8/3/2019 J. Biol. Chem.-1989-Pbo-9709-12
1/4
Minireview Vol. 264, No . 17 , Issue of June 15, p
9709-9712,1989THE JOURNALF BIOLOGICALHEMISTRY0 1989 by The American
Society for Biochemistryand Moykular Biology, Inc.Printed in
U.S.A.
Ancient DNA andthe PolymeraseChain ReactionTHE EMERGINGFIELD O F
MOLECULARARCHAEOLOGY*
Svante Piiiibos,Russell G . Higuchii, andAllan C . WilsonFrom
the Depar tmentof Biochemistry, University ofCalifornia, Berkeley,
California94720 and the Depar tmentof Hu m an Genetics, Cetus
Corporation,Emeryville, California94608
Molecular evolution is a historic process through whichgenes
accumulate changes due to stochastic events aswell asselective
processes. Students of molecular evolution sufferfrom the frustrat
ion of trying to reconstruct his historicprocess from only a
knowledge of the present-day structureof genes. Until recently,
there has been no hope of escapingthis time trap. However, advances
in molecular biologicaltechniques have enabled us to retrieve and
study ancientDNA molecules and thus to catch evolution red-handed.
Inconsequence, we can now study the genealogical relationshipsof
extinct species and vanished populations. In addition, itseems
likely tha t we shall be able to monitor fast geneticprocesses such
as recombinational events. Our review dis-cusses older attempts to
obtain molecular genetic data fromarchaeological remains as well as
recent achievements andemerging vistas.
Studies of Ancient ProteinsThe first indications that molecular
genetic informationmight persist in ancient materials were early
demonstrationsthat the peptide bond can last for up to lo 8 years
in fossilshells and bones (1-3) and that subcellular detail
implyingthe survival of ribosomes and chromatin is evident in
insectsfrom 40 million-year-old amber (4). Indeed, these
findingsinspired the hope that genetic information should be
retriev-able from the amino acid sequences in ancient remains,
andsubstantial efforts over the past two decades went into
suchendeavors.Unfortunately, the major proteins in bone (collagen)
andshell (conchiolin) are likely to be genetically
uninformativebecause collagen has a repetitious primary structure
and isencoded by multiple genes ( 5 ) whereas conchiolin is a
com-plex mixture of proteins whose genetic basis is unknown
(6).
Second, the proteins in ancient remains are structurally
het-erogeneous because of post-mortem modifications (7,8).Even in
exceptionally well preserved remains, such as fro-zenmuscle from an
extinctSiberian mammoth, extensivemodifications were evident from
elemental analysis, electronmicroscopy, and amino acid analysis of
the 40,000-year-oldtissue (9, 10). In the case of albumin, one of
the most stableglobular proteins known in animal tissues (8),only
about 2%of the mammoth molecules could dissolve in water, and 80%of
the lat ter were modified in charge, size, or antigenicity (9).*
Work done in the laboratory of A. C. W. received support fromthe
National Science Foundation and the National Institutes ofHealth.$
Supportedby a long-term fellowship (ALTF 76-1986) from theEuropean
Molecular Biology Organization.
Immunological Distances-The first comparisons of the pri-mary
structure of a gene product of an extinct species to thatof living
species were achieved indirectly by using polyclonalantisera raised
against a homogenate of mammoth muscle.Such an antiserum was tested
for its ability to form precipi-tates and complement-fixing
lattices with a panel of nativealbumins from several living species
(9, 10). The reactionswere strong with Indian andAfrican elephant
albumins, weakwith sea cow albumin, and weaker still with other
mammalianalbumins. Since the cross-reaction specificity of
theanti-mammoth serum matched exactly those of antisera LO thepure
albumins of elephants, the nativealbumins of thesethree species
(mammoth, Indian elephant, and African ele-phant) are nearly
identical in primary sequence (cf. Ref. 11).Radioimmunoassay, which
does not require lattice forma-tion and thus does not demand that a
cross-reactive antigenbear more than one antigenic site, confirmed
the close rela-tionship of elephant and mammoth albumins (12); it
alsoplaced the extinct Tasmanian wolf within the genealogicaltree
for extant carnivorous marsupials (12) and Stellers seacow within
the tree for extant sea cows (13) on the basis oftests with
antisera to their albumins. However, this method,like other
univalentmethods, can be less reliable as a predictorof sequence
divergence than is microcomplement fixation(14).Immunological
methods are especially likely to give mis-leading results when
employing antisera that are raised andlater also tested against
mixtures of poorly defined antigens.For this reason, we consider
most antigenic studies reportedon ancient materials to be of
questionable genetic value.Gene Frequencies-At the population
level, anthropologistsand paleozoologists are often interested in
determining thefrequencies of alleles at polymorphic loci in
ancient popula-tions. Determinants of the AB0 system have been of
mostinterest (15, 16) because they are present on nearly all
cellsin mammalian tissues. However, blood group serology per-formed
on ancient tissues has many pitfalls because Jf thepossibility of
differential degradation of polysaccharide anti-gens as well as
contamination of the old tissues by plant andmicrobial antigens
which may cross-react with the antibodyor bind to blood group
determinants (cf.Ref. 17). In wellpreserved human remains,
serological typing of proteins en-coded in the major
histocompatibility complex (HLA anti-gens) maybe more informative
than ABO. HLA has theadditional advantage of fewer risks of
anomalous cross-reac-tions caused by proteins from other organisms
(e .g. Ref. 18).Nevertheless, the inherent difficulty of
interpreting tht, reac-tion of serological reagents with antigens
that are modifiedby unknown processes remains.
