Multiplex mRNA profiling for the identification of body fluids

Forensic Science International 152 (2005) 1–12 www.elsevier.com/locate/forsciint

Multiplex mRNA profiling for the identification of body fluids Jane Juusola a,c, Jack Ballantyne a,b,c,* a

Graduate Program in Biomolecular Sciences, University of Central Florida, P.O. Box 162366, Orlando, FL 32816-2366, USA b Department of Chemistry, University of Central Florida, Bldg. #5, 4000 Central Boulevard, Orlando, FL 32816-2366, USA c National Center for Forensic Science, P.O. Box 162367, Orlando, FL 32816-2367, USA Received 15 October 2004; received in revised form 16 February 2005; accepted 17 February 2005 Available online 8 April 2005

Abstract We report the development of a multiplex reverse transcription-polymerase chain reaction (RT-PCR) method for the definitive identification of the body fluids that are commonly encountered in forensic casework analysis, namely blood, saliva, semen, and vaginal secretions. Using selected genes that we have identified as being expressed in a tissue-specific manner we have developed a multiplex RT-PCR assay which is composed of eight body fluid-specific genes and that is optimized for the detection of blood, saliva, semen, and vaginal secretions as single or mixed stains. The genes include b-spectrin (SPTB) and porphobilinogen deaminase (PBGD) for blood, statherin (STATH) and histatin 3 (HTN3) for saliva, protamine 1 (PRM1) and protamine 2 (PRM2) for semen, and human beta-defensin 1 (HBD-1) and mucin 4 (MUC4) for vaginal secretions. The known or presumed functions of these genes suggest an extremely restricted pattern of gene expression, which is a basic requirement for incorporation into a tissue-specific assay. The methodology is based upon gene expression profiling analysis in which the body fluid-specific genes are identified by detecting the presence of appropriate mRNA species using capillary electrophoresis/laser induced fluorescence. An mRNA-based approach, such as the multiplex RT-PCR method described in the present work, allows for the facile identification of the tissue components present in a body fluid stain and could supplant the battery of serological and biochemical tests currently employed in the forensic serology laboratory. # 2005 Elsevier Ireland Ltd. All rights reserved. Keywords: RNA; mRNA profiling; Body fluid identification; Blood identification; Saliva identification; Semen identification; Vaginal secretions identification; Menstrual blood identification; Multiplex RT-PCR

1. Introduction Conventional methods of body fluid identification use labor-intensive, technologically diverse techniques that are performed in a series, not parallel, manner and are costly in terms of time and sample. For example, blood identification is carried out by crystal tests such as the Takayama test or by immunochemical or immunochromatographic means in which a human-specific antiserum to serum proteins or hemoglobin produces a precipitated or captured antigen– * Corresponding author. Tel.: +1 407 823 4440; fax: +1 407 823 2252. E-mail address: [email protected] (J. Ballantyne).

antibody lattice upon diffusion or electrophoresis [1,2]. Similarly, semen is identified by the microscopical confirmation of spermatozoa or the immunochemical or immunochromatographic identification of prostate-specific antigen (PSA or p30) [3,4]. Moreover, for some frequently encountered body fluids no confirmatory technique exists. There is no definitive test, for example, for the presence of saliva or vaginal secretions. In seeking to develop novel multiplex (i.e. parallel) analysis procedures for body fluid identification that are compatible with current DNA analysis procedures, we initially considered assays based upon proteins and messenger RNA (mRNA) since both are expressed in a tissue-specific manner. However, multiplex analysis of complex, partially

0379-0738/$ – see front matter # 2005 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.forsciint.2005.02.020

2

J. Juusola, J. Ballantyne / Forensic Science International 152 (2005) 1–12

degraded protein mixtures, such as those present in body fluid stains, awaits further developments in proteomics. Messenger RNA is considered a better option at present because the technologies for massively parallel analysis continue to be developed due to rapid advances in the field of functional genomics. Terminally differentiated cells, whether they comprise blood leukocytes or lymphocytes, ejaculated spermatozoa, or epithelial cells lining the oral or vaginal cavities become such during a developmentally regulated program in which certain genes are turned off whereas others are turned on [5]. Thus, a pattern of gene expression is produced that is unique to each cell type and is characterized by the presence and relative abundance of specific mRNA species [5,6]. The typical body fluids encountered in forensic casework, such as blood, saliva, semen, and vaginal secretions, consist of cells and secretions that originate from multiple tissues. Each tissue type is comprised of cells that have a unique transcriptome, or gene expression (i.e. mRNA) profile [6], but we term the collection of genes that are expressed within the constellation of differentiated cells from the different tissues that makes up a body fluid the ‘multicellular transcriptome’. These genes comprise ubiquitously expressed housekeeping genes, which are responsible for cell maintenance functions, and genes that are specifically expressed in certain tissues only. The mRNA molecules are present in different quantities depending upon the particular species of mRNA and the cell type, and can be classified as abundant, moderately abundant, and rare. Thus, if the type and abundance of mRNAs could be determined in a stain or tissue sample recovered from the crime scene, it would be possible to identify the tissue or body fluid in question. Previously, we have reported the isolation of total RNA of sufficient quality and quantity from recently deposited and aged biological stains to enable subsequent detection of specific mRNA species by reverse transcription-polymerase chain reaction (RT-PCR) methods, and in the process identified a number of genes that appear to be saliva-specific [7]. In the present work, we report the development of a multiplex RT-PCR method for the definitive identification of most of the body fluids commonly encountered in forensic casework, namely blood, saliva, semen, and vaginal secretions. In order to facilitate transfer into routine forensic casework laboratories, the method was designed to incorporate the use of a commonly used capillary electrophoresis system.

