Forensic Science International: Genetics 11 (2014) 96–104
Contents lists available at ScienceDirect
Forensic Science International: Genetics journal homepage: www.elsevier.com/locate/fsig
Direct qPCR quantification of unprocessed forensic casework samples Jason Yingjie Liu * Human Identification, Thermo Fisher Scientific, 180 Oyster Point Boulevard, South San Francisco, CA 94080, USA
A R T I C L E I N F O
A B S T R A C T
Article history: Received 23 November 2013 Received in revised form 16 February 2014 Accepted 2 March 2014
The current short tandem repeat (STR) typing workflow for forensic casework samples involves sample collection, sample screening, DNA extraction, DNA qPCR quantification and STR amplification. Although very effective and powerful, this workflow still has room for improvements. For example, the screening assays in practice do not provide DNA related information and also do not work with touch DNA samples, which make up of the majority of the property crime samples. It is known that not all DNA samples have equal probative values. Considering the DNA backlog situation crime laboratories face today, an effective screening tool would be highly desirable. It would allow forensic scientists to prioritize the DNA samples so that the limited resources would be first spent on samples that would have better chances of producing informative STR profiles. qPCR assay does provide DNA quantity and gender information and would be an ideal screening tool. However, prior to quantification, sample extraction and purification are required. By the time a DNA sample is ready for qPCR quantification, time and resources have already been spent on samples that should have been given low priority or excluded from further processing if DNA quantity and gender information were known. To overcome this problem, a direct quantification technology is developed to allow qPCR quantification of casework samples without the need for DNA extraction and purification. The key to a direct qPCR assay is the PE-Swab, a novel sample collection device. A small sample punch can be generated from a PE-Swab and placed in a qPCR reaction for quantification. After optimizing the punch size and the quantification software baseline setting, accurate DNA quantification can be obtained from a sample without the need to carry out DNA extraction and purification. Proof of concept studies were done with low lever touch samples as well as blood samples. The PE-Swab also allows direct STR amplification of casework samples without the need for DNA extraction. Besides its potential as a screening tool, the direct qPCR assay can also be used to normalize the DNA input for a direct STR amplification reaction. The feasibility of the direct qPCR/direct STR amplification workflow was demonstrated with touch DNA samples and blood stain samples. ß 2014 Elsevier Ireland Ltd. All rights reserved.
Keywords: Direct qPCR Direct quantification Direct real-time PCR Direct STR amplification PE-Swab DNA extraction
1. Introduction The demand for DNA analysis has been driven up by the increased awareness of the power of DNA technology for solving crimes. Law enforcement agencies, which used to only collect DNA evidence in violent crimes such as homicide and sexual assault, have also begun to collect DNA evidence from property crime cases to aide their investigations. However, biological evidence samples from property crimes are mostly touch DNA samples. Unlike blood or saliva stains, touch DNA is not always identifiable by the naked eye. In many cases law enforcement officers swab surfaces which they believe have been touched by a perpetrator at crime scenes. Or the officers collect evidentiary items which they believe the
* Tel.: +1 650 872 7283; fax: +1 650 266 3063. E-mail addresses: Jason.Liu@thermofisher.com, [email protected] http://dx.doi.org/10.1016/j.fsigen.2014.03.003 1872-4973/ß 2014 Elsevier Ireland Ltd. All rights reserved.
perpetrator had physical contact with. Therefore, it is understandable that some of the touch DNA samples collected this way may not contain any DNA at all. Even when a touch DNA sample from perpetrator is collected, there is no guarantee it will contain enough DNA to obtain a probative STR profile. This is because the amount of DNA left behind on a touched surface depends on the surface property of the item, the length and pressure of the contact, and the shedding characteristic of the individual [3]. Although very powerful, the short tandem repeat (STR) DNA profiling technology is also expensive, time consuming and labor intense. The current lack of an effective screening tool and the increasing number of casework samples collected and submitted, especially those from property crimes, have made it impossible to process all of the samples in a timely manner. As a result, DNA backlog has been created. Some laboratories also limit the number and the type of samples that can be submitted in each case. Law enforcement officers have to make difficult decisions on which
J.Y. Liu / Forensic Science International: Genetics 11 (2014) 96–104
DNA samples to submit, and this may leave more probative samples left un-submitted. Therefore, a simple, fast, and inexpensive screening tool capable of measuring DNA quantity and the sample gender is highly desired. Fast and inexpensive presumptive tests do exist for blood and saliva, but they do not provide DNA quality, quantity and gender information. The LGC1 Forensics ParaDNA1 Screening System is the only forensic DNA sample screening tool available that takes an unprocessed forensic sample as an input and also provides gender and DNA related information in just 75 min. This technology is based on the high resolution melting curve analysis of STR loci D16S539 and TH01 and gender locus Amelogenin [4]. The assay produces the gender of the sample’s contributor as well as a percentage score correlated with the chance of obtaining a successful STR profile. The ParaDNA1 Screening System1 has a detection limit of 62.5 pg of DNA, equivalent to approximately 10 human cells. The lowest DNA level tested where repeatable results were obtained was 250 pg. This is equivalent to approximately 40 cells. A positive gender indication was obtained in 80% of samples with at least 250 pg of input DNA. This allows law enforcement and forensic professionals to better prioritize the samples before deciding on which samples should be submitted for DNA analysis. The high resolution melting curve analysis has also been investigated for the feasibility of allowing rapid interrogation of a small panel of STR loci outside of specialized laboratories [5,6], but further work is required to reduce test complexity (e.g. through multiplexing) and to ensure that assays are sufficiently accurate and robust for forensic analysis [5,6]. Although qPCR quantification assay was originally included in the STR typing workflow to ensure the optimal amount of DNA is used in a multiplex STR reaction, forensic scientists realized that the extreme sensitivity of the qPCR assay potentially allows them to identify DNA samples with quantities below STR reaction sensitivity and therefore to forego analysis of samples less likely to amplify, saving both time and money. Several laboratories have investigated the current commercially available qPCR assays for such purpose, but they were not able to find a cutoff quantity value below which no STR profile can be obtained [2,7,8]. The reasons for why some current qPCR assays are not able to serve as STR typing gatekeeper are thoroughly explained by John Butler in his book [1]. In order for a laboratory to determine that an evidentiary sample processing is to be terminated after DNA quantitation, the laboratory shall have a validation study to support that determination. To enhance the utility of a qPCR assay as a more effective screening tool, the most advanced commercial qPCR assays also provide information on the gender, the male/female DNA quantity ratio, the degree of inhibition and the degree of DNA degradation. However, in all current qPCR assay formats, forensic casework DNA samples need to be extracted and purified first before they can be quantified. The DNA sample input also must be in the form of an aliquot. These constraints greatly diminish the value of a qPCR assay as a screening tool, because by the time the DNA sample is ready for qPCR quantification, significant amount of time and resources have been spent on extracting and purifying samples that have very little to no probative value. In this paper we will describe the proof of concept studies of a new forensic DNA sample screening technique called direct qPCR quantification. This technique allows the DNA quantity and the sample gender information to be obtained before the sample is processed for DNA extraction and purification. Such a screening tool would allow forensic scientists to screen and rank DNA samples in a forensic case based on the DNA quantity, the gender information and the case context. Forensic scientists then have the option of processing samples that have higher chances of yielding 1
For Forensic and Human Identity Applications Only.
97
probative information first, and only processing lower ranked samples when necessary. The critical enabling element of the direct qPCR workflow is the PE-Swab, a novel sample collection device. The PE-Swab allows for forensic sample collection and subsequent generation of a very small paper punch for direct qPCR quantification. Samples collected on a PE-Swab can also be directly STR amplified. The feasibility of using the direct qPCR results to normalize the DNA sample input for the direct STR amplification is also demonstrated. Because there is no DNA loss associated with DNA extraction process, the direct quantification/direct PCR workflow is simpler, faster, and would be an attractive alternative to the current DNA extraction and purification based STR typing workflow.
2. Materials and methods 2.1. PE-Swab A PE-Swab consists of three components: a filter paper stripe, a holder and a clip. A PE-Swab was assembled by wrapping a filter paper stripe around the holder and then securing the paper stripe on the holder using the clip at the end of the PE-Swab handle. An example of a functional PE-Swab is shown in Fig. 1A. The prefix before a PE-Swab indicates the height of the active sampling area, which is 5 mm in this example. The width of the active sampling area is defined by the angled fold of the holder. The PE-Swab is 40 mm long measured from the active sampling area to the end of the swab handle, which is 15 mm wide. PE-Swabs were used to swab objects of interest that contain touch DNA samples or blood samples. When a swabbing liquid was used, it was applied to the object of interest before swabbing. After swabbing, the filter paper stripe was detached from the swab holder and was air dried before punching. Using a Harris Uni-CoreTM punch (Ted Pella, Inc., Redding, CA), punches of the desired size were generated from the active sampling area of the filter paper for a direct qPCR assay or a direct STR amplification. 2.2. Touch DNA sample collection and swabbing using a 5 mm PESwab Touch DNA samples were collected on transparency films. In order to increase the amount of DNA present on their fingers, the donors were asked to touch their face first before pressing their fingers on a piece of transparency film to collect the DNA material. The touch samples were stored in a paper envelope to prevent contamination before being tested. 10 mL of ethanol was applied to each touch sample before being swabbed using a 5 mm PE-Swab. 2.3. DNA extraction and purification from PE-Swabs After a 0.5 mm punch was removed from a 5 mm PE-Swab for direct qPCR. The rest of the active sampling area of the 5 mm PESwab was punched out with a 2.5 mm diameter puncher and the punches were transferred into a PrepFiler1 LySepTM columns/ sample tube assembly (Thermo Fisher Scientific, Waltham, MA). 500 mL of PrepFiler1 lysis buffer (Thermo Fisher Scientific) was added to the LySepTM column and incubated on a thermal mixer at 70 8C for 40 min. After lysis, the lysate was transferred into the sample tube by centrifugation at 16,000 g for 2 min. The LySepTM column containing punches was disposed after centrifugation. Sample tubes containing lysate were loaded into an AutoMate ExpressTM Forensic DNA Extraction System.2 DNA from the touch 2 For Research, Forensic or Paternity Use Only. Not for use in diagnostic procedures.
