Regenesis (25 page)

More generally, anything can be done ineptly or expertly, carelessly or carefully, inhumanely or compassionately. This is equally true of attempts to bring back extinct species. There is no reason to think that extinction reversal will be carried out any less humanely than any other medical or
experimental procedure. Indeed, the protocols already in place for the humane treatment of experimental subjects, whether human or nonhuman, can and ought to be extended to members of species that we might choose to bring back.

A final argument against extinction reversal is that to bring species back selectively, according to our own tastes and prejudices, will result in an anthropomorphized, “boutique” environment that reflects human values and judgments and which will result in an artificial construct rather than a natural phenomenon. However, we already live in such a world and have done so ever since the beginning of agriculture, if not long before. Everything from skyscrapers to golf courses, poodles to the Panama Canal, the Hoover Dam, Venice, and Las Vegas all attest to the fact that humans remake nature according to their wishes.
*

Nor are we alone in this respect. Many other animal species also reconstruct the world according to their own wants and needs: birds build nests, beavers build dams, bees build hives, spiders spin webs and cocoons, and ants construct mounds, entire cities, and even ant cemeteries. A propensity for redesigning nature seems to be an inherent part of life itself.

All of that said, how do we bring back animals that have long since vanished from the scene?

That depends on the species and what there is left of it. For species whose intact cells, tissues, or other genetic materials have been preserved—as in the case of the bucardo, for example, or species well represented in frozen zoos, labs, or other storage facilities—revival will be possible through straightforward interspecies nuclear transfer cloning.

A second category consists of species whose genetic material might or might not be too corrupt for cloning. This is true of the woolly mammoth,
for example. Although mammoth specimens buried in permafrost may look remarkably lifelike, their DNA is not in the same condition. The woolly mammoth, however, really falls into a class by itself.

Unlike the Neanderthals, whose remains consist exclusively of teeth, skulls, and other bones, Siberian mammoth carcasses have turned up with hair and even soft tissue on them. In August 1799 a Russian hunter looking for mammoth tusks came across an oddly shaped block of ice on the shoreline of the Laptev Sea in north-central Siberia. The summer sun had melted some of the ice, exposing two projections that later turned out to be tusks.

Two years later, the hunter returned to the area during the summertime and found that one side of the animal had been exposed to view, while the rest of the body was still frozen. It was not until 1806 that Michael Adams, of the Imperial Academy of Science at St. Petersburg, reached the site. By this point, nearby villagers had hacked off some of the flesh and fed it to their dogs. Bears, wolves, and foxes had eaten the rest, leaving only a skeleton. Now known as the Adams mammoth, the skeleton is on display in the St. Petersburg Zoological Museum.

More recently, the Jarkov mammoth was discovered in 1997 by a family of that name who came across a tusk protruding from the frozen ground of the Taymyr peninsula in northernmost Siberia. A group of latter-day mammoth hunters arrived at the scene and speculated that an intact mammoth carcass could be lodged in the ice—an entire mammoth! In October 1999, a helicopter lifted a twenty-three-ton ice block with tusks protruding bizarrely from it up and out of the frozen tundra, and hauled it to an ice cave. There, as recorded by a Discovery Channel film crew, scientists began defrosting the remains with hair dryers.

The Jarkov mammoth turned out to be mostly bones, but even so, a bit of soft tissue remained. It looked like a strip of beef jerky.

Coincidentally, a second defrosting mammoth (the Hook mammoth) happened to be located nearby, and some of the expedition's researchers traveled to the site. One of them, Alexi Tikhonov, cut off a piece of what appeared to be mammoth muscle. Jokingly, he offered it to those present all of whom refused this morsel. And so, braced by a few shots of vodka,
he took a bite himself. “It was awful,” he said. “It tasted like meat left too long in the freezer.” (Mammoth meat is so common in Siberia that fox trappers use it as bait.)

Any mammoth tissue that is fresh enough to eat might harbor intact DNA. Two Japanese scientists have plans for resurrecting the animals. One of them, Kazufumi Goto, proposes finding intact mammoth sperm cells that he will use to inseminate a female elephant to produce a mammoth-elephant hybrid.

Hybrids are usually sterile, but there are known exceptions; whether a mammoth hybrid would be sterile is currently unknown. Assuming it wasn't sterile, then by injecting additional mammoth DNA into the resulting mammoth-elephant cross, you would get a second hybrid that was even more of a mammoth than an elephant. According to Goto, repetition of the process with successive new offspring would yield, within fifty years, an animal that was 88 percent mammoth. (Elephants have a twenty-two-month pregnancy and don't produce offspring until they are ten.)