DNA in Old Tissue RemainsFollowing the realization that DNA may
survive in ancienttissues (19-21), DNA has been extracted from a
wide varietyof such remains (whether d r y , frozen, or preserved
wet inpeat), ranging in age up to 45,000 years (Table I).
AlthoughDNA can be extracted from most soft tissue remains that
arewell preserved morphologically, post-mortem modificationsmade it
hard to clone such DNA in bacteria.Higuchi et al. (22) succeeded in
cloning mitochondrial DNAsequences from the extinct quagga, a
member of the horsegenus (Equus).This represented the first
retrieval of phylo-genetically informative DNA sequences from a
museum spec-imen and allowed the quagga to be placed into
aphylogeny of
9709
8/3/2019 J. Biol. Chem.-1989-Pbo-9709-12
2/4
9710 Minireview:Ancient DNATABLE
DNA from eight k inds of ancient remainsAncient DNA has usually
been detectedy its stainin g with ethid-ium bromide n
electrophoretic gels and its a ctivity as a template thatcan direct
the incorporation of radioactive nucleotides into DNA inthe
presence of DNA polymerase and random primers. Hybridizationwith
DNA from a modern species is used t o determ ine w hether theDNA
detected originates from the species under study or from
acontaminating source. The specific hybridization test has not
beenapplied to the oldest plant remains or the insects in
amber.Sample Maximumage Study
yearsDry remainsMuseum skins 140 Refs. 22, 43"*bNatural animal
mummies 10,000Human mummies 5,000 Refs. 24, 25, 27Plants 45,000
Refs. 41, 44, 45Insects in amber 26 million Ref. 21'
Mammothuscle 40,000 Refs. 21,6'Pickleduseumpecimens 100
*Humanrain 8,000 Refs. 26.32
a R. Higuchi and B. Bowman, unpublished data.
Frozen remainsWet remains
W. K. Thomas, R. H. Thomas, andS. Paabo, unpublished obser-R.
Higuchi and A. Wilson, unpublished data.P. Basasibwaki,T. Kocher,
A. Meyer, and A. Wilson, unpublished
vations.data.a 'uaggaBurchellGrew ZebrasMounfarnWtld ASSHall
ASSDornesftcPrzewalskrr I orses
8 7 6 5 4 3 2 1 0L"l1.I-JPercent dwergence
FIG.1. Phylogenetic tree relatingmitochondrial DNA fromthe
extinct quagga to mtDNAs from other members of thehorse genus.