J.Y. Liu / Forensic Science International: Genetics 11 (2014) 96–104
98
Fig. 1. (A) A 5 mm PE-Swab. The dimension (prefix) before a PE-Swab indicates the height of the filter paper stripe at the active sampling area. (B) The three components of a disassembled 5 mm PE-Swab with two 2 mm punches generated from the active sampling area of the PE-Swab using a Harris Uni-CoreTM punch.
sample lysate was extracted and purified using PrepFiler1 standard protocol. 2.4. Dry blood stain samples preparation and swabbing using a PESwab Dry blood stain samples were prepared by pipetting 20 mL blood from a female donor onto five different types of substrates. The liquid blood was spread out with a pipette tip and was set out to dry for five days. Before swabbing, 40 mL de-ionized water was applied to each dry blood stain. Because the amount of dry blood on the substrate is much higher than on the touch DNA samples, a larger 20 mm PE-Swab was used to swab each dry blood stain. The dry blood stain samples used in the sample uniformity study were prepared by pipetting 10 mL blood from five different male donors onto five separate pieces of microscope slide. The liquid blood was spread out with a pipette tip and allowed to dry over the weekend. Before swabbing using a 5 mm PE-Swab, 10 mL de-ionized water was applied to each dry blood stain. After swabbing, the filter paper stripe was detached from the swab holder and was air dried before punching. 2.5. Real time qPCR quantification of forensic DNA sample collected on a PE-Swab 1
Quantifiler Duo DNA Quantification kit (Thermo Fisher Scientific) is used in the direct qPCR assay. To prepare a direct qPCR reaction plate, a punch generated from the PE-Swab was placed directly into a well of a MicroAmp1 Optical 96-Well Reaction Plate. 25 mL qPCR reaction mix (10.5 mL Quantifiler1 Duo Primer Mix and 12.5 mL Quantifiler1 Duo PCR reaction Mix and 2 mL de-ionized water) was then added to the well containing the sample punch. Quantification reactions were carried out on the Applied Biosystems 7500 Real-Time PCR System3 (Thermo Fisher Scientific) using the manufacture’s recommended protocol. The quantification results were analyzed using SDS Software v2.0.6 (Thermo Fisher Scientific). 2.6. STR amplification and CE analysis The conventional post DNA extraction STR amplification was carried out using GlobalFilerTM PCR Amplification Kit (Thermo Fisher Scientific) by following the manufacture’s recommended protocol. The procedure for direct STR amplification was described below. Based on the direct qPCR result, the desirable input DNA quantity was obtained for each STR reaction by adjusting the punch size and the number of punches generated from a PE-Swab
and released into a well of a MicroAmp1 Optical 96-Well Reaction Plate. 7 mL PCR reaction mix (2.1 mL of GlobalFiler Master Mix,4 3.5 mL of Identifiler1 direct Primer Mix and 1.4 mL of water) was added to each well containing the punches. The reason the Identifiler1 direct Primer Mix was used in place of the GlobalFilerTM Primer Mix is that the GlobalFilerTM Primer Mix is not available to the author when experiments on direct STR amplification were conducted. The thermal cycling conditions were 95 8C/1 m, 28, 29 or 30 cycles of (94 8C/10 s, 59 8C/90 s), 60 8C/10 m and 4 8C – hold. The exact thermal cycle number was specified in each individual experiment described in the results and discussion sections. After thermal cycling, 1 mL PCR product from each sample was mixed with 9 mL GeneScan1500 size standard and deionized formamide. The CE was run on an ABI 3130xl capillary electrophoresis instrument under the following conditions: Oven: 60 8C, Prerun: 15 kV, 180 s, Injection: 3 kV, 10 s, Run: 15 kV, 1500 s, Capillary length: 36 cm, Separation polymer: POP4TM polymer and Dye set: G5. The resulting STR electropherograms were analyzed using the GeneMapper1 ID-X software5 (Thermo Fisher Scientific). 3. Results and discussion 3.1. The effect of the punch size and the baseline setting on qPCR quantitation Because illuminating the sample with light and detecting the fluorescence signal is at the heart of a qPCR assay, it was expected that the presence of a filter paper punch in the reaction well would have an impact on the qPCR assay. This impact becomes clear when one examines the multicomponent plots of the non-template control (NTC) reaction with or without a filter paper punch of different size in the reaction (see Fig. 2). Except for the ROX channel, the passive reference, the presence of a filter paper punch in the reaction well results in elevated background florescent signal in FAMTM (Male target), VIC1 (human target) and NEDTM (IPC) channels and the magnitude of the background elevation is positively correlated with the size of the filter paper punch in the reaction well (see Fig. 2). In addition, the background florescent signal slowly increases after each thermal cycle and the rate of this background fluorescent signal increase is also positively correlated to the size of the filter paper punch (see Fig. 2). It was noted that the elevated baseline and the slope of the baseline affect the Ct value determination, which in turn affects the measured DNA quantity. To determine the Ct value, the SDS software first determines the Rn (the normalized florescent signal) 4
3
When used for purposes other than Human Identification the instruments cited are for Research Use Only. Not for use in diagnostic procedures.
For Forensic or Paternity Use Only. For Research, Forensic, Paternity and Cell Line Authentication. Not for use in Diagnostic and Therapeutic applications. 5
J.Y. Liu / Forensic Science International: Genetics 11 (2014) 96–104
99
Fig. 2. The florescent signals collected from different dye channels during a qPCR reaction without or with a filter paper punch of indicated size in a NTC reaction well. The optimal baseline starting and ending cycle numbers for Ct determination are 15 and 25 for IPC (NEDTM channel).
by dividing the florescent signal in each dye channel by the florescent signal of the passive reference (ROX). The SDS software then uses the Rn values collected from a predefined range of PCR cycles to serve as the baseline. After generating a baselinesubtracted amplification plot of DRn versus cycle number, an algorithm defines the cycle number at which the DRn value crosses the threshold setting as the threshold cycle (Ct). The effect of the paper punch size and the baseline start point/end point setting on IPC Ct value is shown in Fig. 3. Regardless of the baseline setting, the presence of a paper punch in the reaction well causes the IPC Ct value to drift higher. The size of the Ct drift is positively correlated to the size of the paper punch. Therefore, using a small paper punch is essential to minimize the impact of a paper punch on the qPCR assay accuracy. The baseline setting also has impact on the IPC Ct value. After experimenting with different baseline settings, it was discovered that when a paper punch is present in the reaction well, the optimal baseline starting cycle number is 15, which is the highest baseline starting cycle number permitted by the current SDS software. The optimal baseline ending cycle number is right before where the florescent signal starts to break out and rise up above baseline, which is cycle 25 in this case. Compared to Ct values calculated with the baseline setting of 1–25 (the Box plot of crimson color in Fig. 3), the Ct value is less deviated from the true value when the baseline setting is at 15–25, regardless of the punch size (the Box plot of green color in Fig. 3). By using a 0.5 mm
paper punch and the optimal baseline setting in data analysis, we are able to reduce the IPC Ct drift to about 1%. Because the presence of a paper punch in a qPCR reaction causes Ct value to drift higher and the DNA quantity of the sample is determined based on the Ct value of the sample and those of DNA standards, the under estimation of the DNA quantity in the sample will occur when a paper punch is present in the reaction well. However, the DNA quantity under estimation is minimized by the use of the 0.5 mm punch and the optimal baseline setting. To understand this effect more quantitatively, a 0.5 mm paper punch was placed in a reaction well containing a known amount of human DNA standard from Quantifiler1 Duo kit. Six reactions were carried out at each DNA input quantity. The optimal baseline setting was set at 15 and 19 for the Quantifiler1 Duo human and male targets and 15 and 25 for IPC target. The measured DNA quantity against the input DNA quantity was shown in Table 1. As expected, with the presence of a paper punch in the reaction well, the DNA quantity is under estimated for both human and male targets with the DNA input quantity down to 0.42 ng, but the percentage difference is very small. With the DNA input down to 0.42 ng, the percentage difference between the input quantity and the measured quantity is less than 13.7%. Even with Std 7 and 8, the lower end of the DNA quantity standard range, the percentage difference between the input quantity and the measured quantity is still less than 34.5%. The DNA quantity overestimation observed at 0.14 ng and 0.05 ng input level is likely due to the stochastic
J.Y. Liu / Forensic Science International: Genetics 11 (2014) 96–104
100
Fig. 3. The effect of paper punch size and baseline start point/end point setting on IPC Ct value determination. Three replicates for NTC reaction containing a punch of size 0.5 mm, 1 mm or 2 mm. Four replicates for NTC reaction containing no punch. (For interpretation of the references to color in the text, the reader is referred to the web version of the article.)
effect typical of low copy number samples. This result clearly demonstrated the robustness and the accuracy of the qPCR DNA quantification with the presence of a 0.5 mm paper punch. 3.2. Direct qPCR quantification of unprocessed touch DNA samples To demonstrate the feasibility of direct quantification of unprocessed touch DNA samples using a qPCR assay, touch DNA samples from five male and two female donors (total 19 samples) were quantified. The two female donors were known to be low shedders and their samples were included in this study to challenge the detection sensitivity of the qPCR and the conventional post DNA extraction STR typing workflow. The measured human and human male DNA quantities per punch and the corresponding IPC Ct values are shown in Fig. 4. No human or Y DNA was detected in any of the four female samples. Among the 15 male samples (three from each of the five male donors), both human and Y DNA were detected in 13 samples. The two exceptions were sample Male_B_S1 (only Human DNA detected, 41 pg) and Male_D_S3 (only Y DNA detected, 32.5 pg). This could be due to the fact that the quantifiler1 Duo human and quantifiler1 Duo Y targets are located on different chromosomes and there is very little DNA in the sample as shown by the low detected DNA quantity. The failed detection of both targets is likely due to the stochastic effect which is common for low copy number samples.
