TH _ TTH¥HS QOCC ‘I' D c . ‘ _.‘ Jv’ ' This is to certify that the thesis entitled AN EVALUATION OF THE EFFICACY OF USING CELLULAR MATERIAL PELLETED DURING SEROLOGICAL TESTING FOR FORENSIC DNA TYPING presented by JULIE LYNN BARDUGONE has been accepted towards fulfillment of the requirements for M. S . degree in CRIMINAL JUSTICE WITH SPECIALIZATION IN FORENSIC SCIENCE 0M @M/ I a Major prof sor Date E/13}03 0-7639 MS U is an Affirmative Action/Equal Opportunity Institution LIBRARY Michigan State University PLACE IN RETURN BOX to remove this checkout from your re To AVOID FINES return on or befc“ rIare MAY BE RECALLED with earlier as I DA" AN EVALUATION OF THE EFFICACY OF USING CELLULAR MATERIAL PELLETED DURING SEROLOGICAL TESTING FOR FORENSIC DNA TYPING By Julie Lynn Bardugone A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE School of Criminal Justice 2003 ABSTRACT AN EVALUATION OF THE EFFICACY OF USING CELLULAR MATERIAL PELLETED DURING SEROLOGICAL TESTING FOR FORENSIC DNA TYPING By Julie Lynn Bardugone The cellular material conserved during routine tests used to identify the origin of stains encountered in forensic testing could be a rich source of DNA. If scientists were able to obtain DNA profiles from extracted serological material, it would improve overall efficiency and reduce the amount of evidentiary stained material consumed during testing. Preliminary forensic serological examinations rely on soluble proteins associated with cellular secretions, such as seminal acid phosphatase (AF) and prostate specific antigen (p30), both present in semen. Historically, the serologist re-solublized a small portion of the stain to extract the soluble proteins and pelleted cellular debris associated with the stain by centrifugation. This research was designed to evaluate DNA yield and DNA typing results from the pelleted cellular debris associated with semen identification methods, as well as correlate serological test results and recovery of complete DNA profiles. Serial dilutions of neat semen deposited on small clippings of material (swatches) were tested for AP, p30, and visually graded based on the number of sperm present. The DNA in the pelleted cellular debris remaining after these tests was organically extracted, quantified, and typed using STR markers. The ability to generate complete DNA profiles from the pelleted material will help scientists to streamline the processing of evidence between stain identification and DNA testing. ACKNOWLEDGEMENTS I would like to thank the Michigan State Police Lansing Forensic Laboratory for allowing the use of their laboratory to perform this research. All scientists in the Biology Unit offered endless help in answering questions and assisting me through the laboratory work. I would to like to especially thank the Biology Unit Supervisor, Charles Barna, and Forensic Scientist Glen Hall. Dr. Christina Dejong provided knowledge and guidance in the application of a statistical model to the results of this research. I would also like to acknowledge the assistance of Dr. Jay Siegel who guided me through my entire degree as well as the attainment of my firstjob. iii TABLE OF CONTENTS LIST OF TABLES .............................................................................................................. v LIST OF FIGURES ........................................................................................................... vi INTRODUCTION .............................................................................................................. 1 MATERIALS AND METHODS ........................................................................................ 8 RESULTS ......................................................................................................................... 16 DISCUSSION. .................................................................................................................. 29 BIBLIOGRAPHY ............................................................................................................. 39 iv LIST OF TABLES Table 1: Serological Testing and DNA Typing Results for All Samples ......................... 17 Table 2: Slot Blot DNA Quantitation Estimates for each Sample .................................... 20 Table 3: Pearson Correlations Between All Serological Tests, Estimated Quantity of DNA, and DNA Typing ................................................................... 28 LIST OF FIGURES Figure 1: Acid Phosphatase Reaction ................................................................................. 4 Figure 2: COfiIer Electropherogram of Sample 65 with Alleles 50-150 RFUs .............. 21 Figure 3: Results of Serological Tests and DNA Quantity Compared to the Number of Loci with Alleles Above 150 RF Us for Each Dilution. .................. 23 Figure 4: Results of Serological Tests and DNA Quantity Compared to the Number of Loci with Alleles Above 50 RFUs for Each Dilution. .................... 23 Figure 5: Profiler Plus Electropherogram of Sample 96 Showing Secondary Alleles .............................................................................. 25 Figure 6: Profiler Plus Electropherogram of Sample 96 After Re—amplification ..................................................................................... 26 vi Introduction The processing of forensic evidence can be crucial to solving a crime by establishing a link between a suspect and a crime scene or victim. Biological fluids such as blood, semen, and saliva are routinely transferred during the commission of a crime. Forensic DNA testing of biological samples has transcended more conventional methods of genetic characterization (e.g. blood typing); however, collection and identification of biological stains remains unchanged. Several forensic stain identification techniques measure soluble proteins associated with cellular secretions and are non-destructive to DNA. Routinely, cellular debris is pelleted during these stain identification tests to clear the supernatant, which contains the targeted soluble protein. This cellular debris has the potential of being a rich source of cells from which DNA can be extracted. Successful DNA typing is the result of recovering sufficient DNA from a biological stain. Forensic stains are often limited in size; therefore, the conservation of sample is critical for DNA typing. Preliminary identification tests