University of Amsterdam Research Project
Presumptive quantification of DNA and
improving DNA analysis in fingermarks
Magdalena Birkl (11406739)
Supervisor: Dr. Annemieke van Dam
Examiner:
Prof. Dr. Ate Kloosterman
Amsterdam University Medical Center Department of Biomedical Engineering and Physics
1.2.2018 – 7.9.2018 MSc in Forensic Science University of Amsterdam
36 EC
2
1. Abstract
Two significant means of identification in forensics are DNA and fingermarks. To generate a usable DNA profile, a minimum amount of DNA is necessary. Usually, the DNA concentration in biological samples at the crime scene is unknown. A presumptive quantification method for DNA can aid in a pre-selection of crime scene traces and can result in an increased success rate of DNA typing. In many cases, the visual features of fingermarks are used to identify the donor of a fingermark. For the visualisation, different techniques, like for example the use of dusting powder or antibodies, can be used. However, it can also be necessary to generate a DNA profile from a fingermark after the visualisation. It is not yet known if it is possible to generate a DNA profile from fingermarks which were previously treated with antibodies. The first part of the research focussed on the investigation whether, first, the fluorescence intensity of the DNA-binding dyes 4’,6-diamidino-2-phenylindole (DAPI) and Diamond™ Nucleic Acid Dye and, secondly, the amount of the protein markers telomere repeat factor-1 (TRF1) and -2 (TRF2) showed a correlation with the DNA concentration of a sample. It was not yet possible to establish a linear correlation between neither the DNA-binding dyes and the DNA concentration, nor the protein markers and the DNA concentration.
The second part of the research investigated the applicability of DNA typing after the treatment of fingermarks with antibodies. Fingermarks were deposited on different substrates, visualised with antibodies and DNA profiles were generated. It was possible to generate full and partial profiles from 18 of 27 fingermark samples.
3
2. Introduction
Deoxyribonucleic acid (DNA) analysis is a substantial part in the criminal justice system and its relevance has increased over the past years (1). The increasing importance of DNA analysis gave rise for new challenges for forensic laboratories. Due to limited capacities in the laboratories, a selection of suitable samples for DNA analysis became more important (2). To be able to identify the donor of a trace found at the crime scene, in many cases a DNA profile is generated. This DNA profile is compared to possible reference profiles of a suspect or to profiles in a DNA profile databank. A DNA profile is the result of the typing of Short Tandem Repeats (STR), which are repetitions of specific sequences of amino acids in the DNA. For STR profiling of biological samples, a minimum amount of DNA is necessary to retrieve a clear STR profile. However, the quantity and quality of the DNA depends on the source of the material and the used DNA extraction method. The quantity of present DNA subsequently influences the STR typing method and whether the standard STR typing can be performed or if the method needs to be adjusted for a low amount of DNA (3). Usually a minimum amount of 0.1 to 0.5 ng of DNA is required, in the most commonly used STR profiling systems, to allow reliable STR typing. If the amount of DNA is too low, stochastic effects, such as allele drop-out can occur (1). For this reason, a DNA quantification before the STR analysis is desirable.
This research project is divided in two sections. The first section is focusing on the development of a presumptive method to estimate the amount of DNA directly at the crime scene. In section two, the main focus will be on the immunolabelling of fingermarks, subsequently followed by STR profiling.
Part 1: Presumptive DNA quantification
The currently available DNA quantification techniques range from the ultraviolet (UV) spectrometry and gel-based methods, to dye staining and quantitative Polymerase Chain Reaction (PCR) methods (4). The main disadvantage of all these methods is that they are laboratory-bound and often time consuming and costly. In the Netherlands, most of the forensic tests are conducted by the Netherlands Forensic Institute (NFI). Due to limited resources, it is only possible to send in a limited number of samples per case (2). Nevertheless, in forensic cases, it is often necessary for the investigators to get as much relevant information as possible to develop a case scenario (5). Up until know, it is not possible for the forensic investigator to predict if the samples, sent to the NFI, yield enough DNA to get a reliable STR profile. Mapes and others (2) studied the success rate of DNA analysis in the Netherlands and reported that many DNA traces, which were analysed by the NFI, did not result in a DNA profile. A presumptive test for the quantification of DNA can aid in the early phase of the investigation, since it could contribute to a pre-selection of forensic samples which are sent to the NFI for STR-typing (2).
All currently available DNA quantification methods are laboratory-bound. Therefore, a new approach for a presumptive DNA quantification method is investigated. The general aim of this first section of the research is to investigate the possibility of a presumptive quantification method for DNA. Two different strategies were chosen to do so. First, the use of DNA-binding dyes and the correlation to the DNA quantity is investigated. Secondly, immunolabelling of specific DNA binding proteins is examined. The two strategies are discussed in the following part.
4 A common technique to detect and quantify DNA, is the use of fluorescent dyes, which interact with DNA. Depending on the used dye, the fluorescent signal of the sample can increase upon the binding of the dye to DNA or the emission spectrum of the dye can differ between the bound and unbound state of the dye (6).
It is hypothesised that the increased fluorescence intensity of a DNA-binding dye bound to a sample, correlates with the amount of DNA present in the sample. For this purpose, two DNA-binding dyes, namely 4’,6-diamidino-2-phenylindole (DAPI) and Diamond™ Nucleic Acid Dye (Diamond Dye), are tested. DAPI is a fluorescent and DNA-binding dye and binds to the minor groove of double-stranded DNA at A/T rich regions (7,8). The dye is non-destructive for DNA and shows a blue colour upon excitement with UV light (8,9). The binding of DAPI to DNA generates an increase in fluorescence intensity of approximately 20 times (10). DAPI is usually used as counterstaining reagent, for example, for fluorescence microscopy and chromosome staining (11). More recently the DAPI staining of cell nuclei was used as a screening method for successful STR typing of hair roots (8,9,12,13). Cell nuclei were counted and according to the number of stained cell nuclei, a pre-selection of hair was done. Hair with a higher count of stained nuclei were more likely to give a complete STR profile (9). Van Dam and others chose a comparable approach and stained cells in fingermarks with DAPI. A moderate correlation was found between DAPI-stained cells and the amount of DNA (unpublished data). More recently, new DNA-binding dyes have been developed, where Diamond Dye is one of them. Diamond dye binds externally to the DNA strand. Due to its ability to penetrate the cell membrane, it is also possible to stain genomic DNA similar to DAPI (7,14).
The previously mentioned approach by van Dam et al. and Boonen et al. (13) is picked up to answer the hypothesis of a correlation between the fluorescence intensity of a sample and its DNA content. In this study, samples with varying DNA concentrations are stained with the two previously mentioned DNA-binding dyes (DAPI, Diamond dye). The samples are excited with a specific wavelength in a fluorescence spectrometer and the fluorescence intensity is measured.
The second investigated hypothesis is that the amount of a protein marker can be correlated to the amount of DNA present in a biological sample. Potential proteins markers, which can be used for this, are the telomere repeat factor-1 (TRF1) and -2 (TRF2). TRF1 and TRF2 are part of the so called Shelterin complex. This complex has an essential function for the telomeres and both proteins directly interact with the DNA at the telomeres (15). The proteins of the Shelterin complex are abundant at the chromosome ends and are also present throughout the whole cell cycle (16). The concentrations of TRF1 and TRF2 are not correlated to the telomere length and do not show age-dependent changes in concentration (17). This is an indicator that the complex is relatively stable over time and can be measured.
The aim of this first part of the research is the applicability of two different approaches, namely DNA-binding dyes and immunolabelling, and their potential use in a presumptive DNA quantification method. This first part can be divided in two questions, which need to be answered.
