_TH

  

l
w
\

HHIMWHI

WIHWNH|lll|\|(|mll\l\ll\\ll

3

1|

N
OLD

  
  

_(D

LIBRARY
Michigan State
University

This is to certify that the
thesis entitled

GENETIC ANALYSIS OF A MONUMENTAL STRUCTURE WITHIN

THE KAMENICA, ALBANIA TUMULUS

presented by

Stephanie L. Rennick

has been accepted towards fulfillment
of the requirements for the

 

 

MS. degree in Forensic Science
7 / "'

 

Major Professor’s Signature

?//7/0 5’

Date

MSU is an Affirmative Action/Equal Opportunity Institution

—--o-oa—o—o-Q—--o-o-n—u-u-u-a-o-n-.o ..

 

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECAIJED with earlier due date if requested.

 

 

 

 

DATE DUE g” DATE DUE DATE DUE
”HGHng’ T572309
App 9 2 2929 -EP 2 0 2009
. 5 F1 '0 410

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2/05 c:/CIRC/DateDue.lndd-p.15l

 

GENETIC ANALYSIS OF A MONUMENTAL STRUCTURE WITHIN THE

KAMENICA, ALBANIA TUMULUS
By

Stephanie Lynn Rennick

A Thesis

Submitted to
Michigan State University
In partial fulfillment of the requirements
For the degree of

MASTERS OF SCIENCE
School of Criminal Justice

2005

ABSTRACT

GENETIC ANALYSIS OF A MONUMENTAL STRUCTURE WITHIN THE

KAMENICA, ALBANIA TUMULUS
By
Stephanie Lynn Rennick

Analysis of skeletal material can be a useful endeavor in many forensic
investigations. By adding DNA analysis on skeletal material, it may be possible to
provide information that cannot be obtained from anthropological assessment. A
prominent example of this is the determination of relatedness among either several
skeletons, or between skeletons and living individuals. The analysis of skeletal remains
within the tumulus at Kamenica, Albania provided the opportunity to examine both the
relatedness of individuals within the tumulus, and to compare them to individuals living
in Kamenica today. The focus of the analysis was in an area of the tumulus, designated
Monumental Structure 3, in which individuals were buried in rock and surrounded by
large rock walls. The interest was whether or not the individuals in this structure were
maternally related to one another, which was estimated using mtDNA analysis. The DNA
sequences showed that all but four of the individuals shared a strong possibility of
maternal relatedness, in that they had a haplotype of 263-6. These individuals were then
compared to modern day individuals, who showed much more diversity in sequence, and
only had one individual with a sequence consistent with the ancient haplotype. It was also
found that both modern and ancient Albanians had European haplogroups, with

haplogroup H being present in around half of the individuals.

ACKNOWLEDGMENTS

I would like to thank the many individuals that helped to make this project
possible. Virginia Clemmer, Lindsey Murray, and Michael Mutolo all contributed data
that was used to reach the conclusions within this paper. Dr. David Foran was a vital part
of this project, by providing cements and ideas to both the project and the writing. Dr.
Todd Fenton was also vital to the project because of his ability to provide an
anthropologist’s view on the subject and for collecting all of the material. I would also

like to thank Dr. Christina DeJong for her comments and review of this manuscript.

iii

TABLE OF CONTENTS

LIST OF TABLES ............................................................................ vi
LIST OF FIGURES ........................................................................... vii
INTRODUCTION ............................................................................ 1
History of Albania .................................................................... 3
Tumuli in Albania ..................................................................... 4
The structure of the tumulus at Kamenica ......................................... 5
Use of mitochondrial DNA analysis on ancient remains ........................ 9
Complications and precautions associated with ancient DNA ................. 11
MtDNA haplogroup analysis for the classification of individual .............. 16
Project goals ............................................................................. 19
MATERIALS AND METHODS .............................................................. 20
Bone and buccal sample preparation .............................................. 20
DNA extraction ......................................................................... 22
DNA amplification ................. . .................................................. 23
DNA sequencing ..................................................................... 25
SNP analysis for haplogroup H ..................................................... 26
Data analysis ......................................................................... 27
RESULTS ...................................................................................... 28
Bone preparation ..................................................................... 28
DNA extraction ....................................................................... 28
PCR amplification of regions to be sequenced or used in SNP analysis ....... 29
Sequences of mtDNA from bone samples ........................................ 32
MtDNA sequences of buccal samples ............................................... 35
Assessment of haplogroups ......................................................... 35
DISCUSSION ................................................................................. 38
Sample analysis ....................................................................... 38
Mutation effects ...................................................................... 39
Comparison of mtDN A regions .................................................... 4O
Maternal relatedness ................................................................. 41
Modern inhabitants of Kamenica ................................................... 44
Conclusions ........................................................................... 45
APPENDIX A: Sequences from Bone Samples .......................................... 46

APPENDIX B: Consensus Sequences for Each Bunal 51

iv

APPENDIX C: Sequences for Buccal Samples ........................................... 54

BIBLIOGRAPHY ............................................................................. 56

LIST OF TABLES

Table l: Skeletal samples used in analysis ................................................ 21
Table 2: Amplified regions of bone samples ............................................. 33
Table 3: Number of ambiguous bases in each consensus sequence .................... 34
Table 4: Sequences for each bone in region HV2 ......................................... 46
Table 5: Sequences for each bone in region HVl-l ...................................... 47
Table 6: Sequences for each bone in region HVl-l ...................................... 49
Table 7: Consensus sequences in HV2 ..................................................... 51
Table 8: Consensus sequences in HVl—l ................................................... 52
Table 9: Consensus sequences in HV1—2 ................................................... 53
Table 10: Sequences for buccal swabs inHV2 ............................................ 54
Table 11: Sequences for buccal swabs in HVl ............................................ 55

vi

LIST OF FIGURES

Figure 1: Map of the bottom layer of the tumulus ........................................ 6
Figure 2: Map of the top layer of the tumulus ............................................. 7
Figure 3: Photograph of initial grave circle ................................................ 8
Figure 4: Photographs of monumental structures ......................................... 9
Figure 5: Deamination of cytosine .......................................................... 13
Figure 6: Misincorporation of thyamine ................................................... 14
Figure 7: Human mtDNA migrations ...................................................... 18
Figure 8: Single base primer extension process ........................................... 18
Figure 9: An example of typical semi-nested PCR results .............................. 30
Figure 10: Successful amplification after the first round of PCR ....................... 30
Figure 11: Example of inhibition in lane 6 ................................................ 31
Figure 12: An example of SNP results ..................................................... 37

vii

Forensic investigations often involve the analysis of skeletal remains. In these
instances, physical anthropologists may be called upon to use skeletal material to
estimate characteristics such as an individual’s sex, age, population ancestry, and stature,
as well as provide information to the cause or manner of death based on trauma found on
the bones. Further, individual identification is often possible when skeletal remains can
be compared to ante-mortem X-rays or dental records; however, when not available the
chance of identification is limited. In these instances, DNA analysis may be appropriate.
Identification of remains using DNA may be as direct as comparing their profile with
items known to have come in contact with the missing, such as a razor or toothbrush.

If a direct reference sample is not available, identification of remains can be
attempted through comparative analysis with relatives, taking advantage of the heritable
nature of DNA. Beyond identifying remains, genetic analysis of relatedness can serve
other purposes as well, including population and evolutionary investigations. For
instance, relatedness may be determined among several skeletal samples, or between
skeletal remains and living individuals. Determination of relatedness can be on the level
of a family unit, or it may encompass a broader regional identification.

O’Rourke et al. (2000) listed several studies in which DNA analysis is used to
determine relatedness of skeletal samples. DNAs from different Native Americans have
been analyzed and categorized into groups to try and determine similarities or possible
migrations of tribes from different regions. The founding population of Easter Island was
determined to be Polynesian based on DNA evidence. Individuals from the Linzi site in
China were sampled and analyzed across various time points to help explore the site’s

history (Kaestle and Horsburgh, 2002). By determining whether the same genetic

frequencies were maintained over time, one could deduce whether the population had
remained constant throughout time.

Nesser et al. (1995) discussed how DNA from many sources, including bone, can
be used to determine the sex of an individual. This can be especially useful on subadults
because they lack the skeletal characteristics anthropologists use to assess sex, but can
also be used for adults if the sample is fragmentary and the bones more commonly used
for sex determination are missing (Kaestle and Horsburgh, 2002). Even social structure
may potentially be inferred from genetic analysis. For example patrilocal societies (those
in which males stay in a location and females disperse) would be expected to have less
diversity in male inherited markers. On the other hand, matrilocal societies (those in
which females stay in a location and males disperse) should have less diversity in
maternally inherited markers.

Combining these types of analyses can be useful for studying ancient societies.
Determination of the origin of inhabitants, and whether or not they are different from the
people currently living in the area, can provide more information on the history of the
society. Comparison with nearby cities or countries may provide information on how the
groups interacted or help to show differences between the regions. Analysis of burial
grounds can shed light on the burial practices of ancient societies and how graves were
organized. All of this information can help to provide more insight into the way that

ancient societies lived.

History of Albania

The area located around the Adriatic Sea is often considered the ‘cradle of
Western civilization.’ The countries in this region include Greece, Macedonia, Italy,
Serbia and Albania; many of which, like Greece and Italy, have been extensively studied.
Albania, however, has been less studied due to opposition to foreign exploration by the
previous government. However, in the early 1990’s, the government began a transition to
a multiparty democracy, and this gave British and American scientists an opportunity to
learn about this area’s rich history.

