AMERICAN JOURNAL OF PHYSICAL ANTHROPOLOGY 000:000–000 (2008)

Technical Note: Improved Ancient DNA Purification for

PCR Using Ion-Exchange Columns
Kijeong Kim,1,2,3 Kyung-Yong Kim,1,3,4 Eunhee Jeon,2,3 Ariunaa Togloom,2,3 Youn-Ock Cho,3
Min-Soo Lee,3 Gavaachimed Lkhagvasuren,3 Jee-Hye Choi,5 Dashtseveg Tumen,6
Ae Ja Park,7 Keun-Cheol Kim,8 Ki-Won Park,9 Jae-Hyun Kim,10 Maengseok Noh,11
Kwon-Jong Yoo,3,12 and Kwang-Ho Lee3,13*
1
Institute for Medical Sciences, Chung-Ang University, Seoul 156-756, South Korea
2
Department of Microbiology, College of Medicine, Chung-Ang University, Seoul 156-756, South Korea
3
Department of Cultural Properties, Chung-Ang University, Seoul 156-756, South Korea
4
Department of Anatomy, College of Medicine, Chung-Ang University, Seoul 156-756, South Korea
5
Department of Life Science, College of Natural Science, Chung-Ang University, Seoul 156-756, South Korea
6
Department of Anthropology and Archeology, School of Social Sciences, National University of Mongolia,
Ulaanbaatar, Mongolia
7
Department of Laboratory Medicine, Chung-Ang University College of Medicine, Seoul 156-756, South Korea
8
Department of Biology, Institute for Life Science, College of Natural Sciences, Kangwon National University,
Chuncheon 200-701, South Korea
9
DNA Analysis Section, National Institute of Scientific Investigation, Seoul 158-707, South Korea
10
Department of Archaelogy and Art History, College of Humanities, Donga University, Pusan 604-714, South Korea
11
Division of Mathematical Science, College of Natural Sciences, Pukyong National University,
Pusan 608-737, South Korea
12
Department of Philosophy, College of Liberal Arts, Chung-Ang University, Seoul 156-756, South Korea
13
Department of Life science, Chung-Ang University, Seoul 156-756, South Korea

KEY WORDS DNA purification; ancient bone; PCR amplification; ion-exchange columns; PCR
inhibitors

ABSTRACT A novel method of ancient DNA (aDNA)
purification was developed using ion-exchange columns to
improve PCR-amplifiable DNA extraction from ancient
bone samples. Thirteen PCR-resistant ancient bone sam-
ples aged 500–3,300 years were tested to extract aDNA
using a recently reported, silica-based aDNA extraction
method and an ion-exchange column method for the fur-
ther purification. The PCR success rates of the aDNA
extracts were evaluated for the amplification ability of the
fragments of mitochondrial DNA, a high-copy DNA, and
amelogenin, a low-copy DNA. The results demonstrate
that the further purification of silica-based aDNA extracts
using ion-exchange columns considerably improved PCR
amplification. We suggest that the ion-exchange column-
based method will be useful for the improvement of PCR-
amplifiable aDNA extraction, particularly from the poorly
preserved, PCR-resistant, ancient samples. Am J Phys
Anthropol 000:000–000, 2008. V 2008 Wiley-Liss, Inc.
C

The analysis of ancient DNA (aDNA) has become
increasingly popular in both archaeological and evolu-
tionary studies. PCR techniques are essentially required
for such studies due to the limitations of ancient samples
and available amounts of DNA. However, aDNA extrac-
tion suitable for successful PCR amplification is compli-
cated by several intrinsic factors: the aDNA is highly
degraded and damaged (Pääbo, 1989; Hagelberg and
Clegg, 1991; Handt et al., 1994; Höss et al., 1996;
O’Rourke et al., 1996; Keyser-Tracqui and Ludes, 2005;
Mulligan, 2005); presents in low copy number (Handt
et al., 1994; O’Rourke et al., 1996; Poinar et al., 1996; Yang
et al., 1998; Kumar et al., 2000); it is contaminated with
exogenous DNAs that originated from other organisms
and contemporary humans (Kolman and Tuross, 2000;
Pääbo et al., 2004; Calacal and De Ungria, 2005; Kemp
and Smith, 2005; Mulligan, 2005; Hunter, 2006); and it
coexists with a number of potential PCR inhibitors, such
as humic acid, fulvic acid, biologically degraded prod-
ucts, and collagen (Hänni et al., 1995; Kalmar et al.,
2000; Keyser-Tracqui and Ludes, 2005).
Because of these complexities, there have been a variety
of ancient DNA extraction methods attempted, such as
phenol-chloroform extraction (Blake et al., 1992; Kurosaki
et al., 1993; Faerman et al., 1995; Hänni et al., 1995) and
ethanol or isopropanol precipitation (Kurosaki et al., 1993;
Cattaneo et al., 1995, 1997; Hänni et al., 1995), a modified
ethanol precipitation using Dextran Blue (Kalmar et al.,
This article contains supplementary material available via the Inter-
net at http://www.interscience.wiley.com/jpages/0002-9483/suppmat.
Grant sponsor: National Research Institute of Cultural Heritage of
Cultural Heritage Administration (Conservation Technology Research
and Development Project).
*Correspondence to: Kwang-Ho Lee, Department of Life Science
and Department of Cultural Properties, Chung-Ang University, 221,
Heukseok-dong, Dongjak-gu, Seoul 156-756, South Korea.
E-mail: leemanse@cau.ac.kr
Received 1 August 2007; accepted 19 November 2007
DOI 10.1002/ajpa.20782
Published online in Wiley InterScience
(www.interscience.wiley.com).

