Description of the problem

The current confirmatory serology tests used in forensic sciences are used to detect proteins and present several limitations. Proteins are the effector cells in the tissues, and therefore in theory it is expected that their presence and relative abundance can establish which body fluid is present. Proteins can be carriers or catalysts of

biochemical reactions and a test can use the proteins’ substrate as an indicator of an active enzyme. However, for such detection to be possible, the protein must maintain activity otherwise it will consist of a disorganized active center that does not recognize its substrate. Extreme temperatures, changes in pH and moisture can easily denature a protein, rendering its activity non-existent and therefore its detection impossible for body fluid identification. Moreover, some enzymes lose its catalytic capacity due to the

presence of inhibitors present at the crime scene and collected with the body fluid stain. These aspects are extremely important since the samples that arrive at a forensic

laboratory have not been in environments with controlled temperature, pH or clean from other contaminants. Most of the confirmatory methods mentioned above have been validated in mock crime scene samples, most consisting of samples placed in different materials. Some have been exposed to long term storage or high temperatures but such studies are limited.

Forensic serologists have studied the proteins that are specific to a tissue and developed presumptive methods to identify tissues based on the presence or quantity of a

single protein (110). According to the central dogma, proteins are translated from mRNA transcripts. Therefore, measuring specific mRNA transcripts instead of proteins may provide a method with higher sensitivity since unlike proteins, levels of mRNA can be directly assessed without the need for catalytic activity. The transcriptome of a tissue is defined as the pattern of gene expression for the cells that compose such tissue. Forensic research has aimed to determine the transcriptome of body fluids commonly found at crime scenes and identify which mRNA transcripts are specific for which tissues (126). What is attractive about measuring levels of mRNA compared to proteins is the

possibility to use automated methods similar to those used for DNA.

To measure the levels of one specific mRNA transcript, first the total RNA of a tissue must be extracted and quantified. The extracted RNA is then incubated with DNase I, which will degrade any DNA co-extracted with RNA so it would not interfere with the reverse transcriptase reaction. Using random primers that anneal in the poly-A tail of mRNA or in other locations, a reverse transcriptase reaction is performed. The reverse transcriptase reaction is similar to PCR but uses enzymes called reverse transcriptase, which are capable of reading RNA, and synthesizing DNA (called cDNA), which is complementary to the RNA. Following this reaction, a quantitative PCR or end-point PCR will amplify the selected regions that correspond to the mRNA transcripts of interest for the target tissue (127). Additionally, a housekeeping gene is also amplified to allow the normalization of mRNA transcripts quantified. The amount of mRNA produced in a cell or tissue may be momentarily increased as a result of the cell cycle or other factors inherent to the status of the tissue (10). For that reason, the concentration of the specific mRNA transcript is always normalized with the concentration of a housekeeping gene

(128). The transcripts of a housekeeping gene are considered stable because of a constant need of such genes by the cells. Lack of normalization between the mRNA transcript of interest and the transcripts of housekeeping genes would lead to several errors while assessing the concentration or presence of a specific mRNA transcript since some are only expressed at specific points of the cell cycle, and when analyzing a tissue, several cells in different points of cell cycle are present.

For a quantitative PCR approach, the quantity of each transcript after normalization with the housekeeping gene is determined through comparison of fluorescence of unknown samples to those of standard solutions that have known concentrations (128). If end-point PCR is used, the amplicons may be analyzed by capillary electrophoresis (CE) similarly to what is done for STRs. Additionally, the use of high resolution melt analysis has also been tested for analysis of mRNA transcripts in forensic samples (129). Multiplexing several primers is also a possibility for this approach as long as they can be distinguished by the real-time instrument or by the CE application (130, 131).

