wvw.

\1

'5‘11‘

‘\ 2.1. 1.“. ‘ V , warn _ . 1L.
t...}:..L_-f.: {Insurhrfl A. . . @Mflufi-fll—fip 5.. ._
. é. a 53‘s? 31,3 P .1
a L Jun 0%: In.» . 5...

, (tnnfl.
‘1.’-‘51-
is $an
, afi, .
\...o\ .f
.. a, .
storm-fl“
.‘\ -. 1
9.. «giltnfcx...
37%.. {$37.5 5.3%.
g...” n 3A.! .1. .90”
Quaswustfi am».
. a.

i!
.73. flu}!

 

.. 5.!
I n. <3 )I
I. .59... r! .81. 5..
pa .
.

2-13). 1....
% fxfi‘vfim
u.

p: o
a .V'. . u‘
. 3,1 +3“ V x . I.

. .4314.» 1.1.“: 3.x";
”.1. an... id"? 1.5:? n... z (
year; a. «gum ..

. . if. r a. ,
:rlruni {til .«I 3‘ :4 , .
Edi . 3
1.353s. .? 4).; r ..
...¢. , .2?

‘1 a, .
. .. .13

3.19.4 n.
A. is...
u. .0 film!
, a .1...
1w...
{3.1:

.. u,

 

i512 .
flag"?
. , 0.. .
, .53:le‘14-
... P 33..
n .. Jul.» sLunwm-Wyi r.
.2 . ..
1:.»

.2 3r
‘ r“... .t. u. \.
E: .
14.5w? 2., V
«pigtunhhumww “11...... .
. H‘- 13 ‘
Jun?! \‘ufivfliitu Rd.
an.» . 5.3; .‘wdflduq
harry!“- nrtvl‘. .-
nzwxrmuwnuwu .«3 (slit
5'3”, but $73?!
. 1 n-
. {313%
«(gut—.5.” .

\

EEEL)‘ .. it ~ ,
‘ :52. 5%,?«xi 5.31....
autumn $2.451} 1 a.» it.
4.2) 3.32. ”flak. .13

‘ M¢.ui¥Lw..aw. r
a. 1?. .

y a: 33.5.

. «F!..mn9:! I!

‘ .1) I
mau . ..fln.§..?§.u. 1”
, fiuflnnmuhuifin.
fissdfiufi. E.

 

 

 

 

. . 1.:
Langmuuaaqx.» ”.3 1.1.3.... . ... .
. . as” .T . u%awm$

~»J.H,s..,3.... 1.. u

 

4.?me

3%.”
.
C :2. 1 . . (-1

 

This is to certify that the
thesis entitled

SIMPLE TISSUE PRESERVATION METHODS THAT
RESULT IN RELIABLE DNA ANALYSES

presented by

Corinne Lindsey Michaud

has been accepted towards fulfillment
of the requirements for the

MS. degree in Forensic Science

 

 

 

 

Major Professor’s Signature
{79/06

Date

MSU is an Affirmative Action/Equal Opportunity Institution

 

LIBRARY
Michigan State
University

 

 

 

.o-n-.-c-u-‘-n-n-c-c-o-c-u--o-o-.--—v—.-.—-s--o-0-.-I-O-0-O-0-l-O-h-O---n-o-o-l-D-I-o---v—v- — — — 4

 

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

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2/05 p:/ClRC/DaIeDue.indd-p.1

 

SIMPLE TISSUE PRESERVATION METHODS THAT RESULT IN
RELIABLE DNA ANALYSES

By

Corinne Lindsey Michaud

A THESIS

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

MASTER OF SCIENCE
Department of Criminal Justice

2006

ABSTRACT

SIMPLE TISSUE PRESERVATION METHODS THAT RESULT IN RELIABLE DNA
ANALYSES

By
Corinne Lindsey Michaud

In the event of a mass disaster, a large number of victims must be located and
identified, and it is difficult to process the site in a timely and orderly manner. Many
victims may only be identified through DNA analysis; therefore, obtaining viable tissue
samples is of great importance. Likewise, tissue preservation of remains discovered in
very remote areas can also be hindered. The goal of this study was to examine protocols
for on—site preservation of tissues that would later undergo DNA analysis. Tissue
samples were taken over the course of one week from recently killed pig carcasses that
had been placed in a field during the summer. Six preservation methods were evaluated:
storage of tissue in ethanol, isopropanol, RNAIater, and silica, as well as oven drying
(70°C) and freezer storage (~20°C). Muscle and skin samples were collected and
preserved in 0.500g sections. DNA extractions were performed after two weeks and two
months of storage, and DNA quality and quantity were assessed using a series of swine
specific PCR assays. Samples from each of the methods evaluated were able to amplify
nuclear DNA fragments of up to 64pr. Tissue type and level of decomposition
significantly affected DNA quality. Ethanol and RNAlater were shown to preserve DNA
of the highest quality, though differences among preservation methods were not
significant. Field practicality was considered in conjunction with the DNA evaluations,
and the results of this study indicated that each technique could be implemented in the

field and used to preserve DNA of a quality amenable to forensic analyses.

ACKNOWLEDGEMENTS

I would like to thank my committee members, Dr. David Foran, Dr. Christina
DeJong, and Dr. Norman Sauer, for their insight and support through every stage of this
project. I would also like to thank the staff at the MSU Swine Teaching and Research
Facility and Dr. Rich Merritt for their assistance in obtaining the pigs used in this study.

I would like to thank the staff of the MSU Research Technology Support Facility, as well
as Aaron Tarone, for all of their help with the real time PCR experiments. Finally, many
thanks to my family, friends, and fellow Forensic Biology students who have supported

and advised me throughout my time at Michigan State University.

iii

TABLE OF CONTENTS

LIST OF TABLES .................................................................................. vi
LIST OF FIGURES ............................................................................... vii
INTRODUCTION .................................................................................. 1
DNA as an investigative tool ............................................................... l
Obtaining DNA samples in the event of a disaster ...................................... 3
Preservation of tissue for DNA analysis .................................................. 5
Practical applications of this stuay ...................................................... 10
MATERIALS AND METHODS ............................................................... 12
Developing preservation techniques ..................................................... 12
Obtaining and preserving tissue samples ............................................... 12
DNA extraction ............................................................................. 15
Evaluating the quality of DNA ............................................................ 17
Evaluating the quantity of DNA .......................................................... 20
Data analysis ............................................................................... 23
RESULTS ........................................................................................... 25
Tissue collection ........................................................................... 25
Preparation of tissues for preservation ................................................. 29
Digestion of tissue samples and extraction of DNA .................................... 29
Completion of DNA extractions .......................................................... 30
Observations of DNA yield gels .......................................................... 31
Evaluating DNA quality by targeting IGF—l amplicons of increasing sizes ....... 33
Real time quantification of DNA yields using IGF-I and HPRT ..................... 42
DISCUSSION ....................................................................................... 44
Evaluation of the field collection procedure ............................................ 44
Optimization of the organic DNA extraction technique .............................. 45
Dtflerences in DNA quality with regard to preservation method .................... 48
Eflects of tissue decomposition on DNA quality ....................................... 54
Evaluating the eflects of storage time prior to DNA extraction ..................... 56
Eflects of tissue type on DNA quality .................................................... 58
Trends in real time quantification of DNA yields ...................................... 60
Conclusions ................................................................................. 61
REFERENCES ..................................................................................... 64

iv

LIST OF TABLES

Table l — Sequences of PCR primers for DNA quality testing experiments ................ 18
Table 2 -— Grouping of samples for PCR experiments .......................................... 19
Table 3 — Cycling parameters for IGF-l amplification ........................................ 20
Table 4 — Primer sequences for real time PCR experiments .................................. 21

Table 5 — real time PCR Reaction mix concentrations ......................................... 22
Table 6 — Cycling parameters for real time PCR ................................................ 22
Table 7 — Success rates for amplification of the IGF-l 257bp fragment ..................... 34
Table 8 — Significant differences among time points in amplification of the IGF-l 257bp

fragment .............................................................................................. 34
Table 9 — Success rates for amplification of the [GP -1 457bp fragment ..................... 35
Table 10 — Significant differences among time points in amplification of the [GP -1 457bp
fiagment .............................................................................................. 36
Table 11 — Success rates for amplification of the IGF-l 642bp fragment ................... 37

Table 12 — Significant differences among time points in amplification of the IGF-l 642bp
fragment .............................................................................................. 37

Table 13 — Differences in amplification of the IGF-l 642bp fragment due to tissue type
and time of collection ............................................................................... 38

Table 14 — Differences in amplification of the IGF-l 64pr due to tissue type and
preservation method ................................................................................. 39

Table 15 - Overall success of IGF-1 amplification rates between storage time points. ...40

Table 16 — Differences among preservatives in IGF-l amplification rates between storage

time points ............................................................................................ 41
Table 17 — Effects of storage time on amplification of the 642bp IGF-l fragment ........ 41
Table 18 — Real time PCR amplification of IGF-I ............................................. 43

Table 19 - Real time PCR amplification of HPRT ............................................. 43

vi

LIST OF FIGURES

Figure l — Placement of the pigs in the agricultural field ...................................... 13
Figure 2 — Schematic of daily tissue sample collection ........................................ 15
Figure 3 - Day one observations of pig carcasses .............................................. 26
Figure 4 — Day three observations of pig carcasses ............................................. 26
Figure 5 — Day five observations of pig carcasses .............................................. 27
Figure 6 — Day five observations of decomposition in pig one ............................... 27
Figure 7 -— Day eight observations of pigs two and three ...................................... 28
Figure 8 — Day eight observations of pig one ................................................... 28
Figure 9 — Yield gel of D5/2MO extracts ........................................................ 32

vii

INTRODUCTION

DNA as an investigative tool

DNA analysis has become an invaluable tool in the area of identification of
human remains. Whether the case in question concerns a single missing person or a
large-scale disaster, the ability to positively identify an individual’s remains is an
important service that forensic sciences are able to provide. The power of a DNA profile
to identify a single individual makes DNA one of the most influential and unassailable
types of biological evidence. Due to the accuracy and specificity of this type of testing,
DNA analysis is regularly used as a tool in paternity cases, missing person’s cases, rapes,
murders, and virtually every other type of crime. Some cases, such as those involving
paternity disputes or a recently murdered victim, are highly amenable to DNA profiling,
when the sampled material is abundant and may be quickly analyzed.

Advances in technology have also allowed for very small amounts of biological
material to result in successful DNA analysis. Samples such as a single drop of dried
blood, a hair, or even the residue from a fingerprint have been used to generate DNA
profiles. In a recent review of casework involving DNA analysis from hair, Melton et a1.
(2005) reported that full mitochondrial DNA (mtDNA) profiles were obtained from 570
out of 691 hairs. Samples ranging from head hair, pubic hair, and axillary hair have all
been shown to provide enough genetic material for mtDNA sequencing (Pfeiffer et al.,
1999). A quantification study by Andréasson et a1. (2006) revealed that thousands of

copies of nuclear DNA could be extracted from body hairs, fingerprints, and personal

jewelry items. These types of samples may be useful in some cases; however, there is a
great amount of variability in the success of DNA typing using such limited or decayed
samples. Even in instances where biological samples are relatively fresh or plentiful,
DNA analysis is not always successful.

As recent tragedies such as 9/11 and Hurricane Katrina have illustrated, the sheer
enormity of the task at hand may delay the identification process. The number of victims
from the World Trade Center site stands at two-thousand seven hundred and forty-nine.
As of September 11, 2005, positive identifications of 1594 victims had been reported,
approximately 850 of which were based solely on DNA analysis (Biesecker et al., 2005).
Even with advancements in forensic DNA analyses since 2001, such as the development
of mini-STR analyses for degraded samples (Butler, et al., 2003; Coble and Butler, 2005),
this means that 1155 of the World Trade Center victims still have not been identified. In
April of 2005, New York City’s Office of the Chief Medical Examiner announced that
the identification efforts would be put on hold indefinitely (Lipton, 2005). These events
demonstrated the great effort and cooperation that is necessary on the part of the
government, scientists, the public, and various other agencies after a disaster of such
magnitude. Unfortunately, they also underscored the need for improvements to the
system, in the interest of providing timely and accurate identifications for the families

and communities involved.

