. a
5-3. .
is. , m

x . .

 

unmade

SE...

[.31 63.31.

‘13.; {LI , a
t‘i A
$3.15.;

.m 52,};
Was

It)!

{.15).
52.5.

.1tli..,l....
3&1!- :‘193
(r. “111‘
light“ 10 p
,1) .3 .

1.

 

 

 

 

 

 

 

.I .-
Xvsxé.‘ 3.2

1.29.1.3

 

 

t
I.
2:! I}: V

 

‘ \lzl.»
y‘au‘ntu‘l‘r 31%.} ‘ .
. ‘1»: ‘0

THESIS

“at

This is to certify that the

thesis entitled

THE EVALUATION OF INDIVIDUAL LOCUS PERFORMANCE USING THE
PROMEGA POWERPLEX 16 SYSTEM FOR USE WITH SINGLE SOURCE CODIS
SAMPLES

presented by

Teri Lynn Lawton

has been accepted towards fulfillment
of the requirements for

M. S - degree in Wience

Q/M”! ,XglflA
7 fl

  

f'
Major prof or

Date [1/1/3/

0-7639 MS U is an Affirmative Action/Equal Opportunity Institution

 

 

 

LIBRARY

Michigan State

University

 

 

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

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/01 c:/CIFIC/DateDue.p65-p.15

 

THE EVALUATION OF INDIVIDUAL LOCUS PERFORMANCE USING THE
PROMEGA POWERPLEXmlo SYSTEM FOR USE WITH SINGLE SOURCE CODIS
SAMPLES
By

Teri Lynn Lawton

A THESIS
Submitted to
Michigan State University

in partial fulfillment of the requirements
for the degree of

MASTER OF SCIENCE
School of Criminal Justice

2002

ABSTRACT
THE EVALUATION OF INDIVIDUAL LOCUS PERFORMANCE USING THE
PROMEGA POWERPLEX 16 SYSTEM FOR USE WITH SINGLE SOURCE CODIS
SAMPLES
By

Teri Lynn Lawton

In the field of forensic DNA analysis, a genotyping kit (the Promega
PowerPlexWI6 System) has been developed which identifies genotypes at sixteen
different loci. This kit would be ideal for the genotyping of samples that must be entered
into the Combined DNA Index System (CODIS), a DNA database for convicted
ofl‘enders of violent crimes. It would be an improvement over the current method of
analysis, because it only requires one reaction (as compared to two)—but it must first
prove to be an effective and reliable method for the analysis of single source samples.
This project will evaluate the performance of this kit at each locus and determine whether
it would be suitable for single source samples. This project will evaluate locus
performance by comparing values such as peak height ratios and relative fluorescence
units between and within each locus. One hundred and fourteen individuals will be tested
to provide data needed to compare these values and determine whether the
PowerPlexml6 System would be a reliable method for genotyping single source samples

in the forensic DNA laboratory.

ACKNOWLEDGEMENTS

I would like to acknowledge the Michigan State Police Forensics Laboratory
(DNA Unit) in Lansing, Michigan for the opportunity to perform research at their facility.
I would especially like to thank DNA laboratory Supervisor Charlie Barna for giving me
the opportunity to work with the PowerPlexTM 16 System, Forensic Scientists Don Yet
and Glen Hall for their assistance throughout the project, and the rest of the MSP DNA
Unit Staff who were always there to answer endless questions. In addition, I would also
like to thank the Promega Corporation for providing the PowerPlexTM 16 Kits and DNA
extracts that were utilized in this evaluation.

I would also like to acknowledge Dr. Jay Siege] for his encouragement, support,
and persistence throughout my time as a graduate student at Michigan State University.
lastly, I would like to thank my parents and family for always encouraging me to finish

my projects, and for all of the support they have given me the past few years.

iii

TABLE OF CONTENTS

LIST OF TABLES .................................................................................... v
LIST OF FIGURES ................................................................................. vi
INTRODUCTION ................................................................................... 1
THE POLYMERASE CHAIN REACTION AND

THE DEVELOPMENT OF STR MULTIPLEXING .......................................... 11
MATERIALS AND METHODS .................................................................. 21
RESULTS AND DISCUSSION .............................. ' ..................................... 26
CONCLUSION AND FUTURE RESEARCH .................................................. 45
WORKS CITED .................................................................................... 48
ADDITIONAL WORKS CONSULTED ........................................................ 50

iv

LIST OF TABLES

Table 1: Average Peak Heights (R.F.U.) for 9947A,

MBILC, BIS, and H9 at various DNA template amounts .................................... 28
Table 2: Peak Height Ratios of Heterozygotes at

Different Template Amounts (9947A, MBILC, BIS, and H9) ............................... 28
Table 3: Heterozygote Peak Height Ratios ..................................................... 35
Table 4: Average Peak Heights (R.F.U.) Per Locus .......................................... 44

LIST OF FIGURES

Figure l: R.F.U. vs Target DNA, (a) Sample 9947A;

(b) Sample MBILC; (c) Sample BIS; and ((1) Sample H9 ..................................... 29
Figure 2: Peak Height Ratio vs Target DNA (9947A, MBILC, B15, & H9) ............... 31
Figure 3: Heterozygote Peak Height Ratios vs Locus ......................................... 35
Figure 4: R.F.U. vs Fragment Size (Per Locus) ................................................ 36

vi

Introduction

I. The Value of DNA Evidence

The application of deoxyribonucleic acid (DNA) analysis to the field of forensic
science has broadened the horizons of criminal investigation procedures. Evidence
obtained fiom a crime scene that contains DNA can be the prime incriminating or
exonerating factor in a case. DNA is present in most biological fluids such as blood,
semen, vaginal fluid, saliva, and can be occasionally found in urine or feces. Other
samples from which DNA can be extracted include hair, bone, and tooth pulp. Blood
samples tend to produce the best DNA yield (twenty to forty thousand ng/ml), while
urine and bone the least (one to twenty ng/ml) [1]. From their introduction to forensic
science applications, many types of DNA analysis have evolved through time, with the
most recent type termed short tandem repeat (STR) analysis. This process utilizes short
repeating sequences that occur throughout the genome to calculate a frequency of a
particular genetic makeup (genotype).

The Federal Bureau of Investigation (FBI) has established the COmbined DNA
Index System (CODIS)—a database containing genetic profiles of persons convicted of
sexual offenses and other violent crimes. Now, investigators have a useful tool with
which they can compare DNA evidence found at the crime scene with a database of
known convicted offenders. The ultimate goal for CODIS is for every state to have a
database of DNA profiles collected from the scene of the crime, or from the criminals
themselves. This data can then be centralized, allowing each state the ability to search

and compare their data with all of the other states.

The FBI has determined the thirteen STR loci which must be included in every
genetic profile of an individual convicted of a number of offenses, which are determined
by each of the member states in the National CODIS database (the loci being: D381358,
THOl, D2181], D1885], VWA, D881179, TPOX, FGA, D58818, D138317, D78820,
D168539, and CSFIPO). A match with these STR loci between two samples can
produce random match probabilities in the quadrillions.

Currently, most forensic DNA analysts are utilizing the Applied Biosystems
(Foster City, CA) AmpFISTR® Profiler Plus” and CoFilerT“ genotyping kits; these kits
test multiple genetic STR loci in one reaction, and therefore are termed multiplex kits. In
this procedure, two separate runs must be set up to obtain results for all thirteen CODIS
loci. Setting up two separate analyses depletes the original sample, consumes valuable
human and monetary resources, and most importantly—these additional testing steps can
increase the chance of inadvertent sample transfers and other types of sample integrity
concerns.

With the concerns of the forensic scientist in mind, the Promega Corporation
(Madison, WI) has developed the GenePrint® Power Plexml6 multiplex kit which
determines the genetic profile of an individual at sixteen different loci, thirteen of which
are the required CODIS loci. This genotyping system will save the analyst (and the
agency for which he/She works) time, money, and sample—if it is proven to be an
effective, reliable tool for the typing of CODIS samples. The PowerPlexm16 system
would allow for faster processing of convicted offender samples, whose DNA profiles
could be entered into the CODIS database quicker, thus providing the states with another

DNA profile of which to search.

This thesis project is the evaluation of individual locus performance using
Promega’s PowerPlexTM l6 multiplex system for use with single source CODIS samples.
In this thesis, the following issues will be addressed:

1. The optimal target quantity of DNA per reaction

2. The level of performance at each locus

3. The viability of this kit for database samples

This thesis project was part of a larger project by the Promega Corporation and a

number of other forensic DNA testing facilities. The data generated in this evaluation
was included in the “STR primer concordance study” [2], which compared DNA profiles
obtained with the PowerPlexTMl6 typing kit with their corresponding DNA profiles using
Applied Biosystem’s Profiler PlusTM and CoFilerTM typing systems. Results in the STR
primer concordance study indicated that the primers used in the PowerPlex'ml6 Kit,
Profiler Plus, and CoFiler Kits produced reliable, consistent DNA typing results obtained
on reference samples.

