DNA methylation-based forensic tissue identification

Identifying the source tissue of biological material found at crime scenes can be very informative in a number of cases. Despite their usefulness, current visual, catalytic, enzymatic, and immunologic tests for presumptive and confirmatory tissue identification are applicable only to a subset of samples, might suffer limitations such as low specificity, lack of sensitivity, and are substantially impacted by environmental insults. Moreover these assays are incompatible and thus cannot be multiplexed. Thus they are less amenable to automation. In addition their results are operator-dependent. A better alternative approach is tissue identification based on messenger RNA (mRNA) or microRNA (miRNA); however, RNA is not as stable as DNA, and requires the use of non-standard procedures by forensic laboratories. Herein a DNA-based assay is described that enables tissue identification based on detection of tissue-specific methylation patterns. DNA samples are subjected to digestion by a methylation-sensitive restriction endonuclease followed by multiplex amplification of specific genomic targets with fluorescent-labeled primers, capillary electrophoresis of amplification products, and automatic signal analysis by dedicated software, yielding the source tissue of the sample. The single tube assay was designed for easy integration by forensic laboratories (as the assay utilizes the same platforms as current forensic STR profiling). The system is fully automatable, provides operator-independent results, and allows combining tissue identification with profiling in a single procedure. The assay was tested on 50 DNA samples from blood, saliva, semen, and skin epidermis, and the source tissue was successfully identified in all cases. Detection of semen and DNA profiling were combined into one assay and the ability to detect mixtures of semen and saliva in various ratios was demonstrated. The assay correctly detected semen in all samples where it was present, and the calculated percentage of semen was comparable to the fraction of semen in the samples. The results demonstrate that methylation-based tissue identification is more than a proof-of-concept. The methodology holds promise as another viable forensic DNA analysis tool for characterization of biological materials. (C) 2010 Elsevier Ireland Ltd. All rights reserved.