Obtaining DNA samples in the event of a disaster

From the case of a single unidentified body, to the scene of a plane crash or
natural disaster, the forensic goal is the same: to efficiently and accurately process the
scene and provide positive identifications of the persons involved. In order to achieve
this goal, individuals from many different disciplines must work together to locate,
acquire, and process samples from victims. Professionals from pathology, anthropology,
and law enforcement are typically present at a scene, in addition to any special agencies
necessary for a particular situation. Due to the diversity in the training backgrounds of
disaster responders, the procedures that are followed at different locations vary greatly.
Some agencies, such as the National Association of Medical Examiners (NAME)
recommend cold storage for any remains that are found (NAME, 2005). In a recent
special report on mass fatality incidents, the National Institute of Justice (N 11) made the
following statement concerning DNA sample collection: “Collect, place, and
appropriately store samples of suitable size in separately labeled containers Store
samples without preservatives (e. g., formaldehyde)” (N11, 2005). Secured freezer
storage is mentioned as an aside, but not specifically included in the recommended
protocol. The Federal Bureau of Investigation (FBI) also recommends that tissue
samples be preserved only by freezer storage (FBI Forensic Services Handbook, 2003).
The Interpol Handbook on DNA Data Exchange and Practice details the collection
procedure with regards to sterile techniques and storage vessels, but does not outline any

recommendations for subsequent storage of the samples (INTERPOL, 2001). Clearly,

consulting a variety of agencies’ manuals for disaster response will yield an array of
protocols for sample collection and processing.

Disaster response manuals are useful tools, but in reality a case must be managed
according to the situation at hand. Biological samples are obtained and preserved to the
best ability and knowledge of the responding team, and the procedures must be flexible to
meet the demand of a particular event. Depending on the type of tissue available, the
level of decomposition, and the preservation method (if any) used, subsequent DNA
analysis may or may not be successful. Delays between the acquisition of a sample and
the decision to attempt DNA analysis are also cause for concern, since the DNA will
continue to degrade. Although these considerations become integral to cases that hinge
on DNA results, preservation for potential DNA analysis is not the primary focus of a
disaster response team. For example, the NAME disaster response manual places sample
collection for subsequent DNA analysis as the tenth station in the progression through the
morgue, after radiology, photography, anthropological examinations, and other
administrative procedures have been completed (NAME, 2005).

As an investigation proceeds, unidentified remains pass through many stages of
cataloguing and examinations. With large-scale disasters, this is normally done in
temporary morgue facilities that are set up as close to the scene as possible. According to
an NIJ flow chart, a set of remains passes through triage, admitting, radiology,
photography, and cataloguing of personal effects before any anthropological or biological
examinations are conducted (N I], 2005). This typical progression can be shortened at the
discretion of scene supervisors, e.g., if it is apparent that a sample will not be useful for

anthropological or other visual examinations (NAME, 2005). If a case involves a limited

number of victims, the remains are brought to the nearest morgue before any
examinations are begun. When a positive identification can be made based on dental
records or an anthropological examination, further decay of the remains is not an issue of
concern. However, in situations where extensive injuries or decay impede a visual or
anthropological identification, DNA analysis becomes the main tool of the investigation.
If samples are not adequately preserved upon discovery of the remains, the potential for

successful DNA analysis is reduced by the lag time during the initial examinations.

Preservation of tissue for DNA analysis

An effective tissue preservation method must have several attributes. The
foremost concern is how effectively DNA is protected. Basic considerations affecting
this include the pH of a solution, the temperature and moisture level of the storage
conditions, and the ability to retard microbial activity. Once a methodology that is
amenable to DNA protection is developed, the issue of storage time must be addressed.
Will an extraction performed after one year of tissue storage yield the same quality and
quantity of DNA as an extraction performed afier two weeks of storage? If a
preservation method is successful in maintaining a long-term, high quality sample of
DNA, then various secondary factors can be taken into account, including portability of
materials, toxicity of materials, ease of use in the field, and ease of subsequent DNA
extractions.

Current preservation techniques can be classified into three broad categories: cold

storage methods, desiccation methods, and storage in solution. Each of these routes has

advantages and disadvantages, depending on factors such as resources, equipment, and
the future analyses that may need to be performed. When a technique is dependent on
cooling tissue, the necessary electricity and equipment may present problems in remote
locations or areas that have been flooded or lost power. Storing tissue in some type of
solution may require that a large quantity of preservative be kept on hand; since manual
preparation or commercial purchase would use valuable time in the event of a disaster.
Such considerations are integral when implementing a preservation method in the field,
however the first test of a technique’s utility is how effectively it preserves tissue.

Cold storage techniques are widely accepted preservation methods, with storage
at -20°C (a standard freezer) being one of the most commonly used and easily employed.
This method is effective because degradative processes are slowed or fimctionally
stopped by the low temperatures. Freezing limits the amount of chemical or physical
modification that tissue is subjected to, and subsequent DNA extractions do not require
any special considerations, the tissue can simply be thawed and processed. Freezer
storage is also the main technique currently employed in the event of a mass disaster, as
noted in the FBI and NH disaster response manuals (FBI Forensic Sciences Handbook,
2003; NI] 2005).

Storage in liquid nitrogen (-l96°C) is another temperature-based technique,
common in medical and biological research. The long-term benefits of liquid nitrogen
storage are similar to storage at -20°C, since degradation is prevented by the extremely
low temperature. Liquid nitrogen is the industry standard for many medical research
procedures, such as preserving tumor samples (Matsuo et al., 1999; Mutter et al., 2004).

Although liquid nitrogen is a routine preservation method, the regulations and dangers of

transporting and maintaining liquid nitrogen tanks make this method a less than ideal
choice for field applications. Likewise, tissues could be stored on dry ice (-78.5 °C) but
this requires obtaining and transporting the rather bulky material to the scene.

Desiccation represents another common method of tissue preservation. Oven-
drying is one such technique that exposes tissue to a hot dry environment, causing rapid
desiccation. DNA preservation is aided by heat-killing potentially harmful bacteria as
well as by drying out the sample and preventing natural cellular mechanisms from
degrading the DNA. Oven—drying is a common practice in plant research as well as the
food industry. Akpinar (2006) evaluated a range of air drying temperatures for the
preservation of apple, pumpkin, and potato slices; and found that over 70% of the
moisture was removed from the samples after oven—drying between 60 — 80°C for a
period of 8 —- 14h. Although this study was in the interest of food preservation, it shows
the efficacy of oven-drying as a desiccation technique.

Storage in silica mesh also results in rapid moisture removal and preservation.
Silica is a well-known desiccant, and has been included in numerous studies involving
field preservation of potential DNA samples. Silica storage was used to preserve hair
samples in the field during a 1997 study of carnivore populations in the western United
States (F oran et al., 1997). The researchers were successful at amplifying mtDNA
fi'agments of up to 600bp and nuclear DNA fi'agments of up to 440bp using the silica
preserved hair samples. Studies involving DNA analysis from fecal samples have also
utilized silica storage successfully. Roeder et a1. (2004) compared the DNA yield and
quality fiom three fecal sample preservation methods: 90% ethanol storage, silica

storage, and a two-step ethanol and silica preservation technique. Results of their study

showed that the quality of DNA obtained through each method was comparable when
DNA yields were low. When fecal samples contained a high amount of DNA, the silica
storage was not as effective at preserving high quality DNA as the ethanol or two-step
techniques, as measured by the successful amplification of a 757bp fragment. A more
comprehensive study by Murphy et al. (2002) evaluated the performance of 90% ethanol
storage, DETs buffer storage (a solution composed of dimethyl sulfoxide (DMSO),
EDTA, and Tris), silica storage, and two oven-drying methods. Results showed that
storage time significantly affected DNA yields from silica preserved samples, and also
that silica storage yielded a lower quality of nuclear DNA than the other methods. Silica
storage does effectively remove moisture from samples however, is safe and inert,
inexpensive, and is easily transported and utilized in the field, giving it several positive
attributes as a preservation technique.

Storage of tissue in solutions is a third preservation option; alcohols and
formaldehyde are two common preservatives that fall under this category. The
dehydrating and sterilizing properties of alcohols and aldehydes are beneficial to tissue
preservation, by drawing the water out of a sample and killing off bacteria. Although
such organic solvents are effective tissue preservatives, their usefirlness for DNA
preservation varies. Alcohol storage fixes tissue samples without cross-linking DNA,
which is usefiil for future nucleic acid extractions. Aldehydes, including formaldehyde,
permeate tissue and cause chemical changes in proteins and nucleic acids, making future
DNA extractions both cumbersome and potentially unsuccessful (Srinivasan, 2002).

Fixation in alcohol has repeatedly been demonstrated to perform superiorly to

aldehydes and comparably to freezing techniques with regards to DNA quality. In a

study using preserved human prostate samples, Gillespie et al. (2002) found that DNA
from formalin-fixed tissues was highly degraded and achieved limited success with PCR
experiments. Ethanol preserved samples from the same study consistently exhibited
DNA yields of high molecular weight and had high success rates for amplification of
microsatellite markers. Data from an assay of storage time periods by Kilpatrick (2002)
also showed ethanol to be an effective long-term preservation method. The author
analyzed relative DNA yields from murine livers that had been stored for up to two years
in various preservatives. After this time the ethanol stored samples exhibited high
molecular weight DNA when assayed on a yield gel, and a 400bp polymerase chain
reaction (PCR) product could be amplified. These properties, combined with the fact that
alcohols are inexpensive and easily obtained, make the two most safe and readily
available, ethanol and isopropanol, attractive candidates for tissue preservation.

Buffered saline solutions represent another widely used liquid storage method.
Various compositions of salts, chelating agents, buffering agents, and organic solvents
have been documented as successful tissue and DNA preservatives. A study by Seutin et
a1. (1991) was one of the first evaluations of a DMSO-salt solution as a preservative, and
found that a solution of 20% DMSO, 0.25M EDTA, and saturated NaCl preserved tissue
samples as effectively as liquid nitrogen storage. In the comparison of ethanol, DMSO-
salt solution, and lysis buffer (2M Tris-HCl, 0.5M EDTA, 5M NaCl, 20% SDS, pH 8.0)
preservation by Kilpatrick (2002), it was shown that DNA quality and quantity was
consistently greatest in the DMSO-salt solution over a time period of two years. DETs
buffer (containing DMSO, EDTA, and Tris), another DMSO-salt solution, was used to

preserve DNA from fecal samples by Murphy et al. (2002), and was shown to perform

comparatively to ethanol. Amplification of nuclear loci using PCR was successful in 75%
of the DETs stored samples, as compared to 81% of the ethanol stored samples and 43%
of the silica stored samples. All of these solutions have been shown to effectively
preserve tissue and DNA samples, but perfecting the concentrations of the components
and preparing the solutions can be time-consuming. Further, DMSO is slightly toxic.

A commercially prepared tissue preservative RNAIater (Ambion, Austin, TX) is
also available. RNAIater is an aqueous sulfate salt solution that rapidly fixes fresh tissue
and does not require refrigeration of the samples (Ambion). A recent study showed that
storage in RNAIater yielded RNA of a quality and quantity comparable to traditional
fresh or flash-fiozen tissue samples (Mutter, 2004). The commercial availability of
RN Alater has the potential to make it a convenient preservation method that is easily
adapted to use in the field, since the solution does not need to be manually prepared.
However this can also be considered a drawback, since it is only possible to obtain
RNAIater directly fiorn the manufacturer and a stock would need to be maintained in the

event of a disaster.

Practical applications of this study

The goal of this study was to evaluate simple, rapid methods for preserving tissue
samples from unidentified remains, in the interest of aiding future DNA analysis. Based
on a combination of efficacy and practicality, six preservation methods were chosen:
storage in ethanol, isopropanol, and RNAIater solutions, silica mesh, -20°C storage, and

70°C oven-drying. Tissue samples were obtained over the course of one week from three

10

pigs that were placed in a field and allowed to decompose. Samples were then preserved
by the six different methods, and the relative success of each was evaluated by comparing
subsequent DNA analysis results. PCR experiments were designed to evaluate the
quality and quantity of DNA from the samples, by examining both the length of DNA
fragments that could be recovered (DNA quality) and the relative DNA copy number
(DNA quantity). Results were analyzed by looking at trends across preservative type,
tissue type, decomposition level, and storage time. Recommendations for tissue
preservation and further evaluations were developed based on these results. The
molecular assays combined with field practicality considerations provide suggestions for
simple and effective tissue preservation methods that are amenable to subsequent DNA

analysis.