C 2008
V WILEY-LISS, INC.
2 K. KIM ET AL.
TABLE 1. List of ancient human bone samples used in this study
Sex
Sample code Bone Mb Gc Excavation site Estimated agea Reference
d
KR007 left tibia ND ND Neukdo-dong, Sacheon Si, Early Iron Age (2nd c. Shim and Kim, 2001
Gyeongsangnam-Do, BC – 3rd c. AD)
Korea
KR032 left femur ND Fe Neukdo-dong, Sacheon Si, Early Iron Age (2nd c. Shim and Kim, 2001
Gyeongsangnam-Do, BC – 3rd c. AD)
Korea
KR062 right femur Mf F tumuli at Chuam-dong, 6th c. Silla kingdom Kim, 1994
Donghae Si, Gangwon-Do, (5th – 7th c. AD)
Korea
KR064 left femur F F tumuli at Chuam-dong, 6th c. Silla kingdom Kim, 1994
Donghae Si, Gangwon-Do, (5th – 7th c. AD)
Korea
KR067 left femur F M tumuli at Chuam-dong, 6th c. Silla kingdom Kim, 1994
Donghae Si, Gangwon-Do, (5th – 7th c. AD)
Korea
KR072 left femur M ND tumuli at Chuam-dong, 6th c. Silla kingdom Kim, 1994
Donghae Si, Gangwon-Do, (5th – 7th c. AD)
Korea
KR083 right femur M F tumuli at Chuam-dong, 6th c. Silla kingdom Kim, 1994
Donghae Si, Gangwon-Do, (5th – 7th c. AD)
Korea
KR151 left tibia ND M Baekgok-ri Mado-myeon, Chosun dynasty
Hwaseong Si, Gyeonggi- (14th – 19th c. AD)
Do
KR152 left femur ND F Singal-suji block, Gyeonggi- Chosun dynasty
Do, Korea (14th – 19th c. AD)
MN015 right tibia ND M Western Mongolia Bronze Age Tumen, 1978
(13th – 3rd c. BC )
MN032 right femur F ND Central Mongolia Mongol age Tumen, 1982
(13th – 15th c. AD)
MN225 rib M M Bulgan, Buregkhangai, Bronze Age Tumen, 2006
Mongolia (13th – 3rd c. BC )
MN384 right femur F F Sukhbaatar, Asgat, Mongol age Tumen, 2007
Mongolia (13th – 15th c. AD)
a
Based on the archeological findings of discovered artifacts during excavations such as coins, potteries and grave types.
b
Based on morphology.
c
Based on amelogenin sex marker DNA analysis in the present study.
d
Not determined.
e
Female.
f
Male.

2000), an agarose gel extraction (Fischer et al., 2001), the
use of microconcentrators (Hagelberg and Clegg, 1991;
Blake et al., 1992; Yang et al., 1998), Chelex-based meth-
ods (Walsh et al., 1991; Faerman et al., 1995), and silica-
based methods (Höss and Pääbo, 1993; Kurosaki et al.,
1993; Cattaneo et al., 1997; Evison et al., 1997; Prado
et al., 1997; Yang et al., 1998; Kemp et al., 2006; Rohland
and Hofreiter, 2007). Many of these methods were devel-
oped particularly for the removal of potential PCR inhibi-
tors to achieve successful PCR amplification.
Most recently, a highly optimized protocol for aDNA
extraction was introduced by Rohland and Hofreiter (2007)
based on the comprehensive comparison study of aDNA
extraction techniques to date. This protocol is attractively
simpler, quicker, and more economical than any methods
previously reported. It avoids the need for lengthy incuba-
tion for decalcification, use of detergents, phenol-chloroform
solvents, and expensive microconcentrators. Instead, it uti-
lizes silica beads directly in the lysis solution for aDNA cap-
ture and subsequent purification. We applied this new pro-
tocol to aDNA extraction from some ancient bone samples
that were previously PCR-resistant in our laboratory.
The method removed the brownish discoloration in
many of the DNA extracts, which is a marker of insuffi-
cient PCR inhibitor removal; it was not removed well in
our previous study. The PCR was considerably successful
in amplifying a small fragment of mitochondrial DNA
(mtDNA). Even with this method, however, some sam-
ples were still not amplifiable for the small and large
fragments of mtDNA, and many samples were not for a
fragment of amelogenin that is a low-copy gene. These
samples were amplifiable only after further DNA purifi-
cation using ion-exchange columns. The ion-exchange
column purification method has been used for DNA
extraction from soil organisms that are highly contami-
nated with many soil-originated PCR inhibitors (Tebbe
and Vahjen, 1993; Zaporozhenko et al., 2006). To our
knowledge, this method has not been used for DNA
extraction from the ancient samples. Herein, we intro-
duce an ion-exchange column method for the aDNA
extraction to improve PCR amplification.
MATERIALS AND METHODS
Contamination controls
To prevent potential contamination, all procedures
were carried out under sterile conditions, using gowns,
latex gloves, and mouth masks. All appliances, contain-
ers, and the work areas (laminar airflow surface, PCR