To understand the use of STRs with CODIS and how a genetic profile is obtained,
some background material is presented on the structure and fimction of DNA, and the
history of various DNA analysis methods leading up to the most recent application of
DNA analysis—the development of STR multiplexing.

II. The structure of DNA

Deoxyribonucleic acid was first isolated in 1869 by a Swiss chemist Johann
Friedrich Miescher, but it was not until the 1950’s when James Watson and Francis Crick
combined all of the data and created a working model of DNA. DNA is composed of
subunits, with each subunit containing a nitrogenous base, a pentose sugar, and a

phosphate group. The nitrogenous bases fall into two categories: the pyrimidines (a six

member ring) and the purines (structures composed of two rings). There are four types of
nitrogenous bases that make up DNA—two pyrimidines (cytosine and thymine) and two
purines (guanine and adenine).

There are two types of nucleic acids—deoxyribonucleic acid (DNA) and
ribonucleic acid (RNA). What differentiates RNA from DNA is that the sugars contained
in the molecule RNA are riboses (not deoxyriboses, as in DNA). In addition, the bases
that make up RNA include guanine, adenine, cytosine, and uracil (instead of thymine).
Uracil base pairs with adenine (like thymine) but is structurally different than thymine
(lacks a methyl group). Afier formation of the DNA template in a cell, it is transcribed
(or read) into a specific kind of RNA molecule, which is then translated into a specific
protein. DNA is the hereditary material found in all living organisms, while RNA (or
DNA) can be the hereditary material found in viruses.

When a nitrogenous base is bonded to a pentose sugar and a phosphate group, the
structure is termed a nucleotide. Chains of nucleotides are what make up the structure of
nucleic acids. In 1953, Watson and Crick proposed that the DNA molecule was made up
of two polynucleotide chains and that each of the nucleotides was paired such that a
pyrimidine is always opposite a purine. They also determined that the proportion of
cytosine to guanine was always 1:1, as was the proportion of adenine to thymine. This
suggested that in the double helix, the cytosine was always paired with guanine (a purine
with a pyrimidine) and thymine with adenine. This is referred to as complimentary base
pairing. Watson and Crick determined that the two polynucleotide chains stay associated
in the double helix structure by hydrogen bonding of the nitrogenous bases, with three

hydrogen bonds holding cytosine and guanine together and two hydrogen bonds between

the adenine and thymine bases. The structure of the DNA double helix has often been
compared to a ladder, with the rungs of the ladder representing the nitrogenous bases held
together by hydrogen bonds.

The organization of the bases in the DNA molecule form the alphabet for the
genes contained in a particular sequence. The DNA is transcribed into an mRNA
molecule, which is then translated into proteins that are expressed by the cell. Different
forms of the same gene can occur and are termed alleles. An example of a gene with
multiple alleles is that for eye color. It is often useful to measure the variations of an
allele in a population, as it is used in forensic science to determine a frequency of an
individual’s genotype within a distinct population.

111. DNA Inheritance

In 1944, O.T. Avery and colleagues performed laboratory experiments in which
they observed the effect of injecting virulent and avirulent strains of the bacterium
Pneumococcus into mice [3]. They found that by isolating DNA fiom one strain of
bacteria, its traits could be transferred to a new colony of bacteria, thus proving DNA to
be the hereditary material for which genetic information was transferred. Not only is
DNA the hereditary material of bacteria, but for all living organisms, and it resides in the
nucleus of cells, organized into structures termed chromosomes. Humans have forty-six
chromosomes in every cell that contains a nucleus—two sets of twenty-three (one set is
paternally derived, and the other maternally). The only exception to this is that there are
only twenty-three chromosomes in the sex cells (gametes). Upon fertilization of an egg
and a Sperm, the cell then contains the full forty-six chromosomes. In addition to nuclear

DNA, the cells containing mitochondria (organelles in the cytoplasm for cellular

respiration) have their own circular DNA template termed mitochondrial DNA (mtDNA).
Mitochondrial DNA, like nuclear DNA, is inherited, though only maternally. Due to
limited DNA typing capabilities, mtDNA is generally used for the tracking of relatives
and/or when nuclear DNA typing has failed due to minimal quality or quantity.
IV. DNA Replication

A robust method for the amplification of DNA sequences is termed the
polymerase chain reaction (PCR). The principles underlying the polymerase chain
reaction involves the in viva process of DNA replication and how scientists were able to
mimic this process in vitro to produce an invaluable technique for the analysis of DNA
sequences. Since DNA is base-paired in its double helix—each strand can serve as a
template for the synthesis of a complimentary daughter strand, and each strand of the
original double helix is paired with a new strand. In other words, the parental DNA helix
replicates to form two identical daughter duplexes. This type of replication is called
semi-conservative, since one strand of the “original” is conserved in the next round of
replication. This is very similar to the process of PCR amplification—which also utilizes
.the original strands of DNA to replicate the desired sequence millions of times. One
strand of the double-stranded DNA has all the information needed to build its
complimentary strand, but requires an enzyme to catalyze the addition of these
complimentary nucleotides. In vivo, this enzyme is DNA polymerase. DNA polymerase
does not act alone; it requires a primer (3 short sequence of single-stranded DNA) to first
bind the DNA template, which is what initiates the elongation of the complimentary
strand by the polymerase. The in vitro process utilizes a similar DNA polymerase and

will be discussed in the section on PCR analysis.

V. The History of Forensic DNA Analysis

The first type of forensic DNA analysis was Restriction Fragment Length
Polymorphism (RF LP), developed by Alec Jeffries in 1985. In this process, afier DNA is
isolated, it is digested with a restriction enzyme that recognizes specific sequences in the
DNA and cleaves at these sites. The fiagrnented DNA is then separated by size using
agarose gel electrophoresis. The DNA fiagments are then transferred to a nitrocellulose
or nylon membrane (Southern blot). The membrane is then hybridized with a radioactive
or chemiluminescent probe, that identifies the alleles at one locus by complimentary
base-pair binding of the probe to the fragmented DNA. The probe can be stripped from
the membrane and a new locus can be examined. Using multiple probes on a sample
decreases the probability of another person having those alleles. RF LP analysis is
excellent for samples that are not degraded, since the fiagments obtained with this type of
assay are in the order of five hundred base-pairs and larger. RFLP analysis for the
genotyping of degraded samples can be difficult, since the detection of larger alleles
might be missed.

By 1986, Kary Mullis had invented a new typing technique, capable of analyzing
limited or degraded samples and was termed the polymerase chain reaction (PCR). This
process exponentially multiplies the amount of DNA present, therefore making it easier
to analyze. PCR amplification will be described in depth in the discussion of
mulitiplexing.

In addition to RF LP and PCR, DNA sequencing is another method which
can be used by the forensic scientist to determine the genetic profile of an individual.

Unlike the previous methods, there is no comparison of alleles in this method; the process

works by fluorescent based detection of the DNA sequence products. Though this
method provides the examiner with an exact genetic makeup of the individual, the
process is very time consuming and requires costly equipment. Until the day when the
equipment becomes less expensive and the process less cumbersome, other methods of
analysis will generally be employed.

VI. Short Tandem Repeat (STR)

One of the fastest growing methods of DNA analysis in the forensic science

 

community is the analysis of short tandem repeats. STRS are sequences in the genome
made up of approximately three to seven base pairs that repeat as a sequence a variable
number of times and occur repeatedly throughout portions of the genome. Currently,
there are an estimated thirteen hundred different STR loci [1 ], but less than two percent
of those are currently used in the forensic science laboratory. STRS are a type of
Variable Number Tandem Repeat (VN TR), meaning they contain a tandemly repeated
sequence and their fiagment size depends on how many repeats are in the sequence. An
example of an STR is the sequence “AATG”, which can be found at the THOI locus.
This sequence occurs in tandem repeats, where the sequence is repeated as a unit
(AATG-AATG). The number of repeats is responsible for the genetic variation at that
particular locus. For example, a person may have two alleles—one with Six repeats, and
one with ten, each derived from a parent. In addition, STRS fall into non-coding regions
of the genome, but they are still inherited, just like alleles coding for a particular gene.
The most recent method of DNA fragment analysis is based on fluorescence
detection of PCR products. Like DNA sequencing, this process also utilizes DNA

fragments labeled with fluorescent dyes (incorporated during PCR amplification); but

instead of obtaining the actual DNA sequence, one obtains the alleles present for specific
STR loci. One of the first instruments utilized for fluorescent detection of STRS was a
flat bed laser scanning instrument (Hitachi FMBIO), but because of the multiple
problems associated with this cumbersome method, it was soon replaced by capillary
electrophoresis (CE). As where the FMBIO required a polyacrylimide gel for separation,
CE incorporates a polymer-filled capillary (or column) for separation of fluorescently
labeled DNA. Like the FMBIO, CE utilizes a laser to excite the fluorescently labeled
DNA products (which in turn produce a spectrum of light). The different fluorescent
signals are separated according to their wavelength and are displayed by a CCD camera.
A filter allows for separation of the signals, and a matrix is incorporated to normalize the
fluorescent intensities for each dye. The ABI PRISM® 310 Genetic Analyzer (Applied
Biosystems) is an instrument that uses a capillary based separation scheme to
fluorescently detect tagged DNA fragments. The ABI 310 allows the analyst to test
multiple loci in one reaction and allows for examination of alleles with overlapping size
ranges (because of the different dyes employed). In addition, an internal lane standard
(ILS) is incorporated with every sample; this enables automated sizing of fragments
based on a curve derived from the specified ILS fragments. The ILS normalizes any
differences in electrophoretic mobility and is labeled with a different dye than all of the
other loci. This is to ensure that there is no overlap of DNA fragments with the ILS
itself. There are three different software prOgrarns associated with the 310 Genetic
Analyzer: the 310 Data Collection® software (ensures proper operation of the

instrument); Genescan® software (applies the matrix and determines the size of the DNA

fragments using the ILS; and lastly, Genotyper®/PowerTyperT“ software determines the

