SE (“WWI}!IWIIHIWIWWIIIJIHJHHWW‘JIM

THS

    

Iflfi' unmn
I 9.. Michigan State
J? 0 0 5? University

This is to certify that the
thesis entitled

INVESTIGATIVE STUDIES INTO THE RECOVERY OF DNA
FROM IMPROVISED EXPLOSIVE DEVICE CONTAINERS

presented by
Shane Gregory Phillip Hoffmann

has been accepted towards fulfillment
of the requirements for the

MS. degree in Forensic Science

Major Professor’s STgnature
5/30/05

Date

 

 

 

MSU is an affirmative-action, equal-opponunily 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:lProj/Acc&Pres/CIRCIDateDue indd

I_'_/_’_________r____,__——-———r—/

INVESTIGATIVE STUDIES INTO THE RECOVERY OF DNA FROM
IMPROVISED EXPLOSIVE DEVICE CONTAINERS

By

Shane Gregory Phillip Hoffmann

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

MASTER OF SCIENCE
Department of Forensic Science

2008

ABSTRACT

INVESTIGATIVE STUDIES INTO THE RECOVERY OF DNA FROM
IMPROVISED EXPLOSIVE DEVICE CONTAINERS

By

Shane Gregory Phillip Hoffinann

Combating improvised explosive devices (IEDs) and apprehending those
responsible have become national priorities due to their use in the Middle East and the
threat they pose domestically. IEDs are often concealed in containers (e.g., a backpack,
box, or briefcase), as was demonstrated in the Centennial Olympic Park and Madrid train
bombings. The goal of this research was to identify the person(s) responsible for an IED
through post-blast DNA recovery from IED containers. Eight study participants were
asked to use backpacks in everyday activities for eleven days, after which they served as
containers for pipe bombs. Regions likely to be handled by the study participants were
swabbed, and DNA recovered was amplified and typed using miniSTRs. Handler
profiles were called blindly using data from all swabs. Profiles compiled for seven of
eight backpacks matched the handler’s at all nine loci, with DNA recovered fiom all
swabs producing at least the handler’s partial profile. Overall, higher yields of DNA and
more loci with handler alleles were obtained from the straps, top handle, neck region, and
front middle. Recovering DNA fiom IED containers is a practical approach that that can
easily be implemented in IED investigations and vastly improves upon the discriminatory
power achieved in previous studies that concentrated on collecting DNA from IEDs

themselves.

ACKNOWLEDGEMENTS

My sincere appreciation is extended to my committee members for the time they
have devoted to assisting me throughout the research and writing process. To my
advisor, Dr. David Foran, I would like to thank you for three semesters of financial
assistance and for the continuous support of my research. Also, thank you for the
valuable insight you contributed to the project and your many hours of revisions. My
gratitude is extended to First Lt. Shawn Stallworth, who, along with other members of the
bomb squad, put their lives in harms way to make this research possible. Finally, I would
like to thank Dr. Sheila Maxwell for volunteering to serve on the committee of a student
she had just recently met.

My thanks extend to my peers who directly assisted in the project. To Ms.
Stefanie Kremer, whose patience and guidance was instrumental in familiarizing me with
the lab and my research. To Ms. Kamila Gomez, for her assistance in collecting and
separating post-blast fragments. Also, to the study participants who graciously
volunteered to carry around the backpacks for a week and a half.

I would further like to thank my classmates, friends, and teachers who have all
contributed to making my experience at Michigan State one that I will hold fond
memories of. I came here for an education, but am leaving with so much more.

Finally, my gratitude is extended to my family for all the years of unwavering

support and always believing in my abilities despite some difficult times.

iii

TABLE OF CONTENTS

LIST OF TABLES .............................................................................................................. v
LIST OF FIGURES ........................................................................................................... vi
INTRODUCTION .............................................................................................................. 1
Past research in identifying IED handlers using DNA ................................................... 4
miniSTRs ........................................................................................................................ 7
Analysis of trace and low copy number DNA ................................................................ 9
PCR inhibition and DNA quantitation .......................................................................... 10
Analysis of mixtures ..................................................................................................... 12
Study aims: A new approach to identifying those involved in IED campaigns ........... 14
MATERIALS AND METHODS ...................................................................................... 17
Backpack preparation and distribution ......................................................................... 17
Pipe bomb preparation .................................................................................................. 19
Backpack collection and bomb handling ...................................................................... 20
Bomb deflagration and collection ................................................................................. 20
DNA recovery and extraction ....................................................................................... 21
Amplification of miniSTRs using the MinifilerTM kit .................................................. 23
Capillary electrophoresis and miniSTR analysis .......................................................... 24
DNA quantification using real-time PCR ..................................................................... 25
RESULTS ......................................................................................................................... 27
Deflagration observations ............................................................................................. 27
Swabbing backpacks to recover shed cells ................................................................... 28
Isolating DNA from recovered cells ............................................................................. 28
Analysis of STR electropherograms ............................................................................. 29
Categorizing loci based on the presence of handler alleles .......................................... 30
Peaks attributed to increased stutter and the occurrence of mixtures ........................... 34
Comparing regions based on the number of loci with the handler’s alleles ................. 34
DNA quantification ...................................................................................................... 35
Amounts of DNA added to PCR reactions ................................................................... 38
Detection of PCR inhibition via the internal positive control ...................................... 43
DISCUSSION ................................................................ ‘ ................................................... 44
CONCLUSIONS .............................................................................................................. 57
APPENDIX A. Allele calls and loci categorizations ........................................................ 59
APPENDIX B. DNA quantitation data ............................................................................. 70
REFERENCES ................................................................................................................. 74

iv

LIST OF TABLES

Table 1. A review of reported miniplexes. ......................................................................... 9
Table 2. Identifiers for labeling swabs .............................................................................. 19

Table 3. Consensus handler profiles obtained from backpacks and the study participant to

which they corresponded. ................................................................................... 30
Table 4. Number of loci with handler alleles broken down by backpack region. ............ 35
Table 5. Quantities of DNA recovered from backpack swabs. ........................................ 37
Table 6. DNA quantity added to PCR reactions. ............................................................. 38

LIST OF FIGURES

Figure 1. Regions of the backpacks targeted for DNA collection. ................................... 18
Figure 2. Crate used to retain backpack and IED fragments. ........................................... 21
Figure 3. Backpack fragmentation. ................................................................................... 27
Figure 4. Percentage of loci placed in each of the six locus classifications. .................... 31
Figure 5. Examples of four locus classifications from backpack 6 samples. ................... 32
Figure 6. Number of loci containing handler’s alleles from backpack swabs. ................. 33

Figure 7. Electropherogram from 4Z1 UP, representing the recommended amount of input
DNA for typing. ................................................................................................ 40

Figure 8. Electropherogram from 3ZSP, representing LCN typing. ................................. 41

Figure 9. Electropherogram fiom 8LSP, representing too much DNA being added to the
PCR reaction. .................................................................................................... 42

Figure 10. Quantitation results for 3RSP. ......................................................................... 43

Images in this thesis are presented in color.

INTRODUCTION

Improvised explosive devices (IEDs) have gained publicity due to their use in
ongoing conflicts in the Middle East as they have become the weapon of choice for
terrorists. As of November 2007, IEDs have been responsible for 70% of American
combat casualties in Iraq and 50% of combat casualties in Afghanistan (Wilson 2007).
Their effectiveness in the Middle East has caused concern over their implementation
world-wide, especially in the United States. In October 2007, Department of Homeland
Security (DHS) Secretary Michael Chertoff addressed the Center for Strategic and
International Studies regarding IEDs. He emphasized the Department’s focus by saying,
“Our office is 100 percent committed to protecting the people of the United States from
IEDs. All of our counterterrorism efforts focus directly or indirectly on bombing
prevention...” (Chertoff 2007). This concern is well justified as the USA has shown
vulnerability to attacks by IEDs in the past.

An IED is “an explosive device that is placed or fabricated in an improvised
manner; incorporates destructive, lethal, noxious, pyrotechnic, or incendiary chemicals;
and is designed to destroy, incapacitate, harass, or distract” (National Research Council
2007). Counterterrorism consultant David Williams defined IEDs based on their
components and stated that an IED is “any collection of components that are compiled in
conjunction with a power source, an initiator, and an energetic material used as a main
charge” (personal communication). The definitions are rather broad, but they are
reflective of the fact that IEDs can take any form and are only limited by the ingenuity of

the developer.

Besides the flexibility in design that IEDs allow, they have become a p0pular
weapon for a variety of other reasons. The Counterterrorist Handbook (3012 et a1. 2005)

lists the following reasons as to why IEDs are so appealing:

l. The media flocks to events involving IEDs, drawing attention to the
administering group and their cause.

2. The actual bombing can be accomplished by a limited number of
people.

3. There is minimal risk of the bomber being detected.

4. IEDs are inexpensive and cost effective compared to other tactics
(kidnapping, etc.).

5. Random bombings have a considerable psychological impact on the
population.

6. Materials are often readily available and IEDs can be constructed

through the use of legitimately purchased products.
IED components can be brought to the target and assembled on-site.
8. IEDs can be readily concealed and easily transported and delivered.

>1

As is indicated in the last point, IEDs are often concealed and delivered in a
secondary container such as a backpack, purse, or suitcase, as these items do not draw
attention (Burke 2000). Some of the more notable IED events in recent history involved
the use of backpacks to conceal IEDs. For instance, on July 27, 1996, during the
Olympic Games in Atlanta, Georgia, a suspicious green backpack was noticed
underneath a bench in Centennial Park. Authorities observed wires and pipes in the bag
and assumed it was a bomb. As the area was being cleared a bomb threat was called in
that warned of a device like the one that was found. Shortly after 1 am. the bomb inside
the backpack went off, killing two people and injuring 111. The backpack contained
three pipe bombs along with nails and screws to increase the shrapnel (Noe 2007).
Another highly publicized IED attack in which the devices were concealed in backpacks
occurred in Madrid on March 11, 2004. Thirteen IEDs were placed on four commuter

trains during morning rush hour. Ten of these exploded, resulting in 191 deaths and over

1,800 injuries. The three IEDs that did not go off were detonated by police in a
controlled environment (Timeline: Madrid investigation 2004).

The flexibility permitted in IED design allows terrorists to continually adapt to
countermeasures. For example, jarnmers have been developed to block the signals of
radio controlled IED detonators such as cellphones (Wilson 2007). Two such jammers,
the IED Countermeasures Equipment (ICE) and the Warlock, hamper radio controlled
explosive detonators using low power radio fi'equency energy. Terrorists responded to
these countermeasure by simply making bombs with more powerful radio triggers or by
using infiared triggers, rendering the jarnrners less effective (Atkinson 2007a).
Complicating matters more is the fact that the time it takes terrorists to adapt is usually
shorter than the time involved in implementing new countermeasures (National Research
Council 2007). It will be up to the nation’s science and research community to continue
to deveIOp new countermeasures that can detect the most recent IED advancements
(Chertoff 2007).

In an ideal world, all IED events would be impeded before activation of the
device. Consequently, the majority of US research and funding has been targeted for
developing preventive measures against IED attacks (Wilson 2007). However, the
continuous evolution of IEDs, combined with the unlimited number of targets, makes
complete IED prevention impossible. If an IED is activated, focus must turn to
mitigating the effects and apprehending those responsible. As mentioned, IEDs have
become popular weapons among terrorists because there is very little risk of the
perpetrator being detected. Many IEDs can be set to detonate via a timing device, or their

detonation can be controlled from a remote location. Since terrorists are outnumbered by

the forces they are up against it is critical to the longevity of their campaign that they
remain elusive. This is especially true for terrorists who possess bomb making
knowledge (Atkinson 2007a). Not capturing terrorists leads to fear of more attacks
among the targeted population, and gives justification to the success of the attack in the
minds of the terrorists, prolonging the campaign. On the other hand, the capture of one
terrorist can drastically alter a group’s campaign as it often leads to the arrests of other
group members (Sageman 2004). With this in mind, there has been an increased focus on
the forensic analysis of recovered IEDs (Atkinson 2007b). As of September 2007, FBI
director Robert Mueller said that 2,500 latent prints had been developed fiom non-
detonated IED components recovered in the combat theaters, resulting in 60
identifications and more than 1,000 forensic matches among IEDs (Mueller III 2007). To
firrther aid in identifying IED perpetrators, forensic scientists have sought alternative

methods for identifying the manufacturers of IEDs.

Past research in identifizing IED handlers using DNA

Van Oorschot and Jones (1997) found that brief contact between a person and an
object is sometimes all that is needed to recover the handler’s DNA profile fi'om the
object. Since then, scientists have examined the feasibility of recovering DNA from a
variety of objects ranging from shoe insoles (Bright and Petricevic 2004) to deflagrated
pipe bomb components (Esslinger et a1. 2004; Gehring 2004; Kremer and Foran 2008),
the latter which directly pertains to the work presented here. The ability to obtain and
type DNA from pipe bombs is largely dependent on two factors, the number of shed

epithelial cells deposited on the bombs, and the level of DNA degradation. Research has

shown that individuals vary in their tendency to shed epithelial cells (Lowe et al. 2002),
hence, a ‘good’ shedder may leave enough cells to produce a full profile, while a ‘poor’
shedder will leave behind enough cells to produce a partial or no profile. When it comes
to the integrity of the DNA, it is very likely that DNA from the shed cells has already
experienced some level of degradation depending upon the conditions to which it has
been exposed (temperature, moisture, etc.). DNA on an IED is prone to further
degradation due to the heat produced during the explosion, which is problematic when
performing polymerase chain reaction (PCR) based assays, as the fragmented DNA
prevents recovery of larger arnplicons.