Other than the samples from donor Male_E, the IPC Ct values for all samples were within 0.5 Ct of the punch NTC (29.1, average; 0.19, standard deviation) indicating minimal inhibition even with unprocessed sample. Interestingly, the IPC Ct values from all three samples of donor Male_E (27.7, average; 0.09, standard deviation) are more than one Ct lower than the rest of the samples (29.1, average; 0.14, standard deviation). As expected, after DNA extraction and purification using AutoMate ExpressTM, the IPC Ct discrepancy between donor Male_E and rest of the samples disappeared. (The average IPC Ct for all purified samples is 28.6 with standard deviation of 0.11.) The lowered IPC Ct values in unprocessed Male_E samples suggest that some composition in these samples actually enhances qPCR efficiency. The exact reasons are not clear at this point. Nevertheless, this result clearly demonstrated the feasibility of direct qPCR quantification of unprocessed touch DNA samples. 3.3. The correlation between direct qPCR quantification, post extraction quantification and post extraction STR typing One foreseeable application of a direct qPCR assay is the screening and prioritization of a large number of touch DNA samples from property crimes. Such capability would allow limited forensic resources to be spent on processing samples that have the best chance of producing probative information. In order for this approach to be effective one needs to show that the direct
Table 1 The comparison of measured DNA quantity to that of input DNA quantity with the presence of a 0.5 mm filter paper punch in each qPCR reaction.
Std Std Std Std Std Std Std Std a
1 2 3 4 5 6 7 8
STD DNA input (ng)
Duo human (ng)a
Duo human STDEV
Duo male (ng)a
Duo male STDEV
% diff human
% diff male
100.00 33.40 11.12 3.70 1.24 0.42 0.14 0.05
97.38 31.19 11.13 3.59 1.15 0.38 0.14 0.06
4.62 3.16 0.51 0.18 0.03 0.04 0.01 0.01
95.74 30.73 10.58 3.49 1.10 0.36 0.09 0.04
3.29 1.96 0.38 0.16 0.06 0.04 0.01 0.01
2.62% 6.61% +0.13% 2.91% 6.87% 10.53% 3.62% +28.55%
4.26% 7.98% 4.84% 5.64% 11.41% 13.70% +34.48% 5.99%
Average of 6 replicates except Std 3 and Std 8, which are average of 5 replicates.
J.Y. Liu / Forensic Science International: Genetics 11 (2014) 96–104
101
Fig. 4. Direct qPCR quantification of unprocessed touch DNA samples. The touch DNA was introduced into a qPCR reaction in the form of 0.5 mm paper generated from the active sampling area of a 5 mm PE-Swab.
quantification result is correlated to the post extraction quantification as well as to the STR typing result. After taking a 0.5 mm punch for direct qPCR, the remaining sample from all 19 touch samples in the previous section were further extracted and purified using an AutoMate ExpressTM Forensic DNA Extraction System. 2 mL purified DNA was quantified on the same plate for the direct qPCR in the previous section to reduce variation. Based on the post extraction quantification result, 15 mL purified DNA from each of the 19 samples was amplified using GlobalFilerTM PCR kit in a 25 mL reaction for 29 cycles. The quantification results from the direct qPCR, the post extraction qPCR and the percentage alleles recovered from STR typing were plotted in Fig. 5. Since the sample input for the direct qPCR is a 0.5 mm solid punch and that for the post extraction qPCR is 2 mL aliquot, the DNA quantity in the unit of pg/well were plotted so that both the direct qPCR result and the
post extraction qPCR result can be represented by the same Y axis. The samples were ranked from the highest to the lowest based on the direct qPCR human quantity (the solid blue line). The trend of the human quantity from the post extraction qPCR (the dotted red line) follows that of the direct qPCR. The fluctuation of the post extraction qPCR line is likely caused by the non-uniformity of the sample on a PE-Swab and the stochastic effect from low copy number samples. 15 mL purified DNA from each of the 19 touch samples were amplified using a GlobalFilerTM PCR kit. Other than sample Male_C_S2, all the other samples have input DNA quantity less than the target 1 ng. As shown in Fig. 5, the direct quantification results also correlated with the percentage alleles recovered (the solid green line) from the STR analysis. The top five ranked samples on average have 98% of the alleles recovered, while the bottom five ranked samples on average only have 8% of the
Fig. 5. The correlation between direct quantification, post extraction quantification and % alleles recovered from post extraction STR. STR PCR input volume is 15 mL for all samples. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of the article.)
J.Y. Liu / Forensic Science International: Genetics 11 (2014) 96–104
102
Table 2 Sample uniformity characterization of low level touch DNA samples and blood stain samples collected on a 5 mm PE-Swab. Male 1
Touch sample (n = 4) Individual DNA quantity measurement (pg/punch) Average DNA quantity (pg/punch) STDEV CV Blood sample (n = 4) Individual DNA quantity measurement (ng/punch) Average DNA quantity (ng/punch) STDEV CV
Male 2
Male 3
Male 4
Male 5
Human
Y
Human
Y
Human
Y
Human
Y
Human
Y
3.9, 11.2, 8.2, 4.9 7.0
10.1, 12.6, 54.8, 33.5 27.7
35.2, 147.3, 105.2, 129.4 104.3
33.5, 293.1, 159.5, 178.2 166.1
148.5, 173.2, 62.4, 73.0 114.3
213.1, 136.1, 131.2, 28.9, 127.3
16.8, 3.7, 15.9, 35.0 17.8
15.6, 40.5, 34.7, 117.1 52.0
24.3, 28.5, 39.6 18.4 27.7
9.3, 52.0, 39.4, 42.5 35.8
2.8 40%
18.1 65%
42.6 41%
92.1 55%
47.5 42%
65.4 51%
11.1 63%
38.7 74%
7.7 28%
16.0 45%
0.84, 1.97, 2.24, 2.46 1.88
2.27, 3.48, 6.04, 4.55 4.09
0.43, 0.49, 0.82, 0.50 0.56
1.15, 1.31, 2.09, 1.17 1.43
0.78, 1.09, 1.45, 1.57 1.22
1.78, 2.97, 3.76, 4.50 3.25
1.02, 1.57, 1.02, 1.06 1.17
2.14, 3.00, 2.01, 2.58 2.43
0.54, 0.71, 0.49, 0.53 0.57
1.24, 1.88, 1.43, 1.95 1.63
0.72 38%
1.60 39%
0.18 32%
0.45 31%
0.36 29%
1.17 36%
0.27 23%
0.45 18%
0.10 17%
0.34 21%
alleles recovered. This result clearly demonstrated that the direct qPCR with the Quantifiler1 Duo kit is effective at prioritizing a large number of samples. However, the direct qPCR with Quantifiler1 Duo kit should not be used as a gatekeeper for STR analysis. Although the bottom five ranked samples have direct qPCR quantification of zero, only one did not produce any accurate allele. Because of the possibility of false negative with low level DNA samples when a Quantifiler1 Duo kit is used for direct qPCR, one should not discontinue processing of samples, which may prove pivotal in a case solely due to the low or undetected direct quantitation data. 3.4. Sample uniformity on a PE-Swab Since sample is collected along the active sampling area of a PESwab, a certain degree of sample variation on a PE-Swab is expected. The correlation study results between the direct qPCR quantification, the post extraction quantification and the post extraction STR typing described above also indicated that such sample variation exists. To characterize the sample distribution on a PE-Swab directly, five touch samples and five blood stain samples were collected using 5 mm PE-Swabs. Four 0.5 mm punches were generated from each PE-Swab and directly quantified using a Quantifiler1 Duo kit. The individual quantity of each punch, the average quantity of the four punches, the standard deviation and the coefficient of variation (CV) from each sample are listed in Table 2. The result confirms that sample variation does exist for samples collected on a PE-Swab. Blood samples (average CV of 27.8%) are more evenly distributed on a PE-Swab than the low level touch samples (average CV of 42.8%). With low level touch samples, the CV for the Y quantity is larger than that for the human quantity for all five samples. We also examined the IPC Ct values for both low level touch DNA samples and blood samples. The average IPC Ct from all 20 touch samples is 28.7 (Standard deviation of 0.17), which is only 0.1 Ct higher than the IPC Ct for NTC. As expected, inhibition was detected with all blood samples due to the high concentration of heme in blood. The average IPC Ct from all 20 blood sample is 30.2 (Standard deviation of 0.3), which is 1.9 Ct higher than and the IPC Ct (28.3) for NTC. It is also worthwhile to point out that correct gender calls were made with all 40 samples in this study.