 

genotypes based on an allelic ladder from the same run. An allelic ladder, sequenced to
verify fragment lengths and repeat structure, is included with every PowerPlexW16
System. It contains the most common alleles for each locus and by comparisons of sizes
obtained for samples with the known sizes in the allelic ladder, a genotype can be
assigned. The size ranges of each locus are the actual base pair sizes of the sequenced
alleles. Within each locus, each complete four base pair repeat unit is designated by a
whole number, and alleles that contain a partial repeat are the whole number designation,
followed by a decimal point and the number of base pairs in the partial repeat. For
example, the D21811 33.2 allele contains 33 complete four base pair repeats and a partial
repeat of two base pairs. The exception to this method is for loci with five base pair
repeats (such as Penta D and Penta E), where the whole number designation must be a

five base pair unit.

10

The Polymerase Chain Reaction and the Development of STR Multiplexing
I. The Polymerase Chain Reaction

The polymerase chain reaction (PCR) is a technique that copies a specific
sequence of DNA and amplifies that segment exponentially so that millions of copies are
present at the end of the replication procedure. Since PCR needs only a small amount
(0.5-1ng) of template DNA to start the reaction process—this technique is ideal for the
analysis of limited or degraded samples, especially when compared to RFLP analysis,
which can consume up to five hundred nanograrns of high molecular weight DNA to
generate a genetic profile.
II. PCR operation

PCR has three main steps: denaturation, annealing, and extension of the primers.
Denaturation is the first step and takes place at approximately 95° C for one minute. This
cleaves the double stranded DNA molecule into two complementary daughter strands.
This allows the DNA to be accessible for the next step in this reaction. The second step
in the PCR process is the annealing of the primer sequences to the original DNA template
strands. Primers are designed so they are complimentary to the flanking regions of the
desired sequence, allowing them to bind specifically at those locations under thermal
condition where the exact complementary pairing of the primer to the template DNA
strand is favored. The annealing step generally occurs at 55° C to 60° C for 30-45
seconds. A delicate balance must be found to optimize a PCR reaction, a temperature too
low will allow non-specific PCR products (miss-pairing) to occur, and an annealing
temperature too high will inhibit the primers from binding to the template DNA strand.

The final step in PCR is the extension of the primers by the use of a DNA polymerase.

11

 

Taq polymerase (derived from the bacterium Thermus aquaticus) is a commonly used
thermostable polymerase which facilitates the addition of nucleotides to the extending
molecule, while maintaining its viability in temperature ranges required for PCR. The
Taq polymerase attaches to the DNA template at the primers and begins the additions of
the complimentary nucleotides to extend the primers. This step takes place at
approximately 72° C for 30 seconds. This increase in temperature will also aid in the I

dissolution of miss-paired primers from the template DNA. At the end of this step, two

 

identical copies have been made from one original double-stranded sequence. In order to L
achieve millions of copies of the particular sequence, the three steps are repeated for
about 30 cycles. This results in an exponential amplification of the original sequence (the
first cycle produces two copies from one molecule; the next cycle produces four copies
fiom the two present, and so on.).
III. Reagents/instruments needed for PCR

There are many items that work together to optimize the PCR reaction, and these
reagents are combined together in what is termed a “PCR cocktail”. A standard cocktail
contains KCl (a salt), Tris-HCl (a buffer), deoxynucleotide triphosphates (dNTPs; these
are free nucleotides that are incorporated in the extension of the molecule), Mng (the
Mg ions work with Taq polymerase and the dNTPs), BSA (bovine serum albumin helps
prevent PCR inhibitors by binding to them), and Taq polymerase. All of these reagents
must be balanced within the reaction for optimal results, usually by altering one reagent
at a time until an optimal point is reached. With commercial STR typing kits, this
optimization has already been performed and the reagents arrive pre-mixed in a reaction

buffer cocktail—all that is needed is the Taq polymerase and the template DNA. Once

12

the PCR cocktail (including the Taq) has been aliquoted and the DNA templates added,
the samples (usually 25ul reactions) are amplified in a thermal cycler. This instrument
allows for the alternate heating/cooling cycles that PCR reactions must undertake. An
important consideration in the setup of a PCR reaction is the amount of template DNA
required for optimal results. Normally, between approximately 0.5ng and 2.5ng of
template DNA produce the best results, but each kit must be tested with a range of DNA
quantities to determine the amount which produces the best results. Iftoo much DNA
template is added, then an overabundance of PCR product is generated. In fluorescent
detection of PCR products, the ramifications of an excess of PCR product can include
data containing “off-scale” peaks (there is too much fluorescent intensity for detection by
the instrument). Additionally, excess PCR product can cause “pull-up” (an artifact
caused by poor separation of the fluorescent dyes due to the excessive amount of
fluorescently labeled PCR product). On the other hand, if too little DNA is added to the
reaction, unbalanced amplification can occur at heterozygous loci (loci with two different
alleles).
IV. The first types of PCR tests developed for forensic use

Tests that identified sequence polymorphisms were the first forensic tests which
utilized PCR. Sequence polymorphisms occur when there is a mutation in one base-
pairing of a particular sequence. These polymorphisms can be identified through a test
called a reverse dot blot. In a reverse dot blot, PCR product is added to a nylon
membrane which has DNA probes attached to it. The probes and the PCR product are
complimentary sequences of the same locus—the probes are commercially available and

the PCR product is obtained by the analyst via extraction and amplification. Before PCR

13

 

product is added to the probe-coated membrane strips, the strips are white-—upon
addition of the PCR product, complimentary sequences of the probe and the product turn
blue, thus identifying the alleles in the PCR product. The most common forensic DNA
tests developed which identified sequence polymorphisms were DQa (Applied
Biosytems) and the AmpliType®PM system (Roche Molecular Systems).

In addition to tests that identified sequence polymorphisms, PCR was utilized to
amplify length polymorphisms (VNTRs), such as those occurring at the locus D1880. As
described previously, length polymorphisms contain a core repeating sequence; the
number of times the core repeats is determined by its fragment size on a polyacrylamide
gel. Fragments are compared to a molecular ladder and the alleles determined (the alleles
are numbered according to the number of repeats in the sequence). For example, an
individual could be a 10, 12 at this locus, meaning they have an allele with ten repeating
core elements, and an allele with 12 repeating core elements. Unfortunately, this method
of analysis is not highly discriminative, since it only analyzes one locus.

The most recent technology in forensic DNA analysis that applies PCR
technology is in the typing of STRS. STRS, a type of VNTR, are length-based
polymorphisms, but the core repeat is much smaller (three to seven base-pairs) compared
with the sixteen base-pair core repeat in D1880. Though one individual STR locus does
not have a significant amount of variation (usually about five to twenty alleles), STR loci
can be combined in a multiplex reaction—which simultaneously types all of the STR
loci, thereby greatly increasing the variability. Mulitplexing uses the same concept of

PCR as if it were a single locus amplification, but instead of one set of primers in the

14

PCR cocktail, there is a set of primers for each locus (the PowerPlex 7“ 16 kit thus has
sixteen sets of primers).
V. Previous Multiplex PCR Genotyping Kits

Wallin and coworkers [4] developed one of the first commercially available
multiplex kits available for forensic identity casework. It genotyped three STR loci—
D3Sl358, vWA, and F GA, and was called the AmpFlSTR® Blue PCR Amplification
Kit. Blue referred to the color of the dye (S-FAM-S-carboxyfluorescein) used to
fluorescently label the primers. Instead of visualizing the PCR product on an acrylamide
gel, capillary electrophoresis was utilized to detect the alleles present. As described
earlier, capillary electrophoresis is a separation technique in which the DNA sample is
carried through a capillary and the fluorescently labeled fi'agments are excited by a laser
at the end of the capillary journey. The fluorescence is collected and focused by mirrors
onto a CCD camera and visualized as an electropherogram generated by the instrument.
Two instruments used for the detection of fluorescent PCR products are the ABI
PRISM® 310 Genetic Analyzer (Applied Biosystems), a capillary electrophoresis
instrument, and the ABI PRISM® 377 DNA Sequencer (Applied Biosystems), an
instrument which utilizes slab gel electrophoresis. GeneScan® software is utilized to
analyze the electropherograms; it applies a mathematical matrix model to separate
emissions from different dyes in the sample being analyzed, and it creates and applies a
size curve to the data generated from the sample being analyzed. The alleles from the
DNA sample are represented as peaks on a graph, with the size of the peak being related
to how much DNA is present in the amplified sample. The information that can be

derived from an electropherogram includes: the alleles in the sample (peaks), relative