American Journal of Physical Anthropology—DOI 10.1002/ajpa

 IMPROVED ANCIENT DNA PURIFICATION FOR PCR 3
TABLE 2. Primers used for the amplification of mtDNA HV1 and amelogenin DNA fragments
Target Primer name Sequence (50 ?30 ) Product (bp) Reference
mtDNA HV1 F15989 CCCAAAGCTAAGATTCTAAT 263a Edson et al., 2004
R16251 GGAGTTGCAGTTGATGT Edson et al., 2004
mtDNA HV1 F15971 TTAACTCCACCATTAGCACC 440a Edson et al., 2004
R16410 GAGGATGGTGGTCAAGGGAC Edson et al., 2004
Amelogenin F_amel GGGCACCCTGGTTATATCAAC 201,b 195c Present study
R_amel AGGCCAACCATCAGAGCTT Present study
Amelogenin F_amXY GGTTATATCAACTTCAGCTATGAG 190,b 184c Present study
(nested) F_amY GGATTCTTCATCCCAAATAAAGT 92b Present study
R_amXY GCCAACCATCAGAGCTTAAAC Present study
a
Based on the revised Cambridge reference sequence (Andrews et al., 1999).
b
Male only.
c
Female only.

workstation) were cleaned by using an undiluted com-
mercial bleach UV-irradiated at 254 nm for at least 1 h.
All materials that can be applied to autoclaving, such as
mixer mill grinding jars, glassware, metalware, buffers,
and solutions were sterilized by autoclaving at 1218C for
at least 30 min. For plasticware that cannot be auto-
claved, presterilized disposables were purchased. It was
ensured that nonsterile reagents were not used and
were not introduced to all the materials used in this
work. All steps (bone cutting, surface removing, powder-
ing, DNA extraction, PCR preparation, and post-PCR
work) were carried out in a laminar airflow clean bench
in separate clean rooms. Throughout all manipulations,
sterile aerosol-barrier (filter tip) pipettes and pipette tips
were used. To monitor potential material- or worker-ori-
ginated human DNA contamination, reagent blanks
(negative controls) containing everything except bone
powders were included through the entire process.
Sample preparation
The ancient human bone samples used in this study
are shown in Table 1. The bones were fragmented using
sterile saw blades, and several millimeters of the bone
surface were removed using autoclaved sandpaper. To
eliminate any potential DNA contaminated by the bones,
we followed the protocol introduced by Kemp and Smith
(2005). Bone fragments were immersed in undiluted
bleach (
5.4% [w/v] sodium hypochlorite) for 20 min and
rinsed with sterile distilled water. The bones were then
air-dried under UV irradiation at 254 nm for 1 h within
a closed sterile cabinet. The sample-containing grinding
jars were immersed in liquid nitrogen for 10 min and
finely powdered using Mixer Mill MM301 (Retsch).
DNA extraction and purification
aDNA was extracted from the bone powders according
to a recently published protocol (Rohland and Hofreiter,
2007). Approximately 1 g of the sample and 0.5 mg/ml
proteinase K were used. In brief, the bone powders were
incubated in 10 ml extraction buffer (0.45 M EDTA,
0.5 mg/ml proteinase K) at room temperature overnight.
After centrifugation, the supernatant was added to
35 ml binding solution (5.5 M Guanidinium thiocyanate,
20 mM Tris, pH 4.8), 180 ll silica suspension were
added, and the pH was adjusted to 4.0. This mixture
was incubated for 3 h at room temperature under rota-
tion. After centrifugation, the supernatant was dis-
carded. The silica pellet was washed with 1 ml binding
solution and then transferred into a 2 ml tube. After
short centrifugation, the binding solution was removed.
The pellet was resuspended in 1 ml wash solution
(51.3% Ethanol, 125 mM NaCl, 1 mM EDTA, 10 mM
Tris, pH 8.0), and the supernatant was removed by short
centrifugation. The washing step was repeated. Subse-
quently, the silica pellet was air dried at room tempera-
ture for 15 min, resuspended in 200 ll 13 TE and after
8 min of incubation at room temperature, the silica was
collected by centrifugation at maximum speed (16,000g).
The supernatant, representing the extract, was trans-
ferred into a fresh tube. A half volume of the extract
was saved for PCR experiments and the other half was
further purified using ion-exchange columns (Genomic-
tip 20/G, Qiagen) as follows: 100 ll of silica extracts
were mixed with 2 ml QBT buffer (Qiagen), loaded onto
the equilibrated column with 2 ml of QBT buffer, washed
with 15 ml of QC buffer (Qiagen), and then eluted with
7 ml QF buffer (Qiagen). The elutes were concentrated
to about 200 ll using Amicon Ultra-15 Centrifugal Filter
Unit with Ultracel-30 membrane (Millipore). The concen-
trates were further concentrated and washed using
Microcon YM-30 Centrifugal Filter Unit (Millipore) twice
by adding 500 ll 13 TE to decrease the amount of the
salts and isopropanol derived from QF buffer. The final
concentrates (100 ll) were used for PCR experiments.
PCR
The PCR success rates of the silica-based method
alone and the additionally applied ion-exchange column-
based method were assessed based on the amplification
ability of two high-copy DNA targets and a low-copy one:
a 263-bp fragment of mtDNA hypervariable region one
(HV1), a 440-bp mtDNA HV1, and a 201- or 195-bp ame-
logenin DNA. Primers used in this study are shown in
Table 2. PCR amplifications were performed using
GeneAmp1 PCR System 9700 (Applied Biosystems) and
AmpliTaq Gold polymerase (Applied Biosystems) in a
20-ll reaction volume. For the 263-bp mtDNA HV1 amplifi-
cation, the reaction mix consisted of 2 ll 103 PCR buffer
(Applied Biosystems), 0.2 mM of each dNTP (Roche diag-
nostics), 2 mM MgCl2 (Applied Biosystems), 1 lM of
each relevant primer, 1 mg/ml BSA (New England Bio-
labs), and 0.8 U of AmpliTaq Gold polymerase. As the
templates, 3 ll, 1 ll, 1 ll of 1:10 dilution (1/10 ll), and
1 ll of 1:50 dilution (1/50 ll) of the DNA extracts of each
sample were tested. Amplification cycles consisted of an
initial denaturation step at 958C for 10 min, 40 cycles of
30 seconds at 958C, 1 min at 568C, and 1 min at 728C,
followed by a final extension step of 7 min at 728C. Five
microliters of PCR product was separated by electropho-

