Small DNA minicircles are useful for characterizing protein-induced DNA bending and twisting, because obfuscating effects of DNA flexibility are less important than in larger DNA. Our work on DNA geometry and flexibility in protein-DNA complexes has employed T4 ligase-mediated DNA cyclization to make minicircles. Experiments can be carried out as forward ligations, or equivalently protein binding to minicircles can be characterized. In every case we have studied, topological characterization of minicircle synthesis or properties has led to unexpected geometric or mechanistic conclusions. Examples concerning the catabolite activator protein,
RNA polymerase, the Lac repressor, and the TATA-box binding protein are discussed. Topological results have the experimental advantages that they are qualitatively unmistakable and internally controlled: new topoisomers are readily identified even in small amounts, and they are formed in the same reaction as relaxed products. Simulations of topoisomer distributions are quite sensitive to geometrical and flexibility parameters, which helps set stringent constraints on possible structural/dynamic models. However, the disadvantage of a topological measurement is it is consistent with any combination of writhe and twist that sums to the observed Δ
so it is difficult to be confident that a structural/dynamic model is a unique solution.