2. Methods 2.1. Body fluid samples Body fluids were collected using procedures approved by the University’s Institutional Review Board. Blood was collected by venipuncture and 50 ml aliquots placed onto cotton cloth and dried at room temperature. Saliva was obtained in a centrifuge tube and 50 ml aliquots placed onto

cotton cloth and dried at room temperature. Freshly ejaculated semen was collected in plastic cups, and then aliquoted onto cotton cloth in known volumes and dried at room temperature or allowed to dry onto cotton swabs at room temperature. Semen-free vaginal secretions and menstrual blood were collected using sterile cotton-tipped swabs from volunteers who had abstained from sexual contact for a minimum of 3 days. In general and unless otherwise indicated, a 50 ml stain or a single cotton swab was used individually or in combination for RNA isolation. 2.2. RNA isolation Total RNA was extracted from blood, saliva, and semen stains, and vaginal swabs with guanidine isothiocyanate– phenol:chloroform and precipitated with isopropanol [8]. Briefly, 500 ml of denaturing solution (4 M guanidine isothiocyanate, 0.02 M sodium citrate, 0.5% sarkosyl, and 0.1 M b-mercaptoethanol) was preheated in a Spin-EaseTM extraction tube (Gibco BRL, Life Technologies Inc., Gaithersburg, MD) at 56 8C for approximately 10 min. The stain was placed into the extraction tube with the preheated denaturing solution and incubated at 56 8C for 30 min. The gauze or the swab tip was removed from the denaturing solution and placed in a Spin-EaseTM extraction tube filter insert, which was then placed into its cognate extraction tube. The tube with the filter insert was centrifuged for 10 min at 8160
g to remove any absorbed liquid from the gauze or swab. After centrifugation, the filter with associated substrate was discarded. 50 ml of 2 M sodium acetate and 600 ml of phenol:chloroform, isoamyl alcohol (pH 5.3–5.7, equilibrated with succinic acid) were added to the extract, vortexed for several seconds, incubated at 4 8C for 1 h, and then centrifuged at 16,000
g for 20 min. The RNA-containing aqueous phase was transferred to a sterile 1.5-ml microcentrifuge tube, with care taken to avoid the interphase layer. Five hundred microliters of isopropanol together with 2 ml of GlycoBlueTM glycogen carrier (Ambion Inc., Austin, TX) was added to the aqueous layer. RNA was precipitated overnight at 20 8C, after which the sample was centrifuged at 16,000
g for 25 min. Following centrifugation, the supernatant was removed and the pellet was washed once with 1 ml of 75% ethanol/25% DEPCtreated water. After centrifugation at 16,000
g for 10 min, the supernatant was discarded and the pellet was dried in vacuum centrifuge for 5–7 min. The RNA pellet was resolubilized in 12–20 ml of preheated RNA Resuspension Solution (Ambion Inc., Austin, TX) and heated at 60 8C for 10 min to ease re-suspension. The sample was DNase-treated immediately or stored at 20 8C until needed. 2.3. DNase I digestion Six units of DNase I (RNase-free) (2 U/ml) (Ambion Inc., Austin, TX) and digestion buffer provided (10 mM Tris– HCl, pH 7.5, 2.5 mM MgCl2, 0.1 mM CaCl2) was added to

J. Juusola, J. Ballantyne / Forensic Science International 152 (2005) 1–12

the whole RNA extract and incubated at 37 8C for 1 h. The DNase was inactivated at 75 8C for 10 min, after which the samples were quantitated and then stored at 20 8C until needed [9,10]. 2.4. RNA quantitation RNA was quantitated using a sensitive fluorescence assay based upon the binding of the unsymmetrical cyanine dye RiboGreen1 (Molecular Probes, Eugene, OR) [11]. The manufacturer’s instructions were followed for the highrange assay, which detects from 20 ng/ml to 1 mg/ml RNA. Briefly, 200 ml assay volumes were used with 96-well microplates. The final mixture in each sample well consisted of 2 ml of DNase I-treated RNA extract, 98 ml TE buffer (10 mM Tris–HCl, 1 mM EDTA, pH 7.5, in nuclease-free water), and 100 ml of 750 nM RiboGreen1 reagent (diluted 200-fold from the concentrated stock provided in kit). After a 3-min incubation of the samples at room temperature protected from light, fluorescence emission at 535 nm (excited at 485 nm) was determined using a Wallac Victor2 microplate reader (Perkin-Elmer Life Sciences, Boston, MA). The RNA concentrations in the samples were calculated using an appropriate standard curve as described by the manufacturer.

3

2.6.2. Primers PCR primers were labeled with Fluorescent Phophoroamidite dyes (6-FAM: blue; TET: green; HEX: yellow). PCR primer sequences for SPTB, HBD-1, MUC4, and MMP-7 were designed using Primer 3 Online primer design software. PCR Primer sequences for STATH and HTN3 were designed using Oligo1 Primer Analysis Software, Version 6 (Lifescience Software Resource, Long Lake, MN). PCR Primer sequences for PBGD, PRM1, and PRM2 were obtained from published sources [12,13]. Table 1 shows the PCR primer sequences and the expected product sizes. Primers were custom synthesized by Life Technologies, Grand Island, New York. PCR primers were added in the following concentrations: SPTB, 0.8 mM; PBGD, 0.24 mM; STATH, 0.4 mM; HTN3, 0.4 mM; PRM1, 0.04 mM; PRM2, 0.4 mM; HBD-1, 1.6 mM; MUC4, 1.6 mM. 2.6.3. Cycling conditions The standard PCR conditions used for all singleplex reactions and multiplex development consisted of a denaturing step (95 8C, 11 min) followed by 35 cycles (94 8C, 20 s; 55 8C, 30 s; 72 8C, 40 s) and a final extension step (72 8C, 5 min) using the GeneAmp1 PCR System 9700 (Applied Biosystems, Foster City, CA).

