Exploration Of Microbial Community Characteristics In Five Types Of Body Fluids And Their Changes Influenced By Three Factors And Screening Of Potential Microbial Markers For Differences Among Body Fluids

Background:Body fluid samples are one of the biological specimens found at crime scenes.Understanding the origin of these biomarkers in cases of violence such as sexual assault can provide key clues to the investigation of the case.Recently,the emerging full-length 16S rRNA gene sequencing based on single-molecule sequencing technology has the potential to delineate taxonomic and phylogenetic profiles that are closer to the true picture of microbial communities.Therefore,it is important to determine whether full-length 16S rRNA gene sequencing is suitable for forensic body fluid tracing and can yield reliable results by comparing it with the more popularly used 16S rRNA gene partial high-variance region sequencing data.Different types of body fluids contain different microbial communities.The limited populations and sample types covered in the current study could not meet the purpose of forensic body fluid spot tracing in Chinese populations,so five types of body fluid samples from young Han Chinese volunteers were analyzed for microbial community diversities and screened for microbial markers with variability among different types of body fluids.Since menstrual cycle and ethnicity affect the vaginal microbiota,revealing the dynamics of the vaginal flora during the menstrual cycle of the young Han women is a guideline for subsequent vaginal sample collection.In addition,most of the specimens have been exposed to the environment for a period of time when they are collected,so it is important to explore the microbial community characteristics of body fluid samples at different time points of exposure in order to find the most stable microbial markers of differences among body fluids,and to lay the foundation for accurate traceability of body fluids.The saliva of canine and feline,two of our favorite pets,are most likely to be left at the crime scene compared to the types of bodily fluids commonly found in forensics.The revelation of the microbial community characteristics of human,feline and canine saliva and the screening of microbial markers with human species-specific lay a foundation for the accurate identification of the species source of suspicious saliva samples in the crime scene involving pets,and they assist the investigators to quickly find available materials,saving reagents and money,and reducing prosecution time.Methods and contents:In this study,five types of body fluid samples(saliva,peripheral blood,vaginal secretions,menstrual blood and semen)were collected from healthy young Han individuals according to established sampling criteria,and some of the body fluid samples were exposed indoors for different durations to prepare exposure samples.In addition,saliva samples from humans,felines and canines were collected using swabs.Two sequencing methods,16S rRNA genes V3-V4(short-read length)and 16S rRNA genes V1-V9(full-length),were used to investigate the structures of the microbial communities in all of the above body fluid samples.Alpha diversity indices were used to explore community diversity in the sample,and beta diversities were used to detect differences among and within groups.The composition of microbial communities in different samples were display by relative abundance histograms,and various analytical methods such as principal coordinate analysis(PCoA),and non-metric multidimensional scaling analysis(NMDS)and analysis of molecular variance(MOVA)were used to investigate the discrepancies among microbes in different body fluids.Linear discriminant analysis effect size(LEfSe),random forests and receiver operating characteristic curve were utilized to screen for potential microbial markers that differed among body fluids and that could be used for inferencing the types of body fluids.Results and conclusions:(1)Compared with the results of full-length 16S rRNA gene sequencing,the results of short-read length 16S rRNA gene sequencing showed higher number of operational taxonomic units annotated at each taxonomic level,lower species annotation rate,and higher alpha diversity.The bacteria obtained by the two sequencing technologies were consistent at the phylum level but differed in their abundance,and varied at lower taxonomic levels(e.g.,genus or species).LEfSe demonstrated that the differential microbes filtered among the four types of body fluids differentiated in the two sequencing results.In conclusion,the use of full-length 16S rRNA gene sequencing could improve taxonomic precision and more clearly discriminate differences in the composition of microbial communities of various types of body fluid samples than short-read length 16S rRNA gene sequencing,which may be providing a more reliable tool for forensic body fluid identification.(2)PCoA,NMDS and ANOVA based on UniFrac distances all indicated similar microbial community compositions in vaginal secretions and menstrual bloods,and differences in microbial communities among saliva,peripheral blood and vaginal secretion(menstrual blood)samples.Therefore,microbial communities in the four different body fluids were available for retrospective studies of body fluids.(3)Vaginal flora in women varied individually.The microbial community diversities of vaginal samples from menstrual periods were higher than that of non-menstrual samples,and there was no significant difference among vaginal floras within a menstrual cycle,which were in a dynamically stable state.(4)With increasing exposure time,the Proteobacteria gradually increased in negative control and body fluid samples,and the Bacteroidetes and Firmicutes gradually decreased,but the relative abundance of dominant genera in various types of body fluids remained dynamically stable.The microbial community structures of samples from the same individual with different exposure durations were similar,and there were no significant differences in the microbial community structures among three types of body fluids at different exposure timepoints.The orthogonal partial least squares discriminant analysis and random forest analysis revealed that exposed body fluid samples could still be accurately traced through the characteristics of its microbial community.(5)The saliva samples of humans,felines and canines possessed their own unique microbial community compositions.The dominant phyla in both canine and feline saliva samples were Proteobacteria and Bacteroidetes,while the dominant phyla in human saliva were and Proteobacteria.The sex of canines and felines might not affect salivary flora,but breed could have an impact on their salivary microbial communities.Differential microbes among saliva of three species were screened by using multiple differential analysis methods,and the speciesspecific microbes in human saliva were Prevotella melaninogenica and Haemophilus parainfluenzae.The detection of microbes with human species-specificity and high relative abundance offered a possibility to identify saliva of human.(6)Multiple statistical analyses were used to screen for differential,stable and species-specific microbial markers of Haemophilus parainfluenzae in saliva and Lactobacillus iners in vaginal samples.The relative abundance of differential microbes screened from peripheral blood and semen samples were low and vulnerable to the surrounding environment,but the entire microbiome community could be used for traceability studies of peripheral blood and semen samples.Microbial community explorations of multiple types samples revealed microbial markers that can be used for accurate traceability studies of body fluids.