The initial attempt to obtain DNA profiles from exploded pipe bombs
investigated the prospect of typing recovered DNA using short tandem repeat (STR)
methodology (Esslinger et al. 2004). This study had minimal success as only one full
profile (10 loci) was obtained after typing DNA from twenty pipe bombs using Profiler
PlusTM (Applied Biosystems, Foster City, CA). A likely explanation for the results was
that not enough nuclear DNA was recovered, given that DNA quantitation using
Quantiblot® indicated that DNA samples from all twenty bombs were below the lowest
quantitation standard (.03125 ng/uL). Figuring the DNA in the samples was too dilute,
they were concentrated from 80 - 140 uL to 10 uL. This led to the one full profile
observed, however, the rest of samples still resulted in partial or no profiles.

Attempts to obtain DNA profiles from deflagrated pipe bombs turned to the
analysis of mitochondrial DNA (mtDNA) (Gehring 2004). Features of mtDNA,
including high copy number and resistance to degradation (Foran 2006; Holland and

Parsons 1999), made it a promising alternative to nuclear DNA testing. Robin and Wong

(1988) estimated the number of mitochondria per cell in mammals ranges fiom 80 — 680
depending upon cell type, and that the total number of mtDNAs can range from 200 —
1700 per cell. Foran (2006) examined features of mtDNA that make it a better alternative
than nuclear DNA when working with forensic samples that are likely to contain
degraded DNA. He found that mtDNA’s cellular location plays a substantial role in
slowing the rate of degradation, and that the transcriptional activity of the loci being
assayed may also influence their rate of degradation. Using mtDNA to identify IED
handlers allowed for 18 out of 38 bombs to be correctly assigned to a single donor
(Gehring 2004). Further, seven bombs were correctly assigned to a subset of donors,
while twelve and one bomb(s) were not assignable or incorrectly assigned, respectively.
These results were further supported by mtDNA work performed by Kremer and Foran
(2008) where 11 and 10 bombs (n = 34) were correctly assigned to a single donor or
subset of donors, respectively.

Despite improved success in identifying IED handlers through analysis of
mtDNA recovered from deflagrated pipe bombs, the results were not as individualizing
as those obtained using STRs. To try and enhance the discrimination power, Kremer and
Foran (2008) used two sets of ‘miniSTRs’ (see below), NCOl (Coble and Butler 2005)
and miniSGM (STRBase), to obtain handler’s profiles from deflagrated pipe bombs. In
this way, eight bombs were correctly assigned to a single donor while nine bombs were
correctly assigned to a subset of donors. The other 17 bombs in the study were either
incorrectly assigned (14) or not assigned (3). Their results shadowed those obtained
using mtDNA in regards to correctly assigning handlers, as no significant difference was

observed when comparing handler assignments using mtDNA and miniSTRs.

miniSTRs

Since their introduction into the field of forensic biology in the early 1990’s STRs
have developed into the standard marker for human identity testing (Butler 2006). In
1998, the FBI selected thirteen STR loci to be used in their Combined DNA Index
System (CODIS), a national database dedicated to storing DNA profiles from convicted
offenders and forensic casework (Budowle et al. 1999). By incorporating multiply dyes
and carefully designing primers, commercial multiplexes containing as many as 16 loci
(including the 13 CODIS loci) have since been developed (Collins et al. 2004; Krenke et
al. 2002). These multiplexes perform well for DNA typing of non-compromised
samples, however, their performance is hindered with degraded DNA (Butler et a1. 2003;
Coble and Butler 2005). Fragmented DNA often results in the inability to amplify larger
STR amplicons, preventing a full profile from being obtained due to allele and/or locus
drop-out (Whitaker et al. 1995). Researchers found that redesigning primers so they were
closer to the core repeat resulted in increased success in typing compromised DNA
samples (Ricci et al. 1999; Wiegand and Kleiber 2001; Yoshida et al. 1997). The
reduced sized amplicons became known as miniSTRs (Butler et al. 2003).

The flexibility in designing primers that allows for creation of 16 locus
multiplexes is lost when working with miniSTRs (Butler et a1. 2003). However,
researchers were still able to create “miniplexes”—combinations of primers for
amplifying multiply miniSTR loci at one time (Butler et al. 2003). Using redesigned
primers for 16 loci (including 12 CODIS loci) they created five miniplexes (miniplex 1 —
5) consisting of three loci, and one miniplex (Big Mini) containing 6 loci (Table 1). The

different combinations of primers were successfully used to amplify the targeted loci and

produced allele calls that were concordant with those obtained using commercial STR
kits. They also proved effective in amplifying degraded DNA samples. Additional
miniplexes consisting mainly of CODIS loci have since been described including a 6
locus miniplex (Grubwieser et al. 2006) and 5 — 7 loci miniplexes (Parsons et al. 2007)
(Table 1). In 2007, Applied Biosystems released the largest miniplex to date,
MinifilerTM, a miniSTR kit consisting of eight autosomal loci (seven of which are CODIS
loci) and the sex determining locus arnelogenin (Table 1). The kit was marketed for
amplification and typing of compromised (degraded and inhibited) DNA samples.

Despite progress in multiplexing miniSTR loci, the discriminatory power obtained
with commercial STR kits was still lacking. In order to increase the power of
discrimination, Coble and Butler (2005) created two more miniplexes (mini01 and
mini02), each consisting of three non-CODIS loci (Table 1). An additional 20 non-
CODIS miniSTR loci have recently been described in hopes of creating larger

multiplexes (Hill et al. 2008).

Table l. A review of reported miniplexes.

 

 

Number
Name of loci Loci included Source(s)
Miniplex 1 3 csrrpo, THOl, TPOX 1311;30:th
Miniplex 2 3 D58818, D881179, D168539 “
Miniplex 3 3 FGA, D78820, D21811 “
Miniplex 4 3 VWA, D138317, D18851 “
Miniplex 5 3 PentaD, PentaE, D281338 “
Bi Mini 6 CSFlPO, THOl, TPOX, FGA, “
g D78820, D21811
. . Coble and Butler
M1n101 (NCOl) 3 D1081248, D14Sl34, D2281045 (2005); STRBase
Mini02 (NCOZ) 3 DISI677, D28441, D482364 “
. . THO], Amelogenin, D281338,
M‘mSGM 6 D1885], D168539, FGA STRBase
MiniSTR- 6 THO] , Amelogenin, D281338, Grubwieser et a1.
multiplex D18851, D168539, FGA (2006)
MP1 7 THOl, D2181 1, D1885], D168539, Parsons et a1.
Amelogenin, FGA, PentaD (2007)
MP2 6 D21811, D138317, D78820, “
CSF 1P0, D881179, VWA
FGA, CSFlPO, D2181 1, ,,
MP3 5 PentaD, PentaE
D13S317, D78820, D281338, A lied Bios stems
MinifilerTM 9 D21811, D16SS39, D1885], pp (2007)y

CSFlPO, FGA, Amelgenin

 

Bold indicates a non-CODIS locus. The miniSGM miniplex and the miniSTR-multiplex

consist of the same loci, however, some primer sequences varied.

Analysis of trace and low copy number DNA

There are other concerns with recovering and analyzing DNA from IEDs besides

degraded DNA, including working with trace amounts of DNA, PCR inhibition, and

mixtures. Isolating DNA from shed epithelial cells usually results in the recovery of

small quantities of DNA, referred to as trace DNA (W ickenheiser 2002). The term low

copy number (LCN) has been coined to describe analysis of less than 100 pg of DNA

(Gill et al. 2000), which is ten times less than the 1 ng of input DNA that many

multiplexes call for (Gill 2001). Hence, working with and analyzing LCN DNA has
proven to be problematic as analyses can lead to increased allele drop-out due to
stochastic sampling effects, increased stutter, heterozygote peak imbalance, and are more
prone to sporadic contamination (Budowle et a1. 2001; Gill 2001).

Researchers have found ways to combat problems associated with the analysis of
small quantities of DNA leading to increased sensitivity of STR assays (Budowle et al.
2001). By reducing the overall PCR volume while maintaining the amount of input
DNA, the amplification product is more concentrated, resulting in greater peak heights.
Sensitivity can also be increased by filtrating the product after STR amplification, which
removes ions and other small molecules that may be preferentially injected into the
capillary during electrophoresis, interfering with the injection of the STRs. Using
deionized formarnide (required to keep the DNA denatured during capillary
electrophoresis) also helps ensure that DNA will be preferentially injected. Two
additional means of increasing sensitivity with LCN DNA samples include adding more
amplified product to the forrnamide and increasing the injection time. It is also advised
that replicate analyses be performed when working with minute quantities of DNA in

order to confirm allele calls (Gill et al. 2000).

PCR inhibition and DNA quantitation

PCR inhibition is another potential problem when working with DNA recovered
from deflagrated pipe bombs. It is not uncommon for forensic samples to include
substances that co-extract with DNA and inhibit amplification (reviewed by Bessetti

2007). Heme and humic acid, found in blood and soil respectively, have been shown to

10

cause inhibition (Akane et a1. 1994; Watson and Blackwell 2000), and could be sources
of inhibition in the analysis of deflagrated IEDs. Although not documented, another
potential source of inhibition when dealing with IEDs is residue resulting from the
explosive material.

STR electropherograms with small or absent peaks may be indicative of PCR
inhibition. One method to identify PCR inhibition involves spiking the potentially
inhibited sample with high quality DNA before performing PCR (e.g., Shutler et al.
1999). If a PCR product is not obtained it can be concluded that the sample was
inhibited. If product is observed the assumption can be made that the initial PCR
contained insufficient template, meaning that the DNA was of poor quality (degraded) or
that there was no DNA. Building on this approach, a method to recognize PCR inhibition
has been developed using an internal positive control (IPC) incorporated into DNA
quantitation via real-time PCR (Green et al. 2005).

QuantifilerTM (Applied Biosystems) is a standard kit for the quantification of
human DNA using real-time PCR. Quantifying DNA is a routine and integral step in the
processing of DNA samples, and is usually performed before STR analysis so the optimal
amount of DNA can be added to the PCR reactions. In addition to quantifying DNA, the
kit also contains a synthetic strand of DNA as an IPC. Failure of the IPC to amplify is an
indication of the presence of inhibitors.

If inhibition is detected, various measures have proven successful in overcoming
it. Supplementing the PCR reaction with adjuvants such as bovine serum albumin (BSA)
or betaine has been effective in alleviating inhibition caused by heme (Al-Sound and

Radstrom 2000; Kreader 1996). Bourke et al. (1999) found that purifying DNA in the

11

presence of sodium hydroxide leads to improved amplification which is likely due to
neutralizing inhibitors of T aq polymerase. Further, adding aluminum ammonium sulfate
during the extraction of DNA fiom soil samples was shown to reduce the co-purification
of inhibitors (Braid et al. 2003).

There are also numerous ways to overcome PCR inhibition that do not involve
additives (reviewed by Radstrom et a1. 2004). Performing PCR using a diluted DNA
sample results in increased amplification efficiency due to a reduction in inhibitors.
However, this remedy is limited by the amount of DNA recovered, as diluting the sample
too far may lead to an insufficient amount of template. A common method for reducing
the effects of inhibitors is to purify (or repurify) the sample. Some purification options
include spin columns, gel filtration, and DNA binding beads. Oftentimes multiple

measures are utilized to decrease the chances of PCR inhibition (Radstrom et al. 2004).

Analysis of mixtures
A DNA mixture occurs when two or more genotypes are present in an

electrophero gram. Mixtures complicate the analysis of both mitochondrial and nuclear
DNA markers, leading to results with lower discriminatory power, or in some cases, no
evidentiary value. The first indication of a mixture in an STR profile is the observation
of three or more alleles at a locus (Clayton et al. 1998). However, there are a variety of
circumstances that can lead to extra peaks in an electropherograrn besides the presence of
multiple individual’s DNA (Clayton et al. 1998). Stutter is the most common cause for
extra peaks, which is caused by slippage of the polymerase during replication of the STR.

Stutter of tetranucleotide STRs results in peaks that are four nucleotides smaller (most

12

common) or larger than the true allele. The height of stutter peaks is usually less than
15% of the height of the main peak, however, when working with LCN DNA stutter
peaks have even been reported to be larger than the peak from the true allele (Gill et al.
2000). ‘N’ peaks, which are generated when T aq polymerase adds an adenine residue to
the terminal end of a newly synthesized DNA molecule during PCR, are another source
of additional peaks. Incomplete addition of the nucleotide gives rise to peaks that are
separated by one base, however, both peaks represent the same allele. A further cause of
extra peaks stems from non-specific priming of the DNA. Such artifacts can usually be
easily distinguished as they tend to be off ladder, are not reproducible, and have low
intensity. Sporadic contamination also results in superfluous peaks and is exacerbated
when working with LCN DNA. The additional alleles produced have been denoted
‘drop- in’.

It is crucial to account for sources of additional peaks when analyzing STR data to
ensure the proper labeling of a sample as a mixture or non-mixture. Mixture analysis can
be a tedious and time consuming task. To help streamline the process Clayton et al.
(1998) proposed six steps that should be implemented when interpreting mixtures.
Identify the presence of a mixture.

