Вы находитесь на странице: 1из 9

Forensic Science International: Genetics 4 (2010) 95103

Contents lists available at ScienceDirect

Forensic Science International: Genetics


journal homepage: www.elsevier.com/locate/fsig

Authentication of forensic DNA samples


Dan Frumkin a,*, Adam Wasserstrom a,*, Ariane Davidson a, Arnon Grat b
a
Nucleix Ltd., 27 Habarzel St., Tel Aviv 69710, Israel
b
Serious Crime Unit Mobile Lab., Division of Identication & Forensic Science, Israel Police, Israel

A R T I C L E I N F O A B S T R A C T

Article history: Over the past twenty years, DNA analysis has revolutionized forensic science, and has become a
Received 3 April 2009 dominant tool in law enforcement. Today, DNA evidence is key to the conviction or exoneration of
Received in revised form 11 June 2009 suspects of various types of crime, from theft to rape and murder. However, the disturbing possibility
Accepted 16 June 2009
that DNA evidence can be faked has been overlooked. It turns out that standard molecular biology
techniques such as PCR, molecular cloning, and recently developed whole genome amplication (WGA),
Keywords: enable anyone with basic equipment and know-how to produce practically unlimited amounts of in vitro
Articial DNA
synthesized (articial) DNA with any desired genetic prole. This articial DNA can then be applied to
Methylation analysis
surfaces of objects or incorporated into genuine human tissues and planted in crime scenes. Here we
DNA authentication
Forensic science show that the current forensic procedure fails to distinguish between such samples of blood, saliva, and
touched surfaces with articial DNA, and corresponding samples with in vivo generated (natural) DNA.
Furthermore, genotyping of both articial and natural samples with Proler Plus1 yielded full proles
with no anomalies. In order to effectively deal with this problem, we developed an authentication assay,
which distinguishes between natural and articial DNA based on methylation analysis of a set of
genomic loci: in natural DNA, some loci are methylated and others are unmethylated, while in articial
DNA all loci are unmethylated. The assay was tested on natural and articial samples of blood, saliva, and
touched surfaces, with complete success. Adopting an authentication assay for casework samples as part
of the forensic procedure is necessary for maintaining the high credibility of DNA evidence in the
judiciary system.
2009 Elsevier Ireland Ltd. All rights reserved.

1. Introduction by an automatic search of the database (e.g. CODIS). In recent years,


DNA evidence has become the gold standard of forensic testing,
The current forensic procedure that deals with DNA evidence and is an invaluable tool for the criminal justice community [37].
starts at the crime scene where biological samples such as blood The high credibility of DNA evidence in court stems from the fact
and saliva stains are detected, identied, documented, collected, that it uses a statistical approach based on population genetics and
and transferred to the forensic laboratory. In the laboratory, DNA is empirical testing [8], in contrast to other types of forensic
extracted and quantied, usually by real time PCR amplication of evidence, such as ballistics, blood-spatter analysis, and ber
the hTERT locus (Quantiler1) or similar targets [1]. Following analysis, which rely on expert judgment and have limited
quantication, about 1 ng of the DNA is used for a proling connection to established science [7]. It is even considered to be
reaction, in which 915 highly polymorphic short tandem repeat more reliable than eyewitness evidence, which is known to suffer
(STR) loci and the sex-typing marker amelogenin are genotyped. from a relatively high rate of errors [8].
The loci are usually chosen from a standard set of core loci such as The use of DNA recovered at crime scenes as evidence in court
the 13 Combined DNA Index System (CODIS) loci. A detailed relies on the implicit assumption that the DNA is genuine
description of the forensic procedure is provided in Text S1. originating from natural biological material. However, as we
The DNA prole of every person is considered unique (except show here, this assumption may not necessarily be true: DNA
for identical twins) [2], and consequently, this DNA ngerprint is with any desired genetic prole can easily be synthesized in vitro
used in police investigations to link between a crime scene and a using common [9,10], and recently developed [11,12] biological
specic individual, who is either a suspect in the case, or identied techniques, integrated into genuine human tissues or applied to
surfaces of objects, and then planted in crime scenes. When the
current forensic procedure is applied to objects or human tissues
* Corresponding authors. Tel.: +972 3 768 4935; fax: +972 3 768 4945.
that contain synthesized DNA, it fails to recognize the articial
E-mail addresses: dan@nucleix.com (D. Frumkin), adam@nucleix.com origin of the sample, and the resulting prole is indistinguish-
(A. Wasserstrom). able from a genuine DNA prole. Nevertheless, we demonstrate