11

MATERIALS AND METHODS

Developing preservation techniques

Six tissue preservation methods were selected for this study: Storage at -20°C,
oven-drying at 70°C, silica desiccation, storage in 70% ethanol, storage in 70%
isopropanol, and storage in RNAIater. Screw-top, polyethylene 6mL sarnpule vials
(Wheaton, Millville, NJ) were used for sample storage. Samples designated for -20°C
storage were placed in vials and kept in a rack at -20°C. Samples preserved by 70°C
oven-drying were placed on waxed paper in a 70°C incubator for ~72 hours, then
transferred to vials and stored at room temperature. The remaining methods were storage
in either 4mL of 70% ethanol, 4mL of 70% isopropanol, 2.5mL of RNAIater, or 2.0 —
2.5g of silica (28 -— 200 mesh, Fisher Scientific, Pittsburgh, PA). Aside from the samples

stored at -20°C, all samples were kept at room temperature.

Obtaining and preserving tissue samples

The tissue samples for this project were taken fiom swine carcasses, obtained
from the Swine Teaching and Research Center at Michigan State University. Three pigs,
weighing approximately 30, 60, and 901bs, were euthanized and placed in an agricultural
field on the campus of Michigan State University during the month of August (Figure 1).

Initial samples of skin and muscle tissue were taken within an hour of death, by excising

l2

a tissue section fi-om the shoulder of each pig with a disposable scalpel. The tissue
sections measured approximately 10m x 5cm and were approximately 3cm thick.
Separate disposable plastic bags were used to transport the tissues from the field to the
laboratory. Samples were preserved within approximately 24h of collection; storage of

the entire tissue section at -20°C was used in the interim.

Figfi l — Placement of the pigs in the agg'cultural field

 

Pigs were placed directly on the ground within an area of approximately 25
square meters. The terrain varied slightly, from mowed grass to sparse tree cover. All of
the pigs received partial sun and were exposed to the elements, aside fiom the partial
cover that the trees provided over pigs two and three. From the time that the pigs were
initially placed in the field, they were undisturbed except to obtain tissue samples. Pig
one was placed on the edge of a tall grassy field (left) and pigs two and three were placed
in brush under the cover of sparse trees (right).

Using sterile scalpels, forceps, and razor blades in the laboratory, control samples
of 0.500g of tissue (+/- 0.002g) were excised fi‘om each tissue section, and DNA
extractions were begun the same day (see below). Tissue surfaces were shaved with a

disposable razor before taking samples, to minimize the amount of hair included in the

subsequent digestions. Scalpel blades were replaced often, due to the dulling that
occurred while cutting through the pigs’ bristles and skin. The tissues were then divided
into 0.500g aliquots (+/- 0.002g) of skin and muscle to be individually preserved. This
was accomplished by first separating the cutaneous tissue from the underlying muscle,
and removing individual 0.500g samples of each, resulting in a total of six skin and six
muscle samples per pig, on each day of collection (Figure 2). Samples that were to be
preserved in silica or solution (ethanol, isopropanol, or RNAIater) were placed directly
into a vial of preservative after being weighed, and stored at room temperature. Samples
to be stored at -20°C were placed directly into empty vials and frozen. Tissues for 70°C
oven-drying were placed on squares of waxed paper and dried in an incubator for
approximately 72h before being placed into individual vials and stored at room
temperature. During the storage periods all of the samples remained undisturbed at their
respective conditions.

The tissue collection and preservation procedure was repeated on days 3 and 5
post-mortem. By day 8, the 301b and 601b pigs had become mummfied, and the 901b pig
had been removed fiom the field, presumably by foraging animals, thus tissue sampling
ended after the day 5 collection. Preserved samples were stored for time periods of 2

weeks and 2 months, at which point DNA extractions were performed.

14

Figure 2 —- Schematic of daily tissue gimme collection

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

r fl r T r N
3 Pigs Pig One Pig TWO Pig Three
x. 1 P J L J
(”—_\ r fl r N r \ r7 fl r
2 Tissue Types Skin Mame Skin hillside sum Muscle
K____/ x 1 x J

 

 

 

 

 

 

war BIIIIB BIBBBE 311888 WE aural

36 samples per day

 

 

 

 

A total of 36 tissue samples were preserved from each day of field collection.
The collection procedure was repeated on days 1, 3, and 5 post-mortem, resulting in 108
preserved tissue samples. DNA extractions were performed on each of the tissue samples
after 2 weeks of storage and after 2 months of storage, resulting in 216 DNA samples.

DNA extraction

DNA extractions were begun after two weeks by removing samples from storage,
weighing, and processing half of the tissue. The remaining half of the tissue sample was
returned to the corresponding preservative method until the two month storage time
period, when a second DNA extraction was performed on the remaining tissue. Ethanol
and isopropanol stored samples were removed from their storage vials and blotted on
tissue paper before weighing. Preliminary experiments showed that rehydrating the
tissues stored in silica or RNAIater was necessary for tissue digestion; therefore, prior to

the extraction procedure these samples were removed from preservative and rehydrated

15

in 2mL TE (lOmM Tris, lmM EDTA) for approximately 2 hours. RN Alater samples
were returned to storage in their original vials; however, silica stored samples remained
in TE until the second DNA extraction. This was due to the fact that once the sample had
been rehydrated and cut into small fiagments, the remaining tissue would have been
unrecoverable from the silica mesh.

Tissues for DNA extraction were placed in 1.5mL microfuge tubes with 500uL of
digestion buffer (20mM Tris, pH 7.5, IOOmM EDTA, 0.1% SDS) and 5uL of proteinase
K (20mg/mL, Roche, Basel, Switzerland). Samples were macerated for ~10 seconds with
a conical glass pestle, then placed at 55°C for 72 hours. Samples were vortexed for 10 —
15 seconds daily to aid in the digestion of the tissue.

The extractions proceeded by adding 500uL of phenol to the digested tissues,
which were vortexed for 10 seconds and centrifuged at 14,000rpm for 5 minutes. The
aqueous layers were removed and transferred to new 1.5mL microfuge tubes. Five-
hundred microliters of chloroform was added to the extracts, and samples were vortexed
for 10 seconds and centrifuged at 14,000rpm for 5 minutes. It was noted during the
phenol and chloroform steps that the aqueous and organic layers of the RNAIater
samples were inverted after centrifugation. Therefore, samples were monitored to ensure
that the appropriate layer was transferred for the subsequent steps. The aqueous layers of
all samples were transferred to new 1.5mL microfuge tubes, to which two volumes of
cold 95% ethanol and l/10 volume of 3M sodium acetate were added and mixed by
inversion. The samples were placed in -20°C storage to allow the DNA to precipitate.
After 7 — 10 days of storage at -20°C, samples were centrifuged for 15 minutes at

14,000rpm, and supernatant was removed. Samples were vacuum-dried for 15 minutes,

16

and resuspended in SOuL of TE. DNA samples were returned to freezer storage. In the
interest of treating each sample equally, no additional steps were taken to purify

troublesome extracts, and all samples were resuspended in an equal volume of TE.

Evaluating the quality of DNA

Five microliters of each DNA sample were loaded on a 1.0% agarose gel.
Presence of high molecular weight DNA indicated a sample of relatively high quality,
whereas the presence of a wide range of DNA fragments (a smear) indicated a degraded
sample. This technique was used to roughly estimate the differences in yield and
degradation across the 216 samples.

DNA quality was evaluated by attempting to amplify DNA fragments of
increasing length; successful amplification of a large fiagment was used as an indicator
of a high quality sample. Three sets of primers were designed for the porcine Insulin
Growth Factor 1 gene (IGF-l), using the same forward primer but increasingly distant
reverse primers (Table 1). The primer pairs resulted in amplificons of 257bp, 457bp, and
642bp. DNA samples were tested successively for amplification of each target, or until
the sample failed to amplify. For example, if a sample was positive for amplification of

the 257bp target but failed to amplify at 457bp, no further testing was performed.

17

Table 1 — Sequences of PCR primers for DNA quajty testing exm’ments

 

 

 

Primer Amplicon
Primer Sfluence (5’ — 3’) size (nt) size
IGF-l Forward AAT CAT T'I‘G CCC CTC AAG TG 20 N/A
IGF -1 R257 TGA CCC CCT CAT CCT AGT TG 20 257
IGF-l R457 GGC AGG AAG ACA CAC ACA TC 20 457
IGF-l R642 TCT CTC CCT CTT CTG GCA AA 20 642

 

 

The sequences for the IGF-l primers, which pair the same forward primer with
various reverse primers to create amplicons of increasing lengths.

Primers were designed using porcine IGF-l gene sequences obtained fiom the
online National Center for Biotechnology (NCBI) database and the online primer design
program Primer3 (Rozen and Skaletsky, 2000). The primers were synthesized by
Integrated DNA Technologies, Inc. (Coralville, IA), and were received as lyophilized
pellets which were resuspended in TE to a concentration of 200umol/L.

PCR conditions were optimized using the DNA from day 1 unpreserved tissue
samples as a control. The conditions were: IU of Hot Master Taq Polymerase
(Eppendorf, Hamburg, Germany), 1X Hot Master Taq PCR Buffer (Eppendorf), 20uM
dNTPs (Promega, Madison, WI), luM forward IGF-l primer, and luM reverse primer,
per reaction. PCR reactions were set up in lOuL volumes, with luL of DNA sample per
reaction. Unsuccessful PCR due to excess genomic DNA was observed in the majority
of the lGF -1 257bp reactions. Products of these reactions were characterized by high
molecular weight DNA smears and a lack of target amplification when observed by gel

electrophoresis. DNAs were diluted 1/ 100 and re-tested for amplification of the IGF-l

l8

257bp fragment If a DNA extract successfully amplified using the l/ 100 DNA dilution,
all subsequent testing was performed using that dilution.

PCR experiments were divided into six groups, broken down by day of collection
and time point of extraction (Table 2). The six groups were tested separately, including a
positive control (Day 1 unpreserved DNA sample) and negative control (no DNA added)
with each set of reactions. As testing progressed fiom the 257bp test through the 642bp
test, the samples remained grouped in the same manner, so that six separate rounds of

PCR were set up for each fiagment size.

Igble 2 — GroupinLof samples for PCR experiments

 

Q1931; m Collection Day Stmge Time Point
1 D1/2MO Day 1 2 Months
2 Dl/2WK Day 1 2 Weeks
3 D3/2MO Day 3 2 Months
4 D3/2WK Day 3 2 Weeks
5 D5/2MO Day 5 2 Months
6 D5/2WK Day 5 2 Weeks

In order to facilitate the progression of the PCR experiments, samples were
broken into these 6 groupings, and reactions were run for one group at a time, including
positive and negative controls with each set of reactions.

The same thermocycling program was used for the experiments with the 257bp,
45 7bp, and 64pr target amplicons (Table 3). Results were visualized by
electrophoresing 5uL of each sample on a 1.0% agarose gel and staining with ethidium
bromide. Samples were recorded as positive if the reaction produced a band of the

targeted size, regardless of intensity.

19

flble 3 — Cycling parameters for IGF-l amplification

Temperature (°C) Time

94 ' 2m

94 30s

58 1m 35 Cycles
72 1m

72 5m

4 so

The same thermocycling program was used for the 257bp, 457bp, and 64pr
amplification experiments.

Evaluating the quantity of DNA

DNA quantification was attempted using real time PCR analysis. The primer
sequences used were located in the porcine hypoxanthine phosphoribosyltransferase
(HPRT) and lGF-l genes, the sequences of which were obtained fi'om the NCBI online
database. Sequences were entered into Primer Express soflware (Applied Biosystems,
Foster City, CA), which provided primer pairs for small amplicons (50 — 100bp) suitable
for real time PCR (Table 4). Primers were synthesized by Integrated DNA Technologies,
Inc., and were received as lyophilized pellets which were resuspended in TE to a

concentration of 200umol/L.

20

Iable 4 — Primer sequences for real time PCR experiments

 

 

 

Primer Amplicon
Primer Sequence (5'-3') len nt (pp)
HPRT F93 GCT CGA GAT GTG ATG AAA GAG ATG 24 77
HPRT R170 AAA GAA T'I'I‘ ATA GCC CCC CTT GA 23
[GP -1 F679 TCG CCC ATC CTC CAC GTA T 19 69
IGF-l R748 GGC AGG AAG ACA CAC ACA TCT G 22

 

 

Pairs of primers used for the real time PCR experiments, resulting in the
amplification of a 77bp fiagrnent from the HPRT gene and a 69bp fiagment from the
IGF-l gene.