American Journal of Physical Anthropology—DOI 10.1002/ajpa

4 K. KIM ET AL.
TABLE 3. Results of duplicated PCR for mtDNA HV1 and amelogenin DNA using ancient DNAs extracted by using silica-based
method only and ion-exchange column method for further purification of the silica extracts

Template Sample code

PCR target DNA (ll) KR032 KR062 KR064 KR067 KR151 KR152 MN015 MN225 MN384 R.B.a D.W.b
Silica-based DNA extraction
1 mtDNA HV1 (263 bp) 3 –c 1d – – +e/2 + + 2/+ + – –
1 2/1 – – – + + + +/1 + – –
1/10f – 1/2 – – + 1/2 1/2 + + – –
1/50g – 1/2 – – – – – – + – –
2 mtDNA HV1 (440 bp) 3 – +/2 – – 2/1 2/+ 1/2 – + – –
1 +/2 – – – 1/2 + – +/1 + – –
1/10 – – – – 1/2 1/2 – – 1/2 – –
1/50 – – – – – – – – – – –
3 Amelogenin 6 – – – – – +/2 – – + – –
3 – – – – + 1/2 – +/2 + – –
1 – – – – 1/2 1/2 – + + – –
1/10 – – – – – – – – – – –
1/50 – – – – – – – – – – –
Ion-exchange column purification of silica extracts
4 mtDNA HV1 (263 bp) 3 + + + + + + + + + – –
1 + + + + 1/+ + + + + – –
1/10 – – + 1/2 1/+ + – + + – –
1/50 – – + – – 2/1 – 2/1 + – –
5 mtDNA HV1 (440 bp) 3 + + + + + + + + + – –
1 + + + + + + + + + – –
1/10 – – + + 1/+ 2/1 – – 1/+ – –
1/50 – – 1/2 – +/2 – – – – – –
6 Amelogenin 6 + + + + + + + + + – –
3 – – + + + 1/2 1/2 + + – –
1 – – – + – – +/2 + + – –
1/10 – – – – – – – – – – –
1/50 – – – – – – – – – – –

Boldface indicates the same PCR results when repeated; 1/2, 2/1, 2/1, or 1/1 indicates different results when repeated.
a
Reagent blank.
b
Distilled water.
c
PCR product of expected size is not visible based on agarose gel electrophoresis analysis
d
PCR product of expected size is strongly stained based on agarose gel electrophoresis analysis.
e
PCR product of expected size is relatively weakly stained based on agarose gel electrophoresis analysis.
f
1 ll of 1:10 dilution of the original DNA extract.
g
1 ll of 1:50 dilution of the original DNA extract.

resis on a 1.8% agarose gel containing ethidium bromide
and photographed under UV illumination. For the 440-
bp mtDNA HV1 amplification, the conditions were the
same as those for the 263-bp mtDNA HV1 amplification
except the use of the relevant primers and the annealing
temperature (608C). For the identification of the 440-bp
mtDNA HV1 PCR products, the PCR products were puri-
fied using Qiaquick PCR purification kit (Qiagen), and
the PCR products were bidirectionally sequenced.
For the amelogenin DNA amplification, the relevant pri-
mers and 1.2 U of AmpliTaq Gold polymerase were used.
Annealing temperature was 628C; 50 cycles were set up
for the amplification; tested template amounts were 6 ll,
3 ll, 1 ll, 1/10 ll, and 1/50 ll of the DNA extracts of each
sample, and the other conditions were the same as above.
For the identification of the amelogenin PCR products,
a nested duplex PCR was performed as follows: the reac-
tion mix (20 ll) consisted of 2 ll 103 PCR buffer, 0.2 mM
of each dNTP, 1.5 mM MgCl2, 0.25 lM of F_amXY primer,
0.5 lM of R_amXY primer, 0.5 lM of F_amY primer, and
0.6 U of AmpliTaq Gold polymerase. As the template, 1 ll
of 1:100 diluted amelogenin PCR product was tested.
Amplification cycles consisted of an initial denaturation
step at 958C for 10 min, 25 cycles of 30 sec at 958C, 30 sec
at 628C, and 30 sec at 728C, followed by a final extension
step of 7 min at 728C. Five microliters of PCR product
was used for the agarose gel electrophoresis analysis.
Sample reliability determination
It has been increasingly reported that during the an-
cient DNA extraction procedures, DNA contamination
might occur and lead to a biased data interpretation.
As a means of sample reliability assessment with
respect to contamination, we tested sequence reproduci-
bility of mtDNA HV1 440-bp fragment by replicated PCR
and sequencing. With our method, three out of 13 sam-
ples did not generate consistent sequence results. The
silica-based method gave the same results. Thus, those
samples were presumed to be already contaminated
before or during the silica-based DNA extraction proce-
dures and were removed from the study of the DNA
extraction method comparison. One sample (KR083) did
not produce the 440-bp fragment by using both the
methods. Nine out of 13 samples produced the consistent
HV1 440-bp sequences. None of these samples shared
the same sequence profile with any other sample or with
laboratory workers (Table S1). These samples were used
for the purpose of the DNA extraction methods compari-
son in PCR amplicability, and all the PCR experiments
were replicated. As a means of PCR data verification, we
repeatedly sequenced mtDNA HV1 440-bp amplicons
obtained from the duplicate PCR, and the sequences
were analyzed for mtDNA haplogroup determination to
see the phylogenetic significance of the bone samples.

