Schmitz U, Zon G, James T L
Department of Pharmaceutical Chemistry, University of California, San Francisco 94143-0446.
Biochemistry. 1990 Mar 6;29(9):2357-68. doi: 10.1021/bi00461a021.
Exchangeable and nonexchangeable proton and phosphorus resonances (11.75 T) of [d(GTATATAC)]2 in aqueous solution were assigned by using proton two-dimensional nuclear Overhauser effect (2D NOE) spectra, homonuclear proton double-quantum-filtered COSY (2QF-COSY) spectra, proton spin-lattice relaxation time measurements, and 31P1H heteronuclear shift correlation spectra. Due to the large line widths, it was not possible to directly extract vicinal proton coupling constant values from any spectrum including ECOSY or 2QF-COSY. However, comparison of quantitative 2QF-COSY spectral simulations with experimental spectra enabled elucidation of coupling constants. The scope and limitations of this approach were explored by computation and by use of experimental data. It was found that proton line widths exhibit some variability from one residue to the next as well as from one proton to the next within a residue and the exact line width is critical to accurate evaluation of coupling constants. Experimental 2QF-COSY spectra were not consistent with a rigid deoxyribose conformation for any of the nucleotide residues. A classical two-state model, with rapid jumps between C2'-endo (pseudorotation angle P = 162 degrees) and C3'-endo (P = 9 degrees) conformations, was able to account for the spectral characteristics of terminal residue sugars: 60% C2'-endo and 40% C3'-endo. However, the 2QF-COSY cross-peaks from the -TATATA- core could be simulated only if the classical two-state model was altered such that the dominant conformer had a pseudorotation angle at 144 degrees instead of 162 degrees. In this case, the major conformer amounted to 80-85%. Alternatively, the spectral data were consistent with a three-state model in which C2'-endo and C3'-endo conformations had the largest and smallest populations, respectively, but a third conformer corresponding to C1'-exo (P = 126 degrees) was present, consistent with recent molecular dynamics calculations. This alternative yielded populations of 50% (P = 162 degrees), 35% (P = 126 degrees), and 15% (P = 9 degrees) for the -TATATA- sugars. The spectral results indicate little variation of sugar pucker between T and A. Small differences in cross-peak component intensities and characteristic spectral distortions, however, do suggest some unquantified variation. 31P1H heteronuclear chemical shift correlation spectra manifested alternating chemical shifts and coupling constants suggestive of phosphodiester backbone conformational differences between TA and AT junctions.
通过使用质子二维核Overhauser效应(2D NOE)谱、同核质子双量子滤波COSY(2QF - COSY)谱、质子自旋 - 晶格弛豫时间测量以及31P1H异核位移相关谱,对水溶液中[d(GTATATAC)]2的可交换和不可交换质子及磷共振(11.75 T)进行了归属。由于线宽较大,无法从包括ECOSY或2QF - COSY在内的任何谱图中直接提取邻位质子耦合常数的值。然而,通过将定量2QF - COSY谱模拟与实验谱进行比较,得以阐明耦合常数。通过计算和使用实验数据探索了该方法的范围和局限性。发现质子线宽在不同残基之间以及残基内的不同质子之间存在一定变化,确切的线宽对于准确评估耦合常数至关重要。对于任何核苷酸残基,实验性的2QF - COSY谱都与刚性脱氧核糖构象不一致。一个经典的双态模型,在C2'-内端(假旋转角P = 162度)和C3'-内端(P = 9度)构象之间快速跳跃,能够解释末端残基糖的光谱特征:60%为C2'-内端和40%为C3'-内端。然而,只有当经典双态模型改变,使得主要构象体的假旋转角为144度而非162度时,才能模拟出来自 - TATATA - 核心的2QF - COSY交叉峰。在这种情况下,主要构象体占80 - 85%。或者,光谱数据与一个三态模型一致,其中C2'-内端和C3'-内端构象的丰度分别最大和最小,但存在对应于C1'-外端(P = 126度)的第三个构象体,这与最近的分子动力学计算结果一致。对于 - TATATA - 糖,这种替代方案得到的丰度为50%(P = 162度)、35%(P = 126度)和15%(P = 9度)。光谱结果表明T和A之间糖的皱折变化很小。然而,交叉峰成分强度的微小差异和特征性光谱畸变确实表明存在一些未量化的变化。31P1H异核化学位移相关谱显示出交替的化学位移和耦合常数,暗示了TA和AT连接处磷酸二酯主链构象的差异。