1. IntroductionAnalysis of biological evidence typically begins with a preliminary screening for the presence of biological material. Body fluids are commonly identified by chemical analyses, immunological assays, protein catalytic activity tests, spectroscopic methods and microscopy [1Analysis of body fluids for forensic purposes: from laboratory testing to non-destructive rapid confirmatory identification at a crime scene., 2An J.H. Shin K.J. Yang W.I. Lee H.Y. Body fluid identification in forensics.]. New methods include mRNA profiling [3Haas C. Hanson E. Ballantyne J. mRNA and MicroRNA for Body Fluid Identification., 4Messenger RNA profiling: a prototype method to supplant conventional methods for body fluid identification., 5Multiplex mRNA profiling for the identification of body fluids., 6Haas C. Klesser B. Maake C. Bar W. Kratzer A. mRNA profiling for body fluid identification by reverse transcription endpoint PCR and realtime PCR., 7The development of a mRNA multiplex RT-PCR assay for the definitive identification of body fluids., 8mRNA profiling for body fluid identification by multiplex quantitative RT-PCR.], tissue specific DNA methylation [9Gomes I. Kohlmeier F. Schneider P. Genetic markers for body fluid and tissue identification in forensics., 10Frumkin D. Wasserstrom A. Budowle B. Davidson A. DNA methylation-based forensic tissue identification., 11Vidaki A. Giangasparo F. Syndercombe Court D. Discovery of potential DNA methylation markers for forensic tissue identification using bisulphite pyrosequencing.], proteomics [[12]Yang H.Y. Zhou B. Deng H.T. Prinz M. Siegel D. Body fluid identification by mass spectrometry.] and microbial analysis [13Clarke T.H. Gomez A. Singh H. Nelson K.E. Brinkac L.M. Integrating the microbiome as a resource in the forensics toolkit., 14Hanssen E.N. Avershina E. Rudi K. Gill P. Snipen L. Body fluid prediction from microbial patterns for forensic application., 15Hanssen E.N. Liland K.H. Gill P. Snipen L. Optimizing body fluid recognition from microbial taxonomic profiles., 16Dobay A. Haas C. Fucile G. Downey N. Morrison H.G. Kratzer A. Arora N. Microbiome-based body fluid identification of samples exposed to indoor conditions.].Current body fluid identification methods do not reveal any information about the time since deposition (TsD) of biological traces. From a criminalistic point of view, estimation of when the crime was committed would be useful to determine the relevance of trace samples found at the scene, enable the verification of witnesses’ statements, limit the number of suspects, and help corroborate alibis [[17]Bremmer R.H. de Bruin K.G. van Gemert M.J. van Leeuwen T.G. Aalders M.C. Forensic quest for age determination of bloodstains.]. In recent years, several techniques for TsD determination of blood stains have been explored [17Bremmer R.H. de Bruin K.G. van Gemert M.J. van Leeuwen T.G. Aalders M.C. Forensic quest for age determination of bloodstains., 18Anderson S. Howard B. Hobbs G.R. Bishop C.P. A method for determining the age of a bloodstain.]. These techniques include spectroscopy, chromatography, and electron spin resonance. These methods are promising for blood traces, but limited to colored stains and thus not easily transferred to other, white or nearly colorless traces such as saliva, semen, or vaginal fluid.Several attempts have been made to estimate TsD based on RNA analyses. Most often, degradation patterns of different human messenger RNA (mRNA) and ribosomal RNA (rRNA) transcripts in blood were monitored [18Anderson S. Howard B. Hobbs G.R. Bishop C.P. A method for determining the age of a bloodstain., 19Bauer M. Polzin S. Patzelt D. Quantification of RNA degradation by semi-quantitative duplex and competitive RT-PCR: a possible indicator of the age of bloodstains?., 20Anderson S.E. Hobbs G.R. Bishop C.P. Multivariate analysis for estimating the age of a bloodstain., 21A method to estimate the age of bloodstains using quantitative PCR.]. Bauer et al. [[19]Bauer M. Polzin S. Patzelt D. Quantification of RNA degradation by semi-quantitative duplex and competitive RT-PCR: a possible indicator of the age of bloodstains?.] showed that mRNA could be analyzed from blood samples which were 15 years old, and found that mRNA degradation levels differ significantly among samples of different ages. Anderson et al. [18Anderson S. Howard B. Hobbs G.R. Bishop C.P. A method for determining the age of a bloodstain., 20Anderson S.E. Hobbs G.R. Bishop C.P. Multivariate analysis for estimating the age of a bloodstain.] focused on transcripts of two housekeeping genes, ß-actin and 18S-rRNA, and found that the relative degradation patterns of these as examined through quantitative PCR (qPCR) could be used to estimate the age of blood stains. An optimized qPCR approach involving the quantification of small and long fragments within ß-actin mRNA and 