Designate allele peaks
Identify the number of potential contributors
Estimate the relative ratio of individuals contributing to the mixture

Consider all possible genotype combinations
Compare reference samples

P‘V‘PP’PT‘

Additionally, if mixtures are anticipated, steps can be taken to reduce their likelihood
during the processing of the evidence. For example, it has been proposed that portions of
evidence be swabbed independently (reviewed by Wickenheiser 2002), which reduces the

chances of obtaining mixtures and increases the likelihood of deciphering the profiles of

13

the perpetrator(s) and/or victim(s). This methodology aided in the recovery of
perpetrator profiles from a hot dog used in a sexual assault and an electrical cord used in
a strangulation (Wickenheiser 2002).

A study involving mixture analysis from trace DNA produced an interesting
finding that has proved beneficial in the interpretation of mixtures. Van Oorschot and
Jones (1997) showed that when swabbing a surface that had been touched by two people,
a mixture was obtained, but the major profile was that of the second person. This finding
was supported by an observation made in a criminal case where a swab taken from the
steering wheel of the victim’s car produced a mixture of the victim’s and suspect’s DNA,
with the suspect being the major contributor (Wickenheiser 2002). However, despite
success in analyzing mixtures from trace DNA, caution is still urged. When quantities of
DNA approach LCN status, mixture analysis is increasingly difficult due to the

characteristics of LCN DNA outlined above.

Study aims: A new approach to identzfiing those involved in IED campaigns

Past approaches aimed at obtaining handler DNA profiles from deflagrated pipe
bombs have been moderately successful, but they have lacked the discriminatory power
desirable from forensic DNA analyses. To improve the success of identifying those
responsible for IEDs, a new approach was sought. In discussions held with First
Lieutenant Shawn Stallworth of the Michigan State Police Bomb Squad, he noted that
IEDs the Bomb Squad encounters are often concealed in some type of container (personal
communication). It was thought that obtaining handler DNA profiles could be enhanced

if DNA recovery was targeted on these containers, and four reasons were hypothesized as

14

to why this might be a better alternative. First, a perpetrator is likely to have extended
contact with the IED container, allowing ample opportunity to deposit shed epithelial
cells. Second, based on work done by Kisilevsky and Wickenheiser (1999), porous
surfaces of IED containers may retain shed cells better than the smoother surfaces of pipe
bombs. Third, DNA on the container might be less degraded than the DNA on bomb
components leading to improved amplification. Finally, as opposed to the small size of
IED fragments recovered in previous studies (Gehring 2004; Kremer and Foran 2008),
fragments of the container might to be larger and easier to collect.

Potential complications that needed to be considered included the possibility of
encountering LCN DNA, mixtures, degraded DNA, and PCR inhibition. LCN
difficulties can potentially be lessened by implementing LCN procedures described by
Budowle et al. (2001; detailed above). The chance of mixtures can be reduced by
independently swabbing multiple areas on a container instead of using a single swab for
all regions. Processing multiple swabs from the same IED container can also be viewed
as running samples in ‘replicates’, fiilfilling a recommendation for working with LCN
DNA. Degraded DNA and PCR inhibition can potentially be addressed by using Applied
Biosystem’s MinifilerTM kit. The kit was specifically designed for amplification and
typing of degraded DNA, while the buffer (proprietary) was formulated to help overcome
common PCR inhibitors such as heme and indigo (Applied Biosystems 2006).

In the research presented, study participants were asked to use backpacks in
everyday activities for a period of one to two weeks. Backpacks were collected and
served as containers for pipe bombs that were deflagrated in a controlled environment by

members of the Michigan State Police Bomb Squad. All fragments of the backpack were

15

brought to the Michigan State University Forensic Biology Laboratory for DNA
isolation. DNA was extracted from swabs and subsequently amplified, quantified, and
typed. Handler genotypes were predicted blindly and checked for concordance with

participant’s known STR profiles to see if the correct handlers of the backpacks could be

identified.

16

MATERIALS AND METHODS

Backpack preparation and distribution

Ten backpacks (LEED’S, Pittsburgh, PA) designated BP3 — BP12 (BP1 and BP2
were used in preliminary studies for feasibility and optimization purposes) were
autoclaved for 45 minutes at 135°C, followed by a 45 minute drying time at 100°C. The
backpacks were placed in a Spectrolinker XL—1500 UV Crosslinker (Spectronics
Corporation, Westbury, NY) for 15 minutes per side (approximately 7.5 J/cmz). Plastic
shaft, cotton swabs (860PPC, Puritan Medical Products Co. LLC, Guilford, ME) were
used to swab backpacks in areas that were targeted for post-blast DNA recovery. The
double swab technique (Sweet et al. 1997) was incorporated in which a swab moistened
with 150 uL of digestion buffer (20mM Tris, 50mM EDTA, 0.1% SDS, pH 7.5) was
thoroughly passed over the targeted region, followed by a dry swab that was immediately
applied to the same section. Eleven areas were swabbed (Figure 1) including the five
zippers (comprised of the base, metal pull, string, and plastic tab), the top handle, the left
and right strap, the neck region, the front middle region, and the front tab. Swabs were
labeled with the backpack number followed by an identifier for the location swabbed
(Table 2). These swabs served as controls and were stored at -20°C, however, the need

did not arise for analysis.

17

Figure 1. Regions of the backpacks targeted for DNA collection.

  

JOZ ' SO'ZO

 

STRING

 

Eleven areas of the backpacks were targeted for post-blast DNA recovery. Identifiers are
detailed in Table 2. (A) Front view showing the five zippers, top handle, front tab, and
front middle region. (B) Rear view showing the left and right straps and the neck region.
(C) An intact zipper with all four components labeled.

Table 2. Identifiers for labeling swabs;

 

Swab location Identifier

 

Zipper 1 Z1
Zipper 2 Z2
Zipper 3 Z3
Zipper 4 Z4
Zipper 5 ZS
Top handle TH
Lefi strap LS
Right strap RS
Neck region N
Front middle FM
Front tab FT

 

Swabs were labeled with the backpack number followed by the identifier. Swab
locations are illustrated in Figure 1.

Backpacks 3 — 10 were randomly distributed to eight participants who used them
in everyday activities for a period of 11 days. BPll acted as a positive control in which a
participant handled the eleven areas in Table 2 three times a day for three days,
alternating the order in which the regions were handled each time. BP12 served as a
negative control. The use of human subjects followed guidelines established by the

University Committee on Research Involving Human Subjects (IRB # 07-577).

Pipe bomb preparation

Ten pipe bombs (five galvanized steel and five PVC) were assembled. Bombs
were 1 foot in length, one inch in diameter, and had a pair of end caps, one of which had
a ‘A inch hole drilled in the center for fuse placement. Pipes and end caps were soaked
for 1 hour in 10% bleach, rinsed with distilled water, and were placed in a UV crosslinker
for 10 minutes, turning half way through. ELIMINase® (Decon Laboratories, Inc., Bryn

Mawr, PA) was applied to all surfaces according to the manufacturer’s instructions and

19

rinsed twice with sterile water. Pieces were dried in a laminar flow hood. End caps
without the hole were affixed to PVC pipe bombs using PVC cement. Steel end caps
were not fastened. Pipe bombs were individually placed in new paper bags. Steel
bombs were assigned numbers 3 — 7, while PVC bombs were assigned numbers 8 — 12,

corresponding to the numbers given to the backpacks.

Backpack collection and bomb handling

Upon collection of the backpacks, participants were asked to close all the zippers
so the tabs were to the left when looking at the backpack from the front. The eight
participants were randomly assigned a letter (A, B, E — J) that was not known by the main
investigator. A second individual recorded the backpack numbers of the participants.
Subjects were asked to mock assemble, for 30 seconds, the pipe bomb that corresponded
to their backpack number. Participants who handled the steel pipe bombs were instructed
to securely fasten the end caps without the hole. Buccal swabs were obtained from the

participants for DNA reference samples.

Bomb deflagration and collection

Backpacks and pipe bombs were transported to the Lansing Fire Fighting
Training Facility (Lansing, MI), and deflagrations were conducted in the facility’s smoke
room. Immediately preceding deflagration a member of the Michigan State Police Bomb
Squad filled the pipes with 1.5 ounces of Green Dot Smokeless Shotshell Powder (Alliant
Powder Co., Radford, VA) and affixed the end cap with a hole to the device. A firse was

inserted in the hole and the pipe bomb was placed inside the main pocket of the

20

corresponding backpack with only the fuse showing. The same bomb squad member
placed the backpack bomb inside a steel crate (Figure 2) and lit the fuse via the circular
hole in the front of the crate. After deflagration, bomb and backpack fragments were
collected and placed together in a new paper bag. Between deflagrations the steel crate
and the smoke room were swept to remove any uncollected debris. Upon returning to the
lab, pipe bomb fragments were separated from the backpack fragments for use in other
studies. All investigators involved in the deflagration process were sleeves, facemasks,

and gloves.

Figure 2. Crate used to retain backpack and IED fragments.

 

The crate was designed to limit the dispersal of IED and bomb fragments. Walls were
constructed of steel with holes cut in them to relieve pressure from the blast; the floor
was made of wood.

DNA recovery and extraction

Backpacks were processed separately to minimize the chance of cross

contamination. Areas targeted for DNA recovery were individually swabbed in a laminar

21

flow hood that was thoroughly wiped down with 10% bleach and UV irradiated for 10
minutes. Potential DNA contamination of the swabs themselves (860PPC, the initial
swabs utilized), discovered during the research, resulted in backpacks 3 — 6, 8, and 9
being swabbed with different cotton swabs (25-806 2PC, Puritan Medical Products Co.).
These were designated in the same manner as control swabs (above), except that a ‘P’
was added to the end. Also, if the original location of a zipper was unidentifiable, a ‘U’
was incorporated (e.g., the first unidentifiable zipper from BP3 was labeled 3ZlUP).
Both swabs (wet and dry) were placed in the same 1.5 mL tube and stored at -20°C until
extractions were performed. A reagent blank consisting of two unused swabs was also
created (designated by backpack number and the identifier RBSWAB).

DNA extractions were performed by adding 350 uL of digestion buffer (total
volume of 500 uL including the 150 pL previously added to the swabs) and 6 11L of
proteinase K (20 mg/mL) to tubes containing the swabs, which were vortexed and
incubated overnight at 55°C. A second set of reagent blanks was initiated (designated by
backpack number and identifier RB). After incubation, spin baskets were inserted into
2.0 mL tubes. One swab per extraction was centrifuged for 1 minute at 13,000
revolutions per minute (rpm) and discarded. The process was repeated with the second
swabs using the same spin baskets, after which the baskets were removed and the liquid
was pipetted back into the original tubes. An equal volume of phenol (500 uL) was
added to the samples, vortexed, and centrifuged at 13,000 rpm for 6 minutes. The
aqueous layers were transferred to new 1.5 mL tubes and equal volumes of chloroform
were added. The tubes were vortexed and centrifuged for 6 minutes at 13,000 rpm. The

aqueous layers were transferred to Microcon® YM-30 spin columns (Millipore

22

Corporation, Billerica, MA), and 100 11L of TE (10 mM Tris, 1 mM EDTA, pH 7.5) were
added. The columns were centrifuged for 12 minutes at 14,000 x g and the flowthrough
was discarded. The columns were washed with 200 11L TE, with the centrifugation time
reduced to 8 minutes. Twenty microliters of TE were added to the column membranes
and left for 5 minutes. Columns were inverted into new tubes and centrifuged for 3
minutes at 1000 x g. DNA samples were stored at -20°C. Reference buccal swabs were
extracted using a ChargeSwitch® Forensic DNA Purification Kit (Invitrogen, Carlsbad,

CA) as per the manufacturer’s protocol, and stored at -20°C.

Amplification of miniST Rs using the MirrifilerTM kit

DNAs extracted from the backpack swabs were amplified using an AmpFlSTR®
MiniFilerTM PCR Amplification Kit (Applied Biosystems). Reactions were carried out in
10 pL volumes, including 2 11L of the MiniFilerTM primer set, 4 11L of the MiniFilerTM
Master Mix, and 4 11L of DNA template. Reactions of reference samples contained 1 uL
of DNA diluted 1:100 in TE, and 3 uL TE, while positive controls had 3 11L of 007
control DNA (0.1 ng/uL) and 1 uL of TE. All reagents were briefly vortexed and
centrifuged before use. Amplifications were performed using the following thermal
cycling conditions: an 11 minute incubation at 95°C, followed by 30 cycles consisting of
a 20 second denaturation at 94°C, 2 minutes of annealing at 59°C, and a 1 minute

extension at 72°C, followed by a 45 minute final extension at 60°C.

23

Capillary electrophoresis and miniS T R analysis

Two microliters of amplified product (1.5 11L for reference samples) were
combined with 24.5 11L of deionized forrnamide and 0.5 11L of GeneScanTM 500 LIZ®
Size Standard (Applied Biosystems) in 0.5 mL tubes. Allelic ladders contained the same
volumes of forrnarnide and size standard in addition to 1.5 11L of the MinifilerTM allelic
ladder DNA. Tubes were incubated at 95°C for 3 minutes and then placed on ice. Caps
were cut off and one drop of mineral oil was added.

Electrophoresis was performed on an ABI PRISM® 310 Genetic Analyzer
(Applied Biosystems) using the GS STR POP4 (lml) G5 v2.md5 run module. Runs were
conducted using POP4 (performance optimized polymer 4; Applied Biosystems) and 1X
buffer with EDTA (Applied Biosystems). Parameters included a 5 second injection at 15
kV, a 28 minute run time at 15 kV, and a temperature of 60°C. Data were analyzed using
GeneMapper® ID software v3.2.1 (Applied Biosystems). Panels and bins were
downloaded and imported into GeneMapper fiom www.appliedbiosystems.com (Support
> Software Downloads > GeneMapper® ID Software v3.2 >Updaters & Patches). The
analysis method was MiniFiler__G8500_HID_v1, the panel was MiniFiler_G8500_v1,
and the matrix was DS-33 Matrix 7-12-07.

