- Email: [email protected]
Trace evidence characteristics of DNA: A preliminary investigation of the persistence of DNA at crime scenes
Forensic Science International: Genetics 4 (2009) 26–33
Contents lists available at ScienceDirect
Forensic Science International: Genetics journal homepage: www.elsevier.com/locate/fsig
Trace evidence characteristics of DNA: A preliminary investigation of the persistence of DNA at crime scenes Jennifer J. Raymond a,b,*, Roland A.H. van Oorschot c, Peter R. Gunn b, Simon J. Walsh d, Claude Roux a a
Centre for Forensic Science, University of Technology, Broadway, Sydney, NSW 2000, Australia Forensic Services Group, NSW Police Force, 6-20 Clunies Ross Street, Pemulwuy, NSW 2145, Australia c Victoria Police Forensic Services Centre, Forensic Drive, Macleod, VIC 3085, Australia d Forensic and Data Centres, Australian Federal Police, GPO Box 401, Canberra, ACT 2601, Australia b
A R T I C L E I N F O
A B S T R A C T
Article history: Received 1 December 2008 Received in revised form 1 April 2009 Accepted 6 April 2009
The successful recovery of trace or contact DNA is highly variable. It is seemingly dependent on a wide range of factors, from the characteristics of the donor, substrate and environment, to the delay between contact and recovery. There is limited research on the extent of the effect these factors have on trace DNA analysis. This study investigated the persistence of trace DNA on surfaces relevant to the investigation of burglary and robbery offences. The study aimed to limit the number of variables involved to solely determine the effect of time on DNA recovery. Given that it is difficult to control the quantity of DNA deposited during a hand contact, human buffy coat and DNA control solution were chosen as an alternative to give a more accurate measure of quantity. Set volumes of these solutions were deposited onto outdoor surfaces (window frames and vinyl material to mimic burglary and ‘bag snatch’ offences) and sterile glass slides stored in a closed environment in the laboratory, for use as a control. Trace DNA casework data was also scrutinised to assess the effect of time on DNA recovery from real samples. The amount of DNA recovered from buffy coat on the outdoor surfaces declined by approximately half over two weeks, to a negligible amount after six weeks. Profiles could not be obtained after two weeks. The samples stored in the laboratory were more robust, and full profiles were obtained after six weeks, the longest time period tested in these experiments. It is possible that profiles may be obtained from older samples when kept in similarly favourable conditions. The experimental results demonstrate that the ability to recover DNA from human cells on outdoor surfaces decreases significantly over two weeks. Conversely, no clear trends were identified in the casework data, indicating that many other factors are involved affecting the recovery of trace DNA. Nevertheless, to ensure that valuable trace evidence is not lost, it is recommended that crime scenes are processed expeditiously. Crown Copyright ß 2009 Published by Elsevier Ireland Ltd. All rights reserved.
Keywords: Trace DNA Persistence Crime scene investigation Criminalistic Interpretation
1. Introduction The exceptional identifying power of DNA as forensic evidence has been well documented and discussed since its first application to criminal cases. It is now possible to retrieve profiles from the minute amount of skin cells and debris left through brief physical contact with an object or location [1–3]. Equally, the development of national DNA databases has resulted in numerous offender to scene links, many in cases with no previous leads or suspects. The databases have particular
* Corresponding author at: Forensic Services Group, NSW Police Force, 6-20 Clunies Ross Street, Pemulwuy, NSW 2145, Australia. Tel.: +61 2 9688 9201. E-mail addresses: [email protected], [email protected] (J.J. Raymond).
effectiveness in combating ‘volume’ crime, for example burglaries and stolen vehicles, where traditional policing methods are often difficult to apply. The United Kingdom’s (UK) National DNA Database (NDNAD), one of the largest in the world, has found the greatest number of links to be with volume crimes [4,5], as have other jurisdictions including Switzerland and New Zealand [6,7]. Databasing allows forensic evidence to target and link these crimes, and to use forensic intelligence in a proactive manner [8,9]. The positive aspects of these developments are clearly evident, but it is these same developments that have the potential to lower the value of DNA evidence in future. The extreme sensitivity of techniques combined with rapidly expanding database populations results in increasing numbers of ‘links’. Consequently, the chance of an adventitious match, that is, where a person’s DNA is deposited at a crime scene at a time or by means unrelated to the offence, increases correspondingly [10,11]. There has been a
1872-4973/$ – see front matter . Crown Copyright ß 2009 Published by Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.fsigen.2009.04.002
J.J. Raymond et al. / Forensic Science International: Genetics 4 (2009) 26–33
tendency for investigators, on hearing that there is a DNA match in their case, to close down other lines of inquiry and rely on the DNA match without interrogating the circumstances further. False or adventitious matches produced as evidence could result in the devaluing of DNA as forensic evidence. As discussed by Rudin and Inman [12] and Raymond et al. [13], there is a need for forensic DNA scientists to reconsider the basic properties of trace evidence, such as transfer, persistence and abundance. The considerable discriminative power of DNA analysis has led to some neglect of these aspects of the forensic evaluation of evidence. Glass evidence can be used as a comparison, whereby much research and data has been collected to quantify the various types of glass and the trace evidence properties of a glass particle [14– 19]. Not only is the glass from the suspect and control glass compared to determine if they can be discriminated, but the chance that the glass would have actually transferred to and persisted on the suspect, and the likelihood of finding glass on a ‘random man’ are also considered in the analysis. The same could be asked of DNA evidence—what are the chances that the DNA will actually transfer and will persist until detection? Or, to evaluate the possibility of the innocent or adventitious presence of DNA; what is the likelihood of locating DNA on a ‘random’ surface? Combining the particular case circumstances with experimental data on these properties will allow for a more grounded, probabilistic interpretation of a trace DNA profile. The unpredictability of trace DNA analysis is known anecdotally, leaving an opportunity for defence lawyers to claim any number of innocent scenarios leading to the recovery of their client’s profile from the crime scene, as has been the case in R v Joyce [20], R v Hillier [21]. The factors affecting trace evidence recovery are many, existing before, during and after the crime event. With experimental data to estimate the extent to which each factor affects recovery, a holistic assessment of the evidence in terms of the case circumstances can be made. The effect these factors have on trace DNA recovery has been investigated to varying degrees in past research. The chance that a person will leave DNA on a surface after contact, or their ‘shedder’ status, has been discussed [24,25], as have sampling methods [22,23]. However, other factors, such as the length of time that trace DNA is likely to persist in an environment until detection, are yet to be fully examined. The following study focuses on the issue of trace DNA persistence. The experiments were constructed to mimic two specific offence types; residential burglary and street robbery (also known as ‘bag snatches’ or muggings). These two crime types provide scenarios that could be applicable to a variety of other investigations, and the results of this study actually assisted a particular homicide enquiry. An armed robbery of a petrol station became violent and the victim was murdered in the process. Security cameras inside the shop revealed that the offender had pulled shut the entry door during the offence, and opened it on leaving the premises. A DNA profile was recovered from the outside and edge of the door, and matched a suspect. However, the suspect had visited the petrol station two weeks prior to the offence, and the question was asked, could the DNA profile recovered from the crime scene have come from this innocent visit? Intuitively this assertion seems unlikely, however at the time there was limited research to give weight to a theory in court. Circumstances such as this provide a direct justification for the relevance of this criminalistic research. Given the stochastic nature of trace DNA it is often difficult to separate the effect of one variable from the many that affect its recovery. Therefore, this study aimed not to exactly replicate crime scene trace DNA samples, but instead to determine the effect of crime scene environments on pristine DNA samples of known concentration. Analysis of data from casework trace DNA samples is provided to complement the experimental work.
