Attempts to detect Agrobacterium tumefaciens and bacteriophage PS8 DNA in crown gall tumors by DNA · DNA-filter hybridization

Abstract A systematic study of the DNA · DNA-filter reaction is presented which measures its ability to detect small amounts of simple DNA (bacterial or bacteriophage) in model mixtures of DNA immobilized on filters. Saturation curves show qualitatively that significant binding occurs when there is 10% Agrobacterium tumefaciens DNA on the filter but not 1%. PS8 bacteriophage DNA is detectable at a level of 0.1%. True saturation is not attained in the bacterial DNA reaction: radioactivity bound represents only 3% of the theoretical saturation value. The bacteriophage DNA reactions attain 15–30% of the expected saturation value. When crown gall tumor DNA filters were tested for the presence of A. tumefaciens or PS8 bacteriophage DNA by saturation reactions, an apparently significant amount of binding was observed compared with usual background levels for heterologous DNA filters. However thermal dissociation profiles revealed that no well-matched duplexes were formed. Normal tobacco callus DNA filters exhibited the same type of binding of labeled DNA to a similar extent (50–100% as much as tumor DNA filters). Both types of DNA-filters bound Bacillus subtilis and bacteriophage T4 DNA as efficiently as A. tumefaciens and PS8 DNA. The high non-specific background binding of labeled DNA by filters containing DNA isolated from plant tissue culture materials is ascribed to low single strand molecular weight of the filterbound DNA. This study provides no evidence for foreign DNA in crown gall tumors, and raises objections to the interpretation of the data of earlier investigators (Quetier, F., Huguet, T. and Guille, E. (1969) Biochem. Biophys. Res. Commun. 34, 128–133 and Srivastava, B.I.S. (1970) Life Sci. 9, 889–892) who claimed to detect A. tumefaciens DNA in crown gall tumors by DNA · DNA-filter hybridization.