Electropherograrns were manually reviewed and callable alleles were recorded. A
threshold of 50 relative fluorescence units (RF Us) was used. Handler profiles were
compiled blindly using the complete set of swabs from a backpack, and checked for
concordance with the known (buccal) swabs by a second individual.

After the handler’s profiles were known, results for each locus were placed in one

of six categories:

24

p—a

. The locus contained only the handler’s correct alleles.

2. The locus had multiple allele calls, but the handler’s alleles constituted the
major profile.

3. The locus had multiple allele calls, but the handler’s alleles could not be
distinguished as the major profile.

4. The locus had at least one of the handler’s correct alleles. This involved
instances where there may have only been one allele called, or there may have
been multiple alleles called, but only one matched the handler’s profile.

5. None of the alleles matched the handler’s profile.

6. No alleles were called at the locus.

The number of loci from each swab that contained the handler’s correct alleles
was counted; a locus met this criterion if it was in one of the first three categories
described above. Swabs were then classified as producing all 9, 8 or 7, 6 or 5, or less
than five loci with the handler’s correct alleles. The number of swabs in each category
was compared among backpacks. Additionally, the average number of loci with the
handler’s correct alleles was calculated per region swabbed. Loci in categories two or
three above (excluding amelogenin) were reviewed to see if increased stutter could have
accounted for the extra allele calls; if extra peaks were one repeat unit before or alter the
peaks attributed to the handler’s alleles, the locus met this condition. Finally,

electropherograms with six or more loci that had, allele calls in addition to the handler’s

were designated mixtures.

DNA quantification using real-time PCR

DNAs were quantified using a QuantifilerTM Human DNA Quantification Kit
(Applied Biosystems). Amplification and detection were performed on an iCyclerTM
thermal cycler and an iQ5 multi-color real-time PCR detection system (Bio-Rad,
Hercules, CA), respectively. Dye calibrations were completed for VIC and FAM as

instructed in the iQTMS Optical System Instruction Manual (Bio-Rad). The VIC

25

calibrator consisted of a labeled oligonucleotide (VIC-CATTTCCTTC) (Applied
Biosystems) diluted to 300 nM. The FAM solution was from the calibration kit (Bio-
Rad). Calibrations were performed using 15 uL volumes in 0.2 mL dome cap tubes (Dot
Scientific, Burton, MI) and 96 well, half skirted, Thermowell® Gold PCR plates
(Corning Inc., Corning, NY) sealed with microseal® ‘B’ film (Bio-Rad).

The 200 ng/ 11L human DNA stock solution supplied with the QuantifilerTM kit
was serially diluted as per the manufacturer’s protocol providing eight standards ranging
from 50 ng/uL to 0.023 ng/uL. PCR reactions were carried out in 15 uL volumes
consisting of 6.3 uL primer, 7.5 11L reaction mix, and 1.2 11L DNA. Reactions were set
up using 96 well plates and then sealed. Standards were run in duplicate and unknowns
in triplicate. Thermal cycling parameters included a 10 minute incubation at 95°C
followed by 40 cycles of 15 seconds at 95°C and 1 minute at 60°C. Real-time data were
analyzed using software generated thresholds for each fluorophore (PCR base line
subtracted curve fit mode). Threshold cycles (CT) for IPCs were reviewed and any values
above 30 were noted. The software created a standard curve used to determine DNA
quantities and calculated average DNA concentrations and standard deviations for all

replicates.

26

RESULTS

Deflagration observations

Backpacks serving as containers for steel pipe bombs suffered more damage than
those concealing PVC pipe bombs (Figure 3)—evident by an average of 4.7 zippers, or
pieces thereof, recovered from backpacks subjected to PVC pipe bombs as opposed to 2.8
for backpacks containing steel pipe bombs. The majority of backpack fragments (zippers
being the exception) were retained within the crate. Areas of the backpacks that were
stitched or made of stronger material, such as the top handle, withstood the blast better
than others. Backpack 8 briefly caught on fire after the pipe bomb was deflagrated, with

flames shaken out by a member of the bomb squad.

Figure 3. Backpack fragmentation.

   

Examples of post-blast backpacks after serving as containers for PVC (A) and steel (B)
pipe bombs.

27

Swabbing backpacks to recover shed cells

The number of regions swabbed per backpack ranged from seven on backpack 7
to eleven on backpack 10, with most of the variability caused by the failure to retain all
the zippers. Twenty-eight of the forty zippers were recovered, of which only four were
intact. The remaining 24 zippers included: seven that were missing the tab, four that
were lacking the base, six that were missing the base and tab, four that had only a tab,
and three consisting of only a string. The original locations were identified for 7 of the
28 zippers collected due to many becoming detached from the backpack during
deflagration. The fiont middle region of backpack 3 was the only non-zipper portion not

swabbed as it was unable to be identified.

Isolating DNA from recovered cells

A total of 75 swabs was obtained fiom the eight backpacks handled by
participants. However, DNA from swab 8F MP was not included in the study due to a
malfunction in the spin column during processing. Some swabs were soiled with residue
from the explosives resulting in discoloration of the digestion buffer after incubation,
most of which disappeared during the extraction process. Initially, a second chloroform
treatment was administered for those samples in which discoloration persisted (performed
on backpack 11 swabs), however, this was not beneficial and all subsequent samples

were subjected to only one chloroform treatment.

28

Analysis of ST R electropherograms

Handler profiles were compiled for backpacks 3 — 10 after electropherograms
from all the samples were reviewed and allele calls were made (Table 3; Appendix A).
Seven of eight profiles matched a participant’s reference sample at all nine loci (Table 3);
the lone exception was backpack 9, which matched a reference profile at eight loci. This
stemmed from the inability to distinguish between a 30, 31 and a 30 homozygote at the
D21 locus, where thirty was the sole allele in three samples, and 30 and 31 were present
in six samples. In only one instance were 30 and 31 the only two alleles present. The
correct call was later found to be 30, 31.

Reagent blanks that were processed with backpack 3 —- 10 produced a total of 7
callable alleles, with peaks all below 150 RFUs (Appendix A). The swab ‘reagent blank’
for backpack 10 had peaks matching alleles 12 and 48.2 at CSF and FGA, respectively, as
well as alleles X and Y at amelogenin, and a 10 called at CSF. The other two alleles in

reagent blanks came from backpacks 8 (12 at D13) and 4 (11 at D7).

29

Table 3. Consensus handler profiles obtained from backpacks and the study
participant to which they corresponded.

 

 

 

 

Backpack

Locus 3 4 5 6 7 8 9 10
D13 8,12 8,11 11,12 12 11,12 8,12 8,14 11,12
D7 10,11 9,10 11 10,11 8,12 8,11 8,10 9,12

Amel. X X X X X X X X
D2 23 ,24 20,25 23 ,24 20,24 20,21 20,21 17,20 19,25
D21 27,30 27,30 29,32 31,31.2 31,0L 28,30 * 29,30

D16 11,13 12,13 10 9,11 9,11 9,11 12,13 11
D18 15,17 13,16 16 13,14 17 15,16 13,16 13,16
CSF 11 12,13 11,12 11,12 10,11 11,13 12 12,13
FGA 21,24 21,25 21 20,25 22,23 .2 18,24 20,22 22.23

Participant B E G I J F A H

 

A consensus handler profile was determined for each backpack using allele calls from the
complete set of swabs (see Appendix A). *—could not distinguish between a 30
homozygote and a 30, 31. (OL)——the off ladder allele was 33.1.

The ten swabs from positive control backpack 11 led to a profile that matched the
handler despite two samples (11Z2UP and 11Z4UP) experiencing a high level of drop-
out (Appendix A). Negative control backpack 12 had nine swabs that were analyzed.
One full profile (nine loci) was obtained from sample IZZlUP that could not be matched
to anyone directly involved in the study (researchers or participants). The rest of the

samples from backpack 12 produced a total of ten callable alleles, all of low intensity

(Appendix A).

Categorizing loci based on the presence of handler alleles
All 666 loci reviewed were placed in one of six categories (see Materials and
Methods) based upon the extent to which they contained the handler’s alleles (Figure 4;

Appendix A). Examples of four categories can be found in Figure 5. Five hundred and

30

ninety-one loci (88.7%) contained the handler’s actual alleles. Of these, 295 (49.9%) had
callable alleles besides those of the handler. One of the handler’s alleles (for
heterozygotes) was present at 41 loci. For 27 of these the correct allele was the only one
present, while 14 had additional, incorrect alleles. Thirty-one loci did not have any
callable alleles, of which 16 occurred at the D7 locus. Three loci contained only alleles

that did not match the handler’s.

Figure 4. Percentage of loci placed in each of the six locus classifications.

.9
.3;
4:.
\r

:0
, .
9

El Only handler's alleles were
present

           
 
    
    

     
  
 

.fi
0 :0.
o o 9
020203

- ‘v
'0

..

  

:‘
.9
O
.0.

‘0
.9
O
O
.0

 
 
 

v.’
0
6
A

.9

.6

  
 
   
 
 

O
o o
20

.9
A.

Handler‘s alleles constituted
major profile

     

 

           

,
9
0’0
'9
o

O.
at:

O

12.6

A

   

I Handler's alleles present, but
did not constitute major profile

IE Only one of the the handler's
alleles was present

Alleles called, but none
matched the handlers

a No alleles called
The first three categories represent loci in which the handler’s alleles were present in
their entirety (88.7%). The second and third categories indicate loci that had callable

alleles besides the handlers (44.3%). The last three categories represent the lack of
recovery of handler alleles (11.3%).

31

Figure 5. Examples of four locus classifications from backpack 6 samples.

ILL; iva, ,i, ,# v_; E ,,V v ’v,v,__,,,,,

 

 

 

.60 1m in m 21 150 In) no 1') II) 2

13m A 1‘” B

 

 

 

 

 

 

 

 

 

m: rfi‘itt’dII-I IHA‘LJJVI Iv-””*W:*" m: C‘ZETEM I». ~’\WI:\~"IIJ I”; ‘**""“e

 

5
E
s
a
s
3
1
5
s
E
a
a
8
J

 

 

 

 

 

 

,........r+.rw...ltl Inhaler. .,..w~v-..-t :WMW WWWWWN

Ii

FGA locus from four backpack 6 samples. The handler’s correct alleles are 20 and 25
(shaded). (A) Only handler’s alleles present (B) Handler’s alleles as a major profile. (C)
Handler’s alleles present, but did not constitute a major profile. (D) Only one of the
handler’s alleles present (drop-out). The other two categories (alleles called but none
matched the handler’s, and no callable alleles) represent cases of complete drop-out (not
shown).

32

Swabs were categorized based upon the number of loci that contained the
handler’s alleles, and then further analyzed on a per backpack basis. Forty-six of 74
swabs had the handler’s alleles at all 9 loci including ten from backpack 4 and nine from
backpack 8 (Figure 6). Backpack 6 had the lowest number of swabs with the handler’s
alleles at all loci (three), but four swabs had the correct alleles at seven or eight loci.
Eleven swabs produced the handler’s alleles at six or fewer loci, with seven coming from

backpacks 9 and 10.

Figure 6. Number of loci containing handler’s alleles from backpack swabs.

 

 

 

 

 

 

10
Nurrberof
8 lociwith
é 11311116le
5 6 alleles
“5 I9
3 4 B7or8
E 2 BSor6
EI<5
0

 

 

3 4 5 6 7 8 9 10
Backpack number

The majority of swabs (63) had the handler’s alleles present at a minimum of seven loci
(black and gray bars). The highest percentage of swabs with the correct handler alleles at
all loci came from backpacks 4 (100%) and 8 (90%). Eleven swabs had the handler’s
alleles at six or fewer loci (striped and white bars) with seven coming from backpacks 9
and 10.

33

Peaks attributed to increased stutter and the occurrence of mixtures

Eighty-two of 193 loci (42.5%) classified as containing the handler’s alleles as the
major profile (excluding amelogenin) had their extra peaks at positions indicative of
increased stutter. The same analysis performed on loci classified as containing the
handler’s alleles, but not as the major profile, found that 36 of 79 (49.4%) loci had their
extra peaks at stutter locations. Eighteen samples were classified as mixtures based on at
least six loci containing more than two alleles (Appendix A). The most mixtures came

from backpacks 6 and 8 as they had four and five, respectively.

Comparing regions based on the number of loci with the handler ’s alleles

The top handle proved to be the most effective area for recovering a handler’s
alleles as all swabs produced the handler’s full profile (Table 4). Other non-zipper
regions ranged from an average of 8.88 loci (neck region) to 7 loci (fi'ont tab). Overall,
swabs of the zippers averaged the handler’s alleles at 7.35 loci, with intact zippers

producing 8.75 loci and zippers missing both the base and the tab having 5.3 loci.

34

Table 4. Number of loci with handler alleles broken down by backpack region.