27
2. Methods This study consists of two parts; experimental work and the analysis of casework data. 2.1. Experimental 2.1.1. Subject Using hand contact as the DNA deposition method in experiments would prevent a robust quantitative analysis, due to the known variation between shedder types, and other factors such as time since the hands were washed [24,25]. Therefore, buffy coat (the white blood cell and platelet layer of whole blood) with known cell amounts was selected as the DNA source for these experiments. The buffy coat cells used were purchased commercially, purified, and quantitated by direct haemocytometer count and DNA quantitation methods. Cell stocks were stored in 500 mL volumes at 80 8C until use. 3 mL aliquots of buffy coat were spread onto the surfaces with pipette tips over an area approximately 5 mm2. For further accuracy in assessing the quantity of DNA recovered, 3 mL aliquots (387 ng) of control DNA solution (9947A, Promega Corporation, Madison, WI) was also used in one of the experiments. However, it was envisaged that this ‘naked’ DNA would degrade
Fig. 1. (a) Recovery of DNA from swabbed surfaces at various time intervals up to six weeks, after application of a standard amount of human buffy coat cells, presented as a percentage of the amount recovered at t = 0. (b). The recovery of DNA from swabbed surfaces at time intervals up to six weeks after application of DNA control solution, presented as a percentage of the amount recovered at t = 0.
28
J.J. Raymond et al. / Forensic Science International: Genetics 4 (2009) 26–33
more rapidly than the buffy coat DNA protected inside cells, and, therefore, may not be a true representation of the DNA found at crime scenes. 2.1.2. Substrates To model burglary offences, outdoor window frame surfaces in a residential unit building were chosen, being common entry points [26,27]. These surfaces were gloss-painted wooden window frames on a patio, under cover of a one-metre brick balcony above. This set of samples will be referred to as the ‘house’ samples hereinafter. For street robbery offences, vinyl, a common handbag material [28] was purchased and cut into 3 3 cm swatches. This set of samples will be referred to as the ‘bag’ samples. The experiment took place in November and December, which in Sydney, Australia, provides average temperature and relative humidity of 24.1 8C, 63% (day) and 18 8C, 71% (night) [29]. The surfaces were in a partly shady location, and did not receive direct sunlight. As a control set, glass microscope slides were selected as a substrate and stored in the dark in the laboratory in a sterile container, at ambient temperature. This final set will be referred to as the ‘laboratory’ samples. All surfaces were pre-cleaned with 10% bleach and 70% ethanol solutions and surface blanks taken prior to the experiment. 2.1.3. Sampling One set of placements was swabbed within 15 min of placement, using the double swab method [30]. The following sets were swabbed one day, three days, one week, two weeks, four
weeks and six weeks after placement. Three replicates were conducted in each set. Additional to the surface blanks taken after cleaning the surfaces but prior to the placement of the samples, surface blanks were also collected from the ‘house’, ‘bag’ and ‘laboratory’ samples at the two- and six-week time periods to determine if extraneous DNA was present. 2.1.4. DNA and data analysis The samples were extracted using a 20% chelex/Proteinase K method validated by the Division of Analytical Laboratories (New South Wales, Australia), to a final volume of 260 mL. The extracts were quantitated using QuantifilerTM real-time PCR, with an Applied Biosystems 7500 Real-Time PCR System. The samples with the highest quantity from each set of buffy coat samples at time = zero, one, two, four and six weeks were amplified in 25 mL AmpFlSTR1 Profiler PlusTM reactions using a GeneAmp System 9700 thermocycler. The 28 cycle reactions contained up to 10 mL of extract depending on their quantitation, to give up to 1 ng of DNA in the reaction. 1.5 mL of the amplified product with 25 mL sample mix was run on an ABI Prism 310 Genetic Analyser, at 5 s injection. Regression and single-factor ANOVA analysis were performed using Microsoft ExcelTM 2004. 2.2. Casework data To assist the interpretation of the research data, casework data from the New South Wales Police Forensic Services Group were
Fig. 2. Profiler PlusTM profiles of selected buffy coat ‘laboratory’ samples over six weeks.
J.J. Raymond et al. / Forensic Science International: Genetics 4 (2009) 26–33
assessed. 50 trace DNA samples from 46 volume crime cases were compiled with respect to the type of sample, the time lapse between the offence, sample collection and laboratory submission, and the DNA quantity and profile recovered. Using the double swab method described previously, various Scenes of Crime Officers from around New South Wales collected the samples. The swabs were then stored at room temperature until being sent to the Genetic Technologies Limited laboratory, Fitzroy, Victoria where DNA analysis was performed in March and April 2008. At the laboratory, the swabs were stored at room temperature until their extraction using a Qiagen spin column method on a BioRobot 8000, and the extracts stored at 20 8C until amplification. The extracts were not routinely concentrated, but if warranted by a review of results were concentrated to 20 mL using the Qiagen Forensic MicroKit. The extracts were quantified using QuantifilerTM real-time PCR, and 10 mL of extract (diluted if the concentration >0.125 ng/mL) amplified in 25 mL 28 cycle AmpFlSTR1 Profiler PlusTM reactions. 2 mL of the amplified product with 9 mL sample mix was run on an ABI Prism 3130xl Genetic Analyser, at 5 s injection. 3. Results Fig. 1a and b show the amount of buffy coat DNA and control DNA solution recovered from the surfaces over the time period. The results are displayed as a percentage of the amount recovered at time = 0, with the average of the three replicates at each time period.
29
Regarding the buffy coat samples (Fig. 1a) it is apparent that for all but the laboratory samples the amount decreased significantly after two weeks, to a negligible amount recovered after six weeks. The decline in DNA recovery across the six-week period was found to be statistically significant using linear regression analysis for the house and bag samples (p = 2.5 10 3 and 7.0 10 3, respectively), but not for the laboratory samples (p = 0.25). The average amount of DNA recovered at t = 0 was 38.9, 20.4, and 96.3 ng for the laboratory, house and bag samples, respectively. The DNA control solution samples did not decline as rapidly as expected (Fig. 1b). The laboratory samples did not decline below 100% of the starting time period, and the bag samples declined by 50% over the six-week period. Only the house samples were found to decrease significantly in quantity (p = 4.6 10 2). Figs. 2–4 show the Profiler PlusTM amplifications of selected samples from each set of buffy coat samples at time = zero, one, two, four and six weeks. All profiles are displayed at the same scale. No extraneous alleles were located in any of the profiles, indicating that contamination had not occurred. Quantitation of the surface blanks showed that no DNA was present on the surfaces prior to the experiments, and that no additional DNA had been deposited on the surface after the two- and six-week time periods. Table 1 displays the information collated regarding the 50 trace DNA casework samples, arranged sequentially according to the time difference between the offence and the sample collection. The time between the collection of the sample from the crime scene and the submission of the sample to the laboratory is also shown in
Fig. 3. Profiler PlusTM profiles of selected buffy coat ‘house’ samples over six weeks.