The first question is whether DNA-binding dyes can be used in a presumptive DNA quantification method. Specifically, the correlation between the fluorescence signal of the DNA-binding dye and the DNA concentration of a sample is examined. The second section answers the question whether the concentration of the protein markers TRF1 and TRF2, in biological samples, correlates with the DNA concentration of this sample.
5 The conducted experiments are the first step in the direction of a presumptive DNA quantification method and evaluated the usability of DNA-binding dyes and immunolabelling to predict the DNA concentration of samples. A presumptive DNA quantification method would aid in pre-selection of samples directly at the crime scene. This would lead to a reduction in costs and time and could help in increasing the success rate of DNA typing.
Part 2: Immunolabelling of fingermarks and subsequent STR typing
Fingermarks are encountered during many forensic investigations of crime scenes. Latent fingerprints are not visible with the naked eye and require visualisation with different techniques. Currently there are several visualisation techniques available, like for example dusting powders or chemicals (18). Another option is to specifically target chemical components in the fingermarks with antibodies. Antibodies are widely used in the medical field to aid in diagnosis and therapy. However, antibodies can also be used in a forensic context, namely in the visualisation of fingermarks (19). For this purpose, the peptide dermcidin is chosen as a target for the antibodies. Dermcidin is an antimicrobial peptide that is expressed in sweat glands and present on the skin surface (20,21). The peptide can be visualised on the pore-sites of fingermarks. Pores are a level three feature of fingermarks and can be used for the purpose of identification of a person (22). Up till now, it was shown that it is possible to use immunolabelling to target specific components in the fingermarks (19), to develop or re-develop fingermarks with immunolabelling (23) and to use immunolabelling after the visualisation of fingermarks with other techniques (18). However, it is not clear if it is possible to generate a STR profile from fingermarks after immunolabelling.
Therefore, the aim of this section is to investigate whether it is possible to generate a STR profile after the immunolabelling of fingermarks. In this study, fingermarks were left on three forensically relevant substrates, namely glass, metal and tile. Immunolabelling was performed with a subsequent DNA extraction and STR analysis.
However, before this question can be answered, different optimisation steps need to be considered as many factors can influence the success of STR typing of low level traces like fingermarks. The choice of swab for sampling purposes can have an influence on the final yield of DNA, which can be extracted from a sample (24–26). Several different types of swabs are currently available and used in the forensic field. Cotton tipped swabs are most commonly used, but more recently also nylon tipped swabs were introduced in the field (27). Different types of swabs have their advantages and disadvantages, which have to be considered when choosing the right swab. Five different types of swabs, currently used in forensic practise, are selected and analysed for background DNA. Additionally, the most efficient swab in extracting DNA is determined and selected for the final experiment.
The subsequent immunolabelling experiment is conducted on various substrates (glass, metal and tile) to investigate possible differences. The deposited fingermarks are visualised with antibodies, swabbed and a STR profile is generated from each sample. The produced DNA profile is compared to the reference profile of the donor. According to the obtained results, the applicability of STR typing after immunolabelling of fingermarks is evaluated.
The aim of this second part of the research is divided in two questions. The first question is whether it is possible to obtain a useful STR profile after immunolabelling. If STR profiling is possible, the second question investigates whether the obtained STR profiles of treated and untreated fingermarks show significant differences.
6 If it is possible to generate STR profiles after immunolabelling of fingermarks, antibodies can be used to also generate a donor profile before the subsequent STR profiling. This increases the informative value of the fingermark and can help during investigations.
7
3. Material and methods
Part 1: Presumptive DNA quantification
The first part is divided in two sections. In the first section, DNA-binding dyes were tested for their applicability in a presumptive DNA quantification method.
3.1. Detection of DNA with DNA-binding dyes
For the detection of DNA, DNA from whole blood samples of volunteers and human genomic DNA (Promega Corporation, USA) were stained with the two DNA-binding dyes DAPI (Sigma Aldrich, Germany) and Diamond Dye (Promega Corporation, USA). A correlation between the fluorescence intensity and DNA concentration of the samples is investigated. Additionally, after the establishing of a correlation, the limit of detection of this method should be determined by using a dilution series of DNA samples.
3.1.1. Optimisation of the fluorescence spectrometer settings
During a preliminary experiment, the optimal settings for the fluorescence spectrometer were evaluated. The manufacturers’ specifications of the excitation and emission wavelength of the dyes were tested with DNA containing samples. To determine the optimal settings, first the sample was excited with the excitation wavelength proposed by the manufacturer. The emission spectrum was scanned for the wavelength of the maximum peak intensity. At this emission wavelength, an excitation spectrum was generated. The excitation spectrum was again checked for the wavelength of the maximum peak intensity. Afterwards, again an emission spectrum was generated using the excitation wavelength of the excitation spectrum. This process was repeated until the optimal combination of excitation and emission wavelength was found.
Before the measurement, the stock solutions of the two used dyes (DAPI and Diamond Dye) were diluted 1:20 (v/v) in 96% ethanol. For the analysis a mix of 170 µl low Tris-EDTA (TE) buffer, 5 µl sample and 25 µl dye was prepared. After addition of the dye, the sample was mixed and incubated before measurement. The samples were measured according to the previously described pattern. Depending on the readings of the DNA samples, the settings were adjusted to achieve optimal results. In table 1, the settings of the fluorescence spectrometer are described.
Table 1 Setting of the fluorescence spectrometer for DAPI and Diamond Dye
DAPI Diamond Dye
Excitation 348 nm 346 nm
Emission 474 nm 461 nm
Width of emission slit 10 nm 20 nm
3.1.2. Absorption measurement of Diamond Dye
The settings, which gave the best results for the measurements of Diamond Dye, differed from the manufacturer’s specifications. These state an excitation wavelength of 494 nm and an emission wavelength of 558 nm. Because of this difference, between the results of the optimisation and the manufacturer’s specifications, an absorption spectrum of Diamond Dye was made; The absorption spectrum of a substance is corresponding to the excitation
8 spectrum of the sample. For this reason, an absorption spectrum can be used to determine the best excitation wavelength with the maximum absorption as excitation wavelength. The measurements were conducted with the Lambda Bio spectrophotometer (PerkinElmer, USA) and a super micro fluorescence cuvette from Thorlabs (Thorlabs, USA). The samples were measured in a range from 200 nm to 800 nm and low TE buffer was used as a background reference sample. A blank sample, which consists of TE buffer and dye, was measured at six different time points, ranging from t=0 minutes to t=6 minutes. During sample preparation, a mix of 350 µl of TE buffer and 50 µl of dye was added to the cuvette.
3.1.3. Excitation-Emission-Matrix (EEM) of Diamond Dye
To confirm the right settings for the fluorescence spectrometer, an excitation-emission-matrix of Diamond Dye was generated. A range from 300 nm to 545 nm was scanned for the excitation wavelength in steps of 5 nm. This resulted in a total of 50 measurements for each sample. The emission wavelength was scanned in a range from 330 nm to 700 nm. Four different samples were measured: TE buffer, TE buffer + DNA, TE buffer + dye and TE buffer + dye + DNA. From these four samples, four excitation-emission maps were generated. The measurements were conducted with the LS-55 Fluorescence Spectrometer (PerkinElmer, USA) and the super micro fluorescence cuvette from Thorlabs (Thorlabs, USA).
3.1.4. Intensity measurements with the fluorescence spectrometer
The experiment to determine the fluorescence intensity of DNA samples, stained with DNA-binding dyes, was done in triplicates on two different days.