 

Number of loci with handler alleles

 

 

 

Swab location 9 7 or 8 5 or 6 <5 Average
Zippers
Intact 3 1 - - 8.75
no tab 3 2 2 - 7.43
only tab 2 2 - - 8
only string 1 2 - - 7.67
no base 3 - l - 8
no base or tab 2 1 1 2 5.3
Top handle 8 - - 9
Left strap 6 2 - — 8.75
Right strap 4 2 1 1 7.75
Neck region 7 1 - - 8.88
Front middle 5 1 - - 8.83
Front tab 2 3 2 1 7
Total 46 17 7 4 7.97

 

The number of loci containing the handler’s alleles varied based upon the components of
the zippers that were recovered. Intact zippers averaged the most loci with the handler
alleles, while zippers missing the base and tab averaged the fewest. Less variation was
seen in swabs of non-zippers as the top handle averaged 9 loci with the handler’s alleles
while the front tab averaged the fewest (7).
DNA quantification

DNA quantities were determined by taking the average of three real-time PCR
replicates, and ranged from 1.41 pg/uL (lOFTP) to 1.25 ng/uL (4NP) (Table 5; Appendix
B). Fifty-three of 74 samples (71.6%) had DNA quantities less than 0.2 ng/uL, while 11
had quantities above 0.4 ng/uL with nine coming fiom backpacks 4, 7, and 8. The
majority of samples from these backpacks had DNA quantities above 100 pg/uL, with
only three samples below this value (7FTP, SZ4P, and 8FTP). In comparison, no samples
from backpacks 6, 9, and 10 had a DNA quantity in excess of 100 pg/ 11L.

The top handle, straps, and neck region had the highest average DNA quantities

(approximately 300 pg/uL), while the front tab averaged the lowest of the non-zipper

35

regions (66.9 pg/ 11L). DNA amounts recovered from zippers averaged 91.3 pg/uL. The
most DNA was obtained from zippers missing just a base (162.3 pg/uL) and having only
a tab (150.5 pg/uL), while swabbing merely the string produced 41.0 pg/ 11L. The four

zippers that were recovered intact averaged 61.5 pg/ 11L.

36

.530 8: 03 AL $5082 0000525. :80 macaw com 8083.0 .5082 :05 :80 095. a 0%:
no: 20 #8963 e020 2: :20 mofio€fil+v .030: mm 530 80 800525 =< 088:0»: ooh: 0o 008020 20 2.3802 800530

 

 

 

0000.0 5 50.0 0x. 5.0 005.0 0000.0 0 500.0 005.0 000 .0 0000.0 0.70
50.0 005.0 0000.0 - 000.0 0000.0 00v0.0 05.0 - 02.0
000.0 0000.0 0000.0 50.0 00 v.0 500.0 005.0 000. _. 000.0 AZ
50.0 0000.0 0000.0 E00 0 P00 0000.0 005.0 000. _. 0 5.0 00%
000.0 0000.0 «000.0 000.0 00 9.0 55.0 00 _..0 000.0 00.0 mm..—
000.0 500.0 0000.0 000.0 000.0 0000.0 .000 000.0 0 _.0.0 mm:
:Fd 005.0 - 000.0 - - - - 0000 “EN

0000.0 0000.0 0000.0 - - - - - 00 _..0 EN

0000.0 0000.0 - - - - - - - mmN

.. - - 0 50.0 - - - _.00.0 - mDvN

r - 00000.0 03.0 - - - 03.0 005.0 mDmN

.. 00 5.0 00 5.0 000.0 - 00000.0 00000.0 000.0 0000.0 mDNN

a 0000.0 005.0 00 P0 5rd 00000.0 00000.0 00.20 500.0 03 _N

swag... S a w n e m a. m exam
xommxoem

 

60.0.5. zone—25 :8...— 03332 <75 he SEES—O .0 03:.

37

Amounts of DNA added to PCR reactions

Once DNA quantities per uL were known the values were multiplied by four to
obtain the amount of DNA that was added to PCR reactions. Five of 74 reactions had
DNA template quantities that were within the recommended range of 0.5 — 0.75 ng for
the MinifilerTM kit (Applied Biosystems 2006a) (Table 6; Figure 7). The majority of

reactions (45) had input amounts below 0.5 ng.

Table 6. DNA quantity added to PCR reactions.

 

 

Total DNA added to .
. Number of reactrons
PCR reactrons
<.1* 17
0.1 — 0.3 23
.0.3 — 0.5 5
0.5 — 0.75** 5
0.75 - 1 6
1 — 1.5 6
>15 12

 

DNA quantities are given in ng. Total DNA amounts were obtained by multiplying the
amount of DNA per sample by the number of microliters added to the reactions (4).
*— indicates low copy number. **— indicates the recommended range of input DNA.

Seventeen reactions were designated as LCN, accounting for 49 of the 75 (65.3%)

loci that were categorized as having complete or partial drop-out. Twenty-one of the
remaining 26 cases of drop-out were from reactions in which DNA input averaged 169.6

pg. Three LCN samples did not experience drop-out, producing full handler profiles.

Twelve reactions contained more than 1.5 ng of input DNA, or 2 — 3 times greater

than the advised amount. Ten of these were from backpacks 4 and 8. Electropherograms

from all 12 reactions produced the handler’s profile. Excluding amelogenin, greater than

half (55.2%) of the loci had more than two callable alleles.

38

Full profiles were obtained fiom 42 of the 69 (60.9%) PCR reactions where input
DNA was outside the recommended range. However, as the amount of DNA added to
the reactions deviated from this range electropherogram quality was compromised.
Electropherograms fiom reactions with less than the suggested quantity of DNA were
characterized by increased stutter, allele drop-out/drop-in, heterozygote peak imbalance,
and peaks of low intensity (e. g., Figure 8). Electropherograms produced from reactions
in which too much DNA was added had features such as split peaks caused by

incomplete ‘A’ addition, baseline noise, pull-up, and increased stutter (e. g., Figure 9).

39

.00Q 00 00000.8 05 88 00088 .8088 800808 080088 8: :5 :38?

8.0 080 $0000 .8080 8 00080.80 00 03680 :00 £00 a 00 05 2Q 00 00 880800000020 80 8 020:0 860 83 0:0. 020:0
886.8008 80300 000 .0050; EM 0803 .3000 0000200 058 08 08808000800000 8000008 000 8 A0: 00.0 1 0.00 00000 83
<75 00 8:080 00058888 05 e003 .0008m 03 020:0 820003 800008 M00 08 8 00000 <75 00 m0 000 000 0D~Nv 20800

a a 8 m. E
4 l o
q 080
008
0 r r 8 r 8 00
En!

 

0v~ Dom Our 09. 0:. 000 60v Sr 00— 00..

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

do; _
. < I a. a. 1 0 .
00 _ d.
0 0.80 8...
000—.
008—.
3.0 000 000 00.. at. Sr one 30 Sr 000 0: 8w 8 00
E E E E
< 4‘s c
080
003
0.0 000 :9. cap a: 00? 00,—. Div 009 our 0:. our 8 8
E a _H_
4 ..
000
000'
Door
080
05 SN 00.. 00v 0:. 00.. 00v a: one 00— 0: 5.. 8 no

 

.0805 .80 <75 808 00 8880 00.808838 05 08:80.80.— .0Dva 880 88082—08805 .5 0.8000

40

.8 000 0 08:00.0 8 m030> 3.00 3038

0008008 hob 03003 80 8.000 8 088.000 0000 0>00 200 > 80 no 00:0; Dumm— 80 .80 >000 008 000000 8000 008 00000 0000 802
0000888 0000 32 000 Ann— 800: 80080 00000 .ANQ 8000 00000080 0000 800008800 .90 00:80 0000808 .3 0000808000
803 0805 20..— 880 08000800080005 000000 80 000000 00000000 800008 000 05 8 00000 <75 00 00 0.00 000 0mNm 0.0800

«8 “a E E <

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

.
* 9'
80
8Nv
80v
DVN 8N car 00.. 8w ONr 0:. Our 00 8
— E
4t u 4: < 1‘ ¢ j( < Ii 1 11 111‘ d4< a
8
Our
000
OHM
o—N 8N 8r our our our 0: 00.. on 8
_ "ban—a ooh-Una
mm H
c
8‘
02001052.: 08
0000 8:
arN 8N car car our 90—. c: 8r ONr a: Oar on 8
nun—E _
4.. <1 k c
8N
80-080 .
8
SN 8N 8r our 05v 0!. am- Qwr Dar our a: car 8 8
_ E5 _\ a

 

080.3 2010 088009.00.— .0mNm 880 88080008000E .0 0805

41

.Ai 00080 80000000 000 A800 00 00 0000?

000 220 829 P00 002 a? 308 .0220 88 2 05800 as: 2.082 .50 8-0.0 .000 0.08 0% .8 .305 83205
.3 0000808000 803 08000800080005 000000 80 02000 0000000.— .000008 ~50 05 9 00000 <ZQ 0o 00 mud 000 00100 20800
E

  

 

0000000

 

 

.NN a vN 8N on v 8 r on v

 

 

0000
0000000

 

 

 

 

O 8v 2r

3
nun—B—

 

 

 

 

.NN are. 8N Oar Br Br 8r our 8—. car DNr 0:. our on Do

F a _ a

 

 

80:000.. mom 05 3 00000 0800 <75 00:8 08 088000000.— .00A0 8000 88080000500E .0 0.8000—

42

Detection of PCR inhibition via the internal positive control

The majority of IPCs (all but 5) had threshold cycle values within 20 — 30,
indicating that the PCR reactions were minimally or not inhibited. The [PC threshold
cycle for 3RSP replicates 2 (37.31) and 3 (threshold not reached in 40 cycles) were
outside of this range, and replicate 1 also showed poor amplification, indicating partial
inhibition (Figure 10). The IPCs failed to amplify for all three replicates from 7RSP,

however, human DNA amplification was detected.

Figure 10. Quantitation results for 3RSP.

Amplificdion Chart: Data 44-08 an 3andS revised CT opd

§

§

PCR Base Line Subtracted Curve Fit RFU
§ § § §

N
O
O

O

 

Cycle

Amplification of human DNA and [PCs for replicates of 3RSP. The threshold for the
DNA ( 149.30) is represented by the top horizontal line (*) and the IPC threshold (45.21)
is represented by the lower horizontal line (A). Human DNA amplification is represented
by the top three lines (1, 2, and 3), and the corresponding [PCs are denoted 1a, 2a, and 3a.
All replicates show signs of PCR inhibition.

43

DISCUSSION

The goal of this study was to investigate the feasibility of recovering DNA from
post-blast IED containers, and determine if DNA typing using miniSTRs could be used to
correctly identify the handler. This approach stemmed fiom earlier research aimed at
recovering and typing DNA fi'om IED components themselves. Although a direct
comparison to previous studies cannot be made due to different targets of DNA recovery,
the results showed an improvement in the ability to correctly identify the handler.

Correct handler profiles were obtained at all nine loci for seven of the eight backpacks,
with the remaining backpack having a single ambiguous allele.

Discriminatory power was also enhanced compared to previous studies. The most
successful study to date that identified handlers fi'om recovered IED components used
mtDNA (Gehring 2004), but mtDNA polymorphisms are linked and passed along
maternal lineages, which limits statements concerning uniqueness. On the other hand,
the selected STR loci sort independently, allowing individual allele fiequencies to be
multiplied, resulting in random match probabilities that often exceed one in a trillion.

The MinifilerTM kit is reported to have a probability of identity of 8.21 x 10'11 for US.
Caucasians (Applied Biosystems 2007a), which can be further improved if supplemented
with results from standard STR kits. IdentifilerTM, a 16 locus multiplex from Applied
Biosystems, has a probability of identity of 6.10 x 10'8 for US. Caucasians (excluding the
eight STR loci in MinifilerTM). A full profile created using both kits gives a probability

of identity of 5.01 x 10'18 (Applied Biosystems 2006), representative of the

44

discrimination that is desired fiom forensic analyses and an immense improvement over
mtDNA analyses.

There are a variety of possible reasons as to why more success was achieved in
recovering and typing DNA from containers rather than the IEDs themselves, the main
ones of which were portrayed in four hypotheses formulated prior to the research. First,
it was proposed that the perpetrator (handler) would have extended contact with the
container, with ample opportunities to deposit shed cells. All participants who carried the
backpacks indicated that they used them for 7 - 10 days. However, depositing cells also
depends on an individuals’ shedder status, as was demonstrated in studies performed by
Lowe et a1. (2002). Quantifying the DNA from backpack samples revealed that
participants who handled backpacks 3, 4, 7, and 8 were likely better shedders as these
samples consistently had DNA quantities in the hundreds of picograms across all
locations, while the other backpacks had lower values. Despite indications that some
handlers were poor shedders, the period of handling proved sufficient to deposit cells, as
every area swabbed resulted in at least a partial profile of the handler. Based on the
results, it is likely that possessing the backpacks for fewer days would have been
adequate to successfully identify the handler.

The second hypothesis was that the more textured surfaces of the backpacks
would retain shed cells better than the smoother surfaces of the pipe bombs. Critical to
all studies involving touch DNA is the transfer of cells from the handler to the object, and
their persistence. Cells on smoother surfaces, such as pipe bombs, are likely to adhere
loosely, making them more susceptible to removal upon subsequent contact. On the

other hand, cells deposited on more textured substrates are retained better (Kisilevsky and

45

Wickenheiser 1999). All backpack regions targeted for DNA recovery were more
textured than smooth except for the zipper tabs. The improved success in recovering
DNA from backpacks compared to pipe bombs suggests that collection efforts should
first focus on textured surfaces when conducting analyses involving touch DNA.
However, caution should be taken as it may result in more mixtures as cells from
previous handlers are also likely retained. This idea is supported by the observation that
of the six zipper samples that were classified as mixtures, none were from those in which
only the tabs were swabbed, but instead stemmed fi'om zippers in which the textured
strings were recovered and swabbed,. The possibility that smoother surfaces might not
retain cells well could potentially be advantageous based on the thought that cells from
previous handlers would be removed and only those of the last person to touch the area
would remain.