With Greece directly to the south, Macedonia to the east, Serbia to the north and
Italy just across the Adriatic Sea, Albania had influences from many directions; which is
supported by the artifacts from this region showing similarity to those from Greece and
Italy. Another outside influence on Albania was a group of Indo-European shepherds
who migrated into the area near the end of the 11th century BC. or the beginning of the
12th century BC. They apparently did not push out or destroy the native people living in
Albania, but instead integrated their culture with the natives and created a collection of
tribes that all spoke the same language (Cambridge 1982). This group of people has come
to be known as the Illyrians.

The most prevalent form of burial for these people were large mounds, or tumuli,
which have given scientists a wonderful resource for learning more about the ancient
Albanians and where they came from. Bejko (1994) discussed many tumuli in southern
Albania, in which researchers have already found indications of influence from the Ionian

Islands and Thessaly, both in Greece. They have also found influences from Macedonia

in burial mounds in the Ionian Islands. Since Albania lies between these regions,
individuals most likely would have traveled through Albania to get from Macedonia to
the Islands, which suggests a possible influence from Macedonia as well. However, other
areas like Badher and Kros, along the Western shore of Albania, contained only local
pottery and appeared to be uninfluenced by surrounding areas, providing examples of
unaltered Albanian culture. These findings show that although Albania was in contact
with other areas of the world, it was still able to maintain some individualizing

characteristics.

Tumuli in Albania

When the Illyrians came to Albania, they also brought the idea of tumuli, which
became the prominent form of burial throughout the Bronze Age, ranging from
approximately 2100 BC to 1100 BC. (Cambridge, 1982). These are often found in a
high and visible place, and it is most common to find them in groups of 2 — 5, however,
single tumuli are also present. On average, a tumulus is 2 — 6 m high and 15 - 50 m in
diameter, and contains anywhere from 18 to over 300 graves. Common features include
central and outer rings made of mid-size stones, an inner tumulus covered in stone, and a
central grave, which may be under the floor of the tumulus. The composition is generally
a mixture of soil and stone. Burials are most commonly single, primary inhumations,
although some double burials do exist. The position of the individuals is either in a flexed
or fetal position, or in a laid out extended position; however, there appears to be no

distinct preference (Bejko, 1999).

The structure of the tumulus at Kamenica

The tumulus at Kamenica is the largest in Albania, with a diameter of more than
40 m and a center height of 3 m. It is estimated to contain about 800 graves, and is
thought to have been utilized over a time period from the 11th to 6th century BC,
encompassing the Late Bronze Age to Early Iron Age. The arrangement of many of these
graves can be seen in Figures 1 and 2. The tumulus was started with a grave circle about
15 m in diameter that contained 35 burials (Figure 3). The bodies from this initial circle
were all buried in soil and a double ring of large rock encompassed the entire grave
circle. This was thought to be the founders of the tumulus and one person in particular
(burial 395), who was placed at the center of the circle, is thought to be the first burial.
The mound continued with burials on t0p of this circle. These were no longer enclosed
within a large circle of rock, but still seemed to stay in the same area as the original
circle. This continued for several centuries and added about 200 burials to the tumulus.

Around 650 BC, nearly 100 years before the tumulus ceased to be used, the
burial pattern at the tumulus changed. Instead of burying individuals in soil, each
individual was covered with rocks. At first individuals were also encircled by rock walls
that connected together to make large formations, coined “monumental structures”.
Archaeologists uncovered two different areas with these rock wall structures and
designated them Monumental Structure 1.1 and Monumental Structure 3 (refer to Figures
1 and 2). The area between these two structures, however, is being preserved, leaving it
unclear as to whether or not these two structures are connected. On top of these structures

were more burials that continued to be covered in rock but were no longer contained

Figure 1: Map of the bottom layer of the tumulus

wsraoo

 

This map depicts the bottom burials from the tumulus, with the original circle located in
the top left comer of the picture. Two additional structures (discussed above) are located
off to the right and the bottom of the original circle.

Figure 2: Map of the top layer of the tumulus

 

This map depicts the top burials from the tumulus, with the original circle located in the
top left comer of the picture. Two additional structures (discussed above) are located off
to the right and the bottom of the original circle.

Fi ure 3: Photo ra h of initial _rave circle

     

This photograph shows the initial grave circle from the tumulus. Along the outside of the
grave circle is the double ring of large rock. Also in the center of the picture is Burial
395, which is thought to be the first burial of the tumulus.

within the large rock walls. It is unknown whether the use of rocks instead of soil for
burials started because a new culture was utilizing the tumulus, or if the same culture had
a reason to change their burial practices.

The organization within the tumulus is also of interest. The monumental
structures show a tremendous amount of effort required for each burial, so it seems likely
that there was a reason for this change in burial practice. One possibility for the change is
an environmental benefit to the rock foundation, for example, the rock formations would
not wash away as the soil mounds might. It is also possible that being buried in a
structure was an honor given to certain people, such as soldiers or village leaders. If

either of these were the case, adjacent burials would not be expected to be genetically

similar. The change may also have occurred if the village was getting larger and families
decided to begin grouping their relatives together, which may be detectable through
genetic analysis.

Figure 4: Photographs of monumental structures

 

On the left is a photograph of Monumental Structure 1.1 and on the right is a photograph
of Monumental Structure 3. These photographs depict the rock walls that were present
around each individual within these structures.

Use of mitochondrial DNA analysis on ancient remains

DNA can be a useful tool for the analysis of ancient remains, and can provide
information about the relationships between individuals that cannot be obtained
otherwise. There are two different types of DNA in animals, nuclear and mitochondrial
(mtDNA). Nuclear DNA is found within the nucleus of the cell and consists of 46

chromosomes (the mother and father each contributing 23). MtDNA is found within

organelles inside the cell called mitochondria, which are often called the powerhouse of
the cell, as ATP is produced there. There are several benefits to studying mtDNA in
ancient skeletal samples. One is that mtDNA is inherited maternally, because
mitochondria are provided by only the mother (Giles et al. 1980). This form of
transmission means that a child’s mtDNA is identical to the mother’s, allowing for
individuals to be grouped into maternal lineages based on shared mtDNA sequences and
the determination of relatedness even across long periods of time.

A second advantage of mtDNA is that it is often able to be recovered from
samples that do not harbor nuclear DNA, and therefore is often examined in research on
ancient burials. There are a few possible reasons for the persistence of mtDNA. One is
that mtDNA is circular, possibly leaving it less susceptible to exonucleases than the linear
DNA located in the nucleus. Another is that each cell has only one set of the 46
chromosomes which constitute nuclear DNA, but there are hundreds of mitochondria in
every cell, with each mitochondrion having the same mtDNA. This makes the
information in mtDNA highly repetitive, and therefore more likely than single copy
nuclear DNA to have at least one copy persist in ancient samples. It is also possible that
being located within the mitochondria provides mtDNA with better protection than is
available for nuclear DNA (Foran, 2005).

The primary section of mtDNA used for identification is the control region,
which, because it does not encode proteins, is more able to sustain mutation, and thus
vary among unrelated individuals. Within the control region are two segments of mtDN A
where polymorphisms are most prevalent, termed hypervariable region 1 (HVI) and

hypervariable region 2 (HV2). These regions are often chosen for analysis, because the

10

increased variability allows for greater discrimination among individuals. To analyze this
region, amplification of the area is carried out through the use of polymerase chain
reaction (PCR), using a set of primers to designate an area of interest and a polymerase to
amplify the region between them. This process goes through several cycles, with the
amount of DNA doubling after each cycle. The DNA sequence of that region is then
determined, and the sequence is compared to a reference sequence determined by
Anderson et al. (1981), also called the Cambridge Reference Sequence (CRS). An
individual’s mtDNA sequence is then reported as the list of differences between that

individual’s mtDNA sequence and the CRS.

Complications and precautions associated with ancient DNA

Skeletal samples are thought to be an excellent source of ancient DNA because
the DNA binds to hydroxyapatite in bone, slowing down degradation (Tuross, 1994). The
DNA can be retrieved from the ancient bones by converting the bone into a powder, and
using a detergent to break open the cells. By first creating a powder the surface area of
the bone is increased, raising the number of cells that are in contact with the detergent.
However, when working with these samples, several precautions must be taken, because
the DNA may be highly degraded, contain mutations, be contaminated with foreign DNA
by researchers, or contain substances that inhibit PCR.

DNA degradation occurs naturally over time, resulting in less DNA of sufficient
length for amplification or sequencing. Several factors can increase the rate of

degradation, including acidic pH, presence of microbes, and moisture. The level of

11

degradation is also highly dependent on the temperature of the burial and the length of
time since burial (Collins, 2002). Due to the conditions ancient samples endure, the DNA
present is usually limited to segments <300 — 500 base pairs (bp) long (O’Rourke, 2000).
In order to compensate for the decreased amount of DNA present in these samples
it would be ideal to increase the number of cycles used in amplification. The problem
with this, however, is that occasionally primers anneal to non-target areas, amplifying
undesired regions. When the number of cycles is low, this background amplification from
non-specific annealing is usually undetectable, however, as the number of cycles is
increased, it becomes more likely that non-specific amplified regions will interfere with
subsequent analyses. To alleviate this problem a process called nested PCR can be used.
Nested PCR begins with an initial round of amplification across a desired region. If this
does not produce a sufficient amount of DNA for analysis, a second round of
amplification is undertaken using a second set of primers, internal to the previously
amplified area. By using different primers, extraneous amplicons from non-specific
binding during the first round of PCR will not contain the second round bindings sites,
and thus will not be further amplified (Strom, 1998). A variation, called semi-nested
PCR, was used in this research, and involves only one of the primers being moved
internally. O’Rourke (2000) stated that another way to reduce mis-annealing and
dimerization is to use a lower concentration of PCR primers, and should be considered
when working with ancient DNA. A hot start polymerase, which prevents any extension
of mis-annealed primers before the denaturation of the template DNA, will also help to

reduce the amplification of non-target regions (Chou, 1992).