Human genomic DNA (stock concentration 152 ng/µl, Promega Corporation, USA) and extracted DNA from whole blood samples (stock concentration 36.9 ng/µl and 34.8 ng/µl) from two volunteers were used. From the whole blood samples, dilution series were prepared. The dilution series were made with TE buffer with a pH of 7.55 and resulted in the following concentrations: undiluted sample, 10 ng/µl, 5.0 ng/µl, 2.5 ng/µl, 1.0 ng/µl, 0.5 ng/µl, 0.1 ng/µl, 0.05 ng/µl, 0.01 ng/µl and a blank sample (only TE buffer).
Like in the optimisation step, the stock solutions of the two used dyes (DAPI and Diamond Dye) were diluted 1:20 (v/v) in 96% ethanol. For the analysis, a mix of 170 µl TE buffer, 5 µl sample and 25 µl dye was prepared. Here as well, after the addition of the dye, the sample was mixed and incubated for two minutes before measurement. The samples were measured with the optimised excitation wavelengths previously described in 3.1.1. After each sample was measured, the cuvette was cleaned with 100% methanol before reusing.
The resulting data were analysed with GraphPad Prism 7 (GraphPad Software Inc., USA) and Microsoft Office Excel 365 (Microsoft Corporation, USA) and checked for a correlation between the fluorescence intensity and DNA concentration.
3.2. Detection of TRF1 and TRF2 in blood and saliva samples
In this second section of the first part, the two protein markers TRF1 and TRF2 were investigated and their usability in a presumptive DNA quantification method should be evaluated.
3.2.1. Sample preparation
The detection of the TRF1 and TRF2 was divided in three experiments. In each of these three experiments, blood and/or saliva samples were used. The three experiments are built up on one another. After the first experiment did not show promising results, the experimental setup
9 was adapted, and the second experiment was done. The same procedure was followed after the second experiment and resulted in the third experiment. The difference of these experiments is mainly in the sample preparation. These differences are discussed in the following part.
The first experiments were conducted with previously frozen whole blood samples in three different dilutions (1:2, 1:10, 1:100). The whole blood samples were diluted with phosphate buffered saline (PBS) (Lonza, Belgium). Of the blood, 10 µl were pipetted on a glass slide (Superfrost, Gerhard Menzel GmbH, Germany) and spread with a second glass slide.
During the second experiment, blood and saliva samples of one volunteer were spread on glass slides. Saliva was collected by rubbing a cotton swab (Medeco B.V., the Netherlands) against the inner cheek of the volunteer. Blood was collected by using a blood lancet and 10 µl of blood were pipetted on the glass slide. The blood was again spread with a second glass slide.
In the third experiment, blood samples were excluded. Saliva samples of one volunteer were used. The samples were collected in the same way as previously described.
3.2.2. Immunolabelling
The prepared samples were treated with antibodies to detect TRF1 and TRF2. For this purpose the protocol of van Dam et al. (23) was adapted.
The detailed protocol can be found in the supplementary material. The blocking buffer consisted of PBS and 7% of skim milk powder (SMP) (Sigma Aldrich, Germany). After blocking, the primary antibodies (see table 2) were diluted 1:100 in blocking buffer and 150 µl were applied to the sample. For each sample either anti-TRF1 or anti-TRF2 was used as primary antibody.
Table 2 Specifications of the used primary antibodies
Primary Antibody
Host Anti Isotype Company
Anti-TRF1 (TRF-78)
Mouse Human IgG1 Monoclonal Santa Cruz
Biotechnology, INC, USA
Anti-TRF2 (9F10)
Mouse Human IgG1 Monoclonal Santa Cruz
Biotechnology, INC, USA
The samples were left for overnight incubation in a wet chamber. The following day, the samples were washed with PBS and incubated with 150 µl of secondary antibody (see table 3).
Table 3 Specifications of the used secondary antibody
The secondary antibody was diluted 1:250 in blocking buffer. After incubation and washing, a mounting medium (Vectashield mounting medium for fluorescence with DAPI, Vector
Secondary antibody
Host Anti Company
10 Laboratories Inc., USA) was applied to the samples and these were covered with a coverslip (Gerhard Menzel GmbH, Germany).
3.2.3. Additional permeabilization step
The results obtained by following the standard protocol for the three previously mentioned experiments, did not show a specific binding of the primary antibodies. In the following part, the used standard protocol is adapted in two different ways to add a permeabilization step. This additional step should help the antibodies to access the nuclei and subsequently the proteins more easily.
First, the two different solutions Triton X-100 (Sigma Aldrich, Germany) and Tween-20 (Sigma Aldrich, Germany) were added to the blocking buffer. Five different concentrations of Triton X-100 (0.1% to 0.5%) and one concentration of Tween-20 (0.1%) were tested. The experiment was conducted according to the standard protocol.
Secondly, another additional step was included before the fixation step of the standard protocol. Also here, two different solutions were tested, namely a solution of 0.001% pepsin (Sigma Aldrich, Germany) in 10 mM hydrochloric acid (HCl, Millipore, Merck KGaA, Germany) and a PBS solution containing 20 µg/ml of Proteinase K (Roche Diagnostics Deutschland GmbH, Germany). The samples in the pepsin solution were incubated for ten minutes at 37 °C. Samples treated with Proteinase K were incubated for 15 minutes at 37°C.
Additionally, the permeabilization step with Proteinase K was evaluated. For this purpose, Proteinase K was diluted in PBS to obtain three final concentrations of Proteinase K: 10 µg/ml, 20 µg/ml, 30 µg/ml. After incubation for 15 minutes at 37 °C, the samples were covered with a DNase mix (10 µl DNase + 90 µl RDD buffer (Qiagen, the Netherlands)) and incubated for 15 minutes at room temperature. To stop the reaction, the slides were heated up to 75 °C for ten minutes. Afterwards, the slides were covered with mounting medium containing DAPI and analysed using the fluorescence microscope. If the permeabilization step was successful, the DNase could enter the nuclei and cut the DNA into fragments. This can be visualised under the microscope as DAPI still binds to the DNA fragments.
3.2.4. Imaging of samples
The samples were analysed using a fluorescence microscope (Nikon Eclipse E600, Tokyo, Japan), a Cy3 filter (excitation filter: 510-560 nm, dichroic mirror: 575 nm, barrier filter: 590 m) and a Nikon DS-Fi2 camera.
3.2.5. Control experiments
As a positive control, a fingermark sample was stained with an anti-dermcidin antibody to check whether the immunolabelling method itself worked. The antibody was diluted 1:20 in blocking buffer. As mounting medium, the Dako Fluorescence Mounting Medium (Agilent Technologies, Inc., USA) was used. The specifications of the used primary antibody for the positive control experiment can be found in table 4.
Table 4 Specification of the primary antibody used as control experiment
Antibody Host Anti Isotype Company
Dermcidin G-81 Mouse Human IgM Monoclonal Santa Cruz Biotechnology,
11 The negative controls included the exclusion of the primary anti-TRF1 and anti-TRF2 antibody and the exclusion of both the primary antibody and the secondary antibody.
3.2.6. Western Blot
Another technique to identify proteins, is the use of a Western Blot.