15

 

amount (measured by relative fluorescence units (r.f.u.)), and the time with which the
laser detected them (or that they emerged through the capillary). As in other separation
techniques, the smaller fragments elute quicker than the larger fragments. In addition to
the DNA samples run on the 310 or 377, an internal lane standard (1L8) is also run with
each DNA sample. This standard contains labeled fragments of known lengths and thus
provides a manner in which all of the peaks in the run can be determined. Lastly, an
allelic ladder, containing all of the common allele types for the loci is run in a separate
injection, so that the alleles in the DNA sample can be compared and determined, using
either manual determinations or automation software. GenotyperT“ (Applied Biosystems)
is the software program available to perform this task; it uses the allelic ladder to create
allelic matching windows, which are applied to all of the samples in the run. In Wallin’s
validation of the AmpFlSTR® Blue Kit, the 377 DNA Sequencer and the ABI PRISM®
310 were utilized and Genescan® software was utilized for analysis. The loci chosen
for the AmpFlSTR® Blue kit were chosen because the overall size ranges (per locus)
were very small, hence, the primers produced short PCR products (one hundred to three
hundred-fifty basepairs). This is ideal for samples containing degraded DNA, and their
sizes minimized the occurrence of preferential amplification. Preferential amplification
(as defined by Wallin) occurs when there is a difference in amplification of two alleles
within the same locus (evident upon comparison of peak heights of heterozygotes).
Preferential amplification can lead to a loss of information regarding that locus, and a
mistyping could possibly occur. To avoid preferential amplification, STRS with small
size ranges are generally employed (as with the Blue kit). In their validation studies,

Wallin and coworkers noted differential amplification with increasingly degraded DNA,

16

evident by the drop-out of the largest locus (FGA). Wallin described differential
amplification as the occurrence of one or more loci amplifying less than the other loci—a
difference in amplification between loci. Differential amplification may also result in a
loss of information regarding a particular locus, though no mistyping can occur.
Differential and preferential amplification must be considered when developing and
validating multiplex PCR kits. Though a kit may be optimized for minimal preferential
and differential amplification, other factors can promote the same effects, such as
degraded DNA and the presence of inhibitors (soil (which contains metal ions, enzymes,
and other proteins), bleach, dyes, etc.). There is a way to determine if something is
inhibited versus being degraded. In degraded DNA samples, one gets more signal
(higher r.f.u.) when more DNA is added. In the case of an inhibited sample, if it is
diluted, it may promote a more efficient amplification. It is important to determine if the
kit itself promotes preferential/differential amplification, since many of the samples
forensic analysts encounter are either degraded or inhibited. Overall, the AmpFlSTR®
Blue kit had minimal preferential/differential amplification. In addition to testing the
occurrence of differential/preferential amplification, Wallin also compared single locus to
multilocus performance. For the single locus amplifications, individual primers were
used and only one locus per reaction was employed. They obtained the same genotyping
results for both systems and no significant difference in peak heights were noted. They
concluded that the single locus reaction had no benefit over the multiplex reaction.
During this same time, Micka and coworkers [5] had also been validating two
multiplex kits which each typed three STRS. Like Wallin, they determined that the

performance of a single monoplex reaction had no advantage over the use of the

17

multiplex kits. This led researchers to develop larger multiplex kits——such as the one
developed by Lins [6]. Lins and coworkers developed an eight-locus, two-color STR
multiplex system for human identification. This eight-plex combined two four-plex
systems: the CTl‘v and GammaSTRT“ multiplex system. The CTTv multiplex
genotyped STRS at the following loci: CSF 1P0, TPOX, THOI, and vWA, while the

GammaSTRT" system genotyped D168539, D78820, D138317, and D58818. When

combined together, the multiplex was termed the PowerPlex” System. The Power

 

PlexI" System combined the primers for CTTv and GammaSTRTM loci. Originally, all
primers were labeled with fluorescein (FL—recognized as a blue dye), but to achieve
better separation, the CTI‘v multiplex was changed to contain primers labeled with
carboxy-tetramethyl rhodamine (TMR—recognized as a yellow dye). This provided
better resolution of alleles within the one hundred to four hundred base-pair range.
Though labeled with different dyes, the eight sets of primers were combined into one
reaction mix. In addition, allelic ladders were developed, which contained all of the
alleles for both systems (both the FL and TMR labeled products), to aid in the .
determination of allele calls. The primers were modified to eliminate artifactual bands
due to incomplete terminal addition of adenine, also referred to as —A or +A. With
incomplete terminal nucleotide addition, two peaks are present, usually of unequal peak
height. One peak represents the allele without the terminal adenine, and the other peak
represents the STR with the terminal nucleotide addition. In this case, the primers were
modified to help minimize this occurrence. In addition, a thirty minute extension at sixty
degrees Celcius following the PCR cycling process reduces —A/+A. This process

promotes full terminal nucleotide addition to one hundred percent—therefore only one

18

peak is present on the electropherogram. In the development of this system, Lins chose
STRS with low mutation rates and with few microvariants. These factors contribute
greatly to the reliability of the genotyping of the alleles in the system. If an STR has a
high mutation rate, or has many variants of the common alleles, the allelic ladder cannot
be used by the computer software to accurately determine the allele types. Also, the
STRS that were chosen by Lins had relatively low stutter. Stutter is the occurrence of a
smaller band, or peak, one repeat smaller (or larger) than the primary band. A peak is
deemed stutter if it is approximately ten to fifteen percent of the primary peak (this
percentage differs between loci). One theory, supported by Walsh [7], explains that the
occurrence of stutter is due to slipped strand mispairing. During PCR, Taq polymerase
may fall off, giving one strand a chance to loop out before they bind again. Once
together again, one of the strands is shorter by one repeat unit. Stutter is a reproducible
artifact and does not interfere with genotyping of a particular sample, unless the sample is
a mixture. That is why it is important to choose STRs with low stutter percentages, since
many samples encountered in forensic science analysis are mixtures and deciphering the
genotypes would be difficult if stutter were high. Lastly, Lins combined the eight STR
multiplex system with the primers for the Amelogenin locus to obtain gender
information. Amelogenin was labeled with TMR and produced specific fiagments for the
X and Y alleles.

After deveIOpment of an STR multiplex, such as the PowerPlex” System, it must
undergo validation studies to be used for forensic science casework. Micka [8] validated
the GenePrint® PowerPlex” 1.1/Amelogenin System developed by Lins and coworkers

(for the Hitachi F MBIO Fluorescent Scanner). As previously described, this system

19

contained two groups of four STRS—each group labeled with different fluorescent dyes.
Additionally, the locus for gender identification was also incorporated. Micka’s results
indicated no differential amplification (allelic drop—out), or any other artifactual bands.
Other artifactual bands may include primer-dimers (when the primers bind to themselves)
or partial binding of the primer to a complimentary sequence in the DNA. In addition,
Micka did single locus versus multi-locus amplification studies. In both cases, the
systems produced the same genetic typing results.

The development of STR multiplexing kits has made life for the forensic DNA
analyst a little bit easier. Previous kits have shown that multiplexing produces the same
genotyping results as single locus amplifications——meaning one can obtain genetic typing
results for many loci in one reaction, rather than setting up individual reactions for each
locus. This saves sample amount, decreases the chance for contamination, and allows for
quicker genotyping results. Like the multiplexing kits that came before it, the
PowerPlex“ l6 kit would be another useful tool for the forensic scientist. It allows the
genotyping of the thirteen required CODIS loci to be analyzed in one reaction, instead of
setting up two different reactions, as is the current method. To determine whether the
PowerPlex“ 16 system is suitable for single source samples, such as those submitted for

CODIS, evaluations on locus performance must first be made.

20

Materials and Methods

Samples: Promega Corporation (Madison, WI) provided three of the DNA extracts
used in this project: B15 (lng/ul), H9 (lng/ul), and standard DNA template 9947A (10
ng/ul). They had previously been quantified by Promega.