for the conventional DNA extraction based STR typing workflow, the quantification result from a direct qPCR assay could also be used to normalize the DNA input for a direct STR amplification. One could envision the possibility of a simpler STR typing workflow, in which the DNA extraction is eliminated. To demonstrate the direct qPCR/direct STR workflow, touch DNA samples from two female and two male donors were collected on 5 mm PE-Swabs and quantified directly using a Quantifiler1 Duo kit. The measured human and human male DNA quantities along with IPC Ct are shown in Fig. 6. The IPC Ct for all four samples is within 0.5 Ct of NTC IPC Ct. No human male DNA was detected in sample Male 2, which is likely due to the stochastic effect which is common for low copy number samples. Since there is no DNA extraction, the DNA input to a direct amplification reaction has to be in the form of paper punch from a PE-Swab. The design of the 5 mm PE-Swab is such that the touch DNA sample is collected and concentrated on the 1 mm 5 mm stripe of the angled fold area of the PE-Swab. Based on the DNA quantity on each 0.5 mm punch, the optimal input DNA for a direct STR amplification reaction can be obtained by controlling the punch size and the number of punches. Because of the small size of the punches, a smaller PCR reaction volume can be used to increase STR sensitivity, which is critical for low level touch DNA samples. Instead of the standard 25 mL, 7 mL PCR reaction was used in this study and it can accommodate up to eight 1 mm paper punches. After removing a 0.5 mm punch, four 1 mm punches can be practically generated from the remaining active sampling area of a
3.5. Direct qPCR/direct STR PCR of touch DNA samples One additional benefit of collecting DNA sample on a PE-Swab is that the small punch from a PE-Swab is also compatible with direct STR amplification. Therefore, besides its utility as a screening tool
Fig. 6. Direct qPCR quantification of unprocessed touch DNA samples. The touch DNA was introduced into qPCR reaction in the form of 0.5 mm paper generated from the active sampling area of a 5 mm PE-Swab.
J.Y. Liu / Forensic Science International: Genetics 11 (2014) 96–104
Fig. 7. The correlation between DNA input and corresponding average peak height of the STR profile obtained from direct amplification of punches from the same PESwab. Thermal cycle number: 30.
5 mm PE-Swab and all four punches can be used as the input for a direct amplification reaction if needed. The recommended PCR reaction conditions for the GlobalFilerTM PCR kit is 1 ng input DNA in a 25 mL PCR reaction with 29 cycles, which translates to 280 pg input DNA in a 7 mL reaction with 29 cycles. Based on the direct quantification results, the total available amount of DNA on a 5 mm PE-Swab for samples female 1 and male 2 is 208 pg (estimated based on the size of the active sampling area of a 5 mm PE-Swab) and therefore all four 1 mm punches were used in the PCR reaction. For female 2 and male 1, one 0.5 mm punch was used in the STR reaction, which corresponds to 386 pg and 56 pg DNA respectively. More DNA from male 1 could have been used to reach the desired target input of 280 pg. Instead, only one punch was used so that the STR profile quality can be assessed when the suboptimal DNA input is used. The PCR reactions were carried out for 30 cycles. Full STR profiles were obtained for all four touch DNA samples. Considering the existence of the sample variation on a PE-Swab, the PCR input DNA quantity is correlated reasonably well with the average peak height of the STR results (see Fig. 7). Off scale peaks were observed with the profile in female 2, which is in agreement with the amount of input DNA and the fact that PCR was run for an additional cycle.
103
Fig. 8. Direct qPCR quantification of unprocessed blood stains collected on a 20 mm PE-Swab. The blood stains were introduced into qPCR reaction in the form of 0.5 mm paper generated from the active sampling area of a 20 mm PE-Swab. NTC reaction contains a blank 0.5 mm paper punch.
suspect (e.g. his apartment). Although the blood samples were unprocessed, only the IPC Ct for black leather and brown leather is higher than the average IPC Ct value for NTC. The IPC Ct for cement, denim 1 and denim 2 are less than 0.13 Ct higher than the average IPC Ct value of two NTC reactions. Based on the direct qPCR quantitation results, optimal punch size and number of punches were used in the direct STR reactions (28 cycles). Fig. 9 shows the correlation between the DNA input amount and the average peak height of the STR profiles. Full STR profiles with high profile quality were obtained from all five dry blood stain samples. The intra color balance for all four dye channels and for all five samples are above 40%. These results demonstrate that dry blood stains of forensic casework sample types can be directly quantified and the quantification results can be used to estimate the optimal DNA input for direct STR amplification. The ability to obtain both the DNA quantity and the gender information from unprocessed blood stains and the ability to do direct STR amplification have great potential to increase the capability of the crime laboratory for processing forensic casework samples.
3.6. Direct qPCR/direct STR amplification of dry blood stains on various substrates Dry blood stains deposited on various substrates are a common type of crime scene evidence. Besides forensic touch DNA samples, we also demonstrated the feasibility of a PE-Swab enabled direct qPCR/direct STR amplification workflow for processing dry blood stains deposited on various substrates. The measured DNA quantities by the direct qPCR and the corresponding IPC Ct values are shown in Fig. 8. No male DNA was detected in any of the five samples, which agrees with the fact that the blood is from a female donor. If these were samples in a real forensic case and the suspect was a male, all five blood samples could be excluded from the STR typing analysis. However, with the conventional STR typing workflow, valuable resources would have been required to confirm the absence of probative DNA. Another scenario where such direct quantification assay would be value is when attempting to identify the female victim’s blood in a crime scene associated with the male
Fig. 9. The correlation between DNA input and corresponding average peak height of the STR profile obtained from direct amplification of punches from the same PESwab. Thermal cycle number: 28.
104
J.Y. Liu / Forensic Science International: Genetics 11 (2014) 96–104
4. Conclusions
Acknowledgements
A workflow for direct qPCR quantification of unprocessed forensic casework samples has been developed and demonstrated. By collecting the forensic casework sample on a PE-Swab, a paper punch containing unprocessed DNA sample can be generated and placed directly in a Quantifiler1 Duo assay to assess the gender and the DNA quantity of the sample. By optimizing the size of the paper punch and the baseline setting for Ct determination, the presence of a 0.5 mm paper punch in a qPCR assay has minimal effect on DNA quantification. When the direct qPCR quantitation is none zero, correct gender calls were made in 61 of the 64 forensic casework samples analyzed in this paper. The three samples with missed gender calls are all low level touch DNA samples and are likely due to the stochastic effect of low copy number samples. Since the direct qPCR quantitation is correlated to the post extraction qPCR quantitation as well as the post extraction STR typing results, the direct qPCR can be used to prioritize both touch DNA samples and blood stains so that crime laboratories have the choice to process the samples with better chance of obtaining probative STR profiles first. Although there is some sample variation along the active sampling area of a PESwab, the direct qPCR results proved to be capable of estimating the optimal DNA input for the direct STR amplification of the sample from the same PE-Swab. The direct qPCR and the direct STR amplification of unprocessed forensic casework samples collected on the same PE-Swab could become an attractive alternative STR typing workflow.
The author would like to thank Jia Yang Sean Ho for his technical assistance. Jia Yang is a bioengineering student at UC Berkley. Global R&D office at Thermo Fisher Scientific provided the financial support for his summer internship in Human Identification group at Thermo Fisher Scientific in 2013. References [1] J.M. Butler, Advanced Topics in Forensic DNA Typing: Methodology, Academic Press, Waltham, Massachusetts, 2011. [2] C.M. Cupples, J.R. Champagne, K.E. Lewis, T.D. Cruz, STR profiles from DNA samples with undetected or low quantifiler results, J. Forensic Sci. 54 (2009) 103–107. [3] D.J. Daly, C. Murphy, S.D. McDermott, The transfer of touch DNA from hands to glass, fabric and wood, Forensic Sci. Int. Genet. 6 (2012) 41–46. [4] L. Forensics, HyBeacons1 Technology, Technical Note, vol. 2013, LGC Forensics, Teddington UK, 2013, http://paradna.lgcforensics.com/wp-content/uploads/2013/ 07/HyBeacons_6.pdf. [5] D.J. French, R.L. Howard, N. Gale, T. Brown, D.G. McDowell, P.G. Debenham, Interrogation of short tandem repeats using fluorescent probes and melting curve analysis: a step towards rapid DNA identity screening. Rapid typing of STRs in the human genome by HyBeacon melting, Forensic Sci. Int. Genet. 2 (2008) 333–339. [6] N. Gale, D.J. French, R.L. Howard, D.G. McDowell, P.G. Debenham, T. Brown, Rapid typing of STRs in the human genome by HyBeacon melting, Org. Biomol. Chem. 6 (2008) 4553–4559. [7] A. Kremser, B. Bayer, S. Jung, K. Anslinger, Quantifiler Human DNA Quantification Kit (Applied Biosystems) as a screening kit for DNA profiling, in: Supplement Series Progress in Forensic Genetics 13 Proceedings of the 23rd International ISFG Congress, Forensic Sci. Int. Genet. 2 (2009) 106–107. [8] J.T. Thomas, R.M. Berlin, J.M. Barker, T. Dawson Cruz, Qiagen’s Investigator Quantiplex Kit as a predictor of STR amplification success from low-yield DNA samples, J. Forensic Sci. 58 (2013) 1306–1309.
