Sie sind auf Seite 1von 8

Ancient DNA

Genetic information that had seemed lost forever turns out


to linger in the remains of long-dead plants and animals.
Evolutionary change can at last be observed directly

by Svante Pbo

M
ost of our knowledge of the ral History Museum in London estab- cial and peripheral parts, such as the
molecular processes that un- lished a molecular biology research lab- skin on ngers and toes, were much
derlie evolutionary change is oratory. The Smithsonian Institution in better preserved [see illustration on page
based on the comparison of the genes Washington, D.C., the American Muse- 92]. These tissues quite often retained
of living species. From such dierenc- um of Natural History in New York City nuclei that accepted stains specic for
es, molecular evolutionists infer the his- and other museums have since emulat- DNA. The reason appeared to be that
torical changes that gave rise to present- ed the British example. the enzymes that cause dying tissue to
day DNA sequences. Yet these studies I rst became aware of the spectacu- digest itself, a process called autolysis,
are tentative in nature. Unlike the re- lar ease with which molecular tech- require water. Supercial tissues, how-
mains of animals and plants, DNA mol- niques allow us to isolate and study ever, may dry out before the postmor-
ecules do not leave impressions in rock. DNA sequences from humans and oth- tem degradation of DNA is complete.
Biologists therefore despaired of ever er organisms when I was a graduate With great excitement I proceeded to
being able to check their conclusions student in the early 1980s in Uppsala, extract the DNA from the tissues that
against the historical record, as paleon- Sweden. I began to wonder whether the contained cell nuclei. I worked as one
tologists do. same techniques could retrieve mean- would with any modern tissue, digest-
But in the past decade scientists have ingful data from the skins of animals ing away the proteins with enzymes
learned that ancient DNA, though de- and the mummies of humans that and using solvents to extract DNA from
graded, sometimes survives the ravages abound in museums. When a search of sugars, proteins and lipids. I then ana-
of time, and molecular biologists have the literature suggested that no one lyzed the DNA by electrophoresis, a
perfected methods of amplifying these else had considered the possibility, I technique that uses an electric eld to
trace amounts of ancient DNA. Workers embarked on a quest for suitable sam- separate DNA fragments as they mi-
have so far used DNA from bone and ples to test. grate through a gel. The smaller the
soft tissues to establish reliable sequenc- It was no easy task to get materials, fragments are, the farther they migrate,
es for seven extinct mammals. The old- for an archaeological sample is valu- a correlation from which one can calcu-
est was the woolly mammotha frozen able, and inevitably one will have to de- late their size. The results showed that
carcass that was found in the perma- stroy a part of it to extract the DNA. the DNA had been degraded to frag-
frost of Siberia. More such studies are Yet my proposal intrigued curators of ments of only 100 to 200 base pairs
under way, including eorts to decode the University of Uppsalas Egyptian (the four nucleotides that are the build-
DNA extracted from insects entombed antiquities, and they provided me with ing blocks of DNA). Fresh tissues, in
in amber millions of years ago. We can small samples of skin and muscle from contrast, yield DNA fragments more
thus look forward to learning much their mummies. Even more important, than 10,000 bases in length.
more about the genetic relations among they had excellent connections with a

T
extinct species. large museum in what was at that time o study the genetic information
Encouraged by such work, the Natu- East Berlin. of such DNA fragments, one must
I spent four days with the chief con- amplify them into large quanti-
servator of the State Museums in Berlin, ties. At that time, the only available way
looking through the mummies, some to do so was molecular cloning, in which
SVANTE PBO is professor of biolo- of which had been partly destroyed dur- the old DNA is fused to a carrier DNA
gy at the University of Munich. He was ing World War II. With a sterile scalpel I molecule that confers on it the ability
born in Stockholm in 1955 and studied
removed samples of a few grams of tis- to copy itself in bacteria. These hybrid
Egyptology, medicine and the history of
science at Uppsala University, where he
sues from the 23 mummies that ap- molecules are then introduced into bac-
took a doctorate in molecular immunol- peared best preserved. teria, and thousands of bacterial colo-
ogy in 1986. He then worked in Zurich Back in Uppsala I set out to study the nies are grown, each one multiplying a
and, for four years, in the laboratory of tissue, working nights and weekends so single piece of the original DNA. The
the late Allan C. Wilson of the University as not to impede my research in molec- bacterial colonies are then analyzed for
of California, Berkeley. His current re- ular virology, which was supposed to
search into ancient and modern DNA fo- constitute my future thesis. I began by
cuses on the elucidation of evolutionary
examining the specimens under a mi- SKIN AND BONES have yielded priceless
history and the molecular processes un-
derlying genetic change. In 1992 the Ger- croscope. I found that they varied dra- genetic data from such extinct species
man Science Foundation awarded him matically in their state of preservation. as quagga (zebralike animal, center),
the Leibniz Prize. Most of the tissues were badly degrad- moa (skeleton, foreground ) and thyla-
ed, but there were exceptions. Super- cine, or marsupial wolf (front ).