1872-4973/$ see front matter 2009 Elsevier Ireland Ltd. All rights reserved.
doi:10.1016/j.fsigen.2009.06.009
96 D. Frumkin et al. / Forensic Science International: Genetics 4 (2010) 95103

that natural and articial samples can be differentiated based on another is allele 12 of D8S1179, and likewise for the other CODIS
differential methyaltion patterns. Methylation is an epigenetic loci). We note that 1 fg of plasmid in the library contains 160
chemical modication of DNA, occurring in mammals in the copies of its cloned allelethe same copy number that is present
form of a methyl group (CH3) that is enzymatically added to in 1 ng of a haploid genome.
the C5 position of cytosine in some CpG dinucleotides [13]. DNA
methylation is believed to inhibit gene expression in animal 2.3. In vitro synthesis of DNA
cells, probably by affecting chromatin structure [14]. In the
human genome 7080% of all CpGs are methylated, while Articial DNA was synthesized by one of the following
unmethylated CpGs are grouped in clusters called CpG islands methods:
[15]. PCR: For the sample whose prole is shown in Fig. 1, the 10
loci included in the Proler Plus1 kit (Applied Biosystems) were
2. Materials and methods amplied separately from 1 ng of DNA extracted from a cigarette
butt smoked by N400 (PCR conditions were as described in
2.1. Collection of biological tissues Section 2.9; primer sequences are in Text S3). Individual
amplied fragments were puried (QIAquick PCR purication
Samples of blood, dry saliva stains on absorbent paper, skin kit, QIAGEN), quantied (Nanodrop 1000, Thermo Scientic),
scrapings, hair, and smoked cigarette butts were collected from diluted about a million fold (depending on the concentration of
volunteers. Informed consent was obtained from all participants the specic amplicon), and combined in a single test tube. For the
recruited into the study. DNA from these samples was extracted sample whose prole is shown in Fig. 2, 1 ng of N222 DNA
and quantied as described in Section 2.6. (extracted from a saliva stain on absorbent paper) was used as
template in a single PCR reaction using the Proler Plus1 primer
2.2. CODIS allele library mix. A 1:1000 dilution of the PCR reaction was used for
generating the articial sample.
For construction of the library, individual alleles of CODIS WGA: Whole genome amplication was performed by
STRs and the hTERT locus were amplied from pooled DNA multiple displacement amplication [16] with the Repli-g Midi
(Control Human Genomic DNA of the GenomePlex WGA2 kit, kit (QIAGEN) using 10 ng of natural DNA as template.
SigmaAldrich) by separate PCR reactions (primers and condi- Assembly from CODIS allele library: For assembling proles using
tions as described in Section 2.9). Amplied fragments were the CODIS allele library, equal quantities of alleles (cloned into
puried (QIAquick PCR purication kit, QIAGEN), and cloned into plasmids) in the desired prole were picked from the library and
the pGEM-T-Easy vector (Promega). Plasmid DNA was puried by combined in a single tube.
the QIAprep Spin Miniprep kit (QIAGEN) and quantied
(Nanodrop 1000, Thermo Scientic). For genotyping of cloned 2.4. Generation of mock forensic samples
alleles, the PowerPlex16 (Promega) kit was used. Genotyping
was performed in a high throughput manner by simultaneously For generating articial touch DNA samples, in vitro synthe-
genotyping 1015 clones (from different CODIS loci) in a single sized DNA was applied directly to the surface of the object and
PowerPlex16 reaction. In the resulting library each element is a allowed to dry. For generating articial blood samples, red blood
microcentrifuge tube with trillions of copies of a single allele (for cells were isolated from whole blood by centrifugation (1500  g,
example, one element is allele 11 of locus D8S1179, while 10 min), and mixed with in vitro synthesized DNA. Drops of the red

Fig. 1. Proles of in vivo- and in vitro-synthesized DNA are indistinguishable. (A) Prole of natural DNA obtained from the saliva of female donor N400. (BD) Articial proles
of N400 obtained from DNA that was synthesized in vitro by three different methods: PCR (B), WGA (C), and assembly from a library of cloned CODIS alleles (D). (E) Articial
prole of male-N400, which is identical to the prole of N400 at all loci, except for the Amelogenin locus. This prole was created by adding a cloned Y allele (indicated by
arrow) to the mix used to generate the prole in (D). In AE partial proles are depicted; full proles are provided in Text S2.
D. Frumkin et al. / Forensic Science International: Genetics 4 (2010) 95103 97

Fig. 2. Mock forensic samples with articial DNA. (A) Handgun with PCR amplied DNA with the prole of N222 applied to the external surface of its action. (B) Ski-mask with
articial saliva applied to its inner surface. The articial saliva contained an extract of natural saliva from N270 (without DNA) and DNA fragments with the prole of male
N400 assembled from the CODIS allele library. (C) Articial bloodstains containing red blood cells from natural blood of N227 and articial N283 DNA generated by WGA. In
AC, yellow circles depict the areas from which samples were taken for analysis. (D) Proles of the three articial samples. All three proles received a perfect GeneMapper
ID-X score, and are identical to the genotypes of the articial DNA that was used in their production. No traces of DNA from the saliva extract and red blood cells are visible in
the proles from the ski-mask and bloodstains (see E and F). (E) Prole of donor N270, whose saliva extract was used for manufacturing the ski-mask sample. (F) Prole of
donor N227, whose red blood cells were used for manufacturing the bloodstain sample. In DF partial proles are depicted; full proles are provided in Text S2.