Contents lists available at ScienceDirect
Forensic Science International: Genetics journal homepage: www.elsevier.com/locate/fsig
Direct qPCR quantification of unprocessed forensic casework samples Jason Yingjie Liu * Human Identification, Thermo Fisher Scientific, 180 Oyster Point Boulevard, South San Francisco, CA 94080, USA
A R T I C L E I N F O
A B S T R A C T
Article history: Received 23 November 2013 Received in revised form 16 February 2014 Accepted 2 March 2014
The current short tandem repeat (STR) typing workflow for forensic casework samples involves sample collection, sample screening, DNA extraction, DNA qPCR quantification and STR amplification. Although very effective and powerful, this workflow still has room for improvements. For example, the screening assays in practice do not provide DNA related information and also do not work with touch DNA samples, which make up of the majority of the property crime samples. It is known that not all DNA samples have equal probative values. Considering the DNA backlog situation crime laboratories face today, an effective screening tool would be highly desirable. It would allow forensic scientists to prioritize the DNA samples so that the limited resources would be first spent on samples that would have better chances of producing informative STR profiles. qPCR assay does provide DNA quantity and gender information and would be an ideal screening tool. However, prior to quantification, sample extraction and purification are required. By the time a DNA sample is ready for qPCR quantification, time and resources have already been spent on samples that should have been given low priority or excluded from further processing if DNA quantity and gender information were known. To overcome this problem, a direct quantification technology is developed to allow qPCR quantification of casework samples without the need for DNA extraction and purification. The key to a direct qPCR assay is the PE-Swab, a novel sample collection device. A small sample punch can be generated from a PE-Swab and placed in a qPCR reaction for quantification. After optimizing the punch size and the quantification software baseline setting, accurate DNA quantification can be obtained from a sample without the need to carry out DNA extraction and purification. Proof of concept studies were done with low lever touch samples as well as blood samples. The PE-Swab also allows direct STR amplification of casework samples without the need for DNA extraction. Besides its potential as a screening tool, the direct qPCR assay can also be used to normalize the DNA input for a direct STR amplification reaction. The feasibility of the direct qPCR/direct STR amplification workflow was demonstrated with touch DNA samples and blood stain samples. ß 2014 Elsevier Ireland Ltd. All rights reserved.
Keywords: Direct qPCR Direct quantification Direct real-time PCR Direct STR amplification PE-Swab DNA extraction
1. Introduction The demand for DNA analysis has been driven up by the increased awareness of the power of DNA technology for solving crimes. Law enforcement agencies, which used to only collect DNA evidence in violent crimes such as homicide and sexual assault, have also begun to collect DNA evidence from property crime cases to aide their investigations. However, biological evidence samples from property crimes are mostly touch DNA samples. Unlike blood or saliva stains, touch DNA is not always identifiable by the naked eye. In many cases law enforcement officers swab surfaces which they believe have been touched by a perpetrator at crime scenes. Or the officers collect evidentiary items which they believe the
* Tel.: +1 650 872 7283; fax: +1 650 266 3063. E-mail addresses: Jason.Liu@thermofisher.com, [email protected] http://dx.doi.org/10.1016/j.fsigen.2014.03.003 1872-4973/ß 2014 Elsevier Ireland Ltd. All rights reserved.
perpetrator had physical contact with. Therefore, it is understandable that some of the touch DNA samples collected this way may not contain any DNA at all. Even when a touch DNA sample from perpetrator is collected, there is no guarantee it will contain enough DNA to obtain a probative STR profile. This is because the amount of DNA left behind on a touched surface depends on the surface property of the item, the length and pressure of the contact, and the shedding characteristic of the individual [3]. Although very powerful, the short tandem repeat (STR) DNA profiling technology is also expensive, time consuming and labor intense. The current lack of an effective screening tool and the increasing number of casework samples collected and submitted, especially those from property crimes, have made it impossible to process all of the samples in a timely manner. As a result, DNA backlog has been created. Some laboratories also limit the number and the type of samples that can be submitted in each case. Law enforcement officers have to make difficult decisions on which
J.Y. Liu / Forensic Science International: Genetics 11 (2014) 96–104
DNA samples to submit, and this may leave more probative samples left un-submitted. Therefore, a simple, fast, and inexpensive screening tool capable of measuring DNA quantity and the sample gender is highly desired. Fast and inexpensive presumptive tests do exist for blood and saliva, but they do not provide DNA quality, quantity and gender information. The LGC1 Forensics ParaDNA1 Screening System is the only forensic DNA sample screening tool available that takes an unprocessed forensic sample as an input and also provides gender and DNA related information in just 75 min. This technology is based on the high resolution melting curve analysis of STR loci D16S539 and TH01 and gender locus Amelogenin [4]. The assay produces the gender of the sample’s contributor as well as a percentage score correlated with the chance of obtaining a successful STR profile. The ParaDNA1 Screening System1 has a detection limit of 62.5 pg of DNA, equivalent to approximately 10 human cells. The lowest DNA level tested where repeatable results were obtained was 250 pg. This is equivalent to approximately 40 cells. A positive gender indication was obtained in 80% of samples with at least 250 pg of input DNA. This allows law enforcement and forensic professionals to better prioritize the samples before deciding on which samples should be submitted for DNA analysis. The high resolution melting curve analysis has also been investigated for the feasibility of allowing rapid interrogation of a small panel of STR loci outside of specialized laboratories [5,6], but further work is required to reduce test complexity (e.g. through multiplexing) and to ensure that assays are sufficiently accurate and robust for forensic analysis [5,6]. Although qPCR quantification assay was originally included in the STR typing workflow to ensure the optimal amount of DNA is used in a multiplex STR reaction, forensic scientists realized that the extreme sensitivity of the qPCR assay potentially allows them to identify DNA samples with quantities below STR reaction sensitivity and therefore to forego analysis of samples less likely to amplify, saving both time and money. Several laboratories have investigated the current commercially available qPCR assays for such purpose, but they were not able to find a cutoff quantity value below which no STR profile can be obtained [2,7,8]. The reasons for why some current qPCR assays are not able to serve as STR typing gatekeeper are thoroughly explained by John Butler in his book [1]. In order for a laboratory to determine that an evidentiary sample processing is to be terminated after DNA quantitation, the laboratory shall have a validation study to support that determination. To enhance the utility of a qPCR assay as a more effective screening tool, the most advanced commercial qPCR assays also provide information on the gender, the male/female DNA quantity ratio, the degree of inhibition and the degree of DNA degradation. However, in all current qPCR assay formats, forensic casework DNA samples need to be extracted and purified first before they can be quantified. The DNA sample input also must be in the form of an aliquot. These constraints greatly diminish the value of a qPCR assay as a screening tool, because by the time the DNA sample is ready for qPCR quantification, significant amount of time and resources have been spent on extracting and purifying samples that have very little to no probative value. In this paper we will describe the proof of concept studies of a new forensic DNA sample screening technique called direct qPCR quantification. This technique allows the DNA quantity and the sample gender information to be obtained before the sample is processed for DNA extraction and purification. Such a screening tool would allow forensic scientists to screen and rank DNA samples in a forensic case based on the DNA quantity, the gender information and the case context. Forensic scientists then have the option of processing samples that have higher chances of yielding 1
For Forensic and Human Identity Applications Only.
97
probative information first, and only processing lower ranked samples when necessary. The critical enabling element of the direct qPCR workflow is the PE-Swab, a novel sample collection device. The PE-Swab allows for forensic sample collection and subsequent generation of a very small paper punch for direct qPCR quantification. Samples collected on a PE-Swab can also be directly STR amplified. The feasibility of using the direct qPCR results to normalize the DNA sample input for the direct STR amplification is also demonstrated. Because there is no DNA loss associated with DNA extraction process, the direct quantification/direct PCR workflow is simpler, faster, and would be an attractive alternative to the current DNA extraction and purification based STR typing workflow.
2. Materials and methods 2.1. PE-Swab A PE-Swab consists of three components: a filter paper stripe, a holder and a clip. A PE-Swab was assembled by wrapping a filter paper stripe around the holder and then securing the paper stripe on the holder using the clip at the end of the PE-Swab handle. An example of a functional PE-Swab is shown in Fig. 1A. The prefix before a PE-Swab indicates the height of the active sampling area, which is 5 mm in this example. The width of the active sampling area is defined by the angled fold of the holder. The PE-Swab is 40 mm long measured from the active sampling area to the end of the swab handle, which is 15 mm wide. PE-Swabs were used to swab objects of interest that contain touch DNA samples or blood samples. When a swabbing liquid was used, it was applied to the object of interest before swabbing. After swabbing, the filter paper stripe was detached from the swab holder and was air dried before punching. Using a Harris Uni-CoreTM punch (Ted Pella, Inc., Redding, CA), punches of the desired size were generated from the active sampling area of the filter paper for a direct qPCR assay or a direct STR amplification. 2.2. Touch DNA sample collection and swabbing using a 5 mm PESwab Touch DNA samples were collected on transparency films. In order to increase the amount of DNA present on their fingers, the donors were asked to touch their face first before pressing their fingers on a piece of transparency film to collect the DNA material. The touch samples were stored in a paper envelope to prevent contamination before being tested. 10 mL of ethanol was applied to each touch sample before being swabbed using a 5 mm PE-Swab. 2.3. DNA extraction and purification from PE-Swabs After a 0.5 mm punch was removed from a 5 mm PE-Swab for direct qPCR. The rest of the active sampling area of the 5 mm PESwab was punched out with a 2.5 mm diameter puncher and the punches were transferred into a PrepFiler1 LySepTM columns/ sample tube assembly (Thermo Fisher Scientific, Waltham, MA). 500 mL of PrepFiler1 lysis buffer (Thermo Fisher Scientific) was added to the LySepTM column and incubated on a thermal mixer at 70 8C for 40 min. After lysis, the lysate was transferred into the sample tube by centrifugation at 16,000 g for 2 min. The LySepTM column containing punches was disposed after centrifugation. Sample tubes containing lysate were loaded into an AutoMate ExpressTM Forensic DNA Extraction System.2 DNA from the touch 2 For Research, Forensic or Paternity Use Only. Not for use in diagnostic procedures.