may offer insight into the quantity of DNA expected to be recovered from a stain and supply pelleted cellular material not commonly used for DNA recovery. If DNA analysis is probative for the case, it is crucial that an adequate sample of the stain is available for testing. Therefore, a judgment, based on preliminary serology tests, must be made as to the usefulness of the sample for DNA typing. In the Michigan State Police Forensic Laboratory, as well as many laboratories across the country, scientists at different laboratories may perform the serological testing and the DNA typing of a stained evidentiary item. Historically, samples used for preliminary serological tests are not conserved to maintain the pelleted cellular debris. With the advent of the polymerase chain reaction, DNA typing can be successful using a very small quantity of input DNA, making the cellular material associated with preliminary stain identification tests sufficient for DNA typing. Therefore, taking an additional cutting of a stained evidentiary item for DNA typing may be unnecessary. This research evaluates the DNA recovery and profiling results from cellular material pelleted during stain identification testing for semen. Serial dilutions of semen were deposited on swatches of cotton fabric. The swatches were exposed to a presumptive acid phosphatase test, a confirmatory test for p30, and a microscopic examination for sperm, which are typical serological tests for semen. The DNA in the pelleted cells was used for DNA testing instead of removing an additional cutting from the stained item. DNA in the pelleted cells was extracted, quantified, amplified, and a profile was recorded. The completeness of the profile was compared to the serological testing results to conclude the efficacy of using DNA from pelleted cellular material for DNA testing in the Michigan State Police Forensic Laboratory. 1. Acid Phosphatase Test (AP) The AP test is a presumptive, chromogen test used to detect the presence of acid phosphatase (Saferstein, 1988). Kutscher and Wohlbergs (1935— reference in German; Hansen, 1947; Gaensslen, 1983) first described AP as an enzyme found in male ejaculate that hydrolyzes various phosphate esters at an acid pH optimum. They originally named it “prostate phosphatase.” Gutman and Gutman (193 8) studied the increase of AP in patients with metastasizing prostate carcinomas and suggested using an assay for serum AP for diagnosing prostate cancer. Multiple studies in the mid-19405 by scientists in Denmark lead to the suggestion of using the large quantities of AP found in male ejaculates as the basis for identification of seminal fluid in medico-legal situations (Hansen, 1947; Lundquist, 1950; Walker, 1950; Gaensslen, 1983). Acid phosphatase activity found in human semen is much higher than any other bodily secretion. The expected AP levels in seminal fluid range from 70-370 U per 100 cc (Gutman and Gutman, 193 8), while the average AP levels in serum are 0.5 to 2.5 U per 100 cc (Gutman and Gutman, 1940). Although other bodily fluids may also contain AP, the concentration of AP found in semen is so much greater than other fluids that it is sufficient for presumptive testing (Saferstein, 1988). Different methods exist for the detection of AP. The method used in this study involves the interaction of a naphthylphosphate with acid phosphatase, which liberates the phosphate group. Napthol then reacts with a diazonium compound to cause a color change (Kind, 1964; Gaensslen, 1983). In this experiment, seminal AP reacted with sodium alpha napthyl acid phosphate (reagent 1) to liberate free alpha napthol. Diazo Blue-B (reagent 2) was then added to react with the free alpha napthol to immediately form an insoluble purple-colored dye (Figure 1). II. Prostate Specific Antigen (p30) The prostate specific antigen test is a confirmatory test used to detect the presence of p30 (PSA), a soluble protein that originates from the epithelial cells lining the prostate. A Japanese scientist, Hara (1 971 — reference in Japanese), first identified p30, calling it y-seminoprotein (Armbruster, 1993). Li and Beling (1973) also identified p30, calling it Protein E. It was not until Sensabaugh performed a complete characterization of Figure 1: Acid Phosphatase Reaction Ii /P—O- + O \ - Na OH O . 0 Acid Phosphatase H | H+ OH Sodium alpha-napthyl phosphate Alpha-napthol Phosphoric Acid OH NZNH + “CO "‘ ”Z” Alpha-napthol Diazo Blue B Insoluble Dye the protein, that the name p30 was attached (Sensabaugh, 1978; Gaensslen, 1983; Armbruster, 1993). Sensabaugh (1978) identified p30 as a suitable marker for seminal fluid identification. Electrophoresis was used to compare proteins found in seminal fluid to those found in other physiological fluids by separating denatured polypeptide chains based on molecular weight. A strong band was present at 30,000 daltons in seminal fluid, but not in other physiological fluids; therefore, the glycoprotein was named p30. Sensabaugh concluded that p30 was a robust and reliable marker for seminal fluid identification because 1) it demonstrated the biological basis of specificity 2) it was a component of seminal plasma so it could still be detected in vasectomized and aspermic males 3) it was stable in stains and the vaginal tract and detectable at trace levels. P30 detection relies on the formation of an anti gen-antibody complex with anti- p30. Historically, antigen-antibody precipitin methods have been used to visualize these reactions. Different techniques have been applied to this method to enhance visualization and sensitivity (Gaensslen, 1983; Saferstein, 1988). Diffusion techniques that rely on the movement of antisera and samples in an agarose matrix have been used successfully. Also, immunoelectrophoresis techniques (e. g. Cross-over electrophoresis and Rocket immunoelectrophoresis) were employed to improve sensitivity and the ability to detect the precipitate (Saferstein, 1988). Disadvantages of the immunological tests described above were that they were tedious and difficult to visualize with low titer samples (Hochmeister et al., 1999). A new approach was developed by Abacus Diagnostics, OneStepABAcard® p30 Test For the Forensic Identification of Semen, which exploited the formation of the p30 antigen-antibody complex. Benton et al. (1998) and Hochmeister et al. (1999) reported on the novel application of the