Assessing PreCR™ repair enzymes for restoration of STR profiles from artificially degraded DNA for human identification

Forensic Science International: Genetics 12 (2014) 168–180

Contents lists available at ScienceDirect

Forensic Science International: Genetics journal homepage: www.elsevier.com/locate/fsig

Assessing PreCRTM repair enzymes for restoration of STR profiles from artificially degraded DNA for human identification§ James M. Robertson a,*, Shauna M. Dineen b, Kristina A. Scott b, Jonathan Lucyshyn b,c,d, Maria Saeed b, Devonie L. Murphy b, Andrew J. Schweighardt b, Kelly A. Meiklejohn b a

Counterterrorism and Forensic Science Research Unit, Federal Bureau of Investigation Laboratory Division, 2501 Investigation Parkway, Quantico, VA 22135, United States b Counterterrorism and Forensic Science Research Unit, Visiting Scientist Program, Federal Bureau of Investigation Laboratory Division, 2501 Investigation Parkway, Quantico, VA 22135, United States c Armed Forces DNA Identification Laboratory, Armed Forces Medical Examiner System, 115 Purple Heart Ave., Dover Air Force Base, Dover, DE 19902, United States d American Registry of Pathology, P.O. Box 495, Dover, DE 19903, United States

A R T I C L E I N F O

A B S T R A C T

Article history: Received 3 July 2013 Received in revised form 8 April 2014 Accepted 21 May 2014

Forensic scientists have used several approaches to obtain short tandem repeat (STR) profiles from compromised DNA samples, including supplementing the polymerase chain reaction (PCR) with enhancers and using procedures yielding reduced-length amplicons. For degraded DNA, the peak intensities of the alleles separated by electrophoresis generally decrease as the length of the allele increases. When the intensities of the alleles decrease below an established threshold, they are described as drop-outs, thus contributing to a partial STR profile. This work assesses the use of repair enzymes to improve the STR profiles from artificially degraded DNA. The commercial PreCRTM repair kit of DNA repair enzymes was tested on both purified DNA and native DNA in body fluids exposed to oxidizing agents, hydrolytic conditions, ultraviolet (UV) and ionizing radiation, and desiccation. The strategy was to restrict the level of DNA damage to that which yields partial STR profiles in order to test for allele restoration as opposed to simple allele enhancement. Two protocols were investigated for allele restoration: a sequential protocol using the manufacturer’s repair procedure and a modified protocol reportedly designed for optimal STR analysis of forensic samples. Allele restoration was obtained with both protocols, but the peak height appeared to be higher for the modified protocol (determined by Mann–Kendall Trend Test). The success of the approach using the PreCRTM repair enzymes was sporadic; it led to allele restoration as well as allele drop-out. Additionally, allele restoration with the PreCRTM enzymes was compared with restoration by alternative, but commonly implemented approaches using RestoraseTM, PCRBoostTM, bovine serum albumin (BSA) and the MinifilerTM STR system. The alternative methods were also successful in improving the STR profile, but their success also depended on the quality of the template encountered. Our results indicate the PreCRTM repair kit may be useful for restoring STR profiles from damaged DNA, but further work is required to develop a generalized approach. Published by Elsevier Ireland Ltd.

Keywords: DNA repair Degraded DNA PreCRTM Repair Mix Degraded DNA reference sample Short tandem repeat (STR) analysis Forensic biology

1. Introduction In forensic investigations, DNA analysis plays a major role in human identification. However, evidence collected from a

§ Disclaimer: This is publication 13–12 of the Laboratory Division of the Federal Bureau of Investigation (FBI). Names of commercial manufacturers are provided for identification purposes only, and their inclusion does not imply endorsement of the manufacturer or its products or services by the FBI. The views expressed are those of the authors and do not necessarily reflect the official policy or position of the FBI or the U.S. Government. * Corresponding author. Tel.: +1 703 632 4555; fax: +1 703 632 4500. E-mail address: [email protected] (J.M. Robertson).

http://dx.doi.org/10.1016/j.fsigen.2014.05.011 1872-4973/Published by Elsevier Ireland Ltd.

crime scene may have been exposed to environmental and/or chemical stresses that may produce lesions in the DNA. If the DNA has been damaged, the progression of the DNA polymerase during the polymerase chain reaction (PCR) may be prevented. PCR amplification of the short tandem repeat (STR) loci or of the mitochondrial hypervariable 1 (HV1) and HV2 regions are the major tools in human identification. Thus any DNA damage that halts the polymerase can result in a partial or full loss of the STR profile and make obtaining complete HVI and HV2 sequences difficult. The issues involved with analysis of degraded DNA have been collated and presented in a thorough review [1].

J.M. Robertson et al. / Forensic Science International: Genetics 12 (2014) 168–180

DNA damage, such as double-strand breaks, single-strand nicks and modified bases, can occur from numerous processes. For instance, DNA in an aqueous environment can undergo a loss of base residues at high temperatures, leading to base modifications and chain scission [2]. Heat can induce oxidation of the nucleotides and produce base modifications such as 8-oxoguanine, which can result in mispairing [3]. Bleach, which is often used to clean bones and teeth of contaminating environmental DNA [4], produces oxidized residues but also fragments DNA [5]. Exposure to ultraviolet (UV) radiation can introduce lesions in the DNA that, if not repaired, can alter the DNA structure [6,7]. Hydrolytic degradation of DNA can even occur in the dry state [8]. DNA is also susceptible to fragmentation upon repeated freezing and thawing events, which is generally measured by the presence of nick translation [9]. Chemical stresses, such as changes in pH and the presence of metal radicals, can cause base aberrations while reactive aldehydes can produce non-repairable crosslinks [5]. A compilation of the common lesions expected from a variety of damage treatments is given in Supplementary Table S1. In living cells, excision repair pathways can correct lesions in the DNA caused by either endogenous processes (e.g., oxidation) or exogenous agents (e.g., radiation). These repair mechanisms include enzymes such as: glycosylase to excise modified or mismatched bases [10], apurinic or apyrimidinic (AP) endonucleases to remove AP sites [5,11], DNA polymerase to fill in gaps, and ligase to seal nicks [12]. DNA exposed to the environment and certain chemicals, which is common in evidentiary samples, does not have the protection and benefit of these cellular processes since the cells are no longer living. In addition upon cell death, not only is the DNA is exposed to a cascade of enzymes that metabolize the macromolecule into fragments and precursor molecules [1], but it is also susceptible to metabolism by microorganisms. Highly fragmented DNA from nonliving cells may not be able to produce long STR amplification products required for full STR profiles. Similarly, DNA in archived evidence from unsolved crime cases and in stored extracts may suffer damage and be difficult to analyze [13]. Within the last decade, researchers have investigated strategies to repair damaged DNA in non-living cells. Some of this effort was aimed at increasing the ability to sequence ancient DNA (aDNA), which is largely, but not exclusively characterized by modified residues (deaminated pyrimidines), but also fragmentation [14]. The strategies often involved treatment of aDNA with a mixture of repair enzymes, however the success of these repair treatments depends on the extent and type of DNA damage. Initially the only available source of repair enzymes was research laboratorydevised mixtures. Responding to the need for a commercial kit for the repair of damaged DNA, New England Biolabs (NEB) developed the PreCRTM Repair Mix, which includes a polymerase, ligase, endonucleases, and glycosylases. The suite of enzymes was developed to repair abasic sites, nicks, thymidine dimers, blocked 30 ends, oxidized guanine and pyrmidines and deaminated cytosines; however, it does not ligate double-stranded breaks. Considering the PreCRTM repair kit has the potential to repair multiple types of lesions in DNA (Supplementary Table S1), the kit is promising for forensic analysis of compromised samples, as the DNA damage may not be restricted to a specific type of lesion. In this work, the PreCRTM Repair Mix was assessed for its ability to restore DNA profiles from forensically relevant, yet compromised samples. Preliminary studies were focused on optimizing the PreCRTM Repair Mix using purified human DNA, which was artificially degraded by several methods to obtain partial STR profiles. After optimization, the PreCRTM Repair Mix was tested with body-fluid samples of blood, saliva and semen from varying donors. These samples were subjected to various degradation treatments to determine if lost STR alleles can be restored using the