The real time PCR primers were optimized by testing forward and reverse primer
concentrations of 2uM, luM, and 0.2uM. The thermocycling program consisted of a 10
minute denaturation at 95°C, followed by 40 cycles of 95°C for 15 seconds and 60°C for
30 seconds. Once the conditions resulted in a clean, single-banded product, the
experiment was tested under real time PCR conditions on an ABI Prism 7900HT
Sequence Detection System (Applied Biosystems) at the Michigan State University
Research Technology Support Facility. Multiple concentrations of primers were tested,
again using combinations of 2uM, luM, and 0.2uM forward and reverse primers. Final
conditions for the real time PCR experiments were: 1x PowerSYBR Green Master Mix
(Applied Biosystems), luL of DNA template (diluted as appropriate), and varying
concentrations of primers (Table 5). The reaction volume was lOuL, with the remainder
of the mix being made of water. The thermocycling parameters were optimized for

amplification of the small target fragments (Table 6).

21

liable 5 — real time PCR Reaction mix concentrations

£1.11le
Ingredient Concentration Volume
PowerSYBR 1x 5uL
DNA template Unknown luL
IGF-l Forward 0.2Um luL
lGF-l Reverse 1Um luL
HPRT Forward 0.2uM luL
HPRT Reverse 2uM luL

The concentration of PowerSYBR and volume of DNA template added were
consistent across all of the real time PCR reactions. Results of optimization experiments
indicated that the IGF -1 and HPRT primers were most effective at different
concentrations, ranging from 0.2 — 2uM.

Ipble 6 - Cycling parameters for real time PCR

PCR Prom
Temperature (°C) Time

50 2m
95 10m

95 158 40 Cycles
60 303

The real time PCR thermocycling program was designed with short annealing and
extension times to increase specificity and minimize non-specific amplification. The
initial 50°C was required for optimum activity of the AmpErase enzyme (AB1), and the
95°C hold served to activate the room-temperature stable polymerase AmpliTaq Gold
(AB1) contained in the real time PCR reagents.

Reaction mixes for the real time PCR experiments were prepared manually and

aliquoted into 8-strip 0.2mL PCR tubes, and DNA templates were prepared in 96-well

22

plates. A Biomek 2000 Laboratory Automation Workstation (Beckman, Fullerton, CA)
was used to mechanically transfer the master mix and DNA templates into 384-well
plates. DNAs were tested in triplicate, including positive (Day 1 unpreserved DNA
sample) and negative (no DNA) controls. All reactions for IGF-l were prepared and
amplified at one time, and all HPRT reactions were prepared and amplified on a separate
day. The reactions were analyzed using the ABI Prism 7900HT Sequence Detection
System and SDS software version 2.1 (Applied Biosystems). The dissociation curve and
cycle threshold of each sample were obtained for analysis.

Results for the IGF-l real time PCR experiment were ambiguous, so a subset of
DNA samples were re-tested for amplification of the 642bp IGF-l fragment to ensure
that the dilutions being used still produced valid results. This test indicated that the DNA
samples were still of high quality; therefore, the real time PCR testing proceeded with the

HPRT primers.

Data analysis

Samples were scored as either positive (amplification occurred) or negative (no
amplification) for the quality testing results. Positive results for the six time point groups
and the six preservation methods were tabulated, and the mean success rate of each group
for each fragment size was calculated. One-way ANOVA was performed to compare the
differences in mean success rates among the collection time points and among the
preservation methods at each of the IGF-l fiagrnent sizes. The mean success rates

between storage times were also compared through ANOVA, to analyze preservation

23

method performance after two weeks of storage compared to two months of storage.
Differences in the quality of DNA obtained from skin and muscle were analyzed only for
the IGF-l 642bp fragment test results, due to clearly limited differences among success
rates at the 257bp and 45 7bp sizes. A Bonferroni test was performed on all pair-wise
comparisons within a group if ANOVA indicated a significant difference among means
(a=0.05). All statistical analyses were performed using SPSS For Windows, version
11.5.2.1 (SPSS).

Results for real time PCR experiments were analyzed by determining the average
Ct value for samples from each preservation method. Separate averages were calculated
for the IGF-1 and HPRT results. Though DNA extracts were run in triplicate, there were
few cases where all three reactions resulted in amplification. Because of this, three
separate average Ct values were computed for each preservation method—an average
including only samples that worked in triplicate, a second including samples that worked
in duplicate, and a third including all samples that worked at least once. Due to the
limited success of these tests, no further statistical analyses were undertaken. Results of
the real time experiments were compared on the basis of mean Ct values for each
preservation method as well as the number of DNA extracts that were successfully

amplified within each preservation method.

24

RESULTS

Tissue collection

Tissue collection fiom the pig carcasses was performed on days one, three, and
five. Day one samples were taken from the cervical region of each pig within an hour of
death, and no visible decay or insect colonization had occurred (Figure 3). By day three
each of the pigs had begun to show signs of abdominal bloating and putrefaction in their
limbs. Maggot infestation had occurred around the pigs’ mouths, noses, eyes, and ears,
as well as the cut site of the first tissue sampling (Figure 4). Tissue samples were
obtained on day three by removing a section of tissue immediately posterior to the initial
cut site, from an area flee of maggot infestation.

Mummification of the head and forelimbs had occurred on each of the pigs by day
five, and maggot infestation had expanded to the forelimbs and shoulders of each pig
(Figure 5). The abdominal cavity of pig one had ruptured by this time and flies had
begun to congregate in the intestinal area (Figure 6). Tissue samples were again obtained
from an area immediately posterior to the maggot infestation and previous cut sites.
Tissue collection was halted after day five due to the extensive mummification of pigs

two and three, and the disappearance of pig one (Figures 7, 8).

25

Figm 3 — Day ope observatim of pig carcasg

 

Tissue samples were obtained from the pig carcasses immediately after being
placed in the field. Initial samples were removed from the cervical region of each pig, as
illustrated on pig one (left) and pig three (right). No visible decay or insect colonization
had occurred at this time point.

Fi 4—Da bservati of i s

   

Substantial maggot infestation had occurred by day three, centered around the
cranial orifices of each pig (left). Signs of bloating and putrefaction were also observed
in the abdominal regions of each pig (right). Tissue samples were taken from the
shoulder of each pig, immediately posterior to the initial cutting site.

26

FiguLe 5 —- Day five pbsmations pf pig mg

 

The head and forelimbs of each pig had begun to mummify by day five (Pig three,
right; Pig two, left). Maggot infestation had continued to progress posteriorly, and
further putrefaction was observed in the abdomen and limbs.

Figm 6 — Day five observations of decommsitiop in pig one

 

Pig one exhibited the most advanced state of decomposition among the three pigs.
By day five, the abdominal cavity had ruptured, increasing the rate of tissue decay in the
abdomen and hind limbs as compared to pigs two and three.

27

Fi 7—Da ei tobservationsofi two (I

 

The entire carcasses of pigs two and three had become extensively mummified by
day eight. Maggots had completely transgressed the carcass of pig three and were
migrating from the body (left). Pig two (right) had not yet reached that state of
decomposition, but no tissue sections could be collected from either carcass by this date.

Fi 8—Da i t bservationsof i o

 

A mandible and partial cranium were all that remained of pig one after eight days
in the field (left). It appeared that these bones were virtually undisturbed compared to the
location of this carcass on day five (right). No further traces of the carcass could be
located, nor could any determinations be made concerning how it had been removed. It
was presumed that small carnivores such as coyotes had destroyed the remains.

28

Preparation of tissues for preservation

Tissue processing consisted of separating skin and muscle tissues and weighing
individual samples for preservation. The separation of skin and muscle did not present
any difficulties, since the sections obtained in the field greatly exceeded the amount of
tissue needed and precise dissection was not paramount. As the carcasses decayed, this
procedure became easier, with the skin separating readily and intact fi'om the
decomposing fat and muscle layers beneath.

Skin tissue maintained a rubbery and pliable consistency throughout the collection
time period. Cutting through the epidermis continued to require the use of a razor blade
and sharp scalpel. Muscle tissue became progressively more decomposed, to the extent
that samples were slick, pulpy, and beginning to turn green by day five. Care was taken
to avoid clearly rotten samples, by excising any discolored or putrefied sections. The
odor of all samples strengthened as the level of decay increased, and this odor was still

apparent when tissues were taken out of storage for digestion.

Digestion of tissue samples and extraction of DNA

Notable differences among preservation methods were observed throughout the
course of the tissue digestions and subsequent DNA extractions. Oven-dried samples
were difficult to halve without losing tissue, as the samples would crumble when
attempting to cut them with a scalpel. This may have affected their yield when compared

to other preservation methods. By two months of storage, both skin and muscle tissues

29

preserved in ethanol and isopropanol were beginning to fiagrnent in the vials. The
solutions had become cloudy with tissue debris, and samples often shredded when
grasped with forceps. Digestions were prepared using all of the tissue that could be
recovered from the solution, but the comparative DNA yields of these samples may have
been affected by tissue loss. Silica stored samples had fragmented in the TB by the time
the two month DNA extractions were begun. Again, as much tissue as possible was
recovered from the solution, but due to the pulpy nature of the samples there was some
tissue loss. RNAIater samples and -20°C stored samples did not present any difficulties
or disparities between tissue types in this phase of the procedure.

Digestions were continued for approximately 72h at 55°C, due to the difficulty in
getting the oven-dried and RNAIater preserved samples to fully lyse in the digestion
buffer. In the interest of treating all groups uniformly, no samples were allowed to

incubate longer than 72h even if the tissue was not completely digested.

Completion of DNA extractions

Phenol-chloroform DNA extractions were performed on all of the digested
samples with few problems. Silica samples required extra care during transfer of the
aqueous layer to ensure that no residual silica mesh was carried over into the DNA
precipitation. Samples that had been stored in RNAIater were consistently observed to
have the aqueous and organic layers inverted upon addition of the phenol and chloroform.
This required that the top, organic layer be discarded and the lower, aqueous layer be

retained for the next step. Although relative yields may have been affected, these

30

extractions still resulted in DNA, as all extractions were observed to contain DNA when
assayed on yield gels (see below). No further aberrations were observed through this
phase of the extractions.

Following DNA precipitations and drying, the silica samples were found to
contain a small viscous layer (~5 OuL) at the bottom of the microcentrifuge tube, along
with an ill-defined DNA pellet. Care was taken to remove as much of this substance as
possible without interfering with the DNA pellet; after vacuum-drying and resuspending
these samples in TE there was no longer any visible trace of the viscous material.
RNAIater samples also exhibited atypical pellet formation. In some cases a white
interface formed near the bottom of the tube, in other cases a small volume of opaque
solution with no clear pellet was separated out at the bottom of the tube. The supernatant
was removed as efficiently as possible without disturbing either the interface or opaque
solution, which were presumed to contain the DNA. No further anomalies were observed

after vacuum-drying and resuspension in TE.

Observations of DNA yield gels

Analysis of the extracts on yield gels indicated that each extraction had
successfully isolated some quantity of DNA. The concentrations and the extent of
degradation varied widely among all of the extracts (Figure 9). No consistent trends were
observed between particular preservation methods and the apparent DNA yields. These

results became most useful in determining which DNAs would be good candidates for

31

dilutions, in order to overcome the problems observed in subsequent PCR experiments

caused by excess DNA.

Figm 9 — Yield gel of D5/2MO extracts

12 3 4 5 6 7 89101112131415161718

all“.

19 20 2122 23 24 25 26 212%

 

DNA yields varied greatly and without any clear trends across all extracts. Lane
13 is an example of an extract that exhibited a high concentration of high molecular
weight DNA. Lane 16 is an example of an extract with no visible high quality DNA, and
very little DNA overall. Extracts that contained DNA of a high molecular weight and
concentration were diluted 1/ 100 for use in subsequent PCR experiments. The order of
the samples on the gel is Ethanol, Isopropanol, RNAIater, Silica, 70°C oven-drying, and -

20°C, repeated six times from lanes 1 — 36.

