35E UIHHIWillilmllhlmllflWNIH”!WWW

 

 

.7, MLABRARY
. it} i an Stat
2009 Un?versity e

 

This is to certify that the
thesis entitled

A HIGHLY SENSITIVE SEX DETERMINATION ASSAY FOR
LOW QUALITY DNA

presented by

BRIANNE MICHELLE KILEY

has been accepted towards fulfillment
of the requirements for the

MS. degree in Forensic Sciences

 

 

 

 

\—/Major Professor’s Signature

3/7/09

Date

MSU is an Affirmative Action/Equal Opportunity Employer

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

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5/08 K:IProj/Acc&Pres/ClRC/DateDue.indd

A HIGHLY SENSITIVE SEX DETERMINATION ASSAY FOR LOW QUALITY
DNA

By

Brianne Michelle Kiley

A THESIS
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
MASTER OF SCIENCE

Forensic Sciences

2009

ABSTRACT

A HIGHLY SENSITIVE SEX DETERMINATION ASSAY FOR LOW QUALITY
DNA

By
Briaime Michelle Kiley
The sex of an individual is one of the most important attributes when developing a
physical profile for a forensic investigation. Current molecular sexing techniques work
well with high quality samples; however, when analyzing challenging samples, such as
ancient skeletal remains or tissue with little DNA, a more sensitive sexing method is
required. To represent the standard molecular sexing technique, the single copy
amelogenin gene on the X and Y chromosomes was amplified using SYBR® Green real-
time PCR. In a comparative assay, DYZI , a high copy repetitive element on the Y
chromosome, was multiplexed with the Ya5 Alu repetitive sequence as a human and
female control, and amplified using TaqMan® technology. The sensitivity and accuracy
of each method were compared by analyzing both high molecular weight and artificially
degraded DNA, along with low quality DNA extracted from hair shafts and ancient
skeletal remains. The DYZl/Alu assay was successful using as little as 0.0623 pg of high
molecular weight DNA, whereas the amelogenin assay required 1500 times more DNA.
Further, the DYZl/Alu assay required even less artificially degraded DNA for a positive
result, the exact opposite finding using amelogenin. The DYZl/Alu assay correctly sexed
28 of 30 hair shafts and yielded sexing results for all 41 bones, while the amelogenin
assay correctly sexed 16 of 30 hair shafts and yielded sexing results for only 12 of 41
bones. Ultimately, the DYZl/Alu assay proved to be a far more sensitive and accurate

sexing technique than the standard amelogenin sexing method.

ACKNOWLEDGEMENTS

I would first like to thank my advisor, Dr. David Foran, for his guidance and
support throughout my research and graduate education at Michigan State University.
The opportunities and experiences he has provided will benefit my professional career
greatly. I would also like to acknowledge Dr. Todd F enton and Dr. Vincent Hoffman for
sitting on my thesis committee and providing excellent advice for my project. In addition,
thanks goes to Carrie Jackson for her previous work on this assay and Elizabeth Graffy,
Christina Rauzi, and Stephanie Rennick for the DNA they extracted from hair and bone.
Without their work, my research would not have been nearly as successful. For their
support and uplifting attitudes, I would like to thank my fellow graduate students in the
Michigan State University Forensic Sciences Program.

Finally, I would especially like to thank my family, friends and Adam Gobeski.

Without their love, support, and patience, none of this would have been possible.

iii

TABLE OF CONTENTS

LIST OF TABLES ....................................................... vi
LIST OF FIGURES ..................................................... vii
INTRODUCTION ....................................................... 1
Current Sexing Techniques used in Forensic Laboratories ..................... 1
Anthropological Sexing Techniques .................................... 1
Molecular DNA Analysis and Sexing Techniques ........................ 2
Alternative Molecular Sexing Techniques .................................. 3
Modifications to Amelogenin Amplification ............................. 3
Molecular Sex Determination Targeting Different Loci .................... 3
Challenges with Molecular Sex Determination .............................. 4
Targeting Multicopy Loci for Sex Determination ............................ 5
Repetitive Sequences on the Y Chromosome ............................ 6
Real-Time Polymerase Chain Reaction .................................... 7
SYBR® Green Real-Time PCR ...................................... 11
TaqMan® PCR ................................................... 13
Research Goals ...................................................... 14
MATERIALS AND METHODS ........................................... 24
Control DNA ........................................................ 24
DNase Digestion .................................................... 24
Hair Shafi DNA ..................................................... 24
Ancient Skeletal DNA ................................................ 26
Amelogenin Assay ................................................... 28
AMEL-X and AMEL-Y Primer Optimization ........................... 28
Amelogenin Assay Parameters ...................................... 30
DYZl/Alu Assay .................................................... 30
Alu Primer and Probe Optimization ................................... 30
DYZl Primers and Probe Optimization ................................ 31
DYZl/Alu Assay ................................................. 33
Contamination Control ................................................ 33
RESULTS ............................................................. 35
Optimal Sexing Assay conditions ........................................ 35
Amelogenin Assay ................................................ 35
DYZl/Alu Assay ................................................. 37
DYZl/Alu Assay UV Treatment ..................................... 38
HMW and Digested DNA Dilution Series ................................. 39
Sexing of Hair Shaft DNA ............................................. 41
Amelogenin Assay ................................................ 43
DYZl/Alu Assay ................................................. 47

iv

Ancient Skeletal Sexing Results ........................................ 49

Adult Butrint Skeletal Remains ...................................... 51
Adult Karnenica Skeletal Remains .................................... 57
Juvenile Butrint Skeletal Remains .................................... 71
DISCUSSION .......................................................... 73
WORKS CITED ........................................................ 88

Table 1:

Table 2:

Table 3:

Table 4:

Table 5:

Table 6:

Table 7:

Table 8:

LIST OF TABLES

Analyzed Hair Samples ........................................... 25
Analyzed Skeletal Samples ......................................... 26
Amelogenin Primer Sequences ...................................... 29
Alu and DYZl Primer/Probe Sequences .............................. 32
Amplification success of the High Molecular Weight DNA

and DNase Digested Male DNA Dilution Series ....................... 4O
Sexing Results of Hair Shaft DNA .................................. 42

Molecular Sexing Results and Anthropological Sex for Ancient Adult Bones. .50

Sexing Results for Butrint, Albania Juvenile Skeletal Samples ............. 71

vi

Figure 1:
Figure 2:
Figure 3:
Figure 4:
Figure 5:
Figure 6:
Figure 7:
Figure 8:

Figure 9:

Figure 10:
Figure 11:
Figure 12:
Figure 13:
Figure 14:
Figure 15:
Figure 16:
Figure 17:
Figure 18:
Figure 19:
Figure 20:

Figure 21:

LIST OF FIGURES

Location of Repetitive Elements on the Y Chromosome .................. 7
Diagram of the Polymerase Chain Reaction ............................ 8
A Real-time PCR Amplification Curve ............................... 10
SYBR® Green Dye Properties ..................................... 11
SYBR® Green Dissociation Curve .................................. 12
TaqMan® Real-time PCR Probe .................................... 13
Map of the Triconch Palace and Merchant’s House ..................... 16
Bottom Layer of the Kamenica Tumulus ............................. 19
Top Layer of the Kamenica Tumulus ................................ 21
Amelogenin Sequences and Primer Location ......................... 28
Amelogenin Amplification and Dissociation Curve for Male DNA ....... 36
Amelogenin Amplification and Dissociation Curve for Female DNA ...... 37
DYZl/Alu Assay Amplification of Male and Female DNA .............. 38
DYZl/Alu Assay UV Treatment Agarose Gel ........................ 39
Amplification Plots of 20000 pg HMW DNA using the DYZl/Alu Assay . . 41
Inconclusive Sexing Results of Hair using the Amelogenin Assay ........ 43
Female Hair 16 Sexing Results using the Amelogenin Assay ............ 44
Female Hair 38 Sexing Results using the Amelogenin Assay ............ 45
Male Hair 17 Sexing Result using the Amelogenin Assay ............... 46
Male Hair 55 Sexing Result using the Amelogenin Assay ............... 47
Female Hair 19 Sexing Result using the DYZl/Alu Assay .............. 48

vii

Figure 22:
Figure 23:
Figure 24:
Figure 25:
Figure 26:
Figure 27:
Figure 28:
Figure 29:
Figure 30:
Figure 31:
Figure 32:
Figure 33:
Figure 34:

Figure 35:

Female Hair 28 Sexing Result using the DYZl/Alu Assay .............. 49
Sexing Results for Burial 1226 using the Amelogenin Assay ............ 52
Sexing Results for Burial 1518 using the Amelogenin Assay ............ 54
Sexing Results for Burial 1518 using the DYZl/Alu Assay .............. 57
Sexing Results for Burial 142 using the Amelogenin Assay ............. 59
Sexing Results for Burial 156 using the Amelogenin Assay ............. 62
Sexing Results for Burial 97 using the Amelogenin Assay .............. 65
Sexing Results for Burial 97 using the DYZl/Alu Assay ................ 68
Sexing Results for Burial 142 using the DYZl/Alu Assay ............... 69
Sexing Results for Burial 156 using the DYZl/Alu Assay ............... 7O
Sexing Results for Burial 1188 using the DYZl/Alu Assay .............. 72
Magnified View of Burials 142 and 156 ............................. 82
Map and Photograph of Burials 1518 and 1517 ....................... 83
Map and Photograph of Burials 1188 and 1189 ....................... 85

viii

INTRODUCTION

Identifying the sex of unknown persons can be extremely important in a forensic
investigation and is often the first step in creating a physical profile of an individual.
Currently, anthropological and molecular sexing techniques work well with high quality
material; however, when faced with challenging samples, these techniques can become
inadequate. Anthropological sex estimation works well with full adult skeletal remains,
but in other instances, sex estimation can be unreliable. Furthermore, most molecular
sexing techniques target one location in the genome and work best with larger amounts of
high quality DNA. Unfortunately, when DNA is both low in quality and quantity, as from
aged bone or hair shafts, single copy loci may not be sensitive enough for sex
determination. On the other hand, using an assay sensitive enough to detect low copy
number (LCN) and poor quality DNA, sex can potentially be determined from otherwise

challenging sources.

Current Sexing Techniques used in Forensic Laboratories
Anthropological Sexing Techniques

Anthropological analysis is the most common method for sexing skeletal remains.
Without the estimation of sex, other characteristics, such as age and ancestry, are difficult
to determine (Kimmerle et a1. 2008). Typically, the sexually dimorphic characteristics of
bones, most commonly from the pelvis and the skull, are utilized. However, difficulties
can arise that make estimating sex challenging. For example, Komar and Potter (2007)

found that the probability of obtaining an accurate profile, including sex estimations,

decreases from 89% to 58% when less than 50% of a skeleton is available for analysis.
Similarly, subadult remains are difficult to sex. Even though adult sexual morphological
differences are present in juveniles, they are not fully developed, making sex estimates
unreliable (Cox and Mays, 2000). In addition, Kimmerle et al. (2008) discussed that
biological differences in general robustness, sexual dimorphism, and stature vary among

populations and can lead to unreliable estimations of sex.

Molecular DNA Analysis and Sexing Techniques

When anthropological estimations become unreliable, molecular sexing
techniques can be implemented. Forensic biologists analyze DNA, generally from body
fluids, hair, tissues, and touch samples, to form a DNA profile for forensic casework,
missing persons, parentage testing, DNA databasing, and mass disaster victim
identification (Butler, 2007). Sex is often determined by detecting the amelogenin gene
on the X and Y chromosomes. The most common amelogenin sex test was developed by
Sullivan et al. (1993) in which homologous regions of the gene are targeted on the X and
Y chromosomes; however, the sequence on the X chromosome is 6 basepairs (bp) shorter
than the Y chromosome sequence. Therefore, since females only have X chromosomes,
one product of 106 bp will be observed. Conversely, males, with both an X and Y

chromosome, produce 106 bp and 112 bp products respectively.

Alternative Molecular Sexing Techniques
Modifications to Amelogenin Amplification

The amelogenin gene was first sequenced by Nakahori et al. (1991). The X
chromosome amelogenin gene is 2872 bp long and is located at p22, while the Y
chromosome amelogenin gene, located at 11p12.2, is 3272 bp long (N akahori et al.
1991; Haas-Rochholz and Weiler, 1997). There are 19 homologous regions of the
amelogenin gene on the two chromosomes; however, differences between them include 5
deletions on the X gene ranging from 1 to 6 bp in length, and 5 deletions on the Y gene,
ranging fiom 1 to 183 bp in length (Haas-Rochholz and Weiler, 1997).

In addition to the Sullivan et al. (1993) amelogenin sexing technique, additional
sexing methods have been developed that utilize alternative amelogenin regions.
Nakahori et al. (1991) were the first to describe an assay that targeted a 177 bp insertion
on the X chromosome. Bailey et a1. (1992) assayed the first intron of the amelogenin
gene, producing a 542 bp product for the X-chromosome and a 358 bp product for the Y
chromosome. Similarly, Faerrnan et al. (1995) created a technique specific for the 189 bp

deletion on the first intron of the Y-chromosome.

Molecular Sex Determination Targeting Different Loci

Additional molecular sexing techniques have been developed that are specific for
non-amelogenin sequences. For example, the zinc finger regions on the X and Y
chromosomes are identical in size, but differ in sequence. A restriction digest with the
enzyme Hae III digests the X and Y zinc finger sequences at different positions on each

chromosome. When separated by gel electrophoresis, a zinc finger sequence in females

has two bands present, while a male zinc finger sequence has four bands (Reynolds and
Varlaro, 1996; Stack and Witt, 1996). In addition, Witt and Erickson (1989) developed a
sex assessment assay targeting alphoid satellite DNA, consisting of highly repetitive,
non-coding sequences located in the pericentromeric regions on the X and Y
chromosome. This assay targeted a 130 bp sequence of alphoid satellite DNA on the X
chromosome and a 170 bp sequence on the Y chromosome. Similarly, the sex
determining region Y (SRY) gene has also been utilized for sex assessment (Bartha et al.

2003).

Challenges with Molecular Sex Determination

Most molecular sexing techniques, including those detailed above, require less
than 1 ng of DNA; however to increase amplification signal, larger amounts of DNA are
generally used (Nesser and Liechti-Gallati, 1995). Unfortunately, forensic evidence, such
as fingerprints, shed hair, aged bone and teeth, often contain scarce quantities of DNA
that limits analysis (Andréasson and Allen, 2003), making these techniques insufficient
for sex determination. In addition, molecular sexing techniques are affected by DNA
degradation. DNA degrades both biologically and chemically by hydrolysis and
oxidation, creating difficulties in its analysis (Poinar, 2003). Environmental factors, such
as humidity, temperature, and pH can influence the degradation process (Burger et al.
1999; Poinar, 2003). Because degradation lowers the quality of DNA, assaying a single
copy locus or large DNA sequences can be difficult, since an intact copy of the desired

sequence may no longer be available. This can be very problematic in molecular sexing

methods, as failure to detect the Y chromosome will result in erroneously typing a sample

as female.

Targeting Multicopy Loci for Sex Determination

The molecular sexing methods detailed above all assay single copy loci, thus are
susceptible to being missed if very small amounts of DNA are being analyzed. Budowle
et al. (2001) explained that with fewer copies of DNA template, stochastic effects can
occur, resulting in the exclusion of a target locus. Stochastic effect can cause inconsistent
detection of a single copy locus when sexing low quantities of DNA, thereby making sex
determination unreliable.