86 SCIENTIFIC AMERICAN November 1993 Copyright 1993 Scientific American, Inc.


Copyright 1993 Scientific American, Inc. SCIENTIFIC AMERICAN November 1993 87
84 BASE-PAIR 471 BASE-PAIR MUMMY DNA allows amplication of
TARGET SEQUENCE TARGET SEQUENCE only short fragments, whereas modern
DNA yields longer ones. The control
lanes on the electrophoretic gel show
the absence of contaminating human
DNA but show amplication artifacts.
ELECTROPHORESIS RESULTS
(LENGTH IN BASE PAIRS)

471

came clear that there were severe limi-


TARGETS
tations to our ability to study ancient
DNA. The dearth of clones made it ex-
tremely difcult to repeat experiments
84 and made many experiments impossi-
ble to perform in the rst place. The
ARTIFACTS reason for the low efciency of cloning
OF AMPLIFICATION was discovered when the ancient DNA
PROCESS was analyzed biochemically. The aver-
age length of the old molecules (about
100 base pairs) was the same whether
the material came from a 13,000-year-
MUMMY CONTROL MODERN MUMMY CONTROL MODERN
old ground sloth of southern Chile or a
piece of dried pork just four years old.
sequences of interest. In my experiment my results ready for publication, Rus- Most of this fragmentation occurs in
it was crucial to demonstrate that at sell Higuchi and the late Allan C. Wilson the rst hours after death, before the
least some of the cloning products were of the University of California at Berke- tissues dry out. Other types of damage
of human origin and could not have ley published work that also showed appear to come from oxygen radicals,
come from bacteria or fungi that may that DNA could survive after the death which react with bases or the sugar
have lived in the mummy. of an organism. They studied the quag- backbone of the DNA. These reactions
For reasons that became clear only ga, a member of the horse family that cause modications and loss of bases,
later, I obtained far fewer clones carry- lived in southern Africa until going ex- breakage of the strands and cross-link-
ing human DNA than I had expected. tinct at the end of the past century. ing of DNA molecules to one another.
The paucity of clones precluded an at- Samples from a 140-year-old quagga Such changes degrade the DNA and de-
tempt to isolate the most interesting skin in a German museum yielded bac- stroy its genetic information.
genesthose that code for specic pro- terial clones containing sequences from

W
teinsbecause these generally occur in the mitochondrial DNA, which exists in hen ancient DNA damaged in
only two copies per cell. Therefore, to many copies outside the cell nucleus. By these ways is introduced into
show that I had indeed replicated an- comparing these clones with sequences bacteria, the organisms try but
cient human DNA, I had to resort to iso- of modem zebras, Higuchi and Wilson generally fail to make copies. Even when
lating a so-called Alu-repeat, a sequence showed that the quagga was closely re- they succeed, they often introduce er-
that occurs in nearly a million copies in lated to the zebra and much more dis- rors. For example, the quagga sequences
the genome of humans. One bacterial tantly related to other equids. These included two positions that had bases
clone containing two such Alu-repeats were the rst sequences determined dierent from those that occupy the
was isolated from tissues of a mum- from an extinct species. same position in the genomes of other
my reliably dated to between 2,310 and Despite these successes, it soon be- vertebrates. It seemed likely that the
2,550 years ago. This nding made it isolated clones had errors in them at
clear that DNA can survive for long pe- these positions, but because it was so
riods after the death of an individual. difcult to obtain clones from the quag-
In 1984, some months before I had

Ancient DNA Milestones


These extinct organisms have yielded
meaningful genetic sequences.