2.5. cDNA synthesis

2.7. PCR product detection

For the reverse transcriptase reaction, 6 ml RNA template, 0.5 mM each dNTP, 5 mM random decamers, and nuclease-free water were combined to a final volume of 24 ml. This mixture was heated at 75 8C for 3 min to eliminate secondary structure of target mRNA and snapcooled on ice. To the mixture, 3 ml of 10
first-strand buffer (500 mM Tris–HCl pH 8.3, 750 mM KCl, 30 mM MgCl2, 50 mM DTT), 1.5 ml of SUPERase-InTM RNase Inhibitor (20 U/ml) (Ambion Inc., Austin, TX) and 1.5 ml of Moloney Murine Leukemia Virus-Reverse Transcriptase (100 U/ml) (Ambion Inc., Austin, TX) were added to yield a final reaction volume of 30 ml. This reaction mixture was incubated at 42 8C for 1 h, and then at 95 8C for 10 min to inactivate the reverse transcriptase.

2.7.1. Agarose gel electrophoresis RT-PCR products were separated on 4% agarose gels. Electrophoresis was carried out at 100 V for 60 min in TAE buffer (0.04 M Tris–acetate, 0.001 M EDTA). The gels were stained with SYBR1 Gold nucleic acid stain (Molecular Probes, Eugene, OR) and photographed under UV transillumination.

2.6. Polymerase chain reaction 2.6.1. Standard reaction Two microliters of the RT reaction was amplified in a total reaction volume of 25 ml. The standard reaction mixture contained buffer (10 mM Tris–HCl, pH 8.3, 50 mM KCl, 1.5 mM MgCl2), 0.125 mM each dNTP, 0.8 mM PCR primer, and 1.25 units AmpliTaq Gold1 DNA polymerase (Roche Molecular Systems Inc., Branchburg, NJ). The multiplex reaction mixture contained buffer (10 mM Tris– HCl, pH 8.3, 50 mM KCl), 3.5 mM MgCl2, 0.125 mM each dNTP, 5.48 mM PCR primers (see below), and 1.25 units AmpliTaq Gold 1 DNA Polymerase.

2.7.2. Capillary electrophoresis Multiplex RT-PCR products were detected using the ABI Prism 310 capillary electrophoresis system. A 1.5 ml aliquot of each amplified sample was added to 24 ml Hi-Di formamide (Applied Biosystems, Foster City, CA) and 1 ml of GeneScan 500 TAMRA internal lane standard (Applied Biosystems, Foster City, CA). Tubes were heated at 95 8C for 3 min and snap cooled on ice for at least 3 min. Samples were injected through the capillary using the module GS STR POP4 (1 ml) C (5 s injection, 15 kV, 60 8C, run time 30 min, filter set C). Samples were subject to laser induced fluorescence, and analyzed with GeneScan 3.1.2 Software.

3. Results 3.1. Identification of body fluid-specific gene candidates Initially, candidate sets of blood, saliva, semen, and vaginal secretions-specific genes were identified through a

4

J. Juusola, J. Ballantyne / Forensic Science International 152 (2005) 1–12

Table 1 Body fluid-specific genes Body fluid Blood

Saliva

Semen

Vaginal secretions

Menstrual blood

Gene

Primer sequences/dyes 0

Size (bp)

Reference

SPTB

5 -6-FAM-AGG ATG GCT TGG CCT TTA AT 50 -ACT GCC AGC ACC TTC ATC TT

247

Primer 3

PBGD

50 -TET-TGG ATC CCT GAG GAG GGC AGA AG 50 -TCT TGT CCC CTG TGG TGG ACA TAG CAA T

177

[12]

STATH

50 -6-FAM-CTT CTG TAG TCT CAT CTT G 50 -TGG TTG TGG GTA TAG TGG TTG TTC

198

Oligo 6

HTN3

50 -6-FAM-GCA AAG AGA CAT CAT GGG TA 50 -GCC AGT CAA ACC TCC ATA ATC

134

Oligo 6

PRM1

50 -HEX-GCC AGG TAC AGA TGC TGT CGC AG 50 -TTA GTG TCT TCT ACA TCT CGG TCT

153

[13]

PRM2

50 -6-FAM-GTG AGG AGC CTG AGC GAA CGC 50 -TTA GTG CCT TCT GCA TGT TCT CTT C

294

[13]