For the Western Blot analysis, fresh saliva of a volunteer was collected in a 1.5 ml Eppendorf tube. The cell suspension was centrifuged at 2000G for five minutes at 4 °C. The supernatant was discarded and ice-cold PBS was added to the cell pellet to wash the cells. The sample was centrifuged at 2000G for five minutes at 4 °C. Ice-cold Radioimmunoprecipitation assay buffer (RIPA buffer) to lyse the cells, was added to the sample. This mix was than agitated for 30 minutes and subsequently centrifuged at 16000G for 20 minutes at 4 °C. The supernatant was used for the western blot. An in-house protocol was used for the Western Blot (supplementary material). The sample was 10-fold and 100-fold diluted with MilliQ water. A reducing urea sample buffer, containing β-mercaptoethanol, was used. Afterwards, the samples were mixed 1:1 with urea sample buffer and incubated for five minutes at 95 °C. Precasted gels (Mini-PROTEAN® TGX ™ Gels, Bio-Rad Laboratories, Inc., USA) and the Precision Plus Protein ™ WesternC™ Standards (Bio-Rad Laboratories, Inc., USA) were used for electrophoresis. A volume of 20 µl of sample and ladder were added to the gel. The gel blotting was done with the Trans-Blot® Turbo ™ Transfer System (Bio-Rad Laboratories, Inc., USA) using a Midi Transfer Pack. The detection of the blot was done by blocking the membrane for one hour with blocking buffer (Tris-buffered saline + 0.5% Tween-20 + 5% SMP) and afterwards incubating the membrane with primary antibody overnight. The primary antibodies were diluted 1:200 and the specifications can be found in table 5.
Table 5 Specification of the used primary antibodies
Primary Antibody
Host Anti Isotype Company
Anti-TRF1 (TRF-78)
Mouse Human IgG1 Monoclonal Santa Cruz
Biotechnology, INC, USA
Anti-TRF2 (9F10)
Mouse Human IgG1 Monoclonal Santa Cruz
Biotechnology, INC, USA
Table 6 Specifications of the used secondary antibody
Secondary antibody Host Anti Comany
Horse Radish Peroxidase
(HRP)
Goat Mouse Agilent Technologies, Inc., USA
After washing, the membranes were incubated with the secondary antibody (see table 6). The HRP antibody was diluted 1:2000 and a HRP conjugate (Precision Protein ™ StrepTactin-HRP conjugate, dilution 1:10000, Bio-Rad Laboratories, Inc., USA) was added. Enhanced chemiluminescence (ECL) was used to detect the secondary antibody. For this purpose, Clarity™ Western ECL Substrate (Bio-Rad Laboratories, Inc., USA) was used. The two components, peroxide solution and Luminol/enhancer solution, were mixed 1:1 and added to the membranes. The detection of bands was done with the ImageQuant LAS 4000.
12
Part 2: Immunolabelling of fingermarks and subsequent STR typing
3.3. Swab testing
The presence of background DNA was investigated on five different swabs which are described in detail in table 7.
Three different swabs (Sterile Cotton Swab, Disposable micro applicator, Cotton swab) were additionally tested for their efficiency in retrieving known amounts of DNA from glass slides (marked with * in table 7). The motivation to use these three specific swabs is the current and possibly future usage in this specific laboratory. The sterile cotton swab from Deltalab (Deltalab, Spain) and the cotton swab from Medeco (Medeco B.V., the Netherlands) are often used for swabbing different surfaces. The disposable micro applicator with nylon tip (Blink Beauty Parlour, Canada) is tested in the context of a future application in the laboratory.
Table 7 Overview of the five analysed swabs and their tip material
Name of swab Tip
Material
Company
Disposable micro applicator * Nylon Blink Beauty Parlour, Canada
Sterile Flocked Swab Polyester
Flock
Puritan Medical Products Co. LLC, USA
Cotton swab * Cotton Medeco B.V., the Netherlands
Sterile Cotton Swab * Cotton Deltalab, Spain
Sterile DNA-free Standard
Cotton Swab
Cotton Puritan Medical Products Co. LLC, USA
3.3.1. Cleaning protocol
Before the testing of the swabs, glass slides and DNA IQ™ spin baskets (Promega Corporation, USA) were cleaned by adapting the protocol established by Kanokwongnuwut et al. (14). Common household bleach was diluted to a 3% (v/v) solution. The glass slides and spin baskets were immersed in the bleach solution for ten minutes. Afterwards, the spin baskets were dipped in absolute ethanol and irradiated with UV light for 15 minutes from a distance of 3 cm. As UV light source, the Crime-lite 82S with a wavelength of 350 to 380 nm (Foster+Freeman Ltd, United Kingdom) was used.
3.3.2. Sample preparation for efficiency testing
Genomic male DNA (Promega Corporation, USA) with a stock concentration of 192 ng/µl was diluted to a final concentration of 0.1 ng/µl, 0.5 ng/µl, 1.5 ng/µl, 2.5 ng/µl and 10 ng/µl. A volume of 3 µl of DNA solution was pipetted on a cleaned glass slide and left to dry for 15 minutes. The experiment was performed in duplicates for each one of the three tested swabs (n=30).
3.3.3. Sampling background testing
For the analysis of background DNA, the tips of 23 swabs from five different types of swabs were wetted with MilliQ water (Millipore, Merck KGaA, Germany). Afterwards, the tips were removed and placed in 1.5 ml Eppendorf tubes containing DNA IQ™ spin baskets for subsequent DNA extraction. Table 8 shows the sample size for each swab.
13
Table 8 Overview over the sample size of the background testing of swabs
Name of swab Tip Material Sample size
Disposable micro applicator Nylon 5
Sterile Flocked Swab Polyester Flock 4
Cotton swab Cotton 5
Sterile Cotton Swab Cotton 5
Sterile DNA-free Standard Cotton Swab Cotton 4
3.3.4. Sampling efficiency testing
Each swab was wetted with MilliQ water and the sample area was swabbed in a standardised manner. Sampling with the cotton swabs was done by swabbing ten times horizontally and ten times vertically. Due to the smaller size of the swab, sampling with the nylon swab was done by swabbing 15 time horizontally and 15 times vertically. The tip of the swab was removed and placed in a 1.5 ml Eppendorf tube containing a DNA IQ™ spin basket.
3.3.5. DNA extraction
The DNA was extracted using the DNA IQ™ System (Promega Corporation, USA) with DNA IQ™ spin baskets and following the manufacturer’s protocol. Before the extraction, 1,4-Dithiothreitol (DTT, Sigma Aldrich, Germany) was added to the lysis buffer from the kit. The final concentration of DTT was 60 mM. The extracted DNA was eluted in 40 µl of elution buffer. The DNA extraction step was included due to the fact, that it is not possible to do a DNA quantification with a direct PCR. A direct PCR is later used for the generation of STR profiles in section 2.4.
3.3.6. DNA quantification
The extracted DNA was quantified with the Qubit® Fluorometer 3.0 (Thermo Fisher Scientific, USA) and the Qubit™ dsDNA HS Assay Kit (Thermo Fisher Scientific, USA). For each run, 10 µl of sample or negative control were added to 190 µl of working solution. As negative controls, MilliQ and the elution buffer of the DNA IQ™ System kit were used.
3.4. Immunolabelling of fingermarks on 3 different substrates with subsequent STR
analysis
3.4.1. Substrate preparation
Three different types of substrates were used in the experiment and can be seen in table 9. The substrates can be divided in two categories: non-porous surfaces and semi-porous surface.
Table 9 Overview of the used substrates and their classification in non-porous and semi-porous surfaces
Type of surface Substrate
Non-porous surface Aluminium foil
Glass
14 The glass slides were cleaned according to the protocol of Kanokwongnuwut et al. (14). The white tile was wiped with 3% (v/v) bleach solution and absolute alcohol. The substrate aluminium foil was not cleaned since it was directly removed from the packaging.
3.4.2. Fingermark collection
Three volunteers (two male and one female) participated after signing an informed consent form and deposited their fingermarks during working hours. Each volunteer deposited three fingermarks on each of the three different substrates (n=27).Before deposition, no instructions were given to the volunteers.