12

Another concern with ancient DNA is nucleotide mutation, resulting from UV
radiation, chemicals, or hydrolysis. Heavily mutated DNA usually cannot be amplified,
because the enzyme cannot recognize the mutated nucleotide, and therefore cannot
continue amplification; however, Hofreiter et al. (2001) found an instance where
mutation still allows for replication. When cytosine is oxidized, it is deaminated and
results in its conversion to uracil (Figure 5). Since uracil is a recognized base, the
polymerase is able to continue amplification. As the DNA is amplified the deaminated
cytosine residue (now a uracil) is paired with adenine instead of a guanine, leading to
subsequent copies base pairing with thyamine where cytosine originally was (Figure 6).
Because not all strands of DNA are affected, this alteration causes inconsistencies in the
sequences, which appear as GT or G-A disparities between samples from the same

individual or as the presence of both C and T or G and A in a single sequence.

Figure 5: Deamination of cytosine

      

Cytosine
The deamination of a cytosine and replacement with oxygen to produce uracil. Taken
from http://satum.roswellpark.org/cmb/huberman/DNA_Repair/damage_types.html
Hofreiter et al. (2001) also discussed a treatment that can reduce the effects of
these mutations. Since uracil, although recognizable by the enzyme, does not naturally

occur in DNA, the DNA sample can be treated with uracil-N-glycosylase to remove the

13

uracils before amplification. This creates breaks in the DNA strands that have been
mutated and allows only non-mutated samples to be amplified and sequenced. The down
side to this technique is that it reduces the amount of starting material present; however,
the material that is present should provide less ambiguous results. There was not time to

attempt this technique in the research presented here; however, it may be useful for future

work.

Figure 6: Misincorporation of thymine

 
   
   

0’ g c Lightstrand

ta c g tga c Lumen-and

 

hydrolysis to uracil

a
Light-strand cytosine i
analogue

at tgc

1st replicatlon

 

2nd replication

at-tgc

After 2nd round of replication.
light-strand cytosine has become
a mymlne

 

 

 

This figure, obtained from Thomas et al. (2003), shows how the deamination of cytosine
to uracil results in a cytosine to thymine transition in the sequence.

14

Excavators and researchers can also cause potential contamination while handling
skeletal material, particularly if precautions, such as wearing gloves, are not taken to
avoid direct contact. This extraneous DNA is generally less degraded, non-mutated and
may be more prevalent than the ancient DNA that is desired, and may therefore amplify
more easily than the DNA of interest, potentially leading to misidentification. One of the
precautions that can be taken to identify contamination is to analyze multiple bones from
each skeleton. An individual should have only one mtDNA sequence; thus DNA from
every bone taken from an individual should produce the same result. It is best if these
bones are processed at separate times, confirming that the results obtained are repeatable.
It can also be useful to remove any surface contaminants by cleaning the bones,
physically removing the surface of the bone by sanding (Kaestle and Horsburgh, 2002),
or exposing the bones to ultraviolet light (Wilson, 1997).

Finally PCR inhibitors can cause difficulties when working with ancient DNA
samples. Inhibitors that may co-purify with DNA from ancient bone include soil
degradation products such as tannins, humic acids and fulvic acids (O’Rourke et al.
2000). Addition of bovine serum albumin (BSA) which binds inhibitors and prevents
them from inactivating the polymerase, can reduce PCR inhibition (Kreader, 1996).
Making a dilution of a sample will reduce the concentration of inhibitors and may allow
amplification to proceed. The problem with this process, however, is that the DNA is also
being diluted and, since there is often little DNA to begin with, the dilution may not leave

enough DNA for successful amplification (Wilson, 1997).

15

MtDNA haplogroup analysis for the classification of individuals

Anthropologist may provide an assessment of race when analyzing a skeleton;
however this categorization is divided into only a few major groups, mainly identifying a
continent of descent Giuropean, Asian, African, or Native American). In contrast,
molecular analyses have the potential to create finer classifications or groupings, called
haplogroups (based on the haploid, instead of diploid, nature of mtDNA). For example,
European and US Caucasians can be divided into 10 major haplogroups (Allard et al.
2002). This can help to refine identification of a group of individuals and can help to
identify possible relationships among individuals.

Haplogroups are defined by the presence or absence of specific bases in the
mtDN A. These single nucleotide polymorphisms (SNPs) were originally detected
through the use of restriction enzymes (Brown, 1980), which can only cleave the DNA if
the exact restriction site (set of bases) is present. It was found that the presence or
absence of restriction sites occurred in different groups of people, some of which
corresponded to geographical regions. The distribution of haplogroups is shown in
Figure 7.

Today, haplogroups are rarely identified based on restriction sites. Instead, the
actual nucleotide difference is assayed using single-base primer extension, in which a
primer is placed directly upstream of the base of interest, and a single fluorescently
labeled dideoxy nucleotide is added. Each dideoxy nucleotide is labeled with a different

fluorescent dye, so the color can be detected to determine the identity of the base

(Figure 8).

l6

Determining the haplogroup(s) of individuals from the tumulus in Albania may
provide more information about what ancestral group(s) Albanians are similar to, or if
there were changes over time that indicate a new group of people inhabited the area. In
particular, if these ancient Albanians are European, they should include some subset of
haplogroups H, I, J, K, M, T, U, V, W, X, which diverged from other haplogroups about
40-50,000 years ago. The most common European haplogroup is H, and occurs in about
39% of Europeans (www.mitomaggrg).

The haplogroup(s) of modern individuals can also be a useful source of
information. The diversity of haplogroups in a modern population can show what peoples
have been contributing to the history of the society. If the distribution of haplogroup(s) in
a modern society is similar to those of the ancient society, it would indicate that the
inhabitants have not been affected by immigration from other civilizations. If, however,
there are new haplogroup(s) present in the modern society, it not only suggests that there
has been influence from another civilization, but may indicate which societies have

migrated into the area over the passing years.

17

Figure 7: Human mtDNA migrations

Human mtDNA Mlgratlops
mmo WHIP-av

'._—__~ ’
——“

    

Map showing the geographical regions associated with each haplogroup, as well as the
migration pattern for the creation of these haplogroups. Obtained from
www.mitomapcrg

Figure 8: Single base primer extension process

DNA tampla'n DNA template

37:; ,kfn [343. )3. :5 3 ' : . ,a 1‘»; {’6 )3 3" 5'

r l '
5n.) “if v: \? nu.”

 

The process of single base primer extension begins by placing a primer directly upstream
of the base of interest. A single flourescently labeled base is added to the primer and the
dye is detected to identify the base incorporated.

l8

Project goals

The purpose of this research was to address a variety of questions about the
tumulus at Kamenica. The burials from within the tumulus can provide information about
whether or not Monumental Structure 3 may be a family grave site (as defined by
maternal relatedness), if the burials represent a single haplogroup or consist of a diversity
of peoples, and if the haplogroup(s) remained consistent throughout the use of the
tumulus. Analysis of modern samples can add to this by showing whether or not the
current inhabitants of Kamenica are genetically diverse, if they share maternal relatedness
with the burials, and if they are of the same haplogroup(s) as the individuals in the
tumulus. MtDNA hypervariable regions were sequenced and compared within and among
burials in the tumulus and current inhabitants, to investigate the maternal relatedness
among the individuals in Monumental Structure 3, and the diversity of the current day
inhabitants. Single-base primer extension was used for preliminary study of individuals in
the burials and the current inhabitants, by identifying them as either haplogroup H or
other. The resultant data were used to gain information about how the tumulus was
organized and why there were changes in burial practices. By compiling information on
both haplogroups and maternal lineages, a better understanding of the organization of the
tumulus, the people in the tumulus, and the people that are living in Kamenica today, can

be obtained.

19

MATERIALS AND METHODS

A total of 16 burials from the excavation of Monumental Structure 3 were
examined in the research presented here. Three bones from each burial were analyzed,
giving a total of 48 bones, which are listed in Table 1. Femur, the petrous portion of the
temporal bone, and humerus were tested most often; however teeth, tibia, radii and other
skull bones were also analyzed. The bone samples were fragments which ranged from
approximately 2 — 6 cm. Twenty-one buccal swabs, which were collected during the
summer of 2004 from individuals identifying themselves (i.e., their family history) as
being from Kamenica, were designated K—l through K—21 and analyzed. Nine of the
buccal swabs (K-13 — K-21) were extracted and sequenced by another researcher

(Lindsey Murray).