J.J. Raymond et al. / Forensic Science International: Genetics 4 (2009) 26–33
30
Table 1 Results from trace DNA casework samples, sorted by the time difference between the offence and sample collection. Time between offence and collection
Time between collection and lab receipt (days)
Sample
Offence
DNA quantity (ng/mL)
Estimated total DNA in extract (ng)
DNA profile resulta
1 h 40 min 1 h 40 min 1 h 40 min 1 h 45 min 2 h 15 min 2 h 19 min 2 h 23 min 2 h 23 min 2 h 40 min 2 h 40 min 2 h 40 min 2 h 52 min 3 h 10 min 3 h 30 min 3 h 30 min 3 h 30 min 3 h 50 min 4 h 5 min 5 h 15 min 5 h 15 min 5 h 40 min 5 h 50 min 8 h 45 min 9 h 25 min 10 h 5 min 10 h 42 min 11 h 35 min 11 h 35 min 14 h 25 min 14 h 58 min 16 h 17 h 31 min 19 h 26 min 19 h 31 min 19 h 55 min 20 h 55 min 21 h 10 min 1d 1d 2d 3d 3d 3d 5d 9d 11 d 22 d 55 d 62 d 62 d
225 33 15 39 11 37 383 383 9 22 22 7 13 230 230 230 169 4 29 29 29 19 7 31 173 24 121 144 24 5 36 66 14 35 12 131 226 16 219 13 14 6 241 17 304 142 180 32 39 8
FPsb at point of entry Earprint on door FPs on counter Register note clips Handmark on counter Service counter FPs at point of entry Security wire above counter Cash register, FPs Arm mark on counter Arm mark on counter Display mat on counter Perspex partition Shopping basket Freezer door Rear loading dock door FPs on window sill Register keyring Area around reg. plates Area around reg. plates Computer harddrive Plant pots Screwdriver handle Earprint on door FPs at point of entry Safe handle Glovebox handle FPs on security camera FPs at point of entry FPs at point of entry FPs on bumper FPs at point of entry Earprint on door FPs at point of entry Earprint on door Auto gearshift Earprint on door Wallet FPs at point of entry FPs at point of entry Partial FP on camera Gas bottle handle FPs at point of entry Passenger doorhandle Handbag Cheque book Glucodin powder box Bag Ziplock bag FPs on laptop
Burglary Burglary Armed robbery Armed robbery Armed robbery Armed robbery Armed robbery Armed robbery Armed robbery Armed robbery Armed robbery Armed robbery Armed robbery Armed robbery Armed robbery Armed robbery Burglary Theft Theft Theft Burglary Burglary Burglary Burglary Burglary Burglary Robbery Arson Burglary Burglary Theft Burglary Burglary Burglary Burglary Vehicle offence Burglary Burglary/SMV Burglary Burglary Burglary Drugs Theft Assault Robbery Theft Drugs/firearms Drugs Drugs Burglary
0 0.003 0.045 0.002 0.017 0.025 0.025 0 0 0.002 0.008 0.056 0.046 0.178 0.007 0.066 0.033 0.004 0 0.001 0.002 0 0.18 0 0 0.002 0.014 0.003 0.02 0.009 0.001 0 0 0.028 0.002 0.001 0 0.074 0 0 0 0.032 0.001 0.022 0.002 0.004 0.004 0.105 0.001 0.032
0 0.12 1.8 0.08 0.68 1.0 1.0 0 0 0.08 0.32 2.24 1.84 7.12 0.28 2.64 1.32 0.16 0 0.04 0.08 0 7.2 0 0 0.08 0.56 0.12 0.8 0.36 0.04 0 0 1.12 0.08 0.04 0 2.96 0 0 0 1.28 0.04 0.88 0.08 0.16 0.16 4.2 0.04 1.28
Neg Partial <12 Mixture Partial <12 Amel only Amel only Partial <12 Neg Neg Mixture Full profile Mixture Mixture Mixture Partial <12 Mixture Mixture Amel only Neg Neg Partial <12 Neg Partial 12 Neg Neg Neg Mixture Neg Partial 12 Neg Neg Neg Neg Partial <12 Neg Neg Neg Amel only Neg Neg Neg Mixture Neg Full profile Neg Partial <12 Neg Full profile Neg Mixture
a Neg: no profile; Amel only: only the Amelogenin locus present in the profile; partial <12: less than 12 alleles in the profile; partial 12: 12 or more alleles in the profile but less than a full profile; and mixture: a mixture of more than 1 person’s DNA in the profile. b FPs: fingerprints.
days. This time delay ranged from four days to over one year (383 days). The DNA quantity is given as the QuantifilerTM output in nanograms per microlitre, and an estimation of the total DNA quantity in the extract. The DNA profile is the result of a Profiler PlusTM amplification of up to 50% of the extract. The profile results are given in the categories designated by the laboratory; either a full profile, a partial profile of 12 or more alleles, a partial profile of less than 12 alleles, a mixture, the Amelogenin locus only, or a negative result. Fig. 5 displays the percentage of each profile type produced from the samples. The laboratory determined that the mixtures recovered in these cases were not suitable for input onto the national database. It should be noted that these mixtures may still have probative value in terms of inclusions or eliminations, or other non-database use. There was no evidence for a linear relationship between the time delay and quantity of DNA recovered in the DNA casework
data (R2 < 0.03). Likewise, no significant reduction in the quality of profile recovered was noted as the time delay between the offence and sample collection increased. Despite lengthy time delays between the sample collection and laboratory submission, there were again no significant decreases in the DNA quantity and profile quality (p = 0.86 and 0.28). Fig. 6 shows the spread in DNA concentration across five time periods between the offence and the DNA sample collection. Whilst the samples of highest concentration tended towards the lower time frames, there is a large range at each time period. 4. Discussion The experimental data support the hypothesis that the chance of recovering DNA from an outdoor surface decreases significantly over time. If a DNA profile is recovered from a similar location, it is
J.J. Raymond et al. / Forensic Science International: Genetics 4 (2009) 26–33
31
Fig. 4. Profiler PlusTM profiles of selected buffy coat ‘bag’ samples over six weeks.
likely to be from a recent contact. The quantity of DNA deposited in this experiment is up to 300 000 times greater than the amount left during a hand contact, and therefore it is expected that DNA left through hand contact would be undetectable in a much shorter time period. For example, if 150 pg of DNA were left after a hand contact on a bag (it has been estimated that five nucleated cells are present per fingerprint [31]), after two weeks there would be 27% available for recovery. Assuming that there is 100% recovery of the cells, this would leave only 40.5 pg. Whilst this may be feasible in
Fig. 5. Types of profiles recovered from casework trace DNA samples (n = 50).
research scenarios, recovery of such a small amount of DNA at a crime scene would be difficult. As expected, the laboratory-stored samples proved to be more robust than the samples left outdoors. Whilst the amount of DNA declined, full profiles were recovered over the entire six-week period. The results of the laboratory samples indicate that if crime scene samples are left in an environment that is cool, dark and in a low traffic area, profiles may be recovered for a longer period of time.
Fig. 6. Concentration of DNA recovered from casework trace DNA samples, grouped in five time categories of the delay prior to the sample collection.
32
J.J. Raymond et al. / Forensic Science International: Genetics 4 (2009) 26–33
Three points of interest in the experimental data are: the differences in the amount of DNA recovered at t = 0 between the surfaces; the spike in recovery of DNA from the ‘laboratory’ samples at one and two weeks; and the insignificant decline in the ‘laboratory’ and ‘bag’ DNA control solution samples (Fig. 1a and b). These anomalies may be explained in part by an uneven distribution of white blood cells in the buffy coat preparations. Two separate aliquots were each quantitated twice to ascertain the quantity of DNA present, and produced differences of up to 40 ng within and between separate 20 mL aliquots (data not shown). This provides an explanation why the ‘bag’ samples returned significantly more (p = 0.03) DNA than the ‘laboratory’ samples in the first week, and for the spike in recovery of the ‘laboratory’ samples at one and two weeks. In addition, at the one- and two-week time periods the deposit became clearly visible as a dried white spot and therefore was simple to target, whereas prior to this it had not been as easily discernable. The ‘house’ and ‘bag’ samples did not show this spike, and therefore the effect of the outdoor environment on these samples may have negated any targeting advantage provided by the deposits becoming visible. It is supposed that the variation between the ‘house’ and the ‘bag’ samples (Figs. 2–4) may be due to the deposition method. As the house surface was vertical it was more difficult to deposit the fluid sample. Conversely, the ‘bag’ samples were stored horizontally preventing the deposit from spreading, allowing an easier recovery leading to higher quantities and stronger profiles. The casework data in Table 1 indicate the difficulties encountered in forensic trace DNA analysis. The average amount recovered from the trace samples was low at 21 pg/mL or 0.84 ng in the total extract, despite the majority of samples (74%) being recovered within 24 h of the offence. Full profiles were only recovered in three cases (6%), with no profile recovered in approximately half of the cases (Fig. 5). It should be noted that the extraction and amplification methods differ between the experimental and casework samples, which may be a cause of variation in the amount of DNA recovered. Linear regression analysis shows that the time delay prior to sample collection did not significantly affect the quantity of DNA recovered. This may be due to the limited number of samples in this study with a lengthy time delay, and the effect may prove more influencing if further cases could be included. The time