American Journal of Physical Anthropology—DOI 10.1002/ajpa

 IMPROVED ANCIENT DNA PURIFICATION FOR PCR 5

Fig. 1. Agarose gel electrophoresis analysis of PCR products targeting amelogenin DNA fragment (201 bp) from ancient DNAs
extracted by using silica-based method alone (a) and ion-exchange column method for the further purification of the silica extracts
(b). Lane M, 100-bp size marker. Lanes of ancient bone samples: 1, KR032; 2, KR062; 3, KR064; 4, KR067; 5, KR151; 6, KR152; 7,
MN015; 8, MN225; 9, MN384; 10, reagent blank; 11, distilled water.

RESULTS AND DISCUSSION
The amplification of mtDNA HV1 263-bp fragment
using the DNA extracted by the silica-based method
alone [Table 3(1)] showed that six samples were amplifi-
able with 3 ll DNA extracts, six with 1 ll, six with 1/
10 ll, and two with 1/50 ll. Two samples were not ampli-
fiable with any tested amount of template DNA. Three
samples, KR032, KR151, and MN225, were not amplifi-
able or weakly amplifiable with 3 ll template DNA, but
were strongly amplifiable with smaller amounts of tem-
plate DNA. These results indicate that a smaller amount
of template DNA is more favorable for the PCR with
these DNA extracts, and that the potential PCR inhibi-
tors could not be sufficiently removed by using the silica-
based method alone. The amplification of this target
DNA was more successful with the further purified DNA
extracts using the ion-exchange column method than
with DNA extracts from the silica-based method only
[Table 3(4)]. With 3 or 1 ll DNA template, all the sample
DNAs tested were amplifiable; with 1/10 ll template, six
samples were amplifiable; and with 1/50 ll template,
four samples were amplifiable. The two samples, KR064
and KR067, were not amplifiable using the DNAs puri-
fied by the silica-based method alone, but they were
amplifiable using the DNAs further purified by this
method. These results clearly demonstrate that some re-
sidual PCR inhibitors remaining after the use of silica-

American Journal of Physical Anthropology—DOI 10.1002/ajpa

6 K. KIM ET AL.

based method were removed by the further purification
using the ion-exchange column-based method.
The amplification of mtDNA HV1 440-bp fragment
using the DNA extracts of the silica-based method alone
[Table 3(2)] showed that five samples were amplifiable
with 3 ll template DNA, five with 1 ll, three with 1/
10 ll, and none with 1/50 ll. The amplification of this
target DNA from the DNA extracts further purified by
using the ion-exchange columns was more successful
than the amplification from those extracted by the silica-
based method alone [Table 3(5)]. With 3 or 1 ll DNA
template, all the tested sample DNAs were amplifiable;
with 1/10 ll, five samples were amplifiable; and with 1/
50 ll, two samples were amplifiable. It appeared that an
increased amount of template DNA was required for the
amplification of this larger mtDNA target. These results
demonstrate that the larger target is less favorable for
Fig. 2. Agarose gel electrophoresis analysis of the nested
duplex PCR products using the amelogenin PCR products of an-
cient bone samples as the template for their identification. The
samples on lanes 4, 5, 7, and 8 were identified as male, showing
an additional smaller male marker product (92 bp) as well as a
sex-common product (184 bp). Lane M, 100-bp size marker.
Lanes of ancient bone samples: 1, KR032; 2, KR062; 3, KR064;
4, KR067; 5, KR151; 6, KR152; 7, MN015; 8, MN225; 9, MN384;
10, distilled water.
amplification than the smaller one because of the degra-
dation of aDNA. There is also a report that PCR amplifi-
cation of that large DNA was used for the ancient bone
analysis (Keyser-Tracqui et al., 2003).
The amplification results of the amelogenin DNA
showed a remarkable difference between the two DNA
extraction designs. The amplification of the amelogenin
DNA using the DNA extracts of the silica-based method
alone showed that two samples were amplifiable with
6 ll template DNA, four with 3 ll, four with 1 ll, and
none with 1/10 ll and 1/50 ll [Table 3(3), Fig. 1a]. Over-
all, only four samples were amplifiable for the ameloge-
nin from the DNA extracted by this method. Three sam-
ples appeared more successful in the amplification with
the smaller amount of template DNA on the agarose gel
electrophoresis profile. In contrast, all the samples tested
were amplifiable for this gene with 6 ll of the DNA
extracts further purified by the ion-exchange column-
based method [Table 3(6), Fig. 1b]. The identity of the
amelogenin PCR products was confirmed by performing
nested duplex PCR with 1 ll of 1:100 dilutions of the
PCR products as the template. Four samples were deter-
mined as male producing two bands and five as female
producing only one larger band on the agarose gel elec-
trophoresis (Table 1, Fig. 2).
Overall, out of the nine DNA extracts of the silica-based
method only, seven samples were amplifiable for the 263-
bp mtDNA HV1 and the 440-bp mtDNA HV1, and four
samples were amplifiable for the amelogenin DNA. All
nine sample DNA extracts obtained from the additional
ion-exchange column purification were amplifiable for the
three target DNAs. The two sample DNAs, KR0064 and
KR0067, extracted using the silica-based method only,
were not amplifiable for all three target DNAs; these sam-
ples were amplifiable for all the target DNAs only after
the introduction of the ion-exchange column method.
When considering the number of samples producing
consistently positive amplification in PCR duplicates of
the two methods, the addition of the ion-exchange col-
umn method provided much more successful results
than the silica-based method alone in all the three target
PCR amplifications (Table 4). In addition, the results of
the former method did not appear PCR-inhibited by the