32

Evaluating DNA quality by targeting IGF-I amplicons of increasing sizes

Results of the quality testing experiments were analyzed by tabulating which
DNA extracts successfirlly amplified using, the primers for the three IGF -1 fiagments.
Results were analyzed separately for each of the IGF -1 experiments, beginning with
amplification of the 257bp target (Table 7). The overall PCR success rate for this
fragment was 89%, or 192/216 samples. Seventy degree oven-drying had the highest
success rate, with amplification in 94% of the DNA extracts. Ethanol storage and
RNAIater had the second highest success rate, with amplification in 92% of each group.
Isopropanol and -20°C storage had equal amplification success rates of 89%, and silica
storage had amplification in 81% of DNAs. These results were not statistically different
(p=0.509).

Amplification differences based on tissue collection time and length of storage
showed that the D5/2MO and D5/2WK DNA extracts had the fewest positive results
(Table 7). Seventy-five percent of the D5/2MO extracts successfully amplified and 69%
of the D5/2WK DNAs amplified. The D3/2WK group had a success rate of 92%, and the
other three time points had amplification in 100% of the samples. AN OVA indicated
significant differences among the means of time points, and pair-wise comparisons
showed that there were significant differences between seven pairs of time points. The

significant differences all involved comparisons of the D5 time point groups (Table 8).

33

Table 7 — Success mes for amplification ofmg IGF -1 W

 

% f les i'vefrIGF1257b am lifi tin
D5/2MO D5/2WK D3/2MO D3/2WK D1/2MO D1/2WK Pres. Avg._

 

 

 

 

 

 

 

 

Ethanol 0.83 0.83 1.00 0.83 1.00 1.00 0.92
Isopropanol 1.00 0.67 1.00 0.67 1.00 1.00 0.89
RNAIater 0.83 0.67 1.00 1.00 1.00 1.00 0.92
Silica 0.33 0.50 1.00 1.00 1.00 1.00 0.81
70°C 0.83 0.83 1.00 1.00 1.00 1.00 0.94
-20°C 0.67 0.67 1.00 l .00 1.00 1.00 0.89
Time Pt. Avg. 0.75 0.69 1.00 0.92 1.00 1.00

 

 

 

 

 

 

 

 

The results of the IGF-l 257bp test show that oven-drying had amplification in
94% of DNA extracts, the highest success rate among the preservation methods.
Analysis of the time point groups show that both D5 groups had significantly lower
amplification success rates than the other time points.

Table 8 — Sigpificant differences among time mints in amplification of the IGF-l 2529p
figment

 

 

 

 

 

 

 

 

 

Timg Point 1 (meg) Time Point 2 (mg) P valpp
1 D5/2M0 (0.75) D3/2MO (1 .00) 0.004
2 D5/2MO (0.75) D1/2MO (1 .00) 0.004
3 D5/2MO (0.75) D1/2WK (1.00) 0.004
4 D5/2WK (0.69) D3/2MO (1.00) <0.001
5 DS/ZWK (0.69) D3/2WK (0.92) 0.017
6 D5/2WK (0.69) D1/2MO(1.00) <0.001
7 D5/2WK (0.69) D1/2WK (1.00) <0.001

 

 

 

 

 

 

Seven pair-wise comparisons of mean success rates for amplification of the 257bp
fragment were significantly different at the u=0.05 level. This was determined using a
Bonferroni test on mean success rates for all time point groups. All significant
differences involved comparisons between D5 time points and earlier sample collection
time points.

34

The overall success rate for amplification of the IGF-l 457bp fragment was 86%.
Ethanol, isopropanol, and RNAIater had the highest mean success rate amongst
preservation methods (Table 9). Isopropanol samples maintained the same amplification
success rate as in the 25 7bp arnplicon experiment; however, all other preservation
methods exhibited a decrease in amplification success. Silica storage again had the
lowest number of positive samples, with a 75% success rate. None of the differences
among means were statistically significant.

Analysis of the IGF-l 45 7bp test by time point groups showed that the D5/2MO
and D5/2WK sample sets were again the lowest ranked, with mean amplification success
rates of 61% (T able 9). Statistical analysis indicated that the differences between the
means of the two D5 groups and each of the other time points were significant (T able

10).

Table 9 — Success rates for amplification of the IGF-l 457bp ham

 

% of Samplg msitive for IGFl 457bp amplification
D5/2MO D5/2WK D3/2MO D3/2WK Dl/2MO D1/2WK Pres. Avg_._

 

 

 

 

 

 

 

 

 

 

 

 

Ethanol 037* 033* 1.00 033* 1.00 1.00 039*
Isopropanol 1.00 037* 1.00 037* 1.00 1.00 039*
RNAIater 037* 037* 1.00 1.00 1.00 1.00 039*
Silica 0.17* 033* 1.00 1.00 1.00 1.00 075*
70°C 037* 050* 1.00 1.00 1.00 1.00 036*
-20°C 050* 037* 1.00 1.00 1.00 1.00 036*
Time Pt. Avg. 0.61* 031* 1.00 092* 1.00 1.00

 

 

 

 

Mean success rates for amplification declined for all preservation methods except
isopropanol storage. The only significant differences between means were again
attributed to the low success rates of the D5 time point groups. *Samples that were
negative for amplification of the IGF-l 257bp fragment are included as negative results
in these calculations.

35

Table 10 — Sigpificant differences among m e mints in amplification of the IGF-l 457ng

 

 

 

 

 

 

 

 

 

 

meat

Time Point 1 (mean) Time Point 2 (mean) P vglueg
l D5/2MO (0.61) D3/2MO (1 .00) <0.001
2 D5/2MO (0.61) D3/2WK (0.92) 0.001
3 D5/2MO (0.61) D1/2MO (1.00) <0.001
4 D5/2MO (0.61) Dl/ZWK (1.00) <0.001
5 DS/ZWK (0.61) D3/2MO (1.00) <0.001
6 DS/ZWK (0.61) D3/2WK (0.92) 0.001
7 D5/2WK (0.61) D1/2M0(1.00) <0.001
8 D5/2WK (0.61) D1/2WK (1 .00) <0.001

 

 

 

 

 

 

Eight pair-wise comparisons of mean success rates for amplification of the 457bp
fiagment were significantly different at the a=0.05 level. All significant differences once
again involved comparisons between D5 time points and earlier sample collection time
points.

PCR amplification of the IGF -1 64pr fragment was positive in 81% of all
samples. Ethanol and RNAIater had the highest number of successful amplifications
among the preservation methods (Table 11); silica storage had amplification in the fewest
samples. Although the PCR experiments indicated that success continued to decline as
target size was increased, there were no significant differences among the preservation
methods.

The D5/2MO and D5/2WK sample sets ranked lowest among the time points,
with mean success rates of 42% and 56% for amplification of the 64pr fragment (Table
11). The number of samples that were amplified in each of the other time point groups
remained static. Once again, the D5 group means were found to be significantly different

from the other four time point groups (Table 12).

36

Igble ll - Success rates fpr amplification of the IGF-l 64pr went

 

°/g of Samples msitive for IGF l 642bp amplification
D5/2MO D5/2WK D3/2MO D3/2WK Dl/2MO Dl/ZWK Pres. Avg_._

 

 

 

 

 

 

 

Ethanol 050* 033* 1.00 033* 1.00 1.00 036*
Isopmpanol 0.50 067* 1.00 067* 1.00 1.00 031*
RNAIater 050* 067* 1.00 1.00 1.00 1.00 036*
Silica 000* 033* 1.00 1.00 1.00 1.00 072*
70°C 050* 033* 1.00 1.00 1.00 1.00 031*
~20°C 050* 050* 1.00 1.00 1.00 1.00 033*
Tim Pt- Avg- 042* 056* 1.00 092* 1.00 1.00

 

 

 

 

 

 

 

 

Mean success rates continued to decline for each of the preservation methods in
this experiment. Ethanol and RNAIater had the greatest number of positive samples
among the preservation methods, and silica storage had the least number of amplified
DNAs. The mean success rates of the D5 time points continued to decline as well, with
all earlier time points holding static. ‘Samples which failed to amplify the 257bp or
45 7bp fragments are included as negative results in these calculations.

Table 12 — Significant differences among time mints in amplification of the IGF-1 642bp

 

Lament

 

 

 

 

 

 

 

 

 

 

 

 

 

Time Point 1 (mean) Timg Point 2 (mg) P v lues
1 DS/2MO (042) D3/2MO (1.00) 73%??—
2 D5/2MO (0.42) D3/2WK (0.92) <0.001
3 [rs/mo (0.42) D1/2MO (1.00) <0.001
4 D5/2MO (0.42) Dl/ZWK (1.00) <0.001
5 D5/2WK (0.56) D3/2MO (1.00) <0.001
6 D5/2WK (0.56) D3/2WK (0.99 <0.001
7 DS/2WK (0.56) Dl/2MO (1.00) <0.001
8 DS/ZWK (0.56) D1/2WK (1 .00) <0.001

 

 

 

The statistically significant differences in the IGF-1 642bp test were attributed to
the low mean success rates of the two D5 time point groups, which had significantly
different means compared to all other time points.

37

 

Results for the 64pr IGF-1 test were also evaluated on the basis of tissue type.
Skin samples had an average amplification success of 89%, and muscle samples had an
average amplification success of 74% (Table 13). ANOVA indicated that the differences
between these amplification results were significant (p=0.005). The effect of tissue type
on DNA quality was also compared among time points (Table 13). The disparity in
amplification success between tissue types was much greater in the samples fi'om D5
time points. These individual comparisons corroborated the general trend of skin
samples having more positive PCR results than muscle samples. Due to the low number
of samples in each of these sets (n=18) statistical analyses were not attempted on the

individual time point comparisons.

 

Table 13 - Differences in mplification of the IGF-1 6429p Lagment due t9 tissue gym
and time of collection

 

"/9 of samples msitive for IGF-1 642bp amplificatim

 

 

 

 

 

 

S1513; Muscle
D5/2MO 0.61 0.22
D5/2WK 0.72 0.39
D3/2MO 1.00 1 .00
D3/2WK 1 .00 0.83
D l /2MO l .00 1 .00
D l /2WK 1 .00 l .00
Avg. 0.89 0.74

 

 

 

 

Overall, DNA from skin samples amplified with significantly more success than
DNA fi'om muscle samples (p=0.005). Differences between individual time points were

more substantial, but due to small sample sizes no further descriptive statistics were
calculated.

38

The results of the 64pr amplicon experiment were also analyzed with respect to
each preservation method and tissue type (Table 14). Silica preserved samples amplified
with equal success regardless of tissue type. Skin samples amplified with more
consistency than muscle samples in each of the other preservation method sets. ANOVA
indicated that isopropanol was the only preservation method found to have a significant

difference between tissue type (p=0.002).

Table 14 — Differenfi in amplification of thLIGF-l 6429p dm to tissue gm and

 

preservatim method

"/9 of samples msitive for IGF-1 642bp amplification
Skin Musclg P value

 

 

 

 

 

Ethanol 0.94 0.78 0.157
Isopropanol 1.00 0.61 0.002
RNAIater 0.89 0.83 0.641
Silica 0.72 0.72 1.000
70C 0.89 0.72 0.218
-20C 0.89 0.78 0.386

 

 

 

 

 

Tissue type affected the results of PCR amplification of the IGF-l 642bp
fiagment substantially. Silica preservation had the same number of PCR positive skin
and muscle samples, but in every other method assayed the DNA fi'om skin samples was
amplified with a higher success rate than the DNA from muscle samples. Isopropanol
was the only method found to have a significant difference between tissue types.

The affect of storage time on the DNA quality was also analyzed. The overall
mean success rates for DNA extracted after 2 weeks and 2 months of storage were similar
(Table 15). The 2 week extractions had a lower mean success rate for amplification of
the 257bp fiagment, but the success rate remained nearly static as amplicon size

increased. The 2 month extractions initially performed better, but a more substantial

39

decline was observed in the mean success rate as amplicon size increased. The results of
the PCR of the 64pr fiagment were analyzed by ANOVA, which indicated no

significant differences between storage time points (Table 17).

jl'pble 15 -— Overall succesp of IGF-1 amplification rates between storage time mints

 

 

 

 

 

 

IGF-1 2573p IGF-1 4572p IGF-1 6422p ‘
2 weeks 2 months 2 weeks 2 months 2 weeks 2 months
0.87 0.92 0.84 0.87 0.82 0.81

 

 

 

 

 

 

 

 

 

 

 

 

The mean amplification success rates for each fiagment size were similar between
the two storage times. The mean success rate for the 2 week storage time declined 5%
when amplicon size was increased from 257bp to 64pr. The mean success rate for the 2
month storage time declined 9% as target size increased from 257bp to 64pr. The
differences in performance between short and long storage times were not statistically
significant.