delay between the sample collection and the laboratory submission also proved insignificant, despite a third of the samples being held for over three months prior to submission. This concurs with a 2006 study [32] that found no decrease in DNA recovery from swabs stored for different time periods. However, the large number of variables affecting the recovery of DNA from casework samples (such as different donors, differing hand contact, surface types and sampling techniques) greatly inhibits the accurate determination of the effect of a single variable, such as time. These variables are difficult, if not impossible, to ascertain or control in casework which highlights the importance of criminalistic consideration of these factors through research, in order to assist interpretation of casework results. These results emphasise the variability of forensic trace DNA analysis. Whilst all efforts were made to reduce the number of variables in the experimental work, differences were still noted in the outcome. This inherent nature of forensic trace DNA analysis can cause difficulties in the attempt to provide definitive answers to legal questions arising in casework. However, with ongoing research and casework data, guidelines can be offered to assess the likelihood of a suggested scenario, with the potential for Bayesian analysis. This study was used to assist the court’s interpretation of DNA evidence in a murder trial, discussed in Section 1. The defence
asserted that their client’s DNA profile might have been recovered from the crime scene as a result of an innocent visit two weeks prior to the offence. A statement was prepared stating that the ability to recover analysable DNA from outdoor surfaces decreases rapidly over several weeks. With this report countering their argument, and other circumstantial evidence, the accused plead guilty to the murder. 5. Conclusions Human cells deposited on a surface such as a window frame or bag will deteriorate significantly (p < 0.05) over the course of six weeks, to such an extent that the chance of recovering a DNA profile is substantially reduced over this time. However, if the cells are deposited and stored undisturbed in a cool and dark location, as in a laboratory, profiles may still be recovered after six weeks. It should also be noted that these experiments were conducted with a large amount of human cells. Contact between a person’s hand and an object tends to leave far less DNA on the object, as shown by casework data, which would exacerbate the difficulty of recovering a DNA profile over time. It is therefore imperative that trace evidence is collected expeditiously to give the best chance of a successful outcome. This research provides an initial look at the extent of the persistence of DNA in the context of volume crime offences, and the results provided a more objective approach to the interpretation of DNA evidence in a criminal trial. Whilst this is a preliminary and limited study, it demonstrates that research into the trace evidence characteristics of DNA offers a criminalistic approach to biological evidence, of particular importance when the source of the DNA is not in question. Further research of this kind is required to deepen our understanding of persistence of DNA at crime scenes. Acknowledgements The authors would like to thank Ms Wendy Tufts and Dr Tony Raymond, NSW Police Forensic Services Group and Ms Alex Lucca, Genetic Technologies Limited, for providing access and assistance with the trace DNA casework data. References [1] R.A. Van Oorschot, M. Jones, DNA fingerprints from fingerprints, Nature 387 (6635) (1997) 767. [2] M.K. Balogh, J. Burger, K. Bender, P.M. Schneider, K.W. Alt, Fingerprints from fingerprints, International Congress Series, Progress in Forensic Genetics 9 1239 (2003) 953–957. [3] R.A. Wickenheiser, Trace DNA: a review, discussion of theory, and application of the transfer of trace quantities of DNA through skin contact, Journal of Forensic Science 47 (3) (2002) 442–450. [4] B. Gunn, An intelligence-led approach to policing in England and Wales and the impact of developments in forensic science, The Australian Journal of Forensic Science 35 (1) (2003) 149–160. [5] D. Coleman, S. Hyde, I. Gordon, T. Wilson, D. Werrett, S. Bain, et al. The National DNA Database Annual Report 03/04. 2004 [cited 24.2.08]; available from: www.forensic.gov.uk/forensic_t/inside/about/docs/NDNAD_AR_3_4.pdf. [6] S.A. Harbison, J.F. Hamilton, S.J. Walsh, The New Zealand DNA databank: its development and significance as a crime solving tool, Science and Justice 41 (1) (2001) 33–38. [7] P. Voegeli, C. Haas, A. Kratzer, W. Bar, Evaluation of the 4-year test period of the Swiss DNA database, International Congress Series, Progress in Forensic Genetics 11 1288 (2006) 731–733. [8] S.J. Walsh, J.S. Buckleton, DNA intelligence databases, in: J.S. Buckleton, C.M. Triggs, S.J. Walsh (Eds.), Forensic DNA Evidence Interpretation, CRC Press, Boca Raton, FL, 2005, pp. 439–469. [9] P.M. Schneider, P.D. Martin, Criminal DNA databases: the European situation, Forensic Science International 119 (2) (2001) 232–238. [10] G.N. Rutty, An investigation into the transfer and survivability of human DNA following simulated manual strangulation with consideration of the problem of third party contamination, International Journal of Legal Medicine 116 (3) (2002) 170–173.
J.J. Raymond et al. / Forensic Science International: Genetics 4 (2009) 26–33 [11] P. Gill, A. Kirkham, Development of a simulation model to assess the impact of contamination in casework using STRs, Journal of Forensic Science 49 (3) (2004) 7. [12] N. Rudin, K. Inman, Biological evidence as trace evidence: the forensic science of DNA typing, CAC News 4th Qtr. (2002) 19–21. [13] J.J. Raymond, S.J. Walsh, R.A. Van Oorschot, P. Gunn, C. Roux, Trace DNA: an underutilised resource or Pandora’s box? A review of the use of trace DNA analysis in the investigation of volume crime, Journal of Forensic Identification 54 (6) (2004) 668–686. [14] C.I. Petterd, J. Hamshere, S. Stewart, K. Brinch, T. Masi, C. Roux, Glass particles in the clothing of members of the public in south-eastern Australia—a survey, Forensic Science International 103 (1999) 193–198. [15] C. Roux, R. Kirk, S. Benson, T. Van Haren, C.I. Petterd, Glass particles in footwear of members of the public in south-eastern Australia—a survey, Forensic Science International 116 (2001) 149–156. [16] J.M. Curran, C.M. Triggs, J.S. Buckleton, K. Walsh, T. Hicks, Assessing transfer probabilities in a Bayesian interpretation of forensic glass evidence, Science and Justice 38 (1) (1998) 15–21. [17] S.A. Coulson, J.S. Buckleton, A.B. Gummer, C.M. Triggs, Glass on clothing and shoes of members of the general population and people suspected of breaking crimes, Science and Justice 41 (1) (2001) 39–48. [18] F. Taroni, C.G. Aitken, Probabilistic reasoning in the law Part 2: assessment of probabilities and explanation of the value of trace evidence other than DNA, Science and Justice 38 (3) (1998) 179–188. [19] J.M. Curran, T.N. Hicks, J.S. Buckleton, Forensic Interpretation of Glass Evidence, CRC Press, London, 2000. [20] R v Joyce, NTSC: Supreme Court of the Northern Territory, 2002, unreported. [21] Bail application by Hillier, ACTSC, 2003. [22] R.A. Van Oorschot, D.G. Phelan, S. Furlong, G.M. Scarfo, N.L. Holding, M.J. Cummins, Are you collecting all the available DNA from touched objects? International Congress Series, Progress in Forensic Genetics 9 1239 (2003) 803–807.
33
[23] B.C.M. Pang, B.K.K. Cheung, Double swab technique for collecting touched evidence, Legal Medicine 9 (4) (2007) 181–184. [24] A. Lowe, C. Murray, J. Whitaker, G. Tully, P. Gill, The propensity of individuals to deposit DNA and secondary transfer of low level DNA from individuals to inert surfaces, Forensic Science International 129 (1) (2002) 25–34. [25] M. Phipps, S.F. Petricevic, The tendency of individuals to transfer DNA to handled items, Forensic Science International 168 (2–3) (2006) 162–168. [26] J.J. Raymond, S.J. Walsh, R.A. Van Oorschot, P.R. Gunn, L. Evans, C. Roux, Assessing trace DNA evidence from a residential burglary: abundance, transfer and persistence, in: 22nd Congress of the International Society for Forensic Genetics, 2007, Copenhagen, Denmark, 2007. [27] Crime victims and the prevention of residential burglary. 2004 [cited 2008 17/3]; available from: www.jcs.act.gov.au/eLibrary/OtherReports/Burglary_Victims_Report.pdf. [28] F. Kasses, Pers. comm., Sydney Luggage Centre, 2008. [29] Australian Government Bureau of Meteorology: Sydney, NSW, November and December 2007 Daily Weather Observations. 2007 [cited 2008 24.2]; available from: www.bom.gov.au. [30] D. Sweet, M. Lorente, J. Lorente, A. Valenzuela, E. Villaneuva, An improved method to recover saliva from human skin: the double swab technique, Journal of Forensic Science 42 (2) (1997) 320–322. [31] F. Alessandrini, M. Cecati, M. Pesaresi, C. Turchi, F. Carle, A. Tagliabracci, Fingerprints as evidence for a genetic profile: morphological study on fingerprints and analysis of exogenous and individual factors affecting DNA typing, Journal of Forensic Science 48 (3) (2003) 586–592. [32] P. Leemans, A. Vandeput, N. Vanderheyden, J.-J. Cassiman, R. Decorte, Evaluation of methodology for the isolation and analysis of LCN-DNA before and after dactyloscopic enhancement of fingerprints, International Congress Series 1288 (2006) 583–585.