ABAcard as a substitute for the more classical approaches to p30 testing. The ABAcard test employs the use of a mobile monoclonal anti-human p30 antibody-dye conjugate that reacts with p30 to form a mobile antigen—antibody complex. This complex attaches to an immobilized monoclonal anti-human p30 antibody creating an "antibody-antigen-antibody sandwich”, and concentrating dye particles to form a visible line along the reaction zone. Unbound mobile monoclonal anti-human p30 antibodies migrate on the membrane to a control zone, where immobilized antibodies reside, forming a visible line along the control zone (Hochmeister et al., 1999; Abacus Diagnostics, Inc., 2001). A visible line on both the control zone and the reaction zone is indicative of a positive result for p30 and confirmation of human seminal fluid. Hochmeister et a1. (1999) concluded that the test did not interfere with standard DNA testing, making it compatible with forensic biology testing. 111. Slide Staining and Grading The identification of sperm via microscopic examination has been employed since van Leeuwenhoek first described sperm morphology in 1677 (Gaensslen, 1983). The visual identification of sperm cells began to improve when color stains were first used in 1867, and continual improvements of chemical formulations were made in 19705 (Gaensslen, 1983). Oppitz (1969) described a staining procedure utilizing Nuclear Fast Red, “kernechtrot”, and indigo carmine in picric acid. This differential staining technique colors the sperm heads red and the sperm tails green, leading to the common name of Christmas Tree Stain (Oppitz, 1969). IV. DNA profiling Current forensic DNA profiling methods at the Michigan State Police utilize short tandem repeats (STRs; Edwards, et al., 1991; Edwards, et al., 1992; Hammond, et al., 1994; Butler, 2001) in conjunction with the polymerase chain reaction (PCR; Mullis, 1990). PCR is a powerful biotechnology tool that allows for amplification of target DNA sequences. STRs are easily replicated through PCR due to their repeat units of only 2—6 base pairs of length and have been used in nuclear forensic DNA analysis (Lygo et al., 1994; Gill et al., 1995; Moretti et al., 2001). 9 Commercial STR amplification kits include primers that allow multiplexing, amplifying multiple loci in one reaction tube. For purposes of making associations between DNA profiles, 13 loci were established as the core loci acceptable for entry of a profile into the National Combined DNA Index System (CODIS; Budowle et al., 1998; Budowle et al., 1999). PE Applied Biosystems AmpFZSTR® Profiler PlusTM and COfilerTM STR Multiplex kits utilized in this research contain primers that amplify the 13 core loci and have been validated for this use (LaFountain et al., 2001; Frank et al., 2001). This research evaluates the procedure of extracting DNA from pelleted cells after serological tests, and obtaining a profile sufficient for entry into CODIS. Materials and Methods 1. Preparation of samples Seventy-five white, 100% cotton, 1 cm2 swatches were utilized as a medium for semen stains. Fresh liquid semen from one donor was supplied by the Michigan State Police. It was serially diluted with distilled water, creating 14 dilutions ranging from 1:2 to 1:163 84. A control sample was also made with the distilled water used to make the ' dilutions. Fifteen pL of each dilution was applied via pipette to a cotton swatch; five replicates were made for each dilution. Four swatches of each dilution were placed in 1.5mL tubes (lid open) and allowed to air-dry overnight in a clean hood. One swatch per dilution was placed into a manila coin envelope and stored at -20°C. All samples were assigned a computer-generated random number. Concurrently, 15uL of each semen dilution was deposited in the middle of a 15mm-etched circle on a glass microscope slide. The slide, referred to as the pre-pellet slide, was dried on a heat plate at the low setting (approximately 75°C) and stored in a conventional slide box. II. Serological Tests An AP test was performed on each swatch. A moistened swab was pressed to each swatch and 1—2 drops of sodium alpha napthyl acid phosphate dissolved in a acetate buffer (1 g sodium alpha napthyl acid phosphate; 20g sodium acetate, 5mL acetic acid in IL distilled water) were dropped onto the swab. Directly following, 1—2 drops of Diazo Blue B (4.0g napthanil Diazo Blue B dissolved in acetate buffer) were added to the swab. The development of a purple color on the swab after the addition of the reagents was noted. The AP test results were recorded as one of the following: (8+) strong positive ——an immediate, bright purple color change, (+) positive ——a purple color change, (W+) weak positive —a delayed, light purple color change, (-) negative —no color change. The OneStepABAcard® p30 Test For the Forensic Identification of Semen was used after the swatch was tested for AP. The swatch was soaked in 750uL of deionized water in a Spin-EaseTM extraction tube with basket for two hours at room temperature. The tube was vortexed and centrifuged at 3500 X g for 3 minutes. The swatch was placed in the basket and centrifugation was repeated. SOOuL of supernatant was pipetted into a 1.5mL tube and stored at -20°C. 200uL was pipetted into the sample well of the p30 test device. The result was read after ten minutes. Purple lines visible across both the test and control regions of the test device indicated a positive result. The remaining 50uL of liquid was used to resuspend the cellular material pelleted in the bottom of the extraction tube via manual pipette aspiration. Fifteen p.L of the total 50uL (30%) was deposited in the middle of a lSmm—etched circle on a glass microscope slide, called the post-pellet slide, and dried on a heat plate at the low setting. The swatch and remaining liquid and pellet were stored at -20°C until DNA analysis. Both slides (pre-pellet and post-pellet) were stained using the Christmas Tree differential staining method. Red-colored Kemechtrot dye stain was applied to the slide for 30 minutes, and then rinsed with distilled water. Next, green-colored picroindigocarmine dye stain was applied for 30 seconds and rinsed with a solution of 90% EtOH, 5% MeOH, and 5% IPOH. The slides were air dried and then graded. The slides were scanned using a compound microscope at 400x power. A grade was assigned to each slide based on the following