Mean success rates for amplification of each IGF-1 amplicon were also calculated
for the two storage time points within each preservation method. For all preservation
methods except silica, DNAs from the two month storage time had higher amplification
success with the 257bp target than the DNAs from two weeks of storage (Table 16).
However, DNA extracts from two months of storage all exhibited a substantial decline in
success as amplicon size increased. DNA from silica samples stored for two weeks was
amplified with greater success than the DNA from sampr stored for two months for
each amplicon size. Analysis of the 642bp amplification results for individual

preservation methods did not show any significant differences between the short and long

storage periods (T able 17).

40

 

lahle l6 -- Differences among waives in IGF-1 mphfication rates between smpage

 

 

 

 

 

 

time mints

IGF-1 257bp IGF-1 451292 IGF-1 64th

2 weeks 2 months 2 weeks 2 months 2 weeks 2 months
Ethanol 0.89 0.94 0.89 0.89 0.89 0.83
Isopropanol 0.78 1.00 0.78 1.00 0.78 0.83
RNAIater 0.89 0.94 0.89 0.89 0.89 0.83
Silica 0.83 0.78 0.78 0.72 0.78 0.67
70°C 0.94 0.94 0.83 0.89 0.78 0.83
-20°C 0.89 0.89 0.89 0.83 0.83 0.83

 

 

 

 

 

 

 

 

Comparisons between sample storage times indicated that the two month samples
amplified more successfully than the two week samples for the 257bp fragment. As
amplicon size was increased, the DNAs fi'om two months of storage were amplified with
less frequency. DNAs from two weeks of storage exhibited little or no decline in success
rates as amplicon size increased.

Table 17 - Effects of storpge time 911 amplifigm'on of the 64th IGF-1 figment

 

2 weeks 2 months P value

 

 

 

 

 

 

 

Ethanol 039 033 0.641
Isopropanol 0.78 033 0.684
RNAIater 039 033 0.641
Silica 0.78 0.67 0.471
70°C 0.78 033 0.684
-20°C 033 033 1.000
Avg. 0.82 031 0.728

 

 

 

 

 

Differences in 64pr fragment amplification between the 2 week and 2 month
storage times were not statistically significant. Silica exhibited the greatest disparity
among preservation methods, while storage at -20°C had no change in amplification
success between storage times.

41

Real time quantification of DNA yields using IGF-I and HPRT

Results of the real time PCR experiments did not generate enough data to perform
a comprehensive statistical analysis, since amplification occurred sporadically among all
samples. Reactions were run in triplicate for each DNA extract; however, very few
samples amplified in triplicate (4/216 for the IGF-1 assay and 34/216 for the HPRT
assay). Comparisons of preservation methods were tabulated based on the number of
DNAs within each group that successfully amplified in triplicate, the number of DNAs
that successfully amplified in duplicate, and all DNAs that amplified at least once (Tables
18 and 19). Mean Ct values were calculated separately for each of these categories.

Ethanol storage performed well in both assays, achieving amplification rates and
mean Ct values that consistently ranked among the top three methods. Storage in
RNAIater also performed well across all categories, although amplification of the HPRT
locus was notably more successfirl. Storage at -20°C, and 70°C oven-drying had the
greatest number of samples amplify in the IGF-l assay; however, success rates for the
HPRT assay dropped considerably. Sample sizes were highly variable across all of these

categories, so no filrther statistical comparisons were attempted.

42

Table 18 — Real p’me LQR amplification of IGF-1

 

 

 

 

 

 

 

 

 

# Singlets Avg Ct # Duplicates Avg Ct # Triplicates évg Ct
Ethanol 22 24 9 25.2 0 N/A
Isopropanol 21 25.8 6 24.7 0 N/A
RNAIater 26 24.5 8 25.9 0 N/A
Silica 18 25.4 1 34 0 N/A
-20°C 26 24.8 8 24.3 2 22.7
70°C 26 23.3 6 24.6 2 22.2

 

 

 

 

 

 

 

 

Results of the real time PCR experiments were analyzed using data from all

reactions that exhibited amplification in one replicate (“Singlets”), two replicates
(“Duplicates”) and in all three replicates (“Triplicates”). Amplification in the IGF-l real
time PCR was irregular, with only four DNAs amplifying in triplicate. Storage at -20°C
and 70°C oven-drying were the highest ranked preservation methods, based on the
number of reproducible amplifications and the average Ct values. Ethanol and RNAIater
had similar performances based on the singlet and duplicate categories but had no
triplicates.

Table 19 — Real time PCR amplification of HPRT

 

 

 

 

 

 

 

# Singlets Avg Ct # Duplicatgs Avg C; # Triplicates Av Ct
Ethanol 30 24.9 20 25.1 6 22.4
Isopropanol 21 25.6 18 25.8 5 24.5
RNAIater 32 25.7 22 25.4 9 25.4
Silica 25 29.1 17 29.1 7 27.9
-20°C 28 27.5 16 26.7 2 22.7
70°C 27 27.1 17 27.1 5 24.3

 

 

 

 

 

 

 

 

 

Amplification of the HPRT locus was notably more successful than the IGF-1
assay; however, only 34/216 DNAs were amplified in triplicate. DNA from ethanol
samples was amplified with the most consistency in this experiment, followed closely by
RNAIater samples. The performances of -20°C and 70°C oven-drying dropped
considerably compared to results fiom the IGF-1 experiment.

43

DISCUSSION

Evaluation of the field collection procedure

Tissue samples were obtained fi'om the pig carcasses as proposed, without
modification of the field collection procedure. Although the carcasses decayed rapidly, a
wide array of samples were procured in the time period preceding the disappearance of
pig one and the mummification of pigs two and three. Samples were therefore not
diverse from a temporal perspective, but the entire range of decomposition from flesh
through bloat and murnmification had been observed. It appears that placing cages over
the pigs is necessary to prevent interference from scavengers, and that placing pigs at
different times of year would aid the collection process by slowing the rate of
decomposition.

The amount of tissue in each preserved sample needed to be kept constant in order
to generate relative comparisons across the preservative methods. This necessitated
collecting large sections of tissue and meticulously cutting and weighing individual
aliquots in the laboratory, adding time to the procedure that would not occur under
normal (non-experimental) circumstances. There would be no such lag time between
tissue collection and preservation in practical applications, as methods evaluated in this
study could potentially all be employed in the field, and samples would be preserved on-
site without regard for the exact weights or measurements that this study required.

Within 6 — 24h fiom the time of collection, all tissue samples had been processed

and stored. As previously noted, tissue sections were temporarily kept at -20°C during

the delay between collection and processing, which could have influenced the quality of
DNA before the experimental preservation procedures were implemented. This should
not have affected storage comparisons as the tissues were frozen and subsequently
thawed at the same time, ensuring that all samples were exposed to the same pre-
processing conditions.

The actual collection of tissue would be identical for any of the storage methods
tested, but the practicality of implementing a preservation technique must be weighed as
well. Vials of a liquid or dry preservative could be transported to a crime scene or the
site of a mass disaster, and tissue samples could then be stored as soon as they were
acquired. The main consideration for on-site vial storage is maintaining readily available
stocks of the selected preservative, whether that means manual preparation and
dispensation of materials, or advanced ordering from a manufacturer. On the other hand,
storage at -20°C or oven-drying at 70°C would require additional equipment such as
freezers or incubators, as well as electricity. Unless either of these methods is far

superior to all others, they are probably less than ideal for the outcome desired.

Optimization of the organic DNA extraction technique

The first phase of this study where marked variation among the preservation
methods was observed was during the removal of tissues fiom storage for subsequent
DNA extraction. Ethanol, isopropanol, and 2 month silica stored tissues (now in TE) had
begun to fragment in solution, potentially affecting the amount of material available for

DNA extraction. Dividing tissues in half proved difficult for the rubbery RNAIater

45

samples and the extremely desiccated and friable oven-dried samples. Both of these did
not always digest to completion within the 72h incubation. Oven drying was an effective
preservation technique (with 81% of samples having DNA amplified up to 642bp) but the
behavior of the tissues added substantially to the workload and efficiency of the
extraction procedure. However, in casework it is unlikely that dividing and restoring a
sample would occur, thus these techniques should not be eliminated based solely on that
factor.

To circumvent some of the difficulties with tissues after storage in alcohol,
RNAIater, silica, or by heating, samples could be divided initially and preserved,
eliminating the need to isolate sections of tissue at a later date. By using an entire sample
for the DNA extraction, no tissue would be lost and substantial time would be saved.
Digestion of tissues might also be aided by increased maceration or more fiequent
vortexing. These options were not explored in this study as a bias in DNA yield would
have been introduced when treating samples differently. None of the other preservation
methods required special consideration during the digestion process.

Key differences among the preservation methods were also identified during the
phenol and chloroform phases of the DNA extractions. Samples from ethanol,
isopropanol, -20°C storage, and oven-drying methods did not exhibit any unusual
behavior during the DNA extraction procedure. In contrast, the behavior of RNAIater
samples was atypical, due to the inversion of the organic and aqueous layers. It was
presumed that a high salt content in the RNAIater solution was affecting the density of
the aqueous layer and causing it to settle below the organic layer. Although DNA was

extracted successfully from these samples, it is possible potential yields were affected.

All other samples required that an upper aqueous layer be retained, but the RNAIater
samples had the modified treatment of trying to retain a lower aqueous layer while
removing an upper, organic layer. This may have resulted in retention of unwanted
materials in RNAIater DNA extracts due to incomplete removal of the organic liquid. A
potential remedy for this problem would be manipulating the rehydration of the sample
prior to beginning the extraction. By increasing the rehydration time or using several
changes of TE, the amount of RNAIater carried into the subsequent extraction may be
lessened, allowing the aqueous layer to behave normally.

RNAIater samples were also problematic during pelleting and rehydration of the
DNA. All other methods produced visible DNA pellets after centrifugation, as expected.
A white interface or cloudy layer of solution formed during DNA precipitation in
RNAIater stored tissues, indicating that a solution more dense than ethanol was present in
the tube, perhaps preventing the DNA from pelleting. This could be due to carry-over
fiom the organic solvents or the effects of residual RNAIater. As mentioned, an
extended rehydration in TE would help determine if adequately removing the RNAIater
from the tissue prevented subsequent problems during DNA extraction.

Silica preserved samples also proved more diffith during DNA isolation, which
was presumably due to carryover of silica mesh throughout the extraction, or from these
samples generating an inefficient removal of the organic phase of the extraction. Instead
of a distinct white pellet, the silica samples appeared to have a small volume (~50uL) of
viscous solution collected at the bottom of the tube, resulting in malformed pellets that
dried with a grainy appearance. A modified, multi-wash rehydration procedure would

help identify the problem of silica mesh interference, since this would reduce the amount

47

of particles that were introduced into the digestion. Unlike RNAIater samples, silica
does not penetrate the tissue, so increased rehydration time would not necessarily affect

the outcome.

Diflerences in DNA quality with regard to preservation method

Performances of each preservation method with regard to DNA quality were
ranked based on overall results of the IGF-1 PCR experiments, including both the
percentage of samples that amplified, as well as the maximum amplicon that could be
obtained. Samples positive for amplification of the 642bp fiagment were considered to
contain “high quality” DNA; therefore, preservation methods that resulted in the greatest
number of 64pr amplifications were deemed the most amenable to DNA analyses.
There were DNA extracts from each of the preservation methods that were successfully
amplified up to 642bp, validating the belief that each method helps preserve DNA.
However, the sensitive nature of forensic casework requires extreme confidence and
reproducibility in all procedures, thus the comparative amplification success and
consistency of each preservation method are key factors in choosing a ‘best’ preservation
protocol.

Ethanol and RN Alater preserved tissues continually yielded high quality DNA
extracts, as evidenced by the number of samples that were positive for amplification of
the 257bp, 457bp, and 642bp fragments (92%, 89%, and 86%, respectively). The DNA
extracts from these methods had more positive amplifications of the 457bp and 64pr

fiagments than any of the other preservation methods (Tables 9 and 11). These results

48

were not statistically different hour the other storage techniques, but in terms of this
study, ethanol and RN Alater storage were shown to preserve the most “high quality”
DNA.