Contents lists available at ScienceDirect
Forensic Science International: Genetics journal homepage: www.elsevier.com/locate/fsig
Trace evidence characteristics of DNA: A preliminary investigation of the persistence of DNA at crime scenes Jennifer J. Raymond a,b,*, Roland A.H. van Oorschot c, Peter R. Gunn b, Simon J. Walsh d, Claude Roux a a
Centre for Forensic Science, University of Technology, Broadway, Sydney, NSW 2000, Australia Forensic Services Group, NSW Police Force, 6-20 Clunies Ross Street, Pemulwuy, NSW 2145, Australia c Victoria Police Forensic Services Centre, Forensic Drive, Macleod, VIC 3085, Australia d Forensic and Data Centres, Australian Federal Police, GPO Box 401, Canberra, ACT 2601, Australia b
A R T I C L E I N F O
A B S T R A C T
Article history: Received 1 December 2008 Received in revised form 1 April 2009 Accepted 6 April 2009
The successful recovery of trace or contact DNA is highly variable. It is seemingly dependent on a wide range of factors, from the characteristics of the donor, substrate and environment, to the delay between contact and recovery. There is limited research on the extent of the effect these factors have on trace DNA analysis. This study investigated the persistence of trace DNA on surfaces relevant to the investigation of burglary and robbery offences. The study aimed to limit the number of variables involved to solely determine the effect of time on DNA recovery. Given that it is difficult to control the quantity of DNA deposited during a hand contact, human buffy coat and DNA control solution were chosen as an alternative to give a more accurate measure of quantity. Set volumes of these solutions were deposited onto outdoor surfaces (window frames and vinyl material to mimic burglary and ‘bag snatch’ offences) and sterile glass slides stored in a closed environment in the laboratory, for use as a control. Trace DNA casework data was also scrutinised to assess the effect of time on DNA recovery from real samples. The amount of DNA recovered from buffy coat on the outdoor surfaces declined by approximately half over two weeks, to a negligible amount after six weeks. Profiles could not be obtained after two weeks. The samples stored in the laboratory were more robust, and full profiles were obtained after six weeks, the longest time period tested in these experiments. It is possible that profiles may be obtained from older samples when kept in similarly favourable conditions. The experimental results demonstrate that the ability to recover DNA from human cells on outdoor surfaces decreases significantly over two weeks. Conversely, no clear trends were identified in the casework data, indicating that many other factors are involved affecting the recovery of trace DNA. Nevertheless, to ensure that valuable trace evidence is not lost, it is recommended that crime scenes are processed expeditiously. Crown Copyright ß 2009 Published by Elsevier Ireland Ltd. All rights reserved.
Keywords: Trace DNA Persistence Crime scene investigation Criminalistic Interpretation
1. Introduction The exceptional identifying power of DNA as forensic evidence has been well documented and discussed since its first application to criminal cases. It is now possible to retrieve profiles from the minute amount of skin cells and debris left through brief physical contact with an object or location [1–3]. Equally, the development of national DNA databases has resulted in numerous offender to scene links, many in cases with no previous leads or suspects. The databases have particular
* Corresponding author at: Forensic Services Group, NSW Police Force, 6-20 Clunies Ross Street, Pemulwuy, NSW 2145, Australia. Tel.: +61 2 9688 9201. E-mail addresses: [email protected], [email protected] (J.J. Raymond).
effectiveness in combating ‘volume’ crime, for example burglaries and stolen vehicles, where traditional policing methods are often difficult to apply. The United Kingdom’s (UK) National DNA Database (NDNAD), one of the largest in the world, has found the greatest number of links to be with volume crimes [4,5], as have other jurisdictions including Switzerland and New Zealand [6,7]. Databasing allows forensic evidence to target and link these crimes, and to use forensic intelligence in a proactive manner [8,9]. The positive aspects of these developments are clearly evident, but it is these same developments that have the potential to lower the value of DNA evidence in future. The extreme sensitivity of techniques combined with rapidly expanding database populations results in increasing numbers of ‘links’. Consequently, the chance of an adventitious match, that is, where a person’s DNA is deposited at a crime scene at a time or by means unrelated to the offence, increases correspondingly [10,11]. There has been a
1872-4973/$ – see front matter . Crown Copyright ß 2009 Published by Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.fsigen.2009.04.002
J.J. Raymond et al. / Forensic Science International: Genetics 4 (2009) 26–33
tendency for investigators, on hearing that there is a DNA match in their case, to close down other lines of inquiry and rely on the DNA match without interrogating the circumstances further. False or adventitious matches produced as evidence could result in the devaluing of DNA as forensic evidence. As discussed by Rudin and Inman [12] and Raymond et al. [13], there is a need for forensic DNA scientists to reconsider the basic properties of trace evidence, such as transfer, persistence and abundance. The considerable discriminative power of DNA analysis has led to some neglect of these aspects of the forensic evaluation of evidence. Glass evidence can be used as a comparison, whereby much research and data has been collected to quantify the various types of glass and the trace evidence properties of a glass particle [14– 19]. Not only is the glass from the suspect and control glass compared to determine if they can be discriminated, but the chance that the glass would have actually transferred to and persisted on the suspect, and the likelihood of finding glass on a ‘random man’ are also considered in the analysis. The same could be asked of DNA evidence—what are the chances that the DNA will actually transfer and will persist until detection? Or, to evaluate the possibility of the innocent or adventitious presence of DNA; what is the likelihood of locating DNA on a ‘random’ surface? Combining the particular case circumstances with experimental data on these properties will allow for a more grounded, probabilistic interpretation of a trace DNA profile. The unpredictability of trace DNA analysis is known anecdotally, leaving an opportunity for defence lawyers to claim any number of innocent scenarios leading to the recovery of their client’s profile from the crime scene, as has been the case in R v Joyce [20], R v Hillier [21]. The factors affecting trace evidence recovery are many, existing before, during and after the crime event. With experimental data to estimate the extent to which each factor affects recovery, a holistic assessment of the evidence in terms of the case circumstances can be made. The effect these factors have on trace DNA recovery has been investigated to varying degrees in past research. The chance that a person will leave DNA on a surface after contact, or their ‘shedder’ status, has been discussed [24,25], as have sampling methods [22,23]. However, other factors, such as the length of time that trace DNA is likely to persist in an environment until detection, are yet to be fully examined. The following study focuses on the issue of trace DNA persistence. The experiments were constructed to mimic two specific offence types; residential burglary and street robbery (also known as ‘bag snatches’ or muggings). These two crime types provide scenarios that could be applicable to a variety of other investigations, and the results of this study actually assisted a particular homicide enquiry. An armed robbery of a petrol station became violent and the victim was murdered in the process. Security cameras inside the shop revealed that the offender had pulled shut the entry door during the offence, and opened it on leaving the premises. A DNA profile was recovered from the outside and edge of the door, and matched a suspect. However, the suspect had visited the petrol station two weeks prior to the offence, and the question was asked, could the DNA profile recovered from the crime scene have come from this innocent visit? Intuitively this assertion seems unlikely, however at the time there was limited research to give weight to a theory in court. Circumstances such as this provide a direct justification for the relevance of this criminalistic research. Given the stochastic nature of trace DNA it is often difficult to separate the effect of one variable from the many that affect its recovery. Therefore, this study aimed not to exactly replicate crime scene trace DNA samples, but instead to determine the effect of crime scene environments on pristine DNA samples of known concentration. Analysis of data from casework trace DNA samples is provided to complement the experimental work.
27
2. Methods This study consists of two parts; experimental work and the analysis of casework data. 2.1. Experimental 2.1.1. Subject Using hand contact as the DNA deposition method in experiments would prevent a robust quantitative analysis, due to the known variation between shedder types, and other factors such as time since the hands were washed [24,25]. Therefore, buffy coat (the white blood cell and platelet layer of whole blood) with known cell amounts was selected as the DNA source for these experiments. The buffy coat cells used were purchased commercially, purified, and quantitated by direct haemocytometer count and DNA quantitation methods. Cell stocks were stored in 500 mL volumes at 80 8C until use. 3 mL aliquots of buffy coat were spread onto the surfaces with pipette tips over an area approximately 5 mm2. For further accuracy in assessing the quantity of DNA recovered, 3 mL aliquots (387 ng) of control DNA solution (9947A, Promega Corporation, Madison, WI) was also used in one of the experiments. However, it was envisaged that this ‘naked’ DNA would degrade
Fig. 1. (a) Recovery of DNA from swabbed surfaces at various time intervals up to six weeks, after application of a standard amount of human buffy coat cells, presented as a percentage of the amount recovered at t = 0. (b). The recovery of DNA from swabbed surfaces at time intervals up to six weeks after application of DNA control solution, presented as a percentage of the amount recovered at t = 0.