Percent divergence refers to th e estimated number ofbase
substitutions/hundred base pairs compared for any pair of theeight
species. These estimates come from restriction mappingf thewhole
mitochondrial genome 42),sequencing of tw o cloned
segments(23),and, in th e quagga case, sequencingof two
enzymaticallyampli-fied segments (30).t he ho r se a nd its relat
ives, the zebras and assesFig. 1) (23).Th e result s argue fo r a
close relat ionship between the uaggaan d Burchell' s zebra and
again st th e iew th at th equagga isrelated more closely to the
dom estic horse. The cloning ofrepetitive DNA sequences f rom an
anc ient Egypt ian mum my(24) showed tha t far lder DNA may be
preserved inclonableform and tha t nuc lear DNA as well a s
mitochondrial DNAma y p ersist for illenia.Little progress followed
these initial uccesses because, ikeall macromolecules extrac ted
from ld tissue remains , ancientDNA is heavily modified. This ma
nife sts tself by a reductionin size down to an average of only a
few hundre d ba se pa irs(e.g. Refs. 25 an d 26) and by an abu nda
nce f lesions, such asbaseless sites, oxidized pyrim idine s,
andcross-links, m any ofwhich can be attribu ted to oxid ative
processes (27). Th esemodifications ar e so extens ive th a t less
than 1% of th e DN Amolecules extracted from museu m sp ecimens
orrchaeologi-cal f inds can e expected to be undamaged.In a t
temptsto clone ancient DNA in living bacteria, one isconfronted
with this vast excess of dam aged DNA in twoways. First, extremely
low cloning efficiencies are obta inedsince the major i ty f the
vecto rmolecules becomes ligated to
c m XIIYIT I G A T C5-Lqww S A AT CLC rn~1An ATA nt TA RQm MT
ATA m n c T.I. /C l a d quw .....................................
.T. .... "
l l ' ,bn "/c............................................FIG.2.
Mitochondrial DNA sequences obtained in two waysfrom skin of a
140-year-old museum specimen of the extinctquagga. Left p a n e l
(base sequences), the upper result comes from
directly sequencing the p roduct obtained when PCR was applied
to aDNA preparation from quagga skin. The middle result comes
fromsequencing a cloned fragment of unamplified quagga DN A (22).Th
ethat th e DNA was cloned from highly purified mitochondrial
DNAlower result was obtained in the same way as th e middle one
exceptof a Burchell's zebra (23). Dots indicate identity with the
amplifiedquagga sequence. The arrow denotes a position at which a
cloningartifact (a T residue) occurs in th e cloned quagga
sequence. Rightp a n e l (photograp h), part of a sequencing gel h
ighlighting with a narrow the position in the amplified sequence at
which a C (ratherthan a T) residue occurs in the quagga. Modified
from Ref. 30.dam aged molecules whose modifications preclude
replicationin bacteria. Second, som e dam agedmolecules become
repairedin bacter ia , of tenby processes that are erro r-prone and
thuslikely to distor t the inform ation extracted from cloned
se-quences. Recent advan ces stemm ing from Kary Mullis's
in-vention of th e polymerase chain reaction (28) have
eliminatedsome of the problems causedy bo th the ow cloning
efficien-cies and clon ing artifacts arising from modifications in
an-cient DNA.
!l'he Polymerase Chain Reaction inolecularArchaeologyT h e
polymerase chain reaction (PCR)' can amplify prese-lected
segmentsof DN A up to quan ti t ies hich permit directsequencing,
star t ing from extremely small amou nts of DN Aor
evensinglemolecules (for a review, see Ref. 29). T h eamplification
is do ne w ith tw o syn theticligodeoxynucleotide
primers,eachabout25bases long, a thermostable DN Apolymerase,
and the fourdeoxyribonucleotide triphosph ates.The f irst pr imer
matches par t of the Watso n strand at oneend of the segmen t,
while the second p r imer matches theCrick strand at the o ther
end. Since the 3' end s of th e twoprimers po int toward each other
, repeated cycles of heatingan d cooling ead to a chain eaction,
i.e. anexponent ia lsynth esis of man y copies of the specific segm
ent boun ded bythe two pr imers.PC R is an ideal tool to amplify a
small number of intactancient DNA molecules present in a vast
excess of damagedmolecules. In additionto i ts abil i ty o detect
extremely smallquanti t ies of DNA, PCR has the advantage of be ing
an invitro system, which has no capacity for repair or
misrepair.For example, inbacterialclones of quagga mitoch
ondrialDNA, two replacement substi tutions were found when
thequagga equences were comp ared to otherver tebrate se-quences
(22, 23) . Thes e posit io ns in the quagga were latershow n no t
to differ from he general verteb rate sequencewhen they were
directly sequenced from amplif ication prod-ucts ((30) Fig. 2). The
replacements observed in the clonedsequences were hus due to clonin
g ar t ifacts,ossibly resultingfrom misrepairof dam aged m
olecules.In contrast , dur ing enzymaticmplification, most
damagedmolecules will either not be replicated at all, e.g. due to
inter-or intramolecular cross-links, or will be at a replicative
dis-advantage because many lesions, such as baseless sites,
slowdown the DN A polymerase. Inta ct molecules will th us
amplifypreferentially. Som e damaged molecules will of course
haveonlymino r lesions, such asdeaminated bases,which can
'The abbrev iations uaed are:PCR, polymerase chain reaction;
bp,base pair(s).