HBD-1

50 -6-FAM-TTC CTG AAA TCC TGA GTG TT 50 -TAA CAG GTG CCT TGA ATT TT

213

Primer 3

MUC4

50 -HEX-GGA CCA CAT TTT ATC AGG AA 50 -TAG AGA AAC AGG GCA TAG GA

235

Primer 3

MMP-7

50 -TET-TCA ACC ATA GGT CCA AGA AC 50 -CAA AGA ATT TTT GCA TCT CC

240

Primer 3

combination of literature and database searches and consideration of the physiology and biochemistry of each body fluid. Subsequently, candidate genes that had known pseudogenes were rejected due to the difficulty of distinguishing mRNA derived PCR products from contaminating genomic DNA. Primers for each gene were then designed to span different exons such that the genic DNA would produce amplimers that were either larger than that expected for mRNA or would not be amplifiable under the specific PCR conditions used. Preliminary tests of sensitivity and specificity resulted in the identification of a number of putative tissue-specific gene transcripts that demonstrated promise for the identification of blood, saliva, semen, and vaginal secretions. From these the following genes were used to develop the multiplex assay which is the subject of this report: b-spectrin (SPTB) [14,15] and porphobilinogen deaminase (PBGD) [12] for blood, statherin (STATH) [7] and histatin 3 (HTN3) for saliva [7], protamine 1 (PRM1) [13,16] and protamine 2 (PRM2) [13,16] for semen, and human beta-defensin 1 (HBD-1) [17] and mucin 4 (MUC4) [18,19] for vaginal secretions. Table 1 lists the primers, expected amplimer sizes and dye labels used for each of the body fluid-specific genes. 3.2. Confirmation of the body fluid specificity of gene candidates Specificity was established by demonstrating that the mRNA for the candidate gene was present in one type of body fluid stain but absent from all others. The body fluids tested included blood, saliva, semen, and vaginal secretions.

Also included in the testing was menstrual blood, which would be expected to contain not only blood and vaginal secretions components but also some menstrual cycledependent endometrial tissue secretions not present in the other fluids. The testing involved the ‘monoplex’ amplification of gene transcripts by RT-PCR from total RNA isolated from body fluid stains and a subsequent separation and detection of the amplimers by agarose gel electrophoresis. The assays described in this paper were developed using samples obtained from at least three different individuals and, for each series of experiments, representative results are provided. The putative blood-specific SPTB (247 bp) [14,15] and PBGD (177 bp) [12] gene transcripts were detected in blood stain RNA and, to a lesser extent, in menstrual blood RNA (Fig. 1a), but were undetectable in RNA isolated from saliva, semen, and vaginal secretions stains. The saliva gene candidates STATH (198 bp) [20] and HTN3 (134 bp) [21], were amplified from saliva stain mRNA, but were undetectable in blood, semen, and vaginal secretions, and menstrual blood (Fig. 1b and [7]). Due to sequence homology, the larger and less specific Histatin 1 gene transcript (HTN1) (152 bp) was detected using the HTN3 primer set used. HTN1 was detectable in saliva and to a significantly lesser extent in some RNA isolates from semen but not in blood, menstrual blood, or vaginal secretions (Fig. 1b). A smaller STATH transcript (approximately 168 bp), possibly representing an alternatively spliced isoform, was present in saliva stains and in menstrual blood (Fig. 1b). The semen candidate genes PRM1 (153 bp) and PRM2 (294 bp) [13,16] were amplifiable from semen stain RNA, but were undetectable in blood,

J. Juusola, J. Ballantyne / Forensic Science International 152 (2005) 1–12

5

Fig. 1. Identification of body fluid-specific gene transcripts. RT-PCR products for eight body fluid-specific genes using total RNA extracted from blood, saliva, semen, and vaginal and menstrual blood swabs are shown. RNA specificity was demonstrated by processing the same sample with (+) and without () reverse transcriptase. The (a) blood-specific genes SPTB (247 bp) and PBGD (177 bp); (b) saliva-specific genes STATH (198 bp) and HTN3 (134 bp); (c) semen-specific genes PRM1 (153 bp) and PRM2 (294 bp); (d) vaginal secretions-specific genes HBD-1 (213 bp) and MUC4 (235 bp) are indicated. The expected amplimer sizes are indicated in parenthesis and with arrows for samples that are weak or that exhibit multiple products.

saliva, vaginal secretions, and menstrual blood (Fig. 1c). The putative vaginal secretions genes HBD-1 (213 bp) [17] and MUC4 (235 bp) [18,19] demonstrated specificity in that they were amplified from vaginal secretions and menstrual blood RNA, but were undetectable in blood, saliva, and semen (Fig. 1d). In summary, body fluid specificity of the selected gene candidates was confirmed by the presence of appropriately sized RT-PCR products in one body fluid exclusive of the others tested. 3.3. Development of multiplex RT-PCR assays for body fluid identification In order for mRNA profiling to be of use in forensic casework it was important to develop a ‘multiplex’ method for the simultaneous analysis of body fluid-specific genes in

a single sample extract. Initially, a proof of concept tetraplex RT-PCR system was developed by incorporating one tissuespecific gene primer pair per body fluid, namely SPTB for blood, HTN3 for saliva, PRM2 for semen, and HBD-1 for vaginal secretions (Fig. 2). The tissue-specific, and differentially sized, gene products from a mixed stain extract comprising blood, saliva, semen, and vaginal secretions stains were readily visualized by either agarose gel electrophoresis (Fig. 2a) or capillary electrophoresis (Fig. 2b). No products were present in control extracts in which reverse transcriptase was omitted, indicating that mRNA was being detected and not contaminating genomic DNA. Although the tetraplex results were encouraging we hypothesized that normal biological variation in expression levels of any tissue-specific gene could result in the presence of false negative or false positive results. Accordingly, we anticipate that the probability of obtaining such false