criteria: _G_ra_50 RFUs Loci (150 RFU): The number of loci with alleles >150 RFUs Profiler Plus: Results from the Amth’STR Profiler Plus Amplification Kit COfiler: Results for the Amth’STR COfiler Amplification Kit 3p.L sample: The sample was prepared using 3uL of amplified product (instead of 1 uL) 10 sec: The sample was injected for 10 seconds (instead of 5 seconds) Y: Complete profile (all alleles at 13 loci were >150 RFUs) P: Partial profile (alleles at 1—12 loci were >150 RFUs) N: No profile (no alleles at any locus were >150 RFUs) N/A: DNA typing was not performed on that sample 17 Dilution Sample p30 AP pZIIIet :51th 1:13:50 “113,50 P3313?” giggle 10 sec COfiler Sixth 103cc + + 4+ + + + 13 + 3+ 1 + + + 4+ 3+ + + + + 2+ + ++++++++++++++++++ I++I 18 11. DNA Quantitation and Typing The quantity of DNA extracted from each sample in ng/uL is displayed in Table 2. All samples from the 1 :2 dilution set contained >2ng/ttL of DNA. A designation of <0.03125ng/uL was applied to samples that produced a visible band but fell below the lowest standard included in the Quantiblot kit. No DNA was visible for any samples in dilution sets 1:4096 through 1:16384. DNA typing results are displayed in Table 1. Two data sets were developed by declaring alleles using different minimum thresholds, 50 RFU or 150 RFU. Figure 2 shows an electropherogram of sample 65 with peak heights between 50—150 RFUs. Using an allele declaration threshold of 150 RFU, complete 13 loci profiles were constructed for dilution sets 1:2 through 1:64. Sample 15 (l :32 dilution) required 3 pL of amplified product to generate a complete profile. Mixed results were obtained from the 1:128 dilution set. Complete profiles were developed from 3 samples; one of the samples required luL of amplified product, and two samples required 3 1.1L of amplified product. The fourth sample (76) was processed using Profiler Plus only and gave a partial profile using 3uL of amplified product. One of the four samples in both dilution sets, 1:256 and 1:512, had one locus with alleles >150 RF Us. In general, the STR loci with larger size amplicons (e. g. FGA and D18) were the first loci to drop below the allele declaration threshold. No alleles were called for dilution sets 1:1024 through 1:163 84. Using an allele declaration threshold of 50 RFUs, 13 loci were detected for all samples in dilution sets 1:2 through 1:64. Partial profiles were obtained for samples in dilution sets 1:128, 1:256, 1:512, 1:1024, and 122048, two dilutions farther than at 150 19 Table 2: Slot Blot DNA Quantitation Estimates for each Sample Dilution Sample Dilution Sample ng/uL ‘1227 96 , 1:256 23 <0.03125 1:2 . 1:256 53 <0.03125 1:2 17 12256 67 0.03125 1:2 38 . 1:256 65 0.0625 1:4 47 12512 20 <0.03125 1:4 61 12512 77 <0.03125 1:4 1:512 11 ND 1:4 56 . 1:512 58 0.0625 1:8 49 . 1: 1024 97 <0.03125 1:8 8 . 1:1024 26 ND 1:8 79 1:1024 44 ND 1:8 42 1:1024 95 ND 1216 91 1:2048 7 <0.03125 1:16 80 1:2048 24 ND 1:1 1 ~ 1:2048 29 ND 1:1 39 122048 82 ND 40 . 124096 85 ND 6 . 1:4096 45 ND 5 0.5 1:4096 72 ND 98 0.5 124096 63 ND 14 0.0625 128192 78 ND 73 0.1 128192 25 ND 52 0.1 1:8192 57 . ND 0.1 128192 2 ND 0.04 1216384 69 ND 0.03125 1216384 27 ND 0.125 1:16384 1 ND 0.1 5 1:16384 59 ND N/A = Quantitation was not performed on that sample ND = No DNA was visually detected using the slot blot quantitation method 20 Figure 2: COfiler Electropherogram of Sample 65 with Alleles 50-150 RFUs I I I I I I l I I I I I I I ' I I I I I I I I I I I | I I I I I I I I I I I I I I I I I I I I I I 100 120 140 160 180 200 220 240 260 280 300 320 G3-65COJB 43 Blue GSCO-JB 100 50 GS-SSCOJB 43 Green GSCO-JB 150 100 50 223 29 - GS-6500-J8 43 Yellow SSCO-JB 90 60 30 I 1 268 60 276 81 t1 l I 1 :1 (33650048 43 Rod 65CO-JB 2000 1500 1000 500 1 1 _ 21 RFUs, and no alleles were called at 124096, 128192, or 1:16384. No alleles were called in the negative control samples at either threshold of 50 or 150. Comparison of the serological test results, DNA quantities, and number of complete loci in each DNA profile are graphed in Figures 3 and 4. Each point represents an average of the 4 samples tested within a particular dilution set. Results from allele designations made using the declaration thresholds of 150 RFU and 50 RFU are displayed in Figure 3 and Figure 4, respectively. The effectiveness of obtaining a complete profile with alleles >150 RF Us after performing additional manipulations was also evaluated (Table 1). Data were collected on 12 samples for Profiler Plus, and 5 samples for COfiler. Ten of the 12 samples for Profiler Plus gave the same results for both manipulations, either generating full profiles, more alleles above 150 RF Us, or no profile. In the remaining two samples, the 10-second injection generated partial profiles while the 3uL amplified product preparation did not. For the COfiler samples, 3 of the 5 gave the same results for both manipulations. The 10— second injection generated one partial profile and one complete profile in the remaining two samples when compared to the 3uL amplified product preparation. 22 Figure 3: Results of Serological Tests and DNA Quantity Compared to the Number of Loci with Alleles Above 150 RFUs for Each Dilution Dilution +p30+AP+Slide—-—DNA(150RFU)+DNAQuantity Figure 4: Results of Serological Tests and DNA Quantity Compared to the Number of Loci with Alleles Above 50 RFUs for Each Dilution I I I . ' i} r It‘d" .2443“ O fiiuv‘firf‘fif’t‘rm- 1 l 1 Dilution 23 Minor secondary alleles were called in sample 96 when a threshold of 50 RF U was used (Figure 5). Additional alleles were present at D3 (18, 102 RFU), FGA (22, 102 RFU), and D8 (16, 59 RFU). Stutter peaks that were <20% of the major peak at a position of n-4 for any locus were disregarded when profiles were evaluated. Additional peaks, not in the stutter position, that could not be attributed to a nonreproducible artifact (e.g. fluorescent spike) were considered contamination. The source of the secondary allele contribution could not be identified because of insufficient genetic data. Sample preparation was repeated using the original amplified product, and the extracted sample was re-amplified and run on the 310 Genetic Analyzer. The repeated preparation using the original amplified product yielded a profile with the same additional alleles. The secondary alleles were not present when the sample was re-amplified from the original extract (Figure 6). 