When testing small amounts of DNA, overcoming stochastic effects can
potentially be achieved by assaying high copy loci. Unlike single copy loci, multicopy
loci provide a more abundant template for DNA analysis, thereby increasing the chance
of obtaining the desired product. The most abundant repetitive element in human DNA is
the Alu sequence, found on every chromosome. Alu repeats make up 6 — 13% of the
human genome, with an estimated 500,000 to 1,000,000 copies (Mighel et al. 1997).
These repeats are primate specific and make an excellent target for identifying human
DNA (Mi ghel et al. 1997). The Alu sequence is approximately 280 bp long and is made
up of two similar monomers connected by an adenine rich region (Mighel et al. 1997).
The Promega Corporation developed a DNA quantification technique that utilizes the Alu
repetitive regions, the AluQuantTM Human DNA Quantitiation System, which quantifies
human DNA to quantities as low as 100 pg (approximately 20 cells worth) (Mandrekar et

a1. 2001). Similarly, Nicklas and Buel (2003) utilized a specific subfamily (Y a5) of the

Alu sequence for human DNA quantification. By targeting the Ya5 subfamily, the assay
quantified human DNA down to 1 pg, 100 times more sensitive than the AluQuantTM

method.

Repetitive Sequences on the Y Chromosome

High copy loci are also available on the Y chromosome that are ideal for sensitive
sex determination assays. Sixty percent of the Y chromosome is comprised of long
repetitive sequences, with only a small number of functional genes (Roewer et al. 1992).
The DYZ family, numbered in order of frequency, makes up most of the repetitive
elements on the Y chromosome, with the approximate locations for each displayed in
Figure 1. DYZ] and DYZZ are 3.4 kb and 2.4 kb repeated sequences, respectively,
located on the long arm of the Y chromosome (Roewer et al. 1992; Rahman et al. 2004).
DYZ3 is a block of alphoid satellite DNA positioned on the centromeric region of the Y
chromosome and is 5.7 kb long with 170 bp repeating subunits (Tyler-Smith et al. 1988).
On the short arm of the Y chromosome are blocks of low repetitive sequence blocks
called DYZ4 and DYZ5 (Roewer et al. 1992). These sequences have been assayed for
male identification. For example, Nicklas and Buel (2006) utilized the DYZS repeat
sequence for male DNA quantification in which 15 — 50 copies of a 137 bp repeat within
the sequence was targeted. In addition, Rahman et al. (2004) assayed the 2000 — 4000
copies of the DYZl repetitive element for sex determination and diagnostic studies of the

Y chromosome.

Heterochromatin

. 1",”. .l ‘17,. ".4,

    

............. ,_ ,,
0§V9&fifihh u
000000000 ».

0 0 00000000(_ »

0 0 000000000.

000 00 00000<.

0 0 0 00 .........,.

0 0 000 00000000001 »

0000 00 0 00 0000000

§ §UVVUU%§J’ § 00

0000000000

...Q.0.0.0a0.0.0.00.0.0...-.........-......... ...-.

0st DYZ3 DY22 DYZ1

Figure 1: Location of Repetitive Elements on the Y Chromosome

A diagram of the Y chromosome with the approximate location of SRY, amelogenin
(AMEL-Y), and the DYZ family of repetitive elements. The SRY, AMEL-Y, DYZl and
DYZ5 are located on the short arm (Y p). DYZ3 is located in the centromeric region of
the Y chromosome and illustrated with the arrow. DYZl and DYZ2 are located on the
long arm (Y q). The grey box indicates the heterochromatin region on the long arm of the
Y chromosome.

Real-Time Polymerase Chain Reaction

The polymerase chain reaction (PCR) is a process by which a specific region of
DNA is copied (doubled) over and over, which after 30 repetitions, can turn one copy of
DNA into a billion (Saiki et al. 1988) (Figure 2). PCR is initiated by the denaturing step
in which the starting (template) DNA is separated into two strands. The temperature is
lowered and complementary primers anneal to the DNA, flanking the DNA sequence to
be studied. This is followed by enzymatic extension of the primers using a DNA
polymerase extending and free nucleotides, thereby creating a copy of the desired region,

termed an arnplicon.

 

:llllllllllllllllll:

 

 

DENATURE
5'IIIIIIIIIIIIIIIIII3'
3.lllllllllllllllllls.

l ANNEAL PRIMERS

I 5'IIIIIIIIIIIIIIHH3
3.IIIIlLllllllI11J115.

EXTEND PRIMERS

 

 

SIIIIIIIIIIII

  

  

 

Lllllllllllls.

 

Zimnmnnnnn:
ZIIIHHIIIIIIHIII:

 

Figure 2: Diagram of the Polymerase Chain Reaction

(A) Double stranded template DNA is denatured (B) and becomes single stranded. (C)
Complementary primers anneal to the template DNA. (D) DNA polymerase (oval)
extends the primers resulting in (E) a copy of the desired DNA sequence. This cycle is
repeated several times to create billions of copies of the target DNA sequence.

To monitor the amount of DNA being amplified during PCR, 3 process called
real-time PCR (rtPCR) can be implemented. Several rtPCR strategies (detailed below)
exist that utilize different types of fluorescent dyes to monitor DNA amplification;
however, all are based on the same principles. During rtPCR, a laser is used to excite the
dyes, creating a fluorescent signal. This signal is monitored by a camera within the rtPCR
detection system, which then displays the signal on an amplification plot, (Figure 3). The
detected signal is measured in relative fluorescent units (RFUs) and plotted against the
number of PCR cycles.

During the early cycles of rtPCR, before a large amount of desired product has
been produced, a certain amount of background noise can be observed. Therefore, a
threshold RFU value is determined that sits above the background activity. The
relationship of the amplification curve to the threshold can be utilized as a qualitative
pass/fail test: if the signal crosses the threshold, then product DNA was present. The
cycle in which the product signal crosses the threshold is called the cycle threshold (Ct)
value. The Ct value can be used to quantify the amount of DNA present in the PCR
reaction; the more DNA present, the earlier the amplification can be measured, resulting
in a lower Ct value. This value can also be used to directly compare two or more DNA
samples (which has more of the target sequence), or compare Ct values of known

amounts of DNA to get an actual measure of the quantity of DNA product in a reaction.

 

RFUs

 

Ct value

Cycle Number

Figure 3: A Real-time PCR Amplification Curve

A typical amplification plot shows the number of PCR cycles plotted against relative
fluorescent units (RFUs) values. As more DNA is amplified, the fluorescent signal
increases, represented by the sigmoidal curve. When reaction components become
limited, product accumulation stops and the fluorescent signal plateaus. The segmented
line represents the threshold value. The cycle threshold (Ct) value illustrates the cycle
where the amplification signal crosses the threshold.

10

S YBR® Green Real-Time PCR

SYBR® Green is an intercalating dye that fluoresces when bound to double
stranded DNA. As demonstrated in Figure 4, when two complementary strands of DNA
anneal, the SYBR® Green dye binds to the DNA and a fluorescent signal is emitted.

Therefore, as more double stranded DNA is synthesized during PCR, more fluorescence

 

is emitted.
Decrease Increase
Temperature Temperature

\l/ \l/ \l/ \l/ \l/ \l/

fillfillfillftlllallf‘il

/|\ /|\ /|\ /|\ /|\ /|\

   

Figure 4: SYBR® Green Dye Properties

The ovals represent the SYBR® Green dye. As double stranded DNA is synthesized
during PCR, more dye binds, thereby causing an increase in fluorescence. Conversely,
when the reaction temperature increases and the double stranded DNA denatures to
become single stranded, the SYBR® Green dye does not bind and no fluorescence is
emitted.

One disadvantage with SYBR® Green rtPCR is that dye binding is non-specific.
Therefore, if any non-target DNA is present, it too will generate a fluorescent signal (Fey

and Parkes, 2001). To distinguish desired amplicons from non-target DNA, a dissociation

11

reaction is run post-amplification, in which the temperature of the solution is slowly
increased (Figure 5). At a temperature specific for the amplicon, the DNA dissociates,
causing the SYBR® Green dye to fall away, decreasing the fluorescent signal. The real-
time detection system monitors this change in fluorescence and creates a unique
dissociation curve that can differentiate products within the reaction (F oy and Parkes,

2001).

IEA

E
>

RFUs
-d RFU
dT

 

 

 

 

O 9 O >
Temperature (C ) Temperature (C )

Figure 5: SYBR® Green Dissociation Curve

(A) An example of a standard dissociation curve performed after SYBR® Green real-
time PCR. With a slow increase in temperature, strand dissociation occurs, causing the
SYBR® Green dye to fall away, leading to a drop in fluorescence. Amplicons dissociate
at specific temperatures creating a characteristic dissociation curve for a desired products
within the reaction. The dissociation curve is displayed on a plot of RF U verses
temperature (C°). (B) When monitoring fluorescence, the real-time detection system also
plots the change in fluorescence over time (d(RFU)/dT) against an increase in
temperature. This creates a peak at a DNA product’s specific dissociation temperature.

12

T aqMan® PCR

Unlike SYBR® Green rtPCR, TaqMan® rtPCR technology uses fluorescent
probes to monitor the amplification of a specific DNA sequence. The probe is a short
sequence of DNA that is complementary to an internal region of an arnplicon, attached to
either end of the probe are a reporter dye and a non-fluorescent quencher (Figure 6).
When the two are in close proximity, the fluorescence of the dye is quenched, and no
signal is seen by the camera. During the PCR annealing step, the probe binds to the target
sequence, allowing the DNA polymerase to cleave the dye fi'om the probe as it builds the
new strand of DNA. This separates the quencher from the reporter dye, allowing it to

emit a fluorescent signal. The more target DNA that exists, the more probes are cleaved

and fluorescent signal produced (Foy and Parkes, 2001).

IllrrTrrIIHHU

primer

 

 

 

DNA Sequence

Reporter Dye

Quencher

 

Figure 6: TaqMan® Real-time PCR Probe

A TaqMan® probe has a reporter dye on one end and a quencher on the other end. When
the two are in close proximity, the fluorescence of the dye is quenched. After the probe
anneals to the target sequence, polymerase begins to extend the primers. The reporter dye
is then cleaved from the probe, resulting in increased florescence.

l3

Research Goal

The object of the research presented here was to develop a sexing technique that
is more sensitive than the current sexing methods for use on low quality DNA, as seen
with forensic evidence. The standard molecular sexing technique used in forensic
laboratories was represented by amplifying the amelogenin gene using SYBR® Green
rtPCR. In contrast, the DYZl and Alu repetitive elements, being unrelated to one another,
could be typed simultaneously (multiplexed), and were analyzed using TaqMan®
technology. Analysis of high molecular weight (HMW) and artificially degraded DNA,
along with low quality tissue DNA, were performed using the amelogenin and the
DYZl/Alu assay and the sensitivity and accuracy of the methods compared.

The Michigan State University (MSU) Forensic Biology Laboratory regularly
works with sub-optimal DNA samples. For example, Graffy and Foran (2005) obtained
hair from 30 individuals, removed the roots, and isolated hair shaft DNA using two
techniques: a standard glass grinding/organic extraction and a novel alkaline digestion
they developed. Nuclear DNA is notoriously difficult to obtain from hair shafts; therefore
its testing is largely ignored. Since these hair shafts were expected to contain low quality
nuclear DNA and the sex of the hair donors was known, they were ideal for analysis with
the sex determining assays. The DNA from the hair shafts was sexed in a blind study
with the amelogenin and DYZl/Alu assays to compare the accuracy of each sexing
technique.

The Forensic Biology Laboratory also analyzes DNA extracted from human
skeletal remains, which may be fresh or quite ancient. Once again, it is difficult or often

impossible to obtain nuclear DNA results from ancient skeletal material. The bones used

14

in this study were collected from two ancient Albanian burial sites, the Triconch Palace
and Merchant’s House in the ancient city of Butrint, and a large tumulus (burial mound)
located in Kamenica. The Triconch Palace and Merchant’s House, a complex of buildings
near the Butrint city wall on the bank of the Vivari Channel, date to 5th — 7th century AD.
and 13th — 15th century A.D., respectively. Hodges et al. (2000) described the Triconch
Palace as originally a single family dwelling that later became a semi-public building.
During the excavations of the Palace, numerous skeletal remains were unearthed within
the rooms of the structure. A map of the burials in the Triconch Palace and Merchant’s
House is displayed in Figure 7. Multiple burial clusters are located in the palace, with an
additional cluster in the southwest corner of the Merchant’s House. Rauzi (2007)
extracted DNA from several of these remains and analyzed the mitochondrial sequences

to examine the maternal relatedness of the individuals within each cluster.

15

Figure 7: Map of the Triconch Palace and Merchant’s House

Burial locations are displayed along with the palace rooms and the Merchant’s House in
the southwest comer. Letters distinguish the rooms of the palace. Burials are labeled with
a four-digit number and depicted as a bold line ending with a dot showing the location of
the head among the remains.

l6

 

17

The tumulus in Kamenica is the largest tumulus in Albania, with a diameter of 40
m, containing approximately 800 burials. As depicted in the burial maps in Figures 8 and
9, the original structure of the tumulus consisted of a double ring of rocks, the “Great
Circle” 15 min diameter, containing 35 burials covered in soil, and dating to 1200 — 1150
BC. Two hundred additional burials were placed on top of the central circle, and dated to
1150 — 650 BC. Around 650 BC, the burial pattern changed in which individuals were
covered with rocks instead of buried under soil. These were encircled by rock walls that
interconnected to make large formations called “Monumental Structures.” Monumental
Structure 1.1 and 3 are illustrated on the maps in Figures 8 and 9.

Most of the inhumations at Kamenica consisted of only one individual; however,
there were numerous “double burials” or two individuals per grave, some of which are
labeled with stars in Figures 8 and 9 (Bejko et al. 2006). Clemmer (2003) compared the
mitochondrial DNA of several of the double burial remains to examine their maternal
relatedness. In addition, Rennick (2005) extracted DNA from several remains within
Monumental Structure 3 and analyzed mitochondrial DNA to determine if the structure

might be a family grave site, or was a mixture of unrelated individuals.

18

Figure 8: Bottom Layer of the Kamenica Tumulus

The map depicts the lower (oldest) layer of the tumulus (1200 — 650 BC). The double
ring of rocks is displayed in the upper left corner. Monumental Structure 1.1 is located in
the upper right corner and Monumental Structure 3 is in the lower left corner of the map.
Stars indicate double burials within the tumulus tested by Clemmer (2005).

19

l’

 

20

am

o.::;;—*. :*

 

Figure 9: Top Layer of the Kamenica Tumulus

The map depicts the upper layer of the tumulus (1 150 — 650 BC). The double ring of
rocks is displayed in the upper left comer. Monumental Structure 1.1 is located in the
upper right corner and Monumental Structure 3 is in the lower left corner of the map.
Stars indicate double burials within the tumulus tested by Clemmer (2005).

21

 

,1
'6' ., f?
:r ‘
r 'p

22

The hair shafts and skeletal remains provide a good representation of the
condition of some evidence submitted to a forensic laboratory for DNA analysis.
Unfortunately, because of the sub-optimal nature of the DNA, such samples are often
overlooked for DNA analysis. If these samples were analyzed with the standard
molecular sexing techniques, sex determination would be inconsistent and unreliable. The
DYZl/Alu assay, however, yields a sexing result for these low quality samples, providing

valuable information for a forensic investigation.

23

MATERIALS AND METHODS

Control DNA
Single source male DNA (20 ng/uL), developed in the MSU Forensic Biology
Laboratory, was used as the male control. K562 High Molecular Weight DNA (Promega)

at a stock concentration of 10 ng/uL was used as a female control.

DNase Digestion

DNase I was used to digest 3.15 pg of single source male DNA, simulating
degraded DNA. The reaction was performed on ice and included 3 U of Roche’s
recombinant RNase free DNase I, 5 uL of 10x incubation buffer, and sterile water
(Fisher) to a final reaction volume of 50 uL. Fifteen microliter aliquots were removed at
time intervals of 0, 30, and 60 seconds and placed in 5 uL of 20 mM EDTA. Aliquots
were immediately placed in a 95°C water bath for 5 minutes to stop digestion. DNA fiom
each time aliquot was separated on a 2% gel to estimate the level of digestion.

HMW DNA and digested DNA were serially diluted 15-fold and molecularly
sexed in duplicate. HMW DNA ranged from 20 ng to 1.76 fg, while the digested DNA

ranged from 47 ng to 4.13 fg.