18S-rRNA allowed to distinguish fresh from 6-day old blood samples, and 6-day from older blood samples. A subsequent study by Alshehhi et al. [[22]Alshehhi S. Haddrill P.R. Estimating time since deposition using quantification of RNA degradation in body fluid-specific markers.] focused on semen and saliva stains and found that mRNA molecules specific to these body fluids showed a unique degradation pattern over a period of one year, whereas miRNAs and the U6 reference gene were shown to be stable. The relative expression ratio was suggested as a potential method for TsD estimation.Most notably, the introduction of massively parallel sequencing (MPS) has enabled the shift from utilizing small numbers of RNA markers to exploring and using the full range of RNA sequencing transcripts for TsD estimation. However, the low quantity and quality of RNA in forensic samples poses a genuine challenge in practice. Nevertheless, Lin et al. [[23]Lin M.H. Jones D.F. Fleming R. Transcriptomic analysis of degraded forensic body fluids.] successfully sequenced total RNA from forensically relevant body fluids (blood, menstrual blood, oral mucosa/saliva and vaginal secretion) and assessed the global gene expression levels over time. It was shown that sequencing reads with relatively good quality (85% of reads had a Q score >30) can be obtained from both fresh and aged (two and six weeks old) samples. Recently, Weinbrecht et al. [[24]Weinbrecht K.D. Fu J. Payton M.E. Allen R.W. Time-dependent loss of mRNA transcripts from forensic stains.] assessed the global abundance and degradation pattern of mRNA in blood, saliva, semen and vaginal secretion samples, across a one year interval. The abundance of the different transcripts decreased over time but at different rates.In recent years, microbial forensics has emerged as a new research area [25Microbial forensics: new breakthroughs and future prospects., 26Forensic body fluid identification: state of the art., 27Metcalf J.L. Xu Z.Z. Bouslimani A. Dorrestein P. Carter D.O. Knight R. Microbiome tools for forensic science.]. Bacteria colonize vast areas of the human body such as skin, the gastrointestinal and urogenital tracts and the oral cavity. Overall, the number of human and bacterial cells in the human body are estimated to be similar [28Sender R. Fuchs S. Milo R. Revised estimates for the number of human and bacteria cells in the body., 29Sender R. Fuchs S. Milo R. Are we really vastly outnumbered? Revisiting the ratio of bacterial to host cells in humans.]. However, human body sites display large variability in the density as well as the composition of microorganisms. In fact, studies analyzing the human microbiome have revealed that different body sites harbor very distinctive bacterial communities, allowing for body fluid identification based on microbial communities [26Forensic body fluid identification: state of the art., 30Molecular approaches for forensic cell type identification: on mRNA, miRNA, DNA methylation and microbial markers.]. Several bacterial markers have been proposed for the identification of vaginal secretion [31mRNA profiling using a minimum of five mRNA markers per body fluid and a novel scoring method for body fluid identification., 32The use of bacteria for the identification of vaginal secretions., 33Akutsu T. Motani H. Watanabe K. Iwase H. Sakurada K. Detection of bacterial 16S ribosomal RNA genes for forensic identification of vaginal fluid.], saliva [34Power D.A. Cordiner S.J. Kieser J.A. Tompkins G.R. Horswell J. PCR-based detection of salivary bacteria as a marker of expirated blood., 35Nakanishi H. Kido A. Ohmori T. Takada A. Hara M. Adachi N. Saito K. A novel method for the identification of saliva by detecting oral streptococci using PCR.] and feces [[36]Nakanishi H. Shojo H. Ohmori T. Hara M. Takada A. Adachi N. Saito K. Identification of feces by detection of Bacteroides genes.].Traditionally, hypervariable DNA regions of the 16S-rRNA gene are sequenced and used to analyze the bacterial composition of samples [[13]Clarke T.H. Gomez A. Singh H. Nelson K.E. Brinkac L.M. Integrating the microbiome as a resource in the forensics toolkit.]