blood cellDNA mix were dripped from a height of 1 m and 2.7. DNA proling, capillary electrophoresis and signal analysis
allowed to dry. For generating articial saliva samples, saliva
extract (containing no cells) was isolated from the top phase of STR loci were amplied using the Proler Plus1 (Applied
centrifuged natural saliva (1500  g, 10 min), and mixed with in Biosystems) and PowerPlex16 (for preparing the CODIS allele
vitro synthesized DNA. The saliva extractDNA mix was applied library; Promega) kits using a GeneAmp1 PCR System 9700 (Applied
directly to the surface of the object and allowed to dry. A detailed Biosystems). Amplication products were run on an ABI 310 Genetic
description of all samples is provided in Text S4. Analyzer (Applied Biosystems) according to the manufacturers
instructions. The resulting electropherograms were analyzed using
2.5. Identication and collection of mock forensic samples GeneMapper ID-X analysis software (Applied Biosystems).

Stains were identied as human blood using the HEXAGON 2.8. Bisulte conversion and methylation analysis
OBTI kit (BLUESTAR), and as saliva using Phadebas1 Amylase test
(Phadebas). Samples of blood and touch DNA were collected with a Bisulte conversion was performed with the EpiTectTM kit
sterile cotton swab, dampened with distilled water. Saliva samples (Qiagen). Converted DNA was amplied by PCR at the set of loci
were composed of cut-out portions of the ski-mask fabric around used for authentication. In each PCR, 1/10 of the EpiTectTM
the mouth orice. products were used as template and the reaction was performed
as described in Section 2.9. Amplied fragments were puried
2.6. DNA extraction and quantication using the QIAquick PCR purication kit (QIAGEN) and
sequenced.
DNA extraction from all samples was performed according to
an organic extraction protocol [17]. DNA quantication was 2.9. PCR
performed using the Quantiler1 Human DNA quantication
kit (Applied Biosystems). Real time PCR was performed on All non-proling PCRs were performed in a total volume of
a StepOneTM system (Applied Biosystems). 50 ml with 0.2 mM each primer, 0.2 mM each dNTP, 5 U AmpliTaq
98 D. Frumkin et al. / Forensic Science International: Genetics 4 (2010) 95103