The Michigan State Police DNA/Forensic Biology Unit (MSP) provided a blood
stain of known origin and genotype (MBILC). This blood stain sample required an
organic extraction. To perform the extraction, small cuttings of the bloodstain were
placed in a Spin-Ease (Gibco BRL) extraction tube. Approximately 400111 stain
extraction buffer (SDS, EDTA, NaCl, and Tris) and 20ul of Proteinase K (BRL) were
added. After an overnight incubation at 56° C, the cuttings were removed and
approximately 400ul phenol/chloroform/isoamyl alcohol (24:1 :1) was added. After
vigorous vortexing, the sample was then spun down at 14,000 x g to facilitate the
separation of the aqueous and organic layers. The upper layer (the aqueous layer
containing the DNA) was transferred into a Centricon® 100 Concentrator (Millipore
Corp.). The Centricon 100 employs a size exclusion membrane which retained the large
molecules (the DNA) and allowed for passage of the smaller ones (such as salt ions,
detergents, and fragmented proteins). The sample was then rinsed with approximately
400ul TE'4 buffer and spun at 500 x g for 15 minutes. This step was repeated three times.
Lastly, the DNA that was retained by the membrane was captured by inverting the
sample reservoir and spinning the sample at 1000 x g for 3 minutes. The DNA recovered
was then transferred to a new 1.5m] microcentrifuge tube for long term storage at —20 C.
Quantitation: Upon completion of the extraction and purification of the DNA, the

sample was assessed for quality and quantity using a yield gel (an ethiduiurn bromide

21

fortified 1% agarose gel). The yield gel contains a Visual marker (lambda
HINDIII/ECORI) and human DNA quantification standards (BRL) to which one can
compare extracts to determine the relative amount of DNA present. In addition, the yield
gel will identify whether the sample is degraded (observable by a smear instead of a band
when the gel is subjected to short wave UV radiation). Following detection using the
yield gel, slot blot quantitation using the Applied Biosystems QuantiBlot® Kit was
performed to determine a more accurate representation of the DNA present (this method
utilizes the knowledge obtained fiom the yield gel to make an approximate dilution for
the slot-blot). This quantitation method entailed the immobilization of denatured target
DNA on to a charged nylon membrane (Pall-Biodyne B). In addition, this method is
species-specific by utilizing a primate specific DNA probe (D17Z1) in the hybridization
process. Once again, this kit also contains human DNA standards, which are compared
to the unknown samples to determine an approximate quantity of DNA in the sample.
Chemiluminescent detection utilizing ECL (Amersham Pharrnacia) and Kodak XLS film
was utilized to observe the results of the QuantiBlot.

MSP also provided DNA extracts of approximately 114 individuals for analysis
using the Geneprint® PowerPlex” 16 System (Promega Corp.) The extracts were
quantified using the Applied Biosystems Quantiblot® Kit prior to analysis using the
PowerPlex”I 16 kit.

PCR Amplification: Dilutions: Upon extraction of MBILC, dilutions of 2ng/1 0ul,
Ing/IOul, 0.5ng/10ul, 0.25ng/10ul, 0.125ng/10u1, and 0.0625ng/10u1 were set up for each
of the samples (except for 9947A at 0.0625ng/10ul). Dilutions were made with 0.21m

filtered high purity (18MQcm'l) water and placed in new microcentrifirge tubes. The

22

dilutions were based on Promega’s quantitations of B15, H9, 9947A, and the quantitation
of MBILC performed in this study.

Amplification Set-up: Promega Corp. provided the genotyping kits (GenePrint®
PowerPlex" 16 System) utilized in this project. For pre-PCR, the kit includes a buffer
(Gold ST’R 10X Buffer), fluorescently labeled primers (PowerPlexm 16 10X Primer
Pair Mix), and a standard DNA template (9947A). A master mix was set-up containing
nuclease-free water, buffer, primers, and a thermostable DNA polymerase, AmpliTaq
Gold® (Applied Biosystems). 10ul of sample were added to 15ul of master mix.
Negative and positive controls were set up with each amplification run.

Reactions of 25ul were set up in MicroAmp®(Applied Biosystems) reaction tubes
and amplified using the Applied Biosystems GeneAmp® PCR System 2400 Thermal
Cycler. The cycling protocol (from the Geneprint®PowerPlex'ml 6 Technical Manual
[9]) was as follows:

95° C for 11 minutes

96° for 1 minute

Ramp 100% to 94° C for 30 seconds
Ramp 100% to 60° C for 30 seconds
Ramp 23% to 70° C for 45 seconds
Repeat this for ten cycles, then:

Ramp 100% to 90° C for 30 seconds
Ramp 100% to 60° C for 30 seconds
Ramp 23% to 70° C for 45 seconds
Repeat this for 22 cycles, then:

60°C for 30 minutes
4°C soak

Upon completion of amplification, the samples were placed in a —20° C freezer in the

post-amplification room.

23

Capillary Electrophoresis: Instrument Preparation: Before every run on the ABI
PRISM® 310 Genetic Analer (Applied Biosytems), the instrument was thoroughly
cleaned using distilled water and dried using lint fiee wipes and compressed air. This
was to prevent the occurrence of fluorescent spikes (which may be due to crystallized
polymer) and to keep the polymer and buffer fresh throughout the run. In addition, a
matrix was generated on the Genetic Analyzer in order to analyze samples run on that
specific instrument. Promega provided the GenePrint® Matrix F L-J OE-TMR-CXR for
matrix standardization.

Sample Preparation: Once removed fiom the —20° C freezer, the samples were thawed,
vortexed, and centrifirged. A loading cocktail was set up containing 24ul/sample of
deionized formamide (Ultra Pure Grade, Amresco) and lul/sample of Internal Lane
Standard 600 (ILS 600). lul of DNA sample was then combined with 25ul cocktail. In
addition, PowerPlex” 16 Allelic Ladders were set up (two per run). The samples were
then denatured (heated for 3 minutes at 95° C and immediately cooled in an ice bath for
approximately 3 minutes). The samples were then loaded onto the 310 Genetic Analyzer.
The following parameters, found in the ABI 310 Collection Software, were set with

accordance to the parameters specified in the PowerPlexTM16 Technical Manual.

Injection time: 3 seconds
Injection kV: 15.0

Run kV: 15.0

Run °C: 60° C

Run Time: 30 minutes

24

In addition, the “GS STR POP4 (1ml)A” module was employed.

Additional analysis parameters were specified by the PowerPlexml6

Technical Manual:

Analysis Range Start: 3200
Stop: 10000

Data Processing Baseline: Checked
MultiComponent: Checked
Smooth Options: Light

Peak Detection Peak Amplitude Thresholds:
Blue: 50-150 Yellow: 50
Green: 50 Red: 50
Min. Peak Half Width:

Size Call Range Min: 60
Max:600

Size Calling Method Local Southern Method

Split Peak Correction None

To assign a new Size standard for the [LS 600, the peaks were labeled according to Figure
1 Panel D of Section VIII.D in Promega’s Technical Manual for the GenePrint®
PowerPlexTM 16 System. Both the PowerPlexTM 16 matrix and size standard were saved
on the hard drive of the Maclntosh computer for use throughout the project.

Data Analysis: Samples were analyzed using GeneScan® /Genotyper® software
(Applied Biosystems) using a Macintosh G3 233 MHz (for 310 Collection Software) and
a Macintosh G3 350 MHz (for analysis software). These software programs were used in
conjunction with the PowerTyperT" 16 macro (Promega Corp.). Allele tables were
generated containing the allele call, relative fluorescent units, and the fragment size (base
pairs) and exported out to files in Microsoft® Excel (contained in a DELL OptiPlex GXI

400MHz P2 system).

25

Results and Discussion

1. Determination of an optimal target quantity of DNA

The first goal of this project was to determine the optimal target quantity of DNA
per reaction, based on a sensitivity assay using the samples 9947A, MBILC, B19, and
H9. Six target quantities of DNA (0.0625ng, 0.125ng, 0.25ng, 0.5ng, 1.0ng, and 2.0ng)
were amplified using the PowerPlexTM16 System. To determine the optimal target
quantity of DNA per reaction, the peak heights (R.F.U.) obtained at each target DNA
amount were compared. A peak height threshold of 150 R.F.U. was utilized; this is the
minimum threshold that the Michigan State Police (MSP) utilize for reporting alleles.
Anything below 150 R.F.U. would not be reportable (this threshold is represented by a
horizontal line in Figure 1). The Technical Manual for the PowerPlexTM16 System
recommends optimal peak heights less than 2000 R.F.U. Table 1 (Average Peak Heights
(R.F.U.) for 9947A, MBILC, BIS, and H9 at various DNA template amounts) reveals
that for the target DNA quantity of 0.0625ng, the average peak height was 165 R.F.U.—
only 15 R.F.U. greater than the threshold of 150 R.F.U. Similarly, the average peak
height at 0.125ng is still only 345 R.F.U. The average peak height for 0.25ng of target
DNA was 598 R.F.U., which was a little more reasonable and well above the threshold.
At 0.5ng target DNA, the average peak height was 1208 R.F.U., considerably higher than
the value at 0.25ng, but still under the optimal limit of 2000 R.F.U. recommended by
Promega. It is important to note, though, that the homozygote alleles were halved and
counted twice, so that the average R.F.U. for homozygotes at 0.5ng might be closer to
2400 R.F.U., which is still only 400 R.F.U. greater than Promega’s suggested optimal

R.F.U. The average peak heights obtained for 1.0ng (2003 R.F.U.) and 2.0ng (2810

26

R.F.U.) were above the recommended limit of 2000 R.F.U., and those would only be the
heterozygote heights—the homozygote heights would be approximately twice the R.F.U.
For this reason, the target DNA amounts for 1.0ng and 2.0ng are not shown in Figure l:
RF.U. vs Target DNA. By analyzing peak height data alone, the DNA template amounts
of 0.25ng and 0.5ng provided the best results. In addition to observing peak heights,
heterozygote peak height ratios (PHR) were calculated to evaluate the amplification of
heterozygote alleles within a locus. The PHRs were compared to a threshold PHR of 0.7
(a value MSP utilizes in the comparison of heterozygote alleles). Table 2 (Peak Height
Ratios of Heterozygotes at Different Template Amounts) shows the average PHR
obtained for 0.0625ng was only 0.599504—clearly there was not enough DNA template
available to effectively amplify each of the alleles. Average PHRS of 0.76975 and
0.774011 were obtained for template amounts of 0.125ng and 0.25ng. The PHRs were
above 0.70, indicating the template amounts were sufficient in amplifying heterozygote
alleles, but the closer the ratio is to one, the more efficient the amplification. At 0.5ng
template DNA, the PHR calculated was 0.847903—considerably greater than the average
PHRs obtained for 0.125ng and 0.25ng. For the 1.0ng and 2.0ng template amount, PHRs
of approximately 0.88 were obtained, which indicates the most efficient amplification of
heterozygote alleles were at these template amounts. When determining the optimal
quantity of DNA per reaction, both the average peak heights and the peak height ratios
were considered. Based on having an average peak height of 1208 R.F.U. (~2400 for
homozygotes) and a peak height ratio of 0.847903, 0.5ng template DNA was chosen to

be the optimal quantity of DNA per reaction.