J.Y. Liu / Forensic Science International: Genetics 11 (2014) 96–104
98
Fig. 1. (A) A 5 mm PE-Swab. The dimension (prefix) before a PE-Swab indicates the height of the filter paper stripe at the active sampling area. (B) The three components of a disassembled 5 mm PE-Swab with two 2 mm punches generated from the active sampling area of the PE-Swab using a Harris Uni-CoreTM punch.
sample lysate was extracted and purified using PrepFiler1 standard protocol. 2.4. Dry blood stain samples preparation and swabbing using a PESwab Dry blood stain samples were prepared by pipetting 20 mL blood from a female donor onto five different types of substrates. The liquid blood was spread out with a pipette tip and was set out to dry for five days. Before swabbing, 40 mL de-ionized water was applied to each dry blood stain. Because the amount of dry blood on the substrate is much higher than on the touch DNA samples, a larger 20 mm PE-Swab was used to swab each dry blood stain. The dry blood stain samples used in the sample uniformity study were prepared by pipetting 10 mL blood from five different male donors onto five separate pieces of microscope slide. The liquid blood was spread out with a pipette tip and allowed to dry over the weekend. Before swabbing using a 5 mm PE-Swab, 10 mL de-ionized water was applied to each dry blood stain. After swabbing, the filter paper stripe was detached from the swab holder and was air dried before punching. 2.5. Real time qPCR quantification of forensic DNA sample collected on a PE-Swab 1
Quantifiler Duo DNA Quantification kit (Thermo Fisher Scientific) is used in the direct qPCR assay. To prepare a direct qPCR reaction plate, a punch generated from the PE-Swab was placed directly into a well of a MicroAmp1 Optical 96-Well Reaction Plate. 25 mL qPCR reaction mix (10.5 mL Quantifiler1 Duo Primer Mix and 12.5 mL Quantifiler1 Duo PCR reaction Mix and 2 mL de-ionized water) was then added to the well containing the sample punch. Quantification reactions were carried out on the Applied Biosystems 7500 Real-Time PCR System3 (Thermo Fisher Scientific) using the manufacture’s recommended protocol. The quantification results were analyzed using SDS Software v2.0.6 (Thermo Fisher Scientific). 2.6. STR amplification and CE analysis The conventional post DNA extraction STR amplification was carried out using GlobalFilerTM PCR Amplification Kit (Thermo Fisher Scientific) by following the manufacture’s recommended protocol. The procedure for direct STR amplification was described below. Based on the direct qPCR result, the desirable input DNA quantity was obtained for each STR reaction by adjusting the punch size and the number of punches generated from a PE-Swab
and released into a well of a MicroAmp1 Optical 96-Well Reaction Plate. 7 mL PCR reaction mix (2.1 mL of GlobalFiler Master Mix,4 3.5 mL of Identifiler1 direct Primer Mix and 1.4 mL of water) was added to each well containing the punches. The reason the Identifiler1 direct Primer Mix was used in place of the GlobalFilerTM Primer Mix is that the GlobalFilerTM Primer Mix is not available to the author when experiments on direct STR amplification were conducted. The thermal cycling conditions were 95 8C/1 m, 28, 29 or 30 cycles of (94 8C/10 s, 59 8C/90 s), 60 8C/10 m and 4 8C – hold. The exact thermal cycle number was specified in each individual experiment described in the results and discussion sections. After thermal cycling, 1 mL PCR product from each sample was mixed with 9 mL GeneScan1500 size standard and deionized formamide. The CE was run on an ABI 3130xl capillary electrophoresis instrument under the following conditions: Oven: 60 8C, Prerun: 15 kV, 180 s, Injection: 3 kV, 10 s, Run: 15 kV, 1500 s, Capillary length: 36 cm, Separation polymer: POP4TM polymer and Dye set: G5. The resulting STR electropherograms were analyzed using the GeneMapper1 ID-X software5 (Thermo Fisher Scientific). 3. Results and discussion 3.1. The effect of the punch size and the baseline setting on qPCR quantitation Because illuminating the sample with light and detecting the fluorescence signal is at the heart of a qPCR assay, it was expected that the presence of a filter paper punch in the reaction well would have an impact on the qPCR assay. This impact becomes clear when one examines the multicomponent plots of the non-template control (NTC) reaction with or without a filter paper punch of different size in the reaction (see Fig. 2). Except for the ROX channel, the passive reference, the presence of a filter paper punch in the reaction well results in elevated background florescent signal in FAMTM (Male target), VIC1 (human target) and NEDTM (IPC) channels and the magnitude of the background elevation is positively correlated with the size of the filter paper punch in the reaction well (see Fig. 2). In addition, the background florescent signal slowly increases after each thermal cycle and the rate of this background fluorescent signal increase is also positively correlated to the size of the filter paper punch (see Fig. 2). It was noted that the elevated baseline and the slope of the baseline affect the Ct value determination, which in turn affects the measured DNA quantity. To determine the Ct value, the SDS software first determines the Rn (the normalized florescent signal) 4
3
When used for purposes other than Human Identification the instruments cited are for Research Use Only. Not for use in diagnostic procedures.
For Forensic or Paternity Use Only. For Research, Forensic, Paternity and Cell Line Authentication. Not for use in Diagnostic and Therapeutic applications. 5
J.Y. Liu / Forensic Science International: Genetics 11 (2014) 96–104
99
Fig. 2. The florescent signals collected from different dye channels during a qPCR reaction without or with a filter paper punch of indicated size in a NTC reaction well. The optimal baseline starting and ending cycle numbers for Ct determination are 15 and 25 for IPC (NEDTM channel).
by dividing the florescent signal in each dye channel by the florescent signal of the passive reference (ROX). The SDS software then uses the Rn values collected from a predefined range of PCR cycles to serve as the baseline. After generating a baselinesubtracted amplification plot of DRn versus cycle number, an algorithm defines the cycle number at which the DRn value crosses the threshold setting as the threshold cycle (Ct). The effect of the paper punch size and the baseline start point/end point setting on IPC Ct value is shown in Fig. 3. Regardless of the baseline setting, the presence of a paper punch in the reaction well causes the IPC Ct value to drift higher. The size of the Ct drift is positively correlated to the size of the paper punch. Therefore, using a small paper punch is essential to minimize the impact of a paper punch on the qPCR assay accuracy. The baseline setting also has impact on the IPC Ct value. After experimenting with different baseline settings, it was discovered that when a paper punch is present in the reaction well, the optimal baseline starting cycle number is 15, which is the highest baseline starting cycle number permitted by the current SDS software. The optimal baseline ending cycle number is right before where the florescent signal starts to break out and rise up above baseline, which is cycle 25 in this case. Compared to Ct values calculated with the baseline setting of 1–25 (the Box plot of crimson color in Fig. 3), the Ct value is less deviated from the true value when the baseline setting is at 15–25, regardless of the punch size (the Box plot of green color in Fig. 3). By using a 0.5 mm
paper punch and the optimal baseline setting in data analysis, we are able to reduce the IPC Ct drift to about 1%. Because the presence of a paper punch in a qPCR reaction causes Ct value to drift higher and the DNA quantity of the sample is determined based on the Ct value of the sample and those of DNA standards, the under estimation of the DNA quantity in the sample will occur when a paper punch is present in the reaction well. However, the DNA quantity under estimation is minimized by the use of the 0.5 mm punch and the optimal baseline setting. To understand this effect more quantitatively, a 0.5 mm paper punch was placed in a reaction well containing a known amount of human DNA standard from Quantifiler1 Duo kit. Six reactions were carried out at each DNA input quantity. The optimal baseline setting was set at 15 and 19 for the Quantifiler1 Duo human and male targets and 15 and 25 for IPC target. The measured DNA quantity against the input DNA quantity was shown in Table 1. As expected, with the presence of a paper punch in the reaction well, the DNA quantity is under estimated for both human and male targets with the DNA input quantity down to 0.42 ng, but the percentage difference is very small. With the DNA input down to 0.42 ng, the percentage difference between the input quantity and the measured quantity is less than 13.7%. Even with Std 7 and 8, the lower end of the DNA quantity standard range, the percentage difference between the input quantity and the measured quantity is still less than 34.5%. The DNA quantity overestimation observed at 0.14 ng and 0.05 ng input level is likely due to the stochastic
J.Y. Liu / Forensic Science International: Genetics 11 (2014) 96–104
100
Fig. 3. The effect of paper punch size and baseline start point/end point setting on IPC Ct value determination. Three replicates for NTC reaction containing a punch of size 0.5 mm, 1 mm or 2 mm. Four replicates for NTC reaction containing no punch. (For interpretation of the references to color in the text, the reader is referred to the web version of the article.)
effect typical of low copy number samples. This result clearly demonstrated the robustness and the accuracy of the qPCR DNA quantification with the presence of a 0.5 mm paper punch. 3.2. Direct qPCR quantification of unprocessed touch DNA samples To demonstrate the feasibility of direct quantification of unprocessed touch DNA samples using a qPCR assay, touch DNA samples from five male and two female donors (total 19 samples) were quantified. The two female donors were known to be low shedders and their samples were included in this study to challenge the detection sensitivity of the qPCR and the conventional post DNA extraction STR typing workflow. The measured human and human male DNA quantities per punch and the corresponding IPC Ct values are shown in Fig. 4. No human or Y DNA was detected in any of the four female samples. Among the 15 male samples (three from each of the five male donors), both human and Y DNA were detected in 13 samples. The two exceptions were sample Male_B_S1 (only Human DNA detected, 41 pg) and Male_D_S3 (only Y DNA detected, 32.5 pg). This could be due to the fact that the quantifiler1 Duo human and quantifiler1 Duo Y targets are located on different chromosomes and there is very little DNA in the sample as shown by the low detected DNA quantity. The failed detection of both targets is likely due to the stochastic effect which is common for low copy number samples.