Hair Shaft DNA
Volunteers’ hair shafts, along with questionnaires regarding sex, population
ancestry, hair color, and hair treatments (Table 1), had been previously collected (Graffy

and Foran, 2005). DNA was extracted from the hair shafts by one of two methods: a glass

24

grinding/organic extraction or an alkaline digestion. In the current research, the hair shaft

DNA was analyzed blind for sex determination.

Table l: Analyzed Hair Samples
The sex and population ancestry is recorded, along with any hair treatments performed.
(Graffy, 2003). Hair without treatment is denoted by (---).

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Hair Declared Population Hair Treatment
Identifier Sex Ancestry
8 Female African perm/relaxer within year
9 Female Caucasian dye within year
10 Female Caucasian blow dry
1 1 Male Caucasian ---
12 Female Caucasian blow dry
13 Female Hispanic blow dry/dye within month
14 Female Caucasian dye within month
16 Female Asian blow dry/dye within year
1 7 Male Caucasian ---
19 Female Caucasian dye within month
20 Female Asian blow dry
22 Male Caucasian —--
23 Female Afiican blow dry /perm/relaxer within year
24 Male Caucasian blow dry
27 Male Asian blow dry
28 Female Caucasian blow dry/ dye within year
29 Male Caucasian ---
32 Female Caucasian ---
35 Female Caucasian blow dry/ dye/perm/relaxer within
36 Female Caucasian blow dry/ dye within month
37 Female Caucasian blow dry/ dye within month
38 Female Hispanic blow dry/ dye within year
43 Female Asian blow dry/ dye within month
46 Male Caucasian ---
54 Female African blow dry
55 Male African ---
56 Female African perm/relaxer within month
57 Female Asian ---
59 Female Asian ---
60 Female Afiican dye/perm/relaxer within month

 

25

 

Ancient Skeletal DNA

DNA was extracted from skeletal samples excavated from two Albanian
archeological sites: the Triconch Palace (5th — 7th century AD) and Merchant’s House
(13tb — 15‘h century AD) in Burtrint, Albania (Rauzi, 2007), and Monumental Structure 3
within the tumulus in Kamenica, Albania (7th —- 6th century BC) (Rennick, 2005). Table 2
displays the specific bones from which DNA was extracted, the century the remains were

buried, and the anthropological age and sex estimation for each individual.

Table 2: Analyzed Skeletal Samples

An identifier was given to each bone corresponding to the respective burial site. The
specific bones are listed, along with the estimated age of each individual, the century the
remains were buried, and the anthropological age and sex estimates for each burial.
Partial or juvenile bones that were unable to be sexed by anthropological methods are
listed as unknown. Remains from Burtrint, Albania are labeled BA and remains from
Kamenica, Albania are labeled KA. (Rauzi, 2007; Rennick, 2005).

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

I (13:33” Specific Bone Century Age Anthrtgép‘logrcal

BA Bone
1185.1 fibula shaft 5 — 7 AD. Adult Unknown
1187.1 right radius 5‘“ — 7‘“ AD. 30 — 45 Male
1187.2 right fibula 5‘“ — 7‘“ AD. 30 — 45 Male
1188.1 left humerus shaft 5‘“ — 7‘“ AD. 11 — 13 Unknown
1188.2 right petrous 5‘“ - 7‘“ AD. 11 — 13 Unknown
1224.1 left petrous 5‘“ — 7‘“ AD. 3 — 4 yrs Unknown
1224.2 right humerus 5‘“ — 7‘“ AD. 3 - 4 yrs Unknown
1225.1 right petrous 5‘“ -— 7‘“ AD. 3 — 4 yrs Unknown
1226.1 right fibula 5‘“ — 7‘“ AD. 35 - 45 Male
1226.2 left ulna midshaft 5‘“ — 7‘“ AD. 35 — 45 Male
1229.2 humeral shaft 5th — 7ih AD. 3 - 6 mos Unknown
1236.1 fibula midshaft 5‘“ - 7‘“ AD. 23 — 27 Female
1236.2 left ulna midshaft 5‘“ — 7‘“ AD. 23 — 27 Female
1259.1 left femur 5‘“ — 7‘“ AD. 10 — 12 Unknown
1264.1 tibia cortical 5‘“ — 7‘“ AD. 6 - 8 yrs Unknown

fragment fr. Distal

1264.2 fibula midshaft 5th — 7th AD. 6 — 8 yrs Unknown

 

26

 

Table 2 (cont’d)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1518.1 humeral cortical 5‘ — 7 AD. 40+ yrs Male
1518.2 fibula midshaft 5‘h — 7th A.D. 40+ yrs Male
5010.1 fibula midshaft 5‘“ — 7‘“ AD. 35 — 45 Female
5010.2 humerus midshaft 5‘“ — 7‘“ AD. 35 — 45 Female
etrous portion
humerus 7 6 . . n/a
66 femur 7th — 6th B.C. n/a Male
75 tibia 6‘“ BC. 45+ yrs Male
75 femur 6th B.C. 45+ yrs Male
75 petrous portion 6‘“ BC. 45+ yrs Male
82 radius 7‘“ — 6‘“ BC. old Male
82 petrous portion 7th — 6th B.C. old Male
85 femur 7‘“ — 6‘“ BC. 22 — 28 Male
97 tibia 7‘“ — 6‘“ BC. n/a Male
97 femur 7‘“ ~ 6‘“ BC. n/a Male
102 humerus 7th — 6th B.C. middle Female
102 petrous portion 7th — 6Lh B.C. middle Female
104 humerus 7‘“ — 6‘“ BC. 22 — 28 Male
104 femur 7‘“ — 6‘“ BC. 22 — 28 Male
142 pelvis 7‘“ BC. 45+ yrs Male
142 petrous portion 7th B.C. 45+ yrs Male
142 humerus 7‘“ BC. 45+ yrs Male
156 femur 7‘“ BC. 29 — 35 Male
156 petrous portion 7‘“ BC. 29 — 35 Male
156 humerus 7‘“ BC. 29 — 35 Male

 

Each bone was drilled using a Dremel rotary tool (Dremel) and the resultant
powder was digested with proteinase K and digestion buffer. DNA was extracted using a
standard organic extraction and purified using Microcon-YM30 (Millipore) column

filters (Rauzi, 2007; Rennick, 2005).

27

Amelogenin Assay
AMEL-X and AMEL—Y Primer Optimization

The forward primers for the amelogenin assay (Table 3) were designed using
Primer3 software (Rozen and Skaletsky, 2007 ). The AMEL-X forward primer spanned
the 6 bp deletion on the amelogenin gene on the X-chromosome, while the AMEL-Y
forward primer spanned the homologous sequence on the Y-chromosome (Figure 10).
The British reverse (BR) primer (Sullivan et a1. 1993) was used with both forward
primers. The AMEL-X/BR and the AMEL—Y/BR primer sets resulted in 69 bp and 77 hp

products respectively.

x121

x161

 

x201

 

 

 

 

 

 

 

 

 

 

 

 

 

x241

 

 

 

 

 

6%
Figure 10: Amelogenin Sequences and Primer Location
A portion of the amelogenin sequence on the X and Y chromosome is displayed on the
top and bottom rows, respectively. The gray boxes show the 6 bp deleted from the X
chromosome. The lighter arrows denote the position of the X and Y forward primers,
while the darker arrow indicates the British reverse primer. The X chromosome primer
set resulted in a 69 bp product and the Y chromosome product was 77 bp. (Modified from
Sullivan et al. 1993).

28

The optimal annealing/extension temperature was determined using 15 11L
reactions including 1x Fast SYBR® Green Master Mix (Roche), 2 11M AMEL-X or
AMEL-Y, 2 11M BR, 1 ng male or female DNA, and sterile water (Fisher). Cycling
parameters were a 95°C hold for 10 min and 40 cycles of a 95°C denaturing step for 15
sec, and an annealing/extension gradient from 50°C to 60°C for 1 min. Dissociation
parameters were a 95°C hold for 1 min, a 55°C hold for 1 min, and a dissociation range
from 55°C to 95°C. PCR reactions were performed on an iQTMS Multicolor Real-Time
Detection System and iCycler (Bio-Rad), and analyzed with Bio-Rad iQ5 — Standard
Edition software. The products were separated on a 4% agarose gel to determine if
amplicons were the expected size. The annealing/extension temperature that yielded an
amplification curve that crossed the threshold and a peak on the dissociation curve at the
appropriate temperature were considered optimal. Optimization of AMEL-X, AMEL-Y,
and BR primer concentrations was performed at 200 nM, 1 11M, and 2 uM. Cycling and
dissociation parameters outlined above were used with the optimized annealing/extension
temperature. The primer concentrations that yielded an amplification curve that crossed
the threshold and a peak on the dissociation curve at the appropriate temperature were

considered optimal.

Table 3: Amelogenin Primer Sequences

Primer names and sequences are given. AMEL-X and AMEL-Y are the forward primers
targeting the 6 bp deletion region on the X and Y chromosome respectively. BR is the
British reverse primer used with both forward primers.

 

 

 

Primer Identifier Primer Sequence
AMEL-X 5’-TCCCAGATGT'ITCTCAAGTGG-3’
AMEL-Y 5’-CATCCCAAATAAAGTGGTITCTC-3’
BR 5’-ATCAGAGCT1'AAACTGGGAAGCTG-3’

 

 

 

 

 

29

Amelogenin Assay Parameters

The amelogenin assay included 15 11L reactions with either 1x Fast SYBR®
Green Master Mix (Roche) or 1x Power SYBR® Green Master Mix (Applied
Biosystems), 1 11M AMEL-X or AMEL-Y primers, 2 11M BR primer, 1 1.1L DNA, and
sterile water (Fisher). Reactions with samples subject to inhibition, such as hair or bone,
contained 10 pg BSA. Cycling and dissociation parameters were the same as those

optimized for the amelogenin primers. Amelogenin assays were run in triplicate.

DYleAlu Assay
Alu Primer and Probe Optimization

Alu primer and probe sequences (Table 4) were designed by Nicklas and Buel
(2005), yielding a product size of 127 bp targeted as a positive human control. The
TaqMan® Alu probe was labeled with Hex reporter dye on the 5’ end and the non
fluorescent quencher DBHl on the 3’end. Alu forward (Alu-F) and reverse (Alu-R)
primers were optimized at 200 nM, 500 nM, and 900 nM, using 1x Fast SYBR® Green
Master Mix (Roche), 1 ng male DNA, and sterile water (Fisher) to 15 11L. Cycling
parameters were a 95°C hold for 10 min, then 50 cycles of a 95°C denaturing step for 15
sec and 60°C annealing/extension step for 1 min. The dissociation parameters were a
95°C hold for 1 min, a 55°C hold for 1 min, and a melt range from 55°C to 95°C and
were performed to determine if the amplicon dissociated at the expected temperature.
Products were separated on a 4% agarose gel to determine if the amplicon was the
expected size. The primer concentrations and annealing/extension temperature that

yielded an amplification curve that crossed the threshold and a peak on the dissociation

30

curve at the appropriate temperature were considered optimal. An optimization annealing
temperature gradient assay, from 55°C to 63°C, was then performed. Optimized
conditions were used for subsequent experiments.

Alu probe concentration was optimized at 50 nM, 150 nM, and 250 nM. Male and
female DNA at 100 pg and a 1:15 dilution were used in 15 uL reactions including 1x iQ
Supermix (Bio—Rad), 500 nM Alu—F, 900 nM Alu-R, the respective Alu probe
concentrations, and sterile water (Fisher), and cycling parameters outlined above. The
probe concentration that yielded an amplification curve that crossed the threshold for

both male and female DNA was deemed optimal.

DYZI Primers and Probe Optimization

The primer pair and probe targeting the DYZl locus with an expected product
size of 143 bp were designed by William Kiffmeyer, Applied Biosystems Field
Applications Specialist (Jackson, 2005) (Table 4). The DYZl probe was labeled with 6-
FAM reporter dye on the 5’ end along with a minor groove binding (MGB) non-
fluorescent quencher on the 3 ’ end. DYZl primer optimization was performed using
concentrations of 200 nM, 500 nM, and 900 nM and the same cycling were used for Alu
primer optimization. An annealing temperature gradient assay, from 55°C to 63°C, was
also performed for primer optimization. The dissociation parameters were the same as
Alu primer optimization and were performed to determine if the amplicon dissociated at
the expected temperature. Fifteen microliter reactions included 1x iQTM SYBR® Green
Supermix (Bio-Rad), respective concentrations of DYZl forward (DYZl-F) and reverse

(DYZl-R) primers, 1 ng male DNA, and sterile water (Fisher). Products were separated

31

on a 4% agarose gel to determine if the amplicon was the expected size. The primer
concentrations and annealing/extension temperature that yielded an amplification curve
that crossed the threshold and a peak on the dissociation curve at the expected
temperature were considered optimal.

The DYZ1 probe concentration was optimized at 50 nM, 150 nM, and 250 nM
using the same cycling parameters as primer optimization above. Male and female DNA
at 100 pg and a 1:15 dilution were used in 15 11L reactions containing 1x iQ Supermix
(Bio-Rad), 500 nM DYZl-F, 500 nM DYZl-R, the respective DYZ1 probe
concentrations, and sterile water (Fisher). The probe concentration that yielded an
amplification curve that crossed the threshold at the lowest Ct value for both male and

female DNA was deemed optimal.

Table 4: Alu and DYZl Primer/Probe Sequences

Primer and probe names and sequences are given. DHEX and 6FAM are fluorescent dyes
on the 5’ end of the Alu and DYZ1 probe, respectively. DBHl is the non-fluorescent
quencher on the 3’ end of the Alu probe, while MGBNFQ is the minor groove binder
non-fluorescent quencher on the 3’ end of the DYZl probe.

 

 

 

 

 

 

 

 

 

 

Primer Identifier Primer Sequence
Alu-F 5'-GAGATCGAGACCATCCCGGCTAAA-3‘
Alu-R 5‘-CTCAGCCTCCCAAGTAGCTG-3'
DYZ1-F 5'-GGCCTGTCCATTACACTACATTCC-3'
DYZl-R 5'-GAATTGAATGGAATGGGAACGA-3'
Probe Identifier Probe Sequence
Alu-Probe 5'-DHEX-GGGCGTAGTGGCGGG-DBH1-3'
DYZl-Probe 5‘-6FAM-ATTCCAATCCATTCC'I‘TT-MGBNFQ-3'

 

32

DYZI/Alu Assay

Primers and probes for the Alu and DYZ] loci were then multiplexed in a
TaqMan® rtPCR assay. Fifteen microliter reactions included 1x iQ Supermix (Bio-Rad),
500 nM Alu-F and DYZ1-F, 900 nM Alu-R and DYZ1-R, 250 nM Alu and DYZ1
probes, 1 pL DNA, and sterile water (Fisher). Reactions with samples subject to
inhibition, such as hair or bone, contained 10 pg BSA. Each sample was analyzed in

triplicate using the optimized conditions detailed above.

Contamination Control

Pipetters, pipette tips, tubes, racks, and sterile water (Fisher) were placed in a
Spectrolinker TM XL-1500 UV Crosslinker (Spectronics Corporation). Reactions were
prepared in a Clean Spot PCR/UV Work Station (COY Laboratory Products) and a
surgical mask, sleeves and two pairs of gloves were worn during preparation. Sterile
water (Fisher) was filtered through a Microcon YM-3O (Millipore) and ultraviolet (UV)
irradiated for 10 min. Primers for the DYZ1/Alu multiplex and the amelogenin assay
were diluted using the purified sterile water and filtered through a Microcon YM-30
(Millipore).