24 Figure 5: Profiler Plus Electropherogram of Sample 96 Showing Secondary Alleles I I | I I I ' I I I I l I I I I I I j I I I | l I I I I I I I l I I l I I I l 100 120 140 100 100 200 220 240 200 280 300 A5-96-PP-JB a Blue 96-PP-JB 2000 1500 1000 500 AS-96-PP-JB a Green ss-Ppoa 1000 500 A5-96-PP-JB 3 Yellow 96-PP~JB 900 600 A A 300 . a m 1 268 40 -_.J 276 45 [7127-71. ~_;:. AS-QB-PP-JB a Red 96-PP-JB i I 1500 i I 1000 I: II Ii J\ ___J L J I. :11 ' i i:,___i ___jit ‘W .._._..J 25 Figure 6: Profiler Plus Electropherogram of Sample 96 After Re-amplification 100 120 140 100 150 200 220 240 260 280 300 320 H248 96 PP 19 Blue JB 96 PP 2000 1500 1000 500 1000 500 H2-JB 96 PP 19 Yellow JB 90 PP 1000 500 I: 1" l 11:1 .m l", . H208 96 PP 19 Red J8 96 PP ‘ i g i g , 1500 J ‘ l I F 5 l l ,’ g: :3 5 mm :23 ‘3 ll 21 {3‘ li Ii 500 _..,j\._,__.__..___-__~__J L..,_... \.___J‘. .5 1’ \ j k 26 III. Statistical Results Correlations were calculated between serological tests, the estimated quantity of DNA, and the number loci with alleles >50 RFU for each dilution set (Table 3). The strongest correlation was between the AP and p30 tests (app3 0), rapp3o = 0.982 (p<.001). The second strongest correlation was between AP and the slide grade (aps), raps = 0.951 (p<.001). The quantity of DNA estimated for each sample was most strongly correlated to AP test results (apqd), rapqd = .943 (p< .001), and mostly weakly correlated to the number of loci called for each DNA profile (qdd), rqdd = .698 (p=.006). All correlations between serological tests were stronger than the correlation between any of the serological tests and the completeness of the DNA profile. The strongest correlation with the DNA typing was the slide grade (sd) with rsd = .881 (p<.001). The correlation between the p30 test and the DNA (p30d) was rpm = .850 (p=.001), and the weakest correlation with serological tests was between the AP test and the DNA results (apd) with rapd = .815 (p=.001). 27 Table 3: Pearson Correlations Between All Serological Tests, Estimated Quantity of DNA, and DNA Typing p30 AP Slide DNA DNA Grade Quantity Typin p30 Pearson Correlation .982 .935 .891 .850 P-Value <.001 <.001 <.001 .001 AP Pearson Correlation ...... .951 .943 .815 P-Va'ue <.001 <.001 .001 Slide Pearson Correlation .926 .881 Grade p-value <.001 <.001 DNA Pearson Correlation .698 Quantity p-value .006 DNA Pearson Correlation Typing p-value 28 Discussion The research described here was designed to evaluate DNA typing results from pelleted cellular material associated with semen identification methods, as well as correlate quantitative serological test results with DNA yields and the recovery of complete 13 loci DNA profiles. To mimic forensic samples, serial dilutions of semen deposited on swatches were exposed to a presumptive acid phosphatase test, a confirmatory test for p30, and a microscopic examination for sperm. DNA quantitation and typing was performed on the DNA contained within the pelleted cells. The completeness of the profile was compared to the serological testing results to determine the efficacy of using DNA from pelleted cellular material recovered from standard semen identification tests for DNA testing in the Michigan State Police Forensic Laboratory. 1. Correlation of Serological Tests to Completeness of DNA Profile Detectability of both p30 and AP began to decrease in dilution set 1:32. One sample out of four showed a negative (p3 0) or W+ (AP) result. With the exception of one sample, which gave a weak positive, both AP and p30 tests were negative for all samples in dilution set 1:128 and all subsequent sets. These similarities are reflected in the strong correlation between p30 and AP. As shown in Table 3, the identification of spermatozoa on the slide was most strongly correlated with a complete DNA profile. This strong correlation is explainable because DNA is located within the sperm heads, while p30 and AP test for proteins and are not directly associated with sperm. The AP test demonstrated the weakest correlation between serological test results and a complete DNA profile, while the p30 correlation with DNA typing was only slightly stronger. Spermatozoa were observed at dilution sets 29 S 1:128, when AP and p30 tests were negative. While experiments comparing the sensitivities of serological tests demonstrate p30 can be detected when no sperm are visible (Benton et al, 1998; Hochmeister et a1, 1999), those experiments were conducted under different sample conditions. Hochmeister et al. (1999) tested the sensitivity the p30 test using a straight dilution of semen and Benton et a1. (1998) deposited semen dilutions on swabs; both used varying amounts of the dilution to test for p3 0. In this experiment, only 200 of the 75 OuL of water used to extract p30 was pipetted into the test device. This was done to mimic the protocol followed at the Michigan State Police (and also the manufacturer’s guidelines), and a comparison of the sensitivity thresholds of the serological tests was not a main focus of this experiment. It is possible that the excess water diluted the soluble protein; therefore, sperm were present on one of the slides in each dilution set, with the exception of 1 :1024, after p30 tested negative. Counter intuitively, the correlation between DNA quantitation estimates and DNA typing was the weakest. This finding results from the manner in which the data were analyzed. As expected, the quantity of DNA decreased with each dilution (Table 2); therefore, only the first dilution was at the highest level of 4 (Figures 3 and 4). With each subsequent dilution, the quantity of DNA, and therefore the numerical value, decreased stepwise. In contrast, many of these dilutions contained ample DNA (lng) to produce a complete profile; therefore, the DNA typing results remained at the highest level for several dilutions. Because the DNA quantity was only at the highest level (4) for one dilution, it was least correlated with all of the other tests. The DNA yield and DNA typing results show the weakest correlation because the DNA concentrations do not affect the DNA typing result until less than lng of DNA is recovered. 