169

PreCRTM Repair Mix. In addition to this, we evaluated RestoraseTM, a mixture of polymerases reported to improve analysis of compromised DNA [15], for repairing some degraded body fluid samples. Experiments were also performed to distinguish STR restoration by repair from STR improvement by general PCR enhancement, using two approaches: (1) PCRBoostTM, a commercial product reported to enhance PCR; and (2) bovine serum albumin (BSA), which has been documented to relieve PCR inhibition [16] and promote successful genotyping from compromised forensic samples [17]. After either the repair or enhancement treatment of all samples, STRs were generated using the kits AmpF‘STR1 Identifiler1 and AmpF‘STR1 MinifilerTM, the latter of which may provide full STR profiles from compromised DNA without the addition of repair enzymes or enhancers [18]. 2. Materials and methods 2.1. DNA source and sample collection In this work, the term purified-DNA is used for genomic DNA that was purchased as a purified DNA standard or for samples which had been purified before use (i.e., DNA was no longer in whole cells). DNA was purchased from the American Type Culture Collection (HL-60 DNA; Manassas, VA), Promega Corp. (9947A Control DNA; Madison, WI) and Applied Biosystems (AB) (Raji Control DNA; Foster City, CA). Native-DNA is used for DNA obtained from body fluids that had undergone either artificial or environmental degradation. A range of body fluid samples were collected from a set of volunteer laboratory staff according to the FBI Institutional Review Board (IRB) privacy protection procedures. For saliva samples, donors were requested to refrain from eating, drinking, and teeth brushing 30 min prior to collection. Saliva was collected from five donors after they swished 5 mL of room temperature water around in their mouths for 30 s. This was repeated until 3–5 mL saliva was collected from each donor (stored in a 50 mL tube). Blood samples were collected from five donors using two approaches: (1) with a Contact-Activated Lancet (BD Microtainer, Franklin Lakes, NJ) and used immediately for preparing stains; and (2) using a Vacutainer EclipseTM Blood Collection system with a tube containing EDTA as an anticoagulant. For semen samples, each donor was provided with a sterile, leakproof container. Both donors (n, 2) were instructed to refrigerate the sample instead of freezing it and to return the container to the laboratory within 24 h of collection. In addition, semen from three donors was purchased from the Fairfax Cryobank (Fairfax, VA). After collection, body-fluid samples were divided into small aliquots (to minimize freeze/thawing events) and were stored at 4 8C for immediate use or frozen, either at 20 8C (blood and saliva) or in liquid N2 (semen), for long term storage. 2.2. Preparation of artificially damaged purified-DNA and native-DNA in body fluids The objective of the various damage treatments was to generate partial-profile DNA as opposed to DNA with complete profile loss or DNA with all allele peaks present but with only reduced peak heights. With a partial-profile DNA substrate, lost alleles could theoretically be recovered upon repair, helping to distinguish restoration from PCR enhancement and general improvement in peak height. Purchased purified DNA and native DNA samples from donors were artificially damaged in the laboratory following strategies obtained from published procedures as described in Supplementary material 1 [2,19–25]. Samples were subjected to hydrolysis, oxidation by bleach and hydrogen peroxide (H2O2), UVC irradiation, gamma irradiation, desiccation and environmental conditions, as outlined in Table 1

J.M. Robertson et al. / Forensic Science International: Genetics 12 (2014) 168–180

170

Table 1 Summary of the methods used to damage purified and native DNA for use in repair experiments.

Oxidation: H2O2

Bleach

Hydrolysis

Method

DNA source

Substrate

(1) 225 ng DNA incubated for 1 h at 37 8C with 1 volume of H2O2 (30 mM) and 1.25 volumes of Fenton Reagent (FeEDTA) (2) 500 mL of 30 mM H2O2 and 100 mL Fe (III)-EDTA were applied to a stain (from 5 mL of sample) and incubated at 37 8C overnight (3) A stain (from 5 mL of sample) was incubated for 80 min at 37 8C with 8 mL TE4,15 mL of Fenton Reagent and 12 mL of H2O2 (30 mM) (4) 100 mL aliquots of stock H2O2 (30% w/v) were applied to a stain until covered and incubated until the reaction ended (1) 27 mL of sample (250 ng) was incubated at ambient temperature with 1 mL 1:10 v/v bleach for 0, 5 or 10 min (2) Sufficient 5000 ppm bleach was applied to cover a stain (from 5 mL of sample) in a microfuge tube and incubated overnight at ambient temperature (3) 5 mL of sample was incubated at ambient temperature with 25 mL of TE4 and 5 mL of 5000 ppm bleach for 0, 5 or 10 min (4) 25 mL of 5000 ppm bleach was applied to a stain (from 5 mL of sample) and incubated at ambient temperature for 15, 30 or 45 min

Purified: HL-60

Liquid (TE4 buffer)

Native: blood and semen

Dry cotton swatches

Native: blood and semen

Dry cotton swatches

Native: blood and semen

Dry cotton swatches

Purified: HL-60, 9947A

Liquid (TE4 buffer)

Native: blood, saliva and semen

Dry cotton swatches

Native: blood, saliva and semen

Liquid (TE4 buffer)

Native: blood, saliva and semen

Dry cotton swatches

Purified: 9947A and HL-60

Liquid (depurination buffer)

Native: whole blood extracts

Liquid (dH2O)

(1) 200 ng DNA was incubated with 50 mL of depurination buffer at 70 8C for 30 h (2) 1, 5 and 50 ng DNA was incubated at 70 8C for up to 3477 min

UVC irradiation

50 mL of sample in a microfuge tube was placed on the floor of a Spectrolinker XL-1500 and exposed to radiation (2– 4 min, HL-60 ; 16–24 h, blood) at 0.4 mW/cm2

Purified: HL-60; native: blood

Liquid (TE4 buffer)

Gamma irradiation

A stain (from 7.5 mL of sample) was exposed to a 60Co source for 0, 1, 2, 4 or 8 Rads

Native: saliva

Dry cotton swatches

Desiccation

25 mL of sample in a microfuge tube was incubated at 50 8C in an oven for 1 month

Purified: whole blood extracts

Liquid (TE4 buffer)

Environmental conditions

Stains (from 50 mL of sample) were exposed to the environment (on the roof of a building) for up to 2 months during summer

Native: blood, semen and saliva

Dry cotton swatches

Summary of the methods used to damage DNA by oxidation, hydrolysis, UVC irradiation, gamma irradiation, desiccation and exposure to environmental conditions. Information on the DNA (purified and native) subjected to each method and the substrate is also provided. Italicized procedures were the ones selected for use after initial experimentation proved them to be optimal. Two procedures were used for hydrolysis: one was accelerated by use of depurination buffer; whereas, the other was without acceleration.

and detailed in Supplementary material 1. Both a reagent blank and two time zero controls, that included the same DNA and all necessary reagents but were not subjected to the damage, were included in all treatments. These samples were used as a reference to determine the extent of damage from each treatment (i.e., partial-profile or just a reduction in peak heights), but also the subsequent recovery or enhancement of the profiles.

pooled, quantified with Quantifiler1 Duo (AB) and divided into aliquots for storage at 80 8C. This procedure enabled the preparation of a degraded DNA reference sample that that would have different types of lesions like a true environmental sample and that could be reproduced as needed.

2.3. Reference sample for degraded DNA studies

After treatment, DNA was isolated from body fluids following three protocols based on the analyst’s preference. All three protocols implemented are commonly used in laboratories to obtain comparable yields of purified DNA. Cell lysis and DNA extractions were performed using the organic extraction method with phenol/chloroform/isoamyl alcohol and purification/concentration by filtration through Microcon1 100 Ultracel YM-100 centrifugal filters (Millipore Corp., Billerica, MA) [26]. DNA was also isolated with spin-column technology using the QIAmp1 DNA Blood MiniKit (Qiagen, Hilden, Germany) in experiments with few replicates; whereas, the EZ1 DNA Investigator Kit on the EZ1 Advanced XL (Qiagen) was used when there were several replicates.