6

J. Juusola, J. Ballantyne / Forensic Science International 152 (2005) 1–12

Fig. 2. Tetraplex assay for the identification of blood, saliva, semen, and vaginal secretions. RT-PCR products for the tetraplex system using RNA extracted from a co-mixed blood, saliva, semen, and vaginal swab stain with (a) 4% agarose gel electrophoresis and (b) capillary electrophoresis format are shown. Controls without RT (RT) were run in parallel with the RT reactions. The tetraplex with SPTB (247 bp), HTN3 (134 bp), PRM2 (294 bp), and HBD-1 (213 bp) was amplified subsequent to the RT reaction (+). The histatin 1 (HTN1) gene (152 bp), a homolog of HTN3, is also amplified with the HTN3 primer set used and its location in both electrophoretic formats is indicated.

negative or positive results would be effectively eliminated if two independently regulated genes per body fluid were analyzed, although further comprehensive validation of the system would be necessary. Thus, we expanded the tetraplex by incorporating additional blood (PBGD), saliva (STATH), semen (PRM1), and vaginal secretion (MUC4)specific genes. The resulting octaplex produced clear mRNA profiles from all eight incorporated genes when tested with an RNA isolate from an admixed stain extract containing blood, saliva, semen, and vaginal secretions (Fig. 3a). It was important to test the octaplex using single source and admixed body fluid samples since such a range of sample types is encountered in forensic casework. The body fluid specificity of the octaplex was confirmed by performance checks of the system using total RNA extracted from all possible single source and mixed sample combinations of the four body fluids blood, saliva, semen, and vaginal secretions. Single source body fluid stains gave expected results with the octaplex in that only the appropriate two body fluid-specific genes were present in blood (SPTB and PBGD), saliva (HTN3 and STATH), semen (PRM1 and PRM2), and vaginal secretions (HBD1 and MUC4) (Fig. 3b). All six binary mixtures of the four body fluids were analyzed using the octaplex and all gave expected results in that the appropriate pairs of genes were detected with the concomitant absence of the other two pairs. Illustrative examples of the mRNA profiles obtained with binary mixtures are provided for saliva–semen (Fig. 4a), semen– vaginal secretions (Fig. 4b), and blood–semen (Fig. 4c). The octaplex assay performed well independent even if one of the two body fluids was in excess. For example, saliva– semen mixtures exhibited all four expected body fluidspecific genes (HTN3, STATH, PRM1, and PRM2) whether the saliva or semen was in three-fold excess (Fig. 4a). Similar results were obtained for all three possible ternary mixtures. Again only the expected three pairs of body fluidspecific genes were obtained from blood–semen–vaginal

secretions (Fig. 5a), saliva–semen–vaginal secretions (Fig. 5b), and blood–saliva–semen mixtures (Fig. 5c). All of the expected genes were detected in the ternary mixtures even when one of the body fluids was present in three-fold excess as is illustrated for the blood–saliva–semen mixture (Fig. 5c). Thus, the octaplex assay demonstrated body fluid identification specificity in that body fluid-specific genes were present in stains only if the corresponding body fluid was present (i.e. no false positive results) and were absent if the corresponding body fluid was not present in the sample (i.e. no false negative results). Each body fluid was reproducibly identifiable by two body fluid-specific genes and no instance of a single gene drop out was observed with any of the samples tested (some data not shown). 3.4. Octaplex sensitivity The sensitivity of the octaplex RT-PCR analysis and a comparison with the eight cognate singleplex reactions was evaluated using varying quantities of input total RNA ranging from 200 pg to 320 ng and the results are summarized in Table 2. The indicated limits of detection for the octaplex were established by the ability to detect both of the body fluid-specific genes for the given body fluid with a detection threshold set at 150 RFU. The sensitivities of the individual genes were established using singleplex RT-PCR with CELIF for detection. The limits of detection for the octaplex varied in that the semen genes were the most sensitive (detection limit <200 pg of total RNA), the vaginal secretions genes the least sensitive (12 ng), whereas the blood (6 ng) and saliva (9 ng) genes were found to be of intermediate sensitivity. The sensitivity limits for each of the genes did not differ from their singleplex values (Table 2) indicating that the octaplex as formulated demonstrates very good amplification efficiency.

J. Juusola, J. Ballantyne / Forensic Science International 152 (2005) 1–12

7

Fig. 3. Octaplex assay for the identification of blood, saliva, semen, and vaginal secretions. RT-PCR products for the octaplex using RNA extracted from a blood stain, a saliva stain, a semen stain, and a vaginal swab (a) simultaneously and (b–e) individually are shown. Controls without RT were run in parallel with the RT reactions (not shown). The octaplex with SPTB, PBGD, STATH, HTN3, PRM1, PRM2, HBD-1, and MUC4 was amplified subsequent to the RT reaction. The peaks at approximately 152 bp and 165 bp in the blue (top) channel (in (a) and (c)) are products from HTN1 and STATH isoform, respectively. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

3.5. Menstrual blood The ability to distinguish menstrual blood from circulatory blood can be important in certain cases, particularly those involving sexual assault in which the victim is in menses at the time of the incident. Menstruation is characterized by the rapid and incomplete degeneration of the functional layer of the endometrium and consequent exposure of open capillary vessels. What is described as men-

strual blood actually comprises tissue breakdown products from endometrium derived stromal, epithelial, endothelial, vascular smooth muscle, and bone marrow derived cells as well as capillary blood [22,23]. Thus, menstrual blood would be expected to comprise most, but due to degradation and/or dilution effects not all, of the gene products present in circulatory blood plus several unique gene products originating from the endometrium. Menstrual blood gave reactions with the octaplex system consistent with the presence