3.4.3. Immunolabelling of fingermarks on non-porous and semi-porous surfaces
All 27 fingermark samples were included in the immunolabelling process. The used protocol was adapted from the protocol of van Dam et al. (23). The detailed protocol can be found in the supplementary material. For the substrate tile, the fixation step with ice cold methanol was excluded. The blocking buffer consisted of PBS and 7% SMP. After blocking, the mouse monoclonal IgM anti-dermcidin (Santa Cruz Biotechnology, INC, USA) primary antibody was diluted 1:20 in blocking buffer and 150 µl were applied to the sample. The samples were left for overnight incubation in a wet chamber. The following day, the samples were washed with PBS and incubated with 150 µl of secondary antibody. The secondary antibody goat anti-mouse Cy3 (Brunschwig Chemies, the Netherlands) was diluted 1:500 in blocking buffer. After incubation, the samples were washed and air-dried before imaging.3.4.4. Control experiments
As positive control two to three spots of 1 µl of dermcidin peptide (0.1 mg/ml, Abgent Inc., USA) were added on each substrate and let dry (n=3). Untreated fingermarks were included as negative control for the immunolabelling. These fingermarks did not undergo any treatment with antibodies. The untreated fingermarks were also used to check for possible PCR inhibitors during STR typing, which may be introduced by the immunolabelling. For each substrate, one fingermark per donor was included as negative control (n=9).
3.4.5. Imaging of the sample
After the samples have dried, images were taken with a Canon Eos 40D (Canon, Japan). A green Crime-lite® 2 torch (500-560 nm, Foster+Freeman Ltd, United Kingdom) was used to excite Cy3 and fluorescence was detected with a red filter (OG590 Brigth Red, 571 nm; Foster+Freeman Ltd, United Kingdom) in front of the camera. At least one overview image (camera distance to the sample: 65 cm) and one close-up image (camera distance to the sample: 55 cm) was taken from each sample.
3.4.6. DNA sampling
The DNA collection procedure and the protocol for direct PCR were adapted from a previously published protocol by Kanokwongnuwut et al. (14).
After imaging, the fingermarks were sampled for DNA. The nylon tip of a disposable micro applicator (Blink Beauty Parlour, Canada) was wetted with 2 µl of a 2% Triton-X 100 (v/v) solution and the entire fingermark was swabbed.
3.4.7. Direct PCR and STR analysis
After swabbing, a direct PCR was conducted. The direct PCR was performed with the AmpFLSTR™ NGM™ PCR Amplification Kit (Thermo Fisher Scientific, USA) and a total volume of 25 µl. A 0.2 ml PCR tube was prepared for the direct PCR by adding 2 µl Prep-n-Go™ buffer (Thermo Fisher Scientific, USA) and 8 µl of low TE buffer (pH 7.55). The tip of the
15 swab, which was used for DNA sampling, was cut off and placed in the PCR tube. Afterwards, the tube was placed on a shaker for ten minutes to extract DNA from the swab. After shaking, a PCR mix of 5 µl Primer Set and 10 µl Master Mix was added to each tube. The amplification was performed on the Biometra T personal thermocycler (Westburg BV, the Netherlands) with a total of 29 cycles.
Due to the use of a direct PCR, no DNA extraction is required and the swab-containing solution can directly be used for PCR. The choice for a direct PCR is due to the possible loss of DNA during the extraction process. In the case of fingermarks, it is important to use all available template DNA for the analysis.
Control experiments
As positive control, the AmpFLSTR™ DNA Control 007 (Thermo Fisher Scientific, USA) was used. To check for possible PCR inhibitors, which may be introduced during immunolabelling, untreated fingermarks were analysed. Additionally, two swabs without sample material were added to the PCR mix to serve as negative control.
STR analysis
For capillary electrophoresis of the DNA fragments, a mix of 17.4 µl HiDi™-Formamide (Thermo Fisher Scientific, USA) and 0.6 µl of GeneScan™ 500 LIZ™ dye size standard was added to 2 µl of PCR product. Data analysis was done with the GeneMapper™ Software v4.1 (Thermo Fisher Scientific, USA). After the generation of a STR profile, the profiles were compared to reference profiles of the three volunteers.
16
4. Results
Part 1: Presumptive DNA quantification
4.1. Detection of DNA with DNA-binding dyes
4.1.1. DAPI
The first step of the experiment, was the optimisation of the settings of the fluorescence spectrometer to obtain the best results for the measurements. This process resulted in an optimal excitation wavelength of 348 nm and an emission slit width of 10nm. With these settings, the following results were obtained.
After the optimisation, emission spectra of three different sample sets were generated. The analysis of the obtained emission spectra resulted in a range of emission wavelength, where the maximum peak height occurs, from 468.5 nm to 480 nm. Sample set 1 and 2 were measured at the same day. They showed a similar distribution and range of wavelength. A slightly shifted range showed sample set 3, which was measured at another day. The exact ranges of the three sample sets can be seen in the following table 10.
Table 10 Emission wavelength range of DAPI for the three sample sets
Sample set Wavelength range of maximum peak height of all samples
1 468.5 nm - 474.5 nm
2 469 nm - 475.5 nm
3 472.5 nm - 480 nm
In two of the three runs, all samples could reach fluorescence intensities above the level of the blank sample. The sample with a DNA concentration of 2.5 ng/µl could not reach above the emission spectrum of the blank sample in sample set 3 (see supplementary material). For the quantification of the amount of DNA present in a sample, it is necessary to distinguish the sample signal from the background signal of the blank sample. Another requirement for quantifying DNA, is the need for a correlation between the DNA concentration and the fluorescence signal. This can be measured by a correlation coefficient.
To visualise a possible correlation, the fluorescence intensity is plotted against the DNA concentration (see image 1).
17
Image 1 Overview of the maximum intensities of DAPI for each DNA concentration.
The results of the three sample sets vary for each measured DNA concentration. This is also visible in the correlation between the fluorescence intensity and the DNA concentration of each sample set: sample set 1 showed a moderate correlation (R²=0.51), sample set 2 (R²=0.2143) showed a weak correlation and sample set 3 (R²=0.0025) showed no correlation. It is noteworthy, that in sample set 3, samples with a DNA amount of 1 ng/µl or smaller, showed considerable higher fluorescence intensities compared to samples with a DNA amount of 2.5 ng/µl or higher (see image 2).
Image 2 Overview of the maximum intensities of DAPI of sample set 3. Samples below the dashed line cannot be distinguished from the blank sample
For several samples, the variation between the fluorescence intensity for the three sample sets is larger than for other samples. The following image shows an example for a large and a small variation between the emission spectra.
18
Image 3 Variation in measurement between the three sample sets for two different concentrations. The left side shows the comparison of emission spectra of DAPI for a sample with 1.0 ng/µl of DNA present. The right image shows the comparison of emission spectra of DAPI for a sample with 2.5 ng/µl of DNA present. The two DNA concentration show a different variation in measurement of emission.
It was not possible to establish a consistent correlation of the fluorescence intensity and the DNA concentration over all sample sets.
4.1.2. Diamond Dye
Fluorescence measurementAlso, for this dye, the first step was the optimisation of the setting of the fluorescence spectrometer and the determination of the best excitation wavelength. The best results were achieved with an excitation wavelength of 346 nm.
After the generation of emission spectra, the emission wavelength, where the maximum peak height occurs, was evaluated. These wavelengths ranged from 441 nm to 469 nm. Sample set 1 and 2, again, showed a similar distribution and range of wavelength, as previously described in section 4.1.1. The range of sample set 3, which was again measured at another day, showed a slightly larger range. The exact ranges of the sample sets can be seen in the following table 11.
Table 11 Emission wavelength range of Diamond Dye for the three sample sets
Sample set Wavelength range of maximum peak height of all samples
1 442 nm - 459 nm
2 442 nm - 459 nm
3 441 nm - 469 nm
Additionally, a blue shift was observed between the emission spectra of the blank samples and the emission spectra of the DNA-containing samples.