Toward the end of the experimentation reported herein, an environmental chamber (Q-Sun Xe-3 HSC Test Chamber with Chiller, Q-Lab Corp., Cleveland, OH) was purchased to be able to make a ‘baseline’ sample as a reference sample of environmentally exposed DNA for inclusion in allele restoration experiments. Twenty 25 mL aliquots of 1 ng/mL 9947A or Raji purified DNA in TE4 (10 mM Tris–HCl, 0.1 mM EDTA, pH 8) were deposited into the center of a washed, UVC-sterilized glass vial, which had a PTFE rubber lined lid (Wheaton, Millville, NJ). After drying, the samples were exposed to a specific dose of UVA at a set humidity, air temperature, and surface temperature in the environmental chamber. Environmental conditions were programmed to simulate a typical summer, 24-h day in Quantico, VA, with the exposure levels varying during the 15 h of daylight (Supplementary Table S2). Following exposure, the samples were recovered in water,

2.4. DNA extraction

2.5. Restoration of STR profiles Two DNA repair kits were tested in these experiments: the PreCRTM Repair Mix (NEB, Ipswich, MA) and RestoraseTM DNA

J.M. Robertson et al. / Forensic Science International: Genetics 12 (2014) 168–180

171

Fig. 1. PreCRTM Repair Mix optimized protocols: the sequential protocol, the ATF/AFDIL protocol and the modified ATF/AFDIL protocol.

Polymerase (Sigma–Aldrich, St. Louis, MO). The PreCRTM Repair Mix has a variety of enzymes to repair several DNA lesions (Supplementary Table S1). For studies with PreCRTM Repair Mix, we used either a sequential method based on the manufacturer’s protocol or an optimized version of the ATF/AFDIL protocol [27] (subsequently called the modified ATF/AFDIL protocol). In the sequential protocol (Fig. 1), 5 mL of a damaged DNA sample (or extract; 1–5 ng) was mixed with 5 mL of PreCRTM Repair Mix (including 0.5 mL of PreCRTM enzyme) and incubated at 37 8C for 4 h. Upon completion of the incubation, 15 mL of AmpF‘STR1 Identifiler1 STR master mix (containing 5 mL of primers) was added and the DNA was amplified. In the ATF/AFDIL protocol, 5 mL of damaged DNA sample (or extract; 1–5 ng) was added to 15 mL of Identifiler1 STR master mix (containing the PreCRTM enzyme) and incubated at 37 8C for 20 min. Following incubation, 5 mL of AmpF‘STR1 Identifiler1 STR primers were added and the DNA was amplified. The ATF/AFDIL protocol was optimized by examining the benefit of adding 0.5 mL 10 mM dNTPs post incubation and varying the incubation time, volume and temperature. AmpFSTR1 MinifilerTM was run on damaged templates both with (n, 63) or without (n, 84) PreCRTM treatment, and the profiles were compared with those for AmpF‘STR1 Identifiler1 run on the PreCRTM-treated templates. For experiments with RestoraseTM (n, 26), the manufacturer’s protocol was followed.

Enhancement of STR profiles was examined using PCRBoostTM (Biomatrica, San Diego, CA) and BSA (Sigma–Aldrich, St. Louis, MO). The AmpF‘STR1 Identifiler1 STR kit was used in the experiments with artificially damaged DNA because the peak heights of the longer alleles of this system can be lower than those of AmpF‘STR1 Identifiler1 Plus system (as shown in Supplementary Fig. S1), making Identifiler1 peaks more prone to drop-out in damaged DNA. In PCRBoostTM experiments (n, 41), a working concentration of 0.2 ng/mL of DNA in TE4 was diluted to 0.1 ng/mL with one volume of the undiluted PCRBoostTM reagent. Both the NEB buffer used in the sequential protocol and the AmpF‘STR1 PCR buffers used in the modified ATF/AFDIL protocol contain BSA. To determine if an additional amount of BSA would be beneficial, amplifications (n, 46) were performed after adding BSA to a concentration of 0.16 mg/mL in the AmpFSTR1 Identifiler1 or MinifilerTM reaction buffers.

2.6. DNA quantification and profiling

2.8. Statistical analyses

DNA quantification was performed with the Quantifiler1 Duo DNA Quantification Kit (AB) using the 7900 HT Fast Real-Time PCR System (AB). Amplification of the internal positive control (IPC) was monitored to determine if inhibitory residues from the chemicals used to produce lesions and to quench the reactions were successfully removed during the clean-up step. STRs were amplified using the AmpFSTR1 Identifiler1 PCR Amplification Kit (AB) and AmpFSTR1 MinifilerTM PCR Amplification Kit (AB) according to the manufacturer’s instructions. When the amount of DNA after quantification was either very low or could not be

In this study, a partial profile is defined as a profile with at least one allele drop-out. Alleles were considered restored when their peak heights exceeded the threshold to allow a presumptive genotype (>50 RFU for heterozygotic loci). The ultimate goal was to restore the peak heights of heterozygotic alleles to values >200 RFU. The statistical significance of allele restoration using the PreCRTM Repair Mix with any protocol was determined using the Mann–Kendall Trend Test (MKTT) [28,29]. The output of the MKTT is an assessment of the trend direction [Kendall Score (S)] and its statistical significance. In cases where no statistically significant

determined, the maximum sample volume was used in the AmpFSTR1 Identifiler1 reaction. STR profiles were obtained on the 3130 Genetic Analyzer (AB) using POP-4 separation polymer with a 10 s injection at 3 kV and run at 15 kV at 60 8C. Data were analyzed using GeneMapper1ID Analysis Software v 3.2 (AB). Each sample was run in duplicate to determine reproducibility. 2.7. Improvement of STR profiles

J.M. Robertson et al. / Forensic Science International: Genetics 12 (2014) 168–180

172

trends are obtained from the MKTT, the direction of the trend can still be estimated from the sign of the S-value. For example, a negative S-value suggests a decreasing trend and vice versa. The slopes of the regression lines were compared to determine statistical differences in the restoration trends between the protocols [28]. Differences were considered statistically significant at a 95% confidence interval when two-sided p-values were less than 0.05, with no adjustments for making multiple comparisons.

the PreCRTM repair kit, the reaction mix was enhanced with additional dNTPs post incubation (Fig. 1). Improvement in peak heights of the degraded purified-DNA was observed in amplifications containing additional dNTPs, but the effect was sporadic (results not shown). However, given that adding supplementary dNTPs overall did not have a negative impact on allele peak heights, dNTPs were added after incubation in all subsequent experiments.

3. Results

3.1.1.2. Incubation temperature. We examined the temperature dependence of repair on artificially degraded purified-DNA at 4 8C, 20 8C and 37 8C (n, 17). An example of the results with hydrolytically damaged DNA are depicted in Fig. 2, where the allele peak heights of the damaged DNA are plotted versus the allele lengths, from short to long on the X-axis. The bar graph denoting average peak height presents a ski-slope shape (Fig. 2). This result indicates the longer alleles of the artificially degraded DNA were lost prior to the shorter ones, as often observed for DNA from actual forensic samples exposed to the environment [18]. Allele restoration, as observed by peaks amplifying where prior to repair there were none (e.g., D7S820 and CSF1PO), was optimal at 37 8C (Fig. 2). The peak heights of the shorter alleles increased on average by 1.6 fold with incubation at 37 8C (from AMEL to D21S11). The peak heights of the longer alleles increased even more with incubation at the higher temperature, from 3.7 for D2S1338, 2.8 for D18S51, to 2.2 for D16S539 (Fig. 2). Performing the repair reaction at either 4 8C (as recommended by the manufacturer) or 20 8C did not improve the allele peak height (Fig. 2) of these samples. Results similar to those obtained with hydrolytically damaged purified-DNA were also obtained with purified-DNA degraded by UVC and oxidation (results not shown).