The third hypothesis was that DNA on the containers would be less degraded than
that recovered from IEDs. An assay was not performed that specifically addressed DNA
degradation, however, results from this study were vastly improved compared to previous
research in which STR analyses were performed on DNA recovered from pipe bombs
(Esslinger et al. 2004; Kremer and F oran 2008). This improvement may have resulted
from DNA on the containers being less degraded, the use of miniSTRs, or a combination
of both. A reduction in DNA degradation is possible if the containers were not exposed
to the same extreme temperatures as the pipes, thus better preserving the integrity of the
DNA. However, it is unlikely that DNA on the containers did not experience any
degradation, in which case the incorporation of miniSTRs may have been instrumental in

obtaining better results, as they are specifically designed to amplify degraded DNA. An

46

assessment of DNA degradation could be made by analyzing backpack samples with a
conventional STR kit, such as Profiler PlusTM used by Esslinger et al. (2004). If full
handler profiles were still obtained it would support the hypothesis that DNA fiom the
backpacks is less degraded, as Esslinger et a1. (2004) only obtained one full profile.

The final hypothesis was that pieces of the container would be larger than the
IED fragments recovered in previous studies, making them easier to collect and swab.
This was substantiated by the fact that of the non-zipper regions targeted for swabbing
only one was not recovered. Zippers were not retrieved with as much success since they
were susceptible to being dismantled during deflagration, however, 70% of zippers were
still collected. The ease of recovery enabled multiple areas of a backpack to be swabbed,
which served as ‘replicate’ analyses. Rather than processing a single swab from a
backpack, as was done with pipe bombs in previous studies, multiple swabs from
different backpack regions were processed. Confidence was gained in calling handler
profiles when identical alleles repeatedly showed up at a locus. It seems likely that
identifying handlers would have been less successful if only one swab was taken from the
backpacks due to difficulties in differentiating handler alleles from peaks caused by
mixtures and artifacts. Also, instances of allele drop-out would have been hard to assess
with only a single swab.

Recovering DNA from multiple areas of a backpack requires making a decision as
to how many samples need to be analyzed in order to confidently call the handler’s
profile. Eleven regions were targeted for DNA recovery, but all eleven areas were
swabbed for only two backpacks due to the inability to recover some regions. However,

it was found that calling a handler’s profile depended more on the quality of the

47

electropherograms than on the number of regions swabbed. For instance, seven samples
were obtained from backpack 7, but the handler’s alleles were readily determined
because five electropherograms had full, clean, profiles. On the other hand, ten samples
were not sufficient to call the handler’s alleles at the D21 locus of backpack 9 due to
allele drop-out in four samples and at least three callable alleles in five. If
electropherograms are of good quality, swabbing three to five regions may be sufficient
to confidently call the handler’s profile, and additional areas can be swabbed and
analyzed as needed.

The quality of electropherograms was often compromised by allelic drop-out,
which in most cases was attributed to low levels of DNA being added to the PCR
reactions, as 24 of 27 (88.9%) experiencing drop-out had DNA quantities below 275 pg.
The three reactions that experienced drop-out with more than 275 pg of DNA (3NP,
5LSP, and 7RSP) had a total of five loci with drop-out, two each at D16 and D7, and one
at FGA which was attributed to a split speak preventing the 50 RFU threshold from being
obtained. Drop-out at three loci in sample 7RSP likely stemmed from partial inhibition
as the IPC failed to amplify in all three replicates. However, as with degraded DNA,
larger loci are more susceptible to drop-out stemming from PCR inhibition (Bessetti
2007) which does not fit the pattern of drop-out that was observed. Locus D7 is just the
fourth largest in the Minifiler'rM kit and only amelogenin is smaller than D16 based on
size of the largest possible alleles. Of all the loci that had drop-out of the handler’s
alleles, 29.3% came from D7 and 13.3% were from D16. FGA, D21, and D18, the three
largest loci in the kit, contributed to 13.3%, 12%, and 5.3% of total drop-out,

respectively.

48

To the author’s knowledge there has not been any information published
concerning differential amplification success among loci in the Minifiler“M kit. The
increased prevalence of drop-out at D7 compared to the other loci may be a consequence
of sub-optimal primer annealing conditions. Different primers have ideal conditions
(temperature, concentration, etc.) in which they achieve maximum amplification
efficiency. When a variety of primers are incorporated into multiplexes, optimal
conditions for individual primers are usually compromised (Butler 2005). An interesting
observation was that samples in which the D7 locus did not experience drop-out had
relatively low RFU values at the locus. This was especially evident in reactions that had
more than 1.5 ng of DNA added, where RFU values for alleles at D7 typically ranged
from 500 — 1000, while approaching 6000 RFUs for D13 (e.g., Figure 9).

Drop-out was more prevalent in samples with low quantities of DNA, although,
three LCN samples did not experience drop-out at any locus. Two of these were from
backpack 5 (5NP and SFTP) which was used by participant G, but this subject was
homozygous at five of the nine loci, meaning it might be more difficult to detect allele
drop-out. The other LCN sample, lOFMP, had 63.6 pg of DNA added to the PCR
reaction. However, as with the quantitation of other samples containing low amounts of
DNA, there is little confidence in the precision and accuracy of this quantity due to
stochastic sampling effects (i.e., there may have been more DNA). For example, DNA
amounts for 10F MP ranged from 0.0 pg/uL to 32.6 pg/uL (Appendix B). Pipetting low
quantities of DNA increases the chances of unequal sampling, and must be taken into
account when quantifying small amounts of DNA. Quantifying samples in triplicate and

calculating average DNA values is advised to circumvent this phenomenon.

49

Electropherogram quality was also affected by extra peaks as approximately 45%
of the loci contained callable alleles in addition to the handler’s. The majority of the extra
peaks were attributed to stutter and mixtures. The presence of a peak one repeat unit
before, and sometimes after, the true allele is indicative of stutter (e. g., Figure 7).
Software used to analyze electropherograms can be programmed to not call peaks from
this artifact, however, when input DNA deviates from the recommended range (0.5 —
0.75 ng for Minifiler), stutter peaks are known to increase (Applied Biosystems 2007a;
Budowle et a1. 2001), causing the software to call them. A rough assessment of the
prevalence of increased stutter showed that 118 of the 272 (43.4%) loci categorized as
having extra peaks (amelogenin was excluded because it is not an STR) could be re-
categorized as containing only the handler’s alleles after possible stutter was accounted
for. Many of these peaks were likely a result of increased stutter as peak heights were 15
— 20% that of the true allele. However, it is unlikely that peaks in all 118 loci were the
result of stutter. For example, sample 821 UP had three loci with extra peaks that could be
attributed to stutter, however, this reaction was classified as a mixture as the peaks were
one repeat unit greater than the true allele, a rare location for stutter.

Eighteen samples, including SZlUP, were classified as mixtures based on the
presence of callable alleles besides the handler’s at six or more loci. A possible
explanation is that there were multiple handlers of the backpack, but all participants in
the study said that they were the only user of their backpack. Nonetheless, casual contact
is all that is needed to transfer shed epithelial cells (Van Oorschot and Jones 1997). Brief
contact with a passerby, or a classmate simply moving the backpack, might have resulted

in cells being deposited without the handler’s knowledge.

50

Mixtures could have also been caused by contamination from researchers.
Specific measures were implemented to avoid contamination knowing that it had caused
problems in previous studies involving recovering DNA from IEDs (e. g., Gehring 2004;
Kremer and Foran 2008). No direct sources of contamination were found when DNA
profiles of researchers involved in this study were compared with mixtures, although it
cannot be concluded that no researcher contamination occurred. For example, sample
SZ4UP had 15 of the main investigator’s 18 alleles called. It is possible that the three
alleles that were not called were masked by un-called stutter from the handler’s alleles, as
all missing alleles fell one repeat unit shorter than a handler’s allele. It is also
conceivable that the main investigator’s DNA was such a minor component of the
mixture that the three alleles not present in the profile dropped out. Sample 3ZlUP was
the only other mixture where there was a strong indication that the main investigator was
a contributor (15 of 18 alleles called). Five additional mixtures had indications that the
main investigator may have been a contributor, but there were not as many of the
investigator’s alleles called as were seen in the previous two cases. Comparing mixture
profiles with profiles of other investigators involved in the study did not result in any
indications of their contamination.

Contamination fi'om researchers could have occurred at any one of numerous
steps ranging from backpack preparation to PCR amplification. However, the step most
likely vulnerable to contamination was swabbing the backpacks, at which time they were
handled extensively. In particular, the zippers required the most handling during
swabbing because of their small size. Zipper samples accounted for 6 (33%) mixtures,

including both cases where there was a strong indication that contamination stemmed

51

from the main investigator. The straps accounted for 7 mixtures, three of which had
allele calls that suggested the main investigator was the contributor. On the other hand,
no mixtures were obtained from front tab samples, and DNA recovered fiom the top
handles only resulted in one mixture.

A study being conducted concurrently in the lab revealed another potential source
of contamination. Mitochondrial DNA analyses performed on ‘clean’ swabs
(860PPC, Puritan Medical Products Co.) showed the presence of mtDNA, which resulted
in switching to a different type of swab (25-806 2PC, Puritan Medical Products Co.).
However, the five STR analyses of the swab reagent blanks (backpacks 7, 9, and 10 — 12)
made using the potentially contaminated swabs produced a total of two callable alleles,
while the three reagent blanks made using the new swabs produced one (no swab reagent
banks were processed with backpacks 3 and 4). The three callable peaks were attributed
to sporadic contamination, and the concern over swab contamination was lessened given
that nuclear DNA contamination was not detected. This is likely attributable to the
increased sensitivity of mtDNA analyses.

Results from backpack 12 (the negative control) exposed another possible source
of extra peaks in electropherograms—exogenous DNA remaining on the backpacks afier
the decontamination process. Analysis of 1221UP produced a full male profile
(Appendix A) that could not be linked to anyone directly involved in the study
(researchers or participants). Electropherograms from the remaining eight backpack 12
samples produced a total of ten callable alleles, six of which matched alleles from
12Z1UP. This suggested that all areas were not sufficiently exposed to UV irradiation,

and DNA deposited on the backpack during manufacturing and/or delivery persisted.

52

Processing the control swabs that were taken before backpacks were distributed to
participants (see Material and Methods) would have assisted in shedding light on the
extent to which DNA remained on the backpacks after decontamination, however, this
was deemed unnecessary as the peaks stemming from the negative control backpack
samples were mostly of low RFU values.

In a forensic setting, backpacks would not be decontaminated in advance, and
could easily harbor DNA from more than one individual. Even in the current research
where decontamination procedures were implemented, mixtures and loci with more than
two callable alleles were prevalent. However, as noted above, this problem was
overcome by swabbing multiple backpack regions and using information from all the
swabs to create a consensus profile.

Swabbing different backpack regions also revealed that some were better targets
for DNA recovery than others. The top handle produced the handler’s alleles at all nine
loci for every backpack, while DNA recovered from swabs of the left strap, front middle,
and neck region producing similarly good results (Table 4). Conversely, the right strap,
zippers, and fiont tab averaged less than eight loci with the handler’s correct alleles.
Recovering alleles may have been dependent on the surface area of the region swabbed,
given that the small zippers and front tab produced the poorest results. The right strap
was an exception as it averaged one locus less than the left strap, however this number
was skewed due to two right strap samples, 7RSP and 9RSP, which had 6 and 4 correct
loci, respectively. As previously mentioned, problems involving 7RSP were likely due to
PCR inhibition, while an insufficient amount of template DNA probably led to drop-out

at 9RSP.

53

Recovering alleles was also likely influenced by the frequency of which
participants utilized different regions of the backpacks. The overall level of contact is
unknown of course, however, normal backpack use involves the straps being handled in
order to wear it, providing ample opportunities for DNA to be deposited via grabbing the
straps and putting the wearer’s arms through them. Top handles are also commonly used
on backpacks. The success in recovering alleles from the neck region could have
stemmed from its location between the top of the straps and the top handle, making it
susceptible to shed cells when either region was utilized. Further, this region was prone to
contact with the handler’s neck and hair. In this regard, it is recommended that DNA
recovery efforts focus on regions that are likely to be handled or contacted by the user,
and although swabbing multiple areas was shown to be a very successful approach to
identifying handlers, larger regions may be advantageous for recovering a satisfactory
number of cells for analysis.