In all three conducted runs, several samples could not reach fluorescence intensities above the level of the blank sample (see supplementary material). For the quantification of DNA, which is present in a sample, it is necessary to distinguish the sample signal from the background signal of the blank sample.
To establish a correlation between the DNA concentrations and the fluorescence intensity, the maximum peak height is plotted against the DNA concentration (see image 4).
19
Image 4 Overview of the maximum intensities of Diamond Dye for each DNA concentration
All three sample sets show very different correlations, from a strong correlation in sample set 3 (R²=0.97) to a very low and no correlation in sample sets 1 (R²=0.13) and sample set 2 (R²=0.03) respectively. This is also visible in the varying results of the three measurements for each DNA concentration, as the example shows (see image 5)
Image 5 Comparison of emission spectra of samples with a DNA concentration of 10 ng/µl.
However, the appearance of the emission spectra and the fluorescence intensity of the samples in correlation with the fluorescence intensity of the blank sample, needs to be taken into account. In sample set 2, none of the samples could reach above the fluorescence intensity of the blank sample, as is visible in image 6.
20
Image 6 Overview of the maximum intensities of different DNA concentrations in sample set 2. Samples below the dashed line cannot be distinguished from the blank sample
Also several samples in both sample set 1 and 3, could not reach above the fluorescence intensity of the blank sample (see image 7).
Image 7 Overview of the maximum intensities of different DNA concentrations in sample set 1 (left) and sample set 3 (right). Samples below the dashed line cannot be distinguished from the blank sample
Overall, it was not possible to establish a consistent correlation of the fluorescence intensity and the DNA concentration over all sample sets. Currently, the measurements show a wide variation and no reproducibility.
Absorbance spectra
For the measurements of Diamond Dye, the setting of the optimisation step which gave the best results, differed from the manufacturer’s specification. The manufacturer states an excitation wavelength of 494 nm and an emission wavelength of 558 nm.
The settings which gave the best results for the measurements of Diamond Dye differed from the manufacturer’s specification, which state as an excitation wavelength 494 nm and as an emission wavelength of 558 nm. Therefore, an absorption spectrum of Diamond Dye was made. The absorbance measurement did not reveal a clear absorbance spectrum at any time point. It was not possible to determine an excitation wavelength from the absorbance spectrum.
21 Image 8 gives an example of the measured absorbance at t=3 minutes. The absorbance spectra at the remaining time points had a similar appearance.
Image 8 Absorbance spectrum of Diamond Dye at t=3 minutes
Excitation – Emission – Matrix
The EEM shows a clear pattern which indicates an excitation wavelength of around 370 nm with a corresponding emission wavelength of around 465 nm. This is indicated by the red area with a maximum intensity around 600 RFU. The measured EEM of TE buffer, DNA and dye is illustrated in image 9.
Image 9 EEM of TE buffer, DNA and Diamond dye. The red area indicates the wavelengths with the highest fluorescence intensity of around 600 RFU
At the excitation wavelength of 494 nm, which is proposed by the manufacturer, no fluorescence intensity could be measured (crossing of white lines). The EEM of the other samples also showed a measured fluorescence intensity in the same area as is visible in the
22 image. However, these fluorescence intensities are distinguishable lower with a maximum intensity of 90 to 100 RFU (supplementary material).
4.2. Detection of TRF1 and TRF2 in blood and saliva samples
4.2.1. Immunolabelling
The first experiment with previously frozen whole blood showed no specific binding of anti-TRF1 and anti-TRF2 antibodies. However, it did show a high auto-fluorescence of red blood cells and destroyed white blood cells (see image 10). Therefore, the second experiment was conducted with fresh blood and saliva samples.
Image 10 Blood (dilution 1:2) stained with Anti-TRF1, Cy3 and DAPI, 40x magnification. Combined imaged of the Cy3 and UV channel. No specific binding in the nuclei visible
This second experiment with fresh blood and saliva samples, showed similar results to the first experiment. Except the cells, stained with DAPI, were intact. No specific binding could be observed. Again, a high auto-fluorescence of red blood cells was observed. Therefore, whole blood samples were excluded from the following experiments. Image 11 shows a blood sample stained with anti-TRF1. Image 12 shows a saliva sample stained with anti-TRF1.
23
Image 11 Blood stained with Anti-TRF1, Cy3 and DAPI, 40x magnification. Combined image of the Cy3 and UV channel. No specific binding in the nuclei visible. High background fluorescence of the red blood cells
Image 12 Saliva stained with anti-TRF1, Cy3 and DAPI, 40x magnification. Combined image of the Cy3 and UV channel. No specific binding in the nuclei visible.
Saliva samples, which have a high number of nucleated cells, were stained in the third phase of the experiments (see image 13 and 14). Here as well, no specific binding of the antibodies could be observed. Specific binding of TRF1 and TRF2 would occur in the nuclei and should therefore only be visible in the nuclear region of the cell. This was not the case.
24
Image 13 Saliva stained with anti-TRF1, Cy3 and DAPI, 40x magnification. Combined image of the Cy3 and UV channel. No specific binding visible in the nuclei.
Image 14 Saliva stained with anti-TRF2, Cy3 and DAPI, 40x magnification. Combined image of the Cy3 and UV channel. No specific binding visible in the nuclei.
The methanol fixation step in the standard protocol permeabilises the cells. However, different approaches to permeabilise the cell membrane were investigated. None of the tested approaches, described in section 3.2., did result in specific binding and detection of the antibodies.
The testing of the permeabilization step with Proteinase K was investigated by using DNase. If the concentration of Proteinase K is sufficient to permeabilise the cell membrane, DNase can excess the cell and cut the DNA. With a DAPI staining, the different amount of cut DNA per Proteinase K concentration could be visualised (see image 15). This showed that 10 µg/ml of Proteinase K was enough to cut all present DNA. The fact that there were no intact nuclei visible verifies the success of the permeabilization step.
25
Image 15 Saliva samples treated with 10 µg/ml (A), 20 µg/ml (B) and 30 µg/ml (C) Proteinase K, stained with DAPI, visualised in the UV channel with 40x magnification, No intact nuclei visible. The permeabilization step was successful.
4.2.2. Western Blot
The proteins TRF1 and TRF2 could not be detected in saliva samples by using a western blot. No bands were visible (see image 16). TRF1 has a molecular weight of 60 kDa and TRF2 has a molecular weight of 70 kDa.
Image 16 Western Blot of saliva stained with anti-TRF1 (left) and anti-TRF2 (right). No specific bands are visible
Overall, no specific binding of TRF1 and TRF2 could be detected. Therefore, it is not possible to draw a conclusion whether DNA-binding proteins can be used for a presumptive quantification of DNA.
26
Part 2: Immunolabelling of fingermarks and subsequent STR typing
4.3. Swab testing
4.3.1. Analysis of background DNA on five different types of swabs
The swabs are divided in two different categories, cotton tipped and nylon tipped swabs. These two categories showed a difference in presence of background DNA. The cotton swabs showed generally a higher quantity of background DNA than swabs with a nylon tip. The results are represented in table 12.
Table 12 Results of the quantification of background DNA on five different types of swabs.
Tip material
Swab type Company Number
of
samples
Mean DNA concentration [ng/µl] ± SD
Nylon Disposable micro
Applicator
Blink Beauty Parlour, Canada
5 0 ± 0
Sterile Flocked swab
Puritan Medical Products Co. LLC, USA
4 0.0032 ± 0.0065
Cotton Cotton swab Medeco B.V., the
Netherlands 5 0.0481 ± 0.0287 Sterile cotton swab Deltalab, Spain 5 0.0107 ± 0.0091 Sterile DNA-free Standard Cotton Swab
Puritan Medical Products Co. LLC, USA
4 0.0521 ± 0.0293
The only swab without detectable background DNA was the disposable micro applicator from Blink Beauty Parlour, Canada.