Gold (Applied Biosystems), and 5 ml 10 PCR Buffer containing 3.2. The current forensic procedure fails to distinguish between
15 mM MgCl2 (Applied Biosystems). Amplication was performed natural and articial DNA evidence
in a GeneAmp1 PCR System 9700 (Applied Biosystems). The PCR
program used was: 95 8C for 11 min, followed by 35 cycles of 94 8C 3.2.1. Generation of articial DNA evidence
for 1 min, 59 8C for 1 min, 72 8C for 1 min, and followed by a nal We created 10 mock forensic samples with articial DNA, of
extension step of 60 8C for 45 min. PCRs for proling reactions types that may be found in crime scenes, and subjected three of
were performed according to the manufacturers instructions these samples to analysis through the complete forensic procedure
(with 28 cycles). (the rest of the samples are discussed in Section 3.4). These three
samples contained articial DNA that was synthesized using
2.10. Probability of non-existent prole different methods: a handgun sample with PCR amplied DNA, a
ski-mask with saliva containing DNA fragments from the CODIS
The probability that a random unrelated male has the Proler allele library, and bloodstains containing DNA synthesized by WGA
Plus1 prole of male-N400 (a prole identical to that of N400 (Fig. 2AC; see detailed description in Text S4).
with the exception of the Amelogenin locus, in which its genotype
is XY instead of XX) was calculated based on allele frequencies in 3.2.2. Analysis of articial DNA evidence
the US Caucasian population [18]. This probability was multiplied The three samples were processed according to the routine
by 3.5  109 (approximate male population) to yield the approx- forensic procedure performed in crime scenes. Samples were
imate probability that there exists a person with the male N400 collected from the external surface of the handgun action, from the
prole (excluding close relatives of N400). ski-mask fabric, and from the bloodstains. A portion of the ski-
mask sample was tested for presence of saliva using the Phadebas1
2.11. DNA mixtures assay, and the results were positive (data not shown), due to the
presence of amylase in the supernatant of the natural saliva
DNA mixtures were created by combining natural N217 DNA extract. A portion of the bloodstain sample was tested for the
(extracted from blood with the FlexiGene DNA kit, QIAGEN) with presence of human blood DNA using the HEXAGON OBTI assay, and
articial N226 DNA (amplied by Repli-g Midi kit, QIAGEN from the results were positive (data not shown), due to the presence of
DNA extracted by organic extraction from a single hair). hemoglobin in the red blood cells. DNA was extracted from the
samples and proled (Fig. 2D). The genotypes of all three samples
were identical to the genotypes of the articial DNA that was used
3. Results in their production (N222, male N400, and N283, respectively).
Furthermore, in the articial saliva and blood samples there were
3.1. Proles of in vivo- and in vitro-synthesized DNA are no observable traces of natural DNA from the saliva and blood
indistinguishable donors (Fig. 2E and F), and all articial proles received a perfect
GeneMapper ID-X score, consistent with a single contributor.
To demonstrate that DNA can be synthesized in vitro such that
its prole will be indistinguishable from that of DNA of in vivo 3.2.3. Independent analysis of articial blood evidence
origin, we proled a natural DNA sample and compared it to In order to check whether the proling results obtained in our
corresponding proles from DNA that was synthesized in vitro by laboratory were dependant on our specic setup, we sent a
three different methods. Natural DNA was extracted from a saliva duplicate swab of the articial blood sample to a leading forensic
sample of female donor N400 and genotyped using the Proler DNA laboratory for analysis. The procedures employed by this
Plus1 and GeneMapper ID-X (Applied Biosystems); (Fig. 1A). The laboratory have been validated according to standards established
prole obtained from the saliva of donor N400 was perfect, as by the Scientic Working Group on DNA Analysis Methods
indicated by the green bars above all loci. (SWGDAM) and adopted as US Federal Standards. DNA was
Next, we produced articial DNA with the same genotype as extracted from the sample in the laboratory using the EZ1 DNA
N400 using three different types of in vitro synthesis: PCR Investigator Kit (QIAGEN), and quantied using a proprietary real
(Fig. 1B), WGA (Fig. 1C), and molecular cloning (Fig. 1D). The time PCR assay (both extraction and quantication methods were
genotypes of all in vitro synthesized N400 samples were perfect different than those employed in our lab). Genotyping was
according to GeneMapper ID-X analysis and identical to the performed with Proler Plus1 and COler1 (Applied Biosystems).
genotype of natural N400 DNA. Template DNA for PCR was 1 ng of The report received from the laboratory states that The DNA
DNA extracted from a cigarette butt smoked by N400, and the prole obtained from sample 2S09-002-001 [the articial blood
template for WGA was 10 ng of N400 DNA extracted from a dry swab] is consistent with a male contributor, and the proling
saliva stain on absorbent paper. The sample created by molecular results were identical to the genotype of the articial DNA of donor
cloning did not require any N400 DNA as template (only a priori N283, with No Edits (i.e. no anomalies found in any of the
knowledge of her prole) and was assembled using the CODIS analyzed loci; see report in Text S5).
allele library that we created beforehand. The library consists of These results demonstrate that articial DNA can easily be
an array of single CODIS alleles cloned into plasmids, and can be applied to surfaces of objects or incorporated into genuine human
used to generate different desired proles by assembly of their tissues, thereby creating articial forensic evidence that, after
constituent alleles. In order to demonstrate the possibility to create undergoing the entire forensic casework procedure, yields perfect
any desired prole from such a library, we also assembled a prole proles.
of a non-existent person, which we term male N400. This prole is
identical to that of N400, with the exception of the Amelogenin 3.3. Description of the DNA authentication assay
locus, in which its genotype is XY instead of XX (Fig. 1E). We
calculated that the probability that a male unrelated to N400 has a We developed an authentication assay capable of differentiat-
prole identical to that of male N400 is 7.95  10 12, and ing between natural and all types of articial DNA. The assay is
consequently the probability that there does not exist in the world based on the fact that unlike in vitro synthesized DNA which is
population an unrelated male with an identical prole is greater completely unmethylated, in vivo generated DNA contains loci that
than 99.99%. are completely and consistently methylated and other loci that are
D. Frumkin et al. / Forensic Science International: Genetics 4 (2010) 95103 99

Fig. 3. Scheme of the DNA authentication assay. The assay accepts as input DNA that was extracted from the forensic sample in question and outputs whether the DNA is
authentic (in vivo generated) or non-authentic (in vitro synthesized).