Bone and buccal sample preparation

Before the bones were sampled, the drill and drill parts, weigh paper, sterile 1.5
mL tubes, and filter sterilized digestion buffer were exposed to ultraviolet light for 5
minutes to destroy any exogenous DNA. Bone samples were cleaned by pipetting l -— 2
mL of digestion buffer (50 mM EDTA, 0.5% SDS, 20 mM Tris pH 8) onto the sample
and scrubbing with a sterile cotton swab. The samples were rinsed with water purified
through a Milli-Q Ultrapure Water Purification System (Millipore) with a resistivity of
18.2 Mil/cm at 25° C, and set aside to air dry. Bone powder was generated by drilling

with a Dremel rotary tool (Dremel), and titanium l/ 16 inch drill bits. To avoid

20

Table 1: Skeletal samples used in analyses

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Monumental Structure 3 Above Monumental Structure 3 7
urial lBone Type IDate Age [Sex urial [Bone Type IDate Age |$ex
75 [Femur 6 45+ JMale 64H ll-Iumerus 7-6 36-44 [Female
IPetrous portion L4 IPetrous portion
75 (skull) 6 45+ ale 64 (skull) 7-6 36-44 emale
75 Tibia 6 45+ [Male 64 Tooth 7-6 36-44 IFemale
76 emur 6 7-9 Inla 66 [Femur 7-6 a [Male
IPetrous portion JH I“, IN
76 (skull) 6 7-9 n/a 66 umerus 7-6 a ale
Eetrous portion
56 Lroom 6 7-9 n/a 66 skull) 7-6 n/a ale
97 emur 76 n/a Male 78 II-Iumerus 6 29-35 III/[ale
1M Eetrous portion
b7 umerus 7-6 We ale 78 skull) 29-35 ale
97 Tibia 7-6 Wa ITS/[ale 78 JTooth 6 29-35 [Male
142 [l-Iumerus 7 5 + [Male 81 lFemur 7-6 12-16 Inla
142 [Pelvis 7 45+ [Male 81 [Humerus 76 12-16 Inla
[ferrous portion IM ferrous portion
142 skull) 7 45+ ale 81 skull) 7-6 12-16 a
144 IFemur 7 11-13 nla 82 lFemur 7-6 Old [Male
144 [Humerus 7 11-13 Na 82 [Radius 7-6 Old IMale
etrous portion tetanus portion
144 skull) 7 11-13 n/a '82 skull) 7-6 Old ale
149 IFemur 7 15-21 n/a 85 lFemur 7-6 22-28 [Male
149 Tibia 7 1521 n/a 35 Igumcrus 7-6 22-28 [Male
Eetrous portion I;
149 Tooth 7 15-21 lira '85 skull) 7-6 22-28 ale
iddle
156 [Femur l7 29-35 [Male 102 emur 7-6 ed emale
[M IH iddle
156 umerus 7 29-35 ale 102 umerus 5-6 ed emale
etrous portion JM l5 iddle
156 skull) 7 29-35 ale 102 kull 7-6 ged emale
157 lHumerus 7 22-28 emale 104 IFemur 7-6 22-28 [Male
Fotrous portion [5 i“
157 skull) 7 22-28 emale 104 umems 7-6 22-28 ale
etrous portion
157 7 ooth 7 22-28 emale 104 skull) 7-6 22-28 ale

 

 

 

 

 

 

 

 

 

Samples on the left were on the bottom layer in Monumental Structure 3 and the samples
on the right were above the structure. The burial number and bone type are listed for each
sample. The date indicates from which century BC. the bone is thought to have
originated. The anthropological estimate of age and sex are also given when available.

21

contamination that might have been on the surface of the bone, one drill bit was used to
start a hole a few millimeters deep, with the bone powder being collected in a waste tray,
while a second bit was used for continued drilling. The bone powder was collected on a
piece of weigh paper, transferred to a 1.5 mL tube, and weighed.

A disposable scalpel was used to cutoff approximately 1%: of each buccal swab,
which was placed in a 1.5 mL tube. After each cutting the blade was cleaned with 70%

ethanol.

DNA extraction

DNA was isolated from the bone and buccal samples by incubating them in 400
uL of digestion buffer and 4 uL of 20 mg/mL proteinase K at 56°C overnight. Each time
a set of samples was extracted, a reagent blank containing digestion buffer and proteinase
K was also processed. The bone samples were extracted twice with an equal volume of
phenol and once with an equal volume of chloroform, while the buccal samples were
extracted once with each solvent. The aqueous layer was placed onto a Microcon-30
column (Millipore) and centrifuged at 14,000 X g until only a few microliters remained
(approximately 7 minutes). Filtering was repeated twice by adding 300 ill. of 10 mM Tris
pH 7.5, 1 mM EDTA (TE), then the retentate was brought to a volume of 15 pL with TE,

transferred to a clean tube, and stored at —20°C.

22

DNA amplification

Four different mtDNA regions were amplified, three for sequencing and one for
SNP analysis. When necessary, semi-nested PCR was utilized in order to obtain enough
DNA for analysis. All PCR amplifications included a positive control, which contained
Human Genomic DNA-Male (Promega), to ensure that the amplification was working
correctly, a negative control containing no DNA to verify that PCR reagents were not
contaminated, as well as carrying through the reagent blank to test for contamination of
the extraction reagents. The sequenced regions included HV 1-1 (F16144/R16410; 266bp
product), HV 1-2 (F 15989/R16207; 218bp product), and HV2 (F 15/R285; 270bp
product). For semi-nested PCR the same three regions encompassed Fl6190/Rl6410
(220bp product), F16057/R16207 (150bp product), and F82/R285 (203bp product).
Primers F16057, R16207 and F82 were designed in house and the remaining primer
sequences came from Armed Forces DNA Identification Lab (AFDIL) (2005). In both
rounds, the PCR reaction consisted of 2 llL of HotMaster buffer (Eppendorf), 2 uL of 2
)1pr dNTPs, 0.4 uL of a forward and reverse 20 M primer, 6 uL of 10 ug/uL BSA,
and 1 U of HotMaster Taq (Eppendorf), which was brought to 20 uL with sterile water. If
primer activity was faint or not present, the amount of BSA was increased to 8 or 10 uL
to reduce possible PCR inhibition. The cycling parameters for the first round of PCR
included a denaturation step of 94° C for 2 min, followed by 38 cycles of DNA
denaturation at 94° C for 30 sec, primer annealing at 56° C for 1 min, and extension at
72°C for 1 min. Five ILL of PCR product was electrophoresed on a 1.5% agarose gel to

estimate the amount of PCR product present. If there was over-amplification, identified

23

by multiple product bands or bands in the negative controls, the number of cycles was
dropped to 35 or 30; however, if no DNA was visible, semi-nested PCR was conducted,
using the same cycling parameters for fewer cycles. Cycle number was based on the
amount of DNA visualized in the product gel. If DNA was visible, but did not appear to
be enough for sequencing, 8 cycles were done, if there was no visible DNA 20 cycles
were chosen, and intermediate cycle numbers were used when hints of DNA were present
or most lanes showed DNA and it seemed others would produce visible DNA with added
amplification.

SNP analysis was carried out with primers Hfor
(AGACATCGTACTACACGACACG)/ Hrev (TGATGGCAAATACAGCTCCT) (80bp
product) for the first round of PCR followed by Hfor
(AGACATCGTACTACACGACACG)/ HSNP
(TATTGATAGGACATAGTGGAAGTG) (6lbp product) when semi-nested PCR was
necessary. Both rounds of PCR reactions consisted of 1 uL of HotMaster buffer
(Eppendorf), 1 uL of 2 pg/uL dNTPs, 0.2 uL of forward and reverse 20 M primer, 3uL
of 10 ug/uL BSA, and 1 U of HotMaster Taq (Eppendorf), which was brought to 10 uL
with sterile water. The cycling parameters for the first round of PCR included a
denaturation step of 94°C for 2 min, followed by 40 cycles of DNA denaturation at 94°C
for 30 sec, primer annealing at 57°C for 30 sec, and extension at 72°C for 30 sec,
followed by an extension at 72°C for 5 min. A 3% agarose gel using 5 uL of product was
utilized to estimate the amount of DNA present. If there was over-amplification,
identified by multiple product bands or bands in the negative controls, the number of

cycles was dropped to 38, 35, or 30, while if no DNA was visible semi-nested PCR was

24

done, using the same cycling parameters, however, with fewer cycles. The number of
cycles was determined by the amount of DNA in the product gel.

Buccal swab mtDNA was amplified using F15989/R484 or the same HSNP
region as the bone samples (Hfor/ Hrev). The samples that were sequenced by Lindsey
Murray were amplified in the region from F15989/R499. Amplifications were carried out

using the conditions outlined above.

DNA sequencing

Amplified mtDNA was purified through a Montage PCR Microcon (Millipore) by
adding the remaining 15 uL of PCR product and 385 uL of TE, which was centrifuged
for 15 minutes at 1,000 x g. After centrifugation the retentate was brought back to a final
volume of 15 uL with TB. Sequencing reactions consisted of 4 uL of CEO DTCS Quick
Start Mix (Beckman), 1 uL of the appropriate 20 pM primer (those used in PCR
amplification), and approximately 6 ng (50 femtomoles) of DNA for the bone samples
and 33 ng (50 femtomoles) for the buccal samples (0.5 — 5 uL of sample based on the
intensity of the band in the product gel). Water was used to bring the reaction to 10 pL.

Sequencing parameters consisted of 30 cycles with DNA denaturation at 96°C for
20 sec, primer annealing at 50°C for 20 sec and extension at 60°C for 4 min, except when
sequencing with the primer R484, which required an annealing temperature of 40°C.
DNAs were precipitated using 3 pL of stop solution (1.2 uL NaAc, 0.96 uL of water,
0.24 uL 500 mM EDTA, and 0.6 uL of 20 ug/uL glycogen) and 30 uL of 95% ethanol,

which was centrifuged at 14,000 rpm for 15 minutes. The supernatant was removed and

25

the samples were washed twice with 200 pL of 70% ethanol using a 3 minute
centrifugation at 14,000 rpm. The DNA was vacuum dried for at least 30 minutes, and
resuspended in 40 pL of Sample Loading Solution (Beckman). The samples were
sequenced on a CEQ 8000 Genetic Analyzer (Beckman) using the LFR-1—45 program
(capillary temperature 50°C, denature 120 s at 90°C, inject 15 s at 2.0kV, and separate 45
min at 4.2kV). The sequences were uploaded into BioEdit (Hall, 1999) and aligned and
compared against the CRS (Anderson et al., 1981). For a subset of the buccal sequences,
haplogroup determination could be made based on control region sequence (Allard et al.