30 Overall, there was some variability among the 4 samples tested within each dilution (Table 1). Not surprisingly, the results from all tests decreased with serial dilutions. Test results for p30, recorded as only a positive or negative, showed the least variability within a dilution set. The AP results were slightly more variable due in part perhaps to the fact that there were additional gradations of a positive result (S+, +, W+). The pre-pellet slide grade showed low variability within a dilution set, and all grades through the 1:256 dilution set were 4+. This may be due to the fact that a 4+ grade was recorded for all slides that reached or surpassed the criteria of having many sperm in each field of view. The post-pellet slide grade appeared to be the most variable serology test within a dilution set, especially when directly compared to the pre-pellet grade. This may be explained due to the multiple different slide grades available to differentiate lower amounts of sperm on the slide, and the manual aspiration of the pelleted material may not have uniformly mixed the sperm cells. The DNA quantitation results showed some variability within a dilution set (Table 2), but this variability was expected because the quantitation was a visual estimate. The DNA typing results were the most variable at dilution sets less than 1:128. At 1:128 and all subsequent dilutions, the 4 samples differed in the number of alleles that were called above each threshold. II. Issues for Consideration The various serological tests used in this experiment measure different seminal fluid components, and the different measurements may have an impact on the results. For example, the assignment of a strong positive, positive, or weak positive during an AP test was a subjective measurement. This subjectivity is not problematic in forensic testing because AP is only a presumptive test to determine if additional work is warranted. 31 Either a positive p30 test or the microscopic observation of sperm can confirm the presence of seminal fluid or sperm, respectively. The p30 test result was objective; it was either positive or negative when the result was read after 10 minutes. The slide evaluation was also objective because sperm were either observed or they were not. The grading technique, based on the number of sperm per field of view, as adapted from Kind (1964), is subjective, however, because the criteria is at the discretion of the scientist. For example, one scientist may consistently assign slide grades higher than another scientist. The quantity of DNA expected from a 1+ sample, for that scientist, would be less than that expected from the other. Again, this is of little concern in forensics, as the presence of even a single sperm indicates that semen is present. The relationship of the 2 subjective tests and subsequent DNA typing results is important however. The lack of homogeneity of semen stains should also be considered when applying results from this study to forensic evidence. Sperm cells, AP, and p30 are not evenly distributed in semen stains (Hansen, 1947; Mann, 1954); therefore, one region of a stain could have more sperm, containing DNA, than another region. This issue should be taken into account when evaluating the strong correlation between the slide grade and DNA typing. For slide grading, one sperm present on the slide indicates a source of DNA. Both the AP and p30 tests assay soluble proteins instead of a DNA source. Due to the lack of uniformity of semen stains, it is possible to visualize a sperm cell extracted from fabric that tested negative for both AP and p30. Likewise, it is possible to obtain a DNA profile when the stain tested positive for AP and p3 0, but no sperm were observed on the slide. 32 Another potential problem arises from the stochastic effects that can result when less than lng of DNA is added to a PCR reaction. The creation of serial dilutions was performed to mimic the wide range of samples submitted to forensic laboratories for DNA typing. When sub-optimal amounts of DNA are added to a PCR reaction, the equilibrium of the individual chemical reactions associated with copying the target sequence may be disrupted. This can result in allelic imbalance, allele dropout, or complete failure at a locus. General PCR performance observed in this experiment demonstrates that DNA target concentration influenced peak height (RFUs) of the alleles. The samples that contain very small quantities of DNA, and thus most commonly manifest stochastic effects, may be the most useful in determining the efficacy of using the pelleted cellular material for forensic DNA typing 0f evidentiary samples. The generation of a complete DNA profile was initially determined using an allele declaration threshold of 150 RF Us because this is the standard for casework analysis by the Michigan State Police. Lowering the RF U threshold to 50 allowed more differentiation between a partial profile and inconclusive results. For example, one sample in the 1:256 dilution (sample 65) only had 1 allele above the 150 RFU threshold, but the sample had alleles at 12 core loci above the 50 RF U threshold. The COfiler electropherogram for sample 65 (Figure 2) demonstrates alleles that would not be recorded using the 150 RFU threshold. When using the allele declaration threshold of 50 RF Us, at least one allele was detected for one or more samples in the dilution sets down to 1:2048, two gradations more dilute than detected using the 150 RFU threshold (Figures 3 and 4). Declaring alleles using a 50 RF U threshold supplies additional information regarding the potential 33 of generating a full profile. This information was collected in anticipation of technological advances allowing the obtainment of DNA profiles from very small amounts of DNA. The DNA that was recovered from samples that generated partial profiles (alleles 50—150 RFUs) during this experiment may be sufficient to generate complete profiles with the use of future technology. For example, Butler (2001) discussed advancements requiring less input DNA, including single nucleotide polymorphisms (SNPs) which may prove useful in combination with additional markers (e. g. Y-STRs) for use in forensic applications. The use of this technology would allow profiles to be generated from pelleted material that did not generate filll profiles during this experiment. Therefore, the construction of profiles with alleles below 150 RFUs may provide information that will be applicable to future techniques. In addition to the anticipation of technological advances, the knowledge that allele designations were made between 50—150 