PREDICTION OF REGIOSPECIFIC HYDROXYLATION OF CAMPHOR ANALOGS BY CYTOCHROME-P450(CAM)

PREDICTION OF REGIOSPECIFIC HYDROXYLATION OF CAMPHOR ANALOGS BY CYTOCHROME-P450(CAM)
复制标题

DOI:
10.1021/ja00115a009
复制
发表时间:
1995-03-15
影响因子:
15
通讯作者:
LOEW, G
LOEW, G
中科院分区:
化学1区
文献类型:
--
作者:
HARRIS, D;LOEW, G

文献摘要

被引文献

相似文献

的regiospecificities的羟基化的d-樟脑,d-樟脑,d-硫代樟脑,d-和l-降樟脑已预测采用几何标准从分子动力学模拟酶-底物相互作用的热力学标准,相对自由基能量来自从头算和半经验量子力学。使用Amber 4.0程序套件进行分子动力学模拟,该程序具有修改的Lennard-Jones 6-12参数,并明确包含围绕反应性铁基氧中心的16埃“腹部"区域内的所有原子。的预测是在良好的协议与实验的三个基板的羟基化数据可用于解决对映异构体。具体地,樟脑羟基化被预测为完全区域特异性的(在C-5处100%)。樟脑,不能氢键的酶,被发现是羟基化90%的C-5,9%的C-6,和1%的C-3。硫代樟脑羟基化被预测为区域特异性较低(85%在C-5和15%在C-6)比樟脑从一个方向类似的樟脑在P450(凸轮)结合位点。所有三个类似物的预测区域选择性与实验结果非常吻合。区域特异性获得的D-和L-降樟脑的预测,仍有待验证。它们不能直接与实验结果进行比较,因为羟基化产物仅测定了外消旋底物。通过进行全蛋白质模拟以及这两种底物的约束模拟,测试了樟脑和麝香烷所用的约束蛋白质(“belly”)模型的可靠性。这两种模拟的结果导致底物羟基化的区域特异性的预测相当。这些结果说明了实用程序的细胞色素P450(凸轮)的近似模型,限制不受约束的动态运动的结合位点周围的一个区域中作出准确的预测的区域特异性和立体选择性的P450(凸轮)的羟基化为底物的中等大小。
The regiospecificities of the hydroxylations of d-camphor, d-camphane, d-thiocamphor, and d- and l-norcamphor have been predicted by employing geometric criteria obtained from molecular dynamic simulations of enzyme-substrate interactions in conjunction with the thermodynamic criteria of relative radical energetics derived from nb initio and semiempirical quantum mechanics. Molecular dynamic simulations were performed using the Amber 4.0 suite of programs with modified Lennard-Jones 6-12 parameters and explicit inclusion of all atoms within a 16 Angstrom ''belly'' region surrounding the reactive ferryl oxygen center. The predictions were in good agreement with experiment for the three substrates for which hydroxylation data are available for resolved enantiomers. Specifically, camphor hydroxylation is predicted to be completely regiospecific (100% at C-5). Camphane, which cannot hydrogen bond to the enzyme, is found to be hydroxylated 90% at C-5, 9% at C-6, and 1% at C-3. Thiocamphor hydroxylation is predicted to be less regiospecific (85% at C-5 and 15% at C-6) than camphor starting from an orientation similar to that of camphor in the P450(cam) binding site. The predicted regioselectivities for all three analogs are in excellent agreement with experimental results. Regiospecificities obtained for d- and l-norcamphor are predictions that remain to be verified. They are not directly comparable with experimental results because hydroxylation products have been determined only for the racematic substrate. The reliability of the constrained protein (''belly'') model used was tested for camphor and camphane by performing full protein simulations as well as constrained simulations for these two substrates. The results of both simulations led to comparable predictions of regiospecificity of substrate hydroxylation. These results illustrate the utility of an approximate model of cytochrome P450(cam) that confines unconstrained dynamic motion to a region around the binding site in making accurate predictions of regiospecificity and stereoselectivity of P450(cam) hydroxylation for substrates of moderate size.