QM/MM studies show substantial tunneling for the hydrogen-transfer reaction in methylamine dehydrogenase

QM/MM studies show substantial tunneling for the hydrogen-transfer reaction in methylamine dehydrogenase
复制标题

DOI:
10.1021/ja016219a
复制
发表时间:
2001-09-05
影响因子:
15
通讯作者:
Burton, NA
Burton, NA
中科院分区:
化学1区
文献类型:
--
作者:
Faulder, PF;Tresadern, G;Burton, NA

文献摘要

被引文献

相似文献

甲胺脱氢酶(MADH)99.3)催化甲胺氧化去甲基化生成甲醛和氨。1-2最近对该酶的停流动力学研究2显示了异常高的、与温度无关的初级氘动力学同位素效应(KIE),为16.8(0.5),但强烈依赖于温度的反应速率被解释为热诱导振动驱动的极限隧道机制,用于限速的氢提取步骤(图1)。在这里,我们提出了混合量子力学和分子力学(QM/MM)对该反应步骤的计算研究,用典型变分过渡态理论结合多维隧道效应来预测反应速率。我们发现了一个不寻常的反应自由能谱,表明量子隧穿(> 90%)和经典蛋白质运动的结合是造成大KIE的原因。目前,人们对蛋白质柔韧性增强氢隧穿的可能性非常感兴趣,特别是在醇脱氢酶家族的情况下。3-6也许是这个系列中研究最多的酶,实验7,8和理论9,10都是肝脏酒精脱氢酶(LADH),它已经被证明蛋白质的灵活性与嗜热和嗜中温酶的氢化物隧道相关,尽管测量的KIEs相当小(2-3),与这里研究的MADH中的质子位移值相反。混合QM/MM方法11,12非常适合于酶反应路径的研究,因为它们包括对活性位点的精确量子力学描述,其中电子结构很重要,键被破坏和形成,同时用分子力学势近似酶框架的周围效应。这种混合QM/MM方法最近被与半经典变分过渡态理论方法相结合[9,12]来研究氢化物移位反应及其相关的key。从Methylophilus methylotrophus13中提取的MADH晶体坐标,用甲胺取代色氨酸色氨酸醌(TTQ)辅因子的羰基氧,构建了初始亚氨基醌反应物结构,如图1所示(TTQ的两个色氨酸残基分别标记为TRP1和TRP2)。然后整个二聚体被质子化,溶剂化,然后用AMBER分子力学力场最小化。14对于混合计算,QM区域被限制为31个原子(见图1),其中包括催化碱Asp428和TTQ辅助因子中一个色氨酸(TRP2)的催化活性环,这些环经过甲胺还原后。随后,使用我们的QM/MM程序最小化该QM区域以找到优化的反应物复合物,该程序利用Gaussian9415和AMBER16代码并保持连接原子和MM区域固定。17在这项研究中,我们使用了PM3半经验哈密顿量,它已被证明是可靠的,可以研究其他酶的机制,11,12,并使沿着反应路径评估直接动力学计算所需的大量二阶导数。
Methylamine dehydrogenase (MADH)(EC 1.4. 99.3) catalyses the oxidative demethylation of methylamine to formaldehyde and ammonia. 1-2 Recent stopped-flow kinetic studies2 of this enzyme have shown an unusually high, temperature-independent primary deuterium kinetic isotope effect (KIE) of 16.8 (0.5, but strongly temperature dependent reaction rates that were interpreted as indicating a thermally induced vibrationally driven extreme tunneling mechanism for the rate-limiting hydrogen abstraction step (Figure 1). Here we present hybrid quantum mechanical and molecular mechanical (QM/MM) computational studies of this reaction step, with canonical variational transition-state theory incorporating multidimensional tunneling effects to predict reaction rates. We find an unusual reaction free energy profile which indicates that a combination of quantum tunneling (> 90%) and classical protein motion is responsible for the large KIE. There is currently great interest in the possibility of enhanced hydrogen tunneling arising from protein flexibility, particularly in the case of the alcohol dehydrogenase family of enzymes. 3-6 Perhaps the most studied enzyme in this series, both experimentally7, 8 and theoretically, 9, 10 is liver alcohol dehydrogenase (LADH) where it has been shown that protein flexibility correlates with hydride tunneling for thermophilic and mesophilic enzymes, although the measured KIEs are quite small (2-3), in contrast to the value for the proton shift in MADH studied here. Hybrid QM/MM methods11, 12 are well suited to the study of enzyme reaction paths since they include an accurate quantum mechanical description of the active site, where the electronic structure is important and bonds are broken and formed, while approximating the surrounding effects of the enzyme framework with a molecular mechanical potential. Such hybrid QM/MM methods have recently been combined with semiclassical variational transition-state theory methods9, 12 to study hydride shift reactions and their associated KIEs.An initial iminoquinone reactant structure was constructed from the crystal coordinates of MADH taken from Methylophilus methylotrophus13 by replacing the carbonyl oxygen of the tryptophan tryptophylquinone (TTQ) cofactor with methylamine as shown in Figure 1 (the two tryptophan residues of TTQ are labeled separately as TRP1 and TRP2). The entire dimer was then protonated, solvated, and then minimized using the AMBER molecular mechanics force field. 14 For the hybrid calculations, the QM region was limited to 31 atoms (see Figure 1) which includes the catalytic base, Asp428, and the catalytically active rings of one tryptophan (TRP2) of the TTQ cofactor after reduction with methylamine. This QM region was subsequently minimized to find an optimized reactant complex, using our QM/MM program which utilizes the Gaussian9415 and AMBER16 codes and keeping the link atoms and MM regions fixed. 17 In this study we have used the PM3 semiempirical Hamiltonian which has proved reliable to study other enzyme mechanisms, 11, 12 and enables the large number of second derivatives required for the direct dynamics calculation to be evaluated along the reaction path.