RFUs was more informative in determining the probability of obtaining a full genetic profile for CODIS. Currently, additional manipulations, such as longer injection times, may allow a full profile to be obtained from alleles between 50-150 RFUs. Also, if additional evidentiary material exists, the scientist may take another cutting in attempt to construct a profile with more alleles. Although 14 dilutions of semen were tested to compile preliminary data regarding the efficacy of using pelleted material for DNA testing, all samples were serial dilutions of the same specimen. The semen for this research was assumed to contain normal concentrations of p3 0, AP, and sperm. Results regarding the sensitivities of each serological test could provide misleading conclusions when this data is applied to evidentiary samples because only one semen donor was used. While this issue should be 34 considered, using additional donors will not affect the correlation between the slide grade and DNA typing results. Regardless of the sperm concentration of the semen donor, a slide grade (e.g. 1+) will be the same, and the relationship to the amount of DNA present will also be the same. Therefore, the use of a slide grade to predict the completeness of a DNA profile based on this experiment will not change with the addition of multiple semen donors. III. Additional Sample Manipulations The data collected to compare the effectiveness of using a 3uL preparation of amplified product and 10-second injections were an adjunct to the research. The additional manipulation of preparing samples using 3uL of amplified product is a standard protocol at the Michigan State Police. The 10-second injection is an alternative manipulation that may generate the same or better results. Manipulation techniques yielded the same number of loci with allele calls over 150 RFUs in 13 of 17 samples. In the remaining 4 samples, more alleles above 150 RFUs were called using the 10-second injection when compared to the 3uL amplified product preparation. Advantages for using the 10-second injection, in addition to obtaining more alleles, include time efficiency and elimination of additional liquid sample handling steps. A major disadvantage is that the data collected from lO-second injections may have poor peak resolution resulting from long electrokinetic injections. The lack of resolution could cause heterozygous alleles to appear as one peak in situations when the alleles are very close together. For example, alleles that are one base apart, such as 9.3 and 10 at Th0], may not be resolved during longer injection times. Therefore, extra caution must be 35 employed during the interpretation of allele designations that are made from long injections. IV. Conclusions DNA typing from the cellular material pelleted during forensic stain identification testing for semen can yield complete DNA profiles. This indicates that the pelleted cells are rich sources of DNA and are not compromised during stain identification testing. Using the pelleted material will allow less stained evidentiary material to be consumed during the culmination of forensic biological testing for one evidentiary item. The data from this study can be used to guide serologists in determining whether additional cuttings from the stained material are necessary to generate complete genetic profiles. While working with known dilutions of semen is not feasible in forensic situations, the results of this experiment can be used to predict the likelihood of obtaining a DNA profile from conserved material generated during standard serology testing. Using these guidelines, scientists can streamline the processing of forensic evidence. In all circumstances, using the least amount of stained material is desirable in preserving evidentiary material for additional testing. The cells pelleted after forensic serological testing may provide a rich source of DNA for forensic testing, allowing for the conservation of evidentiary items. V. Future Research Based on results from this research, a piece of evidentiary material that has a positive p30, an AP result of 8+, +, or W+, and sperm visible on the microscope slide, will provide pelleted cells with sufficient DNA for generating a full profile. On the other hand, if both the p30 and the AP tests are negative, and the slide grade is a 1+ or lower, a 36 partial profile is expected to result. In this instance, it would be beneficial to take an additional cutting and add it to the pelleted material before submitting the sample for DNA testing. It should be noted that this basic guideline was determined using the results of one semen specimen deposited on cotton material,_following current protocols at the Michigan State Police. In order to generate a more complete set of guidelines, further research should be undertaken. The continued research of this topic using additional substrates would be beneficial to represent practical evidentiary items received at crime laboratories. The cotton material used allowed for the absorption of the semen without the formation of a crusted semen stain. The absorption of the semen may cause the cells containing DNA to be tightly bound in the fabric matrix. Partial extraction of the cells from the fabric during serological testing may result in lower slide grades if the sperm cells are still bound in the fabric. A crusted semen stain superficial to the fabric, that could result with a fabric such as suede, may cause cellular material to be removed from the fabric when the moistened swab is pressed to the swatch during AP testing. The slide grade, in this instance, might be more accurate because the cellular material would be superficial to the fabric, and therefore, more easily removed. The slide stained for microscopic evaluation was made by aspirating the pelleted cellular material. Any sperm that were still bound to the swatch would not have been considered when assigning the slide grade. AP and p30 tests would presumably not be as affected by poor extraction from the fabric because both test for soluble proteins. When using the differential organic extraction protocol, the original swatch as well as the pelleted cellular material was extracted. The DNA typing result was a combination of 37 both the original swatch and the cellular material. Performing this experiment by extraction of only the pelleted cellular material would