TABLE 4. Successful PCR amplification rates of silica-based method only and ion-exchange column method for further purification
of the silica extracts in nine ancient samples with varying amounts of template DNAs for three different amplification targets
mtDNA HV1 263-bp DNA mtDNA HV1 440-bp DNA Amelogenin
Template [ 1a 11b [1 11 [1 11
DNA (ll) Total Total Total Total Total Total
S.B.O.c 6 NDd 7 ND 6 ND 7 ND 3 2 4 1 2
3 6 4 5 1 4 2
1 6 5 5 3 4 2
1/10e 6 3 3 0 0 0
1/50f 2 1 0 0 0 0
I.E.C.g 6 ND 9 ND 9 ND 9 ND 9 9 9 9 9
3 9 9 9 9 7 5
1 9 9 9 9 4 3
1/10 6 5 5 4 0 0
1/50 4 2 2 0 0 0
a
Number of positive samples that produce a electrophoretically detectable product of expected size in at least one of duplicates.
b
Number of positive samples that produce a electrophoretically detectable product of expected size in both of duplicates.
c
DNA extracts using silica-based method only.
d
Not determined.
e
1 ll of 1:10 dilution of the original DNA extract.
f
1 ll of 1:50 dilution of the original DNA extract.
g
Additionally purified silica extracts using ion-exchange column purification method.