4.3.2. Efficiency of three different swabs
The three most commonly used swabs in this laboratory were chosen for efficiency testing. A known quantity of DNA was placed on a glass slide. After swabbing and extraction of DNA, the retrieved amount of DNA was compared to the initial quantity that was placed on the glass. This experiment was conducted in duplicates for each concentration.
None of the three tested swabs was able to retrieve significant amounts of DNA from the glass slides. The retrieved amount of DNA can be seen in table 13 and is not distinguishable from the background DNA, which was discussed in the previous section 4.1.1.
27
Table 13 Results of the testing of efficiency of three different types of swabs
Swab type Company Deposited total DNA quantity Mean retrieved DNA quantity [ng/µl] ± SD Efficiency [%] after correction Cotton swab Medeco B.V., the Netherlands 30 ng 7.5 ng 4.5 ng 1.5 ng 0.3 ng Background 0.0555 ± 0.0064 0.0351 ± 0.0161 0.0166 ± 0.0028 0.0229 ± 0.0073 0.0216 ± 0.0037 0.0481 ± 0.0287 7.39 18.72 14.76 60.93 287.33 Disposable micro Applicator Blink Beauty Parlour, Canada 30 ng 7.5 ng 4.5 ng 1.5 ng 0.3 ng Background 0.0083 ± 0.0117 0.0092 ± 0.0129 0 ± 0 0.0098 ± 0.0138 0.0148 ± 0.0134 0 ± 0 1.11 4.88 0.00 26.00 197.33 Sterile cotton swab Deltalab, Spain 30 ng 7.5 ng 4.5 ng 1.5 ng 0.3 ng Background 0.0367 ± 0.0018 0.0154 ± 0.0028 0.0168 ± 0.0078 0.0185 ± 0.0013 0.0169 ± 0.0048 0.0107 ± 0.0091 4.89 8.19 14.93 49.33 225.3
The mean retrieved DNA quantity was corrected for the added volume of elution buffer (40 µl). The total efficiency was calculated by the following equation:
𝐸𝑓𝑓𝑖𝑐𝑖𝑒𝑛𝑐𝑦 =𝑚𝑒𝑎𝑛 𝑟𝑒𝑡𝑟𝑖𝑒𝑣𝑒𝑑 𝐷𝑁𝐴 𝑞𝑢𝑎𝑛𝑡𝑖𝑦 ∗ 40µ𝑙 𝑡𝑜𝑡𝑎𝑙 𝑑𝑒𝑝𝑜𝑠𝑖𝑡𝑒𝑑 𝐷𝑁𝐴 𝑞𝑢𝑎𝑛𝑡𝑖𝑡𝑦 ∗ 100
The efficiency test, did not reveal a constant retrieval efficiency. The efficiency results of the samples with a low DNA concentration showed a large influence of background DNA on the results, as the efficiency exceeds 100%.
For the following experiments, the disposable micro applicator (Blink Beauty Parlour, Canada) was chosen because, it did not show measurable background DNA. Based on the results of the efficiency determination, no choice could be made.
4.4. Immunolabelling of fingermarks on three different substrates and subsequent
STR profiling
4.4.1. Immunolabelling results
For the immunolabelling experiment three fingermarks were depostited on each substrate (glass, metal, tile) by each volunteer. Two of these three fingermarks were treated with antibodies during the immunolabelling experiment. One fingermark was not treated with antibodies and is referred to as untreated fingermark.
Untreated fingermarks
The untreated fingermarks of volunteer 1 showed a faint auto-fluorescence on glass and metal and no visible auto-fluorescence could be detected on tile. Of volunteer 2, the untreated fingermarks showed a moderate fluorescence on glass and tile and a strong
auto-28 fluorescence on metal. The untreated fingermarks of volunteer 3 showed a faint auto-fluorescence on tile, a moderate auto-auto-fluorescence on glass and a strong auto-auto-fluorescence on metal (see image 17).
Image 17 Untreated fingermarks of volunteer 1 (left), 2 (middel) and 3 (right). The substrates can be seen on the left side.
Treated fingermarks
All treated fingermarks showed a specific binding of anti-dermcidin antibodies to the fingermarks on all substrates. This can be seen in the following images. On the left side, overview images can be seen. In the middle and on the right close-up images can be seen.
30
31
4.4.2. STR profiling results
The analysis of the profiles showed a problem with the fluorescence signal of the PET® dye (red dye). Therefore, the loci, stained with this dye (D2S441, D3S1358, D1S1656, D12S391), were excluded from the interpretation. This resulted in twelve analysed loci with a maximum of 24 possible alleles for one donor.
Control experiment
Both positive controls, consisting of the AmpFLSTR™ DNA Control 007, showed the expected STR profile. The two negative controls, blank swabs, showed a total of 21 and 18 alleles. The majority of these alleles were present in two loci (D10S1248, D22S1045). The negative controls showed different profiles and varied in maximum peak height. The following detail image 21 shows the difference in maximum peak intensity and difference in detected alleles.
Image 21 Locus D22S1045 of negative control 1 (left) and 2 (right)
Samples
A total of 27 profiles from nine untreated fingermarks and 18 treated fingermarks were generated. Each profile was compared to the reference profile of the donor and the allele retrieval was calculated (see supplementary material). Per volunteer and per substrate, each profile was categorised in one of the five categories according to table 14.
Table 14 Five categories to classify the obtained results
Category Features of the category
Full profile All 24 alleles of the donor were detected and
the donor profile can be identified
Partial profile More than 12 alleles of the donor were
detected; other alleles can be present as major or minor contributor; the donor profile can be identified on several loci
Inconclusive profile Less than 12 alleles of the donor were
detected
Degraded profile The profile shows signs of degradation
Contaminated profile A clear contamination is visible, for example
33 A summary of the categorisation of fingermarks can be seen in the following images.
Image 22 Categorisation of the fingermarks per donor
Image 23 Categorisation of the fingermarks per substrate
A total of four (14.81%) full and 14 (51.85%) partial profiles could be identified. Two full profiles were generated from fingermarks deposited on metal and two full profiles were generated from fingermarks deposited on glass. Volunteer 2 was the donor of three of the four full profiles, which is also reflected in the later discussed mean retrieval rate.
Two profiles (7.41%) were inconclusive. Three samples (11.11%) showed signs of degradation. Four samples (14.81%) showed signs of contamination. The four contaminated samples showed a Y chromosome in the profile of a female donor.
The three different substrates showed different retrieval rates as did the three donors. The following table 15 summarises the results.
34
Table 15 Overview of the mean retrieval rate of the three substrates tile, metal and glass
Donor Mean retrieval rate [%]
Tile 62.50
Metal 85.71
Glass 79.69
Substrate Mean retrieval rate [%]
Volunteer 1 62.96
Volunteer 2 89.81
Volunteer 3 71.30
Substrate Donor Mean retrieval rate
[%] Tile Volunteer 1 29.17 Volunteer 2 83.33 Volunteer 3 52.78 Metal Volunteer 1 91.67 Volunteer 2 94.44 Volunteer 3 75.00 Glass Volunteer 1 68.06 Volunteer 2 91.67 Volunteer 3 86.11
Tile showed the worst mean retrieval rate (62.50%) of the three substrates. The substrates glass and metal both had a high mean retrieval rate of 79.69% and 85.71%. The highest mean retrieval rate was 89.80 % of volunteer 2. The deposited fingermarks of volunteer 1 on the substrate tile had the worst mean retrieval rate of 29.17%. The best mean retrieval rate of 94.44% was reached by the fingermarks deposited by volunteer 2 on metal.