8/3/2019 J. Biol. Chem.-1989-Pbo-9709-12
3/4
Minireview:Ancient DNA 9711generate replication errors without
retarding replication.These errors a s well as those introduced by
the polymerase atundamaged sites will be present in the final
population ofmolecules produced by PCR. However, since each error
isspecific to one molecule in the starting population (and
itsdescendants), its contribution to the result, uiz. the
sequencedetermined on the whole descendant population of
amplifiedmolecules, s ikely to be negligible. Such errors would
beencountered only if one were to clone individual moleculesfrom
the final population before carrying out the sequencingreactions
(30,31). An additional advantage of PCR is that itsspeed allows for
easyeproduction of results.
Ancient DNA Sequences Revealed via PCRUsing PCR and direct
sequencing, it has been possible toobtain mitochondrial sequences
from a 7000-year-old brainexcavated in Florida (32). Preserved
brains exist in associa-tion with human skeletal remains at several
sites in Florida(26) and owe their excellent state of preservation
to anaerobicand neutral conditions in the waters of Florida peat
bogs.Extracts from the archaic brain contained DNA that
allowedamplification of 140-bp-long mtDNA ragments. Longer
am-plifications were unsuccessful. This is in sharp contrast
tocontemporary DNA, where ragments of up to 1kilobase andlonger
amplify efficiently. It seems that the damage presentin old DNA
causes a strongly inverse correlation betweenamplification
efficiency and sizeof the amplificationproduct,which sets a limit
to the ength of amplifications hat can beperformed.Three
informative mitochondrial polymorphisms werestudied in the Florida
brain. One is a direct repeat of 9 bpwhich occurs n only one copy
in many Asians s well as someNative Americans(32,33). Furthermore,
a Him11 site atposition 13259, shown to be absent in many
Amerindians (34),and a HaeIII site a t position 8250 were amplified
and se-quenced. From the sequences obtained, the archaic
Floridiancould be comparedo themtDNAs of more than 100 present-
day Native Americans. This ancient sequence does ot matchany of
the hree mitochondrial lineages found to date inAmerica. Further
work is in progressto characterize present-day Amerindian
opulations as well as further rchaic Floridabrains, and these
studies can be expected to yield a fullerpicture of the population
structure and genealogical historyof Amerindians.The same
procedures revealed DNA sequences from ncienthumans that are
preserved in the form of mummies. Forexample, a 4000-year-old
Egyptian priest was showno carryan unusual D-loop sequence (27). t
is not yet known if thiswas a common genotype in ancient Egypt and
whether it isrepresented today in Egypt or elsewhere.Another area
where paleomolecular biology is producinginformation is zoology. In
addition to the quagga discussedabove, the 40,000-year-old Siberian
mammoth (also referredto above) has yielded PCR products
contaminated with hu-man sequences, presumablyoriginating from
handling of themammoth after its discovery. Fortunately, the
mammothsequences wereseparable from the human sequences and,
asexpected, proved o be closely elated to those of elephants?Since
mtDNA is present in many copies/nucleated cell, itcan be assumed
that this high copy number facilitates itssurvival and retrieval.
In fact, attempts to amplify specificnuclear single-copy genes from
ancient remains (e.g. part ofthe @-globin gene fromhuman mummies)
have been unsuc-cessful, presumably due to their lower abundance in
DNAextracts. Furthermore, the fast evolution and maternal modeof
inheritance of mtDNA make t ideal for studying ancestor-descendant
relationships (35,36).