Several factors likely contributed to the effectiveness of ethanol preservation.
First, it is widely used as an antimicrobial agent, and it is logical to conclude that by
killing the bacteria in or on the tissue, the degradation process would be slowed. Second,
the small molecular size of ethanol allows it to rapidly permeate and fix a sample, which
prevents bacterial mechanisms (putrefaction) and host pathways (autolysis) from
functioning and degrading the DNA (Srinivasan, 2002). Since RNAIater preservation
resulted in DNA extracts of the same quality as ethanol, it seems plausible that RNAIater
is also effective at permeating tissue and halting degradation. Although the composition
of RNAIater is proprietary, it is known to be an aqueous salt solution, and prior research
has indicated that such solutions are effective DNA preservatives (Seutin et al., 1991).
Oddly, one area where ethanol and RNAIater were found to have very different
preservative capabilities was between tissue types (skin versus muscle; see below). The
type of tissue available may therefore be an important factor when choosing between
ethanol and RNAIater.

Storage in isopropanol was the next most effective preservative method, based on
the high number of samples that were amplified at each size. DNA extracts from
isopropanol preserved tissues amplified with equal success for the 257bp and 457bp
amplicons (89%), while all other methods experienced a decline in positive results
(Tables 7 and 9). However, amplification of the 64pr fragment from isopropanol

preserved tissues exhibited a decline in mean success rate to 81%. The major reason for

49

the lower overall success of isopropanol storage was the low 67% amplification rate in
day 3, 2 week preserved samples (Table 11). One-hundred percent of the DNAs fi'om
day 3, two month preserved samples amplified in the same assay, which seems
incongruous with the poor results of the earlier storage time. Due to the small number of
samples in this portion of the study, the affects of outliers may have had a
dispmportionate impact on the overall amplification rates.

Storage in isopropanol likely had preservation benefits similar to those of ethanol.
Both of these alcohols have been shown to be effective at killing a range of bacteria and
fungi, which would be useful in preventing putrefaction of a sample (Penna et al., 2001).
Despite similarities, the performance of isopropanol as a DNA preservative was not as
successful as ethanol. A potential reason for this was that the increased molecular weight
of isopropanol hindered its ability to permeate tissue. This could have caused an increase
in lag time between when the tissue was placed in solution and when the tissue was
effectively preserved. Another difference between ethanol and isopropanol is in the
polarity of the molecules. Ethanol is somewhat polar, which has an impact on
interactions with membranes and fluids, causing water to be displaced fiom the surface of
the cell membrane (Chanda and Bandyopadhyay, 2006). Isopropanol is non-polar, thus
some of the preservative benefits of this interaction may be reduced. Though the overall
amplification success rates of isopropanol DNA extracts were inferior to those of DNAs
from ethanol and RNAIater preserved tissues, this was mostly due to the potential effect
of outliers in the D3/2WK sample group, as previously mentioned. The cost,

transportation, and preservation techniques related to ethanol and isopropanol are largely

50

identical, so further analyses are needed to identify specific advantages of one alcohol
over the other.

The two temperature—based storage techniques were shown to have moderate
numbers of high quality DNA extracts (Table 1 1). Oven-dried samples had the greatest
success of all methods for the 257bp amplicon, with 94% of the DNAs being positive
(Table 7). However, these samples experienced the largest decline in success among
PCR assays, with only 81% of the overall DNA extracts being positive for the 64pr
fragment (Table 11). This method also showed a large difference between tissue types;
89% of the oven-dried skin samples amplified in the 64pr test, while only 72% of the
muscle samples amplified. These results indicate that small DNA fragments amplified
consistently from oven-dried samples, but in other areas, including differences between
storage times and between tissue types, the oven-dried samples amplified unpredictably.

It was clear from visual observations that oven-dried samples were well
desiccated, to the point that tissues had to be essentially chipped into fragments before
being digested during DNA extraction. The amount of material that was added to the
digestion may have skewed the PCR results for this preservation method, since difficulty
in dividing the tissue was encountered with all oven—dried samples. The temperature
used in this study was sufficient for sterilization purposes, as per USDA food safety
guidelines (U SFSIS, 2003), so putrefaction was not a likely factor once tissues had been
oven-dried. However, bacterial activity may have continued to occur before the tissues
reached the temperature of 70°C and became sterilized.

DNA may begin to denature at a temperature of 70°C, which means that the two

strands begin to separate from one another, reducing the stability of the molecule.

51

Bruskov et al. (2002) also found that the chemical structures of DNA bases were altered
at increasing rates as temperatures were increased from 37°C to 95°C, and that reactive
oxygen molecules were created, causing mutations in the sequence of DNA or loss of
nucleotides. Without functioning cellular repair mechanisms, these molecular changes
could lead to degradation of the DNA. Oven-drying also requires additional equipment,
an electrical source, and lag time before samples are sufficiently preserved, which
compounds the inefficiency of this preservation method.

Storage at -20°C resulted in a consistent number of DNA amplifications in each
IGF-1 experiment, but as with isopropanol, the number of positive results never
surpassed those of ethanol or RNAIater (T able l6). Eighty-three percent of the DNA
extracts from freezer stored samples were amplified in the 642bp PCR experiment.
Freezing at -20°C was equally effective over the two storage periods, but again the
quality of the DNA was not as high as other preservation methods (Table 12). This
preservation method was poor at preserving the most decayed samples, with only 50% of
all DNA samples from day 5 showing amplification of the 642bp fragment.

Freezer storage is a method that does not alter the permanent chemical or physical
state of a sample. Storage at -20°C essentially slows or halts any metabolic pathways
within the tissue, thereby slowing degradation. Since cold storage may not entirely kill
bacteria as a chemical treatment would, low amounts of putrefaction may have occurred
before tissues were completely frozen, or while they were thawed during the tissue
digestion. One important characteristic of fieezer storage was that it did not exhibit any

substantial decreases in amplification success between storage times (Table 16). If

52

samples need be stored for a long time period, fieezer storage would therefore be an
effective preservation method.

Silica storage was the lowest-ranked preservation method in each of the IGF-l
fragment tests. This method was particularly unsuccessful with highly decayed samples;
no extracts from D5/2MO samples were positive in the 642bp PCR experiment.
However, silica was the only method that did not have any disparity between tissue types,
with amplification in 72% of both skin and muscle tissues in the IGF-1 642bp assay
(Table 14).

Falling under the preservation category of desiccation, silica storage was expected
to preserve tissues in the same manner as oven-drying, by effectively dehydrating
sarnples and preventing microbial or cellular activity. The samples were visibly withered
following storage, but retained more pliability than the oven-dried tissues, indicating that
there was some amount of moisture retained. DNA degradation can occur in a moist
environment, so incomplete desiccation may explain the poor performance of the silica
samples. Adjusting the amount of silica used for preservation may improve the quality of
the DNA, since it is possible that the amount of silica used did not sufficiently desiccate
the tissue.

The process of rehydration and TE storage of silica preserved tissues must also be
considered, as the samples stored for two months were not kept in silica for the whole
time course. Although TB is suitable for DNA storage, autolysis or bacterial activity
could have been reinstated after the rehydration, causing a decline in DNA quality before
the subsequent extractions. This could explain why none of the two month extractions

were positive for the 64pr amplification; however, the two week silica extracts were

53

also the lowest ranked in amplification of the 64pr fragment, which does not involve the
TB storage factor (T able 16).

Due to the exceptionally poor amplification results of the day 5 DNA extracts,
silica samples had a low overall amplification success rate. Since tissues fiom this stage
of collection were the most decayed, it is possible that the samples had higher moisture
content or greater numbers of bacteria than tissues from earlier collection times. Silica
preservation is routinely used for the preservation of feces for DNA analyses (Murphy et
al., 2002; Roeder et al., 2004), which would preclude the notion that bacterial activity
cannot be curtailed by silica storage. The poor performance of silica from the day 5 time
points is most likely due to a combination of inefficient desiccation and the opportunity
for degradation to occur while the samples were stored in TE.

In order to gauge the bias that was introduced by the TB storage, this phase of the
experiment could be repeated without removing samples from storage until DNA
extractions were begun. Further, varying the ratio of silica to tissue volume would show
whether more efficient desiccation results in higher quality DNA samples. These
variables would have to be addressed before a more complete view of silica performance

could be developed.
Eflects of tissue decomposition on DNA quality
The time of tissue collection (1 — 5 days) was found to be the most crucial factor

in DNA quality, and in fact, the only area where significant differences were observed.

Amplification of the 642bp fragment occurred in only 42% of the D5/2MO samples and

54

56% of the D5/2WK samples (Table 11). This is not surprising, since the extent of decay
by day 5 was advanced to the point of mummification (Figures 5 and 6), and DNA would
most likely have been more degraded before preservation was attempted. In addition to
autolysis, putrefaction was likely a significant factor in the degradation of DNA by this
time.

These observations support the notion that early collection and preservation of
tissues is most likely to result in successful DNA analyses. A critical shift in the success
of DNA amplification occurred between the samples from the first and last days of tissue
collection. Amplification of the 64pr fiagment decreased from 100% for the day 1
samples to an average of 49% for the day 5 samples, across all preservation methods
(Table 11). This indicated that with relatively lmdecayed samples, all methods were
equally effective in preserving DNA. Conversely, after a certain stage of decay the
performance of the preservation method became integral to preserving the highest quality
DNA.

In order to elucidate subtle differences in the preservative qualities of these
methods, future studies could incorporate sampling over a longer time period, or
obtaining more tissue samples within the same time frame. Repeating this procedure
during a colder season would allow tissue to be sampled over a broader time period,
which may reveal greater discrepancies among the preservation methods as the initial
DNA quality continued to decline. Taking samples more fiequently within a set time
period could help to correlate the stages of decay with the effectiveness of the
preservation methods. Environmental conditions would also have to be taken into

account, since decomposition rate is largely a function of temperature, moisture, and

55

other seasonal factors. Based on these preliminary results it is clear that the extent of
tissue decay plays a major role in the success of subsequent DNA analysis, and future

investigations could strengthen this connection.

Evaluating the eflwts of storage time prior to DNA extraction

The “shelf-life” of a preserved sample is an important consideration, especially in
situations where the number of victims prevents the timely analysis of samples, or when a
specimen needs to be re-tested at a later date. A preservation method that results in high
quality DNA samples over a wide range of storage times would be most beneficial for
reliable and reproducible analyses. The data from the 642bp amplicon indicate that as
storage time increased, the DNA quality decreased (Table 16). This conclusion is based
on the fact that three out of the six preservation methods (ethanol, RNAIater, and silica)
had more “high quality” samples in the DNAs from two weeks of storage than the DNAs
fiom two months of storage. Frozen samples had no difference in the number of positive
PCR results between storage times. With isopropanol and oven-dried samples,
amplification was more successful using extracts from two months of storage than
extracts from two weeks of storage. The data from the IGF-1 642bp assay indicate that
storage time decreased the quality of DNA; however, results from the 257bp PCR
experiment indicate that DNA from the two storage times performed similarly when
amplifying small targets (Table 16).

These aberrations can be explained by several factors. One of the major variables

in this experiment was the investigator’s prowess with the organic extraction procedure.

56

Several of the preservation methods required special attention during the DNA
extraction; problems which were anticipated and addressed before performing the
extractions for the two month storage time. In addition, the entire procedure went more
smoothly during the later round of DNA extractions, due to the investigator’s increased
experience. Results of the IGF-1 25 7bp assay show that the isopropanol, oven-dried, and
RNAIater samples fiom two months of storage were amplified with greater success than
the similarly preserved samples fiem two weeks of storage (Table 16). This corroborates
the idea that the second round of extractions was performed more efficiently, although
DNA was likely more degraded and only the amplification rates of small amplicons were
affected.

The physical state of the tissues used for the two month DNA extractions was
another factor that could have contributed to the unexpected 257bp test results. It was
noted during the tissue digestions after two weeks of storage that many samples were
difficult to cut and digest, and that after two months of storage the tissues were beginning
to fragment and soften. It is possible that after the longer storage period the tissues were
more amenable to digestion, and therefore more DNA was released fi'om the tissue even
though the time was kept constant. Observations of the ethanol, isopropanol, and silica
samples corroborate this supposition, since it was noted that these samples were visibly
pliable and fiagmented when removed from preservative for the two month extractions.
This may have resulted in more complete tissue digestion; but after the long storage time
the DNA had likely been degraded into smaller fragments, so the quality of the DNA

obtained was not as great as in extracts from the shorter storage time.