American Journal of Physical Anthropology—DOI 10.1002/ajpa

 IMPROVED ANCIENT DNA PURIFICATION FOR PCR
use of increased amounts of DNA templates, or less suc-
cessful with the decreased amounts of templates. Those
of the latter method appeared negatively influenced by
the increasing amounts of templates, presumably due to
the PCR inhibitory effect of potential impurities in the
extracts. From the viewpoint of PCR reproducibility
(PCR amplification in duplicated reactions shows the
same negative or the same positive results), the former
method was also superior to the latter one, showing
93.7% (134 identical results out of 143 duplicates) and
85.3% (122 out of 143) reproducibility, respectively.
These results suggest the higher reliable positive PCR
amplification is expected from particularly PCR-resistant
ancient DNAs with the former method than with the lat-
ter one.
It was likely that the PCR outcome of a template
resulted from a competition between the amounts of tar-
get DNA and PCR inhibitors coexisting in the specified
volume of template used in the reaction mixture. Some
samples required a smaller volume of the template for
PCR success, whereas some samples required a higher
volume. This phenomenon was observed in the PCR
results from both methods, but was less prominent with
the purified DNAs using the ion-exchange columns. The
PCR success rates seemed rather to be template-volume
dependent with this method. This may indicate that the
ion-exchange column method considerably contributed to
the removal of the residual PCR inhibitors from the an-
cient DNA silica extracts; it would be easier to deter-
mine the optimal volume of template required for the
successful PCR with the ion-exchange column-purified
DNA extracts.
It is an important issue to be solved before proceeding
to aDNA analysis that contamination from the modern
DNA should be avoided during the procedures and
aDNA should be authentic. Generally, there are three
major possible contamination sources: modern DNAs
attached to the ancient bone samples that were not com-
pletely removed, laboratory worker DNAs during the
experiments, and sample DNAs crosscontaminated
between the samples, including PCR products. The
newly introduced ancient DNA extraction method
includes increased purification steps due to the addition
of the ion-exchange column method. This may add to the
chance of contamination and potentially lead to a spuri-
ously increased PCR success rate. However, the repli-
cated sequencing showed that the duplicated sequences
were identical to each other, there were no identical
sequences in the samples, and no identical sequences to
those of laboratory workers were found (Table S1). The
sequences of amplicons from the silica-based method
were identical to those from the additional ion-exchange
column method in four available samples. These might
support that additionally increased contamination did
occur despite increasing purification steps in the present
study, and the improved PCR results shown by using the
developed method were not attributed to the increased
contamination.
Three samples excluded in the comparison study of
PCR success might not have been contaminated during
ion-exchange column procedures but might have been
contaminated before or during the silica-based method
application. This is based on the fact that the inconsis-
tent sequencing results of these samples were also found
in the silica-based method only. It is reasonable that the
additionally purified extracts with the ion-exchange col-
umn method also showed inconsistent sequencing results
American Journal of Physical Anthropology—DOI 10.1002/ajpa
7
in these samples. Many of the bone samples were heav-
ily discolored (brown), evidencing poor preservation; and
there was no amplifiable DNA from the multiple reagent
blanks and water controls. It is more likely that the
decontamination procedures were insufficient for the re-
moval of contaminated DNAs.
The sequences of mtDNA HV1, except for one sample,
were among the reported haplogroups of their excavated
countries. Exceptionally, the haplogroup of one ancient
bone sample, KR067, was determined as haplogroup T2
that originated in the Near East approximately 46,500
years ago but is now most prevalent in Europe (Richards
et al., 1998). This haplogroup is also reportedly found
among the Middle East countries including Iran (for-
merly Persia) (Abu-Amero et al., 2007). The sample
KR067 was excavated from a tumulus of Silla in Korea,
and its archeologically estimated age is the 6th c. Silla
kingdom. This age is very close to the golden age of Silla
(7th c.–9th c.), when Koreans actively started the trade
with the Persian and Arabian world, according to Ko-
rean archeologists (personal communication). Unfortu-
nately, the ethnic morphological estimation was not
available due to the severe destruction of the skull.
When considering the facts that the DNA sequence was
reproducible, that the haplogroup is reportedly not found
among modern Koreans, that no foreigner has been
involved since the moment of excavation, and that his-
torically the age is very close to the actively trading pe-
riod with the Middle East, the sequence result seems not
to be attributed only to the contamination, and further
authentic investigation is on its way. Interesting findings
could be drawn in conducting further various molecular
and archeological studies on its origin and the relation-
ship with the Silla kingdom.
Our results demonstrate that the additional ion-
exchange column purification of silica-extracted aDNA
considerably improved PCR performance, presumably by
an enhanced removal of potential PCR inhibitors. We
suggest that the developed method will provide a useful
tool for an improved PCR-amplifiable ancient DNA
extraction, particularly from poorly preserved ancient
samples in the fields of archaeology, anthropology, and
forensic science. The additional procedures inevitably
require more attention for the contamination risk.
ACKNOWLEDGMENTS
This study, which forms a part of a larger project, has
been achieved with the support of the Conservation
Technology Research and Development project, which
has been hosted by the National Research Institute of
Cultural Heritage Administration. We express our grati-
tude to it.
LITERATURE CITED
Abu-Amero KK, Gonzalez AM, Larruga JM, Bosley TM, Cabrera
VM. 2007. Eurasian and African mitochondrial DNA influ-
ences in the Saudi Arabian population. BMC Evol Biol 7:32.
Andrews RM, Kubacka I, Chinnery PF, Lightowlers RN, Turn-
bull DM, Howell N. 1999. Reanalysis and revision of the Cam-