27

Table 1: Average Peak Heights (R.F.U.) for 9947A, MBILC, B15, and H9
at various DNA template amounts.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

[Locus 0.0625ng sizing 0.250n1 0.50ng 1.0n1 2.0ng__
[0381358 136 182 385 838 1562 2691
[Tl-101 115 182 337 771 1900 2777
[02181 1 195 355 519 1150 2143 2959
[018851 148 377 613 1269 1691 2228
[Penta E 177 353 588 1 120 1373 1605
[058818 106 238 427 881 1636 2635
[0138317 105 274 459 944 1909 2984
[078820 168 359 522 1075 1845 2563
[0168539 178 405 723 1408 2214 3151
[C8F1PO 232 509 733 1380 2199 3053
[Penta D 203 442 863 1391 1907 2125
[MA 122 272 499 1033 1887 3213
[0881 179 134 343 627 1633 2881 3952
[TPOX 300 432 840 1546 2510 3181
FGA 163 454 828 1683 2393 3026
Ave Peak Height

Per DNA Template

Amount 165 345 598 1208 2003 2810
Standard Deviation 53 98 169 292 394 551

 

 

 

 

 

 

 

 

 

Note: Homozygote peak heights were halved and counted twice.
Peak heights have been rounded to whole numbers.

Table 2: Peak Height Ratios of Heterozygotes at Different Template
Amounts (9947, MBLIC,B15, 8 H9)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

[tacos 0.0625n 0.125% 0.25ng 0.5LnL 1.1m 2.0ng
[0381358 0.61869 0.727 0.851667 0.843133 0.906 0.8222-
[T1101 0.670033 0.757475 0.6432 0.957225 0.804025 0.945675
[021811 0.536567 0.839133 0.8977 0.815533 0.911633 0.8427
[018851 0.505467 0.896525 0.873225 0.89955 0.89535 0.827275
[Penta E 0.645867 0.638525 0.744675 0.8658 0.83725 0.7833
[058818 0.5327 0.81115 0.6761 0.652 0.9226 0.96835
[0138317 0.834233 0.768933 0.892567 0.8533 0.872867 0.958233
[078820 0.646033 0.861475 0.87225 0.75435 0.809525 0.83665
[0168539 0.688133 0.765075 0.746975 0.794775 0.917775 0.85485
[CSF1PO 0.32945 0.82165 0.652825 0.923075 0.82575 0.8988
[PentaD 0.631533 0.554367 0.664333 0.830567 0.903367 0.8856
hIWA 0.77765 0.946167 0.788567 0.951467 0.951033 0.916767
0881179 0.5055 0.77 0.90945 0.8802 0.9414 0.9277
TPOX 0.567 0.6181 0.72685 0.87195 0.91425 0.9333
FGA 0.5037 0.770675 0.669775 0.825625 0.905725 0.873225
Average

Peak Hiight Ratio 0.599504 0.76975 0.774011 0.847903 0.887903 0.884975
Standard Deviation 0.12382 0.10469 0.100545 0.077821 0.047036 0.056151

 

 

 

 

 

 

 

 

Note: Homozygote alleles were omitted.

28

 

 

Figure 1: R.F.U. vs Target DNA, (a) Sample 9947A; (b) Sample MBILC;
(c) Sample 815; and ((1) Sample H9

 

R.F.U. vs Target DNA

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Sample 9947A
2000
1750
1500
_ 1250
E; 1000
x 750
500
250
0
0 0.125 0.25 0.375 0.5
Target DNA (119)
(a)
R.F.U. vs Target DNA
Sample MBILC
1750
I
1500 r #—
' t
1250 ,
I 8
5 1000 x WT ‘ i
u; ’ vi
a: 750 ~ -- 1%» A +
500 fig ~—
250
O V I T V
0 0.125 0.25 0.375 0.5
Target DNA (ng)
PANAMA

 

 

 

(13)

Note: Homozygote alleles were halved and counted twice.

29

 

Figure 1: (cont)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

R.F.U. vs Target DNA
Sample B15
2000
1750
1500
a, 1250
5 1000
750
500
250
0
0 0.125 0.25 0.375 0.5
Target DNA (ng)
(c)
R.F.U. vs Target DNA
Sample H9
1750
I
1500 »— I A
1250 a
5 1000 i
:i 750 I
E“ ', I I. i
500 - -- -~ -——e
I
25° _‘_.i [
0 T . . ,
0 0.125 0.25 0.375 0.5
Target DNA (ng)

 

 

 

 

 

 

(d)

Note: Homozygote alleles were halved and counted twice.

30

Figure 2: Peak Height Ratio vs Target DNA (9947A, MBILC, 815, & H9)

 

 

Peak Height Ratio vs.Target DNA
(9947A, MBILC,B15,&H9)

9.0
coco-s

Peak Height Ratio
o
01

0 0.25 0.5 0.75 1
Target DNA (ng)

 

Note: Peak height ratios were calculated with the exclusion of homozygote alleles.

31

 

 

J

II. Observation of performance at each locus

After the optimal target quantity of DNA was determined to be 0.5ng, a set of one
hundred and fourteen samples were run (with 0.5ng template DNA) with the
PowerPlexm16 Amplification System. The results are shown in Figure 3: Heterozygote
Peak Height Ratio vs Locus, Table 3: Heterozygote Peak Height Ratios, Table 4:
Average Peak Heights (R.F.U.) Per Locus, and Figure 4: R.F.U. vs Fragment Size (Per
Locus). The second goal of this project was to evaluate the level of performance at each
of the PowerPlexTM16 loci (excluding Amelogenin). To observe locus performance, one
can determine heterozygote peak height ratios (PHR) for each locus and compare them to
an optimum value (in this project, heterozygote peak height ratios were compared to a
value of 0.70). The MSP DNA Unit utilized this value in validation studies of previous
genotyping kits. Evaluating PHRS helps identify whether a DNA genotyping kit has the
tendency to preferentially amplify alleles (this occurs when there is a difference in
amplification of two alleles within the same locus). Preferential amplification is evident
when two alleles within a locus differ greedy by peak heights—hence, by PHR. An ideal
PHR is equal to one (or 100%), meaning the peaks are exactly the same height. The
smaller the values are from 100%, the greater the difference in heterozygote peaks. MSP
has determined in previous studies that an acceptable PHR for heterozygotes is 0.70
(70%) or greater. It is useful to evaluate PHRS with a new genotyping kit, especially if
the kit will be used in analyzing casework samples (which are often degraded and contain
mixtures). Since there should not be any mixtures or degradation with single-source
database samples, it may be a little easier to evaluate the viability of this genotyping kit.

From Table 3: Heterozygote Peak Height Ratios, it is evident that the average PHRS for

32

 

the fifteen loci range from 0.800437 (Penta E) to 0.876938 (TPOX). This range of
average PHRs is well above 0.70, indicating that there is minimal preferential
amplification occurring at each locus. In addition, the percentages of PHRs that fell
under 0.70 were also calculated. Percentages ranged from 2.2% (T'HOI) to 23% at Penta
E. These percentages reveal that though the average PHRs are above 0.70, there is some
preferential amplification occurring. This also shows that a threshold of 0.70 for
heterozygote PHRs may be too high for this particular genotyping kit, and that a value
slightly lower than 0.70 may be more suitable.