Quantitating the DNA provided another means of comparing the different regions
targeted for DNA recovery. Quantitation results showed all of the non-zipper regions,
except for the front tab, averaged close to 300 pg/uL of DNA, supporting the hypothesis
that textured regions and/or surface area are important factors in the recovery of cells.
However, it is also possible that certain parts of the backpack, such as the front tab, were
simply not utilized by the handler. Lower quantities of DNA were also recovered from
the zippers, but there was variation in DNA amounts depending on the components that
were swabbed. The average amount of DNA from zippers that retained tabs (124.76
pg/uL) was greater than those that did not have them (45.2 pg/uL). This disparity among

zippers might have been caused by the ease of swabbing the plastic surface of tabs

54

compared to the strings, which were often singed. Even if DNA persisted on the singed
strings, difficulties in swabbing them, such as the cotton tip of the swabs becoming
unwound, likely hindered recovery. Strings that were not singed also presented problems
when swabbing as they were small and intertwined with the pull. Like the tabs, the metal
pulls were easy to swab, but the zippers were constructed in a manner that favored the
user gripping the tab or the string, hence these areas likely harbored more DNA

Using Quantifiler'rM to quantitate the DNA also provided a means of assaying
PCR inhibition. Along with sample 7RSP, 3RSP showed signs of inhibition based upon
poor amplification of the IPCs, however, in both cases genomic DNA was still detected
Sample 3RSP was likely subjected to partial inhibition as, unlike 7RSP, a full handler’s
profile was obtained. It has been reported that when the IPC does not amplify, but human
DNA does, the IPC results should be disregarded, however, this is only valid for high
amounts of DNA (>10 ng) as competition between the genomic DNA and the IPC leads
to the suppression of IPC amplification (Applied Biosystems 2006a; Katz 2007).
Initially, inhibition concerns centered on post-blast residue, but inhibition most likely
would have been more prevalent if that was the cause. Inhibition in samples from right
straps suggests that there was some feature about the straps that hindered PCR. Straps
were made fiom a different material than the rest of the backpack, and pre—blast control
swabs of straps sometimes turned black. This substance may have been deleterious to
PCR, although the left strap samples did not show any signs of inhibition. The exact
reasons for PCR inhibition were not resolved, but its low occurrence in general resulted
in minimal impact on the study. However, this does not mean PCR inhibition will be

absent in all analyses of IED containers. Variation in the composition of containers (e. g.,

55

cardboard, fabric, wood, or metal), and the elements to which the containers have been
exposed to before and during deflagration (e.g., soil, blood, water), provide many diverse

situations, each with the potential of producing different PCR inhibitors.

56

CONCLUSIONS

The impact that IEDs have had in the Middle East has caused growing concern
that these devices will become more prominent domestically, with the potential to have
profound implications on American society. Two experts, counterterrorism consultant
David Williams and Michigan State Police Bomb Squad Commander Lt. Shawn
Stallworth, both say that they have already seen an increase in IED usage in the United
States (personal communications). Even with the lack of a large scale domestic IED
event since 9/11, IEDs are implemented daily throughout the United States.

Results of this study show that post-blast DNA recovered from IED containers
can be used to correctly identify the handler. Analyzing multiple samples from the same
backpack and creating a consensus profile proved to be an effective approach to
identifying handlers and was necessary to circumvent drop-out and mixtures. All areas
targeted for swabbing produced a handler’s full profile from at least one backpack, but
overall, regions that were larger and likely used more often resulted in higher quantities
of DNA and more loci with the handler’s alleles. With this in mind, it is recommended
that swabbing first be directed towards areas that are likely to be handled, particularly
larger areas initially and then smaller regions (e. g., zippers), as need, resources, and time
permit. However, if only small pieces of a container are available, these should not be
deemed irrelevant because they can still produce informative data.

The research presented offers a feasible and practical approach towards national
efforts aimed at countering lEDs, and answers DHS Secretary Chertoff 5 call to the

scientific community to develop means to counter IEDs (Chertoff 2007). Although much

57

of the focus of countering IEDs has been on developing measures to prevent them from
detonating, the fact that this research focused on post-blast analysis does not take away
fi'om its value, as it is unlikely that IED countermeasures will ever thwart all IED events.
This study demonstrated an effective way to identify those responsible for IEDs, and aids
in their apprehension, which will most likely prevent future attacks and allow citizens to
feel more secure. By disseminating the findings of the research to law enforcement
personnel and emergency responders, the importance of properly collecting post-blast
fragments can be stressed, and DNA analyses of IED containers can become a valuable

tool in IED investigations.

58

APPENDIX A. Allele calls and loci categorizations

The tables show allele calls made for the backpacks involved in the study. Tables
for the control backpacks (11 and 12) are included, however, they served solely as
references and their data were not incorporated into analyses. Loci from each sample are
colored to signify the category they were placed in (see Results). DNA profiles ‘Call’
were blindly made based upon the review of all loci, and checked for concordance with
reference profiles from study participants ‘Subject’. The following key is for interpreting

all tables.

 

S mbol Description
The locus contained only the handler’s correct alleles.
The locus had multiple allele calls, but the handler’s alleles constituted the
major profile.
The locus had multiple allele calls, but the handler’s alleles could not be
distinguished as the major profile.
The locus had at least one of the handler’s correct alleles.
None of the alleles matched the handler’s profile.
No alleles were called at the locus.
BOLD Peak was between 100 and 149.9 RFUs
" Peak was between 50 and 99.9 RFUs
+ Classified as a mixture

 

 

 

59

 

0m. 5.

I.

5.6—.

mt:

omKN

0N

a

mlm

3.2

Nfim

0 50.330

 

0N. _.N

I.

Nfimr

mt:

8K0

VNdN

:dr

Nwd

=00

 

 

0:03

 

m xomaxomm

 

6O

.rrrmmaw
00:0

 

mim—
mfimw

0000
0:00

mfimw
mtmw

omKN
omKN

dem
0050

x
x

otm
one
.PF

:6
:6

 

008..

m «00.305
=00

w xowaxomm

61

 

 

FN

N..._._.

or

o_.

Nm.mw

VNdN

C.

00:

0 8.330

 

 

Fm

mm:

9.

or

Nm.mm

vmdm

I.

N...:.

:00

 

 

 

0:00..

mmm
m<>>mmmm
n_._h_m
n=>_n_m
mZm
mmmm
mmrfi
n_I._.m
mDNNm
m3 _.Nm
ntm

m xooaxomm

 

62

 

mNdN
mNdN

Ni:
NP...—

3.2
032.

:6
Sum

0.00.00
0;0;0

vmdm
cmdm

x
x

0000
000—

w...
NF

 

 

0:00..

_ «003.0
=00
mmo

m<>>wmm©

nzkw

+n=2n_o
+n_Z®
+mwm®

+1040

n_I._.o

0:0N0
anrNo
0020

0 xomaxomm

 

63

#00 00 2 0000.008 003 0.2.0 .0000. to 05 .0000: 40.00 000 mm 00.2.0 .2 00.0 05 0002000 0: 003000 2.020.008 0.0.2....

 

 

 

NMNNN :6F 5 :6 1.00.5 ONKF x Nfim NF. : 0 80.300
NdNNN :dw 5 Sum :00. 5 ON: x Nfim N—.: :00
mmh

m<>>wmmn

 

 

0003

n 0.000008

 

 

v0.0. 00: 9...: :0
v0.2 00: 002 2.0

 

omdw PNdN X :6 N _._m u. 50.33m

omfim FNdN x :d N; :00
mmm
.N _. m<>>mmmm
mhuw
me

mwmm
+n_m._w
+n=.....m

mmNm
+n.30Nm

+08%
020.00
+0: £0
£5

 

0:00..

0 xomaxoam

65

.3 6m 0 0o Bomzuoaoa em 0 0003000 0008 00 30000908 80 0:50 36':

 

dem N? w flaw m .._N_. 5.0m CNN: x 0 Wm v in < «003.5

0 - . ilwume - n —
NNON NP ormw mwNF =mooz 0N: x owm 3.0 :00
mmm

m<>>wmmm
mbum
120m

+n.zm

mmmm
+n.m1.m

1:5

 

 

 

+n_va

 

038..

m xomaxomm

 

66

 

 

mwmw 000, 0000 PF 0000 0000 x 000 0000 0.00030
0N00 000? 000F PF 0000 0N0? x 000 00;? :00
.OF >xx 000.
.000 0r 902.0002
00000
02000
020_

+mmmow

 

+quow

0:Fo_
0000_
00000
00000

03009

anrNoF
0020

 

 

030°;

owxumaxomm

67

0002000 20. 000000980. 80 000; 0.009.000 0.5 E0: 0.00 062

 

vaM 00 0F 00: 0. F0 00.00 x 000 2.0 .220:
0.0.00 00 0F 00: 0. F0 00.00 x 000 2.0 :8
00:
055000:
at:
02“.:
a2:
0003
004:
0:0:
030:
0:00:
0:00:
020:
0020

 

 

0:004
.0008 +

2 0.0000000

68

0002000 20. 00.05980. .00 0002, xo0ax000 0.5 So: 0.00 ”062

 

*om

mm. 5.
<0“.

0:.

.3

«to?
n50

000...?
9.

00...: Sum
30 000

Pndm
FNO
0:004

3.
9.

vududv
No

.>

>.X
..m.S.<

«or
no

.2.

3
9.0

mmmw
m<>>wmmmw
QENF
n=2u.N_.
QZNF
awmmv
awqmr
QIPNP
QVNNF
QDNNNF
05 .NN..
0.03m

.9008 ..
NF x00200m

69

APPENDIX B. DNA quantitation data

DNA samples were quantified in triplicate using QuantifilerTM. The averages and
standard deviations were calculated for all samples and are reported in nanograms.
Reactions in which the threshold cycle of the internal positive control was above 30 are
noted. Areas swabbed were labeled as; backpack#/identifer/P. The identifiers used were:
Z# for zippers (# replaced by 1, 2, 3, 4, or 5 depending upon the zipper’s origin; if the
origin of the zipper was unidentifiable a ‘U’ was added after the number), TH for top
handle, LS and RS for left and right strap, respectively, FM for front middle, FT for front
tab, and N for neck region. For example, 3THP indicates the swab is from the top handle

of backpack 3, while BZlUP indicates that the swab is from the first unidentifiable zipper

 

 

of backpack three.
Backpack 3
Replicate
Standard
Swab 1 2 3 Average Deviation
321 UP 0.113 0.0874 0.0970 0.0991 0.0129
322UP 0.0689 0.0909 0.0461 0.0686 0.0224
3Z3UP 0.0451 0.0671 0.0310 0.0476 0.0184
3Z4P 0.0943 0.116 0.118 0.109 0.0131
3Z5P 0.0389 0.0160 0.0189 0.0246 0.0125
3THP 0.196 0.271 0.179 0.215 0.0490
3LSP 0.238 0.241 0.192 0.224 0.0275
3RSP 0.641 0.347* 0.267“ 0.418 0.197
3NP 0.322 0.302 0.251 0.292 0.0366
3FT P 0.0416 0.0172 0.0248 0.0279 0.0125

 

*—IPC had a threshold cycle of 37.3 1.

**—IPC did not reach the threshold cycle by 40 cycles.

70

 

Backpack 4

 

 

 

 

 

 

 

Replicate
Standard
Swab 1 2 3 Average deviation
4Z1UP 0.154 0.152 0.188 0.165 0.0202
4Z2UP 0.427 0.387 0.364 0.393 0.0319
4Z3UP 0.153 0.192 0.103 0.149 0.0446
4Z4UP 0.291 0.202 0.261 0.251 0.0453
4THP 0.803 0.650 0.743 0.732 0.0771
4LSP 0.926 0.963 0.787 0.892 0.0928
4RSP 1.25 1.16 1.29 1.233 0.0666
4NP 1.06 1.62 1.07 1.25 0.320
4FMP 0.895 0.890 0.759 0.848 0.0771
4FTP 0.410 0.369 0.375 0.385 0.0221
Backpack 5
Replicate
Standard
Swab 1 2 3 Average deviation
521 UP 0.00638 0.00972 0.01 19 0.00933 0.00278
522UP 0.00660 0.00915 0.0129 0.00955 0.00317
5THP 0.326 0.339 0.306 0.324 0.0166
5LSP 0.152 0.187 0.208 0.182 0.0283
5RSP 0.0596 0.0466 0.0402 0.0488 0.00989
5NP 0.0171 0.0194 0.0143 0.0169 0.00255
5FMP 0.0633 0.0285 0.0388 0.0435 0.0179
5FTP 0.0300 0.0173 0.000 0.0158 0.0151
Backpack 6
Replicate
Standard
Swab 1 2 3 Average deviation
621 UP 0.00221 0.0104 0.0131 0.00857 0.00567
622UP 0.00270 0.000 0.00419 0.00230 0.00212
6THP 0.0554 0.0476 0.0681 0.0570 0.0103
6LSP 0.0524 0.0279 0.0400 0.0401 0.0123
6RSP 0.0538 0.061 1 0.0438 0.0529 0.00869
6NP 0.0651 0.0735 0.0537 0.0641 0.00994
6F MP 0.0484 0.0320 0.0172 0.0325 0.0156
6FTP 0.000 0.0211 0.00320 0.00810 0.0114

 

71

 

Backpack 7

 

Replicate
Standard
Swab 1 2 3 Average deviation
721 UP 0.0846 0.132 0.0955 0.104 0.0248
7THP 0.293 0.280 0.347 0.307 0.0355
7LSP 0.182 0.187 0.198 0.189 0.00819
7RSP* 0.260 0.162 0.217 0.213 0.0491
7NP 0.158 0.238 0.180 0.192 0.0413
7FMP 0.368 0.536 0.618 0.507 0.127
7FTP 0.0907 0.0522 0.0482 0.0637 0.0235

 

*——IPC did not reach the threshold cycle by 40 cycles for all three replicates.

