The interaction of actinomycin C3 and actinomine with DNA. A small-angle x-ray scattering study.

Abstract

Small-angle X-ray scattering was applied to solutions of calf thymus DNA and calf thymus DNA complexed with various amounts of actinomycin C3 or actinomine in phosphate-saline buffer at pH 6.9 and I equals 0.2. From the measurements of DNA in the absence of dye, two cross-section radii of gyration of R-c equals 0.875 plus or minus 0.015 nm and R-c2 equals 0.81 plus or minus 0.02 nm, and a mass per unit length of M/l equals 1906 plus or minus 43 daltons/nm resulted. The investigation of DNA complexed with dye revealed a decrease of the cross-section radii of gyration as compared to those for the DNA in the absence of dye and a relatively low increase of the mass per unit length on binding of actinomycin and a slight decrease of M/l on binding of actinomine. The latter results are interpreted on the basis of a length increase of the DNA double helix by 0.47 plus or minus 0.03 nm per actinomycin molecule and by 0.355 plus or minus 0.03 nm per actinomine molecule bound. The results for R-c and R-c2 obtained for the various samples of complexed DNA were extrapolated to the limiting binding ratio where each dye molecule is associated with a minimum of six nucleotide pairs. According to this extrapolation, the cross-section radii of gyration of such a complex would amount to (R-c)b equals 0.805 plus or minus 0.015 nm and (R-c2)b equals 0.76 plus or minus 0.015 nm for the complex with actinomycin, and to (R-c)b equals 0.77 plus or minus 0.015 nm and (R-c2)b equals 0.75 plus or minus 0.01 nm for the actinomine complex. On the basis of a core and shell model for solvated DNA, these results can be understood as to indicate a decrease of the radial dimensions of both the core and the shell when the dye is bound. The experimental results are compared with theoretical data calculated from the atomic coordinates of the detailed intercalation model for the actinomycin - DNA complex as recently proposed by Sobell and Jain. The model proves to be consistent fairly well with our data on the length increase of the double helix, but it appears to be unable to explain the experimentally observed decrease of R-c2 on binding of dye.