Aguilar C F, Thomas P J, Moss D S, Mills A, Palmer R A
Department of Crystallography, Birkbeck College, University of London, U.K.
Biochim Biophys Acta. 1991 Dec 11;1118(1):6-20. doi: 10.1016/0167-4838(91)90435-3.
The X-ray structures of two complexes of bovine ribonuclease-A produced by soaking pre-grown crystals in solutions of the inhibitors cytidylyl-2',5'-guanosine (2',5' CpG) and deoxycytidylyl-3',5'-guanosine (3',5'dCpdG) have been determined at 1.5 A resolution and refined by restrained least squares to R = 21.0% for 17,855 reflections, and R = 19.1% for 16,347 reflections, respectively. Binding of the substrate analogs to the protein has taken place in a completely unexpected and previously unreported manner. In each case the guanine base occupies the well characterized B1 pyrimidine binding site adjacent to Thr-45 (described by Richards, F.M., Wyckoff, H.W., Carlson, W.D., Allewell, N.M., Lee, B. and Mitsui, Y. (1971) Cold Spring Harbor Symp. Quant. Biol. 36, 35-54, and others including Palmer, R.A., Moss, D.S., Haneef, I. and Borkakoti, N. (1984) Biochim. Biophys. Acta 785, 81-88) having entered through a secondary channel external to the active site itself. We designate this reversed non-productive mode as retro-binding. In this mode of binding the SO4(2-) anion bound in the active site of the native protein crystals (Borkakoti, N., Moss, D.S. and Palmer, R.A. (1982) Acta Crystallogr. B38 2210-2217) has not been displaced by the phosphate of the inhibitor molecule as originally anticipated and observed in other studies. Instead the CMP or dCMP moiety of the inhibitor molecule is held loosely in a channel running towards the surface of the protein molecule and is thus completely external to the active site. Consequently, although it has been possible to model them, no attempt has been made to refine either the disordered cytosine in the CpG complex or the deoxycytosine in the dCpdG complex. The traditional B2 purine binding site of RNase (Richards et al., 1971) is unoccupied by the soaked inhibitors. Important changes that have taken place in the protein structure include: stabilization of both Lys-41 and Gln-11 via H-bonding to SO4(2-); stabilization of His-119 in the A conformation (Borkakoti, N., Moss, D.S. and Palmer, R.A. (1982) Acta Crystallogr. B38 2210-2217); and stabilization of SO4(2-) by H-bonds formed with the retro-bound guanine base. Binding of the inhibitors and stabilization of the active site is accompanied by displacement and redistribution of solvent molecules.
通过将预先生长的晶体浸泡在抑制剂胞苷酰-2',5'-鸟苷(2',5' CpG)和脱氧胞苷酰-3',5'-鸟苷(3',5'dCpdG)溶液中,得到了牛核糖核酸酶-A两种复合物的X射线结构,分辨率为1.5 Å,并通过约束最小二乘法进行了精修,对于17855个反射,R = 21.0%,对于16347个反射,R = 19.1%。底物类似物与蛋白质的结合以一种完全出乎意料且以前未报道的方式发生。在每种情况下,鸟嘌呤碱基占据了与苏氨酸-45相邻的特征明确的B1嘧啶结合位点(由理查兹、F.M.、怀科夫、H.W.、卡尔森、W.D.、阿勒韦尔、N.M.、李、B.和三井、Y.(1971年)《冷泉港定量生物学研讨会》36卷,35 - 54页,以及包括帕尔默、R.A.、莫斯、D.S.、哈尼夫、I.和博尔卡科蒂、N.(1984年)《生物化学与生物物理学报》785卷,81 - 88页等描述),它是通过活性位点本身外部的一条二级通道进入的。我们将这种反向非生产性模式称为反向结合。在这种结合模式中,天然蛋白质晶体活性位点中结合的SO4(2-)阴离子(博尔卡科蒂、N.、莫斯、D.S.和帕尔默、R.A.(1982年)《晶体学报》B38卷,2210 - 2217页)并没有像其他研究中最初预期和观察到的那样被抑制剂分子的磷酸基团取代。相反,抑制剂分子的CMP或dCMP部分松散地位于一条通向蛋白质分子表面的通道中,因此完全处于活性位点外部。因此,尽管有可能对它们进行建模,但并未尝试对CpG复合物中无序的胞嘧啶或dCpdG复合物中的脱氧胞嘧啶进行精修。核糖核酸酶传统的B2嘌呤结合位点(理查兹等人,1971年)未被浸泡的抑制剂占据。蛋白质结构中发生了重要变化,包括:通过与SO4(2-)形成氢键使赖氨酸-41和谷氨酰胺-11稳定;使组氨酸-119处于A构象稳定(博尔卡科蒂、N.、莫斯、D.S.和帕尔默、R.A.(1982年)《晶体学报》B38卷,2210 - 2217页);以及通过与反向结合的鸟嘌呤碱基形成氢键使SO4(2-)稳定。抑制剂的结合和活性位点的稳定伴随着溶剂分子位置的改变和重新分布。