28
J.J. Raymond et al. / Forensic Science International: Genetics 4 (2009) 26–33
more rapidly than the buffy coat DNA protected inside cells, and, therefore, may not be a true representation of the DNA found at crime scenes. 2.1.2. Substrates To model burglary offences, outdoor window frame surfaces in a residential unit building were chosen, being common entry points [26,27]. These surfaces were gloss-painted wooden window frames on a patio, under cover of a one-metre brick balcony above. This set of samples will be referred to as the ‘house’ samples hereinafter. For street robbery offences, vinyl, a common handbag material [28] was purchased and cut into 3 3 cm swatches. This set of samples will be referred to as the ‘bag’ samples. The experiment took place in November and December, which in Sydney, Australia, provides average temperature and relative humidity of 24.1 8C, 63% (day) and 18 8C, 71% (night) [29]. The surfaces were in a partly shady location, and did not receive direct sunlight. As a control set, glass microscope slides were selected as a substrate and stored in the dark in the laboratory in a sterile container, at ambient temperature. This final set will be referred to as the ‘laboratory’ samples. All surfaces were pre-cleaned with 10% bleach and 70% ethanol solutions and surface blanks taken prior to the experiment. 2.1.3. Sampling One set of placements was swabbed within 15 min of placement, using the double swab method [30]. The following sets were swabbed one day, three days, one week, two weeks, four
weeks and six weeks after placement. Three replicates were conducted in each set. Additional to the surface blanks taken after cleaning the surfaces but prior to the placement of the samples, surface blanks were also collected from the ‘house’, ‘bag’ and ‘laboratory’ samples at the two- and six-week time periods to determine if extraneous DNA was present. 2.1.4. DNA and data analysis The samples were extracted using a 20% chelex/Proteinase K method validated by the Division of Analytical Laboratories (New South Wales, Australia), to a final volume of 260 mL. The extracts were quantitated using QuantifilerTM real-time PCR, with an Applied Biosystems 7500 Real-Time PCR System. The samples with the highest quantity from each set of buffy coat samples at time = zero, one, two, four and six weeks were amplified in 25 mL AmpFlSTR1 Profiler PlusTM reactions using a GeneAmp System 9700 thermocycler. The 28 cycle reactions contained up to 10 mL of extract depending on their quantitation, to give up to 1 ng of DNA in the reaction. 1.5 mL of the amplified product with 25 mL sample mix was run on an ABI Prism 310 Genetic Analyser, at 5 s injection. Regression and single-factor ANOVA analysis were performed using Microsoft ExcelTM 2004. 2.2. Casework data To assist the interpretation of the research data, casework data from the New South Wales Police Forensic Services Group were
Fig. 2. Profiler PlusTM profiles of selected buffy coat ‘laboratory’ samples over six weeks.
J.J. Raymond et al. / Forensic Science International: Genetics 4 (2009) 26–33
assessed. 50 trace DNA samples from 46 volume crime cases were compiled with respect to the type of sample, the time lapse between the offence, sample collection and laboratory submission, and the DNA quantity and profile recovered. Using the double swab method described previously, various Scenes of Crime Officers from around New South Wales collected the samples. The swabs were then stored at room temperature until being sent to the Genetic Technologies Limited laboratory, Fitzroy, Victoria where DNA analysis was performed in March and April 2008. At the laboratory, the swabs were stored at room temperature until their extraction using a Qiagen spin column method on a BioRobot 8000, and the extracts stored at 20 8C until amplification. The extracts were not routinely concentrated, but if warranted by a review of results were concentrated to 20 mL using the Qiagen Forensic MicroKit. The extracts were quantified using QuantifilerTM real-time PCR, and 10 mL of extract (diluted if the concentration >0.125 ng/mL) amplified in 25 mL 28 cycle AmpFlSTR1 Profiler PlusTM reactions. 2 mL of the amplified product with 9 mL sample mix was run on an ABI Prism 3130xl Genetic Analyser, at 5 s injection. 3. Results Fig. 1a and b show the amount of buffy coat DNA and control DNA solution recovered from the surfaces over the time period. The results are displayed as a percentage of the amount recovered at time = 0, with the average of the three replicates at each time period.
29
Regarding the buffy coat samples (Fig. 1a) it is apparent that for all but the laboratory samples the amount decreased significantly after two weeks, to a negligible amount recovered after six weeks. The decline in DNA recovery across the six-week period was found to be statistically significant using linear regression analysis for the house and bag samples (p = 2.5 10 3 and 7.0 10 3, respectively), but not for the laboratory samples (p = 0.25). The average amount of DNA recovered at t = 0 was 38.9, 20.4, and 96.3 ng for the laboratory, house and bag samples, respectively. The DNA control solution samples did not decline as rapidly as expected (Fig. 1b). The laboratory samples did not decline below 100% of the starting time period, and the bag samples declined by 50% over the six-week period. Only the house samples were found to decrease significantly in quantity (p = 4.6 10 2). Figs. 2–4 show the Profiler PlusTM amplifications of selected samples from each set of buffy coat samples at time = zero, one, two, four and six weeks. All profiles are displayed at the same scale. No extraneous alleles were located in any of the profiles, indicating that contamination had not occurred. Quantitation of the surface blanks showed that no DNA was present on the surfaces prior to the experiments, and that no additional DNA had been deposited on the surface after the two- and six-week time periods. Table 1 displays the information collated regarding the 50 trace DNA casework samples, arranged sequentially according to the time difference between the offence and the sample collection. The time between the collection of the sample from the crime scene and the submission of the sample to the laboratory is also shown in
Fig. 3. Profiler PlusTM profiles of selected buffy coat ‘house’ samples over six weeks.