3.1. Investigations with purified-DNA 3.1.1. Optimization of repair protocol with purified-DNA Experiments were performed with the NEB PreCRTM Repair Kit to ascertain whether the manufacturer’s protocol was adequate or could be improved for samples of forensic interest that generally contain low-quality DNA. Initially, all repair reactions were performed on purified-DNA using the sequential approach, with lengthening the incubation time to 4 h being the only change to the manufacturer’s guidelines. During the course of this work, a group reported better STR results when the repair reaction was performed in the AmpF‘STR1 Identifiler1 PCR buffer, followed by the addition of STR primers to the treated DNA mixture for amplification (called in this study the ATF/AFDIL protocol) [27]. Since this protocol was originally adopted for UV-damaged DNA, we performed optimization studies of this protocol using oxidatively and hydrolytically damaged purified-DNA. 3.1.1.1. Addition of supplementary dNTPs post incubation. To test whether the dNTPs present in the Identifiler1 Master Mix may be reduced to a suboptimal level during the initial incubation with

2000 No Repair

1800

4°C Repair

1600

20°C Repair

Average peak height (RFU)

37°C Repair

1400 1200 1000 800 600 400 200 0

Alleles Fig. 2. Temperature dependence on repair of a hydrolytically damaged purified-DNA (HL-60) using the ATF/AFDIL protocol. The bar height is the average of two independent repair reactions and Identifiler1 amplifications. The error bars represent the standard deviation of the mean. The results shown are a good representation of the trends seen in the other 17 temperature dependant experiments, performed on a variety of degraded purified-DNAs. Two labels are given for heterozygotic loci, one for each allele. Peaks that are labeled twice denote the different alleles of a heterozygotic pair.

J.M. Robertson et al. / Forensic Science International: Genetics 12 (2014) 168–180

3.1.1.3. Incubation time and volume of enzyme mix. A range of incubation times from minutes to several hours was tested on artificially damaged purified-DNA (n, 16): 0 min, 20 min, 1 h, 3 h and overnight. However, it was difficult to establish an optimal incubation time as this was influenced by the extent of DNA damage (quality) and amount of DNA in a given sample. For example, the reported optimal incubation time of the original ATF/ AFDIL protocol was 20 min (at 37 8C) for UVC–treated samples [27]. However, the time required for maximum allele restoration could be much longer for samples damaged by other conditions, such as for oxidized (Fe(III)/H2O2) purified DNA (depicted in Supplementary Fig. S2). For this damaged DNA, an overnight incubation at 37 8C resulted in restoration of the dropped-out alleles except for the larger one at the CFS1PO locus. However, from a series of 27 preparations of damaged purified-DNA, it was concluded that the restoration of lost alleles generally did not increase after 4 h incubation at 37 8C. Finally, a series of studies was performed to determine the optimal volume of the enzyme mix (n, 19). Using 0.25, 0.5, 1 and 2 mL, we determined there was no apparent advantage in using a larger or smaller volume of the PreCRTM repair kit enzyme mix than the 0.5 mL, as suggested by the manufacturer (results not shown). 3.1.1.4. Comparison between sequential and modified ATF/ATDIL protocols. Using a baseline sample for degraded purified-DNA exposed in an environmental chamber to the synergistic effects of heat, humidity, and UV radiation, the sequential and modified ATF/ ATDIL protocols were compared using 10 repair reactions performed on different days. The average allele peak heights using the modified ATF/ATDIL protocol were higher, but the standard deviations from the mean values were also higher (Fig. 3). Using the MKTT, the 2sided p-value was 0.029732 for the modified ATF/ATDIL protocol, but lower for the sequential protocol, 0.0072008 (with S-values of 111 and 137, respectively). This result indicates that the sequential protocol yields the more consistent allele peak heights upon repair with the PreCRTM enzymes. When the peak heights were plotted

173

against the allele lengths, the slopes of the two regression lines were almost identical (Supplementary Fig. S3) and a slopes comparison analysis indicated the differences between them were insignificant (p-value, 0.9967085). After several experiments with different artificially damaged purified-DNA samples, it was evident that both the sequential and modified ATF/ATDIL protocols repaired damaged DNA. This conclusion was reached from studies of 348 repair reactions following the sequential protocol and 368 reactions with the modified ATF/ATDIL protocol. Differences in allele restoration efficiencies between the two protocols with the same substrate were sometimes observed by a single operator. However, before one can make a generalized statement or a recommendation for one protocol over the other, more experimentation needs to be performed than done here to understand the role of stochastic effects and operator performance. 3.1.2. Observations for repair of artificially damaged purified-DNA In this study the purified DNA was subjected damage via hydrolysis, oxidation (bleach and H2O2) and UVC irradiation. The results from 716 repair reactions that included 27 individual artificial degradation preparations of purified DNA are collated in Table 2. Traversing a row for a particular treatment, i.e., UVC damage, a notation mark () is provided for each locus for which the alleles either dropped out partially or fully in at least one preparation; however, the presence of this mark does not mean the alleles dropped out in all of the preparations. A mosaic pattern is a good description for the results observed for the medium-sized loci (listed smallest to largest from left to right). Also noted in Table 2 are the loci that were restored with the PreCRTM enzymes (denoted by U). Incubation with the PreCRTM enzymes also led to allele drop-out in 48 reactions (denoted by ). Both allele restoration and drop-out were observed in 27 reactions. As depicted in Fig. 2 and Supplementary Fig. S2, longer alleles more commonly drop out in degraded DNA when compared to the shorter alleles. In addition, loci with two alleles may drop out sooner, if the intensities of the alleles are lower than those of

450

Average peak height (RFU)

400 350 300

sequential 250

modified ATF/AFDIL

200

unrepaired

150 100 50 0

Alleles Fig. 3. Comparison of the sequential and modified ATF/AFDIL protocols for repairing a degraded DNA baseline sample subjected to UV irradiation under controlled heat and humidity in an environmental chamber. Ten independent repair reactions were performed on the baseline sample at 10 different times by a single researcher using each protocol. Bars represent the mean of all 10 reactions. The error bars represent the standard deviation of the means. Drop-outs in the unrepaired baseline sample included alleles at FGA, D7S820, D18S51, and D2S11338 and are masked by the averaging except for the second FGA allele. Two labels are given for heterozygotic loci, one for each allele.

174

Table 2 Loci influenced by degradation and PreCRTM treatments of purified-DNA. AMEL

D19S433

 U

D3S1358

D8S1179

D5S818

 U





 U 

 U   U 

 U 

 

vWA

TH01

 U 

D21S11

  U  U 

D13S317

TPOX

FGA

D7S820

D16S539

D18S51

D2S1338

CSF1PO



 U

 U





 U  

 U

 U   U 



 U

 U  

 U  

 U 

 U 

 U   U   U 

 U 

 U 



 U 

Depicted in the table are results for 105 experiments (716 reactions) of damaged purified DNA repaired with PreCRTM enzymes using both the sequential and the modified ATF/AFDIL protocols. Loss of alleles upon degradation (), gain of alleles upon repair (U) and loss of alleles during repair () are represented for three degradation treatments (UVC [n, 3]; oxidation through H2O2 or with bleach [n, 12]; and hydrolysis by heating at 70 8C in low pH [n, 12]). Heterozygotic loci of the larger alleles were the most prone to drop out due to their inherently lower peak heights. The percentage of alleles dropped-out, restored, or lost was not estimated because of the stochastic nature of the reactions: the effects observed varied within and between experiments probably because the lesions were random in number and in placement.