The DYZl/Alu multiplex, the iQ Supermix, primers and probes were UV
irradiated for 0 seconds, 30 seconds (2.73 J/em‘), 1.0 min (5.45 J/em“), 1.5 min (8.18
J/cmz), 2.0 min (10.90 J/cmz), 2.5 min (13.63 J/cmz), and 3.0 min (16.36 J/em‘). Fifteen
pL reactions included each UV treated master mixes, 10 pg BSA, and either 100 pg male

DNA as a positive control or sterile water as a negative control. A UV treatment that

33

yielded amplification of both loci for the positive control and no amplification for the

negative control was used for subsequent multiplex reactions.

34

RESULTS

Optimal Sexing Assay Conditions
Amelogenin Assay

Optimal primer concentrations for the amelogenin assay were 1 pM for both
AMEL-X and AMEL-Y and 2 pM for the BR primer. The optimal annealing/extension
temperature for was 5 8.5°C. Using the Fast SYBR® Green Master Mix (Roche), the
AMEL—X dissociation curve demonstrated a peak between 735°C and 745°C, while
AMEL-Y had a peak between 73°C and 75°C. Both amelogenin products yielded
dissociation peaks between 745°C and 7 5.5°C using the Power SYBR® Green Master
Mix (Applied Biosystems).

AMEL-X and AMEL-Y signals crossed the threshold using similar Ct values for
male control DNA (Figure 11A) and the dissociation curves demonstrated peaks at the
correct temperatures (Figure 11B). Conversely, only the AMEL-X signal crossed the
threshold using female DNA (Figure 12A) and the respective peak on the dissociation

curve was present (Figure 12B).

35

Amplification [El
Curve

E

Dissociation
Curve

 

Figure 11: Amelogenin Amplification and Dissociation Curve for Male DNA

The AMEL-X and AMEL-Y target sequences from male control DNA were amplified.
(A) The AMEL-X and AMEL-Y amplification plots (x-axis = cycle number, y-axis =
RFU value, AMEL—X and AMEL-Y Ct value = 33 cycles). The bold line represents the
threshold value of 10 RF Us. (B) Dissociation curves for the AMEL-X and AMEL-Y
products (x-axis = temperature (°C), y—axis = -d(R_FU)/dT, AMEL-X and AMEL-Y melt
temperature = 755°C).

36

 

 

 

 

 

Amplification E]
Curve
l

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

_
E AMEL-)(J Ema:
_—, ‘F
c E \A‘ ’l
09 v :1 I!
to <0 E
.. a , , p
0 3 , v. 7 -
° C ( IAMEL-YI
m 0 " , 1"; ‘V
.9 I z ,. _, -.—- ,. ,1
5.- 0w- _ _ , - .
Q Lil—W 43—. - . -‘_.. ah i r P, A q _‘__A A p A 1*: 4—A J—L #“ J L

 

 

 

 

 

 

 

 

 

 

 

Figure 12: Amelogenin Amplification and Dissociation Curve for Female DNA

The AMEL-X target sequence from female control DNA was amplified. (A) The AMEL-
X amplification plot (x-axis = cycle number, y-axis = RFU value, AMEL-X Ct value =
31 cycles). The bold line represents the threshold value of 10 RFUs. (B) The AMEL-X
dissociation curve for female DNA, using no peak for the AMEL-Y product (x-axis =
temperature (°C), y-axis = -d(RFU)/dT, AMEL-X melt temperature = 755°C).

DYZ1/Alu Assay

Optimal primer concentrations for DYZ1 were 500 nM for DYZ1-F and 900 nM
for DYZ1-R, using a probe concentration of 250 nM. Optimal Alu primer concentrations
were 500 nM for Alu-F and 900 nM for Alu-R, using a probe concentration of 250 nM.
Optimized cycling parameters included a 95°C hold for 10 min, then 50 cycles of a 95°C
denaturing step for 15 sec and 60°C annealing/extension step for 1 min. When the assay
was run using control DNA, male DNA displayed an amplification curve crossing the
threshold for both DYZl and Alu at similar Ct values (Figure 13A), while only the Alu

signal crossed the threshold using female DNA (Figure 13B).
37

 

 

Male A'" / DYZ1

loopg "1:97 A

 

 

. .-
- -_ ~
in

 

 

 

 

 

 

 

 

 

 

 

 

 

_.- q. M
.1-
'1
.
L
s

 

 

Female ,7

 

1 00pg ‘ if

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 13: DYZ1/Alu Assay Amplification of Male and Female DNA

Amplification of 100 pg of male and female control DNA (x-axis = cycle number, y-axis
= RF U value). The bold line represents the threshold value of 10 RFUs. (A) Male DNA
displays a curve for both DYZ] and Alu crossing the threshold at similar Ct values of 27
and 26 cycles, respectively, (B) while only Alu curve crossed the threshold at 24 cycles
using female DNA.

DYZ1/Alu Assay UV Treatment

Negative controls produced a DYZ1 product during assay development, therefore
the supermix and primers were UV irradiated. Probes were not UV irradiated so as not to
cause interference using the fluorescent dyes. UV treatments of 30 seconds and 1 minute
yielded DYZl product for the positive male control and no product for the negative

control. Treatments of 1.5 minutes and higher produced no DYZl product for either the

38

positive or negative control. Alu product was persistent using both negative and positive
controls after all UV treatments, even at the longest irradiation of three minutes. Figure
14 represents a 4% agarose gel separating DYZ1 and Alu products for reactions UV

irradiated for 30 seconds and non-UV treated reactions.

UV Treatment (2.73 J/cmz) Non-UV Treatment

M M
1:15x 1:15x

 

Figure 14: DYZl/Alu Assay UV Treatment Agarose Gel

Four percent agarose gel showing amplification products of the DYZ1/Alu assay using
UV treatment of 30 seconds (2.73 J/cmz) (lanes 2 - 8) and non-UV treatment (lanes 9 —
12). (std = 100 bp size standard, neg = negative control, F = female DNA, M = male
DNA, 1x = 100 pg HMW DNA, 1:15x = 15-fold dilution of 100 pg HMW DNA). The
male control UV treatment was run in duplicate. The upper band represents the DYZ1
product and the lower band represents the Alu product. The UV treated reactions yielded
a DYZ1 product using male DNA and no product using the negative and female DNA.
The non-UV treated reactions yielded DYZl product from the negative and male DNA
and no product for female DNA. Alu product was present in all reactions.

HMW and Digested DNA Dilution Series

The dilution series of HMW and digested male DNA demonstrated the sensitivity
of the amelogenin and DYZ1/Alu assays. Table 5 shows that the amelogenin assay
generated a positive result in at least one replicate containing as little as 88.9 pg of HMW
DNA, while the DYZl/Alu assay gave a positive result using as little as 0.0623 pg of

HMW DNA. The dilution series of DNase digested male DNA demonstrated that the

39

amelogenin assay generated a positive result in at least one replicate containing as little as
208.9 pg of digested DNA, while the DYZl/Alu assay gave a positive result using as

little as 0.00413 pg of digested DNA.

Table 5: Amplification Success of the High Molecular Weight DNA and DNase
Digested Male DNA Dilution Series

Diluted DNAs that produced a correct result for each DNA concentration are shown for

the amelogenin and DYZ1/Alu assays. The amelogenin assay detected both the AMEL—X

and AMEL-Y loci in HMW DNA quantities as low as 88.89 pg, and 208.9 pg of digested

DNA. HMW DNA was detected for both the Alu and DYZ1 loci down to 0.0623 pg,

while digested DNA was detected down to 0.00413 pg.

 

 

 

 

 

 

 

 

 

 

20000 pg 2 2 l 1
1330 pg 2 2 2 2
88.89 pg 2 1 2 1
5.93 pg 0 O 2 2
0.393 pg 0 0 2 2
0.0623 pg 0 0 2 1
0.00176 pg 0 0 2 O
Digested
DNA AMEL-X AMEL-Y Alu DYZ1
47000 pg 2 2 2 2
3130 pg 2 2 2 2
208.9 pg 2 2 2 2
13.93 pg 0 2 2 2
0.928 pg 0 0 2 1
0.0619 pg 0 0 2 1
0.00413Jg O 0 2 l

 

The AMEL-Y assay of 88.89 pg HMW DNA generated positive results for only
one of two replicates, while 88.89 pg and 0.0623 pg of HMW DNA and 0.928 pg, 0.0619
pg, and 0.00413 pg of digested DNA produced positive results for the DYZ1 locus in one
of two replicates. In addition, 20000 pg of HMW DNA produced a positive result in one

out of two replicates of the DYZ1/Ala assay. Figure 15 demonstrates the amplification

40

plot of both replicates for 20000 pg HMW DNA using the DYZ1/Alu assay. Replicate
one produced a curve for the Alu product that did not cross the threshold and no curve for
DYZl. Replicate two produced a non-sigmoidal curve and a lower Ct value for the DYZl

product than the Alu amplification curve (Figure 15).

 

 

 

 

 

 

 

 

 

 

. g » , 2 i ‘ AIU
One l , i- ; .
IMde ; 1 2 Q , _..;f’/T- ,
20000pg : ; , : 2 i l
Replicate . DYZI ; : ~ . 7 V ‘ '
-\. ,3... .
Male , _ I Y
20000pg if g l i ,

 

 

Figure 15: Amplification Plots of 20000 pg HMW DNA using the DYZ1/Ala Assay
The plot for replicate one displays a curve for the Alu product which does not cross the
threshold, and no curve for DYZ1. The plot of replicate two displays a non-sigmoidal
curve with a Ct value of 7 cycles for DYZland an Alu amplification curve that crossed
threshold at a higher Ct value of 16 cycles (x-axis = cycle number, y-axis = RFU value).
The bold line represents the threshold value of 10 RFUs.

Sexing of Hair Shaft DNA

Sex determination for 30 hairs using the amelogenin assay and the DYZ1/Alu

assay, and the known sex for each hair donor, are displayed in Table 6.

41

Table 6: Sexing Results of Hair Shaft DNA

Sex determined using the amelogenin and DYZ1/Alu assays are compared to the known
sex of each hair donor. (*) represent sexing results that do not match the known sex. An
inconclusive result was given to hairs when a sex result could not be determined. The
organic DNA extractions for each hair were sexed and hairs with incorrect sexing results
were reanalyzed using the alkaline digestion, denoted by (I).

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Hair Donor Amelogenin DYZl/Alu Known
Identifier Assay Assay Sex
8 female female female
9 female female female
1 0 inconclusive female female
1 1 male male male
1 2 inconclusive female female
1 3 female female female
14 female female female
16 male* female female
17 female* male male
1 9 female male*I female
20 female female female
22 inconclusive male male
23 inconclusive female female
24 inconclusive male male
27 inconclusive male male
28 inconclusive male*I female
29 inconclusive male male
32 female female female
3 5 inconclusive femaler female
36 female female female
37 female female female
3 8 male* female female
43 female female female
46 male male male
54 female female female
55 female* male male
56 female femaleI female
57 female female female
59 inconclusive female female
60 female female female

 

 

Amelogenin Assay

The amelogenin assay correctly sexed 16 hairs (53%), 14 females and 2 males.
Additionally, ten hairs (33%) produced inconclusive results, as exemplified in Figure 16.
The AMEL-X and AMEL-Y amplification curves crossed the threshold, but the peaks on

the dissociation curve were lower than the correct temperature for sex determination.

Hafi10
AMELY

Amplification IE]
Curve

E

Dissociation
Curve

 

Figure 16: Inconclusive Sexing Result of Hair using the Amelogenin Assay

A female hair that yielded no sexing result. (A) The AMEL-X and AMEL-Y
amplification plots (x-axis = cycle number, y-axis = RFU value). Both amplification
curves crossed the threshold of 10 RFUs (bold line). The AMEL-X Ct value was 36
cycles and the AMEL-Y Ct value was 34 cycles. (B) The AMEL-X and AMEL-Y peaks
on the dissociation curve (x-axis = temperature (°C), y-axis = -d(RFU)/dT, AMEL-X
melt temperature = 72°C, AMEL-Y melt temperature = 69°C). The peaks for both
products lower than the correct temperatures for sex determination.

43

Hairs 16 and 38 were incorrectly sexed as male using the amelogenin assay. An
amplification and dissociation curves for one replicate from both are illustrated in Figures
17 and 18, respectively. For each, the AMEL-X and AMEL-Y amplification curves

crossed the threshold and displayed a dissociation curve at the correct temperatures.

Hah16

Amplification LEI
Curve

 

E

Dissociation

 

Figure 17: Female Hair 16 Sexing Result using the Amelogenin Assay

Hair 16 typed as male when the known sex was female. (A) The AMEL-X and AMEL-Y
amplification plots (x-axis = cycle number, y-axis = RFU value). Both amplification
curves crossed the threshold of 10 RFUs (bold line). The AMEL-X Ct value was 37
cycles and the AMEL-Y Ct value was 35 cycles. (B) The AMEL-X and AMEL-Y peaks
on the dissociation curve were at the correct temperatures (x-axis = temperature (°C), y-
axis = -d(RFU)/dT, AMEL-X melt temperature =7 3.5°C, AMEL—Y melt temperature =
73 °C. ).

..._ '4-

Hair38 _

whom-wr- r._._- 0
1
t

l
miAMEbX

    

I5!

Amplification '
Curve

 

_, ._._,-___..,.._ .2.__‘....‘_.... ..-.v.._-_._.._v . aw—-._..--

E

Dissociation
Curve

 

Figure 18: Female Hair 38 Sexing Result using the Amelogenin Assay

Hair 38 typed as male when the known sex was female. (A) The AMEL-X and AMEL-Y
amplification plots (x-axis = cycle number, y-axis = RFU value). Both amplification
curves crossed the threshold of 10 RFUs (bold line). The AMEL-X Ct value was 30
cycles and the AMEL-Y Ct value was 32 cycles. (B) The AMEL-X and AMEL—Y peaks
on the dissociation curve were at the correct temperatures (x-axis = temperature (°C), y-
axis = -d(RFU)/dT, AMEL-X and AMEL-Y melt temperature = 735°C).

Hairs 17 and 55 incorrectly sexed as female using the amelogenin assay.
Amplification and dissociation curves for one replicate from both are illustrated in
Figures 19 and 20, respectively. For hair 17, only the AMEL-X product yielded an
amplification curve that crossed the threshold and a dissociation curve at the correct
temperature. No curve was present for the AMEL-Y product (Figure 19). Hair 55
generated an amplification curve for AMEL-X and a non-sigmoidal curve for AMEL-Y.
The dissociation curve of the AMEL-X product produced a peak at the correct

temperature, but the peak for the AMEL—Y dissociation curve was too low (Figure 20).

45

Haul?

 

 

[Fl

 

 

 

 

 

Amplification —‘
Curve
>
E
'T
. _......-.' L f L .

 

 

 

 

 

 

 

 

E

Dissociation

 

Figure 19: Male Hair 17 Sexing Result using the Amelogenin Assay

Hair 17 typed as female when the known sex was male. (A) The AMEL-X amplification
curve (x-axis = cycle number, y-axis = RFU value). Only AMEL-X produced an
amplification curve that crossed the threshold (bold line) at a Ct value of 41 cycles. (B)
The AMEL-X peak on the dissociation curve was at the correct temperature (x-axis =
temperature (°C), y—axis = -d(RFU)/dT, AMEL-X melt temperature = 735°C). No signal
was produced for AMEL-Y.

46

Hair 55

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

—A" "Wit“ll I
I_r: AMEL-X
E 023 .y i X [AMEL-YJ
its 1 \r /
e \- w
< “r y . ,x"
I...- .._ -...-. 1 ._ : a...“ . ‘A. _.._. .. .. .v ._ III—'- _.a .l
.L_.lc
.9
a 0
.8 g
E 0
D

 

Figure 20: Male Hair 55 Sexing Result using the Amelogenin Assay

Hair 55 typed as female when the known sex was male. (A) The AMEL-X and AMEL-Y
amplification curve (x-axis = cycle number, y-axis = RFU value). AMEL—X and AMEL-
Y produced an amplification curve that crossed the threshold of 10 RFUs (bold line). The
AMEL-X Ct value was 32 cycles and the AMEL-Y Ct value was 46 cycles. Note the
non-sigmoidal curve and the high Ct value of the AMEL-Y signal. (B) The AMEL-X
peak on the dissociation curve was at the correct temperature, while the AMEL-Y peak
was at a temperature too low for sex determination (x-axis = temperature (°C), y-axis = -
d(RFU)/dT, AMEL-X melt temperature = 735°C, AMEL-Y melt temperature = 70°C).