8

J. Juusola, J. Ballantyne / Forensic Science International 152 (2005) 1–12

Fig. 4. Detection of two-body fluid mixtures using the octaplex. RT-PCR products for (a) 3:1 and 1:3 saliva–semen mixtures, (b) semen–vaginal secretions mixture, and (c) 1:1 blood–semen mixture are shown. Expected amplimers were detected in all samples. Controls without RT were run in parallel with the RT reactions (not shown).

of vaginal secretions, namely the presence of the genes HBD-1 and MUC4, but failed to detect significant levels of the blood-specific genes PBGD and SPTB (Fig. 6b) despite the low level detection of the latter genes in menstrual blood by singleplex analysis (Fig. 1a). In contrast, as expected binary blood and vaginal secretions mixtures exhibited both the blood and vaginal secretions-specific genes (Fig. 6a). The cause of this difference in the detection of blood-specific markers in menstrual blood samples between the singleplex and octaplex assays is unknown but is probably due to decreased efficiency of the amplification of blood-specific markers in the multiplex format that was not noted previously with circulatory blood (Table 2). Thus, the octaplex as formulated would not be capable of distinguishing menstrual blood and vaginal secretions. In order to improve the specificity of the octaplex to include menstrual blood detection for those specialized instances in which menstrual blood may be useful, we investigated the possibility of adding a menstrual bloodspecific marker to the octaplex. Enzymes of the matrix metalloproteinase (MMP) multigene family that play a role in the tissue remodeling that takes place during menstruation are spatiotemporally expressed in endometrial tissue [24,25]. One of these enzymes, MMP-7, exhibits a restricted pattern of gene expression in that its mRNA is detectable in menstrual blood but not in circulatory blood, semen, saliva, or vaginal secretions (Fig. 6c). MMP-7 was incorporated into the octaplex and the resulting nonaplex was able to

distinguish between menstrual blood and a circulatory blood–vaginal secretions admixture. Menstrual blood expressed MMP-7 and the vaginal secretions genes HBD1 and MUC4 but no blood genes (Fig. 6d) whereas the circulatory blood–vaginal secretions admixture, as before (Fig. 6a), exhibited HBD-1 and MUC4 plus the blood genes SPTB and PBGD (data not shown).

4. Discussion In this paper, we describe genes that are specifically expressed in each of the forensically relevant body fluids, including b-spectrin (SPTB) and porphobilinogen deaminase (PBGD) for blood, statherin (STATH) and histatin 3 (HTN3) for saliva [7], protamine 1 (PRM1) and protamine 2 (PRM2) for semen, and human beta-defensin 1 (HBD-1) and mucin 4 (MUC4) for vaginal secretions. The known or presumed functions of these genes suggest an extremely restricted pattern of gene expression, which is a basic requirement for incorporation into a tissue-specific assay. b-Spectrin is a unit of the major protein component of the erythrocyte membrane skeleton, spectrin, and is processed in a tissue-specific manner in erythroid cells [14,15]. Porphobilinogen deaminase is the third enzyme of the heme biosynthesis pathway, and has an erythrocyte-specific isoform [12]. Statherin is an inhibitor of the precipitation of calcium phosphate salts in the oral cavity [20]. Histatin 3 is a

J. Juusola, J. Ballantyne / Forensic Science International 152 (2005) 1–12

9

Fig. 5. Detection of three-body fluid mixtures using the octaplex. RT-PCR products for (a) blood–semen–vaginal secretions mixture, (b) saliva– semen–vaginal secretions mixture, and (c) blood–saliva–semen mixtures in different ratios (1:1:1, 3:1:1, 1:3:1, and 1:1:3) are shown. Expected amplimers were detected for all samples. Controls without RT were run in parallel with the RT reactions (not shown).

histidine rich protein involved in the non-immune host defense in the oral cavity [21]. Protamine 1 and Protamine 2 are DNA-binding proteins involved in condensation of sperm chromatin [13,26]. Human beta-defensin 1 is an antimicrobial peptide involved in host defense of the urogenital epithelium [17]. Mucin 4 is a major membranespanning mucin of the endocervix that protects epithelial surfaces of the reproductive tract against pathogens and controls sperm entry into the uterus [19,27]. The principle aim of this work was the development, using the tissue-specific genes described above, of a multiplex mRNA method for the identification of the body fluids and tissues commonly encountered in forensic casework. We deliberately incorporated a number of design factors and assay quality controls to ensure that the tissue-specific RTPCR assays detect mRNA and not contaminating genomic DNA. Firstly, the gene-specific primers were designed to span different exons such that the genic DNA would produce amplimers that are either larger than that expected for

mRNA or would not be amplifiable under the PCR conditions used. Secondly, the particular genes chosen demonstrated no evidence for the presence of processed pseudogenes as assessed by testing the primers with DNA ([7], data not shown). Finally, RNA isolates are treated with DNase I prior to RT-PCR and the detection of mRNA is only deemed positive when no PCR product is observed in the absence of reverse transcriptase (RT). Our initial parallel assay was a tetraplex system, which included one body fluid-specific gene each for blood, saliva, semen, and vaginal secretions. The amplified products from the tetraplex were distinguishable on agarose gels; however, in order to determine the size of the PCR products more accurately, especially since some of the sizes of the amplimers only differed by as little as 30 bp, to expand the multiplex to include more genes per body fluid and to facilitate technology transfer to operations, we developed the multiplex for use on a standard capillary electrophoresis (CE) platform. In both the agarose and CE formats, the RT

