Identification Phylogenetics and Population Genetics

Correctly identifying the insect specimen can be of crucial importance for a forensic entomological analysis (Smith 1986). One might think that this is self-evident for a forensic entomologist, but the situation is complicated. Many young immature stages of several forensically relevant taxa (like e.g. the Muscidae) are difficult to identify and would usually be reared to the adult stage for getting a safe ID. For some groups, like the sarcophagids, the situation is even worse, and morphological characters of the immature don't allow for identification.

Here, DNA methods may be a powerful tool. They have been used to identify specimens that are impossible or difficult to distinguish based on morphology. However, the work done so far has been very uneven in its coverage of forensically important insect taxa. The calliphorids are, by far, the most represented group in species identification (Harvey et al. 2003; Ratcliffe et al. 2003; Chen et al. 2004; Wells and Williams 2007). This reflects their importance as indicator species of the postmortem interval and, probably, the relative ease in identifying adult specimens to be used for DNA extraction. The sarcophagids, also important for postmortem interval estimation, require more specialized knowledge compared to the calli-phorids, and fewer DNA data exist for these flies (Wells et al. 2001a; Ratcliffe et al. 2003; Zehner et al. 2004b). Scattered DNA records exist for a few species from other forensic insect groups, such as the phorid Megaselia scalaris (GenBank record AF217464, unpublished) and the muscid Musca domestica (AY818108, unpublished). Expanding the taxonomic breadth of DNA data for specimen identification will probably require that such studies be done more in cooperation with taxonomic experts who are not now involved in forensic entomology research.

In most published papers on DNA-based specimen identification, only a small number of individuals from each species were analyzed. The issue of replication when evaluating a species-diagnostic test has received little attention, even though the reliability of a procedure probably cannot be known until this issue is addressed (Wells and Williams 2007). In contrast to the DNA methods used to identify an individual (Evett and Weir 1998), there is no theoretical standard for estimating the replication needed to validate a species-diagnostic test. The development of such an analytical framework is clearly needed, not only for forensic entomology, but also for other forensic fields that involve species identification such as consumer fraud (Lenstra 2003), enforcement of conservation laws (Wan and Fang 2003), and bioterrorism (Jones et al. 2005).

Several loci have been used to make species determinations of forensically related fly species. The most popular mitochondrial DNA markers used for this purpose include the cytochrome oxidase subunit one (COI) gene (Wells and Sperling 2001; Zehner et al., 2004a, b; Wells and Williams 2007) and the cytochrome oxidase subunit two (COII) gene (Sperling et al. 1994; Wells and Sperling 2001; Wallman and Donnellan 2001). Other mitochondrial markers that have been used include the control region (Thyssen et al. 2005), subunit 5 of the NAD dehydrogenase (ND5) gene (Zehner et al., 2004a, b), and the tRNA-leucine gene (Wells and Sperling 2001). Nuclear markers have been utilized, as well, such as the internal transcribed spacer regions (Ratcliffe et al. 2003), and the gene for 28S ribo-somal RNA (Stevens and Wall 2001). Although COI and COII are recommended by many authors as the best marker for identification of forensically important insects, some closely-related species cannot be differentiated in this manner (Wallman and Donnellan 2001; Wells et al. 2004). Future research should investigate other markers for closely-related species.

The methodologies vary, as well, and they include direct sequencing (Harvey et al. 2003; Chen et al. 2004; Zehner et al. 2004b; Ames et al. 2006), polymerase chain reaction-restriction fragment length polymorphism (PCR-RFLP) analysis (Sperling et al. 1994; Malgorn and Coquoz 1999; Thyssen et al. 2005), random amplified polymorphic DNA (RAPD) analysis (Benecke 1998), and allozyme electrophoresis (Wallman and Adams 2001). Although mtDNA sequencing is a particularly robust technique, the lesser-used methods (i.e., PCR-RFLP, RAPD, and allozyme electrophoresis) can be fast, inexpensive, and highly discriminating. Therefore, they deserve further evaluation as forensic tools.

Most of the genetic research in forensic entomology has focused on individual or species identification, as previously described. Investigations of larger phyloge-netic patterns are limited (Stevens and Wall 1996; Stevens and Wall 1997; Wells and Sperling 1999; Bernasconi et al. 2000; Stevens and Wall 2001; Stevens et al. 2002; Wells et al. 2004, 2007). However, these basic studies can reveal forensically important information by uncovering a so-called "cryptic" species, one that could not be recognized by traditional means (Wallman et al. 2005). This kind of analysis may require a greater amount of genetic information than is normally used for simple identification. One approach is to use several mitochondrial genes (Wallman et al. 2005). In fact, it is possible to sequence the entire blowfly mitochondrial DNA molecule (Junqueira et al. 2004).

Perhaps more promising would be an analysis based on a combination of mito-chondrial and nuclear loci. This has been done to a limited extent, for example using COI and 28S (Stevens et al. 2002). At present, however, it is much more difficult to generate useful nuclear DNA sequences compared to mitochondrial DNA when examining a previously unstudied taxonomic group. Fortunately, the phylogeny of the Order Diptera is an extremely active are of research (see www. Forensic entomologists should pay close attention to new protocols developed for other fly taxa, because that technology might be applied to forensically important groups with little, or no, modification.

The population genetics of forensically important insects have been almost untouched. Yet such information is relevant to forensic entomologists for at least two reasons (Böhme 2006). The discovery of regional genetic differences would suggest that regional differences could exist in a forensically important characteristic of a species, such as development rate as a function of temperature. Regional genetic differences could also make it possible to infer the postmortem movement of a corpse if a larva from the body has a genotype characteristic of another location.

For these reasons, we expect to see an expansion in population genetic studies of carrion insects. Although STR genotypes are particularly valuable for this purpose, the development of STR methods for groups such as blowflies has been slow because such loci appear to be much less common in insect genomes compared to those of vertebrates (Ji et al. 2003). Nevertheless, some STR primers have been developed for some calliphorid species (Florin and Gyllenstrand 2002; Torres et al. 2004; Torres and Azeredo-Espin 2005). This demonstrates that it can be done, and it may be possible to apply these methods with little or no change to other, closely-related, species.

All of the aforementioned studies have generated valuable data and stand as the foundation of genetic research in forensic entomology. However, validation studies are essential before these methods are used for casework. That is, how confident can we be that a proposed method will accurately identify a specimen? In part, this depends on adequate replication, something that has not yet been determined (see above). However, this doesn't prevent validation efforts from being carried out. For example, Wells and Williams (2007) tested the accuracy of a published method for identifying individuals in the blowfly subfamily Chrysomyinae in North America (Wells and Sperling 2001). They obtained genetic data for hundreds of additional identified specimens, and analyzed them as if they were unknowns. All were correctly "identified" by the genetic method. More validation studies, such as this, are necessary.

0 0

Post a comment