AMBER
FOSSIL LEAVES MAMMOTH
40 MILLION YEARS OLD 17 MILLION 40,000 13,000

88 SCIENTIFIC AMERICAN November 1993 Copyright 1993 Scientific American, Inc.


AMBER TIME CAPSULE for a prehistoric
scorpion may provide an ideal environ-
ment for the preservation of DNA. Am-
ber, the solidied resin of ancient trees,
generally protects its contents from the
eects of air and water.

ga skin, there was no practical way to


go back and repeat the experiments.
Similarly, I could not again isolate the
same Alu-repeats from the mummy in
order to verify their sequence. Molecu-
lar evolutionists who were keen on time
travel therefore found themselves in a
depressing situation. Because they could
not verify their results by repeating an
experiment, the study of ancient DNA
could not qualify as a fully respectable
science.
A dramatic change came in 1985,
when Kary B. Mullis described a test-
tube cloning technique of extraordina-
ry sensitivity and power [see The Un-
usual Origin of the Polymerase Chain
Reaction, by Kary B. Mullis; SCIENTIFIC
AMERICAN, April 1990]. This method, strands. This separation enables the two Our eorts were soon rewarded.
the polymerase chain reaction (PCR), primers to sit down on their respective When we designed primers for the seg-
allows the multiplication in the test target sequencesone on one strand, ments of quagga DNA that Higuchi had
tube of a particular piece of DNA. The the other on the complementary strand. cloned in bacteria a few years before,
piece is specied by two short DNA Next, an enzyme begins adding bases we were able to amplify the pieces of
fragments, or primers, which are chem- at the ends of the primers, extending DNA from extracts of the skin that he
ically synthesized to match the base the double helices that the primers had had used. Our experiment revealed that
sequences anking the sequence one initiated. Because each single strand the positions in which the cloned quag-
wishes to study. The primers initiate produces a new double helix, the de- ga sequences had diered from other
the test-tube amplication, which pro- sired DNA sequence doubles in quanti- vertebrates had in fact carried the ex-
ceeds in successive cycles that each ty with each cycle. pected general vertebrate sequence.
double the number of copies of the Thus, the bacterial clones at these posi-

I
specied sequence. t immediately became clear that tions carried errors, presumably in-
A cycle begins with the melting of PCR could open great possibilities duced by damage in the ancient DNA
the double-strand template into single for the study of ancient DNA. The that the bacteria had incorrectly re-
main reason is that the technique is ex- paired. PCR, however, was able to re-
tremely sensitive. One can literally gen- trieve the correct sequence.
erate billions of copies starting from as We have since shown that PCR brings
little as a single DNA molecule. Even one back the correct ancient sequence be-
or a few intact molecules that survive cause its errors, though frequent, are
in an ancient tissue might be amplied rarely serious. It is accurate because, in
by PCR, whereas the damaged mole- practice, PCR starts the amplication
cules, which may be present in an ex- from at least a few dozen or a few hun-
cess of several 1,000-fold, would not dred ancient DNA fragments. When the
disturb the experiment. In the hope end product of an amplication is ana-
that this potential would be realized, I lyzed all at once, any error in a particu-
joined Wilsons group at Berkeley, which lar position in the sequence of one ini-
was one of the rst to use PCR. tial template will be canceled by the

MOA QUAGGA THYLACINE


4,300 140 80 PRESENT

Copyright 1993 Scientific American, Inc. SCIENTIFIC AMERICAN November 1993 89


EVOLUTIONARY TREE OF THE QUAGGA is deduced from dier- with the zebra than with the others. The evolutionary tree ( far
ences between DNA from the extinct equid and from three mod- right) reconstructs genetic changes, such as two inferred in the
ern species. The quagga shares much more sequence similarity ancestor of the quagga and the zebra ( yellow). The quagga dif-

other templates that do not carry mis- very little oxygen. We extracted and at- man remains is still impeded by techni-
takes in that position. Bacterial cloning, tempted to amplify pieces of mitochon- cal difculties that are to some extent
in contrast, starts from a single mole- drial DNA, whose rapid rate of evolu- inherent in PCR: its exquisite sensitivity
cule and is so more prone to make er- tion makes it very useful for studying can be a weakness as well as a strength.
rors of consequence. relations among populations. Any trace of modern DNA that makes
Further studies have shown that PCR Our rst attempts failed because the it into the experiment will be amplied
can also reconstruct intact DNA from extracts contained unknown factors that if it carries sequences to which the prim-
several partially degraded ancient mol- somehow inhibited the enzyme used in ers can bind. One must wonder wheth-
ecules. In such cases, the two primers PCR to replicate DNA. This problem an- er a similarity between oneself and ones
sit down on partial molecules. They are noyed us for several months, until we putative ancestors is real or merely the
extended up to a damaged site or to found out that albumin, a protein found result of sloppy laboratory technique.
an end of the template molecule. In the in the blood, appears to relax the inhi-