DYZ1/Alu Assay

Twenty-eight hairs (93%), 20 female and 8 male, matched the known sex when
analyzed using the DYZ1/Alu assay. No hairs were incorrectly typed female or were
considered inconclusive. An incorrect sexing result was made for hair 35 and 56 using
organically extracted DNA. When the DNA from the alkaline digestion was analyzed,
both hair 35 and 56 yielded the correct female sexing result. An incorrect male sex result

was obtained from hairs 19 and 28 using DNA from both the organic extraction and the

47

alkaline digestion techniques. Hair 19 produced a signal for both Alu and DYZ1 that
crossed the threshold; however the Ct values were 15 cycles apart for both the organic
extraction and alkaline digestion (Figure 21). DNA organically extracted from Hair 28
produced a signal that crossed the threshold for Alu and non-sigrnoidal curve for DYZ1
(Figure 22). The alkaline digestion technique yielded DNA that produced a signal for
both Alu and DYZl that crossed the threshold; however the Ct values were 15 cycles

apart.

Han19

 

 

 

 

 

 

 

 

Organic
Extraction
/
IF
'. \ \X
\
\‘1

l :

 

 

 

 

 

 

 

 

 

.,
f
\

J u..-“

 

 

 

F
l

l

l

1

 

 

 

 

..............

Alkaline
Digestion

 

 

 

 

 

 

 

Figure 21: Female Hair 19 Sexing Result using the DYZ1/Alu Assay

The DYZ] and Alu products were present in the DNA from the (A) organic extraction
and (B) alkaline digestion of hair 19. (x-axis = cycle number, y-axis = RFU values). The
bold line represents the threshold value of 10 RFUs. The DYZ1 Ct value for the organic
extraction was 42 cycles, while the Alu Ct value was 31 cycles. For the alkaline
digestion, the Alu Ct value was 31 cycles and the DYZ1 Ct value was 43 cycles. Note the
large separation of Ct values for Alu and DYZl for both isolation techniques.

48

 

Hafi28

 

 

 

 

 

 

[El 4:1
Alu 1 1 wet”
c: T l "T

O O i. \H ,.
E "3 t =3 /'f
gm 1 . , ,ay loafl
L0 g ‘2‘ ”‘1; I———-l
O m ...__._.. “it. __ ....-. __ .. lax—’- ”1.-

\m l l

l i

 

 

 

 

 

 

 

 

 

 

 

 

 

ll
i
iv
3
N_.
A
\

Alkaline
Digestion
.l
I "\

 

 

 

 

 

 

 

 

 

 

 

Figure 22: Female Hair 28 Sexing Result using the DYZ1/Alu Assay

The DYZl and Alu products were present in the DNA from the (A) organic extraction
and (B) alkaline digestion of hair 28. (x-axis = cycle number, y-axis = RFU values). The
bold line represents the threshold value of 10 RFUs. The DYZ] and Alu Ct value for the
organic extraction was 28 cycles. For the alkaline digestion, the DYZ1 Ct value was 46
cycles and the Alu Ct value was 29 cycles. Note the non-sigmoidal curve for the DYZ1
signal of the organic extraction and the separation of Ct values for Alu and DYZ] of the
alkaline digestion.

Ancient Skeletal Sexing Results
Table 7 displays the sexing results and anthropological sex estimations for the
Butrint and Kamenica, Albania adult skeletal remains. Two or three bones were analyzed

per individual.

49

Table 7: Molecular Sexing Results and Anthropological Sex for Ancient Adult

Bones

Sexing results using the amelogenin and DYZ1/Alu assays were compared to the

anthropological estimated sex for the Butrint, Albania (BA) and Kamenica, Albania (KA)

adult skeletal samples. An asterisk (*) represents results that do not match the

anthropologically estimated sex. An inconclusive result was given to bones when a sex

determination could not be made.

 

 

 

 

 

BA Adults
1187.1 right radius midshaft inconclusive male male
1 187.2 ri t fibula midshaft inconclusive male male
1226.1 right fibula inconclusive male male
1226.2 left ulna midshaft female* male male
1236.1 fibula midshaft inconclusive female female
1236.2 left ulna midshaft inconclusive female female
1 5 l 8. 1 humeral cortical frag inconclusive female* male
1 5 1 8.2 fibula midshaft female* male male
5010.1 fibula midshaft inconclusive female female
5010.2 humerus midshaft inconclusive female female

 

 

3023.1 etrous ortion inconclusive male male
KA Adults

 

 

 

 

 

 

 

 

 

 

 

 

 

66 humerus inconclusive male male
66 femur inconclusive male male
75 tibia inconclusive male male
75 femur inconclusive male male
75 petrous portion male male male
82 radius inconclusive male male
82 petrous portion male male male
85 femur inconclusive male male
97 tibia male female* male
97 femur female* male male
1 02 humerus female female female
102 petrous portion inconclusive female female
1 04 humerus inconclusive male male
1 04 femur inconclusive male male
142 pelvis inconclusive female* male
142 petrous portion inconclusive male male
142 humerus female* male male
156 femur female* female* male
156 petrous portion inconclusive female* male
156 humerus female* male male

 

50

 

Adult Butrint Skeletal Remains

Six individuals were analyzed from the Butrint adult skeletal remains (Table 7).
Two bones were analyzed per individual, except 3023 where only DNA from the petrous
portion was available. The amelogenin assay produced no results for nine of the bones,
while individuals 1226 and 1518 demonstrated bones that sexed as female, the 1226 left
ulna midshaft and the 1518 fibula midshaft, while the anthropologically estimated sex
was male. The amplification and dissociation curves for individuals 1226 and 1518 are
illustrated in Figures 23 and 24, respectively. Only an amplification curve for AMEL-Y
crossed the threshold (Figure 23A), while no dissociation curve was present for either
product for the 1226 right fibula (Figure 23B). The left ulna midshaft produced an
amplification curve that crossed the threshold for both AMEL-X and AMEL-Y (Figure
23C), but only the peak for the AMEL-X dissociation curve was at the correct
temperature (Figure 23D). The AMEL-Y dissociation curve was approximately ten
degrees higher than the correct temperature. The 1518 humeral cortical fragment
amelogenin sexing results demonstrated an amplification curve that barely crossed the
threshold for both the AMEL—X and AMEL-Y products (Figure 24A), but neither product
displayed a dissociation curve (Figure 24B). Only the AMEL-X product produced a
signal that crossed the threshold and a correct melting temperature for bone 1518 fibula

midshaft (Figure 24 C and D).

51

Figure 23: Sexing Results for Burial 1226 using the Amelogenin Assay

The anthropological sex for burial 1226 was male. The right fibula produced no sexing
result and the left ulna midshaft typed as female. (A) The amplification plot for 1226
right fibula (x-axis = cycle number, y-axis = RFU value, AMEL-Y Ct value = 47 cycles).
The bold line represents the threshold value of 10 RFUs. (B) The dissociation curve for
1226 right fibula. Neither product produced a dissociation curve (x-axis = temperature
(°C), y-axis = -d(RFU)/dT). (C) The amplification plot for 1226 left ulna midshaft (x-axis
= cycle number, y-axis = RFU value, AMEL-X Ct value = 41 cycles, AMEL-Y Ct value
= 37 cycles). The bold line represents the threshold value of 10 RFUs. (D) The
dissociation curve for 1226 left ulna midshaft (x-axis = temperature (°C), y-axis = -
d(RFU)/dT, AMEL-X melt temperature = 735°C, AMEL-Y melt temperature = 855°C).
Note the dissociation curve for AMEL-Y 10 degrees higher than the correct temperature.

52

Bone 1226.1
right fibula

 

 

 

Curve

 

Amplification El

 

 

 

 

 

 

 

E

Dlssociation
Curve

__....-._

Bone 1226.2
left ulna m'dShafl

_....-

 

 

 

 

 

 

 

 

 

Curve

:\ "

/

 

 

Amplification E

 

 

 

 

 

 

 

 

 

Curve

DissociationEI

53

 

Figure 24: Sexing Results for Burial 1518 using the Amelogenin Assay

The anthropological sex for burial 1518 was male. The humeral cortical fragment
produced no sexing results and the fibula midshaft typed as female. (A) The amplification
plot for 1518 humeral cortical fiagment (x-axis = cycle number, y-axis = RF U value).
Neither the AMEL-X nor AMEL-Y produced an amplification signal that crossed the
threshold of 10 RFUs (bold line). (B) The dissociation curve for 1518 humeral cortical
fragment (x-axis = temperature (°C), y-axis = -d(RFU)/dT). (C) The amplification plot
for 1518 fibula midshaft (x-axis = cycle number, y-axis = RFU value, AMEL-X Ct value
= 39 cycles). The bold line represents the threshold value of 10 RFUs. (D) The
dissociation curve for 1518.2 (x-axis = temperature (°C), y-axis = -d(RFU)/dT, AMEL-X
melt temperature = 735°C).

54

Bone 1518.1

humeral cotical fragment

 

 

 

 

 

I_ ‘ Y ‘
A . :11: l
._ ‘: 1 I
3 i 3 ;

1--.;1. ._‘ . 1 . . .1 ._ _. 7‘ " ' --;'; . .L‘L;..'.:.- .-. it
i ' AMEL-

Curve

..,1-.

 

Amplification ‘—

 

 

 

 

 

 

 

 

 

 

 

Dissociation

 

 

 

Curve

 

Bone 1 518.2
fibula midshaft

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

l . ‘I 1‘ I ' r I " I
IE] I: i i ,1 lAMEL-Xl l
g l. 7 . I
13 o . i l 3 i i . - .
_ E i E i .
{:5 3 fl " 1
£510 5 _
l l
. l ;:
< ‘ 5
i 3
a" .
Vl- if 4 ‘ E‘ 5 I ‘ I + * a
- I
in . ~
c: t
.9 I '3
"" t
.9 “3’ E ;
8 a C
m .
.9 ’
D r
- -1 ;_ ,
..r. 4... ... 0: c... L. .. J» J. -._._.... 1.-

55

 

 

When analyzed using the DYZ1/Alu assay, all six individuals yielded a sexing
result consistent with the anthropological sex, encompassing two females and four males,
and no inconclusive results. Bones from each individual demonstrated concordant results,
except burial 1518, from which the humeral cortical fragment sexed as female and the
fibula midshaft sexed as male. The amplification plots for burial 1518 are illustrated in
Figure 25. The Alu product produced a signal that crossed the threshold for bone 1518
humeral cortical fragment, while the DYZ] curve was well below the threshold (Figure
25A). The fibula midshaft of burial 1518 demonstrated an amplification curve that
crossed the threshold for both the Alu and DYZl products (Figure 25B); however, the

DYZ1 Ct value was approximately eight cycles higher than that of Alu.

56

 

Burial1518
[E “W“ 7

Bone
1518.1

humeral ;
cortical ;_ 1 .
fragment-.. _- __-._.__. T 1 . \ ,,____1

 

 

 

i
l
l
I
1
I

 

 

,l
I
l
3

 

 

 

 

 

 

 

 

L-—-_J

Bone Ns..l name“ 1
1518.2 g" “\ -

fibula . 1
midshaft DYZ1

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 25: Sexing Results for Burial 1518 using the DYZl/Alu Assay

The anthropological sex for burial 1518 was male. The humeral cortical fragment typed
as female and the fibula midshaft typed as male (x-axis = cycle number, y-axis = RFU
values). The bold line represents the threshold value of 10 RFUs. (A) The plot for 1518
humeral cortical fragment demonstrates a signal crossing the threshold for Alu at 39
cycles and a non-sigrnoidal curve that does not cross the threshold for DYZl. (B) The
plot for 1518 fibula midshaft demonstrates a curve for both DYZ] and Alu that crosses
the threshold. Note the Ct values differ by approximately eight cycles with the DYZ1 Ct
value at 40 cycles and the Alu Ct value at 33 cycles.

Adult Kamenica Skeletal Remains

Nine individuals were analyzed from Kamenica, Albania, using two or three
bones sexed from each, except individual 85 where only DNA from a femur was
available (Table 7). Twelve out of twenty bones produced no sexing results and of the
remaining eight, four typed consistently with the anthropological sex using the
amelogenin assay. Four bones did not yield sexing results consistent with the

anthropological sex. Individuals 75, 82, and 102 yielded one bone (75 and 82 petrous

57

portion, and 102 humerus) consistent with the anthropological sex, while the remaining
bone produced no sexing result. The 75 and 82 petrous portions produced amplification
signals for both the AMEL-X and AMEL-Y products that crossed the threshold and
dissociation curves at the correct temperatures, thereby typing male (data not shown).
The 102 humerus only produced an amplification signal that crossed the threshold and a
dissociation curve at the correct temperature for the AMEL-X product, thereby typing
female (data not shown). The petrous portion of burials 142 and 156, and 156 femur,
sexed as female although the estimated anthropological sex was male. Figure 26A and B
shows the result for the petrous portion of burial 142, in which only the AMEL—X product
yielded an amplification curve that crossed the threshold and had a dissociation curve at
the correct temperature. No signal was detected for the AMEL-Y product. The additional
bones sexed from burial 142, the pelvis and humerus, gave no sex result. The pelvis
demonstrated no amplification or dissociation curve for either the AMEL—X or AMEL-Y
products (Figure 26 C and D). The humerus yielded a signal for AMEL-X and AMEL-Y
that crossed the threshold on the amplification plot (Figure 26E), but demonstrated no

dissociation curves (Figure 26F).

58

Figure 26: Sexing Results for Burial 142 using the Amelogenin Assay

The anthrOpological sex for burial 142 was male. The petrous portion typed as female
and the pelvis and humerus produced no sexing results. (A) The amplification plot for
142 petrous portion (x-axis = cycle number, y-axis = RFU value, AMEL-X Ct value = 37
cycles). (B) The dissociation curve for 142 petrous portion (x-axis = temperature (°C), y-
axis = -d(RFU)/dT, AMEL-X melt temperature = 755°C). (C) The amplification plot for
142 pelvis (x-axis = cycle number, y-axis = RFU value). The bold line represents the
threshold value of 10 RF Us. Note that no signal was produced for either AMEL-X or
AMEL-Y. (D) The dissociation curve for 142 pelvis (x-axis = temperature (°C), y-axis =
-d(RFU)/dT). (E) The amplification plot for 142 humerus (x-axis = cycle number, y-axis
= RFU value, AMEL—X Ct value = 7 cycles, AMEL-Y Ct value = 12 cycles). The bold
line represents the threshold value of 10 RFUs. Note that neither AMEL—X nor AMEL-Y
produced sigmoidal curves (F) The dissociation curve for 142 humerus (x-axis =
temperature (°C), y-axis = -d(RFU)/dT).

59

Bone 1423p

 

 

 

 

r i. ll 9%. .llsllale

- . .

..
n w w.

. 2.

.M .. u
.
.fi. X ...- llll.+ 2...! 2.... .. ....-.. .v

w . __

L ..

M E

O _. _.

ll _ a
.n n
0 will-ll... ... f. t h a» ..ii 1.. o
p _. .M.
S ,_ w
r
t .
m

 

Bone 142 P

 

 

 

 

k
a!
a.
o . a ..l. 2 n a ll. 0
H .
. _
Va I“. -lv| l..-
—
.. h. .
— .
p .
— — ‘
. .
Fl... 2.“ la!

 

 

m>50
cozmosacz

26.835

S

pe

 

 

 

 

 

 

 

~33... .. llnl .

828535...

