had not previously been seen in Native Americans. Although I found the sequences twice in independent experiments, I had realized by now that contamination with modern DNA was a very common problem, particularly when ancient human remains were being studied. So I cautioned in the paper that “indisputable proof that the amplified human sequences reported here are of ancient origin must await more extended work.”
{10}

Nevertheless, this research seemed promising; perhaps I needed to learn more about human population genetics. When Ryk Ward, a theoretical population geneticist from New Zealand who worked in Salt Lake City, contacted Allan’s lab to learn about the PCR, I volunteered to work with him. This resulted in a monthly commute to Utah, where I taught people in Ryk’s lab how to do PCRs. Ryk, an excellent population geneticist, was also pleasantly eccentric, given to wearing shorts and knee socks even in cold weather and taking on projects and various administrative tasks without finishing them. This latter habit did not endear him to his university, but on the plus side he loved to discuss science and had an almost infinite patience for explaining complicated algorithms to people like me who sadly lacked formal mathematical training. Together we studied mtDNA variation in the Nuu-Chah-Nulth, a small First Nations group on Vancouver Island with whom Ryk had worked for many years. Amazingly, we found that the few thousand individuals in this group contained almost half the mtDNA variation that exists among native people throughout the North American continent. This finding suggested to me that the common belief that such tribal groups in the past were genetically homogeneous was a myth, and that instead humans may always have lived in groups that contained substantial amounts of genetic diversity.

Back in Berkeley, it seemed that almost everything we tried worked. When Richard Thomas, a Canadian postdoc, came to learn PCR in the lab and needed a project, I suggested he take a turn working on
Thylacinus cynocephalus,
the marsupial wolf that had frustrated me during my sojourn in Zurich. The thylacine, native to Australia, Tasmania, and New Guinea, looked very much like a wolf but was a marsupial, like kangaroos and several other Australian animals. It was therefore a textbook example of  convergent evolution, the process whereby unrelated animals in similar environments and subject to similar pressures often evolve similar forms and behaviors. By sequencing small pieces of mtDNA from the marsupial wolf, we showed that it was closely related to other carnivorous marsupials in the region, such as the Tasmanian devil, but distant from South American marsupials, although some extinct marsupials there had been very wolf-like. This meant that wolf-like animals evolved not only twice but three times, once among placental mammals and twice among marsupials. Thus evolution was, in a sense, repeatable—an observation that had already been made, and would be made again, in studies of other groups of organisms. We wrote this up for
Nature,
and Allan graciously allowed me to be last author, the place occupied by the scientist who has led the work.
{11}
This was a first for me, and I knew that my situation in science was beginning to change. Until now, I had been someone who did the work at the lab bench, producing results by doing experiments myself the whole day and often much of the night; even when the ideas were my own I was often helped and inspired by discussions with a supervisor. Now I realized that this was beginning to change. I was not doing all experiments myself anymore. Gradually, I would have to be the one to lead and inspire others. While this prospect seemed daunting when I thought about it in the abstract, I nevertheless found that doing so often came naturally to me.

Whereas I worked with others on many applications of the PCR to ancient DNA, I concentrated my own efforts on understanding the technical intricacies of ancient DNA retrieval. I summarized the knowledge accumulated during my work in Uppsala, Zurich, London, and Berkeley in a paper in the
Proceedings of the National Academy of Sciences,
showing that DNA in ancient remains was generally short in length, contained many chemical modifications, and sometimes exhibited cross-links between molecules.
{12}
The degraded state of the DNA had several implications for work done with the PCR. Its main consequence was the unfeasibility of using the PCR to retrieve long pieces of ancient DNA. Anything above 100 or 200 nucleotides was generally impossible. I also found that when there were few or even no molecules long enough for the DNA polymerase to operate continuously from one primer to the other, the polymerase would sometimes stitch shorter pieces of DNA together, producing Frankenstein’s monster–like combinations that did not exist in the original genome of the ancient organism. Formation of such hybrid molecules through this  process, which I called “jumping PCR,” is an important technical complication that can confuse results when amplifying ancient DNA, and I described it in two papers, but I totally overlooked its broader implications. As it happens, a few years later the same basic stitching process was used by a more practically oriented scientist, Karl Stetter, to combine pieces of different genes to generate new “mosaic” genes that made proteins with new properties. This idea—which, being totally focused on my forays into the past, I had utterly failed to consider—formed the basis for a whole new branch of the biotech industry.

While many things were working well in Allan’s lab, the limitations of the new techniques and of DNA preservation also began to be discernible to me. First, not all ancient remains contained DNA that could be retrieved and studied, even by the PCR. In fact, apart from museum specimens that had been prepared rapidly after an animal’s death, few old samples yielded DNA that could be amplified. Second, in old samples that did yield DNA, its degraded state meant that one could generally amplify only pieces that were 100 or 200 nucleotides long. Third, it was often next to impossible to amplify nuclear DNA from old specimens. My dream at Uppsala of finding long pieces of nuclear ancient DNA seemed to be just that—a dream.

My life in the Bay Area was intense and satisfying, outside as well as inside the lab. I had always been attracted to men as well as women and had been active in the gay-rights movement in Sweden. In the Bay Area the AIDS epidemic was growing exponentially and took the lives of thousands of young men. Feeling I had to do something to help, I had joined the AIDS Project of the East Bay as a volunteer. There I encountered two of the most beautiful aspects of American society: self-organization and volunteerism, habits often lacking in Europe. Yet, in spite of this welcoming atmosphere and the scientific opportunities I encountered in the United States, I wanted to return to Europe eventually. It was a girlfriend who was to have a decisive influence on the route the rest of my life would take. Barbara Wild, a German graduate student in genetics, was visiting Berkeley, and I was introduced to her by Walter Schaffner, who had arranged her stay there. She was energetic, beautiful, and smart. We had a short but intense affair that continued even after she returned to her native Munich. I took every opportunity to visit Europe; on one occasion we met for an almost ridiculously romantic weekend in Venice. Much of my emotional life since my teenage years had been centered on infatuations with heterosexual men,  many of whom ended up as little more than friends. It was an exhilarating experience to walk around in Venice with Barbara and behave publicly in a way I had never dared with my erstwhile boyfriends.