Table 3 Loci influenced by degradation and PreCRTM treatments of native-DNA. AMEL H2O2 loss/saliva H2O2 restored/saliva Bleach loss/saliva Bleach restored/saliva H2O2 loss/blood H2O2 restored/blood Bleach loss/blood Bleach restored/blood Environmental exposure loss

D19S433

D3S1358

D8S1179

 U

 U 

D5S818

vWA

TH01

D21S11

D13S317

TPOX

FGA

D7S820

D16S539

D18S51

D2S1338

CSF1PO

 U 

 U 

 U 

 U 



 U 

 U 

 U 

 U 



 U 



 U 

 U

 U

 U 

 U 

 U 





 U 

 U 

 U

 U 

  se

 se  bl

 se  bl U bl

 se  se  bl  bl Environmental exposure restored U se U se U bl U bl Loss of alleles upon degradation () and gain of alleles upon repair with PreCRTM enzymes (U) are represented for artificial degradation of liquid saliva (n, 17) or liquid blood (n, 17) with H2O2 and bleach. Alleles lost to bleach treatment could not be restored. Results are also shown for semen (se) and blood (bl) stains subjected to environmental exposure on the roof of the building for 1 or 2 months before retrieval for analysis (n, 4).

J.M. Robertson et al. / Forensic Science International: Genetics 12 (2014) 168–180

UVC loss UVC restored Lost during repair (UVC) Oxidation loss Oxidation restored Lost during repair (oxidation) Hydrolysis loss Hydrolysis restored Lost during repair (hydrolysis)

J.M. Robertson et al. / Forensic Science International: Genetics 12 (2014) 168–180

neighboring homozygotic loci. These trends however may depend on the type of lesion and its frequency. For example, with UVC irradiation, alleles of TPOX, FGA, D7S820, D18S51, and D2S1338 were lost before the longer CFS1PO locus (results not shown); however, the D21S11 alleles appeared to be very resistant to UVC damage. The D7S820 locus, which has alleles of medium length, was usually the first to drop out after UVC irradiation and hydrolysis. These observations are in concordance with those previously reported for UVC-treated DNA samples [27]. The D18S51 and D2S1338 loci of hydrolytically damaged DNA were repaired efficiently; whereas D2S1338 was rarely repaired in oxidized DNA preparations. 3.2. Investigations with native-DNA from body fluids 3.2.1. Repair of oxidized native-DNA In this study, we used several protocols to oxidize native DNA in body fluid samples because either the treatment was too harsh, resulting in no STR profiles for the extracted DNA, or was too gentle, resulting in full STR profiles because the reagent apparently did not modify the chromatin-packaged DNA in the cells. In experiments where DNA was damaged by the oxidizing agents yielding partial profiles (n, 17 for both blood and saliva), we observed an allele drop-out pattern (presented in Table 3) similar to that observed for purified-DNA (Table 2). Differences were observed at D3S1358 (sensitive to H2O2 in saliva but not blood) and D8S1179 (sensitive to H2O2 in blood but not saliva). D7S820, D18S1338 and CSF1PO alleles damaged by H2O2 were either not restored or only in some replicates; whereas, alleles lost due to

175

bleach treatment could not be restored in these experiments. It was not unusual to lose alleles from oxidized samples upon treatment with the PreCRTM enzymes. An example of this is shown for a blood sample that was treated with H2O2 (Fig. 4). A reduction in peak height and allele drop-out was observed upon incubation with the PreCRTM enzymes (bottom panel, Fig. 4). This example emphasizes that further damage to the degraded native-DNA is possible upon treatment with the PreCRTM enzymes. A reduction in peak height after incubation with PreCRTM enzymes was a general trend noted for the majority of the oxidized blood samples (results not shown). Blood treated with bleach under conditions that produced partial profiles was especially sensitive to the enzymes, as indicated by severe allele loss and the observation of off-ladder peaks (Supplementary Fig. S4). 3.2.2. Repair of native-DNA from g-irradiated saliva In this set of experiments, dried saliva from five donors on cotton swabs was exposed to gamma rays for different time intervals up to the known level required to kill pathogens. At the highest dose studied (8 Rads or 0.08 Gy), allele peak heights decreased and substantial allele drop-outs were noted when compared to the unexposed sample (Fig. 5). Using the sequential protocol, incubation with the PreCRTM enzymes led to complete allele restoration (with only D7S820 below 100 RFU; Fig. 5). After repair using the modified ATF/ATDIL protocol, D18S51 and CSF1PO fell below 100 RFU (results not shown). This type of result was often observed for the two methods of repair, showing the stochastic nature of the repair reaction.

Fig. 4. Example of STR profile of native-DNA (blood) prior to oxidation with H2O2 and after oxidation, both without and with repair with PreCRTM enzymes. Alleles for D21S11, D7S820, D13S317, D5S818, CSF1PO, FGA dropped out (indicated by the arrows in the second panel).

176

J.M. Robertson et al. / Forensic Science International: Genetics 12 (2014) 168–180

Fig. 5. Example of STR profiles of native-DNA (saliva) after being exposed to ionizing radiation (8 Rads) with no repair and after repair with PreCRTM enzymes (using the sequential protocol).

3.2.3. Repair of native-DNA from environmentally exposed body fluids Blood and semen stains were exposed to ambient humidity, heat, and sunshine for up to 2 months on the roof of the laboratory building. For blood samples, partial profiles were obtained only after 2 months of exposure. Dropped-out alleles at D18S51, D2S1338 and CSF1PO were restored after repair with PreCRTM enzymes, but the D7S820 locus was refractory (Table 3 and Supplementary Fig. S5). For semen samples, partial profiles were observed after 1 month of exposure (data not shown). Following exposure, alleles were lost at four loci (FGA, D7S820, D18S51, and D2S1338, Table 3). FGA and D2S1338 could be restored with the PreCRTM enzyme treatment (Table 3). Saliva stains were not investigated. 3.2.4. UVC-treated blood samples Experiments were performed with whole blood exposed to UVC irradiation to mimic environmentally exposed body fluid (n, 7). Only full STR profiles were obtained for the DNA extract from the samples, even after 24 h exposure. The peak heights of the alleles were reduced relative to the controls that had not been exposed to the irradiation. These samples were not investigated further, since there were no allele drop-outs for testing the efficiency of the PreCRTM enzymes. 3.3. Alternatives for repairing damaged DNA 3.3.1. Repair of damaged DNA with RestoraseTM enzymes RestoraseTM is a mixture of polymerases reported to repair damaged DNA. RestoraseTM was tested on native-DNA from semen stains, which were exposed to environmental conditions for 2 months (laboratory roof). When compared to PreCRTM enzyme repaired samples, RestoraseTM restored one more allele (at D18S51; n, 2; results not shown). In experiments using the degraded DNA reference (baseline) sample, RestoraseTM and PreCRTM enzymes exhibited similar restoration efficiencies (n, 7; results not shown). 3.3.2. Enhancement of STR profiles from damaged DNA with PCRBoostTM and BSA PCR enhancers were also tested on the artificially degraded DNA to determine whether they improved the STR profiles relative to the PreCRTM enzymes. It is important to clarify that the enhancers cannot repair lesions in damaged DNA, but they should faciliate an increase in allelic peak heights. PCRBoostTM is described by the