R. Higuchi andA. Wilson, manuscript in preparation.
Authenticity of Amplified SequencesContamination by Modern
DNA-The main concern perti-nent to the amplification of ancient DNA
sequences is con-tamination of the DNA extracts or reagents by
contemporaryDNA. When extinct species are studied, such
contaminationcan often be detected by mere inspection of the
sequenceswhen a phylogenetic analysis fails to place the species
understudy close to its biological relatives. For example, the
au-thenticity of the DNA sequences obtained by bacterial
cloningfrom the quaggawas demonstrated by the fact that theyproved
the quagga to be a close relative of extant zebras butnot identical
to any of those tested (22,23). Similarly, mtDNAsequences from the
Siberian mammoth show the mammothto be closely related to elephants
but not identical to eitherelephant species.Although contaminating
sequences of human origin areeasily detected in ancient remains of
nonhuman species, con-tamination by nonhuman sequences may be arder
to detect.Therefore, the phylogenetic criterion of authenticity
discussedabove cannot be solely relied upono distinguish
contaminat-ing DNA.Accordingly, the following additional criteria
must be set
up and adhered to in order to detect any source of
contami-nating DNA. Wehave found it imperative always to do
controlextracts (i) in parallel with the extracts of the old
specimensin order to detect contamination in solutions and
reagents(32, 37). Furthermore, several independent extracts (ii)
fromevery individual should be prepared, and the sequences
ob-tained should be unambiguousand identical.The strong inverse
correlation between amplification effi-ciency and size of the
amplification product (iii), which isobserved for ncient but not
odern DNA, an to some extentserve as a further criterion of
authenticity. In our experience,archaeological remains generally do
not yield any productsabove 150 bp in size, whereas better
preserved specimens,
Lag phase:I *-A b
Exponential phase:A
B4Amplification of the region boundedby A and B
to make >lob copiesFIG. . Concept of jumping PCR. n this
hypothetical exam-ple, two primers ( A and B ) use undamaged parts
of five damagedtemplates to amplify a mosaic product. In an
amplification reactionwhere each template DNA s so damaged that no
molecules allowheDNA polymerase to proceed directly from one primer
site (A , blue)to the other ( B , ellow),the primers will be
extended duringhe firstPCR cycle up to points where lesions (green)
or ends of fragmentscause the polymerase to stop. During subsequent
cycles, these ex-tended primers can annealo other template
molecules and be furtherextended. After a sufficient number of
cycles the two primers have
grown so long that their 3 ends overlap and a full-length
double-strandedmolecule is formed. This moleculecannow erveas
atemplate for a conventional chain reaction.
8/3/2019 J. Biol. Chem.-1989-Pbo-9709-12
4/4
9712 Minireview: Ancient D N Asuch as museum skins, may allow
for the amplification of up plants (41). A particularly important
goal for the future is toto 500-bp pieces. Sequences longer than
this have invariably refine the polymerase chain reaction so that
it becomes POS-proved to originate from contamination of the
specimens by sible to study ancient nuclear single copy genes as
well asmodern DNA. DNA sequences from bones, teeth, shells, and
amber.When the three above criteria (i, ii, and iii) are fulfilled,
agiven sequence is considered likely to be of ancient origin.
Acknowledgments-We thank numerous colleagues for discussionErrors
because of Post-mortem Chnges-As noted above, and sharing
unpublished data with US, in particular C. E. Beck, D.specific
errors stemming from post-mortem changes are not M. Irwin, T. D.
Kocher, C . Orrego,E. M. Prager, W. K. Thomas, andexpected to
predominate in an amplified population of mole-cules. If they were
to predominate and thus cause incorrectbe affected in a predictable
way. The substitutions would beIn fact, however, the ratio of
silent to replacement changes in 3. Curry, G. B. (1988) in
Molecular Euolution and the ~ o ~ ~ i lecord (Broad-
F. X. Villablanca, as well as E. Myers for expert graphical
work.REFERENCESsequences to be determined, the pattern Of
f3ubstitution would 1. Wyckoff, R, W, G, (1972) he Biochemktry o f
~ n i m l ~ ~ ~ i h ,cientech-
distributed at random with respect to position within codons. 2.
Weiner, s.,Lowenstam, H. A., and Hood, L. (1976) h o c . Natl.
Acad. Sei.DNA from th e extinct quagga " Ompared to modern DNA 4.
Poinar, G. O ., Jr., an d Hess, R. (1982) Science 216 ,1241 -
1242from the extant mountain zebra is 10 to 0. If the diff~ ~ences
. Yamada, Y., Awedimento , E., Mudryj, M., Ohkubo, H., Vogeli, G.,
Irani ,
nica, Bristol, United KingdomU. S. A. 73 , 2541-2545head, T. W.,
ed) pp. 20-33, Knoxville, TN
had occurred at random, a ratio of 2 to 8 would have been M.,
Pastan, I. , and de Crombrugghe, B. (1980) Cell 22 ,887 -
892expected. 6. Weiner, S. (1983) Biochemistry 22,4139-41457.
