[Bacteriocins as cytotoxic agents in experimental neoplasms: considerations on the role of possible contaminants]

In most recent papers some Authors have demonstrated that several bacteriocins are able to markedly inhibit the division of various neoplastic mammalian cell lines; also in vivo the inhibition of tumor growth or even regression relative to controls are statistically significant. Since such preparations are "crude" and it's possible that contaminating agents may represent a possible source of variability in the results, mostly in vivo; we have examined a preparation of Pyocin employed in this study for the possible endotoxin contamination using the Limulus assay. Such test was positive in triplicate and semiquantitative determination has shown and endotoxin amount of 5 to 10 microgram/ml. This level is, in vitro, not anough to give cytotoxic effect, but in vivo is able to influence experimental results. In addition, since some bacteriocins and enterotoxins are produced under exactly identical conditions, there is this a possibility that the enterotoxin are coproduced in these preparations and are causing some of the cytotoxic effects on experimental tumors. A more complete separation and purification of preparations will be required before definitive comments.