2002).

SNP analysis for haplogroup H

Haplogroup H is defined by a C at position 7027, whereas other haplogroups have
a T at this location. This haplogroup was identified via SNP analysis (for assignment of
other haplogroups, see below). Following PCR amplification of region 6991 - 7051,
primers and dNTPs were removed by adding 2 U of shrimp alkaline phosphatase (SAP)
and 10 U of exonuclease 1, and incubating at 37°C for 1 hour, followed by enzyme
denaturation at 75°C for 15 minutes or 80°C for 20 minutes (as per manufacturers
instructions). SN P reactions contained (added in order, per manufactures instructions) 3
uL of water, 5.5 uL of pre-mix from the CEQ SNP-Primer Extension Kit (Beckman;
reagent concentrations not given), which contains 1 [AL of reaction buffer, 1 uL each of
ddGTPs and ddATPs (the other nucleotides were not required and were replaced with 2
uL of water), and 0.5 pL of polymerase, followed by 0.5 uL of sample and 1 [IL of 2 uM

primer. Cycling parameters were 25 cycles of DNA denaturation at 96°C for 10 sec,

26

primer annealing at 55°C for 5 sec, and extension at 72°C for 30 sec. Unincorporated
ddNTPs were removed by adding 1 U of SAP and incubating at 37°C for 1 hour,
followed by 75°C for 15 minutes to denature the enzyme. The samples were analyzed on
a CEQ 8000 Genetic Analyzer by combining 39 [L of Sample Loading Solution, 0.5 11L
of size standard 80 (Beckman), and 0.5 uL of sample, and running the SNP-long program
(capillary temperature 50°C, denature 60 s at 90°C, inject 30 s at 2.0kV, and
electrophoresed for 18 min at 6.0kV). Data were analyzed using the program Fragment

Analysis associated with the CEQ 8000.

Data analysis

Consensus sequences were generated for each of the burials. When three bones
provided sequence at a given location and one of the results was different from the other
two, the most prevalent base was used in the consensus sequence. When two bones
showed different sequences at a given location, this was reflected in the consensus
sequence as both bases. There were also regions (near the ends of the sequence) in which
only one bone from the burial provided sequence. In these instances the consensus
sequence reflects the one sequence obtained.

Buccal swab mtDNA sequences were used for individual identification and
haplogroup determination. Control region polymorphisms that define each haplogroup
(Allard et al. 2002) were compared to the buccal sequences. Many haplogroups were
characterized by multiple polymorphisms, all of which had to be present for haplogroup

assignment.

27

RESULTS

Bone preparation

A list of all bones processed can be found in Table 1. During the sample
preparation, it became evident that certain bone types had characteristic problems. The
teeth were the most challenging to drill. Their small size made them difficult to hold
while drilling and the drill did not easily cut through the sample because of its hardness.
This problem was more noticeable with the crown of the tooth than the root. The petrous
portion was also a hard bone, and sometimes caused the drill to overheat if too much
pressure was applied. The one pelvis sample was also challenging because the porous
nature made it difficult to clean with the digestion buffer and cotton swab, resulting in
dirt mixed in with the bone powder. The bones that were easiest to drill were long bones
in which one end of the bone was cleanly cut. This gave a smooth surface and allowed for

the drill to travel lengthwise through the bone, producing more bone powder per hole.

DNA extraction

During the extraction process, the DNA samples varied in color, ranging from
clear to dark pink to brown. Often this discoloration was removed when the sample was
Microcon filtered; however, in five samples (76 tooth, 81 femur, 102 humerus, 142
pelvis, and 157 petrous portion) the contaminants were too large to be removed. When
this occurred there were generally clumps on the filter, however any attempt to wash over
the filter and retrieve the DNA caused them to redissolve. Attempts were made to

centrifuge the DNA samples to see if the contaminant could be pelleted, but these were

28

unsuccessful. Since the discolored samples seemed likely to cause inhibition during
amplification, 1 uL of the DNA was diluted with 9 uL of TE, and the dilution was used

for all subsequent amplifications.

PCR amplification of regions to be sequenced or used in SNP analysis

Approximately 20% of the DNA samples generated a visible product after the
initial round of PCR (see details below); however, the majority of samples required semi-
nested PCR to generate sufficient product for sequencing. Figure 9 shows a typical
outcome in which the first round of PCR showed no visible DNA on an agarose gel, but
after a semi-nested round of PCR, DNA was present. A faint band was also noticeable
above the desired product; this was a common observation from semi-nested PCR and
results from remnants of the first round primers, which continue amplification of this
larger region. Approximately 10 — 15% of the DNA samples did not amplify even after
the semi-nested reaction, and although PCR was attempted on all DNA samples multiple
times, those that did not amplify were discarded (see sequencing results below).

Bone types differed in the success rate of DNA amplification. Of the three bone
types that had a large sample size (36 — 39 attempted amplifications), femur DNA
amplified the most often (92%) followed by petrous portion (90%) and humerus (81%).
None of these differences were statistically significant. First round amplification success,
however, was more frequent in petrous portion (28%) than it was in humerus (22%) or
femur (11%). Interestingly, petrous portions from burials 75, 142, 156, and 157, which
represented 31% of the petrous portions examined, repeatedly amplified without using

the semi-nested technique (Figure 10), as did the tibia from burial 97 and the humerus

29

Figure 9: An example of typical semi-nested PCR results

 

Left—a 1.5% agarose gel showing no PCR products after the first round of amplification
(mtDNA region 15 - 285). The sample lanes in order are: negative control, reagent blank,
66 femur, 78 tooth, 85 femur, 104 femur and positive control. The remaining lanes are
empty. Right— a 1.5% agarose gel of the same samples (in the same order) after the
semi-nested round of amplification (mtDNA region 82 — 285). PCR products are now
visible.

Figure 10: Successful amplification after the first round of PCR

 

 

A 1.5% agarose gel showing successful amplification after a single round of PCR
(mtDNA region 15989 — 16207). The lanes from left to right are: negative control,
reagent blank, 75 petrous portion, 142 petrous portion, 156 petrous portion, 157 petrous
portion, and positive control.

30

from burials 66 and 85. Other DNA samples that had one mtDNA region amplify without
using the semi-nested technique are displayed in Table 2.

PCR inhibition also occurred during amplification attempts. BSA was added to
each reaction in order to reduce inhibition, but approximately 10% of the DNA samples
still showed inhibition, identified as a lack of primer activity on the agarose gel (Figure
11). Note that primers are evident in all lanes (lowest arrow); however, primer activity
(middle arrow) is not present in lanes six or seven. The lack of primer dimerization in
lane six is the indication of inhibition, which was addressed either by diluting the DNA
1:10, or increasing BSA concentration. The lack of primer activity in lane 7 results from

the large amount of product utilizing most of the primers and leaving few to dimerize.

Figure 11: Example of inhibition in lane 6

 

 

Non-nested product
Nested product

Primer dimers
Primers

 

A 1.5% agarose gel showing the PCR products from amplification of mtDNA region 82 -
285. Samples are: negative control, reagent blank, 64 petrous portion, 81 petrous portion,
82 petrous portion, 102 petrous portion, and positive control. Lack of primer activity in
the sixth lane (102 petrous portion) indicates PCR inhibition.

31

Sequences of mtDNA from bone samples

Table 2 lists the skeletal samples used in these experiments, as well as the range
of sequence obtained from each bone. When a DNA region amplified, it was generally
also successfully sequenced; however, approximately 15 — 25% of the time the
sequencing reaction needed to be repeated due to excess DNA, a lack of DNA, or
equipment failure. Amplification was not successful on all DNA samples, resulting in a
lack of sequences from some of the bones. However at least two bones from each burial
were successfully sequenced in each of the three mtDNA regions attempted. These were
aligned and compared to the CRS. Appendix 1 displays the differences from the CRS
found in each bone, and a consensus sequence that was generated for that burial.
Appendix 2 contains only the consensus sequences.

All of the burials contained the common polymorphism 263-G in HV 2 (except
for burial 78, where the sequence was not available) as well as other polymorphic sites.
Many of these were classified as ‘ambiguous bases’, meaning two or more bases existed
at a single location, or that the base at that location was unidentifiable. These bases were
denoted as Y (C and T present), M (C and A), R (A and G), W (A and T), or N (any
base). The frequency of ambiguous bases was consistent across the three mtDNA regions
that were sequenced; however, the number in each burial’s consensus sequence varied
greatly, from 1 to 11 (Table 3). The type of ambiguity also varied greatly, with 57%

being Y’s, 18% W’s, 14% R’s, 10% M’s and 1% N’s.