Gillespie, J. M. (1970) Science 170 ,1100 - 1102Mosaic Sequences
via Jumping PCR-An additional prop- 8. Sensabaugh, G. F., Jr.,
Wilson, A. C., and Kirk, P. L. (1971)Int. J . Biochern.erty Of PC R
may affect the authenticity Of the amp1ified 9. Prager, E. M.,
Wilson, A.C., Lowenstein, J. M., andSarich, V. M. (1980)2 ,545 -
568sequences. If no template molecules spanning the entire seg-ment
defined by the primers exist in the old extract, the 10 . Prager,
E. M., Taylor, D. A., Sarich, v. M.3 and Wilson, A.C. (1981)
inScience 209 ,287 - 289amplification may in fact start from
shorter fragments of the Magadan Baby Mammoth (Vereschagin, N. K.,
and Mikhelson, V.M.,DNA that are complementary to one or the other
of the 11 . Benjamin, D.C., Berzofsky, J. A., East, I. J., Gurd, F.
R. N., Hannum, C.,
eds) pp. 178-190, Nauka, Leningrad, U. S. S. R.primers (Fig. 3 )
. These fragments serve as templates in the Leach,,S. J.,
Margoliash, E., Michael, J. G., Miller, A., Prager, E. M.,first
extension, and thepartially extended primers can in the Reichhn,
M., Sercarz, E. E., Smlth-Gd1, S. J., Todd, P. E., and Wilson,A. C.
(1984) Annu. Reu. Immunol. 2 ,67 - 101next cycle be further
extended upon hybridization to other 12 . Lowenstein, J. M.,
Sarich, V. M., and Richardson, B. J. (1981) Naturefragments of the
region that is to be amplified. Only when 13 . Rainey, W. E.,
Lowenstein, J. M., Sarich, V.M., and Magor, D.M. (1984)2 9 1 , 4 0
9 - 4 1 1Naturwksemcha ten71 ,586 - 588extended primer roducts
overlap with their 3' ends will a 14-18conventional exponential
mplification reaction ensue. These 15 . Harrison, R. G., Connolly,
R. c., nd Abdalla, A. (1969) Nature 2 2 4 , 3 2 5 -initial steps
may allow the amplification of regions that are 16 . Berg, K.,
Rosing, F. W., Schwartzfischer, F., and Wischearth, H. (1975)longer
than the ongest intact molecule in an extract. Homo 26, 148-153The
building of a mosaic sequence via PCR poses no prob- 17 . Flaherty,
T., and Haigh, T. J. (1986) in Science in Egyptology (David, A.R.,
ed) pp. 379-382, Manchester Unlversity Press, Manchester, Unitedlem
in the Of mtDNA sequences for 18. Hansen, H. ., and Giirtler, H.
(1983) A m . J. Phys. AnthropoL 6 1 , 4 4 7-Kingdomwhich homoplasmy
is the rule (35). However, in the case ofchromosomal genes from
ancient remains of heterozygous 19 . Wan% G. H., andLu, c. C. 1981)
S h e w w u HUQ Hsw h S h e w wu Li452individuals, "jumping PCR'
may generate erroneous se- 20. phabo, s. 1984) D~ Altertum 3 0 ,
213-218Chin Chon3 9 , 7 0quences tha t result from recombination
during amplification. 21 . Higuchi, R., and Wilson, A. C. (1984)
Fed. Proc. 4 3 , 1 5 57
enough extension steps have been performed so that the two 14 .
Hoffmann, D., a n d v a n Regenmortel, M. H. V. (1984) J.Mol.Euol.
2 1 ,326
22. Higuchi, R. G., Bowman, B., Freiberger, M., Ryder, 0. A.,
and Wilson, A.Vistas forMolecularArchaeology 23 . Higuchi, R. G.,
Wrischnik, L.,, Oakes, E., George, M., Tong, B., andC. (1984)
Nature 312 ,282 - 284Wilson, A. C. (1987)J.Mol. Euol. 26 , 283-
287The recently achieved ability to study DNA from museum 24 .
Paabo S. (1985) Nature 314 ,644 - 645specimens and archaeological
finds via PCR (27, 30, 32, 38) 25 . Piiabo: s. 1986) cold Spring
Harbor Symp. Quant. Biol. 61 ,441 - 446
opens up the possibility of studying molecular evolution
byactually going back in tirne and directly approaching DNA 27 .