completely and consistently unmethylated. A scheme of the assay between these two types of DNA is achieved by sequencing the
is presented in Fig. 3. DNA from a forensic sample in question is four non-CODIS amplicons and analyzing their methylation
treated with sodium bisulte, which converts all unmethylated pattern. The DNA is determined to be of in vivo origin if its
cytosines to uracils, while leaving the methylated cytosines methylation pattern is consistent with that of in vivo generated
unaffected [19]. Following bisulte conversion, the DNA is DNA (i.e. complete methylation of all CpGs in NT18 and ADD6
amplied by PCR at a set of loci, containing one reference CODIS alongside with complete non-methylation of all CpGs in MS53 and
locus (FGAref), and four non-CODIS loci (NT18, ADD6, MS53, SW14), otherwise it is determined to be of in vitro origin.
SW14; Text S6). The set of loci consist of high-complexity, non-
repetitive DNA sequences (FGAref consists of the non-repetitive 3.4. Demonstration of the DNA authentication assay
part of the FGA locus). They were chosen because NT18 and ADD6
are consistently methylated, while MS53 and SW14 are consis- We applied the DNA authentication assay to 20 mock forensic
tently unmethylated in human tissues such as blood, saliva and samples, 10 with natural DNA, 10 with articial DNA (three of
epidermis (the source of touch DNA). For increasing the reliability these samples were described in Section 3.2), and a negative
of the assay, the primers for these loci were designed to amplify control sample without DNA (detailed description in Text S4). All
with equal efciency both converted and unconverted DNA, thus samples with natural DNA showed successful amplication of all
enabling detection of incomplete bisulte conversion. Following loci, and the FGAref amplicon was present in all samples, both
PCR, the presence or absence of amplicons is determined by natural and articial (Fig. 4). Samples 13,14,16,17,19,20 which
electrophoresis (alternatively, real time PCR can be used). contain articial DNA synthesized by PCR or molecular cloning,
Complete absence of amplicons (including FGAref) indicates a failed to amplify the four non-CODIS loci, since the DNA in these
problem in the procedure due to PCR inhibitors, insufcient samples contains only CODIS loci. These samples were therefore
template, etc. Successful amplication of FGAref with concomitant determined to be non-authentic and were not processed further.
failure of amplication of the non-CODIS loci indicate that the DNA The remaining articial DNA samples (11,12,15,18) contained
is articial and was synthesized by one of the methods that WGA-synthesized DNA and in these samples all loci amplied
generate only a subset of genomic loci (e.g. PCR or cloning of CODIS successfully, similar to natural DNA.
loci). Successful amplication of all loci indicates that the DNA The natural and WGA-synthesized DNA samples were processed
contains a full representation of the genome and is either natural further by sequencing the four non-CODIS loci and analyzing the
DNA or articial DNA synthesized by WGA. Differentiation methylation status at all CpG positions (Table 1). All natural DNA
100 D. Frumkin et al. / Forensic Science International: Genetics 4 (2010) 95103

Fig. 4. Amplication products in natural and articial mock forensic samples. Aliquots of PCR products were run on a 2% Agarose gel. The FGAref locus is amplied in all
samples (both natural and articial), but not in the negative control sample. Non-CODIS loci are amplied in all natural (110) and in WGA-based articial samples
(11,12,15,18), but are absent in PCR- and cloning-based articial samples (13,14,16,17,20).

samples showed complete methylation of all CpG positions in NT18 mixture, which was identied as a single contributor sample
and ADD6, and no methylation in any of the CpG positions in MS53 with a prole identical to that of the natural DNA (Text S7). In
and SW14. In contrast, all WGA-synthesized samples showed methyaltion analysis of the mixtures, natural and articial
complete lack of methylation in all loci (Fig. 5). Based on this sequence signals are superimposed and signals from both types
analysis, the 10 natural samples were determined to be authentic, of DNA are observed even in the 10% articial DNA mixture (Text
and the four WGA-synthesized samples were determined to be non- S7). These results indicate that articial DNA can be detected in
authentic. Therefore the assay was successful in determining the DNA mixtures, even when it constitutes a minor component.
correct status of all 20 samples (Table 1).
4. Discussion
3.5. Naturalarticial DNA mixtures
4.1. Producing articial DNA evidence requires only basic equipment
We checked the authentication assay on mixtures of natural and know-how
N217 DNA and articial N226 DNA with various ratios of DNA
(the percentage of articial DNA ranged from 10 to 67%). The We demonstrated the ease at which articial DNA evidence can
mixtures were proled with GeneMapper ID-X software using the be produced, and that such evidence passes the current forensic
mixture analysis mode and all mixtures were correctly identied procedure as genuine. The fact that an independent forensic
as two-contributor mixtures, except for the 10% articial DNA laboratory, which provides services to United States law enforce-

Table 1
DNA authentication results on natural and articial mock forensic samples.