 

_
Ill .5. ll villi-81.111!!-

02:0

 

ll: .. 1 1.1-11221-2!!- l;
M
r
. _.
n .
m w
. a
M ..
. fl
. .
>|l L r
11
. H
.
m .
. m
. .
_
.. _
r
_.
-Il-clliil 1....O‘llll.l\ll..illl. . 2b : I. .2 1:0
. _
. H.
.
.
u
. ~
. M
-. .. ..-..-.-.... l 2 . ..l .1. 1+.
. w
.
. k
. . __
H . . ._
. . V .
U .. ..
l l .. O 4.. all P .4- .. -.. fl .
..
.
._ .
. ..
_ m u.
. . x
. u \
r-lF 19:-.9 .- I i.. 1...... l. . r. A.
H . _
. .a .
_ .
. _
a . J
y? (.5...9.¢.05.I:l :0- !.l- ‘1
. . ,
. , ,.
u _
_ .
H _. n.
.. .
_
.
q . .4
. w
. ..
. ._
. _
_ m
. m
_ 4
a _.
. .
.. _
.. M
u u
. u m
. .
. . M
..O. I It o...._lolb-t- t

-......_..-. — .1 2 ...,

 

 

 

_. .
.
..
.
.
h
.

..

1...-..

~

a

a ..

_ .

_ .

. w

.

. .

.-- - _._..-._..-_..-_.T-_..-.-.-

. . .9 a . :1 a. 2 5 2 .4
J . .. . ..
a _

. . _ . . .
_ m . ..
. .. .e A . . .. w
. W .

.. m
n . ..

_ _

m ..

. u.

. ._

_ m

_ . .

. ..
.- . u
« lll4. .

’ . .
a _ ..
_ . M m
_ _ L
_ . I
. ..
M 2
.
l v %- 4w 1.4-. Y-le v .m
_ Z
w m m n
. . M .
. u —
a .
_ .
.m ..
m ..
m ‘
r e- 1; . .5. u. ,v. m
_ .
. .
_ a
1
r.
. ..
. . n.
v I

 

¢
1

0
l

o

. .._._..___. ._._....__..,.- .. '.. ..

-._.__- _'. _ -_.__..~ __.-...+

 

 

 

...T o i...
.. H H r
.. _. s
_ . .
.. . . u
I
m H“
n . h
- — .
~
m . _ .. .1.
will-01. 27.1....- l-..+-ll¢. . -.t+l.-.|Vll.dl .112 - .
_ w m , U
u. . _ _ w .
— ,. . . n
... . W I
N

...v..._.._ _ _ _._ _~_._ __ _

-"

 

m
_

 

__,___,.,._‘____ -.-....-....-..-. .-_-._._--
-._..
\ - . ..
_ .2 . h
. , .. .
...- ..-....... ,.__ -...,

   

9:30
5:98me

60

Figure 26 (cont’d)

Bone 142 H

humerus

 

 

“11'
l

l
--—--l

Amplification ‘
Curve
\-
l
13

 

 

w Mu- M” a
{—2 I
> 3
l t
, .
g -
P -2: i.
U
. gran-nun... 0" ”Hv’ .4 . u .,-_. n... - 4
l

 

 

 

w
l
. l

l

l

 

 

 

1r: .~., . -_ -
q
1 .

 

 

El

Dissociation

 

 

 

Sexing results for the all bones of burial 156 are presented in Figure 27. The femur
demonstrated an AMEL-X amplification curve barely crossing the threshold and a small
dissociation curve at the correct melt temperature. The AMEL—Y product yielded a non-
sigrnoidal curve that produce a very low Ct value and no dissociation curve (Figure 27 A
and B). The sexing result for the petrous portion yielded only an amplification curve that
crossed the threshold and dissociation curve at the correct temperature for AMEL—X
(Figure 27 C and D). The additional bone for burial 156, the humerus produced no
amplification or dissociation curve for either the AMEL-X or AMEL—Y product (Figure

27 E and F) and no sex result could be determined.

61

Figure 27 : Sexing Results for Burial 156 using the Amelogenin Assay

The anthropological sex for burial 142 was male. The femur and petrous portion typed as
female and the humerus produced no sexing result. (A) The amplification plot for 156
femur (x-axis = cycle number, y-axis = RF U value, AMEL-X Ct value = 10 cycles,
AMEL-Y Ct value = 48 cycles). The AMEL-Y produced a non-sigmoidal curve and the
AMEL-X produced a signal barely crossing the threshold (bold line). (B) The
dissociation curve for 156 femur (x-axis = temperature (°C), y-axis = -d(RFU)/dT,
AMEL-X melt temperature = 755°C). Note the small dissociation curve for AMEL—X at
the correct temperature. (C) The amplification plot for 156 petrous portion (x-axis = cycle
number, y-axis = RF U value, AMEL-X Ct value = 36 cycles). The bold line represents
the threshold value of 10 RFUs. No signal was produced for the AMEL-Y product. (D)
The dissociation curve for 156 petrous portion (x-axis = temperature (°C), y-axis = -
d(RFU)/dT, AMEL—X melt temperature = 755°C). (E) The amplification plot for 156
humerus (x-axis = cycle number, y-axis = RF U value). Note that neither AMEL—X nor
AMEL-Y produced sigmoidal amplification curves. The bold line represents the
threshold value of 10 RFUs. (F) The dissociation curve for 156 humerus (x-axis =
temperature (°C), y-axis = —d(RFU)/dT).

62

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Bone 156 F

femur

 

 

 

 

 

. i I‘ I- lull-ul- ‘ u‘IA‘ ile-‘J ...‘IW: A. ‘ 3‘2! A‘] 41.4 iJiJ l.1_.'¢I-l III A J '1.
. . . .
_ m M H . . m w .
17 H .. N
. . H a... --- t - ..--ii...“ .... . .
_ .. .. M A . . w a . ._
u .. _ .. u h _ ‘ . .
_ u _ . h u. M
.\ ...... ...-1... A . ... . f ..
. __ n _ . . . H m M
E . . .u a _ .
M . _ . m H a u
_A . .. . . , w . .... m M m
I . . o .......... .2- TI +1....-o...iii- . -11: l... ._ . ...- .1 -e - iv. lH. .-if - . r. .
. .. . , u
.. .. ... _ .. . a. ~ ..
.e . _ . . h . . ... w . V.l
.. : .. .l-. i. 2.. . : .1... . . ..r . . O m m .i . ... .. . .2 w ..
. . .. n .
. .. . .v
_ . . W .. . . E
. w . _ _
.Q- 0 w h ...i .r..§2 ...» M .1 {AT
__ .. . . 4 E _ .... .. A
.- r i M . .. m
tl fl 4 . M l ‘ It: E ill 1 till} .2). .. 5 - v ...l... I .«d .. w
. a . u. ~
m H. n I . p n m ... _. m _ ..
.. I .. ... ..
.. .... _ _ S o m m m . u .
r . .. M 0' . . w _ . _ .
~ w A. a I? + ». Emir! 4r .. .... 6 fl _ . ._ D it s . ....» o .
.- _. _.. 5cm.--.-..-- ......) _.
.. u n . a w. . . _ H
H m ,_ u ,. 1. m w H. . . m 1 M .
m . .. . .. u n
. .. . m . M S . _ . . m N .
m . W .. M e U W X . w . w
. . y . . — _ v
.. -. . -. .. .. . .- . .- .. o z . .
. _ . . _ 0 t . L . _
.. . m _ M _. v n . ~
_ . w _ _ B e u E _ a _ .
.. m " fl . _
.. . . _ .. m _ ._ W .
. _ H _ A . . . ..
i m M u .iliil.+.l!ll.l- -. lili- .. L M A m
_ i ‘ . l. i ..-.
_ .. . . m T w ... . N n. .
_ n m . L . ... . 2 . .M
. . a _ .. . _ .. ... H . _.
,_ . _ . _ _. .. U. . u
. H u ‘ . .. H.. . n
. _ .1 ._ . _ _ .. . u
. . . w .. . .. . ..
. Y 2 -_ . . - .. w . .
_ u . . _ _ H. ._ n .. . ..
n E . . m A ... M H m
.71: ... . _ ~ . a .. .... _ ~
.. . . _ v V. 0 l A .... .0 .. ~ _
M . M _ _. _ H m . n
m A 2.. . _ . .. . . ; _ n ” .... . . .
. . . . . . _ ”v y
_ . a . . .. . . . . . W .... _ .. w
_ i H . _ _. . . . . m .... .._ m
. — _ ~ . . _ _ r a _ w
h . .. .. . . .. u. . _
Free. {turvrpbc grip-“ . . . ..f: . -lilti -1.-..¢. vii-tr? . . f. I. r» it... ... :1...Lf££.s.rv .

 

 

 

 

 

mzno

 

 

 

 

mzzo

 

 

 

 

 

 

9:30

 

 

 

 

 

 

9:30

 

 

I \ ~- ¢.—-L -+-l..n.1

 

63

ll co_fioE_aE<

,co=m_oowm_o
E

E

8:8533

_m_

seafloomma

 

 

 

 

 

 

Figure 27 (cont’d)

Bone 156 H

humerus

 

 

l'l'l

 

 

 

 

Amplification
Curve

 

 

 

 

 

 

 

 

 

 

 

 

Curve

Dissociationa

 

The tibia of burial 97 typed as male, while the femur generated a female result
(Figure 28). Both the AMEL—X and AMEL-Y products produced an amplification curve
that crossed the threshold for the 97 tibia (Figure 28A), along with dissociation curves at
the correct temperature (Figure 28B). The 97 femur yielded only the AMEL—X product
that presented an amplification curve that crossed the threshold (Figure 28C) and a

dissociation curve at the correct temperature (Figure 28D).

64

Figure 28: Sexing Results for Burial 97 using the Amelogenin Assay

The anthropological sex for burial 97 was male. The tibia was typed as male and the
femur was typed as female. (A) The amplification plot for 97 tibia (x-axis = cycle
number, y-axis = RFU value, AMEL-X Ct value = 29 cycles, AMEL-Y Ct value = 39
cycles). The bold line represents the threshold value of 10 RF Us. Note the Ct values are
different by 10 cycles. (B) The dissociation curve for 97 tibia (x-axis = temperature (°C),
y-axis = -d(RFU)/dT, AMEL-X melt temperature = 75.5 °C, AMEL-Y melt temperature =
75°C). (C) The amplification plot for 97 femur (x-axis = cycle number, y-axis = RFU
value). The bold line represents the threshold value of 10 RFUs. An amplification curve
crossed the threshold for AMEL-X product at 39 cycles and no curve crossed the
threshold for the AMEL-Y. (D) The dissociation curve for 97 femur (x-axis =
temperature (°C), y-axis = -d(RFU)/dT, AMEL-X melt temperature = 74°C).

65

Bone 97 Ti
tibia

- ...“-..---.._--._....--...... l_—|
rm“. ‘ | AMEL—Xl
‘ l

 

 

 

 

 

  

 

 

Curve

 

Amplification FE]

 

 

 

 

 

 

 

 

 

'
l 7
i E
t :
l -' i
i .l l
t ... a - .....- l - r - ~ 9 «— ~ > ~L— A - - +~ J

 

Curve

Dissociationm

 

Bone 97 F

femur

 

 

 

 

[El :5 * T ,——.‘
AMEL-X

c: T : ‘5 ,
m ‘D i a
g E i .‘ i
= 3 ', . .. .. .‘i... i . .L
‘10 2 ‘ 2 2
E .-
< ,_ ...—..i. , '
. V‘-_,-_ i' l

1' \-. ... ’g g

L A 1 '

 

 

 

 

 

 

 

 

 

Dissociationg
Curve

 

66

DYZl/Alu sexing results from all nine adult Butrint individuals, one female and
eight males, were consistent with the anthropological sex estimate. The multiple bones
analyzed fiom five individuals (66, 75, 82, 102, 104) yielded consistent sexing results.
The three remaining individuals each had one or more bones (97 tibia, 142 pelvis, and
156 femur and humerus) that typed as female, while the other bones (97 femur, 142
humerus and petrous portion, and 156 petrous portion) typed as male. Figure 29 displays
the sexing result for the tibia and femur from burial 97. The tibia only yielded an
amplification curve crossing the threshold for the Alu product (Figure 29A) and the
femur yielded both the Alu and DYZl products (Figure 29B). The sexing results for
burial 142 are illustrated in Figure 30. The pelvis yielded only an Alu amplification curve
that crossed the threshold (Figure 30A), while the hmnerus and the petrous portion
demonstrated a curve for both DYZ1 and Alu products that crossed the threshold (Figure
30 B and C). In addition, the sexing results from burial 156 are shown in Figure 31. The
femur yielded only an amplification curve for the Alu product that crossed the threshold.
The DYZ1 signal was below the threshold (Figure 31A). The humerus produced an
amplification curve crossing the threshold for only the Alu product (Figure 3 18), while
both the Alu and DYZl produced a signal that crossed the threshold for the petrous

portion (Figure 31C).

67

Bufial97

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Bone \.
97 TI ,
tibia l
. .
'ffir—vt'f‘ 5"".fo "—- ““C‘r‘r 3;, . . . .. . . . . . ~. . . F ""7“.’T‘.‘T"'I"“‘1l".' -.-...
.; l l -..-.- “...-xxx 1
J l 1 ‘ ,_, _
.~ ..-»!
i -' I r”?
97 F l l r .x'
’f 7 l i /" K
femur : - \
..I
g g/; [own]
i .’ _/ l ,.
-...2.-_-.-.....-=i-__--._--_..-_-....._....--_---.....--...---.--2. __.._._...._-_....- _ .-...-“ -.-
L L: x‘i Wm" ‘ J 2:

 

 

 

 

 

 

 

 

 

 

 

Figure 29: Sexing Results for Burial 97 using the DYZ1/Alu Assay

The anthropological sex for burial 97 was male. The tibia typed as female and the femur
typed as male. The bold line represents the threshold value of 10 RFUs (x-axis = cycle
number, y—axis = RF U values). (A) The plot for 97 tibia demonstrates a signal crossing
the threshold for Alu at 23 cycles and no signal for DYZl. (B) The plot for 97 femur
demonstrates a curve crossing the threshold for both DYZ1 and Alu with a difference in
Ct values of five cycles (DYZ1 Ct value = 38 cycles, Alu Ct value = 33 cycles).

68

 

Burial 142

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

[a --.... . in
Bone l \\ .- .
142 P
pelvis "
~— l :.-...-. ~ r 7:. ...-_-;::.-f at»... - - ; 3; :::'.‘:':”..‘..1L tar: 11:. r. ~
l DYZ1 ff, we’ff
Bone l - X/ffw ”
142 H : + s 7
l I l / i
humerus . l 1 3 ,’
1‘ z 3 ..o‘r‘ Alu
l l {’I .‘l
l... .-- -.- - - 1i ._ .. - - . l . f). _ _
.“If‘\ A 3 1 15.5.!
[a Cl -~
1 DYZ1 fg,-47": If“
Bone i \\ , \
142 sps: ~ s.
l .r .
petrous f Alu
portion .-
(’ .
- if. '. .‘_. l‘ __\ l: A_
i l ‘

 

 

 

 

 

 

 

 

 

 

 

Figure 30: Sexing Results for Burial 142 using the DYZ1/Alu Assay

The anthropological sex of burial 142 was male. The pelvis typed as female and the
humerus and petrous portion typed as male (x-axis = cycle number, y-axis = RFU
values). The bold line represents the threshold value of 10 RF Us. (A) The plot for 142
pelvis demonstrates a curve for Alu that crosses the threshold at 36 cycles and no curve
for DYZ]. (B) The plot for 142 humerus demonstrates a signal crossing the threshold for
both DYZl and Alu at 34 and 36 cycles, respectively. (C) The plot for 142 petrous
portion demonstrates a signal crossing the threshold for both DYZ1 and Alu at 32 and 33

cycles respectively.

