Radiation-induced cell killing is highly dependent upon buffer treatment (filtration compared to autoclaving) due to metal-catalyzed formation of hypochlorite: a cautionary note.

Buffer solutions used in experiments in radiation biology may be sterilized by either autoclaving or filtration. We show here that for phosphate-buffered saline such differences in buffer treatment may result in widely differing dose-effect curves for cell killing. The temperature-dependent transformation of monophosphate ions into di- or polyphosphate evidently proceeds to an appreciable extent upon autoclaving the buffers at 120 degrees C for 10 to 20 min. This increases the capability of the buffer to chelate spurious metal contaminations and, as a consequence, to reduce the amount of cytotoxic hypochlorite being produced. Depending on conditions of buffer treatment we have observed dose modification factors for the colony-forming ability of yeast cells up to the order of 3. Thus effects due to buffer treatment might easily outweigh the effect which the experiment was originally designed to determine. We strongly advise, therefore, that results of parallel sets of experiments in which different methods of buffer sterilization have been used should not be compared directly.