It is necessary to mention, that an allele is considered as detected when it is present in the profile independent of its RFU. In a number of profiles, the retrieved alleles showed a lower RFU value than other alleles present in the profile. To following image 24 illustrates this case. The green boxes mark the alleles of the donor. In locus vWA, the major profile would be 14/17 instead of the expected 16/17 of the donor. Therefore, the profile of the donor cannot be detected on this locus, but both alleles could be retrieved.
35
Image 24 Example of sample Glass 3_1. Green boxes symbolise the alleles of the donor.
In those cases, the profile of the donor cannot be detected as major donor on all loci. Untreated fingermarks
Untreated fingermarks showed a significantly higher RFU value than the treated fingermarks from the same volunteer on the same substrate. On the substrate glass, the RFU values of all three samples reached above 10000 RFU. On metal and tile, the untreated fingermarks of volunteer 2 reached above 20000 RFU, whereas the untreated fingermarks of the other volunteers on the same substrate showed RFU values of 2000 to 3000 RFU. Image 25 shows an example profile of an untreated fingermark of volunteer 2 on metal. The profile was in agreement with the reference profile of the donor and all donor alleles could be retrieved. The interpretation of the profile can be seen in the following image 25 and table 16. On all loci the donor alleles could be identified as the major contributor.
36
Image 25 STR profile of the untreated samples Metal 2_3
Table 16 Interpretation of the profile Metal 2_3 and the comparison to the reference profile of volunteer 2
D10S1248 vWA D16S539 D2S1338 AMEL D8S1179 D21S11 D18S51 Sample profile 14/15 16/20 9/11 17/25 X/Y 9/14 29/30 12/17 Reference profile 14/15 16/20 9/11 17/25 X/Y 9/14 29/30 12/17 D22S1045 D19S433 TH01 FGA Sample profile 11/16 12/15 6/9.3 22 Reference profile 11/16 12/15 6/9.3 22
37
5. Discussion
Like the previous parts, the discussion is divided in two separate parts: part 1, presumptive DNA quantification and part 2, immunolabelling of fingermarks and subsequent STR typing.
Part 1: Presumptive DNA quantification
In this first part, the results of the experiments for a presumptive DNA quantification method are discussed. This section is split in two parts, namely the DNA-binding dyes and the detection of the proteins TRF1 and TRF2.
DNA-binding dyes
In the first section, the research question focussed on the applicability of DNA-binding dyes and whether they can be used in a presumptive quantification method. During the experiments, promising results could be achieved. However, it is not yet possible to establish a linear correlation between the fluorescence signal of DAPI or Diamond Dye and the DNA concentration of a sample. This will now be discussed in more detail.
To quantify DNA by using a DNA-binding dye, a correlation between the fluorescence signal and the amount of DNA should exist. Therefore, the two different dyes, DAPI and Diamond Dye, were tested for their possible correlation and the results are discussed in the following section.
The correlation between the fluorescence signal of DAPI and the DNA content of cells, was already discussed and investigated 20 years ago. Kobayashi et al. (28) investigated human megakaryocytes and their DNA content in healthy people and patients with myelodysplastic syndrome. In these diseased people, the DNA amount is reduced in the megakaryocytes and a difference should be visible in the measured fluorescence intensity. With the use of a DAPI staining and the measurement of the fluorescence intensity, a linear relation between the amount of DNA, present in the megakaryocytes, and the fluorescence intensity could be established (28). This linear relation is supported by the specification of the manufacturer of DAPI (29).
However, this linear correlation could not be detected in the results of the conducted measurements. The fluorescence intensities, measured within the samples with the same DNA concentration, showed a high variation. Therefore, also the correlation of the fluorescence intensities and the DNA amounts varied between the three conducted measurements. DAPI has two different binding modes to DNA (30). DAPI primarily binds to the AT-rich regions of the DNA. But it can also bind to GC-rich regions of the DNA. The binding to AT-rich regions results in the fluorescence signal, which is measured during the experiments. When the dye binds to GC-rich regions, no fluorescence signal can be measured. Banerjee and Pal (30) have proven that the intercalative binding to GC-rich regions does not play a major role in the binding of DAPI to human DNA. Therefore, the different binding modes can be excluded as the cause of the varying results.
An issue, which always needs to be taken into account, is the possibility of contamination with foreign DNA. During the preparation of the samples, foreign DNA could have been introduced and increased the total amount of DNA present in the samples.
The current results lead to the following recommendations for future experiments; it is highly recommended to establish a consistent experimental setup to avoid variations within samples,
38 with the same DNA concentration. Furthermore, additional measurements and a larger number of samples with a greater variety of DNA concentrations are recommended for the future. The reason for this is that according to the literature (28,29), a linear relationship between the fluorescence signal and the DNA concentration of a sample should exist. With the recommended adaptions and the promising results from this experiment, it might be possible to establish a linear relationship in future experiments.
Other than DAPI, Diamond Dye is a relatively new DNA-binding dye. First of all, it was striking that Diamond Dye did not show the maximum fluorescence intensity at the expected emission wavelengths (31), when generating an EEM. At the excitation and emission wavelength, proposed by the manufacturer Promega, no fluorescence intensity could be detected. Also the absorbance measurement did not reveal a clear absorbance peak, which can indicate the best excitation wavelength. To check for a possible influence of the spectrometer and the cuvette on the measurement, the absorbance of indigo carmine, a blue solution, was measured. The absorbance spectrum was as expected. This could exclude a failure of equipment and the divergent results are due to the measured sample solution. The excitation wavelength of 346 nm, which showed the best results during the optimisation steps, generated clear peaks and was therefore used for the following experiments.
The exact binding chemistry and composition of Diamond Dye is protected by the manufacturer. Therefore, it is not possible to make a statement about an influence of the dye chemistry on the results. So far, little is known about the behaviour of the dye in the presence of varying amounts of DNA. As a result, no statement can be made about a possible linear correlation of the fluorescence signal and the DNA concentration. Different studies used Diamond Dye to stain hair follicles (7) and touch DNA (32) or to stain gels for gel electrophoresis (33) and use it in a quantitative PCR (34). Until now, there was no study found which used the measurement of the fluorescence signal of the dye to establish a correlation to the DNA concentration.
Comparable to the previous results of DAPI, the conducted measurements of diamond dye did result in different correlations between the fluorescence signals of the samples and the DNA concentrations. However, the measurement of sample set 3 showed a very good correlation, which is promising for the future. It should be recommended to investigate a smaller range for a linear correlation between the fluorescence signal and the DNA concentration, as done by Bruijns et al. (35). They showed that some cyanine dyes only showed linearity in a certain range of DNA concentration. This can also be the case for Diamond Dye. With the used sample set, it was not possible to investigate a smaller range due to the limited number of samples. It is recommended to expand the experiment to more samples with varying concentrations of DNA. Also, the use of TE buffer, which was used for the dilution of the DNA, needs to be reconsidered, as it showed a rather high fluorescence signal at the same emission wavelength as Diamond Dye. The cause for the differences in proposed excitation and emission wavelength by the manufacturer and the measured excitation and emission wavelength needs to be investigated.
Protein marker detection
After the approach to quantify DNA by DNA-binding dyes, the use of protein markers for the quantification of DNA is discussed. The following part investigates the question, whether it is possible to use the two proteins markers, TRF1 and TRF2, in a presumptive quantification method for DNA. To use these markers for quantification, it is necessary that the amount of