10

J. Juusola, J. Ballantyne / Forensic Science International 152 (2005) 1–12

Table 2 Sensitivity of octaplex system and individual genes Octaplex Blood (ng) Saliva (ng) Semen (pg) Vaginal secretions (ng)

6 9 <200 12

Blood SPTB (ng) PBGD (ng)

6 6

Saliva HTN3 (ng) STATH (ng)

6 9

Semen PRM1 (pg) PRM2 (pg) Vaginal secretions HBD-1 (ng) MUC4 (ng)

<200 <200 12 <2

The sensitivity of the octaplex system is defined as the lowest amount of input RNA, which permitted the detection of both body fluid-specific genes.

sample was negative, indicating that mRNA and not contaminating DNA was being detected. Additional genes were then incorporated into the CE assay, with the goal of including two markers per body

fluid into one multiplex reaction. The built-in redundancy of two genes per body fluid was designed to account for possible biological variation in gene expression levels. While an individual may possess a mutation that reduces or eliminates the expression of one of the two body fluidspecific genes, the possibility of two independent loss of function mutations is expected to be remote. The resulting octaplex system was able to reproducibly identify each of the four body fluids when present as single or mixed body fluid stains. The sensitivity of the system (<200 pg–12 ng of total RNA input) is suitable for forensic casework use since typical sized body fluid stains (50 ml) yield hundreds of nanograms of total RNA even in aged stains ([7], data not shown). The species specificity of the assay has not yet been fully determined and will be the subject of another manuscript detailing a fuller developmental validation study carried out in accordance with guidelines recommended by the Scientific Working Group on DNA Analysis Methods (SWGDAM) [28]. However, there is good reason to be optimistic that all or most of the genes used will be human-specific. The primers used in the octaplex do not show significant homology with any other species in Genebank, including non-human primates (data not shown). We have a high level of confidence in the human blood specificity since preliminary testing of the octaplex with cat, dog, deer, horse, cow, sheep, and spider monkey blood produced negative results (data not shown).

Fig. 6. Identification of menstrual blood stains. RT-PCR products for (a) blood-vaginal secretions mixture and (b) menstrual blood using the octaplex are shown. Expected amplimers were detected for the blood–vaginal secretions mixture. Only the vaginal secretions markers were detected in menstrual blood. (c) RT-PCR product for menstrual blood-specific gene, MMP-7 (240 bp), is shown. (d) RT-PCR products for menstrual blood using the nonaplex. Controls without RT were run in parallel with the RT reactions ((c) and not shown).

J. Juusola, J. Ballantyne / Forensic Science International 152 (2005) 1–12

In some cases, it may be necessary to definitively identify menstrual blood and distinguish between menstrual blood and a vaginal secretions/venous blood mixture. Although the octaplex clearly differentiates these two body fluid types, it would not be possible to distinguish vaginal secretions from menstrual blood (which contains vaginal secretions). Thus, we developed a nonaplex, which incorporates the menstrual blood-specific gene MMP-7 for use in those cases that warrant it. We believe that messenger RNA profiling will play a major role in the future of forensic biology, not only for body fluid and tissue identification but also for providing other probative information such as the time since deposition of a stain and the determination of the phenotypic traits of a stain donor such as the individual’s age. Our long-term aim is to develop mRNA profiling assays to detect abundant or moderately abundant tissue-specific genes to automate the identification of body fluids that are significant to forensic investigations, including blood, saliva, semen, vaginal secretions, menstrual blood, urine, skin, muscle, adipose, and brain. Automation could be accomplished by using a ‘body fluid identification chip’ containing a microarray of cDNAs that would be able to recognize each of the candidate genes.

[3]

[4]

[5]

[6]

[7]

[8]

[9]

5. Conclusion [10]

We have identified tissue-specific genes characteristic of the body fluids that are most commonly encountered in forensic casework, namely blood, saliva, semen, and vaginal secretions. It is recognized that this report constitutes a ‘proof of concept’ of the proposed methodology and more extensive validation studies including, for example, environmental impact and species specificity studies need to be carried out. Indeed, such studies are in progress and will be the subject of a separate report. An mRNA-based approach, such as the multiplex RT-PCR method described here, allows the facile identification of the tissue components present in a body fluid stain and, conceivably, could supplant the battery of serological and biochemical tests currently employed in the forensic serology laboratory.

Acknowledgement Support for this project was provided by the State of Florida through the UCF Center for Forensic Science.

[11]

[12]

[13]

[14]

[15]

[16]

[17]

[18]

References [1] H.C. Lee, Identification and grouping of bloodstains, in: R. Saferstein (Ed.), Forensic Science Handbook, Prentice Hall, Engelwood Cliffs, NJ, 1982, pp. 267–331. [2] M.N. Hochmeister, B. Budowle, R. Sparkes, O. Rudin, C. Gehrig, M. Thali, L. Schmidt, A. Cordier, R. Dirnhofer,

[19]