# Sample FGAref amplied Methylated CpG positionsa Decision

Source of DNA NT18 ADD6 MS53 SW14

1 In vivo (blood) Yes 12/12 11/11 0/6 0/17 Authentic


2 In vivo (blood) Yes 12/12 11/11 0/6 0/17 Authentic
3 In vivo (blood) Yes 12/12 11/11 0/6 0/17 Authentic
4 In vivo (blood) Yes 12/12 11/11 0/6 0/17 Authentic
5 In vivo (saliva) Yes 12/12 11/11 0/6 0/17 Authentic
6 In vivo (saliva) Yes 12/12 11/11 0/6 0/17 Authentic
7 In vivo (saliva) Yes 12/12 11/11 0/6 0/17 Authentic
8 In vivo (skin) Yes 12/12 11/11 0/6 0/17 Authentic
9 In vivo (skin) Yes 12/12 11/11 0/6 0/17 Authentic
10 In vivo (skin) Yes 12/12 11/11 0/6 0/17 Authentic
11 In vitro (WGA) Yes 0/12 0/11 0/6 0/17 Non-authentic
12 In vitro (WGA) Yes 0/12 0/11 0/6 0/17 Non-authentic
13 In vitro (PCR) Yes No amp. No amp. No amp. No amp. Non-authentic
14 In vitro (Cloning) Yes No amp. No amp. No amp. No amp. Non-authentic
15 In vitro (WGA) Yes 0/12 0/11 0/6 0/17 Non-authentic
16 In vitro (PCR) Yes No amp. No amp. No amp. No amp. Non-authentic
17 In vitro (Cloning) Yes No amp. No amp. No amp. No amp. Non-authentic
18 In vitro (WGA) Yes 0/12 0/11 0/6 0/17 Non-authentic
19 In vitro (PCR) Yes No amp. No amp. No amp. No amp. Non-authentic
20 In vitro (Cloning) Yes No amp. No amp. No amp. No amp. Non-authentic
21 Negative Control No No amp. No amp. No amp. No amp. No decisionb
a
Number of methylated CpG positions out of total number of CpG positions in each locus. No amp. = No amplicon observed; bold indicates results inconsistent with DNA of
in vivo origin.
b
No decision is outputted when there is no amplication in any of the loci. Possible reasons may be insufcient/degraded template DNA, PCR inhibitors, etc.
D. Frumkin et al. / Forensic Science International: Genetics 4 (2010) 95103 101

Fig. 5. Methylation analysis of natural and articial samples. Partial sequences of DNA from natural and articial blood samples (samples 2 and 11, respectively) at non-CODIS
loci (CpG dinucleotides are underlined). The sequences of unconverted DNA are identical at all loci, demonstrating that natural and articial samples cannot be distinguished
on the basis of sequence alone. Following bisulte conversion, the differential methylation pattern of natural vs. articial DNA is exposed: natural DNA is methylated at NT18
and ADD6, and unmethylated at MS53 and SW14, while articial DNA is unmethylated at all four loci.

ment agencies, analyzed our articial blood sample yielding a DNA can be performed overnight, using basic laboratory equip-
perfectly normal, single contributor DNA proleattests to the ment and commercial kits, requires only basic knowledge in
problem. molecular biology, and little nancial expense. There is a very large
In this case the articial DNA was designed to have the prole of and growing number of people with the necessary expertise and
donor N283, and was amplied from a minute amount of DNA access to the required equipment, such as scientists, research
extracted from a single hair of this donor. Similarly, we produced students, lab technicians in hospitals, pharmaceutical or biotech
articial samples of DNA amplied from a cigarette butt and a dry companies, etc. Such people might manufacture articial DNA and
saliva stain on absorbent paper. Such common everyday objects, use it maliciously themselves, or transfer it to other people who do
which can be used to obtain source DNA for producing articial not have the ability to manufacture the DNA. Moreover, since
samples, can be obtained from practically anyone. Even this commercial molecular biology services are becoming widespread
constraint is removed when considering the possibility to produce and DNA with any sequence can be ordered online, manufacturing
articial evidence using the CODIS allele library, since any prole an articial DNA sample does not require much more than a
can be assembled without the need for source DNA, only requiring personal computer and link to the internet.
knowledge of the desired prole. A library containing 425 clones
corresponding to all known CODIS alleles (including all rare micro- 4.2. Authentication is necessary for preventing false DNA matches
variants) is sufcient to generate any desired prole, while a much
smaller library is sufcient to generate the proles of the vast The DNA proles of millions of people are registered in rapidly
majority of the human population. growing national databases, and the current trend around the
Once source DNA from a person or knowledge of his/her prole world is to include more and more proles in them, not only of
is obtained, the actual manufacturing of the articial sample is convicted offenders, but also of arrestees. Proles from casework
simple and straightforward. Generating large amounts of articial samples are routinely searched against these databases (e.g. by
102 D. Frumkin et al. / Forensic Science International: Genetics 4 (2010) 95103