Other than the samples from donor Male_E, the IPC Ct values for all samples were within 0.5 Ct of the punch NTC (29.1, average; 0.19, standard deviation) indicating minimal inhibition even with unprocessed sample. Interestingly, the IPC Ct values from all three samples of donor Male_E (27.7, average; 0.09, standard deviation) are more than one Ct lower than the rest of the samples (29.1, average; 0.14, standard deviation). As expected, after DNA extraction and purification using AutoMate ExpressTM, the IPC Ct discrepancy between donor Male_E and rest of the samples disappeared. (The average IPC Ct for all purified samples is 28.6 with standard deviation of 0.11.) The lowered IPC Ct values in unprocessed Male_E samples suggest that some composition in these samples actually enhances qPCR efficiency. The exact reasons are not clear at this point. Nevertheless, this result clearly demonstrated the feasibility of direct qPCR quantification of unprocessed touch DNA samples. 3.3. The correlation between direct qPCR quantification, post extraction quantification and post extraction STR typing One foreseeable application of a direct qPCR assay is the screening and prioritization of a large number of touch DNA samples from property crimes. Such capability would allow limited forensic resources to be spent on processing samples that have the best chance of producing probative information. In order for this approach to be effective one needs to show that the direct
Table 1 The comparison of measured DNA quantity to that of input DNA quantity with the presence of a 0.5 mm filter paper punch in each qPCR reaction.
Std Std Std Std Std Std Std Std a
1 2 3 4 5 6 7 8
STD DNA input (ng)
Duo human (ng)a
Duo human STDEV
Duo male (ng)a
Duo male STDEV
% diff human
% diff male
100.00 33.40 11.12 3.70 1.24 0.42 0.14 0.05
97.38 31.19 11.13 3.59 1.15 0.38 0.14 0.06
4.62 3.16 0.51 0.18 0.03 0.04 0.01 0.01
95.74 30.73 10.58 3.49 1.10 0.36 0.09 0.04
3.29 1.96 0.38 0.16 0.06 0.04 0.01 0.01
2.62% 6.61% +0.13% 2.91% 6.87% 10.53% 3.62% +28.55%
4.26% 7.98% 4.84% 5.64% 11.41% 13.70% +34.48% 5.99%
Average of 6 replicates except Std 3 and Std 8, which are average of 5 replicates.
J.Y. Liu / Forensic Science International: Genetics 11 (2014) 96–104
101
Fig. 4. Direct qPCR quantification of unprocessed touch DNA samples. The touch DNA was introduced into a qPCR reaction in the form of 0.5 mm paper generated from the active sampling area of a 5 mm PE-Swab.
quantification result is correlated to the post extraction quantification as well as to the STR typing result. After taking a 0.5 mm punch for direct qPCR, the remaining sample from all 19 touch samples in the previous section were further extracted and purified using an AutoMate ExpressTM Forensic DNA Extraction System. 2 mL purified DNA was quantified on the same plate for the direct qPCR in the previous section to reduce variation. Based on the post extraction quantification result, 15 mL purified DNA from each of the 19 samples was amplified using GlobalFilerTM PCR kit in a 25 mL reaction for 29 cycles. The quantification results from the direct qPCR, the post extraction qPCR and the percentage alleles recovered from STR typing were plotted in Fig. 5. Since the sample input for the direct qPCR is a 0.5 mm solid punch and that for the post extraction qPCR is 2 mL aliquot, the DNA quantity in the unit of pg/well were plotted so that both the direct qPCR result and the
post extraction qPCR result can be represented by the same Y axis. The samples were ranked from the highest to the lowest based on the direct qPCR human quantity (the solid blue line). The trend of the human quantity from the post extraction qPCR (the dotted red line) follows that of the direct qPCR. The fluctuation of the post extraction qPCR line is likely caused by the non-uniformity of the sample on a PE-Swab and the stochastic effect from low copy number samples. 15 mL purified DNA from each of the 19 touch samples were amplified using a GlobalFilerTM PCR kit. Other than sample Male_C_S2, all the other samples have input DNA quantity less than the target 1 ng. As shown in Fig. 5, the direct quantification results also correlated with the percentage alleles recovered (the solid green line) from the STR analysis. The top five ranked samples on average have 98% of the alleles recovered, while the bottom five ranked samples on average only have 8% of the
Fig. 5. The correlation between direct quantification, post extraction quantification and % alleles recovered from post extraction STR. STR PCR input volume is 15 mL for all samples. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of the article.)
J.Y. Liu / Forensic Science International: Genetics 11 (2014) 96–104
102
Table 2 Sample uniformity characterization of low level touch DNA samples and blood stain samples collected on a 5 mm PE-Swab. Male 1
Touch sample (n = 4) Individual DNA quantity measurement (pg/punch) Average DNA quantity (pg/punch) STDEV CV Blood sample (n = 4) Individual DNA quantity measurement (ng/punch) Average DNA quantity (ng/punch) STDEV CV
Male 2
Male 3
Male 4
Male 5
Human
Y
Human
Y
Human
Y
Human
Y
Human
Y
3.9, 11.2, 8.2, 4.9 7.0
10.1, 12.6, 54.8, 33.5 27.7
35.2, 147.3, 105.2, 129.4 104.3
33.5, 293.1, 159.5, 178.2 166.1
148.5, 173.2, 62.4, 73.0 114.3
213.1, 136.1, 131.2, 28.9, 127.3
16.8, 3.7, 15.9, 35.0 17.8
15.6, 40.5, 34.7, 117.1 52.0
24.3, 28.5, 39.6 18.4 27.7
9.3, 52.0, 39.4, 42.5 35.8
2.8 40%
18.1 65%
42.6 41%
92.1 55%
47.5 42%
65.4 51%
11.1 63%
38.7 74%
7.7 28%
16.0 45%
0.84, 1.97, 2.24, 2.46 1.88
2.27, 3.48, 6.04, 4.55 4.09
0.43, 0.49, 0.82, 0.50 0.56
1.15, 1.31, 2.09, 1.17 1.43
0.78, 1.09, 1.45, 1.57 1.22
1.78, 2.97, 3.76, 4.50 3.25
1.02, 1.57, 1.02, 1.06 1.17
2.14, 3.00, 2.01, 2.58 2.43
0.54, 0.71, 0.49, 0.53 0.57
1.24, 1.88, 1.43, 1.95 1.63
0.72 38%
1.60 39%
0.18 32%
0.45 31%
0.36 29%
1.17 36%
0.27 23%
0.45 18%
0.10 17%
0.34 21%
alleles recovered. This result clearly demonstrated that the direct qPCR with the Quantifiler1 Duo kit is effective at prioritizing a large number of samples. However, the direct qPCR with Quantifiler1 Duo kit should not be used as a gatekeeper for STR analysis. Although the bottom five ranked samples have direct qPCR quantification of zero, only one did not produce any accurate allele. Because of the possibility of false negative with low level DNA samples when a Quantifiler1 Duo kit is used for direct qPCR, one should not discontinue processing of samples, which may prove pivotal in a case solely due to the low or undetected direct quantitation data. 3.4. Sample uniformity on a PE-Swab Since sample is collected along the active sampling area of a PESwab, a certain degree of sample variation on a PE-Swab is expected. The correlation study results between the direct qPCR quantification, the post extraction quantification and the post extraction STR typing described above also indicated that such sample variation exists. To characterize the sample distribution on a PE-Swab directly, five touch samples and five blood stain samples were collected using 5 mm PE-Swabs. Four 0.5 mm punches were generated from each PE-Swab and directly quantified using a Quantifiler1 Duo kit. The individual quantity of each punch, the average quantity of the four punches, the standard deviation and the coefficient of variation (CV) from each sample are listed in Table 2. The result confirms that sample variation does exist for samples collected on a PE-Swab. Blood samples (average CV of 27.8%) are more evenly distributed on a PE-Swab than the low level touch samples (average CV of 42.8%). With low level touch samples, the CV for the Y quantity is larger than that for the human quantity for all five samples. We also examined the IPC Ct values for both low level touch DNA samples and blood samples. The average IPC Ct from all 20 touch samples is 28.7 (Standard deviation of 0.17), which is only 0.1 Ct higher than the IPC Ct for NTC. As expected, inhibition was detected with all blood samples due to the high concentration of heme in blood. The average IPC Ct from all 20 blood sample is 30.2 (Standard deviation of 0.3), which is 1.9 Ct higher than and the IPC Ct (28.3) for NTC. It is also worthwhile to point out that correct gender calls were made with all 40 samples in this study.