69

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Bone g \\ 1 ~'
156 F r y . .2
femur 4’ I \ /./
lilfihfi+ .fii- N \ - / I
t _: ...};1: 1 41175. ._.:.'_'_‘:;.:. 1... t .:. 1 '.:..-i'l,('t’(:';:.. 1
‘ i l '5 "\. '
'~———— ; Alui 1 ‘"
'- 1 ,
: i .- ~ .
Bone , \ ;
156 H 2 ‘ f
humerus . 3 f
.. -.. - - ..-. , - -..- -.i -... .. .. -
f -- ' My" 1
.{ ,r‘l’“ 2‘
Bone \‘ * i if, ”'x
156 sp , \\
petrous 1l g l
l 1
portion 6. . DYZ1]
l . ‘

 

 

 

 

 

 

 

 

 

 

Figure 31: Sexing Res

ults for Burial 156 using the DYZ1/Alu Assay

The anthropological sex for burial 156 was male. The femur and humerus typed as
female and the petrous portion typed as male (x-axis = cycle number, y-axis = RFU

values). The bold line represents the threshold value of 10 RFUs. (A) The plot for 156
femur demonstrates a signal crossing the threshold for Alu at 34 cycles, but the DYZl
curve did not cross the threshold. (B) The plot for 156 humerus demonstrates a curve for

Alu crossing the threshold at 29 cycles and no curve for DYZ1. (C) The plot for 156

petrous portion demonstrates a signal crossing the threshold for both DYZ] and Alu at 34
and 32 cycles, respectively.

70

Juvenile Butrint Skeletal Remains

Seven juveniles from Butrint that could not be sexed anthropologically were
sexed molecularly using the amelogenin and DYZ1/Alu assays (Table 8). One bone from
juvenile burials 1185, 1225, 1229, and 1259, and two bones from 1224 and 1264, were

analyzed.

Table 8: Sexing Results for Butrint, Albania Juvenile Skeletal Samples

Sexing results from the amelogenin and the DYZ1/Alu assays are compared for the
Butrint, Albania (BA) juvenile skeletal samples. The asterisk (*) signifies sexing results
that did not match the alternate bone for the same individual. An inconclusive result was
given to bones where a sex result could not be determined.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Bone S 'f‘ B Amelogenin DYZ1/Alu
Identifier pecr 1c one Assa Assa
BA Juveniles

1 185.1 fibula shafi inconclusive male
1188.1 lefi humerus shaft inconclusive female*
1188.2 right petrous portion inconclusive male*
1224.1 left petrous portion inconclusive female
1224.2 right humerus midshaft female female
1225.1 right petrous portion inconclusive female
1229.2 humeral shalt inconclusive male
1259.] left femur midshaft inconclusive male
1264.1 nbla g0 3:111:13“ ent inconclusive male
1264.2 fibula midshaft male male

 

 

When sexed using the amelogenin assay, one bone fi'om 1224 and 1264 yielded a
sexing result; the rest were inconclusive. Using the DYZ1/Alu assay, juveniles 1224 and
1264 typed as female, while 1185, 1229, 1259, and 1264 typed as male. Results were
congruent between both bones from juvenile 1224 and 1264. Juvenile 1188 did not
produce two bones the same sexing result, therefore an overall sex determination could

not be made (Figure 32). The left humerus shaft yielded an for the Alu product; however,

71

the DYZ1 amplification produced an amplification curve below the threshold (Figure
32A). The right petrous portion demonstrated an amplification curve for both the DYZ1

and Alu products that crossed the threshold (Figure 328).

. ... . . . .. - 5911311 18.8. - -
El LT" Alu

Bone \ i I DYZ1 I

1 188.1 1 )1
left ., ...... _ -.-, _ \ ff,»-

humerus , 4‘ . ,

shafi

 

 

 

 

 

 

 

 

 

 

 

 

 

Bone
1 188.2
n'ght
petrous
portion

 

Figure 32: Sexing Results for Burial 1188 using the DYZ1/Alu Assay

Burial 1188 was a juvenile, therefore an anthropological sex could not be determined.
The left humerus shaft typed as female and the right petrous portion typed as male (x-axis
= cycle number, y-axis = RFU values). The bold line represents the threshold value of 10
RFUs. (A) The plot for 1188 left humerus shaft demonstrated a signal crossing the
threshold for Alu at 36 cycles, but the DYZ1 signal did not cross the threshold. (B) The
plot for 1188 right petrous portion demonstrated a signal crossing the threshold for Alu

and DYZ1 at 36 and 35 cycles, respectively. Note the non-sigmoidal curve of the DYZ1
signal.

72

DISCUSSION

Generating a physical profile can be extremely important in a forensic
investigation of human remains, one foundation of which is accurate sex determination.
Standard anthropological sexing methods are generally reliable with high quality
materials; however, when remains are incomplete or degraded, molecular sexing
techniques may be necessary. In the current study, a more sensitive molecular sexing
assay was sought in an attempt to overcome the shortfalls of the standard assay based on
the single copy locus amelogenin. The repetitive elements DYZ1 and Alu proved to be a
more sensitive sexing tool than amelogenin, both in control assays and when sexing
samples of forensic interest such as hair shafts and aged skeletal remains.

Optimization of the amelogenin assay showed that male and female control DNA
produced standard amplification curves along with dissociation curves at the correct
temperatures for the respective X and Y chromosome products (Figures 11 and 12). This
allowed the assay to be used for direct comparison to that with DYZ] and Alu. Similarly,
optimized results for the DYZ1/Alu assay showed standard curves for both loci using
control DNA (Figure 13), thereby allowing it to be compared with the amelogenin assay.

While the DYZ1/Alu assay correctly typed male and female control DNA, both
markers amplified in some of the negative controls. This was likely caused by
contamination resulting from handler DNA on supplies or within reagents, especially
primers and probes purchased from venders with unknown quality control procedures.
Given this, all supplies and workstations were extensively irradiated to destroy any DNA.

In addition, the supermix and primers were irradiated. Periods longer than 30 seconds

73

apparently damaged the primers enough that amplification did not occur, so 30 second
irradiation became the standard. The probes were not irradiated because the UV affected
the fluorescent dyes such that no signal was detected in the rtPCR reaction (data not
shown). Once these precautions were executed, the DYZ1 locus did not amplify in the
negative controls; however, the Alu contamination persisted and produced signals that
crossed the threshold at 40 cycles or more. Although Alu amplification in the negative
control was not ideal, the DYZl/Alu assay was still viable for unambiguous male sex
determination as long as DYZl amplified only with male control DNA and not in
negative or female controls. Furthermore, the Alu signal for known female DNA crossed
the threshold at much lower Ct values than in the negative control, thereby differentiating
between a female sexing result and DNA contamination. On the other hand, with
extremely low quantities of DNA, an amplification curve could cross the threshold at a
high Ct value, making differentiation of contamination and a true female sex result
difficult. Therefore, in an actual investigation, such small amounts of DNA would not be
reliable for sex determination if contamination existed. When an actual female sample
with a high Ct value was observed, additional reactions would need to be performed
using an increased amount of DNA to ensure the sex determination was accurate.

Sexing of the DNA dilution series showed that the amelogenin assay was
successful using as little HMW DNA as found in approximately 15 cells. Artificial
degradation resulted in more than twice as much DNA being required for amplification.
These findings make sense given that degradation of the DNA would destroy some of the
target loci, thus requiring more input DNA for amplification. This would be the case for

forensic samples as well, since DNA found in forensic evidence is often degraded. The

74

DYZ1/A1u assay was successful using HMW DNA quantities down to 1% of a cell’s
DNA (0.0623 pg), approximately 1500 times more sensitive than the amelogenin assay.
Counterintuitively, the DYZ1/Alu assay was even more sensitive when DNA was
artificially degraded, an exact opposite finding of the amelogenin assay. The presumed
reason for this is that the DYZ] repetitive elements are all linked at the heterochromatin
region of the Y chromosome (Roewer et al. 1992), therefore acting as a single locus in
intact DNA. When analyzing only a few cells’ worth of DNA, stochastic sampling effects
presumably occurred, in which the DYZ1 region was sometimes sampled and other times
it was missed. However, when DNA was degraded, these repeat sequences became
fragmented, making it more likely to sample the DYZ1 locus fi'om thousands of repeat
copies instead of from one intact sequence. Conversely, the Alu repeat elements are
dispersed throughout the genome, making it possible to sample the locus with both HMW
and degraded DNA, even at low quantities.

The dilution series of degraded DNA experiments also had several concentrations
where DYZ1 amplified in only one replicate. This was likely caused by the lack of
optimal digestion of the DYZ1 region, once more resulting in stochastic sampling effects.
Further, the DYZ1 locus may not fi'agment evenly; if the DYZl sequence were digested
into 10 kb blocks on average, approximately 7 O repeats would be linked together, again
making stochastic sampling much more likely than with Alu. However, if these repeats
were further segmented and dispersed throughout the sample, the target locus would be
easier to sample. Conversely, because of the high distribution of Alu repeats, the loci
would already be adequately dispersed. This, paired with the slightly higher copy number

of Alu repeats, helps explain why the DYZ1 locus did not perfectly mimic Alu

75

amplification, resulting in a difference in their Ct values. The difference could pose a
problem when analyzing forensic samples with extremely small amounts of DNA. In this
study, a few of the correctly sexed male hairs demonstrated DYZ1 signals with Ct values
5 - 10 cycles higher than the Alu signal (data not shown). Nuclear DNA from hair shafts
is highly degraded, but due to the highly compact nature of the Y chromosome
heterochromatin region, the DYZ1 sequence may be protected from extensive
fragmentation (Roewer et al. 1992), resulting in stochastic sampling and thus the
separation of the DYZl and Alu Ct values. In contrast, it is also possible that the DYZl
locus is more susceptible to degradation, resulting in its higher Ct value. Given these
findings, analysis of forensic samples with the DYZ1/Alu assay may result in the
amplification curves for the two loci being separated more than normal. However, when
using control DNA this artifact was only seen when less than 1% of one cell’s worth of
DNA was assayed. Therefore, seeing such separation may simply indicate that a forensic
sample contains an insufficient amount of DNA for definitive sex determination.

Very large amounts of control DNA also produced unanticipated rtPCR results in
some instances. Only one replicate amplified both the DYZl and Alu for the largest
HMW DNA quantity (20000 pg) (Figure 15). The DYZ1 amplification signal crossed the
threshold at an extremely low Ct value with a curve that was not sigmoidal, which was
likely attributable to the large quantity of DNA itself. If too much DNA is present in PCR
reactions, reagents can become limited within the first several cycles resulting in
inadequate amplicon production. However, because the DYZ1/Alu assay displayed
sigmoidal curves for both loci at similar Ct values at lower DNA quantities, simply

diluting the large amount of DNA would produce viable sex determination results.

76

Comparison of the amelogenin and DYZ1/A1u assays using hair shaft DNA
showed that two-thirds analyzed with the amelogenin assay produced X and Y products,
and 80% of these yielded a correct sexing result. The remaining third generated either no
product or produced a signal for both loci with incorrect dissociation temperatures,
resulting in no sex call (e. g., Figure 16). Furthermore, two DNAs typed as male when the
donor was female, both of which demonstrated what seemed to be reasonable
amplification and dissociation curves (Figure 17 and 18). A possible explanation for this
was cross contamination between other samples in the reaction set. However, if
contamination was the cause, the Ct values would presumably be far apart, but in both
cases similar amelogenin Ct values were observed, meaning contamination was unlikely.
A sample mix-up when setting up the reaction would also explain the incorrect results,
but the hairs typed correctly using the DYZ1/Alu assay, as detailed below. Additionally,
two hairs analyzed with the amelogenin assay typed as female although the donor was
male, generating an X product with a correct dissociation temperature. The dropout of the
Y chromosome could be related to amplifying different sized amplicons in highly
degraded DNA. Because the amelogenin Y product is larger than the X sequence,
assaying an intact copy of it could be more difficult, explaining why a male hair typed as
a female.

When hair shaft DNA was analyzed using the DYZl/Alu assay, all hairs yielded a
sexing result. Ninety-three percent of these were correct, or about twice as many as the
amelogenin assay. In addition, the four hairs that were incorrectly called with the
amelogenin assay produced accurate sexing results with the DYZl/Alu assay. However,

two other hairs were incorrectly sexed as male, as the DYZl curve crossing the threshold.

77

It should be noted that neither of these hairs (19 and 28) were “clean” results. Hair 19
generated a DYZ1 signal with a Ct value 15 cycles higher than the Alu signal (Figure
21), while hair 28 displayed a non-sigmoidal curve for DYZ1 from the organic
extraction, along with a Ct value also 15 cycles higher than Alu with the alkaline
digestion, making the male results dubious. Potential explanations for these data are that
both female donors actually had a Y chromosome, or at least a segment of DNA similar
enough to DYZ1 to allow amplification. Schmid et a1. (1992) described females who
harbored a Y chromosome along with two X chromosomes. However, there have been
only five reported cases of XXY females, making it extremely unlikely that two female
hair donors displayed this characteristic. Furthermore, because the Alu sequence is not
related to the number of X chromosomes, the amplification plot for an XXY female
would still present DYZ1 and Alu signals that would cross the threshold at similar Ct
values. A more likely explanation is contamination with a small amount of male DNA,
which would explain why the DYZ1 amplification curves crossed the threshold much
later than theAlu product. In addition, Graffy (2003), who originally purified the hair
DNAs, had difficulty analyzing the mitochondrial DNA from Hair 19, in that it had many
ambiguous base calls. It appears, owing to the inconsistent mtDNA results and the
DYZ1/Alu results, there is a good chance that the female hair DNA sample was
contaminated with a small amount of male DNA.

Similar amelogenin and DYZl/Alu results were seen in assays of the ancient
skeletal remains. Amelogenin typing showed just a third of the adult bones yielded a
sexing result; four of which were the same as the anthropological estimated sex, and the

six bones that were inconsistent typing as female. Most of these demonstrated an X

78

product with a clear sigmoidal amplification curve and a dissociation peak at the correct
temperature. One of the inconsistent bones (1226 left ulna midshaft) did produce a Y
signal (Figure 23); however, because the dissociation temperature was ten degrees too
high, it was not considered to be from the Y product. Similar to the hairs, these bones
were highly degraded, potentially making assaying the larger Y target sequence more
difficult than the X sequence. In addition, the burial 156 femur had an unusual female
result (Figure 27 A and B): the curve from the X product barely crossed the threshold,
and did so at an extremely high Ct value, while the Y signal produced a non-sigmoidal
amplification curve. Normally these results would be considered inconclusive; however
because the X signal did produce a small dissociation peak at the correct temperature, the
bone was typed as female for this study. Realistically, it would be impossible to make a
definitive sex determination with such data. Burial 97 also had two bones with
inconsistent amelogenin sexing results (Figure 28), although neither of these yielded
unambiguous data. The tibia showed amplification and dissociation curves consistent
with male DNA, but the X and Y Ct values were separated by ten cycles. This makes it
hard to determine if the Y signal was fiom the bone or contamination. The femur, on the
other hand, displayed curves consistent with female DNA, but the X curve crossed the
threshold at approximately 40 cycles, indicating that only a small amount of the X
product amplified, making a definitive sex determination difficult. Due to the
unreliability of either sexing result for the tibia and femur, a sex determination would not

be possible for this individual.

79

Only two juvenile bones were sexed using the amelogenin assay. The left ulna
midshaft of burial 1224 produced an X signal that crossed the threshold and had a
dissociation curve at the correct temperature. Burial 1264 fibula midshaft produced both
an X and Y signal that crossed the threshold and yielded dissociation curves at the correct
temperature. In contrast, the 1224 left petrous portion and the 1264 tibia cortical fragment
did not produce a signal for either locus (data not shown). In a forensic investigation
multiple bones fi'om an individual would ideally be assayed to make a definitive sex
determination. If only one bone is available, multiple DNA extractions should be
performed and analyzed separately to ensure the accuracy of the sex result.

