Experiments are done with DU-145 human prostate carcinoma cells. A novel co-culture system was designed for bystander effect detection: cells growing on the bottom mylar membrane are directly exposed to alpha particles; cells co-cultivated on an inserts, positioned 3 mm above the bottom cell layer and out of the range of the alpha particles, serve as the non-targeted bystander cells. DNA damage, as measured by micronucleus (MN) formation in the non-targeted DU-145 cells was used as a measure of the bystander effect. MN frequency in binucleated cells growing on the insert increased ~ 1.5 fold when the cells on the mylar were irradiated for 1 min. Irradiation of medium alone produced a 1.1 fold increase in MN frequency in the non-targeted cells. The results suggest that the irradiated tumor cells release some factor(s), which can diffuse through the medium and cause damage to the non-targeted tumor cells. Further more, the mechanism of bystander effect is explored. Addition of 1% DMSO (a free radical scavenger) to the medium reduced MN frequency to control level, indicating that free radical(s) are involved in the bystander effect signaling. To narrow down the free radial candidates, PTIO, a nitric oxide specific scavenger, was added to the medium and reduced MN frequency to ~1.1 fold of control, which suggests that nitric oxide plays a role in bystander effect. Transfer of medium from irradiated cells onto un-irradiated cells is being used to test the stability of the released factor(s). The preliminary results show that the factor(s) are stable in the medium for up to 2 hours. More experiments are underway to determine the identity and the stability of the released factor and to explore whether this released factor may exhibit synergy with added chemotherapy agents in these non-targeted tumor cells.