Tonelli M, James T L
Department of Pharmaceutical Chemistry, University of California, San Francisco 94143-0446, USA.
Biochemistry. 1998 Aug 18;37(33):11478-87. doi: 10.1021/bi980905d.
Sequence-dependent structures of DNA duplexes in solution can be reliably determined using NMR data if care is taken to determine restraint bounds accurately. This entails use of complete relaxation matrix methods to analyze nuclear Overhauser effect (NOE) spectroscopic cross-peak intensities, yielding accurate distance restraints. NMR studies of various DNA duplexes have suggested that there may be some limited internal motions. First, it is typically not possible to reconcile all vicinal proton coupling constants in deoxyribose rings with a single conformer. In addition, with the increased accuracy of interproton distance measurements afforded by the complete relaxation matrix algorithm MARDIGRAS, we find inconsistencies in certain distances which can most readily be ascribed to limited conformational flexibility, since conformational averaging is nonlinear. As base-sugar interproton distances depend on both sugar pucker and glycosidic torsion angle chi, motion involving these structural variables should be reflected by experimental data. Possible motional models have been considered to account for all of the data for three DNA duplexes. Analysis of intraresidue base-sugar interproton NOE bounds patterns suggests a motional model with individual sugars in equilibrium between S (2'-endo) and N (3'-endo) conformations, with S being preferred. As sugar repuckering is correlated with changes in glycosidic torsion angle chi, different sugar conformers imply different values for chi, but this is insufficient to account for all data. A two-state jump between anti and syn glycosidic conformers was considered, but it was incapable of accounting for all data. However, a model with restricted diffusion (rocking) about the glycosidic bond in addition to sugar repuckering was capable of accommodating all experimental data. This motional model is in qualitative agreement with experimental 13C relaxation-derived order parameter values in a DNA duplex.
如果注意准确确定约束边界，溶液中DNA双链体的序列依赖性结构可以通过核磁共振（NMR）数据可靠地确定。这需要使用完整的弛豫矩阵方法来分析核Overhauser效应（NOE）光谱交叉峰强度，从而产生准确的距离约束。对各种DNA双链体的NMR研究表明，可能存在一些有限的内部运动。首先，通常不可能用单一构象来协调脱氧核糖环中所有邻位质子耦合常数。此外，随着完整弛豫矩阵算法MARDIGRAS提供的质子间距离测量精度的提高，我们发现在某些距离上存在不一致，这最容易归因于有限的构象灵活性，因为构象平均是非线性的。由于碱基 - 糖质子间距离取决于糖的折叠和糖苷扭转角χ，涉及这些结构变量的运动应该由实验数据反映出来。已经考虑了可能的运动模型来解释三种DNA双链体的所有数据。对残基内碱基 - 糖质子间NOE边界模式的分析表明，存在一种运动模型，其中单个糖在S（2'-内型）和N（3'-内型）构象之间处于平衡状态，S构象更受青睐。由于糖的重新折叠与糖苷扭转角χ的变化相关，不同的糖构象意味着χ有不同的值，但这不足以解释所有数据。考虑了反式和顺式糖苷构象之间的两态跳跃，但它无法解释所有数据。然而，除了糖的重新折叠外，关于糖苷键的受限扩散（摆动）模型能够适应所有实验数据。这种运动模型与DNA双链体中通过实验13C弛豫得出的序参值在定性上是一致的。