. In order to distinguish body sites using such datasets, different statistical approaches have been proposed. For instance, Hanssen et al. [[14]Hanssen E.N. Avershina E. Rudi K. Gill P. Snipen L. Body fluid prediction from microbial patterns for forensic application.] showed that a combination of principal component analysis (PCA) and linear discriminant analysis (LDA) can successfully be used to differentiate samples from saliva being deposited on skin and samples of skin only. In addition, Tackmann et al. [[43]Tackmann J. Arora N. Schmidt T.S.B. Rodrigues J.F.M. von Mering C. Ecologically informed microbial biomarkers and accurate classification of mixed and unmixed samples in an extensive cross-study of human body sites.] have shown that machine learning algorithms trained on large heterogeneous datasets provide high accuracy when predicting body fluids, not only from single source samples but also from mixtures generated in-silico. In another study, Hanssen et al. [[15]Hanssen E.N. Liland K.H. Gill P. Snipen L. Optimizing body fluid recognition from microbial taxonomic profiles.] developed a prediction model using partial least squares (PLS) in combination with LDA using data from the Human Microbiome Project for the identification of samples originating from the oral, nasal and vaginal cavity as well as skin and feces.The human microbiome not only varies across body sites but also across individuals, across intrinsic conditions (e.g. diet, medication and diseases), and across environmental/extrinsic conditions [36Nakanishi H. Shojo H. Ohmori T. Hara M. Takada A. Adachi N. Saito K. Identification of feces by detection of Bacteroides genes., 37Human Microbiome Project Consortium Structure, function and diversity of the healthy human microbiome., 38Integrative H.M.P. Research Network Consortium The Integrative Human Microbiome Project., 39Integrative H.M.P. Research Network Consortium The Integrative Human Microbiome Project: dynamic analysis of microbiome-host omics profiles during periods of human health and disease., 40Lloyd-Price J. Mahurkar A. Rahnavard G. Crabtree J. Orvis J. Hall A.B. Brady A. Creasy H.H. McCracken C. Giglio M.G. McDonald D. Franzosa E.A. Knight R. White O. Huttenhower C. Erratum: Strains, functions and dynamics in the expanded Human Microbiome Project., 41The human microbiome: at the interface of health and disease., 42Gilbert J.A. Blaser M.J. Caporaso J.G. Jansson J.K. Lynch S.V. Knight R. Current understanding of the human microbiome., 44David L.A. Maurice C.F. Carmody R.N. Gootenberg D.B. Button J.E. Wolfe B.E. Ling A.V. Devlin A.S. Varma Y. Fischbach M.A. Biddinger S.B. Dutton R.J. Turnbaugh P.J. Diet rapidly and reproducibly alters the human gut microbiome., 45Costello E.K. Lauber C.L. Hamady M. Fierer N. Gordon J.I. Knight R. Bacterial community variation in human body habitats across space and time., 46Caporaso J.G. Lauber C.L. Costello E.K. Berg-Lyons D. Gonzalez A. Stombaugh J. Knights D. Gajer P. Ravel J. Fierer N. Gordon J.I. Knight R. Moving pictures of the human microbiome.]. While intrinsic conditions have been explored extensively, little is known about the changes in microbial composition over time once a sample is exposed to the environment outside the human body. Since samples collected at crime scenes are subjected to the extracorporeal environment and are stored for some time before analysis, the composition of the microbial community might change with time, depending on environmental conditions. Dobay et al. [[16]Dobay A. Haas C. Fucile G. Downey N. Morrison H.G. Kratzer A. Arora N. Microbiome-based body fluid identification of samples exposed to indoor conditions.] collected samples of forensically relevant body fluids (blood, saliva, skin, semen, menstrual blood and vaginal secretion) and assessed the microbial composition directly after collection and after one month. They found that characteristic signatures allowing the identification of the body site could be obtained even after prolonged storage time. Subsequent studies by Diez-Lopez et al. [47Díez López C. Montiel González D. Haas C. Vidaki A. Kayser M. Microbiome-based body site of origin classification of forensically relevant blood traces., 48Díez López C. Vidaki A. Ralf A. Montiel González D. Radjabzadeh D. Kraaij R. Uitterlinden A.G. Haas C. Lao O. Kayser M. Novel taxonomy-independent deep learning microbiome approach allows for accurate classification of different forensically relevant human epithelial materials.] also indicate the stability and reliability of such signatures over longer periods of time when the samples were stored at 4 °C as well as room temperature for up to seven years.Microbial sequencing for body fluid prediction has already been applied to the investigation of two criminal cases in the Netherlands [[49]Quaak F.C.A. van de Wal Y. Maaskant-van Wijk P.A. Kuiper I. Combining human STR and microbial population profiling: Two case reports.]. In these two cases, human STR profiling and microbial population profiling were applied to the same trace sample in order to simultaneously investigate the donor as well as the bodily origin of the sample. In both cases, the analyses were helpful in the court proceedings, and the suspect was eventually convicted.In a recent study, we published an optimized RNA-Seq workflow for forensic samples (blood, menstrual blood, saliva, semen, skin and vaginal secretion) [[50]Salzmann A.P. Russo G. Aluri S. Haas C. Transcription and microbial profiling of body fluids using a massively parallel sequencing approach.]