Another method to evaluate locus performance is to compare peak heights
(R.F.U.) and fragment sizes of the alleles at each locus. Graphs showing the relationship
of peak height and fragment size were generated (Figure 4: R.F.U. vs Fragment Size (Per
Locus)). From these graphs, one can determine whether differential amplification is
evident (the occurrence of larger sized alleles amplifying less than the smaller sized
alleles). Often, in multiplex PCR kits, the larger sized alleles (or loci) get amplified less
than the smaller fragments—which results in lower peak heights for the larger alleles. In
this project, minimal differential amplification was observed at each locus and between
loci, as can be observed in Figure 4———the graphs which compare peak height and
fiagment size, and Table 4: Average Peak Heights Per Locus. Some of the graphs did
produce a slight downward trend, indicating a reduction of peak height with an increase
in size (D1885 1 , Penta E, and FGA). To better evaluate the relationship of peak height
and fiagment size, more samples should be run so that an equal representation of all

alleles would be present.

33

Lastly, Table 4: Average Peak Heights (R.F.U.) Per Locus compares the average
peak heights at each locus. Peak heights ranged from 962 R.F.U. (Penta E) to a height of
1819 R.F.U. at D881179. As noted previously, Promega suggests an optimal peak height
under 2000 R.F.U. The average peak heights obtained for this project fell within that
optimal range, though it is important to note that the homozygote alleles were halved and
counted twice. Therefore the range for homozygotes may be closer to 1800-3600 R.F.U.

(a little higher than recommended by Promega).

34

Figure 3: Heterozygote Peak Height Ratios vs Locus

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

HotsmzygotePeak Height RatiosvsLocus o 1=D3S1358
a 2=TH01
A 3:021s11
x 4=018851
x 5=PentaE
3 s 6-053818
& + 7-0133317
3g - 88078820
g - 9:0183539
a: o 10=CSF|PO
& I 11=PentaD
A 12=vWA
0W3 . x13=oes1179
123456789101112131415 ”“me
PowerPlex 16 Loci (see legend) . ngGA
Table 3: Heterozygote Peak Height Ratios
Locus 0331358 T1101 021311 018351 PentaE 053818 0133317
Samples 88 91 90 97 100 90 88
Avera 8 0.871856 0.875807 0.868036 0.848737 0.800437 0.857274 0.864429
Median 0.891664 0.882656 0.877778 0.863177 0.821655 0.880224 0.863921
Std Dev 0.093666 0.081407 0.079097 0.110684 0.133053 0.10576 0.092161
<70% 5.68% 2.20% 4.44% 10.31% 23% 6.67% 4.55%
073820 0168539 CSFtPO Penta 0 MA 0881179 TPOX FGA
97 93 82 100 96 88 73 95
0.87198 0.853188 0.851628 0.859954 0.872133 0.866916 0.876938 0.845391
0.882129 0.865478 0.862699 0.8744 0.878005 0.882867 0.891164 0.859031
0.08237 0.097068 0.108027 0.090788 0.087882 0.097309 0.097608 0.099755
4.12% 10.75% 10.98% 9% 3.13% 7.95% 6.85% 8.42%

 

 

 

 

 

 

 

 

 

Note: Values have been calculated with the exclusion of homozygotes.

Average: The average value of peak height ratios for that locus.

Median: The number In the set of values which has one half of the values
hlgher and one half lower.
Std. Dev: The standard deviation measures how wldely values are dispersed
from the average value.

(70%: The percentage of peak height ratios that fell under 0.70.

35

 

 

Figure 4: R.F.U. vs Fragment Size (Per Locus)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0381 358
R.F.U. vs Fragment Size
(1 14 samples)
3000
2500 e Allele 1 1
:- 2000 lAllele 13
n; 1500 A Allele 14
0‘ 1000 e Allele 15
500 x Allele 16
0 e Allele 17
100 110 120 130 140 150 +Allele18
Fragment Size (bp)
THO1
R.F.U. vs Fragment Size
(114 samples)

-- e Allele 6

I Allele 7

A Allele 8

e Allele 9
. . . x Allele 9.3

150 160 170 180 190 200 ' ”'9'9 1°
Fragment Size (bp)

 

 

 

 

 

Note: Homozygote alleles were halved and counted twice.

36

 

Figure 4: R.F.U. vs Fragment Size (Per Locus)

 

D21 S1 1
R.F.U. vs Fragment Size

(1 14 samples) ’ N'e'e 2‘5
I Allele 27

A Allele 28

x Allele 29

x Allele 30

e Allele 30.2
x Allele 31

- Allele 31.2
- Allele 32
200 210 220 230 240 250 260 0 Alllele 32.2
- Allele 33.2
A Allele 34.2

 

   

Fragment Size (bp)

 

 

 

 

 

 

 

 

 

 

018851
. R.F.U. vs Fragment Size
(1 14 samples)

 

e Allele 11
3000 lAllele 12
2500 AAlleie 13

. 2000 IAllele 14
E 1500 xAllele 15
F 1000 eAllele 16

500 +Allele 17
o -Allele 18
280 290 300 310 320 330 340 350 360 370 xAllele 19

eAliele 20

I Allele 22

 

Fragment Size (bp)

 

 

 

 

 

 

Note: Homozygote alleles were halved and counted twice.

37

Figure 4: R.F.U. vs Fragment Size (Per Locus)

 

Penta E
R.F.U. vs Fragment Size
(1 14 samples)

 

 

 

 

 

 

 

 

 

o I T i I T I I T I I
370 380 390 400 410 420 430 440 450 460 470 480

Fragment She (bp)

 

 

 

eAllele 5

IAllele 7

AAllele 8

xAlleIe 9

xAllele 10
eAlleIe 11
+Al|ele 12
eAllele 13
xAllele 14
eAlleie 15
IAllele 16
AAllele 17
“Web 18
oAllele 22

 

 

 

 

 

058818
R.F.U. vs Fragment Size
(1 14 samples)

3000
2500
5 2000
u; 1500

500

 

110 120 130 140 150 160
Fragment Size (bp)

 

Note: Homozygote alleles were halved and counted twice.

38

 

e Allele 7
I Allele 8
AAllele 9
-Allele 10
)KAllele 11
OAllele 12
+ Allele 1 3
O Allele 14

 

 

 

 

Figure 4: R.F.U. vs Fragment Size (Per Locus)

 

 

 

 

 

 

 

 

Fragment Size (bp)

 

 

 

 

D1 3831 7
R.F.U. vs Fragment Size
(1 14 samples)
3500

3000 e Allele 8

=5 20533 I Allele 9
I; 1500 A Allele 10
1000 I Allele 1 1
500 x Allele 12
0 e Allele 13
160 170 180 190 200 210 +Alle|e 14

Fragment Size (bp)
078820
R.F.U. vs Fragment Size
(1 14 samples)
3000

2500 e Allele 7

. 2000 I Allele 8

a 1500 A Allele 9
(i 1000 IAllele 10
500 x Allele 1 1
0 eAlIeie 12
210 220 230 240 250 + “'9"? ‘3
- Allele 14

 

 

Note: Homozygote alleles were halved and counted twice.

39

 

 

 

 

Figure 4: R.F.U. vs Fragment Size (Per Locus)

 

 

 

 

 

 

 

 

 

 

 

0168539
R.F.U. vs Fragment Size
(1 14 samples)
3000
2500 e Allele 8
. 2000 I Allele 9
E 1500 A Allele 10
.5 1000 I Allele 11
500 x Allele 12
0 e Allele 13
260 270 280 290 300 310 * N'e'e ‘4
Fragment Size (bp) -Al|ele 15
CSF1 P0
R.F.U. vs Fragment Size
(1 14 samples)
3000 _______..
2500 e Allele 8
. 2000 I Allele 9
a 1500 AAllele 10
g 1000 I Allele 11
500 x Allele 12
0 e Allele 13
320 330 340 350 360 * N'e'e ‘4
- Allele 15

 

 

 

 

Fragment Size (bp)

 

Note: Homozygote alleles were halved and counted twice.

40

 

Figure 4: R.F.U. vs Fragment Size (Per Locus)

 

 

 

 

 

 

 

 

 

 

 

 

 

Penta D
R.F.U. vs Fragment Size
(1 14 samples)

0 Allele 5

3000 I Allele 7

2500 A Allele 8

. 2000 O Allele 9
E 1500 x Allele 10
500 + Allele 12
o - Allele 13
370 380 390 400 410 420 430 440 450 I Allele 14
Fragment Size (bp) ° Allele 15
I Allele 16

vWA
R.F.U. vs Fragment Size
(1 14 samples)
3500
g l e Allele 14
=5 2000 I Allele 15
u, AAllele 16
“‘ 13% I Allele 17
500 * xAllele 18
0 IAllele 19
120 130 140 150 160 1 70 180 X Allele 20
Fragment Size (bp)

 

 

Note: Homozygote alleles were halved and counted twice.

41

 

 

 

 

 

Figure 4: R.F.U. vs Fragment Size (Per Locus)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Fragment Size (bp)

 

 

 

 

D881 1 79
R.F.U. vs Fragment Size
(1 14 samples)

0 Allele 8

I Allele 9
A Allele 10
. I Allele 1 1
x Allele 12
e e Allele 13
I + Allele 14
200 210 220 230 240 250 * £39 1:

I e
Fragment Size (bp) . Allele 17
TPOX
R.F.U. vs Fragment Size
(1 14 samples)
3000
2500

. 2000 e Allele 6

i! 1500 I Allele 8

m‘ 1000 A Allele 9
500 I Allele 10
0 X Allele 1 1
260 270 280 290 300 ' N'e'e ‘2

 

 

 

 

 

Note: Homozygote alleles were halved and counted twice.