manufacturer as a reagent specifically formulated to enhance multiplex amplification of templates from degraded samples. PCRBoostTM was compared with PreCRTM enzymes using purifiedDNA degraded by two different methods. Firstly, purified-DNA was dessicated by heating at 50 8C in the dark for sufficient time to yield alleles with lower peak heights, but not long enough to produce allele drop-outs (1 month). For this template, PCRBoostTM enhanced the peak heights of all alleles more than PreCRTM enzymes did (n, 7; Supplementary Fig. S6). PCRBoostTM was also tested with DNA exposed in the environmental chamber to produce a degraded template with reduced peak heights but no allele drop-outs. PCRBoostTM was only slightly successful in enhancing the amplification of these templates (results not shown). In additional experiments, BSA was investigated for its multiplex PCR enhancing capability using templates degraded by various procedures; the results were inconclusive (not shown). 3.3.3. AmpF‘STR1 MinifilerTM as an alternative to repair with PreCRTM enzymes Another way to obtain genotypes from compromised samples is to use STR kits with reduced length amplicons, such as AmpF‘STR1 MinifilerTM. Experiments with the same artificially damaged purified-DNA or environmentally exposed native-DNA from body fluids were run using both AmpF‘STR1 Identifiler1 and AmpF‘STR1 MinifilerTM. Both kits contain primers for the midlength loci (D21S11, D13S317, FGA) and long-loci (D16S539, D7S820, D18S51, D2S1338, CSF1PO). Since the MinifilerTM system is designed to amplify shorter products than Identifiler1, it may be able to yield full profiles with degraded templates that have the distance between the lesions close to the length between the Identifiler1 primer binding sites. The results indicated that DNA yielding partial AmpF‘STR1 Identifiler1 profiles would sometimes provide full profiles with AmpF‘STR1 MinifilerTM. As an example, genotype comparisons of semen exposed to the environment for 2 months, for Identifiler1 without repair, Identifiler1 with repair and MinifilerTM without repair, are shown in Fig. 6 and Supplementary Table S3. Without repair, allele drop-out at D7S820, D18S51, D2S1338, and FGA was observed for AmpF‘STR1 1 Identifiler1; whereas a full profile was obtained with AmpF‘STR1 MinifilerTM. After treatment with the PreCRTM enzymes, restoration of all but one allele (at D18S51) was obtained with AmpF‘STR1 Identifiler1. It is important to highlight that the success with AmpF‘STR1 MinifilerTM depended on the severity of the damage, since the

J.M. Robertson et al. / Forensic Science International: Genetics 12 (2014) 168–180

177

Fig. 6. Comparison of AmpF‘STR1 Identifiler1 profiles both with and without repair (using PreCRTM enzymes and the modified ATF/AFDIL protocol) to AmpF‘STR1 MinifilerTM profile without repair. The sample was a semen stain on white cotton exposed to the ambient conditions for 2 months on the roof of the laboratory building. The top panel depicts the partial profile (23 of 29 alleles among 16 loci) of AmpF‘STR1 Identifiler1 before repair. The middle panel shows the partially restored profile (4 alleles gained, 1 allele lost) of AmpF‘STR1 Identifiler1 after repair and the bottom panel shows a full profile (17 of 17 alleles among 9 loci) for AmpF‘STR1 MinifilerTM. The genotypes for the three samples are given in Table S3.

MinifilerTM kit did not yield full profiles for some of the samples tested (results not shown). When AmpF‘STR1 MinifilerTM was run with damaged DNA samples that had been treated with the PreCRTM enzymes, some improvement of the profiles was noted (results not shown). However, in most of these tests, PreCRTM enzymes led to failure of the AmpF‘STR1 MinifilerTM amplification, as noted by several off-ladder alleles at the shorter length loci (results not shown). Addition of PCRBoostTM may be an option to improve AmpF‘STR1 MinifilerTM results because it enhanced the MinifilerTM profile of the baseline reference sample prepared in the environmental chamber (n, 7; results not shown). Addition of BSA enhanced the MinifilerTM profile of templates damaged by hydrolysis (results not shown; n, 14). 3.4. Quality control issues related to repair of artificially degraded DNA When attempting to repair compromised DNA, it is important to have an assay for enzyme performance. Knowledge of enzyme activity was crucial, because the repair reaction on artificially degraded DNA did not work if the treatments were too harsh. The manufacturer provides a gel-based repair assay for damaged lDNA, and this was used on a regular basis to determine whether the enzyme mix was still active in repair. Five lots of the PreCRTM enzyme mix were used in the study. Negative control blanks run in each assay indicated no human DNA contamination in the manufacturer’s preparations. With 1 ng or less input DNA in the

repair reaction, only two instances of allele drop-in were observed during this study with artificially damaged DNA. No allele drop-in was observed for the DNA extracted from the body fluids. In a few reactions, there were several extraneous off-ladder peaks, indicative of a failed reaction. Split peaks were observed twice involving +A/A: once with bleach treated DNA and once with DNA heated in water at 80 8C. 4. Discussion The results of this study indicate that some types of damaged DNA may be repaired with the NEB PreCRTM repair kit. Lost alleles may be restored and STR peak heights may increase upon repair with this kit. However, the results were sporadic and sometimes deleterious. Increases in STR allele peak heights may be body fluid-, type of damage-, and frequency-dependent. For example, in several experiments with hydrolytically damaged DNA, there was an increase in the peak heights after repair with PreCRTM enzymes. In several other experiments, especially with oxidatively damaged DNA, the opposite trend was observed whereby peak heights often decreased upon incubation with the repair enzymes. The cause for this discrepancy may be related to the formation of double-strand breaks during the repair process. These may arise by the enzymes creating a single-strand break opposite one on the complimentary strand that was created during the damage treatment. The repair enzymes could introduce the single-strand break when an endonuclease and glycosylase excise a base lesion.

178

J.M. Robertson et al. / Forensic Science International: Genetics 12 (2014) 168–180

If the DNA damaged by oxidation had more such lesions than DNA damaged by other insults, it would be more susceptible to allele loss upon the PreCRTM enzyme treatment. The mechanism of profile restoration may include two different types of processes. The obvious process is excision of lesions and repair of any introduced gaps. This will result in the appearance of alleles after amplification of the repaired template, which had originally dropped out following the damage treatment. The less obvious process is enhancement of the PCR by the presence of proteins or reagents that optimize the molecular interactions, thus producing more amplicons and higher peak heights of all the alleles. Experiments that included PCR enhancers in the amplification reaction sometimes resulted in an increase in the number of the alleles from compromised samples. Incubation with the PreCRTM enzymes led to increased peak height of several of the shorter- and mid-length alleles. This observation leads to the hypothesis that the PreCRTM enzymes were not merely repairing lesions, but were also enhancing the PCR. The mechanism of allele enhancement by protein addition (BSA) may warrant further research, since this could provide a simple approach for improving typing success of compromised samples [17]. It is important to note however, that the addition of BSA was not always beneficial. It is possible that the sporadic results obtained when using the PreCRTM repair kit may be attributed to the amount of DNA used in the repair reaction, if it were too low for the repair reaction to perform optimally. In this work, optimization experiments were performed with 1–5 ng DNA. However, the actual amount of DNA for STR analysis may have been less. It is possible that the short probe used for DNA quantification may have over-represented the amount of quality DNA if the distance between the lesions was shorter than the distance required for a full Identifiler1 profile, but long enough to work in the Quantifiler1 Duo assay. From conversations with NEB scientists, it was learned that the PreCRTM kit was designed for samples containing larger amounts of DNA (up to 500 ng). Thus for the kit to work effectively, it is plausible to assume that the enzymes must be present in high concentrations. When the enzymes are in molar excess due to low DNA concentrations, it is possible that a minor enzymatic activity that causes a molecular aberration of the DNA may be amplified. This activity could arise from a contaminant of an enzyme preparation. The NEB standard protocol suggests a titration be performed in the range of 25–200 ng. Considering the yields obtained from forensically compromised samples generally are between 0.25 and 5 ng, the suggested titration is impractical and was not completed in this study. We did not test the effect of using higher amounts of DNA on repair. Thus, our evaluation of the PreCRTM repair kit may not be a true reflection of its performance, given the low DNA concentration of our samples and that the titration was not completed. Further experimentation might include testing whether the addition of non-human carrier DNA might prove beneficial to the repair process. In the 716 reactions with purified-DNA and 103 reactions with native-DNA, that encompasses all treatments for purified DNA, only two allele drop-ins were observed (D5S818 and D19S433). However, in some cases the profile had several off-ladder peaks. It was straight-forward to dismiss such peaks as allele drop-ins, as multiple peaks were present instead of one or two peaks, indicative of allele drop-in. We observed that when the repair reaction did not occur properly, chimerical-type products of all sizes were produced and appeared as multiple peaks in the subsequent profile. The cause for such an event is not clear, but multiple peaks were present in the profile more often when the DNA was highly degraded, as determined by high cycle threshold (Ct) values in the Quantifiler1 Duo assay. Multiple, off-ladder bands also occurred in some experiments with AmpF‘STR1 MinifilerTM on PreCRTM