automatic software such as CODIS), and when an identical prole is 4.6. Other approaches to DNA authentication
found, a DNA match is made, making the identied person a
suspect in the case and usually leading to his arrest [20]. In some Analysis of methylation patterns represents only one of several
jurisdictions, DNA evidence alone can lead to conviction without possible approaches that can be used for DNA authentication.
the requirement of any corroborating evidence [21]. However, Alternative methods may be based on analysis of stutter products,
even when supporting evidence is required by law, there is little representation bias, distribution of DNA fragment sizes, and
doubt that the presence of DNA evidence from a crime scene presence of non-genomic sequences. Stutter products are artifacts
against a defendant places him/her at a dire position. caused by slippage of the DNA polymerase on repeated sequences
The combination of the ease at which articial DNA samples can [27]. Upon proling, articial DNA that was pre-amplied by PCR
be manufactured, with the fact that a registered DNA prole found undergoes more amplication cycles than natural DNA (pre-
at a crime scene will automatically lead to a database match, and amplication also includes several cycles of PCR) and because a
the heavy weight of DNA evidence in the courtroom, creates a higher number of cycles is associated with elevated stutter levels
problematic situation which we believe should be addressed by the [28], such pre-amplied DNA may be distinguished from natural
forensic community by adopting a DNA authentication assay for DNA by higher stutter percentages. Representation bias refers to
casework samples. differences in copy number between different genomic loci that are
Articial DNA with unregistered prole can also be produced, an inherent consequence of in vitro amplication of DNA [11].
and in such a case it will not result in a match. Nevertheless, this Since this bias might not necessarily be apparent in the small set of
articial DNA evidence might hamper the investigation and CODIS loci used for proling, it may be necessary to analyze a wider
authentication of this DNA could aid in focusing the investigation set of loci. Analysis of the distribution of fragment sizes can also
to relevant directions. reveal the origin of the DNA: in natural DNA, the distribution has
an expected stereotypical pattern (which is a function of the
4.3. SNP based proling approaches are also susceptible to fabrication extraction method used and the extent of degradation), different
from the patterns observed in various types of in vitro synthesized
Recently, alternatives to STR based proling have been DNA. Non-genomic sequences such as primer dimers, plasmid
proposed, primarily single nucleotide polymorphism (SNP) based sequences, articial oligonucleotide linkers, etc., are not expected
approaches [22], which may be advantageous over STR proling to be found in natural DNA (with the possible exception of bacterial
[2325]. Similar to STR based proling, SNP based approaches are sequences), but are expected to be found in various types of in vitro
also susceptible to fabrication by the methods described here. Even synthesized DNA.
if a very large number of SNPs are to be used in proling, this will
not effectively deal with the problem of WGA-based fabrication, 5. Conclusion
since WGA produces a full representation of the genome, and
therefore is expected to produce a perfect SNP prole. In this work we address the disturbing possibility that DNA
evidence can be faked and planted in crime scenes, and the current
4.4. Integrating DNA authentication into the forensic procedure inability of the forensic procedure to detect such articial
evidence. We present a solution to this problem in the form of
The assay described here employs bisulte sequencing, a a DNA authentication assay that can distinguish between natural
procedure that is relatively labor intensive, time consuming, and and articial DNA evidence. In order to preserve the high
requiring specic expertise, and therefore may be best suited as a credibility of DNA evidence in the courtroom, if there is a concern
service provided by dedicated labs to the forensic community. regarding the possible authenticity of DNA evidence, an approach
However, in order to reduce costs and possible backlogs, and to such as presented here may be taken to test for the presence of
reduce the risks of errors related to lengthening of the chain of articial DNA.
custody, it may be advantageous to develop an integrated DNA
authentication assay that will be performed in existing forensic
Acknowledgements
laboratories, as part of the regular forensic procedure. The question
of integrating DNA authentication into the forensic procedure also
We thank Howard Cedar for his insights on methyaltion
has legal aspects, and therefore we hope this work will invoke a
analysis, Eran Eldar for performing blood extraction, Shira Grat
discussion in legal as well as in scientic circles.
for help with the gures, and Avraham Levy, Yuval Dor, Yaakov
Benenson, Eliahu Wasserstrom, Elon Ganor, and Meirav Chovav for
4.5. DNA mixtures
critical review of this manuscript.

Articial DNA evidence can contain pure articial DNA or a


mixture of articial and natural DNA. For example, such a mixture Appendix A. Supplementary data
may contain articial DNA incorporated in or applied onto
genuine tissues from the victim (e.g. blood, ngernails). As we Supplementary data associated with this article can be found, in
demonstrated, mixture samples can be authenticated in the same the online version, at doi:10.1016/j.fsigen.2009.06.009.
manner as single source samples, and the articial DNA is
detected even when it is a minor component of the mixture. The References
automatic software used for sequencing assigns a nucleotide at a
certain position when the template is pure or contains a major [1] J.M. Butler, Forensic DNA Typing. Biology, Technology and Genetics of STR
component, and outputs N when there is ambiguity. Therefore, Markers, 2nd ed., Elsevier Academic Press, Burlington, 2005.
[2] A.J. Jeffreys, V. Wilson, S.L. Thein, Individual-specic ngerprints of human DNA,
articial DNA can be detected automatically in samples when it is
Nature 316 (1985) 7679.
the major component, using existing sequencing software. [3] M.A. Jobling, P. Gill, Encoded evidence: DNA in forensic analysis, Nat. Rev. Genet. 5
However, the interpretation of mixtures in which the articial (2004) 739751.
DNA is a minor component is more complex and may require the [4] N. Morling, Forensic genetics, Lancet 364 (Suppl. 1) (2004) s10s11.
[5] P. Reilly, Legal and public policy issues in DNA forensics, Nat. Rev. Genet. 2 (2001)
development of guidelines, similar to those that have been 313317.
suggested for proling [26]. [6] A. Opar, Crime and punishment, Nat. Med. 12 (2006) 11101111.
D. Frumkin et al. / Forensic Science International: Genetics 4 (2010) 95103 103