42

 

 

 

 

 

Figure 4: R.F.U. vs Fragment Size (Per Locus)

 

FGA
R.F.U. vs Fragment Size
(1 14 samples)

3500
3000
. 2500
a 2000
m- 1500
1000
500

O
320 330 340 350 360 370 380 390 400

Fragment Size (bp)

 

 

 

e Allele 18
I Allele 19
A Niels 20
x Allele 21
x Allele 22
e Allele 23
+ Allele 23.2
- Allele 24
I Allele 25
e Allele 26
I Allele 27

 

 

 

Note: Homozygote alleles were halved and counted twice.

43

 

 

 

Table 4: Average Peak Heights (R.F.U.) Per Locus

Peak |Devhflku1
11 1 509
1291 1130 602
1481 1398 535
1277 1151 47
962 962 296
1108 1011 479
1299 1202 535

1243 1163 425

1451 1362 473
1151 1117 403
1255 1198

81 1746 739
1488 1411 478
1340 1300 465

   

Note: Homozygote alleles were halved and counted twice.

44

 

Conclusion and Future Research

The third goal of this project was to evaluate the viability of the PowerPlexTM 16
Amplification Kit for database samples. As stated previously, database samples (as those
used for CODIS) should all be single source samples. No degraded samples or mixtures
would be included in a database such as CODIS, therefore lessening the chance for
preferential amplification. In addition, by observation of the heterozygote PHRs
(approximately 0.8 to 0.86), preferential amplification was minimal. Peak heights were
also compared with fragment size of alleles at each locus; this would reveal any
differential amplification trends (within loci). Differential amplification would be
evident by decreasing peak height with increasing size at each locus, or between loci.
There was minimal differential amplification within loci (as can be noted in Figure 4),
meaning the PowerPlele6 Kit amplified all of the alleles at each locus equally. Also,
the average peak heights per loci did not display any signs of decreasing peak height with
increasing locus size, meaning the kit had amplified each of the loci equally. In addition,
the average heterozygote peak heights obtained per locus were under the recommended
peak height limit recommended by Promega, with the height of the homozygotes a little
higher than the suggested optimal height.

It should be noted that overall, the kit performed equally within and between loci,
with the exception occurring at the locus Penta E. The Penta E locus was slightly
problematic, being such a large locus and containing the largest sized alleles in the
PowerPlexTM16 allelic ladder. From the results, preferential amplification is observed at
the Penta E locus, with 23% of the peak height ratios falling under the 70% threshold.

This may be due to the overall size and base pair lengths of this locus. The occurrence of

45

 

 

preferential amplification indicates that the heterozygote alleles are not amplifying
equally, which again could be due to the size of this locus, meaning the larger sized
alleles would be amplifying less than the smaller alleles. In addition, a downward trend
was noted in the graph of Penta E (Figure 4), indicating slight differential amplification
(a reduction of peak height with an increase in fragment size). Though some anomalies
were observed in the Penta E locus, it should be stressed that it is not one of the thirteen
required CODIS loci, that it was added simply to increase specificity of the DNA typing
system.

Future research for the PowerPlexTM16 System include developing an appropriate
threshold for acceptable peak height ratios. The value of 0.70 was utilized in this project,
but that value was generated from validation studies of a genotyping kit of only ten loci,
whereas the PowerPlexm16 kit has sixteen. Research using single source samples
should be done to observe amplification trends per locus and to develop a suitable cutoff
for determining sister alleles. Additional research should also be done to observe any
trends in differential amplification within and between loci. This should be done using a
large number of single source samples so that all of the PowerPlexTMl6 alleles are
represented.

Overall, the PowerPlexTMl6 System is a robust method for obtaining the DNA
profiles for single source samples to be entered into the CODIS database. Though this
project did not include a concordance study to verify the DNA profiles of the one
hundred and fourteen samples, it should be noted that the known samples used in this
evaluation (9947A, MBILC, B19, and H9) all typed as expected, and that the allelic

ladders utilized in this project produced the expected results as well (as compared to the

46

 

allelic ladder in the PowerPlexml6 Technical Manual). As stated previously, this
evaluation of locus performance was part of a larger project by Promega and a number of
other forensic DNA testing facilities. The data generated in this evaluation was included
in a concordance study with the Applied Biosystem’s Profiler PlusTM and CoFilerTM
typing systems, which evaluated the concordance of the primers used in each kit. Results
in the “STR primer concordance study” indicated that the primers used in the
PowerPlexm16 System and the Profiler PlusTM/CoFilerTM systems produced reliable
results on reference samples.

The PowerPlexml 6 system is a powerful method of obtaining the DNA profile
of an individual at sixteen different loci in one reaction. Not only does this system save
the analyst time, money, and resources, but also the agency for which he/she works. This
typing system would allow for faster processing of convicted offender samples, whose
DNA profiles could be entered into the CODIS database quicker, thus providing

investigators with another DNA profile of which to search.

47

 

WORKS CITED

48

WORKS CITED

. Lee, et al. “DNA Typing in Forensic Science, I. Theory and Background.” The
American Journal of Forensic Medicine and Pathology 1994; 15(4), 2699-282.

. Budowle, et al. “STR primer concordance study.” Forensic Science
International. 2001; in press.

. Avery, MacLeod, and McCarty. “Studies on the chemical nature of the substance
inducing transformation ofpneumococcal types.” Journal of Experimental
Medicine 1944; 98, 451-460.

. Wallin, et al. “TWGDAM Validation of the AmpFLSTRm Blue PCR
Amplification Kit for Forensic Casework Analysis.” Journal of Forensic Science
1998; 43(4), 117-133.

. Micka, et al. “Validation of Multiplex Polymorphic STR Amplification Sets
Developed for Personal Identification Applications.” Journal of Forensic
Sciences 1996; 41(4), 582-590.

. Lins, et al. “Development and Population Study of an Eight-Locus Short Tandem
Repeat (STR) Multiplex System.” Journal of Forensic Science 1998; 43(6),
1 1 68-1 1 80.

. Walsh, Fildes, and Reynolds. “Sequence analysis and characterization of stutter
products at the tetranucleotide repeat locus vWA,” Nucleic Acid Research 1996;
24(14), 2807-2812.

. Micka, et al. “TWGDAM Validation of a Nine-Locus and a F our-Locus
Fluorescent STR Multiplex System.” Journal of Forensic Science 1999; 44(6),
1243-1257.

. Promega Corporation. “Technical Manual, GenePrint® PowerPlexml6 System.”
Copyright © 2000 Promega Corporation, Madison, Wisconsin.

49

ADDITIONAL WORKS CONSULTED

Balding and Donnelly. “Evaluating DNA Profile Evidence When the Suspect Is
Identified Through a Database Search.” Journal of Forensic Science 1996; 41(4),
603-607.

De Pancorbo, et al. “Population Genetics and Forensic Applications Using Multiplex
PCR(CSF1PO, TPOX, and THOl) Loci in the Basque Country.” Journal of
Forensic Science 1998; 43(6), 1 181-1 187.

Dictionary of Scientists. Johann Friedrich Miescher. University Press © Market
House Books Ltd 1999.

Dieffenbach and Dveksler. “PCR Primer, A Laboratory Manual.” Copyright © 1995
by Cold Spring Harbor Laboratory Press, Plainview, New York.

Erlich. “PCR Technology, Principles and Applications for DNA Amplification.”
prright © 1992 by Oxford University Press, Inc., New York, New York.

Gill, et al. “Automated short tandem repeat (STR) analysis in forensic casework—a
strategy for the future.” Electrophoresis 1995: 16, 1543-1552.

Gill, Jeffreys, and Werrett. “Forensic application of DNA ‘fingerprints’.” Nature
1985; 318, 577-579.

Inman and Rudin. “An Introduction to Forensic DNA Analysis.” Copyright © 1992
by CRC Press, Boca Raton, Florida.

Lazaruk, et al. “Genotyping of forensic short tandem repeat (STR) systems based on
sizing precision in a capillary electrophoresis instrument.” Electrophoresis 1998: 19,
86-93.

Lewin. “Genes VI.” Copyright © 1997 by Oxford University Press, Inc., New York,
New York.

Lorente, et al. “Sequential Multiplex Amplification: Utility in Forensic Casework
with Minimal Amounts of DNA and Partially Degraded Samples.” Journal of
Forensic Science 1997; 42(5), 923-925.

Mullis. “The unusual origin of the polymerase chain reaction.” Scientific American
1990; 262(4), 56-65.

50

 

PE Applied Biosystems. “AmpFlSTR Profiler PlusTM PCR Amplification Kit User’s
Manual.” Copyright © 1998 by the Perkin-Elmer Corporation, Foster City,
California

Schmitt and Benecke. “Five cases of forensic short tandem repeat DNA typing.”
Electrophoresis 1997: 18, 690-694.

Watson and Crick. “A structure for deoxyribose nucleic acid.” Nature 1953; 171,
737-738.

51