treated templates, indicating that the multiple peaks were not introduced by the profiling kit. Paabo and colleagues [30] described a similar trend for aDNA and suggested that DNA damage promotes the jumping of enzymes between templates, resulting in chimeric products. As both experienced (n, 2) and novice (n, 4) researchers performed the experiments in this study and obtained positive results, it is clear that repairing damaged DNA using the methods outlined is straightforward and reproducible for a range of skill levels. Performing the repair reaction does not significantly alter the subsequent STR procedure. Thus it would be straight-forward to add a repair step into an automated procedure for STR amplification when multiple, degraded samples from a mass fatality incident need to be analyzed. When the sequential protocol was implemented, an incubation for as short as 20 min returned positive results for most damaged samples. However, for some more heavily damaged samples, an overnight incubation was beneficial. As it cannot be known what kind of damage is present or how severe the damage is for a given forensic sample, incubation for a minimum 20 min should be performed. This is the only recommended change in the manufacturer’s sequential protocol. However, a modification of the protocol (ATF/AFDIL protocol), optimized for STR analysis of repaired UVC-damaged DNA, has been reported [27]. The authors suggested the NEB buffer is incompatible with the efficiency of the AmpF‘STR1 Identifiler1 primers and proposed performing the repair in the AmpF‘STR1 Identifiler1 PCR buffer and adding the primers after the repair step. In this study, we further optimized the ATF/AFDIL protocol, by increasing the incubation time and adding additional dNTPs after the repair reaction (modified ATF/AFDIL protocol). Statistical analyses of repair and STR amplification performed by the sequential and modified ATF/AFDIL protocols on a purified-DNA subjected to heat, humidity, and UV in an environmental chamber indicated the modified ATF/AFDIL protocol yielded alleles with higher peak heights, in agreement with the original report for UVtreated samples [27]. However, a clear distinction between the two protocols was not seen when examining a variety of differently degraded samples, providing further evidence of the sporadic nature of the repair reactions. Our results demonstrate that the PreCRTM repair kit can be used to successfully restore dropped-out alleles. However more data needs to be obtained for both purifiedand native-DNA exposed to a variety of degradation regimes before a strategy can be recommended for the standardized repair of compromised samples. Difficulty was encountered when attempting to damage nativeDNA in whole cells. Whereas purified-DNA may react rapidly with oxidizing agents [31], several hours were required to induce sufficient damage to produce partial profiles from highly structured DNA in chromatin. This observation might warrant further investigation to determine the kinetics of allele drop-out from evidence collected after attempts to destroy it with bleach. In concordance with the study by Hall and Ballantyne [7], dried blood also required over 24 h of exposure to UVC irradiation to affect the STR profiles. However, additional studies should be completed with DNA exposed to natural sunlight, as the type of lesions produced by artificial UVC irradiation may be different [32]. We observed successful repair of damaged native-DNA extracted from blood less often than from semen and saliva. This observation may be associated with metal ions producing localized reactive species that can cause more lesions due to the proximity to the DNA. Exposing human cells to g-irradiation can produce lesions in native-DNA such as loss of bases, oxidized bases, and single- and double-strand breaks [33]. We observed that a low dose of ionizing radiation (8 Rads) yielded partial profiles and the PreCRTM enzymes could repair several of the aberrations, as we observed allele restoration. This result suggests fragmentation, common for

J.M. Robertson et al. / Forensic Science International: Genetics 12 (2014) 168–180

DNA damaged by ionizing radiation [5], was not a major product of the irradiation under our conditions. Alternatives to repair compromised DNA with PreCRTM enzymes were examined in this study. These included allele recovery with RestoraseTM and PCR enhancement with PCRBoostTM and BSA. In concordance with another study, we found that RestoraseTM may repair lesions almost as well as PreCRTM enzymes [15]. Enhancement of the STR profiles of artificially damaged DNA was observed with both PCRBoostTM and BSA in some cases. AmpF‘STR1 MinifilerTM performed well with some of the artificially degraded DNA studied in this investigation that did not work with AmpF‘STR1 Identifiler1. However, AmpF‘STR1 MinifilerTM did not work with all the damaged templates produced during this investigation. As shorter templates are required for the AmpF‘STR1 MinifilerTM, these templates are likely to have fewer lesions thus giving AmpF‘STR1 MinifilerTM an advantage with the compromised DNA [34]. It is important to highlight that the alternatives to repair examined were not tested as extensively as the PreCRTM enzymes, due to the scope of the project. However, the beneficial effects observed for the alternatives were positive and suggest that further research may be warranted. An alternative not tested here was the use of conventional STR kits with single nucleotide polymorphisms (SNPs) for genotyping of the degraded DNA [15,35]. As SNPs do not require long pristine templates, highly degraded DNA may still be able to provide SNP results that can be used for identifications. A novel aspect of this investigation is the use of a baseline sample for damaged DNA that can act as a reference in the repair reactions. This baseline sample was prepared under simulated environmental conditions that should result in multiple and disparate types of lesions, after exposure to UV irradiation, heat, and humidity. In this way, the DNA damage should reflect natural processes, as the lesions are produced from the treatments acting synergistically as opposed to individually. As a reference for DNA repair, the baseline sample can be used for testing new formulations of repair enzymes and in validation studies. The baseline sample can be run in each DNA repair reaction like DNA controls supplied in the STR reaction kits. The STR profiles for the baseline sample should be similar in successive experiments. Thus, repair of this damaged DNA should provide assurance that the repair reactions are performing properly.

5. Conclusion The NEB PreCRTM Repair Mix has been tested with artificially degraded purified-DNA and native-DNA from body-fluid specimens exposed to the environment, to determine whether it improves AmpF‘STR1 Identifiler1 STR profiles from such compromised samples. The STR profiles from degraded purified-DNA could be improved by the PreCRTM Repair Mix; however, the procedure was less successful with damaged native-DNA obtained from body fluids. For both types of damaged DNA, the success of repair depended on the type of damage and its severity. In a few experiments, the PreCRTM Repair Mix was compared with another repair mix, RestoraseTM, along with the PCR enhancers, PCRBoostTM and BSA and another kit for generating STR profiles, AmpF‘STR1 MinifilerTM. All of the tested systems and reagents did improve the STR profiles in some experiments, but the results were sporadic and in some instances the profiles were diminished or even lost after treatment. Based on this, it is not clear which approach would be best for compromised forensic evidence. The successful allele restoration for many of the damaged templates observed in our investigation is encouraging and possibly will generate interest and support for further research in this important area of forensic biology.