One of the first reports to quantify mRNA for body fluid characterization in forensics was published in 1999 by Bauer et al. (132). In that report, the authors aimed to determine the presence of epithelial cells in menstrual blood in stains stored for 6 months. The positive results achieved initiated a body of research in stability and usefulness of mRNA transcripts for the identification of body fluids. Juusola and Ballantyne (126, 130, 133) were very active in the identification and in performing multiplex reactions of several mRNA transcripts specific for different tissues. Initially the analysis of transcripts was merely to determine its presence in only the target body fluid and it was achieved by gel electrophoresis (126), later with capillary electrophoresis (130) and in the last report

by the two authors the use of a real-time instrument with TaqMan® _{technology (Figure}

2.2). In real-time PCR, the determination of CT values for all transcripts in all tissues of

forensic interest allows the establishment of a reference table that would permit forensic analysts to identify the tissue of interest. The number of fluorophores available and the possibility that a mixture of body fluids and not a single source sample is present may limit the analysis of mRNA transcripts by this method. Additionally, individual variation in the mRNA expression of certain genes may also lead to false negatives (133).

Other reports (134) have used a multiplex approach similar to that of Juusola and Ballantyne (2007) using different mRNA transcripts with varying specificities for the target body fluid. With the aim of identifying the transcriptome of forensically-relevant tissues, in 2013, Park and colleagues (135) performed a whole-genome microarray for blood, saliva, semen and vaginal secretion in order to detect which mRNA transcripts were present in higher quantities in which body fluids. In total, they were able to select 137 candidate genes from which a second selection of 41 was made. Some of the body fluid-specific mRNA transcripts already reported by other groups also showed as candidate genes for those body fluids in the microarray, further confirming the previous work by others.

The main concerns with the use of RNA is the low stability of the molecule over time and when exposed to the elements, variation in its abundance on the tissue of origin and the need to use extra portions of evidence to extract RNA for body fluid

identification (136). Most of the forensic research using mRNA transcripts have done some level of stability studies and developed mock crime scene samples to determine the ability to detect mRNA in challenging samples. Despite the overall results being positive,

environments with high humidity can cause issues in the detection of mRNA and more studies are needed until this approach is robust enough for body fluid identification. Regarding the use of extra evidence to perform RNA extraction, even though the simultaneous extraction of DNA and RNA from the same piece of evidence is possible (137, 138), handling RNA in a laboratory must be accompanied with extra care because of the ubiquitous presence of RNases (including in human skin). For example, all water used, including that used to dilute buffers and other solutions must be treated with diethyl pyrocarbonate (DEPC), which includes long incubations so that the DEPC may combine with all RNases and a final autoclave step to remove the DEPC from the water(128). The limitation in the use of mRNA can also result from where the sample itself was collected, since the presence of RNases is ubiquitous in the human body (139).

Other issues related to the use of mRNA for body fluid identification include the use of DNase I to remove any DNA contaminating the extracted RNA (133, 140). The accidental release of this enzyme in a forensic laboratory could signify the impairment of DNA typing methods for several samples, even for samples processed several weeks after the release happens. To tackle some of the stability issues inherent to mRNA transcripts, some research has suggested the use of microRNA (112, 141). Because they are smaller, microRNA molecules present higher stability when compared to mRNA (142, 143), however the issues with DNase I and the ubiquitous presence of RNases are still a concern even for these smaller RNA molecules.

Using the central dogma as a reference, it makes sense to look at DNA instead of RNA or proteins to the identification of body fluids. DNA presents higher stability, it is the source of the code for all genes and for differential gene expression in tissues.

Moreover, the extraction, use and storage of DNA in forensic laboratories has been already validated for several years (68, 72).

The reliability of a confirmatory test for body fluid identification for forensic purposes depends on its capacity to accurately determine which body fluid is present regardless of its concentration, its purity and whether its mixed or not with other body fluids. The accurate determination of which body fluid is in the origin of a DNA samples can be extremely important to solve some crimes, namely those that rely on

determination of the suspect’s involvement in the crime. If a serology test shows false positives or false negatives in one of the validation studies performed, it creates reasonable doubt for any piece of evidence analyzed with such test. Adding to this, the amount of evidence recovered from the crime scene may already be very limited, therefore any portion of it being used for a test known to show false results may be considered a waste of sample. For this reason there is the need to establish a body fluid determination test that can achieve similar degree of certainty as those methods currently observed in DNA typing technologies. Due to the fact that all forensic laboratories are well equipped to collect, extract, store and analyze DNA, it seems logical to choose DNA as the biomolecule for body fluid discrimination in forensics.

CHAPTER III – FORENSIC EPIGENETICS