C
next cycle the resulting molecules sep- bition by binding to impurities that dis- ontaminating DNA can come
arate into single strands; these primers turb the polymerase. The poor state of from many sources: skin cells
are longer than the original primers the old DNA allowed only short pieces shed by the archaeologists or mu-
were. They can therefore bind to other to be amplied, but even so, the frag- seum keepers as they handle the sam-
template molecules that are not dam- ments sufced to reveal a mitochondri- ples; dust particles; or minute amounts
aged where the previous templates were. al type that had not been found in con- of DNA from earlier experiments con-
This process of jumping from tem- temporary Native Americans. ducted in the same room. This problem
plate to template can piece together in- We could not be certain that this ge- has forced workers to establish cum-
formation from many partially degrad- netic variant had gone extinct in the bersome procedures. All reagents must
ed pieces. Under favorable conditions, Americas, because we did not know be treated with the utmost care. Extrac-
it can enable PCR to amplify sequences much about the mitochondrial sequence tions and amplications must be per-
that are longer than any single frag- variations in contemporary Native Amer- formed in a room separate from all oth-
ment in the original sample. We are at- icans. Fortunately, several groups are er work. Ventilation systems must not
tempting to use this property of PCR to now engaged in studies to provide such connect one PCR laboratory with anoth-
develop procedures that would allow data, not only for Native Americans but er using similar material. Yet despite
the reconstruction of even more exten- for other populations in the Old and such precautions, contamination contin-
sively damaged DNA than can now be New Worlds as well. ues to be a serious concern, especially
analyzed. Other experiments are direct- The objects of such investigations are when one tries to amplify DNA from
ed at repairing the damage in the old many. Comparisons of DNA sequences ancient human remains in which con-
DNA before it can be amplied by PCR. in dierent populations help workers tamination from modern humans can-
Yet it can be argued that the most im- not only to detect shared variants but not easily be ruled out.
portant advantage of PCR is its techni- also to estimate the degree to which For this reason, the rst study of old
cal simplicity, which made it possible to the sequences are related. This infor- DNA from populations focused not on
amplify the same sequence from dier- mation can be used to gauge how pop- humans but on a remotely related spe-
ent quagga extracts in just a few days. ulations are related to one another and ciesthe kangaroo rat. My colleagues
We and other laboratories could thus the time that has elapsed since they Kelley Thomas, Francis Villablanca and I
reproduce and verify the results: mo- shared an ancestor. One can even nd studied museum collections containing
lecular archaeology could for the rst signs of ancient expansions and con- 48 skins collected during the early part
time claim to be a respectable branch of tractions in population size, such as of this century from three populations
science. might have followed epidemics, fam- in the Mojave Desert. All the skins con-
ines, wars or migrations. tained mitochondrial DNA good enough

C
ould we extend the technique to Genetic studies strongly imply an to yield products 200 base pairs long.
materials that were thousands Asian origin for living Native Americans. After we had determined the sequenc-
of years old and were retrieved They also suggest that a fairly small es from the museum skins, we made
from archaeological excavations? To nd group of Asians colonized the New use of old eld maps and collected sam-
out, we tested the new technique on a World, perhaps in just a few waves of ples from kangaroo rats that live at the
human brain that, amazingly, had sur- immigration. For the rst time, we can same localities today. We discovered that
vived intact for 7,000 years at Little Salt begin to test these theories by analyz- the contemporary rats carry sequences
Spring, a sinkhole in Florida. William W. ing the DNA of ancient populations in identical or closely similar to the rats
Hauswirth and his collaborators at the the New World. This possibility is ex- that lived in the same places 60 to 80
University of Florida at Gainesville had tremely exciting in itself. It may also years ago and that the populations con-
already shown by other techniques that enable us to determine whether a pop- tain as many dierent sequences now
DNA was preserved in a brain found at ulation was replaced by another one or as then. Thus, the contemporary kan-
a similar site, presumably because the simply adopted a new material culture. garoo rats are the direct descendants
water was neutral in pH and contained Unfortunately, the study of old hu- of the previous populations, and no ma-