32

s. -- .. 3 35065 2: $68: “o: E: $5 $383 <20 .8 .3 38:2: 2: mos—2:52 323563 6

cm: 05 Song; umEEEw 2.23303: 35 «295$ .28 5 ._m .o :oflou:< 2 ocficoamctoo «cones: 5:5 698.qu

2 :29: some :_ 352:: 8:263 Co :38: 9:. .20: some :2 .m >I cc: .mé >I 62>: 8:059. <29E 625 :o
38:33: me; 96:309. :5 :ozmoEaEm <20 .8 E: F $59“. 39 Eu: 9: :o 2.353 on. Co :98. :2 2: Eat
22.5: :5 £2 :5 no 2.383 05 mo :32 Boxes 05 Bob 22.52 5:.» .3832: on: 095 0:0: 28 238:: 3:5

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

-- :22 :82 288 2: ~22 :22 -- 58... 5
_22 222 -82 2:68.... 2: .62-; :22 2:2 Boss: 5
:22 :22 -- as: :2 «2:2 :22 -222 mauasm E
22.2 :22 :22 :5: 22 82-2 :22 2:2 38...: :2

-- :22 :82 8652 22 32-2 :22 :22 335.2 :2
2.2.2-2 :22 :82 as: 22 «2:2 :22 2:2 556: :2
42.2 -22 2:2 goes: 2 2-2 -- -- 58.: 22
22-2 -222 :82 mass: 2 22-2 :22 :82 see 2..
22:: -222 ::2 as: 2 -- :22 :82 5&6: 2;
«2-2 :22 :82 36:6: 2 22:2 :22 -- sees: 3;
2.2.5 -22 -- see 2 2.2.2 :22 :82 ”265:: 32
22:2 :22 :82 as: 2 -- :22 :82 ass: 3;
:22 :22 :82 888 a 22-2 :22 -82 32.6: 22

-- :22 -- mass: 5 22:2 :22 :82 2%: 22
22:2 222 -_ _ _2 as: a 22-5 -- -- ”sues: a:
2-: -22 2:2 569 2 22.5 222 :82 65:. 8
22-2 -- 5:2 36:8 2 -- :22 -- masses: 3

-- :22 :82 265:: 2 22-2. :22 :82 356: 8
2.2.: :22 5:2 2:8: 8 -- :22 :22 58... 8
2.22-2 :22 5:2 2652 8 22-2 :22 $82 388 2
M28 :22 :82 see 8 N24: :22 -- ass: 8
:22: :22 :82 569 8 22-2 -22 -- «5:. 2
22-2 :22 -282 886: 8 22.2 :22 2:2 36:8 2
22:2 -- 8:2 masses: 8 22-2 :22 :82 .28: 2

«>2 5: 3E 6:: 68m 3.5: SE S: SE 8: 25m 35m

 

 

 

 

 

 

 

 

 

 

 

8.958 one: .«o 2.83.. 3538< "N :33.

33

Table 3: Number of ambiguous bases in each consensus sequence

 

burial HVl-l HV1-2 HV2 Total
64
66
75
76
78
81
82
85
97
102
104
142
144
149
156
157

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

A—owNNNOOHA—NNn—o
wOHO-hUIt—NUJNOOOHOUJ
—-owwoooc~wo~ooo
oomwmooxooammwo-bwm—z

The number of ambiguous bases is given for each burial in each of the three mtDNA
regions, followed by a total for the entire sequence.

When ambiguous bases were not included, 12 of the burials had a sequence of
263-G, which is a very common sequence, however four unambiguous polymorphisms
were also identified. The first was 72—C in burial 142, which was present in one of the
bones, with the others having no sequence at that location. Burial 149 also had a
polymorphism (16221-T) present in one of the bones, with no others having sequence at
that location. Burial 81 contained 16363-T, which was present in two of the individual
bones; however the third bone had a C at that location. Burial 66 was the only bone to
show a consistent polymorphism in all three bones (16353-T). Among all samples there
were also several locations (approximately 25), in which one bone had a base difference
that was not seen in the other two. In these instances the consensus sequence for that

burial did not reflect the difference.

34

MtDNA sequences of buccal samples

Sequencing was attempted on twenty-one buccal samples; 20 were successful.
The remaining sample (K19) was extracted on two separate occasions and neither DNA
prep resulted in successful amplification. The DNA region examined from the buccal
swabs was larger (15989 — 16410 and 15 - 484/499) than that obtained for the bone
samples. Polymorphisms observed are shown in Appendix 3. All 20 of the sequences that
were obtained were different from one another. There were a total of 63 polymorphic
sites, with 38 of them occurring in only 1 individual. Three differences from the CRS that
are very common in Caucasian individuals, 263-G, 315.1-C and 309.1-C, were also
present in 20, 19 and 13 of the modern Kamenica samples respectively. When only the
bases that were available in both the bone and buccal samples were considered, one of the
buccal samples (K-1 1) had a sequence of 263-G, which is the same as the majority of the

burial sequences.

Assessment of haplogroups

Two methods were used for haplogroup determination of the modern day (buccal
swab) Albanian samples. The first was based on fixed differences in the control region
that define certain haplogroups (Allard et al. 2002). Four samples were typed in this way:
K-2 was identified as haplogroup V (defined as l6223-C and 72—C), K4 was identified
as haplogroup U5 (16270-T), K-15 was identified as haplogroup J (16069-T and 16126-
C), and K-20 was identified as haplogroup T (16126-C and 16294-T). Allard et al. (2002)

defined the H haplogroup as having 72-A (which is the base present in the CRS) and

35

none of the other polymorphisms, but noted that haplogroup H is poorly defined in the
control region. Because of this, none of the sequences were assigned to haplogroup H
using this technique, even if those criteria were met. Instead, single base primer extension
was used to analyze base 7027, which better defines haplogroup H (Figure 12). In doing
so, nine of the buccal swabs were found to be in haplogroup H (K-3, K—5, K-6, K-7, K-8,
K-10, K-1 1, K-13, and K-21), and the remaining 11 were not in haplogroup H. Sample
K-19 could not be amplified.

The bone samples were also analyzed using the same single base primer extension
as the buccal samples. Of the 16 burials, 8 produced a consistent result from multiple
bones, with five burials (66, 82, 97, 104, and 156) in haplogroup H and three (64, 75, and
81) not in haplogroup H. There were two other burials (142 and 144) in which DNA from
one bone was successfully analyzed, but DNA from a confirmation bone was not able to
be amplified; both indicated the individual was not in haplogroup H. Four of the burials
(76, 85, 102, and 157) produced ambiguous results, and the final two burials (78 and 149)

were unable to be amplified without the reagent blank also producing product.

36

Figure 12: An example of SNP results

 

 

 

 

 

 

 

On the top is the electropherogram of sample K-20, which incorporated a G (green peak)
at approximately 22bp, indicating that the sample is in haplogroup H. On the bottom is

mm):-
11m“-
,—
)—
_
1m -:
I
cacao-3
._
..
I an
mo...
..
b
h—
.—
anew--
_.
..
_
._
g mutu-
h
3' m:—
E
«Illa--
..
h
cacao-3
._
W's- -
: u
1m.: {I -
- IL . YI- a J
-— 11111 J ll__a.’ L L A 111 n:n__n_a J'lelun
I T I T 1 T I
D 10 ID no ‘I u) no on D
—M
IAI'IO'IIJ
h-
_
sac—a -:
b
E
120000 a;
E
«and-
._
com .rE a.“
Z
cacao -—
Z
P
I—I --_—
._
._
s v— -:
-
g. cacao a:
._
a—I 4:-
...
_
«coco-1:.
:
.0000 -l-
..
._
a.— 15
: -
snow-=-
= I' .L- -=.
_ - can ‘
°_llJL:‘LI—‘—?I‘LL_IJ1:A‘LII1:711Pl:lllj: -_'i-__p__.-"_4.1_M/J_I_Ll
I l
D 10 ID I) N a: no 70 on no
—0‘)

the electropherogram of sample K-21 which incorporated an A (red peak) at

approximately 22bp, indicating the sample is not in haplogroup H. Images in this thesis

are presented in color.

37

 

 

DISCUSSION

Genetic analysis of ancient cultures can provide unique historical insight that
cannot be obtained by other means. Examination of the genetic diversity that existed in a
local people, as well as how that compares to today’s inhabitants, can help us understand
the movements of peoples, their possible interactions, and the cultural makeup of their
society. For the burials in the tumulus at Kamenica, the meaning behind the burial
patterns, particularly the switch to Monumental Structures, is unclear. The DNA
sequences from individuals within the tumulus shed light on the relatedness of
individuals within Monumental Structure 3, which can potentially explain why these
structures were made. By examining mtDN A haplogroups, the ancestral history of
Albanians could be studied, helping discern if the same group of people always existed at
that location, and into the modern day. The modern samples also help provide
information on the diversity that is present today in Kamenica. At the same time,

observations were made that will help future DNA studies on ancient skeletal material.
Sample analysis

It was evident throughout this study that some bone types were easier to work
with than others. There were two types of bones that were most often analyzed. One was
long bones, particularly femur, which were chosen because Misner (2004) suggested
using long bones instead of spongy bones like pelvis for DNA analyses. Whenever
possible, these samples were drilled lengthwise down the bone, which produced a greater

quantity of whiter bone dust than drilling perpendicular to the shaft of the bone. The other

38

material that was used frequently was the petrous portion of the temporal bone. This bone
was selected because its unique hardness meant it might better protect the DNA and
produce more robust results. Both of these bones had high amplification success, and the
petrous portion showed high quality of DNA, evident by the more frequent amplification
without using the semi-nested technique. One bone type that was avoided was tooth,
which has been shown to give acceptable amplification success, but due to its size and the

hardness of the crown is difficult to drill.

Mutation efi'ects

The majority of differences found in theconsensus sequences of the burials (79%)
were instances in which multiple bases were present at the same location, either from the
same bone or in multiple bones from the same individual. Of these, 57% were Y’s, with
both C and T present. One possible explanation for multiple bases in a sequence is
contamination with exogenous DNA. This, however, seems unlikely given the frequency
of CIT transitions observed. If contamination were an issue, one would expect to see
more random ambiguities. Further, contamination seems unlikely in that the multiple
differences among sequences would require that a large number of separate contaminants
be involved, since the same locations were not ambiguous in each of the sequences.

