Scenarios involving intimate contact between partners are frequent in cases of homicide, sexual assault and gender-based violence. Saliva, a relevant vehicle for DNA transfer, is commonly encountered in forensic contexts. In addition, cigarette butts are among the most frequently collected forensic exhibits in crime scene investigations, where they are routinely subjected to DNA profiling. Interpreting mixed DNA traces is as relevant in terms of source attribution as it is with a view to underlying transfer mechanisms, particularly when activity-level propositions are disputed. However, limited data are available regarding the interpretation of mixed saliva traces on cigarette butts, which may lead to different interpretative scenarios due to secondary transfer within an activity-level framework. This study addresses this gap. 

Our study investigates DNA transfers, persistence, prevalence and recovery in two simulated scenarios: in the first, each partner of a couple smoked a cigarette at different time intervals after an intense kiss (immediately and after 15, 30, 60, 120, and 180 minutes); in the second, both partners alternately smoked the same cigarette. The primary objective is to evaluate how mixed DNA profiles change over the time frame. A second goal is to assess whether absence of a minor contributor reflects true donor disappearance or degradation. Finally, we aim to assess the influence of smoking devices, so that empirical data may determine whether discrimination is possible at the activity-level of proposed scenarios.

Ten volunteer pairs were recruited and a total of 130 cigarette butts were collected. Each sample underwent planned automated extraction (immediately, after 2 weeks, 3 weeks and 3 months) using QIAamp DNA Investigator on QIAcube Connect. A slice of cigarette filter paper was used to directly extract DNA, resulting in 520 extracts quantified with PowerQuant System. DNA profiles were obtained with PowerPlex Fusion 6C System and PowerPlex Y23 System on SeqStudio Genetic Analyzer for HID. LRmixStudio v.2.1.5 was used for profile comparisons and discriminative alleles counting, whereas sub-source LR evaluations were performed with the ShinyRFU2 tool based on EuroForMix v4.2.5. Finally, fictional scenarios were hypothesized and Bayesian networks were used to evaluate results in light of activity-level propositions.

Results suggest that transfer scenarios may be reconstructed. Moreover, a time window of non-self DNA disappearance may be assessed, with respect to impact of processing times and smoking devices as well. Our findings confirm that practical knowledge of transfer dynamics is important for activity-level evaluations highlighting how implementation in real caseworks is still challenging.

Authors

• Denise Gianfreda (Institute of Legal Medicine, Department of Biomedical, Metabolic and Neural Sciences, University of Modena and Reggio Emilia, Italy)
• Duncan Taylor (Forensic Science SA; College of Science and Engineering, Flinders University, Australia)
• Beatrice Corradini (Institute of Legal Medicine, Department of Biomedical, Metabolic and Neural Sciences, University of Modena and Reggio Emilia, Italy)
• Gianmarco Ferri (Institute of Legal Medicine, Department of Biomedical, Metabolic and Neural Sciences, University of Modena and Reggio Emilia, Italy)
• Francesca Ferrari (Institute of Legal Medicine, University Hospital of Modena, Italy)
• Elisa Manganelli (Institute of Legal Medicine, Department of Biomedical, Metabolic and Neural Sciences, University of Modena and Reggio Emilia, Italy)
• Egizia Baglieri (Institute of Legal Medicine, Department of Biomedical, Metabolic and Neural Sciences, University of Modena and Reggio Emilia, Italy)
• Anna Laura Santunione (Institute of Legal Medicine, Department of Biomedical, Metabolic and Neural Sciences, University of Modena and Reggio Emilia, Italy)
• Rossana Cecchi (Institute of Legal Medicine, Department of Biomedical, Metabolic and Neural Sciences, University of Modena and Reggio Emilia, Italy)