179

Acknowledgements This research was supported in part by an appointment to the Visiting Scientist Program at the Federal Bureau of Investigation Laboratory Division, administered by the Oak Ridge Institute for Science and Education, through an interagency agreement between the U.S. Department of Energy and the FBI. We thank Susanne E. Howlett for performing work on the alternative approaches to PreCRTM repair, Joseph Donfack for performing statistical analysis of the data of Supplementary Fig. S3, Robert Bull for arranging the g-irradiation experiments at the NBACC and Rhonda Craig for help planning the investigation. We are grateful to Mark Kavlick and Joseph Donfack for critically reviewing the manuscript and offering suggestions for its improvement, Jennifer Coleman for editorial assistance and Eugene Peters for support. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.fsigen.2014.05.011. References [1] R. Alaeddini, S.J. Walsh, A. Abbas, Forensic implications of genetic analysis from degraded DNA – a review, Forensic Sci. Int. Genet. 4 (2010) 148–157. [2] T. Lindahl, B. Nyberg, Rate of depurination of native deoxyribonucleic acid, Biochemistry 11 (1972) 3610–3618. [3] V.I. Bruskov, L.V. Malakhova, Z.K. Masalimov, A.V. Chernikov, Heat-induced formation of reactive oxygen species and 8-oxoguanine, a biomarker of damage to DNA, Nucleic Acids Res. 30 (2002) 1354–1363. [4] B.M. Kemp, D.G. Smith, Use of bleach to eliminate contaminating DNA from the surface of bones and teeth, Forensic Sci. Int. 154 (2005) 53–61. [5] E.C. Friedberg, G.C. Walker, W. Siede, R.D. Wood, R.A. Schultz, T. Ellenberger, DNA Repair and Mutagenesis, second ed., ASM Press, Washington, DC, 2006. [6] R.P. Sinha, D.-P. Hader, UV-induced damage and repair: a review, Photochem. Photobiol. Sci. 1 (2002) 225–236. [7] A. Hall, J. Ballantyne, Characterization of UVC-induced DNA damage in bloodstains: forensic implications, Anal. Bioanal. Chem. 389 (2004) 72–83. [8] A. Marrone, J. Ballantyne, Hydrolysis of DNA and its molecular components in the dry state, Forensic Sci. Int. Genet. 4 (2010) 168–177. [9] K.L. Thomson, S.D. Flemming, K. Barone, J.-A. Zieschang, A.M. Clark, The effect of repeated freezing and thawing on human sperm DNA fragmentation, Fertil. Steril. 93 (2010) 1147–1156. [10] S. Boiteux, E. Gajewski, J. Laval, M. Dizdaroglu, Substrate specificity of the Escherichia coli Fpg protein (formamidopyrimidine-DNA-glycosylase): excision of purine lesions in DNA produced by ionizing radiation or photosensitization, Biochemistry 31 (1992) 106–110. [11] T. Lindahl, Instability and decay of the primary structure of DNA, Nature 362 (1993) 709–715. [12] C.M. Pusch, I. Giddings, M. Scholz, Repair of degraded duplex DNA from prehistoric samples using Escherichia coli DNA polymerase I and T4 DNA ligase, Nucleic Acids Res. 26 (1998) 857–859. [13] K.L. Patchett, K.J. Fox, D.M. Burns, Recovery of genomic DNA from archived PCR product mixes for subsequent multiplex amplification and typing of additional loci: forensic significance for older unsolved criminal cases, J. Forensic Sci. 47 (2002) 786–796. [14] T. Lindahl, Suppression of spontaneous mutagenesis in human cells by DNA base excision-repair, Mutat. Res. 262 (2000) 129–135. [15] A.A. Westen, T. Sijen, Degraded DNA sample analysis using DNA repair enzymes, mini-STRs and (tri-allelic) SNPs, Forensic Sci. Int. Genet. Suppl. Ser. 2 (2009) 505– 507. [16] C.A. Kreader, Relief of amplification inhibition in PCR with bovine serum albumin or T4 gene 32 protein, Appl. Environ. Microbiol. 62 (1996) 1102–1106. [17] J.M. Robertson, J.B. Sgueglia, C.A. Badger, A.C. Juston, J. Ballantyne, Forensic applications of a rapid, sensitive, and precise multiplex analysis of the four short tandem repeat loci HUMVWF31/A, HUMTH01, HUMF13A1, and HUMFES/FPS, Electrophoresis 16 (1995) 1568–1576. [18] C. Luce, S. Montpetit, D. Gangitano, P. O’Donnell, Validation of the AmpF‘STR1 MinifilerTM PCR amplification kit for use in forensic casework, J. Forensic Sci. 54 (2009) 1046–1054. [19] H. Ide, K. Akamatsu, Y. Kimura, K. Michiue, K. Makino, A. Asaeda, Y. Takamori, K. Kubo, Synthesis and damage specificity of a novel probe for the detection of abasic sites in DNA, Biochemistry 32 (1993) 8276–8283. [20] J. Nelson, Repair of damaged DNA for forensic analysis, 2009, http:// www.ncjrs.gov/pdffiles1/nij/grants/227498.pdf (accessed 0.6.18.14). [21] M. Whiteman, A. Jenner, B. Halliwell, Hypochlorous acid-induced base modifications in isolated calf thymus DNA, Chem. Res. Toxicol. 10 (1997) 1240–1246.

180

J.M. Robertson et al. / Forensic Science International: Genetics 12 (2014) 168–180

[22] A. Hussain, P. Trudell, A.J. Repta, Quantitative spectrophotometric methods for determination of sodium hypochlorite in aqueous solutions, J. Pharm. Sci. 59 (1970) 1168–1170. [23] Biological agent decontamination technology testing, http://cfpub.epa.gov/si/ si_public_record_report.cfm?dirEntryId=225844 (accessed 06.18.14). [24] J. Roskams, L. Rodgers, Lab Ref A Handbook of Recipes, Reagents, and Other Reference Tools for Use at the Bench, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY, 2002. [25] F.S. Baechtel, K.W. Presley, J.B. Smerick, D1S80 typing of DNA from simulated forensic specimens, J. Forensic Sci. 40 (1995) 536–545. [26] C.T. Comey, B.W. Koons, K.W. Presley, J.B. Smerick, C.A. Sobieralski, D.N. Stanley, F.S. Baechtel, DNA extraction strategies for amplified fragment length polymorphism analysis, J. Forensic Sci. 39 (1994) 1254–1269. [27] T.M. Diegoli, M. Farr, C. Cromartie, M.D. Coble, T.W. Bille, An optimized protocol for forensic application of the PreCRTM Repair mix to multiplex STR amplification of UV-damaged DNA, Forensic Sci. Int. Genet. 6 (2012) 498–503. [28] K.W. Hipel, A.I. McLeod, Time Series Modeling of Water Resources and Environmental Systems. Available from: http://www.stats.uwo.ca/faculty/aim/1994Book/ (accessed 0.6.18.14). [29] H.B. Mann, Nonparametric tests against trend, Econometrica 13 (1945) 245–259.

[30] S. Paabo, D.M. Irwin, A.C. Wilson, DNA damage promotes jumping between templates during enzymatic amplification, J. Biol. Chem. 265 (1990) 4718– 4721. [31] H. Hayatsu, S.-K. Pan, T. Ukita, Reaction of sodium hypochlorite with nucleic acids and their constituents, Chem. Pharm. Bull. 19 (1971) 2189–2192. [32] J.-H. Yoon, C.-S. Lee, T.R. O’Connor, A. Yasui, G.P. Pfeifer, The DNA damage spectrum produced by simulated sunlight, J. Mol. Biol. 299 (2000) 681–693. [33] B.M. Sutherland, P.V. Bennett, O. Sidorkina, J. Laval, Clustered damages and total lesions induced by ionizing radiation: oxidized bases and strand breaks, Biochemistry 39 (2000) 8026–8031. [34] C.S. Oh, S.J. Lee, J.B. Park, S.D. Lee, S.B. Seo, H.Y. Kim, J. Kim, Y.-S. Kim, D.H. Shin, Autosomal short tandem repeat analysis of ancient DNA by coupled use of miniand conventional STR kits, J. Forensic Sci. 57 (2012) 1556–4029. [35] L.A. Dixon, A.E. Dobbins, H.K. Pulker, J.M. Butler, P.M. Vallone, M.D. Coble, W. Parson, B. Berger, P. Grubwieser, H.S. Mogensen, N. Morling, K. Nielsen, J.J. Sanchez, E. Petkovski, A. Carracedo, P. Sanchez-Diz, E. Ramons-Luis, M. Brio¯n, J.A. Irwin, R.S. Just, O. Loreille, T.J. Parsons, D. Syndercombe-Court, H. Schmitter, B. Stradmann-Bellinghausen, K. Bender, P. Gill, Analysis of artificially degraded DNA using STRs and SNPs-results of a collaborative European (EDNAP) exercise, Forensic Sci. Int. Genet. 164 (2006) 33–44.