90 SCIENTIFIC AMERICAN November 1993 Copyright 1993 Scientific American, Inc.


wis came to New Zealand later than the
moas didprobably after the island
separated from Australia some 80 mil-
lion years ago. Was the kiwi, like its
Australian cousins, ightless when it
came? If so, it must have swum to New
fers from all other vertebrates at one position Zealand, perhaps following a chain of
(red), an error introduced by cloning and cor- islands that has since disappeared. Al-
rected by the polymerase chain reaction. ternatively, it may have come by air,
losing the power of ight only after ar-
riving on a landmass where no ground-
jor migrations or other disturbances dwelling carnivorous animals lived.
have occurred in the intervening de- Another insight emerging from the
cades. This work demonstrated that DNA sequences of the dierent genera
museum collections could be used to of moas is that they diverged from one
monitor the genetics of a population another rather recently. Cooper has sug-
over time. gested that this timing can be explained
The urgency of these developments by the geologic history of New Zealand.
stems from the realization that the Much of the island sank beneath the
study of both extinct and living popu- Victoria University in Wellington ana- waves 30 million years ago, apparently
lations is important now that so many lyzed both the bones and the mum- driving most of the ancestral moas to
species are in danger of extinction. It is mied soft tissues of moas found in extinction. Then, the rise of mountain
already known, for instance, that as a caves and swamps. He managed to am- ranges restored the islands surface area
population shrinks its genetic diversity plify mitochondrial DNA from four and provoked a burst of speciation in
declines, making it more susceptible to of the six genera that existed. The se- the survivors, which quickly occupied
infections and other stress, as in the quences reveal that the moas were all the new niches.
case of the cheetah [see The Cheetah closely related to one another but not

F
in Genetic Peril, by Stephen J. OBrien, to the kiwis, a group of ightless birds rom how far back in time can we
David E. Wildt and Mitchell Bush; SCI- that still exist on New Zealand. Kiwis expect to retrieve DNAfrom the
ENTIFIC AMERICAN, May 1986]. Studies are instead more closely related to the progenitors of modern humans
of ancient DNA may deepen our under- emus and cassowaries of Australia. some 200,000 years ago, from the
standing of how a gene pool changes This discovery indicates that the ki- founders of the hominid line some
over time and may show whether a re-
duction in genetic diversity generally
precedes extinctions.
At rst, workers dared to seek ancient
DNA only in soft tissues, which form
but a small fraction of the remains
turned up in archaeological excavations.
Then, in 1989, Erika Hagelberg and
Bryan Sykes of the University of Oxford
succeeded in extracting usable DNA
from ancient bones. Such DNA presum-
ably stems from osteoblasts and osteo-
clasts, cells that reside within the com-
pact bone and continuously remodel it
throughout life. Perhaps because DNA
binds to the minerals that make up
bone, skeletal remains may actually sur-
pass soft tissue as a repository for DNA.
A good example is provided by bones
of moas, an extinct group of ightless
birds of New Zealand. Some species
grew to 3.5 meters in height and 200
kilograms in weight; they lived by
browsing until humans arrived on the
island about 1,000 years ago and hunt-
ed them to extinction. Alan Cooper of

FLIGHTLESS BIRDS of New Zealand, the


kiwi and the extinct moa, are not direct-
ly related. The moa ancestor probably
arrived rst (A ), possibly when the is-
land separated from Gondwana some
80 million years ago. The kiwi ancestor
seems to have arrived later (B ), per-
haps by ying.

Copyright 1993 Scientific American, Inc. SCIENTIFIC AMERICAN November 1993 91


of our recent and more distant ances-
tors, recall the moas or quaggas to life
or even establish breeding farms for
dinosaurs?