11

Validation studies of an immunochromatographic 1-step test for the forensic identification of human blood, J. Forensic Sci. 44 (1999) 597–602. S. Baechtel, The identification and individualization of semen stains, in: R. Saferstein (Ed.), Forensic Science Handbook, Prentice Hall, Englewood Cliffs, NJ, 1988, pp. 369–374. M.N. Hochmeister, B. Budowle, O. Rudin, C. Gehrig, U. Borer, M. Thali, R. Dirnhofer, Evaluation of prostate-specific antigen (PSA) membrane test assays for the forensic identification of seminal fluid, J. Forensic Sci. 44 (1999) 1057–1060. B. Alberts, D. Bray, J. Lewis, M. Raff, K. Roberts, J.D. Watson, Molecular Biology of the Cell, Garland Publishing Inc., New York, 1994. H. Caron, B. van Schaik, M. van der Mee, F. Baas, G. Riggins, P. van Sluis, M.C. Hermus, R. van Asperen, K. Boon, P.A. Voute, S. Heisterkamp, A. van Kampen, R. Versteeg, The human transcriptome map: clustering of highly expressed genes in chromosomal domains, Science 291 (2001) 1289– 1292. J. Juusola, J. Ballantyne, Messenger RNA profiling: a prototype method to supplant conventional methods for body fluid identification, Forensic Sci. Int. 135 (2003) 85–96. P. Chomczynski, N. Sacchi, Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction, Anal. Biochem. 162 (1987) 156–159. Z. Huang, M.J. Fasco, L.S. Kaminsky, Optimization of Dnase I removal of contaminating DNA from RNA for use in quantitative RNA-PCR, Biotechniques 20 (1996) 1012–1014, 1016, 1018–1020. I. Wiame, S. Remy, R. Swennen, L. Sagi, Irreversible heat inactivation of DNase I without RNA degradation, Biotechniques 29 (2000) 252–254, 256. L.J. Jones, S.T. Yue, C.Y. Cheung, V.L. Singer, RNA quantitation by fluorescence-based solution assay: RiboGreen reagent characterization, Anal. Biochem. 265 (1998) 368–374. A.N. Gubin, J.L. Miller, Human erythroid porphobilinogen deaminase exists in 2 splice variants, Blood 97 (2001) 815– 817. K. Steger, K. Pauls, T. Klonisch, F.E. Franke, M. Bergmann, Expression of protamine-1 and -2 mRNA during human spermiogenesis, Mol. Hum. Reprod. 6 (2000) 219–225. K.M. Amin, A.L. Scarpa, J.C. Winkelmann, P.J. Curtis, B.G. Forget, The exon–intron organization of the human erythroid beta-spectrin gene, Genomics 18 (1993) 118–125. Z.L. Chu, A. Wickrema, S.B. Krantz, J.C. Winkelmann, Erythroid-specific processing of human beta spectrin I premRNA, Blood 84 (1994) 1992–1999. M. Bauer, D. Patzelt, Protamine mRNA as molecular marker for spermatozoa in semen stains, Int. J. Legal Med. 117 (2003) 175–179. E.V. Valore, C.H. Park, A.J. Quayle, K.R. Wiles, P.B. McCray Jr., T. Ganz, Human beta-defensin-1: an antimicrobial peptide of urogenital tissues, J. Clin. Invest. 101 (1998) 1633–1642. I.K. Gipson, S.B. Ho, S.J. Spurr-Michaud, A.S. Tisdale, Q. Zhan, E. Torlakovic, J. Pudney, D.J. Anderson, N.W. Toribara, J.A. Hill 3rd, Mucin genes expressed by human female reproductive tract epithelia, Biol. Reprod. 56 (1997) 999– 1011. I.K. Gipson, S. Spurr-Michaud, R. Moccia, Q. Zhan, N. Toribara, S.B. Ho, A.R. Gargiulo, J.A. Hill 3rd, MUC4 and MUC5B transcripts are the prevalent mucin messenger ribo-

12

[20]

[21]

[22]

[23]

J. Juusola, J. Ballantyne / Forensic Science International 152 (2005) 1–12 nucleic acids of the human endocervix, Biol. Reprod. 60 (1999) 58–64. L.M. Sabatini, Y.Z. He, E.A. Azen, Structure and sequence determination of the gene encoding human salivary statherin, Gene 89 (1990) 245–251. L.M. Sabatini, T. Ota, E.A. Azen, Nucleotide sequence analysis of the human salivary protein genes HIS1 and HIS2, and evolution of the STATH/HIS gene family, Mol. Biol. E 10 (1993) 497–511. L.A. Salamonsen, D.E. Woolley, Menstruation: induction by matrix metalloproteinases and inflammatory cells, J. Reprod. Immunol. 44 (1999) 1–27. S. Tabibzadeh, The signals and molecular pathways involved in human menstruation, a unique process of tissue destruction and remodelling, Mol. Hum. Reprod. 2 (1996) 77–92.

[24] L.A. Salamonsen, D.E. Woolley, Matrix metalloproteinases in normal menstruation, Hum. Reprod. 11 (Suppl. 2) (1996) 124– 133. [25] J.C. Dong, H. Dong, A. Campana, P. Bischof, Matrix metalloproteinases and their specific tissue inhibitors in menstruation, Reproduction 123 (2002) 621–631. [26] L. Domenjoud, G. Nussbaum, I.M. Adham, G. Greeske, W. Engel, Genomic sequences of human protamines whose genes, PRM1 and PRM2, are clustered, Genomics 8 (1990) 127–133. [27] I.K. Gipson, Mucins of the human endocervix, Front. Biosci. 6 (2001) D1245–D1255. [28] Scientific Working Group on DNA Analysis Methods (SWGDAM), Revised validation guidelines, Forensic Sci. Comm. 6 (2004).