FORENSIC MOLECULAR BOTANY: IDENTIFICATION OF PLANTS FROM TRACE EVIDENCE

The database genbank at the National Institutes of Health contains numerous plant DNA sequences that can be used to identify botanical trace evidence and include both nuclear and chloroplast genes. The most extensively characterized locus across all plant species is the large subunit of ribulose 1,5- bisphosphate carboxylase (rbcL), encoded in the chloroplast genome. Due to its cellular abundance and the extensive collection of DNA sequences contained in the NIH database, the rbcL locus was selected as the model system to develop methods for molecular analysis of trace evidence. One locus to complement rbcL for the investigation of trace evidence includes the ITS (internal transcribed spacer) region. The ITS region of the nuclear genome flanks the 5.8S ribosomal DNA coding region, is multi- copy, and has more sequence variability as compared to the rbcL gene. Using the basic local alignment search tool (BLAST; 3), an evidentiary sequence can be compared to thousands of accessioned sequences from previously identified plants. BLAST is a web-based program that yields a list of organisms contained in the database with a sequence similar to the sequence of interest; the retrieved data starts with sequences having the greatest similarity. The matches can either be 100% or some lower percentage. Depending on the sequence of interest and the plant group that it belongs to, a high percentage match (near 100%) suggests a possible identification to the species level. However, if the match percentage is less than 100%, phylogenetic analysis is required to better characterize the sequence. Phylogenetic methods are based on models of DNA sequence evolution and are used to determine relationships among different organisms. These models are used in analysis programs such as PAUP (4). Because the NCBI database contains a sample of representative plant species, phylogenetic methods are useful for determining or confirming the genus or family to which a sequence belongs. By constructing a nucleotide alignment that includes the evidentiary sequence and closely related sequences (of known identity) the evidentiary sequence can be classified within a genus and / or family. The most information is obtained when a sequence is identified to the species level, but useful data is also gleaned from genus or family level identification. Sequences recovered from trace evidence may provide insight to the geographic origin of the evidence. Depending on the sample composition, a range of one to many different sequences can be recovered. Based on the ecology and habitat of the plants identified with these sequences, the geographic location can be inferred. The level of precision to which a sample's origin can be identified is often most influenced by the rarity and combination of the plants recovered. This combination of plants provides a 'description' for a sample that can match or exclude reference samples from a crime scene or other evidentiary materials. To begin the validation process, we report two investigations that were conducted on botanical mixtures. The first mixture was made by combining purified DNA from eight previously identified plants and the second mixture was dust obtained from a piece of mock evidentiary clothing of unknown origin. We