57

Further investigation of the affect of storage time on DNA quality could include
more time points and a greater sample size. Assaying DNA quality over a longer storage
time may support the theory that low quality DNA is more abundant as storage time is
increased. This would be determined by keeping the tissue digestion procedure static,
and comparing the quality of DNA extracts fi'om samples that were digested after a range
of storage times. It would be expected that increasing storage time would yield DNA that
was successfully amplified for small targets. If degradation was occurring during
storage, then these extracts should continue to show a decrease in quality, as assayed by
amplification of larger amplicons. An increased sample size would benefit the storage
time experiment by reducing the impact of problematic extractions on the overall
performance of a preservation method. The small number of samples that had to be
specially treated during extractions may have disproportionately affected the average for
a particular method since the sample size was only eighteen per tissue type and
preservation method for a particular time point. A larger sample size would also benefit
the experiment in general by helping to reduce any random effects introduced by the

limited number of pigs, tissue types, and samples used.
Eflects of tissue type on DNA quality

The disparity between the PCR results of skin and muscle samples was substantial
for nearly every preservation method (Table 14). Silica storage was the only technique

that presented no difference in DNA quality between skin and muscle samples; DNA

from skin was amplified with more success than DNA from muscle in all other cases.

58

The difference in amplification rates between tissue types ranged from 39% (isopropanol)
to 6% (RNAIater). The overall success rate for amplification of the 64pr fragment was
89% for skin samples, compared to 74% of muscle samples (Table 13), which was found
to be statistically significant (p=0.005). These results indicate that tissue type may be an
important consideration when choosing samples for subsequent DNA analyses.

One theory for the discrepancy between tissue types is that skin is naturally more
resistant to decomposition, supported by the fact that skin acts as a physical barrier
against pathogens and moisture loss (Elias and Friend, 1975). Even after death, these
were likely factors in slowing the degradation process. Muscle tissue, in contrast, is
located in the moist environment below the epidermis. Once metabolic by-products and
bacteria began leaching throughout the body cavities, muscle tissue would have been
affected before the skin tissue that was sampled. The universality of the tissue type
discrepancy found in this study presents an interesting bias that warrants investigation.

A more comprehensive analysis of this trend could greatly influence decisions
about which tissues to collect in the event of a disaster. The ideas presented previously
may be tested by quantifying the DNA yield from fresh skin and muscle tissues, in order
to determine if muscle is simply a poorer source of DNA. Second, samples of skin could
be taken from areas of the carcass that were kept moist, such as the surface of the body
that was in contact with the ground. By contrasting these samples with skin that had been
exposed to the air, it would be possible to ascertain whether or not skin was being
preserved to an extent by air-drying and exposure to sunlight. Although all preservation
methods yielded higher quality DNA fiom skin samples, the disparity between tissue

types for each method varied widely. Determining which methods are most versatile

59

across tissue types would be beneficial for situations where skin may not be available,

and a different tissue type must be sammed.

Trends in real time quantification of DNA yields

The results of the real time PCR experiments were unreliable gauges of DNA
quantity. The goal of this phase of the project was to determine which preservation
method(s) yielded the highest amounts of DNA, and to correlate those results to the
corresponding DNA quality. Unfortunately, the number of replicates that consistently
amplified was very low; and the results of each preservation method varied between the
two genes amplified as well. This made it difficult to place a high level of confidence in
the assays. Each DNA extract was assayed in triplicate for the two real time experiments.
Of the 432 triplicate reactions prepared, only 38 amplified in triplicate (Tables 18 and
19), and 130 replicates did not amplify at all. Samples which did amplify in duplicate or
triplicate often did not have similar Ct values, which made interpretation of the results
problematical. Due to the irreproducible nature of the real time PCR experiments, the
results fiom this phase of the project were not factored into the evaluation of preservation
methods.

All reaction mixes and DNA templates were aliquoted by robot for the real time
experiments, which was intended to reduce measuring errors. A mechanical malfunction
in the measurement or dispensation of DNA or reaction mix may explain why replicates
were not exhibiting similar amplification results, although errors in the manual

preparation of the reagents and DNA templates could also have occurred, resulting ill

60

heterogeneity among the sources of input and causing inconsistent results. No
amplification would have occurred if the correct reagents were not present, so the
problem was not likely due to a simple error of omission. The data from real time PCR
experiments may be useful for subsequent molecular analyses; however, further
investigation is needed to identify the sources of error and determine if the assay could be

optimized to achieve consistent amplification results.

Conclusions

The results of this study accent the importance of rapid sampling of tissue from
crime scenes or sites of mass disaster. Regardless of tissue type, storage time, or
preservation method, samples that were obtained a short time after death were more
likely to yield high quality DNA than samples from several days after. Tissue type was
also found to have a significant effect on DNA quality, with skin being the clear choice
for reliable DNA analyses. The effect of storage time was less defined, but in general,
the cause for rapid tissue collection and DNA analysis was supported.

The combination of these factors helped to distinguish the utility of each
preservation method. Preservation in ethanol and RNAIater resulted in the most high
quality DNA extracts, based on amplification of the 64pr fragment (Table 12). Both of
these methods also exhibited minimal decreases in efficacy over a storage time of two
months (Table 16). RNAIater was the most versatile preservation method for different
tissue types, exhibiting a decrease in mean amplification rate of 6% between skin and

muscle (Table 15). As previously discussed, the general availability of RNAIater is one

61

of the main chawbacks with its use in the field, and the decision to use this method would
have to be made in advance in order to maintain a supply on hand. Ethanol has an
advantage over RNAIater in the sense that it is available from a wide range of
manufacturers and venues.

Oven-drying, freezer storage, and isopropanol storage were all found to have
various levels of success; however, DNA extracts from these three methods never
exceeded the quality of DNA from ethanol or RNAIater samples. DNA from oven-dried
samples was inconsistently amplified across amplicon sizes, storage time points, and
tissue types. Coupled with the fact that this method would require extensive equipment,
the field applications may be limited. Storage at -20°C was among the top three methods
in performance for each factor evaluated, but like isopropanol, this method did not have
any clear benefits over the quality of the ethanol and RNAIater samples under these
conditions. Since the transportation of cold storage equipment would be more
cumbersome than vials of ethanol or RNAIater, the utility of -20°C storage must be
considered on a site by site basis. Isopropanol was similar to ethanol in practical
considerations (composition, cost, etc.) yet did not perform as well with regard to DNA
quality, although this may have been largely a factor of sample size. Silica storage was
also one of the less effective preservation methods, but further optimization may increase
the quality of DNA extracts.

The preservation methods evaluated in this study have been used in the field or
the laboratory for years. Under the conditions tested in this study, each method was
shown to have strengths and weaknesses in regard to preserving high quality DNA. The

recommendations presented in this study provide useful information on field and

62

laboratory techniques associated with these methods; firrthermore, many potential factors
have been identified that may result in the preservation of higher quality DNA using each
method. The trends identified can be investigated further and may be useful in any field

applications that require tissue preservation for subsequent DNA analysis.

63

REFERENCES

Akpinar, E.K. 2006. Determination of suitable thin layer drying curve model for some
vegetables and fiuits. Journal of Food Engineering, 73: 75 — 84.

Ambion. 2006. RNAIater Product Description. Accessed online April 23, 2006.
<http://www.ambion.com/techlib/prot/bp__7020.pdf>

Andréasson, H., M. Nilsson, B. Budowle, H. Lundberg, and M. Allen. 2006. Nuclear and
mitochondrial DNA quantification of various forensic materials. Forensic Science
International, in press.

Biesecker, L.G., J .E. Bailey-Wilson, J. Ballantyne, H. Baum, F.R. Bieber, C. Brenner, B.
Budowle, J.M. Butler, G. Carmody, P.M. Conneally, B. Duceman, A. Eisenberg, L.
Forman, K.K. Kidd, B. Leclair, S. Niezgoda, T.J. Parsons, E. Pugh, R. Shaler, S.T.
Sherry, A. Sozer, and A. Walsh. 2005. DNA identifications after the 9/11 World Trade
Center attack. Science, 310: 1122 — 1123.

Bruskov, V.I., L.V. Malakhova, Z.K Masalimov, and AN. Chemikov. 2002. Heat-
induced formation of reactive oxygen species and 8-oxoguanine, a biomarker of damage
to DNA. Nucleic Acids Research, 30: 1354 — 1363.

Butler, J.M., Y. Shen, and DR. McCord. 2003. The development of reduced size STR
amplicons as tools for analysis of degraded DNA. Journal of Forensic Sciences, 48: 1054
— 1064.

Chanda, J., and S. Bandyopadhyay. 2006. Perturbations of phospholipids bilayer
properties by ethanol at high concentration. Langmuir, 22: 3775 - 3781 .

Coble, M.D., and J .M. Butler. 2005. Characterization of new MiniSTR loci to aid
analysis of degraded DNA. Journal of Forensic Sciences, 50: 43 — 53.

Elias, P.M., and D.S. Friend. 1975. The permeability barrier in mammalian epidermis.
The Journal of Cell Biology, 65: 180 — 191.

FBI. 2003. Handbook of Forensic Services.

Foran, D.R., S.C. Minta, and KS. Heinmeyer. 1997. DNA-based analysis of hair to
identify species and individuals for population research and monitoring. Wildlife Society
Bulletin, 25: 840 — 847.

Gillespie, J.W., C.J.M. Best, V.E. Bichsel, K.A. Cole, S.F. Greenhut, S.M. Hewitt, M.
Ahram, Y.B. Gathright, M.J. Merino, R.L. Strausberg, J .I. Epstein, S.R. Hamilton, G.
Gannot, G.V. Baibakova, V.S. Calvert, M.J. Flaig, R.F. Chuaqui, J .C. Herring, J. Pfeifer,
E.F. Petricoin, W.M. Linehan, P.H. Duray, G.S. Bova, and MR. Emmert-Buck. 2002.

Evaluation of non-formalin tissue fixation for molecular profiling studies. American
Journal of Pathology, 160: 449 — 457.

INTERPOL. 2001 . Recommendations from the Interpol DNA Monitoring Expert Group.

Kilpatrick, CW. 2002. Noncryogenic preservation of mammalian tissues for DNA
extraction: An assessment of storage methods. Biochemical Genetics, 40:53 — 62.

Lipton, E. 2005. At limits of science, 9/11 II) effort comes to end. New York Times,
April 3, p29.

Murphy, M.A., L.P. Waits, K.C. Kendall, S.K. Wasser, J.A. Higbee, and R. Bogden.
2002. An evaluation of long-term preservation methods for brown bear (Ursus arctos)
faecal DNA samples. Conservation Genetics, 3: 435 — 440.

Mutter, G.L., D. Zahrieh, C. Liu, D. Neuberg, D. Finkelstein, H.E. baker, and J .A.
Warrington. 2004. Comparison of frozen and RNALater solid tissue storage methods for
use in RNA expression microarrays. BMC Genomics 5:88 —— 94.

NAME. 2005. National Association of Medical Examiners Mass Fatality Plan.
NI]. 2005. Mass Fatality Incidents: A Guide for Human Forensic Identification.

Penna, T.C.V., P.G. Mazzola, and A.M.S. Martins. 2001. The efficacy of chemical agents
in cleaning and disinfection programs. BMC Infectious Diseases 1: l - 16.

Pfeiffer, H., J. Hflhne, C. Ortrnann, K. Waterkamp, and B. Brinkrnann. 1999.
Mitochondrial DNA typing from human axillary, pubic, and head hair shafts—success
rates and sequence comparisons. International Journal of Legal Medicine, 112: 287 —
290.

Roeder, A.D., F.I. Archer, H.N. Poinar, and RA. Morin. 2004. A novel method for
collection and preservation of faeces for genetic studies. Molecular Ecology Notes, 4:
761 — 764.

Seutin, G., B.N. White, and RT. Boag. 1991. Preservation of avian blood and tissue
samples for DNA analyses. Canadian Journal of Zoology, 69: 82 — 90.

Srinivasan, M., D. Sedmak, and S. Jewell. 2002. Effect of fixatives and tissue processing
on the content and integrity of nucleic acids. American Journal of Pathology, 161: 1961 -
1971.

United States Food Safety Inspection Service. 2003. Safety of Fresh Pork. Accessed
April 23, 2006.
<http://www.fsis.usdagov/Frame/FrameRedirect.asp?main=http://www.fsis.usda.gov/oa/
pubs/pork.htm>

65

lift/Will!!!Will/II/lllI/M/lll/llllllI/l/l/I