Structure of poly(dA).poly(dT) is not identical to the AT rich regions of the single crystal structure of CGCGAATTBrCGCG. The consequence of this to netropsin binding to poly(dA).poly(dT).

The basic assumption of Dickerson and Kopka (J. Biomole. Str. Dyns. 2, 423, 1985) that the conformation of poly(dA).poly(dT) in solution is identical to the AT rich region of the single crystal structure of the Dickerson dodecamer is not supported by any experimental data. In poly(dA).poly(dT), NOE and Raman studies indicate that the dA and dT units are conformationally equivalent and display the (anti-S-type sugar)-conformation; incorporation of this nucleotide geometry into a double helix leads to a conventional regular B-helix in which the width of the minor groove is 8A. The derived structure is consistent with all available experimental data on poly(dA).poly(dT) obtained under solution conditions. In the crystal structure of the dodecamer, the dA and dT units have distinctly different conformations-dA residues adopt (anti, S-type sugar pucker), while dT residues belong to (low anti, N-type sugar pucker). These different conformations of the dA and dT units along with the large propeller twist can be accommodated in a double helix in which the minor groove is shrunk from 8A to less than 4A. In the conventional right handed B-form of poly(dA).poly(dT) with the 8A wide minor groove, netropsin has to bind asymmetrically along the dA strand to account for the NOE and chemical shift data and to generate a stereochemically sound structure (Sarma et al, J. Biomole. Str. Dyns. 2, 1085, 1985).