J.J. Raymond et al. / Forensic Science International: Genetics 4 (2009) 26–33
30
Table 1 Results from trace DNA casework samples, sorted by the time difference between the offence and sample collection. Time between offence and collection
Time between collection and lab receipt (days)
Sample
Offence
DNA quantity (ng/mL)
Estimated total DNA in extract (ng)
DNA profile resulta
1 h 40 min 1 h 40 min 1 h 40 min 1 h 45 min 2 h 15 min 2 h 19 min 2 h 23 min 2 h 23 min 2 h 40 min 2 h 40 min 2 h 40 min 2 h 52 min 3 h 10 min 3 h 30 min 3 h 30 min 3 h 30 min 3 h 50 min 4 h 5 min 5 h 15 min 5 h 15 min 5 h 40 min 5 h 50 min 8 h 45 min 9 h 25 min 10 h 5 min 10 h 42 min 11 h 35 min 11 h 35 min 14 h 25 min 14 h 58 min 16 h 17 h 31 min 19 h 26 min 19 h 31 min 19 h 55 min 20 h 55 min 21 h 10 min 1d 1d 2d 3d 3d 3d 5d 9d 11 d 22 d 55 d 62 d 62 d
225 33 15 39 11 37 383 383 9 22 22 7 13 230 230 230 169 4 29 29 29 19 7 31 173 24 121 144 24 5 36 66 14 35 12 131 226 16 219 13 14 6 241 17 304 142 180 32 39 8
FPsb at point of entry Earprint on door FPs on counter Register note clips Handmark on counter Service counter FPs at point of entry Security wire above counter Cash register, FPs Arm mark on counter Arm mark on counter Display mat on counter Perspex partition Shopping basket Freezer door Rear loading dock door FPs on window sill Register keyring Area around reg. plates Area around reg. plates Computer harddrive Plant pots Screwdriver handle Earprint on door FPs at point of entry Safe handle Glovebox handle FPs on security camera FPs at point of entry FPs at point of entry FPs on bumper FPs at point of entry Earprint on door FPs at point of entry Earprint on door Auto gearshift Earprint on door Wallet FPs at point of entry FPs at point of entry Partial FP on camera Gas bottle handle FPs at point of entry Passenger doorhandle Handbag Cheque book Glucodin powder box Bag Ziplock bag FPs on laptop
Burglary Burglary Armed robbery Armed robbery Armed robbery Armed robbery Armed robbery Armed robbery Armed robbery Armed robbery Armed robbery Armed robbery Armed robbery Armed robbery Armed robbery Armed robbery Burglary Theft Theft Theft Burglary Burglary Burglary Burglary Burglary Burglary Robbery Arson Burglary Burglary Theft Burglary Burglary Burglary Burglary Vehicle offence Burglary Burglary/SMV Burglary Burglary Burglary Drugs Theft Assault Robbery Theft Drugs/firearms Drugs Drugs Burglary
0 0.003 0.045 0.002 0.017 0.025 0.025 0 0 0.002 0.008 0.056 0.046 0.178 0.007 0.066 0.033 0.004 0 0.001 0.002 0 0.18 0 0 0.002 0.014 0.003 0.02 0.009 0.001 0 0 0.028 0.002 0.001 0 0.074 0 0 0 0.032 0.001 0.022 0.002 0.004 0.004 0.105 0.001 0.032
0 0.12 1.8 0.08 0.68 1.0 1.0 0 0 0.08 0.32 2.24 1.84 7.12 0.28 2.64 1.32 0.16 0 0.04 0.08 0 7.2 0 0 0.08 0.56 0.12 0.8 0.36 0.04 0 0 1.12 0.08 0.04 0 2.96 0 0 0 1.28 0.04 0.88 0.08 0.16 0.16 4.2 0.04 1.28
Neg Partial <12 Mixture Partial <12 Amel only Amel only Partial <12 Neg Neg Mixture Full profile Mixture Mixture Mixture Partial <12 Mixture Mixture Amel only Neg Neg Partial <12 Neg Partial 12 Neg Neg Neg Mixture Neg Partial 12 Neg Neg Neg Neg Partial <12 Neg Neg Neg Amel only Neg Neg Neg Mixture Neg Full profile Neg Partial <12 Neg Full profile Neg Mixture
a Neg: no profile; Amel only: only the Amelogenin locus present in the profile; partial <12: less than 12 alleles in the profile; partial 12: 12 or more alleles in the profile but less than a full profile; and mixture: a mixture of more than 1 person’s DNA in the profile. b FPs: fingerprints.
days. This time delay ranged from four days to over one year (383 days). The DNA quantity is given as the QuantifilerTM output in nanograms per microlitre, and an estimation of the total DNA quantity in the extract. The DNA profile is the result of a Profiler PlusTM amplification of up to 50% of the extract. The profile results are given in the categories designated by the laboratory; either a full profile, a partial profile of 12 or more alleles, a partial profile of less than 12 alleles, a mixture, the Amelogenin locus only, or a negative result. Fig. 5 displays the percentage of each profile type produced from the samples. The laboratory determined that the mixtures recovered in these cases were not suitable for input onto the national database. It should be noted that these mixtures may still have probative value in terms of inclusions or eliminations, or other non-database use. There was no evidence for a linear relationship between the time delay and quantity of DNA recovered in the DNA casework
data (R2 < 0.03). Likewise, no significant reduction in the quality of profile recovered was noted as the time delay between the offence and sample collection increased. Despite lengthy time delays between the sample collection and laboratory submission, there were again no significant decreases in the DNA quantity and profile quality (p = 0.86 and 0.28). Fig. 6 shows the spread in DNA concentration across five time periods between the offence and the DNA sample collection. Whilst the samples of highest concentration tended towards the lower time frames, there is a large range at each time period. 4. Discussion The experimental data support the hypothesis that the chance of recovering DNA from an outdoor surface decreases significantly over time. If a DNA profile is recovered from a similar location, it is
J.J. Raymond et al. / Forensic Science International: Genetics 4 (2009) 26–33
31
Fig. 4. Profiler PlusTM profiles of selected buffy coat ‘bag’ samples over six weeks.
likely to be from a recent contact. The quantity of DNA deposited in this experiment is up to 300 000 times greater than the amount left during a hand contact, and therefore it is expected that DNA left through hand contact would be undetectable in a much shorter time period. For example, if 150 pg of DNA were left after a hand contact on a bag (it has been estimated that five nucleated cells are present per fingerprint [31]), after two weeks there would be 27% available for recovery. Assuming that there is 100% recovery of the cells, this would leave only 40.5 pg. Whilst this may be feasible in
Fig. 5. Types of profiles recovered from casework trace DNA samples (n = 50).
research scenarios, recovery of such a small amount of DNA at a crime scene would be difficult. As expected, the laboratory-stored samples proved to be more robust than the samples left outdoors. Whilst the amount of DNA declined, full profiles were recovered over the entire six-week period. The results of the laboratory samples indicate that if crime scene samples are left in an environment that is cool, dark and in a low traffic area, profiles may be recovered for a longer period of time.
Fig. 6. Concentration of DNA recovered from casework trace DNA samples, grouped in five time categories of the delay prior to the sample collection.
32
J.J. Raymond et al. / Forensic Science International: Genetics 4 (2009) 26–33
Three points of interest in the experimental data are: the differences in the amount of DNA recovered at t = 0 between the surfaces; the spike in recovery of DNA from the ‘laboratory’ samples at one and two weeks; and the insignificant decline in the ‘laboratory’ and ‘bag’ DNA control solution samples (Fig. 1a and b). These anomalies may be explained in part by an uneven distribution of white blood cells in the buffy coat preparations. Two separate aliquots were each quantitated twice to ascertain the quantity of DNA present, and produced differences of up to 40 ng within and between separate 20 mL aliquots (data not shown). This provides an explanation why the ‘bag’ samples returned significantly more (p = 0.03) DNA than the ‘laboratory’ samples in the first week, and for the spike in recovery of the ‘laboratory’ samples at one and two weeks. In addition, at the one- and two-week time periods the deposit became clearly visible as a dried white spot and therefore was simple to target, whereas prior to this it had not been as easily discernable. The ‘house’ and ‘bag’ samples did not show this spike, and therefore the effect of the outdoor environment on these samples may have negated any targeting advantage provided by the deposits becoming visible. It is supposed that the variation between the ‘house’ and the ‘bag’ samples (Figs. 2–4) may be due to the deposition method. As the house surface was vertical it was more difficult to deposit the fluid sample. Conversely, the ‘bag’ samples were stored horizontally preventing the deposit from spreading, allowing an easier recovery leading to higher quantities and stronger profiles. The casework data in Table 1 indicate the difficulties encountered in forensic trace DNA analysis. The average amount recovered from the trace samples was low at 21 pg/mL or 0.84 ng in the total extract, despite the majority of samples (74%) being recovered within 24 h of the offence. Full profiles were only recovered in three cases (6%), with no profile recovered in approximately half of the cases (Fig. 5). It should be noted that the extraction and amplification methods differ between the experimental and casework samples, which may be a cause of variation in the amount of DNA recovered. Linear regression analysis shows that the time delay prior to sample collection did not significantly affect the quantity of DNA recovered. This may be due to the limited number of samples in this study with a lengthy time delay, and the effect may prove more influencing if further cases could be included. The time