provide data as to the efficiency of cellular extraction from the fabric during serological testing. Additionally, to minimize the impact of the fabric type on the results, the post- pellet slide could be made after the initial differential extraction. The detergent and proteinase K added during DNA extraction are intended to lyse the epithelial cells and remove additional cellular material from the fabric. If the post-pellet slide was made after this step, the slide grade would be representative of all the sperm cells associated with that sample, not only the cells extracted from the fabric during serological testing. Another issue to consider is that the conclusions from this research are based on semen dilutions made from one donor. Using multiple donors would be useful to account for differences in AP, p30, and sperm cell concentrations among individuals. The use of multiple semen donors would allow for accumulation of additional data that would support this experiment. In summary, additional research investigating the use of multiple substrates and semen donors would be beneficial to support the results of this experiment. 38 BIBLIOGRAPHY Abacus Diagnostics, Inc. 0neStepABAcard® p30 Test for the Forensic Identification of Semen. Technical Information Sheet. 2001. Armbruster, D. A. Prostate—Specific Antigen: Biochemistry, Analytical Methods, and Clinical Application. Clin Chem 1993; 39(2): 181-195. Bachman, R. and R. Paternoster. Statistical Methods for Criminology and Criminal Justice. The McGraw-Hill Companies, Inc. 1997. Benton, 'K. A., Donahue, J. A., and M. Valadez, Jr. Analysis of the ABAcard® OneStep PSA Test For Use in the Forensic Laboratory. Presented at the 1998 Spring Meeting of SWAFS. Budowle, B., Moretti, T. R., Niezgoda, S. J ., and B. Brown. CODIS and PCR-based Short Tandem Repeat Loci: Law Enforcement Tools. In: Second European Symposium on Human Identification. Promega Corporation, Madison, Wisconsin. 1998; 73-88. Budowle, B., Moretti, T. R., Baumstark, A. L., Defenbaugh D. A., and K. M. Keys. Population Data on the Thirteen CODIS Core Short Tandem Repeat Loci in African Americans, US. Caucasians, Hispanics, Bahamians, Jamaicans, and Trinidadians. J Forensic Sci 1999; 44(6): 1277-1286. Butler, .1. M. Forensic DNA Typing: Biology & Technology behind STR Markers. Academic Press. 2001. Edwards, A., Civitello, A., Hammond, H., and C. Caskey. DNA Typing and Genetic Mapping with Trimeric and Tetrameric Tandem Repeats. Am J Hum Genet 1991; 49: 746-756. Edwards, A., Hammond, H., J in, L., Caskey, C., and R. Chakraborty. Genetic Variation at Five Trimeric and Tetrameric Tandem Repeat Loci in Four Human Population Groups. Genomics 1992; 12: 241-253. Frank, W. E., Llewellyn, B. E., Fish, P. A., Riech, A. K., Marcacci, T. L., Gandor, D. W., et al. Validation of the AmpF€STRTM Profiler Plus PCR Amplification Kit for use in Forensic Casework. J Forensic Sci 2001; 46(3): 642-646. Gaensslen R. Sourcebook in Forensic Serology, Immunology, and Biochemistry. Section 10: Identification of Semen and Vaginal Secretions. Washington D.C.: US Department of Justice. 1983; 149-180. 39 Gill, P., Kimpton, C. P., Urquhart, A., Oldroyd, N., Millican, E. 8., Watson, S. K, and T. J. Downes. Automated Short Tandem Repeat (STR) Analysis in Forensic Casework — A Strategy for the Future. Electrophoresis 1995; 16: 1543-1552. Gutman, A. B. and E. B. Gutman. An “Acid” Phosphatase Occurring in the Serum of Patients with Metastasizing Carcinoma of the Prostate Gland. J Clin Invest 1938; 17: 473-478. Gutman, AB. and E. B. Gutman. Quantitative Relations of a Prostatic Component (Acid Phosphatase) of Human Seminal Fluid. Endocrinology 1941; 28(1): 115-118. Hammond, H., J in, L., Caskey, C., and R. Chakraborty. Evaluation of 13 Short Tandem Repeat Loci for Use in Personal Identification Applications. Am J Hum Genet 1994; 55: 179-189. Hansen, P. Determination of the Acid Prostatic Phosphatase as a New Method for Medicolegal Demonstration of Sperm Spots. Acta Path 1947; 23(2): 187-214. Hara, M., Koyanagi, T., Inoue, T., and T. Fujuyama. Some Physiochemical Characteristics of Gamma-Seminoprotein: An Antigenic Component Specific for Human SeminaI Plasma. Jpn J Legal Med 1971; 25: 322-324. Hochmeister, M. N., Budowle B., Rudin, 0., Gehrig, C., Borer, U., Thali, M., and R. Dimhofer. Evaluation of Prostate-Specific Antigen (PSA) Membrane Test Assays for the Forensic Identification of Seminal Fluid. J Forensic Sci 1999; 44: 1057-1060. Kind, S. S., The Acid Phosphatase Test. Methods of Forensic Science. 1964; 3: 267-287. Kutscher, W., and H. Wohlbergs. Prostataphosphatase. HoppeSeyler’s Z Physiol Chem 1935; 239: 109-126. LaFountain, M. J ., Schwartz, M. B., Svete, P. A., Walkinshaw, M. A., and E. Buel. TWGDAM Validation of the AmpFESTR Profier PlusTM and AmpFéSTR CofilerTM STR Multiplex System using Capillary Electrophoresis. J Forensic Sci 2001; 46(5): 1191- 1198. Li, T. S., and CG. Beling. Isolation and Characterization of Two Specific Antigens of Human Seminal Plasma. Fertil Steril 1973; 24: 134-43. Lundquist, F. Medicolegal Identification of Seminal Stains Using the Acid Phosphatase Test. Archives of Path 1950; 50(4): 395-399. Lygo, J. E., Johnson, P. E., Holdaway, D. J ., Woodroffe, S., Whitaker, J. P., Clayton, T. M., Kimpton, C. P. and P. Gill. The Validation of Short Tandem Repeat (STR) Loci for use in Forensic Casework. Int J Leg Med 1994; 107: 77-89. 40 Mann, T. The Biochemistry of Semen. John Wiley & Sons, Inc., 1954. Moretti, T. R., Baumstark, A. L., Defenbaugh, D. A., Keys, K. M., Smerick, J. B., and B. Budowle. Validation of Short Tandem Repeats (STRs) for Forensic Usage: Performance Testing of Fluorescent Multiplex STR Systems and Analysis of Authentic and Simulated Forensic Samples. J Forensic Sci 2001; 46(3): 647-660. Mullis, K. B. The Unusual Origin of the Polymerase Chain Reaction. Scientific American 1990; 262 (4): 56-65. Oppitz E. Eine neue Farbemthode zum Nachweis der Spermien bei Sittlichkeitsdelikten. Arch Kriminol 1969; 144: 145-148. Saferstein, R. Forensic Science Handbook 1988; 2: 347-392. Sensabaugh, G. F. Isolation and Characterization of Semen-Specific Protein from Human Seminal Plasma: A Potential New Marker for Semen Identification. J Forensic Sci 1978; 23 (1): 106-115. ’ Walker, J. T. A New Test for Seminal Stains. New Engl J Med 1950; 242: 110-1 1 1. 41 ililHllr'lllllIIlIll!HIIIIIMIII'II1| 3 1293 02460 4211