for the conventional DNA extraction based STR typing workflow, the quantification result from a direct qPCR assay could also be used to normalize the DNA input for a direct STR amplification. One could envision the possibility of a simpler STR typing workflow, in which the DNA extraction is eliminated. To demonstrate the direct qPCR/direct STR workflow, touch DNA samples from two female and two male donors were collected on 5 mm PE-Swabs and quantified directly using a Quantifiler1 Duo kit. The measured human and human male DNA quantities along with IPC Ct are shown in Fig. 6. The IPC Ct for all four samples is within 0.5 Ct of NTC IPC Ct. No human male DNA was detected in sample Male 2, which is likely due to the stochastic effect which is common for low copy number samples. Since there is no DNA extraction, the DNA input to a direct amplification reaction has to be in the form of paper punch from a PE-Swab. The design of the 5 mm PE-Swab is such that the touch DNA sample is collected and concentrated on the 1 mm 5 mm stripe of the angled fold area of the PE-Swab. Based on the DNA quantity on each 0.5 mm punch, the optimal input DNA for a direct STR amplification reaction can be obtained by controlling the punch size and the number of punches. Because of the small size of the punches, a smaller PCR reaction volume can be used to increase STR sensitivity, which is critical for low level touch DNA samples. Instead of the standard 25 mL, 7 mL PCR reaction was used in this study and it can accommodate up to eight 1 mm paper punches. After removing a 0.5 mm punch, four 1 mm punches can be practically generated from the remaining active sampling area of a
3.5. Direct qPCR/direct STR PCR of touch DNA samples One additional benefit of collecting DNA sample on a PE-Swab is that the small punch from a PE-Swab is also compatible with direct STR amplification. Therefore, besides its utility as a screening tool
Fig. 6. Direct qPCR quantification of unprocessed touch DNA samples. The touch DNA was introduced into qPCR reaction in the form of 0.5 mm paper generated from the active sampling area of a 5 mm PE-Swab.
J.Y. Liu / Forensic Science International: Genetics 11 (2014) 96–104
Fig. 7. The correlation between DNA input and corresponding average peak height of the STR profile obtained from direct amplification of punches from the same PESwab. Thermal cycle number: 30.
5 mm PE-Swab and all four punches can be used as the input for a direct amplification reaction if needed. The recommended PCR reaction conditions for the GlobalFilerTM PCR kit is 1 ng input DNA in a 25 mL PCR reaction with 29 cycles, which translates to 280 pg input DNA in a 7 mL reaction with 29 cycles. Based on the direct quantification results, the total available amount of DNA on a 5 mm PE-Swab for samples female 1 and male 2 is 208 pg (estimated based on the size of the active sampling area of a 5 mm PE-Swab) and therefore all four 1 mm punches were used in the PCR reaction. For female 2 and male 1, one 0.5 mm punch was used in the STR reaction, which corresponds to 386 pg and 56 pg DNA respectively. More DNA from male 1 could have been used to reach the desired target input of 280 pg. Instead, only one punch was used so that the STR profile quality can be assessed when the suboptimal DNA input is used. The PCR reactions were carried out for 30 cycles. Full STR profiles were obtained for all four touch DNA samples. Considering the existence of the sample variation on a PE-Swab, the PCR input DNA quantity is correlated reasonably well with the average peak height of the STR results (see Fig. 7). Off scale peaks were observed with the profile in female 2, which is in agreement with the amount of input DNA and the fact that PCR was run for an additional cycle.
103
Fig. 8. Direct qPCR quantification of unprocessed blood stains collected on a 20 mm PE-Swab. The blood stains were introduced into qPCR reaction in the form of 0.5 mm paper generated from the active sampling area of a 20 mm PE-Swab. NTC reaction contains a blank 0.5 mm paper punch.
suspect (e.g. his apartment). Although the blood samples were unprocessed, only the IPC Ct for black leather and brown leather is higher than the average IPC Ct value for NTC. The IPC Ct for cement, denim 1 and denim 2 are less than 0.13 Ct higher than the average IPC Ct value of two NTC reactions. Based on the direct qPCR quantitation results, optimal punch size and number of punches were used in the direct STR reactions (28 cycles). Fig. 9 shows the correlation between the DNA input amount and the average peak height of the STR profiles. Full STR profiles with high profile quality were obtained from all five dry blood stain samples. The intra color balance for all four dye channels and for all five samples are above 40%. These results demonstrate that dry blood stains of forensic casework sample types can be directly quantified and the quantification results can be used to estimate the optimal DNA input for direct STR amplification. The ability to obtain both the DNA quantity and the gender information from unprocessed blood stains and the ability to do direct STR amplification have great potential to increase the capability of the crime laboratory for processing forensic casework samples.
3.6. Direct qPCR/direct STR amplification of dry blood stains on various substrates Dry blood stains deposited on various substrates are a common type of crime scene evidence. Besides forensic touch DNA samples, we also demonstrated the feasibility of a PE-Swab enabled direct qPCR/direct STR amplification workflow for processing dry blood stains deposited on various substrates. The measured DNA quantities by the direct qPCR and the corresponding IPC Ct values are shown in Fig. 8. No male DNA was detected in any of the five samples, which agrees with the fact that the blood is from a female donor. If these were samples in a real forensic case and the suspect was a male, all five blood samples could be excluded from the STR typing analysis. However, with the conventional STR typing workflow, valuable resources would have been required to confirm the absence of probative DNA. Another scenario where such direct quantification assay would be value is when attempting to identify the female victim’s blood in a crime scene associated with the male
Fig. 9. The correlation between DNA input and corresponding average peak height of the STR profile obtained from direct amplification of punches from the same PESwab. Thermal cycle number: 28.
104
J.Y. Liu / Forensic Science International: Genetics 11 (2014) 96–104
4. Conclusions
Acknowledgements
A workflow for direct qPCR quantification of unprocessed forensic casework samples has been developed and demonstrated. By collecting the forensic casework sample on a PE-Swab, a paper punch containing unprocessed DNA sample can be generated and placed directly in a Quantifiler1 Duo assay to assess the gender and the DNA quantity of the sample. By optimizing the size of the paper punch and the baseline setting for Ct determination, the presence of a 0.5 mm paper punch in a qPCR assay has minimal effect on DNA quantification. When the direct qPCR quantitation is none zero, correct gender calls were made in 61 of the 64 forensic casework samples analyzed in this paper. The three samples with missed gender calls are all low level touch DNA samples and are likely due to the stochastic effect of low copy number samples. Since the direct qPCR quantitation is correlated to the post extraction qPCR quantitation as well as the post extraction STR typing results, the direct qPCR can be used to prioritize both touch DNA samples and blood stains so that crime laboratories have the choice to process the samples with better chance of obtaining probative STR profiles first. Although there is some sample variation along the active sampling area of a PESwab, the direct qPCR results proved to be capable of estimating the optimal DNA input for the direct STR amplification of the sample from the same PE-Swab. The direct qPCR and the direct STR amplification of unprocessed forensic casework samples collected on the same PE-Swab could become an attractive alternative STR typing workflow.
The author would like to thank Jia Yang Sean Ho for his technical assistance. Jia Yang is a bioengineering student at UC Berkley. Global R&D office at Thermo Fisher Scientific provided the financial support for his summer internship in Human Identification group at Thermo Fisher Scientific in 2013. References [1] J.M. Butler, Advanced Topics in Forensic DNA Typing: Methodology, Academic Press, Waltham, Massachusetts, 2011. [2] C.M. Cupples, J.R. Champagne, K.E. Lewis, T.D. Cruz, STR profiles from DNA samples with undetected or low quantifiler results, J. Forensic Sci. 54 (2009) 103–107. [3] D.J. Daly, C. Murphy, S.D. McDermott, The transfer of touch DNA from hands to glass, fabric and wood, Forensic Sci. Int. Genet. 6 (2012) 41–46. [4] L. Forensics, HyBeacons1 Technology, Technical Note, vol. 2013, LGC Forensics, Teddington UK, 2013, http://paradna.lgcforensics.com/wp-content/uploads/2013/ 07/HyBeacons_6.pdf. [5] D.J. French, R.L. Howard, N. Gale, T. Brown, D.G. McDowell, P.G. Debenham, Interrogation of short tandem repeats using fluorescent probes and melting curve analysis: a step towards rapid DNA identity screening. Rapid typing of STRs in the human genome by HyBeacon melting, Forensic Sci. Int. Genet. 2 (2008) 333–339. [6] N. Gale, D.J. French, R.L. Howard, D.G. McDowell, P.G. Debenham, T. Brown, Rapid typing of STRs in the human genome by HyBeacon melting, Org. Biomol. Chem. 6 (2008) 4553–4559. [7] A. Kremser, B. Bayer, S. Jung, K. Anslinger, Quantifiler Human DNA Quantification Kit (Applied Biosystems) as a screening kit for DNA profiling, in: Supplement Series Progress in Forensic Genetics 13 Proceedings of the 23rd International ISFG Congress, Forensic Sci. Int. Genet. 2 (2009) 106–107. [8] J.T. Thomas, R.M. Berlin, J.M. Barker, T. Dawson Cruz, Qiagen’s Investigator Quantiplex Kit as a predictor of STR amplification success from low-yield DNA samples, J. Forensic Sci. 58 (2013) 1306–1309.