[7] M. Lynch, Gods signature: DNA proling, the new gold standard in forensic [19] M. Frommer, L.E. McDonald, D.S. Millar, C.M. Collis, F. Watt, G.W. Grigg, P.L.
science, Endeavor 27 (2003) 9397. Molloy, C.L. Paul, A genomic sequencing protocol that yields a positive display of
[8] M.J. Saks, J.J. Koehler, The coming paradigm shift in forensic identication science, 5-methylcytosine residues in individual DNA strands, Proc. Natl. Acad. Sci. U.S.A.
Science 309 (2005) 892895. 89 (1992) 18271831.
[9] D. Stirling, J.M.S. Bartlett (Eds.), PCR Protocols (Methods in Molecular Biology), [20] J.W. Bond, C. Hammond, The value of DNA material recovered from crime scenes,
2nd edition, Humana Press, New Jersey, 2003. J. Forensic Sci. 53 (2008) 797801.
[10] J. Sambrook, Molecular Cloning: A Laboratory Manual, Cold Spring Harbor [21] M. Levitt, Forensic databases: benets and ethical and social costs, Br. Med. Bull.
Laboratory Press, New York, 2001. 83 (2007) 235248.
[11] R.S. Lasken, M. Egholm, Whole genome amplication: abundant supplies of DNA [22] B. Sobrino, M. Brion, A. Carracedo, SNPs in forensic genetics: a review on SNP
from precious samples or clinical specimens, Trends Biotechnol. 21 (2003) 531535. typing methodologies, Forensic Sci. Int. 154 (2005) 181194.
[12] S. Panelli, G. Damiani, L. Espen, G. Micheli, V. Sgaramella, Towards the analysis of [23] J.M. Butler, M.D. Coble, P.M. Vallone, STRs vs. SNPs: thoughts on the future of
the genomes of single cells: further characterisation of the multiple displacement forensic DNA testing, Forensic Sci. Med. Pathol. 3 (2007) 200205.
amplication, Gene 372 (2006) 17. [24] K. Babol-Pokora, J. Berent, SNP-minisequencing as an excellent tool for analysing
[13] T.B. Miranda, P.A. Jones, DNA methylation: the nuts and bolts of repression, J. Cell degraded DNA recovered from archival tissues, Acta Biochim. Pol. 55 (2008) 815
Physiol. 213 (2007) 384390. 819.
[14] T. Hashimshony, J. Zhang, I. Keshet, M. Bustin, H. Cedar, The role of DNA [25] H. Nakahara, N. Hosono, T. Kitayama, K. Sekiguchi, M. Kubo, A. Takahashi, Y.
methylation in setting up chromatin structure during development, Nat. Genet. Nakamura, Y. Yamano, K. Kai, Automated SNPs typing system based on the
34 (2003) 187192. invader assay, Leg. Med. 11 (2009) S111S114.
[15] A. Bird, DNA methylation patterns and epigenetic memory, Genes Dev. 16 (2002) [26] P. Gill, C.H. Brenner, J.S. Buckleton, A. Carracedo, M. Krawczak, W.R. Mayr, N.
621. Morling, M. Prinz, P.M. Schneider, B.S. Weir, DNA commission of the International
[16] F.B. Dean, S. Hosono, L. Fang, X. Wu, A.F. Faruqi, P. Bray-Ward, Z. Sun, Q. Zong, Y. Society of Forensic Genetics: recommendations on the interpretation of mixtures,
Du, J. Du, M. Driscoll, W. Song, S.F. Kingsmore, M. Egholm, R.S. Lasken, Compre- Forensic Sci. Int. 160 (2006) 90101.
hensive human genome amplication using multiple displacement amplication, [27] D. Shinde, Y. Lai, F. Sun, N. Arnheim, Taq DNA polymerase slippage mutation rates
Proc. Natl. Acad. Sci. U.S.A. 99 (2002) 52615266. measured by PCR and quasi-likelihood analysis: (CA/GT)n and (A/T)n microsa-
[17] E.F. Fritsch, T. Maniatis, J. Sambrook, Molecular Cloning: A Laboratory Manual, tellites, Nucleic Acids Res. 31 (2003) 974980.
2nd ed., Cold Spring Harbor Laboratory Press, New York, 1989. [28] L. Forster, J. Thomson, S. Kutranov, Direct comparison of post-28-cycle PCR
[18] J.M. Butler, R. Schoske, P.M. Valone, J.W. Redman, M.C. Kline, Allele frequencies for purication and modied capillary electrophoresis methods with the 34-cycle
15 autosomal STR loci on U.S. Caucasian, African American, and Hispanic popula- low copy number (LCN) method for analysis of trace forensic DNA samples,
tions, J. Forensic Sci. 48 (2003) 908911. Forensic Sci. Int. Genet. 2 (2008) 318328.

Вам также может понравиться