Introduction

Physical evidence, especially DNA, is increasingly important in the investigation of criminal cases and in the court of law in today's society. Diagnostic DNA analysis, including forensic applications, is often limited by components that interfere with the amplification, so-called PCR inhibitors (1). Several substances have been identified as PCR inhibitors, and some have been characterized with respect to their PCR-inhibitory mechanism(s) (2). The problem of inhibition is especially prominent in forensic DNA analysis, due to the nature of the sampling environment at crime scenes (3,4,5). The consequences of PCR inhibition in forensic casework may be severe. Failure to produce a DNA profile from a crime scene stain may leave a case unsolved or a person wrongly accused. As an illustration, a police force in the UK reported that 57% of swabs from bottles and cans failed to produce acceptable DNA profiles (6), which could be explained by a lack of sufficient amounts of DNA or the presence of PCR inhibitors. The common approach to overcoming PCR inhibition is extensive DNA purification (3,4,5). In special cases, isolation of single cells using laser-capture microdissection can be used (7). Simple dilution of the extract may be applied (8,9), but only if the DNA concentration is sufficiently high.

It has previously been observed that thermostable DNA polymerases of different origins may have different abilities to withstand the effects of various PCR inhibitors (10). For example, DNA polymerases from Thermus aquaticus (Taq) are more sensitive to the PCR inhibitors in human blood [i.e., hemoglobin, immunoglobulin G, and lactoferrin (11,12)] than DNA polymerases from Thermus thermophilus (Tth). Protein engineering has recently been used to improve the tolerance of Taq to PCR inhibitors in blood and soil in an attempt to develop a better Taq DNA polymerase for diagnostic PCR (13). A derivative of Taq, AmpliTaq Gold (Applied Biosystems, Foster City, CA, USA), is currently the standard DNA polymerase in several forensic DNA typing kits in use worldwide, including the widely used kits AmpFlSTR SGM Plus (Applied Biosystems) and PowerPlex 16 (Promega, Madison, WI, USA).

The objective of this study was to systematically investigate the potential of several DNA polymerase–buffer systems not currently used in forensic DNA analysis. The success rate of casework DNA samples was also investigated to evaluate the impact of PCR inhibition in routine analysis at the Swedish National Laboratory of Forensic Science (SKL). We present an approach to improve the quality of forensic DNA analysis and at the same time circumvent PCR inhibition in crime scene DNA samples. Modifying the PCR chemistry by employing alternative DNA polymerases and PCR facilitators was found to be a successful approach to generate high-quality DNA profiles without using additional sample preparation. The results were evaluated on two different instrumental platforms, and a statistical model for unbiased quality control of forensic DNA profiles was developed to quantify the results.

Materials and methods

The amplification efficiency (AE), dynamic range of amplification, and detection limit of nine DNA polymerase–buffer systems were evaluated on mock crime scene saliva samples using a standardized forensic singleplex real-time PCR assay. The three best-performing systems were then assessed under routine conditions at SKL. Short tandem repeat (STR) analysis was performed on 32 inhibited real crime scene saliva samples with low/medium DNA concentrations which had previously failed to produce complete forensic DNA profiles, to assess possible improvements compared with the standard method using AmpliTaq Gold. Finally, a statistical tool was developed to assess the quality of the forensic DNA profiles (i.e., capillary electrophoresis electropherograms) in an unbiased way.

Success rate of routine DNA profiling of crime scene saliva stains

We investigated the success of DNA typing of 1936 crime scene saliva samples (with DNA concentrations of 0.025–0.25 ng/µL) from volume crimes analyzed at SKL during 2007. DNA extraction was performed using Chelex beads (Bio-Rad Laboratories, Hercules, CA, USA) (14) with the addition of Centricon (Millipore, Billerica, MA, USA) purification (15) for visibly dirty samples. Ten microliters of DNA template was used in 25 µL AmpFlSTR SGM Plus PCR reactions. Mastermix preparation and PCR programming was performed according to the manufacturer's recommendations (AmpFlSTR SGM Plus PCR Amplification Kit User's Manual).