Another possible reason for the ambiguous bases is heteroplasmy, or a natural
occurrence of two bases at a single location within an individual. Again the high
frequency of CIT ambiguities makes heteroplasmy unlikely since other base

combinations would be expected. Also, in modern samples, heteroplasmy has been found

39

to be a rare event, and is unlikely to be occurring in every individual in the tumulus.
Further, when heteroplasmy has been observed in modern material, it is virtually always
found at a single location. Many of the skeletal sequences displayed ambiguous bases at
multiple locations, a fact that points away from heteroplasmy.

A third possibility for the CH“ ambiguities is the mutation of cytosine to uracil
(see Introduction). Since this process specifically results in OT transitions, it would
explain their prevalence. Hofreiter et al.(2001) mentioned that the post-mortem C/T
mutations may be more common in “hotspots” that already tend to vary among
individuals. Indeed, all of the GT ambiguities from HV 1, and five of the eight from
HV2, have been previously reported as transitions (www.mitomap.org). Two of the
samples containing C-T ambiguities in the current study were treated with uracil-N-
glycosylase (UNG) (Hofreiter et al. (2001), by another scientist in the laboratory
(Michael Mutolo). This resulted in some, but not all, of the questionable bases being
more definitively identified. More testing will need to be done to determine if UNG
treatment can help eliminate all ambiguous bases; however the method seems to hold
promise. The results of the UNG treatment further support the idea that the OT
ambiguities are most likely due to post-mortem mutations and do not reflect the sequence

of the individual when they were living.

Comparison of mtDNA regions

Three different regions of the mtDNA, two of which were located in HVl, were

analyzed from each of the bone samples. Belidi et al. (2000), analyzed mtDNA sequences

40

of 42 modern Albanians from “24 Albanian districts and in the adjacent regions of
Macedonia” from position 15996 - 16401, and found that 31 different haplotypes existed.
Five of these shared the same HV1-2 sequence, and eight shared the same HVl-l
sequence. This information made it evident that it would be necessary to look at both of
these regions to identify diversity among individuals in the current study. Surprisingly,
none of the burials differed in their HVl-l, and there were only two repeatable
differences found in all tumulus samples, both in HVl-2. All three regions had
approximately the same occurrence of ambiguous or unidentifiable bases, suggesting that
no one region was more or less susceptible to post-mortem mutations. The only
noticeable difference among regions was that twelve samples allowed HV2 amplification
on the first round of PCR, while HVl-l and HV1-2 only had 7 and 6 respectively (Table
1). This was unexpected since the amplicon for HV2 was the largest. A possible reason
for less amplification success from HVl-l is that the reverse primer for this region is
within a C stretch making it difficult for the polymerase to amplify. The primers for the
HV1-2 region, however, are prevalently used in this lab, and it is unclear as to why this

would be producing fewer amplification products than the HV2 primers.

Maternal relatedness

MtDN A analysis can be used to identify maternal lineages. The monumental
structures of the Kamenica tumulus provided an opportunity to use genetic tools to
examine the relatedness of the individuals buried there, and indicated that the burials in

Monumental Structure 3 could be maternally related. Aside from the ambiguous bases,

41

which are most likely postmortem mutations, there were four burials with differences
from the others. Two of the differences were found in a single bone, and were located on
the bottom layer of the tumulus, with one burial (149) in the center of the structure and
the other burial (142) just north of the first. The other two differences were confirmed in
multiple bones, and were located on the top layer of the tumulus, with one burial (66) at
the north end of the region and the other burial (81) just south east of the first. If this were
a family plot, the burials with different sequences could be due to individuals who had
married into the family and did not share the same mtDNA sequence. It is also possible
that, as time went on, individuals that were not part of the family were also buried in this
structure, since the two consistently different burials were from the upper layer of the
tumulus and were not surrounded by rock walls.

Another possibility for these results is that individuals not closely related also had
similar mtDNA sequences, since the sequences obtained were common. This seems
unlikely, however, since analysis of different regions of the tumulus, done by other
researchers (Lindsey Murray and Virginia Clemmer), has shown burials with differences
from both the CR8 and each other. This indicates that the burials in Monumental
Structure 3 are genetically different from previously studied regions of the tumulus. This
may be due to the structure being a family plot, or since the structure was utilized after
the other regions, there could be a time dependent reason for the lack of diversity.
Examples of this include a less genetically diverse group of individuals taking over use of
the tumulus, or a genetic bottleneck causing a loss of diversity. Analysis of other rock
burials, those in Monumental Structure 1.1 or the unexcavated area, may help to provide

more information on the change in diversity, since they were created during the same

42

time period as Monumental Structure 3. If the burials from the other rock burials are
genetically different from those in Monumental Structure 3, it would support the idea that
burials in Monumental Structure 3 share a maternal relationship. If all of the burials are
very similar to one another it would still leave two possibilities. One is that the entire
region is from one family unit and the second is that there was a loss of diversity
throughout the entire population.

Another possibility is that although these individuals do share the same sequence
in the regions analyzed, there are other areas of the mtDNA sequence which are different
among the individuals. This possibility seems more likely when the haplogroup
information is considered. The haplogroup analysis indicates that about half of the
individuals are from haplogroup H, which is similar to the percentage of modem
Europeans. Since maternal lineages would be of the same haplogroup, these results
suggest that the individuals are not maternally related, however, it is important to keep in
mind that four of the burials produced ambiguous results. The reason for this discrepancy
may be because haplogroup H determination is based on a CIT transition. As discussed
previously, CIT ambiguities were prevalently found in burial DNAs. If these transitions
were due to post-mortem mutations, they could also be occurring at the SNP base, which
would alter the results of the haplogroup H determination. UNG treatment of these

samples may be of interest since it may produce less ambiguous results.

43

Modern inhabitants from Kamenica

The analysis of the modern day inhabitants of Kamenica tells us about the genetic
diversity of individuals living there today, and allow for comparisons between the current
inhabitants and inhabitants 2500 years ago. The most notable difference between the
modern and ancient samples is that the current day inhabitants show more genetic
diversity than the ancient individuals. None of the modern samples shared a common
sequence, indicating that the individuals were not maternally related. However, it is also
important to note that the region examined in the modern samples was larger than the
region examined in the bone samples. If only the regions that were analyzed in bone
samples were considered, then two of the buccal samples (K-5 and K-6) would share a
mtDNA sequence. The modern samples were from at least five different haplogroups,
with just under half of them being in haplogroup H. The distribution of haplogroups in
the modern samples is similar to the haplogroup distribution in other areas of Europe.
Comparison of modern samples to the samples within the tumulus shows that there is one
modern sample (K-1 1) that shares the same sequence as the majority of the tumulus
samples, when only those bases examined in the tumulus samples are considered. Also
there were four differences in the tumulus samples, and only one (72-C) also occurred in
the modem samples. However, if the tumulus burials are from a family unit, they may not
be portraying an accurate view of the polymorphisms that were present in the ancient
individuals.

The mtDNA sequences obtained from current Kamenica inhabitants were also

compared to the Albanian sequences obtained by Belidi et al. (2000). Buccal sequences

K-4 (l6192-C, 16256-T, and l6270—T) and K-7 (16287-T) were the same as individuals
27 and 12 respectively from Belidi et al. (2000). However, 70% of the polymorphic sites
found in current day inhabitants of Kamenica were also found in the Albanians studied by
Belidi et al.(2000); suggesting that the current residence of Kamenica are as diverse as

Albanians in other regions.

Conclusions

The inhabitants of Kamenica, both in modern times and in the 6‘h to 7th century
BC, appear to have European haplogroup H as their most common haplogroup. Taking
into consideration that the burials in other regions of the tumulus, and the current day
Albanians have diverse mtDNA sequences, the burials within Monumental Structure 3
indicate that this structure contains individuals that are maternally related. The increased
diversity of the modern samples could suggest that at some point in history there was an
influence from outside populations. However, since the monumental structure could be a
family grave, it would be necessary to compare the modern samples to a more diverse

sampling of the entire tumulus before any conclusions could be made.

45

 

     

APPENDIX A:
Sequences from Bone Samples

Table 4: Sequences for each bone in region HV2

 

     

Table 4

     

     
   

          

   
 

      

The burials are arranged so the first burials are from the bottom the and the last are top. the sequence each separate bone is a sequencl: labeled
With a C, which is the consensus sequence for that burial. The gray regions indicate areas in which Sequence was not obtained. Any dlfferences are noted by letters lndlcatlng the 1ase .d
present at that spot. The letters stand for A—adenine, C—cytosine, T—thymine, G—guanine, Y—both C and T, R both A and G, W both T and A, M both C and A, or N an unknown nuc eotl e.

    

    

Table 5: HVl-l

 

for each bone in

 

Table 5

   
  
  

 

        

burials are so that the are from the bottom of the and last eight are the top. After the sequence each separlate bone dllsC :tisgqqfigqjealsaebeled
with a C, which is the consensus sequence for that burial. The gray regions indicate areas in which sequence was not obtained. Any differences ari:1 réotedibAy :Etgsailnunknovgn nucleotide.
present at that spot. The letters stand for A-adenine, C—cytosine, T-thymine, G-guanine, Y—both C and T, R both A and G, W both T and A, M bot an a

48

 

Table 6: for each bone in

 

C C T A C A A

49

 

 

Table 6

         

The are arranged so that the are from the bottom of '
with a C, which is the consensus sequence for that burial. The gray reglons 1ndlc
present at that spot. The letters stand for A—adenine, C—cytosrne, T—thymlne, G-g

. labeled
. f separate bone 18 a sequence
umu IIS alltl he 3.31 e ght are the tOp. After the sequence 0 . ' ‘ t
:1 6 21116218, .11 thllClll sequlence was not obtained. Any differences are noted by letters 1ndlcatlng he blaSCti I
t . 1‘ t R both A and G both and A, M both C and A, or N an un
uanlne, 1 b0 ['1 I and I, , W I known nuc 60 de

50

 

62:83:: 9505?: a: Z .8 .< 2:: 0 53-2 .< 2:: H 53.3 .0 :5: < 53.x d. 2:: U 53-» 6:222»:
-0 85:53? 62828-0 .2:Eofi:-< :8 ES: 28:2 2:. .89: an: 2. 808:: one: on: wedge—2 28:2 .3 :82: 0:: maggobme
be .Bngno 8: we? 8:038 533 E 80:: 38:22 2:0me 8% 25. .325 new: 8.: 00:33: 82:02:00 : 825:: 03:: mat.