delay between the sample collection and the laboratory submission also proved insignificant, despite a third of the samples being held for over three months prior to submission. This concurs with a 2006 study [32] that found no decrease in DNA recovery from swabs stored for different time periods. However, the large number of variables affecting the recovery of DNA from casework samples (such as different donors, differing hand contact, surface types and sampling techniques) greatly inhibits the accurate determination of the effect of a single variable, such as time. These variables are difficult, if not impossible, to ascertain or control in casework which highlights the importance of criminalistic consideration of these factors through research, in order to assist interpretation of casework results. These results emphasise the variability of forensic trace DNA analysis. Whilst all efforts were made to reduce the number of variables in the experimental work, differences were still noted in the outcome. This inherent nature of forensic trace DNA analysis can cause difficulties in the attempt to provide definitive answers to legal questions arising in casework. However, with ongoing research and casework data, guidelines can be offered to assess the likelihood of a suggested scenario, with the potential for Bayesian analysis. This study was used to assist the court’s interpretation of DNA evidence in a murder trial, discussed in Section 1. The defence
asserted that their client’s DNA profile might have been recovered from the crime scene as a result of an innocent visit two weeks prior to the offence. A statement was prepared stating that the ability to recover analysable DNA from outdoor surfaces decreases rapidly over several weeks. With this report countering their argument, and other circumstantial evidence, the accused plead guilty to the murder. 5. Conclusions Human cells deposited on a surface such as a window frame or bag will deteriorate significantly (p < 0.05) over the course of six weeks, to such an extent that the chance of recovering a DNA profile is substantially reduced over this time. However, if the cells are deposited and stored undisturbed in a cool and dark location, as in a laboratory, profiles may still be recovered after six weeks. It should also be noted that these experiments were conducted with a large amount of human cells. Contact between a person’s hand and an object tends to leave far less DNA on the object, as shown by casework data, which would exacerbate the difficulty of recovering a DNA profile over time. It is therefore imperative that trace evidence is collected expeditiously to give the best chance of a successful outcome. This research provides an initial look at the extent of the persistence of DNA in the context of volume crime offences, and the results provided a more objective approach to the interpretation of DNA evidence in a criminal trial. Whilst this is a preliminary and limited study, it demonstrates that research into the trace evidence characteristics of DNA offers a criminalistic approach to biological evidence, of particular importance when the source of the DNA is not in question. Further research of this kind is required to deepen our understanding of persistence of DNA at crime scenes. Acknowledgements The authors would like to thank Ms Wendy Tufts and Dr Tony Raymond, NSW Police Forensic Services Group and Ms Alex Lucca, Genetic Technologies Limited, for providing access and assistance with the trace DNA casework data. References [1] R.A. Van Oorschot, M. Jones, DNA fingerprints from fingerprints, Nature 387 (6635) (1997) 767. [2] M.K. Balogh, J. Burger, K. Bender, P.M. Schneider, K.W. Alt, Fingerprints from fingerprints, International Congress Series, Progress in Forensic Genetics 9 1239 (2003) 953–957. [3] R.A. Wickenheiser, Trace DNA: a review, discussion of theory, and application of the transfer of trace quantities of DNA through skin contact, Journal of Forensic Science 47 (3) (2002) 442–450. [4] B. Gunn, An intelligence-led approach to policing in England and Wales and the impact of developments in forensic science, The Australian Journal of Forensic Science 35 (1) (2003) 149–160. [5] D. Coleman, S. Hyde, I. Gordon, T. Wilson, D. Werrett, S. Bain, et al. The National DNA Database Annual Report 03/04. 2004 [cited 24.2.08]; available from: www.forensic.gov.uk/forensic_t/inside/about/docs/NDNAD_AR_3_4.pdf. [6] S.A. Harbison, J.F. Hamilton, S.J. Walsh, The New Zealand DNA databank: its development and significance as a crime solving tool, Science and Justice 41 (1) (2001) 33–38. [7] P. Voegeli, C. Haas, A. Kratzer, W. Bar, Evaluation of the 4-year test period of the Swiss DNA database, International Congress Series, Progress in Forensic Genetics 11 1288 (2006) 731–733. [8] S.J. Walsh, J.S. Buckleton, DNA intelligence databases, in: J.S. Buckleton, C.M. Triggs, S.J. Walsh (Eds.), Forensic DNA Evidence Interpretation, CRC Press, Boca Raton, FL, 2005, pp. 439–469. [9] P.M. Schneider, P.D. Martin, Criminal DNA databases: the European situation, Forensic Science International 119 (2) (2001) 232–238. [10] G.N. Rutty, An investigation into the transfer and survivability of human DNA following simulated manual strangulation with consideration of the problem of third party contamination, International Journal of Legal Medicine 116 (3) (2002) 170–173.
J.J. Raymond et al. / Forensic Science International: Genetics 4 (2009) 26–33 [11] P. Gill, A. Kirkham, Development of a simulation model to assess the impact of contamination in casework using STRs, Journal of Forensic Science 49 (3) (2004) 7. [12] N. Rudin, K. Inman, Biological evidence as trace evidence: the forensic science of DNA typing, CAC News 4th Qtr. (2002) 19–21. [13] J.J. Raymond, S.J. Walsh, R.A. Van Oorschot, P. Gunn, C. Roux, Trace DNA: an underutilised resource or Pandora’s box? A review of the use of trace DNA analysis in the investigation of volume crime, Journal of Forensic Identification 54 (6) (2004) 668–686. [14] C.I. Petterd, J. Hamshere, S. Stewart, K. Brinch, T. Masi, C. Roux, Glass particles in the clothing of members of the public in south-eastern Australia—a survey, Forensic Science International 103 (1999) 193–198. [15] C. Roux, R. Kirk, S. Benson, T. Van Haren, C.I. Petterd, Glass particles in footwear of members of the public in south-eastern Australia—a survey, Forensic Science International 116 (2001) 149–156. [16] J.M. Curran, C.M. Triggs, J.S. Buckleton, K. Walsh, T. Hicks, Assessing transfer probabilities in a Bayesian interpretation of forensic glass evidence, Science and Justice 38 (1) (1998) 15–21. [17] S.A. Coulson, J.S. Buckleton, A.B. Gummer, C.M. Triggs, Glass on clothing and shoes of members of the general population and people suspected of breaking crimes, Science and Justice 41 (1) (2001) 39–48. [18] F. Taroni, C.G. Aitken, Probabilistic reasoning in the law Part 2: assessment of probabilities and explanation of the value of trace evidence other than DNA, Science and Justice 38 (3) (1998) 179–188. [19] J.M. Curran, T.N. Hicks, J.S. Buckleton, Forensic Interpretation of Glass Evidence, CRC Press, London, 2000. [20] R v Joyce, NTSC: Supreme Court of the Northern Territory, 2002, unreported. [21] Bail application by Hillier, ACTSC, 2003. [22] R.A. Van Oorschot, D.G. Phelan, S. Furlong, G.M. Scarfo, N.L. Holding, M.J. Cummins, Are you collecting all the available DNA from touched objects? International Congress Series, Progress in Forensic Genetics 9 1239 (2003) 803–807.
33
[23] B.C.M. Pang, B.K.K. Cheung, Double swab technique for collecting touched evidence, Legal Medicine 9 (4) (2007) 181–184. [24] A. Lowe, C. Murray, J. Whitaker, G. Tully, P. Gill, The propensity of individuals to deposit DNA and secondary transfer of low level DNA from individuals to inert surfaces, Forensic Science International 129 (1) (2002) 25–34. [25] M. Phipps, S.F. Petricevic, The tendency of individuals to transfer DNA to handled items, Forensic Science International 168 (2–3) (2006) 162–168. [26] J.J. Raymond, S.J. Walsh, R.A. Van Oorschot, P.R. Gunn, L. Evans, C. Roux, Assessing trace DNA evidence from a residential burglary: abundance, transfer and persistence, in: 22nd Congress of the International Society for Forensic Genetics, 2007, Copenhagen, Denmark, 2007. [27] Crime victims and the prevention of residential burglary. 2004 [cited 2008 17/3]; available from: www.jcs.act.gov.au/eLibrary/OtherReports/Burglary_Victims_Report.pdf. [28] F. Kasses, Pers. comm., Sydney Luggage Centre, 2008. [29] Australian Government Bureau of Meteorology: Sydney, NSW, November and December 2007 Daily Weather Observations. 2007 [cited 2008 24.2]; available from: www.bom.gov.au. [30] D. Sweet, M. Lorente, J. Lorente, A. Valenzuela, E. Villaneuva, An improved method to recover saliva from human skin: the double swab technique, Journal of Forensic Science 42 (2) (1997) 320–322. [31] F. Alessandrini, M. Cecati, M. Pesaresi, C. Turchi, F. Carle, A. Tagliabracci, Fingerprints as evidence for a genetic profile: morphological study on fingerprints and analysis of exogenous and individual factors affecting DNA typing, Journal of Forensic Science 48 (3) (2003) 586–592. [32] P. Leemans, A. Vandeput, N. Vanderheyden, J.-J. Cassiman, R. Decorte, Evaluation of methodology for the isolation and analysis of LCN-DNA before and after dactyloscopic enhancement of fingerprints, International Congress Series 1288 (2006) 583–585.