Automation of the standard DNA differential extraction on the Hamilton AutoLys STAR system: A proof-of-concept study

For several decades, a common approach for processing sexual assault evidence has been to use the "standard" differential extraction to separate the evidence into a non-sperm-cell fraction and a sperm-cell fraction for further analysis. In this standard approach (P. Gill et al., Nature 318 (1985) 577-579), an initial mild chemical lysis step preferentially digests the mainly epithelial cells, which allows for removing this lysate as a non-sperm-cell fraction. The undigested sperm cells in the remaining fraction may then be purified by a series of wash and centrifugation steps, after which more robust lysis conditions are used to digest this sperm-cell fraction. Although this standard approach has been generally effective, it has been difficult to fully automate, due to the variety of different types of manipulations required for sample processing (e.g., incubation, shaking, substrate separation by centrifugation, and multiple liquid transfers for sperm-pellet centrifugation and washing steps). We describe here a fully automated standard differential extraction procedure that uses the Hamilton AutoLys STAR liquid handling assay-ready workstation, which is configured with on-deck components for sample incubation, shaking and centrifugation steps, and works with unique AutoLys-A^® sample tubes for front-end sample processing. In this proof-of-concept procedure, up to 24 samples may be processed, "hands-free," in a single automated workflow. The automated procedure was tested by performing differential extractions on mock sexual assault swabs. For comparison, manual differential extractions were performed on identically prepared swabs in a side-by-side manner. DNA quantification and STR typing results showed that similar levels of separation efficiency were achieved for the sperm-cell fractions using both automated and manual procedures, although the results suggest that somewhat higher male DNA yields may be achieved for samples with extremely low semen levels (<∼0.1 μL) using the manual processing procedure. In addition to these mock samples, automated differential extractions were also performed on a set of authentic post-coital swabs (24, 48, 72, and 96 hours, post-coitus); primarily male STR profiles for the sperm-cell fractions were obtained for each sample in this set.