 

 

 

 

 

Backpack 8
Replicate
Standard
Swab 1 2 3 Average deviation
821 UP 0.214 0.158 0.124 0.165 0.0454
822UP 0.260 0.265 0.202 0.242 0.0350
8Z3UP 0.150 0.146 0.152 0.149 0.00306
824UP 0.0559 0.125 0.0939 0.0916 0.0346
825P 0.282 0.273 0.334 0.296 0.0329
8THP 0.404 0.595 0.699 0.566 0.150
8LSP 0.775 0.874 0.767 0.805 0.0596
8RSP 0.271 0.221 0.322 0.271 0.0505
8NP 0.573 0.487 0.541 0.534 0.0435
8FTP 0.0245 0.00489 0.0180 0.0158 0.00999
Backpack 9
Replicate
Standard
Swab 1 2 3 Average deviation
921 UP 0.0105 0.0252 0.0182 0.0180 0.00735
922UP 0.0171 0.0151 0.000 0.0107 0.00935
923UP 0.0109 0.0101 0.000 0.00700 0.00608
924P 0.0332 0.0312 0.0161 0.0268 0.00935
9THP 0.0492 0.00866 0.0326 0.0302 0.0204
9LSP 0.0560 0.0735 0.0308 0.0534 0.0215
9RSP 0.0429 0.0306 0.0332 0.0356 0.00648
9NP 0.0529 0.0860 0.0348 0.0579 0.0260
9FMP 0.0520 0.0269 0.0287 0.0359 0.0140
9FTP 0.0241 0.0199 0.00759 0.0172 0.00858

 

72

 

Backpack 10

 

Replicate

Standard

Swab 1 2 3 Average deviation
1OZIUP 0.0457 0.0160 0.0152 0.0256 0.0174
1022UP 0.0209 0.00918 0.00752 0.0125 0.00729
1023P 0.0147 0.0436 0.0276 0.0286 0.0145
1024P 0.0188 0.0229 0.0493 0.0303 0.0166
1025P 0.0139 0.0208 0.00486 0.0132 0.00799
10THP 0.0606 0.0725 0.108 0.0804 0.0247
10LSP 0.0817 0.0564 0.0593 0.0658 0.0138
10RSP 0.0341 0.0490 0.0888 0.0573 0.0283
10NP 0.0315 0.063 0.0815 0.0586 0.0253
10FMP 0.0150 0.0326 0.000 0.0159 0.0163
10FTP 0.000 0.00422 0.000 0.00141 0.0024

 

73

REFERENCES

Akane A, Matsubara K, Nakamura H, Takahashi S, Kimura K. 1994. Identification of the
heme compound c0purified with deoxyribonucleic acid (DNA) from bloodstains, a
major inhibitor of polymerase chain reaction (PCR) amplification. J Forensic Sci.
39:362-72.

Al-Soud WA, Radstrom P. 2000. Effects of amplification facilitators on diagnostic PCR
in the presence of blood, feces, and meat. J Clinical Microbiology. 38:4463-70.

Applied Biosystems. 2006, October. The next generation of STR technology begins with
the AmpFlSTR® MiniFilerTM PCR amplification kit, available soon fi'om Applied
Biosystems. Forensic News. Retrieved February 11, 2008, fiom
http://marketing.appliedbiosystems.com/images/enews/ForensicNews_Vol7/PDF/00_
ForensicNews.pdf

Applied Biosystems. 2006a. QuantifilerTM human DNA quantification kit and
QuantifilerTM Y human male DNA quantification kit. User guide.

Applied Biosystems. 2007. AmpFlSTR® MiniFilerTM PCR amplification kit. Product
Bulletin.

Applied Biosystems. 2007a. AmpFlSTR® MiniFilerTM PCR amplification kit. User
guide.

Atkinson R. 20073, October 1. ‘There was a two-year learning curve...and a lot of people
died in those two years.’ Washington Post. p. A01. Retrieved March 18, 2008, fi’om
http://www.washingtonpost.corn/wp—
dyn/content/ article/2007/09/ 3 0/AR2007093 001 675_2.htrnl?sid=ST2007092900754

Atkinson R. 2007b, October 3. ‘If you don’t go after the network, you’re never going to
stop these guys. Never.’ Washington Post. p. A01. Retlieved March 18, 2008, from
http://www.washingtonpost.com/wp-
dyn/content/article/2007/ 10/02/AR2007100202366.htm1?sid=ST2007092900754

Bessetti J. 2007, March. An introduction to PCR inhibitors. Profiles in DNA. 1029-10.

Bolz F, Dudonis K, Schulz DP. 2005. The counterterrorism handbook. CRC Press.
Boca Raton, FL.

Bourke MT, Scherzinger CA, Ladd C, Lee HC. 1999. NaOH treatment to neutralize
inhibitors of Taq polymerase. J Forensic Sci. 44: 1046-50.

Braid MD, Daniels LM, Kitts CL. 2003. Removal of PCR inhibitors from soil DNA by
chemical flocculation. J Microbiological Methods. 52:389-93.

74

Bright J, Petricevic SF. 2004. Recovery of trace DNA and its application to DNA
profiling of shoe insoles. For Sci Int. 145:7-12.

Budowle B, Moretti TR, Baumstark AL, Defenbaugh DA, Keys J M. 1999. Population
data on the thirteen CODIS core short tandem repeat loci in Afiican Americans, US.

Caucasians, Hispanics, Bahamians, Jamaicans, and Trinidadians. J Forensic Science.
44:1277-86.

Budowle B, Hobson DL, Smerick J B, Smith JAL. 2001. Low copy number-
consideration and caution. Presented at Twelfth International Symposium on Human
Identification. Biloxi, MS, USA.

Burke R. 2000. Counter-terrorism for emergency responders. CRC Press LLC. Boca
Raton, FL.

Butler JM. 2005. Forensic DNA typing 2nd ed. Elsevier Academic Press. Boston, MA.

Butler JM. 2006. Genetics and genomics of core short tandem repeat loci used in human
identity testing. J Forensic Sci. 51:253-65.

Butler J M, Shen Y, McCord BR. 2003. The development of reduced size STR amplicons
as tools for analysis of degraded DNA. J Forensic Sci. 48(5):1054-64.

Chertoff M. 2007, October 19. Preventing IED attacks. DHS Leadership Journal
(discussion forum). Retrieved February 25, 2008, from
http://www.dhs. gov/ j oumal/leadership/2007_1 O_01_archive.html

Clayton TM, Whitaker JP, Sparkes R, Gill P. 1998. Analysis and interpretation of mixed
forensic stains using DNA STR profiling. For Sci Int. 91:55-70.

Coble MD, Butler J M. 2005. Characterization of new miniSTR loci to aid analysis of
degraded DNA. J Forensic Sci. 50(1):43-53.

Collins PJ, Hennessy LK, Leibelt CS, Roby RK, Reeder DJ, Foxall PA. 2004.
Developmental validation of a single-tube amplification of the 13 CODIS STR loci,
D281338, D19S433, and amelogenin: the AmpFlSTR Identifiler PCR amplification
kit. J Forensic Sci. 49:1265-77.

Esslinger KJ, Siege] JA, Spillane H, Stallworth S. 2004. Using STR analysis to detect
human DNA from exploded pipe bomb devices. J Forensic Sci. 49:481-4.

Foran DR. 2006. Relative degradation of nuclear and mitochondrial DNA: an
experimental approach. J Forensic Sci. 512766-70.

Gehring M. 2004. The recovery and analysis of mitochondrial DNA from exploded pipe
bombs. Master’s thesis. Michigan State University.

75

Gill P. 2001. Application of low copy number DNA profiling. Croatian Medical J.
42:229-32

Gill P, Whitaker J, Flaxman C, Brown N, Buckleton J. 2000. An investigation of the rigor
of interpretation rules for STRs derived fi'om less than 100 pg of DNA. For Sci Int.
1 12:1 7-40.

Green RL, Roinestad IC, Boland C, Hennessy LK. 2005. Developmental validation of the
QuantifilerTM real-time PCR kits for the quantification of human nuclear DNA
samples. J Forensic Sci. 502809-31.

Grubwieser P, Muhlmann R, Berger B, Niederstatter H, Pavlic M, Parson W. 2006. A
new “mini-STR multiplex” displaying reduced amplicon lengths for the analysis of
degraded DNA. Int J Legal Med. 120:1 15-20.

Hill CR, Kline MC, Coble MD, Butler J M. 2008. Characterization of 26 miniSTR loci for
improved analysis of degraded DNA samples. J Forensic Sci. 53(1):73-80.

Holland MM, Parsons TJ. 1999. Mitochondrial DNA sequence analysis — validation and
use for forensic casework. For Sci Rev. 1 1: 21-51.

Katz DE. 2007, March. The hidden benefits of real-time PCR: Assessing and addressing
qualitative challenges. Profiles in DNA. 1026-8.

Kisilevsky AE, Wickenheiser RA. 1999. DNA PCR profiling of skin cells transferred
through handling. Proceedings of the Annual Meeting of the Canadian society of
Forensic Science. Edmonton, Alberta, Canada.

Kline MC, Duewer DL, Redman JW, Butler J M. 2005. Results from the NIST 2004
Quantitation Study. J Forensic Sci. 50:1-7.

Kreader CA. 1996. Relief of amplification inhibition in PCR with bovine serum albumin
or T4 gene 32 protein. Appl Environ Microbiology. 62:1102-06.

Kremer S, Foran D. 2008. Identification of improvised explosive device assemblers
using miniSTRs. Presented at the American Academy of Forensic Sciences 60th
annual scientific meeting. Washington, DC, USA.

Krenke BE, Tereba A, Anderson SJ, Buel E, Culhane S, Finis CJ, Tomsey CS, Zachetti
J M, Masibay A, Rabbach DR, Amiott EA, Sprecher CJ. 2002. Validation of l6-locus
fluorescent multiplex system. J Forensic Sci. 47:773-85.

Lowe A, Murray C, Whitaker J, Tully G, Gill P. 2002. The propensity of individuals to
deposit DNA and secondary transfer of low level DNA from individuals to inert
surfaces. For Sci Int. 129:25-34.

76

Mueller 111 RS. 2007, March 28. Speech fi'om the director of the FBI. Retrieved March
21, 2008, fi'om http://www.fbi.gov/pressrel/speeches/muellerO32807.htrn

National Research Council: Committee on defeating improvised explosive devices. 2007.
Countering the threat of improvised explosive devices: basic research opportunities,
abbreviated version. National Academies Press. Washington, DC. Retrieved October
23, 2007, fi'om http://www.nap.edu/catalog/1 1953.htm1

Noe D. 2007. Eric Rudolph: serial bomber (the Olympics bombed). Retrieved February
11, 2008, fiom
http://www.crimelibrary.corn/terrorists_spies/terrorists/eric_rudolph/1 .htrnl

Parsons TJ, Huel R, Davoren J, Katzmarzyk C, Milos A, Selmanovic A, Smaljovic L,
Coble MD, Rizvic A. 2007. Application of novel “mini amplicon” STR multiplexes
to high volume casework on degraded skeletal remains. For Sci Int Genetics. 1:175-
79.

Radstréim P, Knutsson R, Wolffs P, Lovenklev M, Lofstrom C. 2004. Pre-PCR
processing: Strategies to generate PCR-compatible samples. Molecular
Biotechnology. 26: 1 33-46.

Ricci U, Giovannucci Uzielli ML, Klintschar M. 1999. Modified primers for D128391
and a modified silver staining technique. Int J Legal Med. 112:342-44.

Robin E, Wong R. 1988. Mitochondrial DNA molecules and virtual number of
mitochondria per cell in mammalian cells. J Cell Physiol. 136:507-13.

Sageman M. 2004. Understanding terror networks. U. of Penn. Press. Philadelphia, PA.

Shutler G, Gagnon P, Verret G, Kalyn H, Korkosh S, Johnston E, Halverson J. 1999.
Removal of a PCR inhibitor and resolution of DNA STR types in mixed human-
canine stains fi'om a five year old case. J For Sci. 44:623-26.

Stallworth, F/Lt. Shawn E. Bomb squad cormnander, Michigan State Police.

STRBase. Retrieved March 6, 2008, from http://www.cstl.nist.gov/biotech/strbase/

Sweet D, Lorente M, Lorente J A, Valenzuela A, Villanueva E. 1997. An improved
method to recover saliva from human skin: the double swab technique. J For Sci.

42:320-2.

Timeline: Madrid investigation. 2004. Retrieved February 11, 2008, from
http://news.bbc.co.uk/2/hi/europe/3597885.st:m

Van Oorschot RAH, Jones MK. 1997. DNA fingerprints from fingerprints. Nature. 387:
767.

77

Watson RJ, Blackwell B. 2000. Purification and characterization of a common soil
component which inhibits the polymerase chain reaction. Can J Microbiol.46:633-42.

Whitaker JP, Clayton TM, Urquhart AJ, Millican ES, Downes TJ, Kimpton CP, Gill P.
1995. Short tandem repeat typing of bodies from a mass disaster: high success rate

and characteristic amplification patterns in highly degraded samples. BioTechniques.
18(4):670-7.

Wickenheiser R. 2002. Trace DNA: a review, discussion of theory, and application of the
transfer of trace quantities of DNA through skin contact. J Forensic Sci. 472442-50.

Wiegand P, Kleiber M. 2001. Less is more—length reduction of STR amplicons using
redesigned primers. Int J Legal Med. 1 14:285-7.

Williams D. Counterterrorism consultant, Counter Terrorism Consulting Group LLC,
Harwood, Maryland. November 6, 2007.

Wilson, C. 2007, November 21. Improvised explosive devices in Iraq and Afghanistan:
effects and countermeasures. CRS Report for Congress. Retrieved February 9, 2008,
fiom http://assets.opencrs.com/rpts/R822330_20071 121.pdf

Yoshida K, Sekiguchi K, Kasai K, Sato H, Seta S, Sensabaugh GF. 1997. Evaluation of
new primers for CSF 1 PO. Int J Legal Med. 110236-38.

78

 

 

l"iliillgiiglljiiiiii

 

-‘ “.3--_