Piiiibo, S. (1989) Proc. Natl. Acad. Sci. u.S.A. 86 ,1939 -
1943sequences that are ancestral to their present-day counter- 28 .
Saiki, R. K., Scharf, s.,Faloona, F.,Mullis, K. B., Horn, G. T.,
Erlich, H.parts. This allows us to address many questions involving
the 29 . White, T . J., Arnheim, N., and Erlich, H. A. (1989)
Trends Genet. 5, inidentity and relationship of extinct species to
other extinct 30 . p-bo, s., nd wilson, A. c. (19%) Nature 33 4 ,
387 -388and extant species. In these cases, even rare examples of
31 . Saiki R. K. Gelfand D.H., Stoffel,S., Scharf, S.J., Higuchi,
R., Horn, G.single preserved specimens are of p ea t value. Second,
the 32 . Piabo, s.,Gifford, J. A,, and Wilson, A. C. (19%) NLlcletc
A C ~es. 1 6 ,evolution of populations can be studied and will most
often 9775-9787involve the comparison of ancestral and descendant
popula- 33 . Wrischnik, L. A,, Higuchi, R. G., Stoneking, M.,
Erlich, H. A., Arnbeim,tions, attributing changes in gene
frequencies to factors such 34 . Wallace, D.C., Garrison, K., and
Knowler, W. C. (1985) A m . J . Phys.as variations in population
size* and 35 . Wilson, 1. ., Cann, R.L., Carr, S . M., George, M.,
Gyllensten, U . B.,genetic drift. Helm-Bychowskl, K. M. Higuchl
R.G., Palumbi S. R. Prager E M.,In addition to Organisma' and
population questions, prom- 36 . Wilson: A. C., Zimmer, E. A.,
Prager, E. M., and Kocher, T. D. (1989) inising avenues of research
on molecular evolution may now be The Hierarchy of Life (Fernholm,
B., Bremer, K., and Jornvall, H., eds)We refer to the study Of
those mo- 37 , Pa&o, S,(1989) in PCR-Pmtocok &Appl icat ,p
.-A Moratory ! a n a l ,lecular genetic processes that may occur
rapidly enough to be (Innis, M. , Gelfand, D., Sninsky, J.,
andWhlte, T., eds) Academlc Press,accumulation of base
substitutions, which in nuclear and 39 . Jeffreys, A. J., Royle, N.
J., Wilson, v., and Wong, Z. (1988) Nature 332.organelle genomes is
generally less than 1% of sequence 278-281divergence/1O5 years.
Faster processes include recombination 40 . Gyllensten, U. B.,
Jakobsson, S., Temrin, H., and Wilson, A. C. (1989)in minisatellite
sequences which cause apparent mutation 41 . Helentjaris, T. (1988)
Maize Genetics COOP.News Letter 62 ,104 - 105Nucleic Acids Res.
17,2203-2214frequencies on the order of 10-2~locus~genera tion (39,
40). 42 . Geor e M., J r and Ryder 0. . (1986) Mol. Biol. Euol.
3,535-546Other cases where rapid evolution may allow molecular
proc- 44 . Rogers, S. O ., and Bendich, A. J. (1985) Plant Mol.
Biol. 6 , 6 9 - 7 6esses to be observed directly can be provided by
human : ~~~~~~&1 9 a ) camp, B k h m ,parasites (especially
viruses) and he evolution of domestic Physiol. 81B, 1045-1051
26 . Doran, G. H., ickel, D. N., Ballinger, W. E., Jr., Agee, 0.
F., Laipis, P. J.,and Hauswirth, W. W. (1986) Nature 323 ,803 -
806A., and Arnheim, N. (1985) Science 230 ,1350 -
1354pressT.,'Mullis', K. B., in d Erlich, H. A. (1988) Science 2 3
9 , . 4 8 7 ~ 4 9 1
N., and Wilson, A. C. (1987) Nucleic Acids Res. 15 ,529 -
542Anthro 01 68 ,149 - 155Sage R. D., and Stoneking, M. (1b85)
Biol. J. Linh. S O C . ' ~ ~ ,76-400p 407 419, Elsevier Science
Publi shers B.V., AmsterdamNew York, in pressstudied On a time Of
lo4 ears' in contrast to the 'low 38 . Rollo, F.,Amici,A., Salvi,
R., and Garbuglia, A. (1988)Nature 3 3 6 , 7 7 4
43 . Hou&:P., and'Braun, M. i. 1988) Auk 105 ,773 - 776