. Body fluid specific bacterial signatures could be identified in fresh and aged samples. In addition, aged samples showed a higher level of RNA degradation and decreased bacterial diversity. The aim of this study was to gain further insights into the changes of microbial communities of forensically relevant body fluids over time and to determine the utility of microbial RNA markers for TsD estimation. Blood, menstrual blood, saliva, semen, and vaginal secretion samples were exposed to different environmental conditions (indoor, dry, room temperature versus outdoor, exposed to the environment but protected from rain). Total RNA was analyzed at seven different time points (time point 0/deposition, 1 day, 7 days, 4 weeks, 6 months, 1 year, 1.5 years). The microbial composition at different taxonomic levels was assessed across all time points, for indoor and outdoor conditions. To our knowledge, this is the first study evaluating microbial markers for TsD estimation and in particular microbial RNA markers. In a parallel manuscript, we assessed the degradation of human mRNA transcripts over time as an indicator of the time since deposition in biological crime scene traces [[51]Salzmann A.P. Russo G. Kreutzer S. Haas C. Degradation of human mRNA transcripts over time as an indicator of the time since deposition (TsD) in biological crime scenetraces.].4. Discussion

Determination of TsD is a desired addition to the current forensic toolbox and will aid investigators and courts of justice in assessing the relevance of biological traces. It is expected to contribute to a more accurate identification of true perpetrators as well as to clear the wrongfully accused. To date, there is no reliable method available to estimate the age of stains deposited at crime scenes despite efforts being made in this field of work. The aim of our study was to determine the utility of microbial RNA markers for TsD estimation. Hence, we examined RNA-Seq data from blood, menstrual blood, saliva, semen, and vaginal secretion over seven time points, ranging from fresh to 1.5 years. One set of samples was stored indoors, the other set was exposed to the outdoor environment. The microbial composition at different taxonomic levels was assessed across all time points for outdoor and indoor conditions.

Inspection of the taxonomic composition of both outdoor and indoor samples at the domain level (Fig. 1, Fig. 5) as well as at the phylum level (Fig. 3, Fig. 6) indicated that the five investigated body fluids can be categorized into three groups based on similarity: (a) blood and semen, (b) menstrual blood and vaginal secretion and (c) saliva. This finding is most likely due to the origin of the different body fluids and their respective microbial composition. For example, blood and semen from healthy individuals contain a minimal bacterial fraction, while the bacterial load of menstrual blood, vaginal secretion and saliva is high under normal conditions. Menstrual blood and vaginal secretion are more similar to each other due to their common body site of origin.Overall, in outdoor samples there was a shift from prokaryotes to non-human eukaryotes between 4 weeks and 6 months (Fig. 1). This change is driven primarily by the increase of Quercus (Oak tree) and Panicum (switch grass) in all samples (Fig. 2). Oak trees are common in the park surrounding the university building, where the study was conducted. The presence of switch grass was unexpected at first because it is not native to the area, but we discovered that it is commonly utilized for decorative purposes in gardens. The body fluid samples were placed outdoors in March / April. Oak trees blossom around April to May, while switch grass starts to blossom in August. In addition, signatures of Olea (Olive tree) and Cercospora (ascomycete fungi) were identified. Interestingly, the pollen from Olea species has been found to travel long distances. For instance, Olea europea pollen most likely originating in Italy has been detected in Hungary [[63]Tedeschini E. Udvardy O. Sofiev M. Palamarchuk J. Makra L. Magyar D. Long distance transport of Olea europea pollen over Central Europe.]. Our results indicate, therefore, that outdoor samples contain not only environmental RNA from the immediate vicinity but possibly also from more distant locations, for example as a result of pollen transportation from other areas. Overall, we hypothesize the increase in environmental RNA to be useful as a first indicator for sample age, particularly if signatures specific to the sample location can be identified. It is, however, important to note that these signatures may change across seasons. Overall, further studies are needed in order to verify these findings and also to further assess the effect of different seasons and conditions on the samples.When focusing on the prokaryotic fraction, we found three phyla that were of particular i