> M 3 M :2
M 0.2

3
Q:
vo

. .. _ 2:252:00 no
2.2828: mm
mm

2 w
2.2525: x h

26:02:00 on
2.2888 mm
w 383:8 co
, , 2.2528: we
0 P U H O H H O < H H O < H O U mmu

Sum ~ cam mmm 9m mm 2: mm: N2 31 a S. we mocom
N>I

O
O
O
O
O
O
O
O
O
O
O

 

«g E 89:252.. 2.25250 "A. 93:.
325 new :8 mooaoswom 25:02:00

um XHQmemA:

51

.8828: E88 8 Z 8 .< :8 0 53-2 .< :8 H 58.? a0 :8 < 58-x .H :8 U 58% 6:88
-0 6588:- H 68898.0 688:.< :8 :58 8.8— 8H 8% 86 8 “:88: 83 05 8:885 838— 3 88: 8 80888:
8.: 80:58 8: 8B 8888 833 E 88: 8888 828.. 8:» 8H .38: 88 8m 88:8 888:8 a 828:: 033 8:.

2 > 2 888:8 Hm—
888:8 of

888:8 o3
888:8 v3
888:8 NE

2:

mo fl
888:8 3

888:8 mm

888:8 mm

888:8 _m

888:8 an

on

mu

888:8 we

3 3 > 888:8 vw
H H < H < H < O mmo

033 mag 88. 3:2 _3 $29 mfg H32 mflf 2 933 g Sog Keg

 

fir»! E 88888 88880 "a 03: H

52

 

8:828: 8505.8 8 Z :0 .< :8 0 88-2 “.4: :8 H 88.? .0 :8 < 58.8 .H :8 0 58% 6588
-0 6583... H 65830.0 6:808-< :8 :88 v88. 2: .88 88 8 880:: 88 2: w58om:5 808— 3 :80: 08 88888:
:28 .8588 :o: 83 8888 853 5 808 085:5 888: 88 8H 88: 88 :8 8888 888:8 a 8:38: 033 8:

 

888:8 Hm.

 

888:8 cm—

 

>
>
>

888:8 av—

 

888:8 3.—

 

mzmcomcoo N?—

 

>3

888:8 X:

 

888:8 No—

 

888:8 3

 

888:8 mm

 

888:8 mm

 

888:8 ;

 

888:8 ”H

 

888:8 8.

 

888:8 mH

 

888:8 e:

 

 

>

>

888:8 v:

 

0

H

U

U

0

mxu

 

name:

mom:—

Nome.

 

Hmmm:

 

enme—

 

m9”:—

 

2m:—

 

acme.

 

3mg

 

com:—

 

mom:—

 

gmf

 

3m:—

 

«Hmo—

 

vane—

 

mmme _

 

EN:—

 

3N2

 

8:8

 

 

 

 

«A; 5 898.58 88880 “a 03: H

53

 

APPENDIX C:

Sequences for Buccal Samples

Table 10: Sequences for buccal swabs in HV2

41 42 43 6O 72 73 95 120 146 152 153 162 195 204 210 245 253 263 272 295 309.1 309.2 310 315.1 343 373 384 395 456 462
A C -— —— T C A A A C C

CRS C T C T T A A C T T A C T T A T C
K1 G C G
K2 C
K3

K4

K5

K6

K7

K8

K9

K10

K11

K12

K13

K14

K15

K16

K17

K18

K20

K21

Y

QQQQQQQQQQOQQQQQQQQQ>

OOOOOOOOOOOOOOOOO

 

The sequences for each of the buccal swabs. The gray regions indicate areas in which sequence was not obtained. Any differences are noted by letters indicating the base present at that
spot. The letters stand for A-adenine, C-cytosine, T—thymine, G—guanine, Y-both C and T, R—both A and G, W—both T and A, M-both C and A, or N an unknown nucleotide.

54

 

 

Table 11: Sequences for buccal swabs in HVl

6

 

The sequences for each of the buccal swabs. The gray regions indicate areas in which sequence was not obtained. Any differences are noted by letters indicating the hase present at that
Spot. The letters stand for A-adenine, C—cytosine, T—thymine, G—guanine, Y-both C and T, R—both A and G, W-both T and A, M—both C and A, or N an unknown nucleotlde.

55

 

BIBLIOGRAPHY

Allard, M.W., K. Miller, M. Wilson, K. Monson, and Budowle B. 2002. Characterization
of the Caucasian Haplogroups Present in the SWGDAM Forensic mtDNA Dataset

for 1771 Human Control Region Sequences. Journal of Forensic Sciences.
47(6):1 - 9.

Armed Forces DNA Identification Laboratory. Armed Forces Institute of Pathology.
2005. <http://www.afip.org/Departments/oafine/dna/> Accessed 2005.

Anderson, S., A.T. Bankier, B.G. Barrel], M.H.L. deBruijn, A.R. Coulson, J. Drouin, I.C.
Eperon, D.P. Nierlich, B.A. Rose, F. Sanger, R.H. Schreier, A.J.H. Smith, R.
Staden, and LG. Young. 1981. Sequence and organization of the human
mitochondrial genome. Nature. 290:457 - 465.

Bejko, L. 1999-2000. Mortuary customs in the late bronze age of south-eastem Albania.
Iliria. 29(1— 2): 155 —- 181. .

Bejko, L. 1994. Some problems of the Middle and Late Bronze Age in southern Albania.
Bulletin of the Institute of Archeology. 31 :105 — 126.

Brown, W. 1980. Polymorphisms in mitochondrial DNA of humans as revealed by
restriction endonuclease analysis. Proc. Natl. Acad. Sci. 77(6): 3605-3609

Butler J .M. 2005. Forensic DNA Typing: biology, technology, and genetics of STR
markers. London: Academic Press.

Cambridge Ancient History v. 3(1) 2'“1 edition. 1982 The Prehistory of the Balkans; and
the Middle East and the Agean World, Tenth to Eighth Centuries B.C. Cambridge
University Press.

Chou Q, M. Russell, D.E. Birch, J. Raymond, W. Bloch. 1992. Prevention of pre-PCR
mispriming and primer dimerizatin improves low-copy number amplifications.
Nucleic Acids Res. 20: 1 717-1723

Collins, M.J., C.M. Nielsen-Marsh, J. Hiller, C.I. Smith, J .P. Roberts, R.V. Prigodich,
T.J. Wess, J. Csapo, A.R. Millard, and G. Turner-Walker. 2002. The survival of
organic matter in bone: A review. Archaeometty. 44(3): 383 — 394.

Foran, David. E-mail to author. 9/1/2005
Gilbert, M.T., A.J. Hansen, E. Willerslev, L. Rudbeck, I. Barnes, N. Lynnerup, and A.

Cooper. 2003. Characterization of genetic miscoding lesions caused by
postmortem damage. American Journal of Human Genetics. 72(1):48 — 61.

56

Giles, R.E., H. Blanc, H. Cann, and D. Wallace. 1980. Maternal Inheritance of Human
Mitochondrial DNA. Proceedings of the National Academy of Sciences of the
United States of America. 77 (11): 6715-6719

Hall, T.A. BioEdit: a user-friendly biological sequence alignment editor and analysis
program for Windows 95/98/NT. Nucl. Acids. Symp.. 41 :95 — 98.

Hofreiter, M., V. Jaenicke, D. Serre, Av. A. Haeseler, and S. Paabo. 2001. DNA
sequences from multiple amplifications reveal artifacts induced by cytosine
deamination in ancient DNA. Nucleic Acids Research. 23:4793 — 4799.

Kaestle, FA, and K.A. Horsburgh. 2002. Ancient DNA in anthropology: methods,
applications and ethics. Yearbook of Physical Anthropology. 45:92 - 130.

Kreader, CA. 1996. Relief of Amplification Inhibition in PCR with Bovine Serum

Albumin or T4 Gene 32 Protein. Applied and Environmental Microbiology. 62
(3): 1102-1106

Nesser, D, and S. Liechti-Gallati. 1995. Sex determination of Forensic Samples by
Simultaneous PCR Amplifiation of a-Satellite DNA from Both the X and Y
Chromosomes. Journal of Forensic Sciences. 40 (2): 239-241

O’Rourke, D.H., G. Hayes, and SW. Carlyle. 2000. Ancient DNA Studies in Physical
Anthropology. Annual Review of Anthropology. 29:217 — 242.

Strom, C .M., and S. Rechitsky. 1998. Use of nested PCR to identify chaired human
remains and minute amounts of blood. Journal of Forensic Sciences. 43(3):
696 — 700.

Tuross, N. 1994. The biochemistry of ancient DNA in bone. Experianta 50: 530-535

Wilson, I. 1997. Inhibition and Facilitation of Nucleic Acid Amplification. Applied and
Environmental Microbiology 63(10) 3741-3751

57

RA

II"LilliflfllilfliiiililillliHIiii!

2736 7014