I
t is my rm conviction that such
dreams (or nightmares) will never
be realized. We have no idea how to
piece together the millions of DNA frag-
ments that we extract from an animal
into chromosomes in a functional cell,
nor can we set in motion the thousands
of genes that regulate development. If
we cannot even take a fresh cell from
an adult vertebrate and use it to clone
another individual, how can we imag-
ine cloning an extinct species from the
otsam and jetsam of ancient DNA?
The resurrection of lost species will
remain beyond our power. The furthest
step toward reanimation that one can
imagine would be the isolation of a sin-
gle ancient gene. Such a gene, intro-
duced into existing species, might cre-
ate animals that mimic an aspect of an
extinct species. In this way, one could
GENETIC TREASURE TROVE is provided by mummies from the Americas, Europe test the function of an ancient gene.
and Egypt. The author is shown holding the foot of an Egyptian mummy similar to Yet such experiments do not in any
that which inspired his rst foray into the study of ancient DNA. sense preserve or re-create lost species
or ecosystems. Extinction is and will al-
ways be forever.
three million years ago or from the di- seum of Natural History. They and their What we can hope is that the study
nosaurs themselves? Certain physical collaborators study insects that have of ancient DNA will help us learn more
limits seem inescapable. In approxi- been entombed in amber, a dry environ- about the dynamics of genetic change
mately 50,000 years, water alone strips ment that may also be protected from in populations over time. With such
bases from the DNA and leads to the oxygen if the amber is deposited under knowledge we may be able not only to
breakage of strands into pieces so small suitable conditions. They have extracted understand better the history of our-
that no information can be retrieved DNA from insects imprisoned in amber selves and other species but also to
from them. Oxygen also contributes to about 40 million years ago. DeSalle and frame more rational strategies to limit
the destruction of DNA. Even in ideal his colleagues have shown that their se- the ongoing erosion of biodiversity.
conditionsin the absence of water and quences, from termites, are compatible
oxygen and at low temperatureback- with those from a modern termite.
ground radiation must nally erase all The next few years will undoubtedly
genetic information. show much activity in this eld. Above FURTHER READING
Despite these considerations, recent all, it will be crucial to nd out whether DNA SEQUENCES FROM THE QUAGGA, AN
results have encouraged hopes of go- other laboratories can conrm the re- EXTINCT MEMBER OF THE HORSE FAMILY.
ing further into the past than many be- sults reported for the very old DNA se- R. Higuchi, B. Bowman, M. Freiberger, O.
lieved possible. In 1990 Edward M. Gol- quences. If DNA sequences millions of A. Ryder and A. C Wilson in Nature, Vol.
enberg and Michael T. Clegg of the Uni- years old can indeed be determined re- 312, No. 5991, pages 282284; November
15, 1984.
versity of California at Riverside pub- producibly, one of the most important
ANCIENT DNA: EXTRACTION, CHARACTERI-
lished a DNA sequence from a magno- opportunities that would result is the ZATION, MOLECULAR CLONING AND ENZY-
lia leaf deposited in clay on the bottom ability to measure the rate of molecu- MATIC AMPLIFICATION. Svante Pbo in
of a lake in northern Idaho some 17 mil- lar evolution directly. Yet some prob- Proceedings of the National Academy of
lion years ago. They were able to ampli- lems will remain. It will always be hard Sciences, Vol. 86, No. 6, pages 1939
fy a fragment as long as 800 base pairs. to tell whether an old specimen is truly 1943, March 1989.
Pamela S. Soltis and Douglas E. Soltis an ancestor of a present-day species INDEPENDENT ORIGINS OF NEW ZEALAND
of Washington State University repeat- or whether the ancient and modern MOAS AND KIWIS. A. Cooper, C. Mourer-
Chauvir, G. K. Chambers, A. von Haesel-
ed this feat with specimens of another sequences merely shared an ancestor er, A. C. Wilson and S. Pbo in Proceed-
plant species found at the site, but ex- much further back in time. The practi- ings of the National Academy of Sciences,
tracts from many other leaves from the tioners of molecular evolution will thus Vol. 89, No. 18, pages 87418744; Sep-
same site have failed to yield DNA. The still be able to disagree fruitfully. tember 15, 1992.
clay was wet, however, and one won- If DNA provides a recipe for building DNA SEQUENCES FROM A FOSSIL TERMITE
ders how DNA could have survived the an organism, and ancient DNA stores IN OLIGO-MIOCENE AMBER AND THEIR PHY-
damaging inuence of water for so long. the recipe with delity, then a question LOGENETIC IMPLICATIONS. R. DeSalle, J.
Gatesy, W. Wheeler and D. Grimaldi in
Perhaps more promising are the re- naturally arises: Can we hope to reverse
Science, Vol. 257, pages 19331936; Sep-
sults of George O. Poinar, Jr., of Berkeley extinction by resuscitating vanished tember 25, 1992.
and Rob DeSalle of the American Mu- species? Could we clone identical twins

92 SCIENTIFIC AMERICAN November 1993


Copyright 1993 Scientific American, Inc.