While this scenario is technically feasible, regenerating a woolly mammoth in this way is a speculative possibility at best because intact mammoth sperm cells have never been found—and might never be. The second Japanese researcher, Akira Iritani, is chairman of the Department of Genetic Engineering at Kinki University, near Osaka. He plans to find a mammoth cell with intact chromosomes and then fuse them with an egg cell from an Asian elephant and let it divide to an early embryonic stage. Finally he plans to implant the embryo into the womb of an Asian elephant and hope for the best.
*

Richard Stone, who wrote a book about resurrecting mammoths, calls this “the Mount Everest of biology experiments.” Everything hinges on locating a fresh supply of exceptionally well preserved frozen mammoth tissue, which scientists would then defrost under carefully controlled laboratory conditions. This combination of circumstances has not occurred
to date, despite the fact that mammoth hunters have conducted several field searches. Russian mammoth expert Sergey Zimov takes a dim view of these searches. “Frozen mammoths find you, not the opposite,” he says. “Directed searches have almost no chance of success.”

A third class of extinct species is represented by DNA samples that are so fragmented and corrupt that their genomes must be laboriously reconstructed from innumerable isolated pieces. This is true of Neanderthal man, whose draft genome Svante Pääbo reconstructed in that manner. Unfortunately, the draft genome doesn't exist physically as actual chromosomes or genes, but only as strings of DNA sequences stored in computers.

Theoretically it is possible to convert those sequences into a physical, real-life genome by synthesizing short sequences (oligos) in DNA synthesis machines and then stitching them together into chromosomes. In 2010 Craig Venter created his so-called synthetic
Mycoplasma
bacterium by chemically synthesizing its entire genome, oligo by oligo. However, there is a huge difference between synthesizing a bacterial genome and synthesizing the genome of an animal as large and complex as Neanderthal man. While Venter's
Mycoplasma
genome was 1.08 million base pairs in length, the Neanderthal's genome consists of 3 billion base pairs, as long as that of a modern human. Synthesizing such an object oligo by oligo would take forever—or at least a very long time.

Fortunately, there's another way to accomplish the same objective: start with a physical genome that closely resembles the Neanderthal's and then change it, piecemeal, into the genome of a Neanderthal. Reverse-engineer it into existence.

What genome closely resembles the Neanderthal's? The modern human genome. In fact, the genomic difference between a modern human and a Neanderthal is about threefold more than between one modern human and another—about 10 million base pairs. Indeed some Melanesian genomes consist of up to 8 percent more closely related to Neanderthal and Denisova genomes than to the African genomes and hence would
make a slightly (500,000 base pairs) better starting point.
*
(The chimpanzee is the next closest contender with about 30 million differences.)

Millions of alterations is a lot, and making them is all but out of the question for traditional genetic engineering methods, which introduce modifications one at a time, serially. We generally try to bring down costs before we undertake a large project (as happened with human genome sequencing). Recent work in my Harvard lab shows that you can reduce the costs of the process considerably by introducing the necessary changes on a batch basis, modifying multiple genetic sites at a time, and in parallel. This is the MAGE method, introduced in
Chapter 3
. One way to use it here would be to break up the human genome into 10,000 pieces of 300,000 base pairs each, and then replicate them in
E. coli
as bacterial artificial chromosomes (BACs) or in yeast as yeast artificial chromosomes (YACs). Then each piece can be reprogrammed in parallel by MAGE (about 1,000 changes each). By using chip syntheses, 10 million oligos can be printed at a cost of about $5,000 and at least double that to get them amplified and error-corrected in appropriate subpools. The twenty-three human chromosomes could be reconstructed in parallel (about 500 steps each) and then combined by chromosome transfer using cell or microcell fusion methods and multiple positive and negative selection markers. An example of a positive selection employs a drug resistance gene like neomycin phospho-transferase. When this resistance gene is attached to the BAC and exposed to a cell, then only the cells that take up the BAC will survive in the presence of the drug. Later, if you want to remove the piece of DNA that was needed temporarily, you can use a negative selection. An example used widely in mammalian genetics is a viral thymidine kinase gene not normally found in human cells that makes them sensitive to an antiviral drug.

Supposing, then, that we have recreated the physical genome of Neanderthal man in a stem cell, the next step would be to place it inside a
human (or chimpanzee) embryo, and then implant that cell into the uterus of an extraordinarily adventurous human female—or alternatively into the uterus of a chimpanzee. Admittedly, this will only ever happen if human cloning becomes safe and is widely used and if the possible advantages of having one or many Neanderthal children are expected to outweigh the risks.

This same technique could be applied to the wooly mammoth once its genome was fully sequenced. In 2008 a scientific team headed by Stephen Schuster and Webb Miller of Pennsylvania State University reported in
Nature
that they had reconstructed a substantially complete draft sequence from clumps of mammoth hair. But that was sufficient for them to calculate that the mammoth genome differed at 400,000 sites from the genome of the African elephant. This shows us what the road to regenerating the wooly mammoth is.

You would begin with an intact African or Asian elephant genome (both animals are phylogenetically close to the mammoth), and then by using MAGE technology you'd introduce the modifications that would turn it into a mammoth genome. Finally you would implant that genome into an elephant's embryonic cell in the now familiar way, and implant it into the womb.

And then, twenty-two months later . . .