J. Mol. Biol. (1974) 86, 109-127 Supercoiled Circular DNA in Crown-gall Inducing Agrobacterium Strains i. ZA~N, N. vA~ L~m~BEKE, It. TEUCH¥I" , M. VAN MONTAGU'~ AND J. SCH~ELL Laboratorium veer Genetika and Laboratorium veer Histologiet Rijksuniversiteit Gent, Gent, Belgium (Received 16 October 1973) By alkaline sucrose gradient, neutral sucrose gradient and dye-buoyant density centrifugation a large plasmid was shown to be present in the crown-gall indue- ing Agrobacterium tumefaciens strain B6-83. Measurements of contour lengths carried out on electron micrographs resulted in a mean length of 54.1 ~m, corresponding to a molecular weight of 112 × 10 ~. Only one or a few copies of this plasmid are present per bacterial chromosome. Mitomycin C induction has no influence on the amount of plasmid I)NA in the cell. At present this plasmid must be considered as cryptic, for no genetic markers on it are known. Furthermore, large plasmids were isolated from crown-gall inducing strains belonging to seven Agrobaeterium groups described by Kersters et al. (1973). Contour length measurements carried out on the plasmids isolated from the various crown-gall inducing strains fell in the range from 54.1 ~m to 75.4 ~m, depending on the strain examined. We were not able to find such plasmids in eight non-pathogenic strains belonging to four of the same groups. The hypothesis is formulated that the large plasmid present in crown-gall inducing bacteria could be the "tumor-inducing principle". 1. Introduction Agrobacterium tumefaciens (Smith & Townsend, Corm.), for which Heberlein et al. (1967) propose the name of Bhizobium radiobacter, vat. tumefaciens, induces crown- gall tumours in many, mostly dicotyledonous, plants. The bacterium itseff cannot be detected intraeelhilarly, either in the plant cells that are being transformed, or in the cells of sterile crown-gall turnouts grown in vitro (Tourneur & Morel, 1970). Once the tumour induction has taken place, the autonomous proliferation of the tumour cells becomes entirely independent of the bacteria. This observation led Braun (1947) to introduce the concept of a tumour-inducing principle, postulating that this principle was transferred by the bacteria to the plant cell to induce trans- formation (Kern, 1965). Since then many attempts have been made to elucidate the nature of this tumour- inducing principle. Various sterile bacterial fractions were tested for pathogenicity with no reproducible positive results (Manil et al., 1955; Gribnau, 1965; Bieber & Sarfert, 1968; Gribnau & Veldstra, 1969). The possible role of growth substances 109