The DYZ1/Alu assay of skeletal remains produced far more robust results. All
adult and juvenile bones yielded sexing data, and 84% of the adult remains had sexing
results consistent with the anthropological sex, along with five that differed among four
burials. Each of the latter group demonstrated clear sigmoidal curves for Alu and no
DYZ1 result, thereby typing as female. Furthermore, two or three bones were analyzed
for 16 adult and juvenile burials, 11 of which produced consistent sexing results among
all bones. Five burials produced inconsistent sexing results, including Kamenica burials
(97, 142 and 156). Of these, the 97 tibia, 142 pelvis, and 156 femur and humerus typed as
female, while the remaining bones typed as male. One explanation for such findings is
that some bones produce better DNA results than others. Edson et al. (2004) discussed
the success rate of analyzing mtDNA fi'om different bone types. Wei ght-bearing bones,
such as the tibia, have the densest cortical material and generated the best DNA results
(approximately 90%), while spongy bones, such as the pelvis, demonstrated a success

rate of approximately 65%. This could explain why only Alu amplified from the 142

80

pelvis, while both loci amplified from the humerus and petrous portion DNAs. However,
because both the tibia and femur are weight-bearing bones, it is unlikely that the
inconsistent sexing results for burial 97 and 142 were caused by the difference in bone
density. A second explanation for inconsistent sexing results is that the bones were not
actually from the same individual. The Kamenica tumulus includes double burials (Bejko
et al. 2006) several of which were analyzed by Clemmer (2005). In those burials it was
clear that two sets of skeletal remains were present. However, many of the Kamenica
burials contained only fragmented remains. If skeletal fragmentation is high enough and
the bulk of the bones are missing, it would be difficult or in some instances impossible to
determine if remains were from one or more than one individuals. Likewise, the nature of
a burial mound means that graves occur over other graves, and that commingling may
happen over time, either during subsequent burials as the ground shifts, or as the
skeletons are unearthed. For example, by observing the diagram of burials 142 and 156
(Figure 33) from the Kamenica map of the lower level of the tumulus, neither burial
depicts a full, undisturbed skeleton. It seems plausible that bones from two individuals

were unknowingly collected from a grave, accounting for the inconsistent sexing results.

81

 

Figure 33: Magnified View of Burials 142 and 156
Partial skeletons of burials 142 and 156 from the map of the lower layer of the tumulus in
Kamenica, Albania. See Figure 8 for the full map.

Inconsistent results were also observed from two of the Butrint burials (1518 and
1188). The humeral cortical fragment of 1518 showed a clear sigmoidal curve for Alu,
while the fibula midshaft produced signals for both DYZ1 and Alu (Figure 25). However,
the Ct values for the latter were approximately ten cycles apart. Once again, this could
result fi'om contamination with male DNA or the other factors detailed above. A potential
explanation for incompatible sexing results for bones of burial 1518 is the commingling
of remains. Individual 1518 was buried in the same grave as 1517, mapped in Figure
34A. Figure 348 shows that the burial contained a large skeleton of an approximately 40
year old male (individual 1518), while the bone fragments at the foot of the grave
(indicated by the arrow) belonged to 1517, an infant of 3 — 6 months. When the bones
were first unearthed, archaeologists bagged both skeletons together (Fenton, personal
communication). The bones were then rinsed with water by locals from the village of

Butrint, and anthropologists later separated the remains for further analysis; it is possible

82

that during the cleaning process bones from multiple burials were commingled,
explaining the different sex results for individual 1518 (F enton, personal
communication). Alternatively remains from individual 1518 are in fact female. The
humeral cortical fragment produced a clear Alu amplification curve in three replicates
(data not shown) giving confidence to a female sex call. However, the DYZ1
amplification curves for the two additional replicates of the fibula midshaft were non-

sigmoidal and may not be the DYZ1 product (data not shown), thereby misidentifying the

remains as male.

       

.\ \ . ‘ 1.. “A

. . A ' .. . ,
Figure 34: Map and Photograph of Burials 1518 and 1517
(A) A magnified view of burials 1518 and 1517, indicated by the circle, taken from the
Butrint map (Figure 7). (B) A photograph of burials 1518 and 1517. The large skeleton is
individual 1518, an estimated 40 year old male. The fragments at the foot of the grave

belong to individual 1517, an infant of 3 — 6 months (arrow). (Modified fiom Rauzi,
2007).

83

Similarly, juvenile burial 1188 (Figure 32) presented a left humerus shaft that
typed as female with a clear sigmoidal curve for Alu and a DYZ1 signal that did not cross
the threshold, while the right petrous portion produced a signal for both loci. However,
the curve for DYZ1 was not sigmoidal for either bone, bringing the male result into
question. The Butrint map displayed that individual 1 188 was buried along with
individual 1189 (Figure 35A), and the photograph of this burial (Figure 35B) shows two
skeletons present, the smaller of which is indicated by the arrow. Because of the
commingling of remains, the humerus shaft or the petrous portion analyzed for burial
l 188 could belong to either individual, leading to the inconsistent sexing results

observed.

84

 

 

 

 

 

Figure 35: Map and Photograph of Burials 1188 and 1189

(A) A magnified view of burials 1188 and 1189, indicated by the circle, taken from the
Butrint map from Figure 7. (B) A photograph of burials 1188 and 1189. It is unclear
which skeleton belongs to either individual. The smaller skeleton is indicated by the
arrow. (Modified from Rauzi, 2007).

The high sensitivity of the DYZ1/Alu assay is also its major drawback, resulting
in susceptibility to contamination; however, it is possible for contamination to occur in
one direction. Since Alu is present in both males and females, only DYZ1 can
contaminate female DNA resulting in an incorrect sex determination, while a sexing
result from contaminated male DNA would not type as female. Fortunately, DYZ1
contamination is likely to be present at only low levels, leading to an amplification curve
with higher Ct values than that for Alu. How this might actually present itself could be

examined by testing mixtures of female/male DNA at differing ratios to see how little

85

male DNA is required to produce a male sexing result. Furthermore, observing the
difference in Ct values for the curves at different ratios would aide in the identification of
male contamination within a female sample.

There are several precautions that must be taken while using the DYZ1/Alu assay.
Proper controls must be run with every set of samples to ensure that the sexing results are
reliable. Both female and male control DNA must demonstrate amplification curves with
Ct values that demonstrate the assay is working properly. Ideally, negative controls
would show no amplification for either locus to demonstrate that the reaction solutions
are not contaminated. While the Alu signal in the negative control did not affect the
results of the present experiments, in a non-experimental situation one would want no
amplification for both DYZ1 and Alu. In this regard, it is important to obtain reagents
from suppliers with proper quality control to avoid all contamination during their
production. Each reagent should be analyzed using the DYZ1/Alu assay to make sure it is
free of contamination. High performance liquid chromatography purification of primers
and probes may help to combat contamination as well. While setting up the assay, control
DNA should be added after sample DNA, to minimize the chance of cross contamination.
Furthermore, analyzing multiple DNA extractions from one piece of evidence in separate
runs would help to monitor cross contamination among the samples. When sexing ancient
remains with the DYZ1/Alu assay, it would be most advantageous to analyze multiple
bones fi'om an individual, thereby ensuring the accuracy of the sex determination. It is
also important to become comfortable with interpreting the amplification curves of the
DYZ1/Alu assay. When the curves are sigmoidal and clearly cross the threshold, it is

easy to make a call. However, some results may require a certain amount of

86

 

interpretation, and experience analyzing these types of data will make the sex
determination simpler and more accurate.

Considering the results presented here as a whole, it is clear that the DYZl/Alu
assay outperformed the amelogenin assay with both control and forensic DNA samples.
The DYZl/Alu assay was more sensitive with low quantities of HMW and degraded
DNA, more precise with DNA from hair shafts of known sex, and more accurate when
sexing ancient skeletal remains. While the extreme sensitivity of the DYZ1/Alu assay
may create challenges, by taking the aforementioned precautions, the assay makes an
excellent tool for sexing the most challenging samples. Ultimately, this assay can assist in
determining the sex of an individual for forensic investigations and anthropological

research when standard sex determining methods are insufficient.

87

WORKS CITED

Andréasson H, Allen M. 2003. Rapid quantification and sex determination of forensic
evidence materials. Journal of Forensic Sciences 48(6): 1280 — 1287.

Bailey DMD, Affara NA, Ferguson-Smith MA. 1992. The X-Y homolgous gene
amelogenin maps to the short arms of both the X and Y chromosomes and is
highly conserved in primates. Genomics 14: 203 - 205.

Bartha J L, Finning K, Soothill PW. 2003. F etal Sex Determination from Maternal Blood
at 6 Weeks of Gestation when at Risk for 21-Hydroxylase Deficiency. Obstetrics
and Gynecology 101(5):]135 —- 1136.

Bejko L, Fenton T, Foran D. 2006. Recent advances in Albanian mortuary archaeology,
human osteology, and ancient DNA. In L Bejko and R Hodges (Ed.), New
Directions In Albanian Archaeology: Studies presented to Muzafer Korkuti. 309 —
322. Tirana, Albania: International Centre for Albanian Archaeology.

Budowle B, Hobson DL, Smerick JB, Smith JAL. 2001. Low copy number —
considerations and cautions. Proceedings of the Twelfth International Symposium
on Human Identification. Madison, Wisconsin: Promega Corporation.
<http://www.promega.com/geneticidproc/ussymp12proc/contents/budowle.pdf>.

Burger J, Hummel S, Herrmann B, Henke W. 1999. DNA preservation: A microsatellite-
DNA study on ancient skeletal remains. Electrophoresis 20: 1722 -- 1728.

Butler J M. 2007. Short tandem repeat typing technologies used in human identity testing.
BioTechniques 43 (4):ii — v.

Clemmer VL. 2005. Maternal relatedness within double burials of an ancient Albanian
tumulus. Thesis for the Degree of MS. School of Criminal Justice. Michigan
State University.

Cox M, Mays S. 2000. Human Osteology: In Archaeology and Forensics. Cambridge
University Press, New York, New York.

Edson SM, Ross JP, Cobble MD, Parsons TJ, Barritt SM. 2004. Naming the dead —
confronting the realities of rapid identification of degraded skeletal remains.
Forensic Science Review 16(1): 64 — 88.

Faerman M, Filon D, Kahila G, Greenblatt CL, Smith P, Oppenheim A. 1995. Sex
identification of archaeological human remains based on amplification of the X
and Y amelogenin alleles. Gene 167: 327 - 332.

88

F enton, T. Personal communication. 2008.

Foy CA, Parkes HC. 2001. Emerging homogeneous DNA-based technologies in the
clinical laboratory. Clinical Chemistry 47(6): 990 — 1000.

Graffy EA, Foran DR. 2005. A Simplified Method for Mitochondrial DNA Extraction for
Head Hair Shafts. Journal of Forensic Sciences 50(5):] 119 — 1122.

Graffy EA. 2003. Development and Validation of an Alkaline Extraction Method for
Isolating Mitochondrial DNA from Human Hair Shafts. Thesis for the Degree of
MS. School of Criminal Justice.

Haas-Rochholz H, Weiler G. 1997. Additional primer sets for an amelogenin gene PCR-
based DNA-sex test. International Journal of Legal Medicine 110: 312 - 315.

Hodges R, Bowden W, Gilkes O, Lako K. 2000. Late Roman Butrint, Albania: survey
and excavations, 1994 -— 98. Archeologia Medievale, XXV II: 241 — 257.

Jackson CB. 2005 . A more sensitivity sex determination assay. Thesis for the Degree of
MS. School of Criminal Justice.

Kimmerle EH, Ross A, Slice D. 2008. Sexual Dirnorphism in America: Geometric
Morphometric Analysis of the Craniofacial Region. Journal of Forensic Sciences.
53(1):54 - 57.

Komar DA, Potter WE. 2007. Percentage of Body Recovered and Its Effect on
Identification Rates and Cause and Manner of Death Determination. Journal of
Forensic Sciences. 52(3):528 — 531.

Mandrekar MN, Erickson AM, Kopp K, Krenke BE, Mandrekar PV, Nelson R, Perterson
K, Shultz J, Tereba A, Westphal N. 2001. Development of human DNA
quantitiation system. Croatian Medical Journal 42(3):336 — 339.

Mighell AJ, Markham AF, Robinson PA. 1997. Alu sequences. Federation of European
Biochemical Society Letters 417(1):] —— 5.

Nakahori Y, Harnano K, Iwaya M, Nakagome Y. 1991. Sex identification by polymerase
chain reaction using X-Y homologous primer. American Journal of Medical
Genetics 39: 472 — 473.

Nesser D, Liechti-Gallati S. 1995. Sex determination of forensic samples by simultaneous
PCR amplification of a-Satellite DNA from both the X and Y chromosomes.
Journal of Forensic Sciences 40(2): 239 — 241.

89

Nicklas J A, Buel E. 2006. Simultaneous determination of total human and male DNA

using a duplex real-time PCR assay. Journal of Forensic Sciences 51(5): 1005 —
1 01 5.

Nicklas JA, Buel E. 2005. An Alu-based, MGB EclipseTM real-time PCR method for
quantitation of human DNA in forensic samples. Journal of Forensic Sciences
50(5): 1 — 10.

Nicklas JA, Buel E. 2003. Quantification of DNA in forensic samples. Analytical
Bioanalytical Chemistry 376: 1160 — 1167.

Poinar HN. 2003. The top 10 list: criteria of authenticity for DNA from ancient and
forensic samples. International Congess Series 1239: 575 — 579.

Rahman MM, Basharnboo A, Prasad A, Pathak D, Ali S. 2004. Organizational variation
of DYZ1 repeat sequences on the human Y chromosome and its diagnostic
potentials. DNA and Cell Biology 23(9): 561 — 571.

Rauzi CM. 2007. Genetic analysis of burials from the Butrint, Albania Triconch Palace
and Merchant’s House. Thesis for the Degee of MS. School of Criminal Justice.

Rennick SL. 2005. Genetic analysis of a Mounumental Structure within the Kamenica,
Albania tumulus. Thesis for the Degee of MS. School of Criminal Justice.

Reynolds R, Varlaro J. Gender determination of forensic samples using PCR
amplification of ZFX/ZFY gene sequences. Journal of Forensic Sciences 41(2):
279 — 286.

Roewer L, Amemann J, Spurr NK, Grzeschik KH, Epplen JT. 1992. Simple repeat
sequences on the human Y chromosome are equally polymorphic as their
autosomal counterparts. Human Genetics 89:389 — 394.

Rozen S, Skaletsky H. 2007. Primer3 v.0.4.0. <http://frodo.wi.mit.edu/>.

Saiki RK, Gelfand DH, Stoffel S, Scharf SJ, Higuchi R, Horn GT, Mullis KB, Erlich HA.
1988. Primer-directed enzymatic amplification of DNA with a thermostable DNA
polymerase. Science 239: 487 - 491.

Schmid M, Guttenbach M, Enders H, Terruhn V. 1992. A 47,XXY female with unusual
genitalia. Human Genetics 90: 346 — 349.

Stacks B, Witt MM. 1996. Sex determination of dried blood stains using the polymerase

chain reaction (PCR) with homologous X-Y primers of the Zinc Finger Protein
Gene. Journal of Forensic Sciences 41 (2): 287 -— 290.

90

Sullivan KM, Mannucci A, Kimpton CP, Gill P. 1993. A rapid and quantitative DNA sex
test: fluorescence-based PCR analysis of X-Y homologous gene amelogenin.
BioTechniques 15(4):636 — 641.

Tyler-Smith C, Taylor L, Miiller U. 1988. Structure of a hypervariable tandemly repeated
DNA sequence on the short arm of the human Y chromosome. J ouma] of
Molecular Biology 203:837 — 848.

Witt M, Erickson RP. 1989. A rapid method for detection of Y-chromosomal DNA from

dried blood specimens by the polymerase chain reaction. Human Genetics 82: 271
— 274.

91