bridge reference sequence for human mitochondrial DNA. Nat
Genet 23:147.
Blake E, Mihalovich J, Higuchi R, Walsh PS, Erlich H. 1992.
Polymerase chain reaction (PCR) amplification and human
leukocyte antigen (HLA)-DQ alpha oligonucleotide typing on
8 K. KIM ET AL.
biological evidence samples: casework experience. J Forensic
Sci 37:700–726.
Calacal GC, De Ungria MC. 2005. Fungal DNA challenge in
human STR typing of bone samples. J Forensic Sci 50:1394–
1401.
Cattaneo C, Craig OE, James NT, Sokol RJ. 1997. Comparison
of three DNA extraction methods on bone and blood stains up
to 43 years old and amplification of three different gene
sequences. J Forensic Sci 42:1126–1135.
Cattaneo C, Smillie DM, Gelsthorpe K, Piccinini A, Gelsthorpe
AR, Sokol RJ. 1995. A simple method for extracting DNA
from old skeletal material. Forensic Sci Int 74:167–174.
Edson SM, Ross JR, Coble MD, Parsons TJ, Barritt SM. 2004.
Naming the dead—confronting the realities of rapid identifi-
cation of degraded skeletal remains. Forensic Sci Rev 16:64–
89.
Evison MP, Smillie DM, Chamberlain AT. 1997. Extraction of
single-copy nuclear DNA from forensic specimens with a vari-
ety of postmortem histories. J Forensic Sci 42:1032–1038.
Faerman M, Filon D, Kahila G, Greenblatt CL, Smith P, Oppen-
heim A. 1995. Sex identification of archaeological human
remains based on amplification of the X and Y amelogenin al-
leles. Gene 167:327–332.
Fischer GF, Fae I, Mann D, Kriks D, Jager W, Platzer W, Mayr
WR, Volc-Platzer B. 2001. An HLA class-II allele frequent in
Eskimos and Amerindians is found in the Tyrolean Ice Man.
Ann Hum Genet 65:363–369.
Hagelberg E, Clegg JB. 1991. Isolation and characterization of
DNA from archaeological bone. Proc Biol Sci 244:45–50.
Handt O, Höss M, Krings M, Pääbo S. 1994. Ancient DNA:
methodological challenges. Experientia 50:524–529.
Hänni C, Brousseau T, Laudet V, Stehelin D. 1995. Isopropanol
precipitation removes PCR inhibitors from ancient bone
extracts. Nucleic Acids Res 23:881–882.
Höss M, Jaruga P, Zastawny TH, Dizdaroglu M, Pääbo S. 1996.
DNA damage and DNA sequence retrieval from ancient tis-
sues. Nucleic Acids Res 24:1304–1307.
Höss M, Pääbo S. 1993. DNA extraction from Pleistocene bones
by a silica-based purification method. Nucleic Acids Res 21:
3913–3914.
Hunter P. 2006. Ancient DNA research goes nuclear. A new
technique to extract sequence data from nuclear DNA may
reveal exciting new insights into evolution and phylogeny.
EMBO Rep 7:136–139.
Kalmar T, Bachrati CZ, Marcsik A, Rasko I. 2000. A simple and
efficient method for PCR amplifiable DNA extraction from an-
cient bones. Nucleic Acids Res 28:E67.
Kemp BM, Monroe C, Smith DG. 2006. Repeat silica extraction:
a simple technique for the removal of PCR inhibitors from
DNA extracts. J Archaeol Sci 33:1680–1689.
Kemp BM, Smith DG. 2005. Use of bleach to eliminate contami-
nating DNA from the surface of bones and teeth. Forensic Sci
Int 154:53–61.
Keyser-Tracqui C, Crubezy E, Ludes B. 2003. Nuclear and mito-
chondrial DNA analysis of a 2,000-year-old necropolis in the
Egyin Gol Valley of Mongolia. Am J Hum Genet 73:247–260.
Keyser-Tracqui C, Ludes B. 2005. Methods for the study of an-
cient DNA. Methods Mol Biol 297:253–264.
Kim J-H. 1994. The burial process of the Chuam-Dong tumuli,
Dong-Hae city, Korea: in the case of the KA-31 location B
tomb. J Soc Kyushu Prehist Ancient Cultur Stud (KOBUNKA
DANSO. JAPAN) 33:205–221.
American Journal of Physical Anthropology—DOI 10.1002/ajpa
Kolman CJ, Tuross N. 2000. Ancient DNA analysis of human
populations. Am J Phys Anthropol 111:5–23.
Kumar SS, Nasidze I, Walimbe SR, Stoneking M. 2000. Brief
communication: discouraging prospects for ancient DNA from
India. Am J Phys Anthropol 113:129–133.
Kurosaki K, Matsushita T, Ueda S. 1993. Individual DNA iden-
tification from ancient human remains. Am J Hum Genet
53:638–643.
Mulligan CJ. 2005. Isolation and analysis of DNA from archaeo-
logical, clinical, and natural history specimens. Methods
Enzymol 395:87–103.
O’Rourke DH, Carlyle SW, Parr RL. 1996. Ancient DNA: meth-
ods, progress, and perspectives. Am J Hum Biol 8:557–571.
Pääbo S. 1989. Ancient DNA: extraction, characterization, mo-
lecular cloning, and enzymatic amplification. Proc Natl Acad
Sci USA 86:1939–1943.
Pääbo S, Poinar H, Serre D, Jaenicke-Despres V, Hebler J, Roh-
land N, Kuch M, Krause J, Vigilant L, Hofreiter M. 2004.
Genetic analyses from ancient DNA. Annu Rev Genet 38:645–
679.
Poinar HN, Höss M, Bada JL, Pääbo S. 1996. Amino acid race-
mization and the preservation of ancient DNA. Science
272:864–866.
Prado VF, Castro AK, Oliveira CL, Souza KT, Pena SD. 1997.
Extraction of DNA from human skeletal remains: practical
applications in forensic sciences. Genet Anal 14:41–44.
Richards MB, Macaulay VA, Bandelt H-J, Sykes BC. 1998. Phy-
logeography of mitochondrial DNA in western Europe. Ann
Hum Genet 62:241–260
Rohland N, Hofreiter M. 2007. Comparison and optimization of
ancient DNA extraction. Biotechniques 42:343–352.
Shim B-K, Kim, J-H. 2001. A Significance of Nuk-Do Site,
SaCheon. The Research report of Graduate School of Kyushu
University, Japan. p 139–153.
Tebbe CC, Vahjen W. 1993. Interference of humic acids and
DNA extracted directly from soil in detection and transforma-
tion of recombinant DNA from bacteria and a yeast. Appl En-
viron Microbiol 59:2657–2665.
Tumen D. 1978. Paleoanthropology of western Mongolia. Studia
Archeologica, Institute Historiae Academiae Scientiarum,
Mongolia. T-VII (in Russian).
Tumen D. 1982. Paleoanthropological study of human remains
from Kara Korum (Mongolian Period—13–15th c.). Studia
Archeologica, Institute Historiae Academiae Scientiarum,
Mongolia. T-XI (in Russian).
Tumen D. 2006. Paleoanthropology of ancient popultion of Mon-
golia. Mongolian J Archaeol Anthropol Ethnol 2. p 90–108.
Tumen D. 2007. Anthropology of human remains from Mongo-
lian Period in Sharga mountain site, Asgat sum, Sukhbaatar
aimag. Research report of eastern Mongolia: Archaeology and
Anthropolog Perspective. p 15–16.
Walsh PS, Metzger DA, Higuchi R. 1991. Chelex 100 as a me-
dium for simple extraction of DNA for PCR-based typing from
forensic material. Biotechniques 10:506–513.
Yang DY, Eng B, Waye JS, Dudar JC, Saunders SR. 1998. Tech-
nical note: improved DNA extraction from ancient bones using
silica-based spin columns. Am J Phys Anthropol 105:539–543.
Zaporozhenko EV, Slobodova NV, Bulygina ES, Kravchenko IK,
Kuznetsov BB. 2006. Method for rapid DNA extraction from
bacterial communities of different soils. Mikrobiologiia 75:
127–134.