萜烯是以异戊二烯单体组成的烯烃，在生物体中是由异戊二烯焦磷酸通过一系列生物合成反应生成的。由于其单体异戊二烯本身的化学性质，以及萜烯合酶进行催化的活性口袋的灵活性，催化这自然界中最复杂的催化反应。底物在进入萜烯合酶后，往往要经历离子化、质子化、环化、碳正离子重排等一系列催化过程，从而产生结构功能不尽相同的各类萜烯产物，因此其在天然产物中数量极多，至今已经发现超过60000种。 本文研究的α-蒎烯合酶与柠檬烯合酶是单萜烯合酶，参与以GPP为底物的催化反应。为了探究两者的催化机理，为将来改造萜烯合酶打下理论基础，我们从突变分析和理论计算两方面入手对这两种酶进行研究。 同源建模得到的α-蒎烯合酶三维结构是我们的研究基础。我们向该结构分子对接底物中间体，以对接后的底物中间体为中心，突变筛选活性口袋中心周围的氨基酸位点。我们发现F482位点在突变成其他非芳香族的中性氨基酸后，α-蒎烯合酶的主要产物从α-蒎烯变为了桧烯，这说明该位点的苯丙氨酸对于碳正离子重排至关重要。而在F482位点突变成其他芳香族氨基酸后，α-蒎烯合酶几乎失去了绝大部分催化活性，由此我们证明482位点的苯丙氨酸对于α-蒎烯合酶专一性地产生α-蒎烯至关重要。 另外，我们还发现I335位点在突变成空间位阻更小的氨基酸后，也能改变产物分布产生桧烯，这意味α-蒎烯合酶在335位点的空间位阻能改变产物。同时我们还发现当I335位点突变为极性氨基酸后催化活性有明显下降，这说明该位点的极性能影响α-蒎烯合酶的催化活性。 除此以外，通过对α-蒎烯合酶的动力学模拟，我们证明了单纯计算活性口袋的体积并不能预测催化活性或酶的相关催化性质，这说明空间位阻可能在某几个位点起到了更重要的作用。而对氨基酸位点范德华能和库伦能的计算结果与野生型突变体催化结果的数据并不能很好拟合，说明对于预测像α-蒎烯合酶催化这样复杂的过程，还需要更精确的计算。 虽然分子动力学模拟在预测萜烯合酶催化方面还有待优化，但是在解释突变结果方面还是起到了一定的作用。在对4S-柠檬烯Y573位点的突变结果进行解释的过程中，分子动力学模拟发现了Y573和D496之间的氢键作用，以及Y573有向螺旋型LPP（cisoid）靠拢的趋势，这有可能说明Y573和D496共同对LPP的异构化起到了贡献。当然以NPP为底物的催化实验结果显示，Y573起码并不是唯一参与LPP异构化的氨基酸。这虽然没有从根本上解决Y573催化作用的问题，但也为探究其催化机理起到了一定的指导和佐证作用。 本文以定点突变和理论计算证明了α-蒎烯合酶中F482和I335对于产物分布和催化活性的作用。也通过分子动力学模拟与实验结果的拟合过程中，从侧面揭示了单萜烯合酶催化的复杂性。另外，分子动力学模拟还从一定程度上解释了4S-柠檬烯合酶中Y573位点起到的催化作用。我们的研究为改造萜烯合酶提供了理论上的指导。

Terpene is a kind of olefin composed of isoprene monomers which are produced by isoprene pyrophosphate in a series of biosynthesis reactions in organisms. Due to the chemical property of its isoprene monomers and the flexibility of the active pockets of terpene synthase, terpene catalyzes the most complex catalytic reaction in nature. A series of catalytic processes happen, such as ionization, protonation, cyclization, and carbocation rearrangement after substrates dock into the active pocket, producing various terpene products with different structures and functions. Therefore, they are extremely abundant in nature, and more than 60,000 species have been found so far. The α-pinene synthase and limonene synthase are researched in this paper, which are monoterpene synthase and participate in the catalytic reaction using GPP as the substrate. In order to explore the catalytic mechanism of the them, and lay a theoretical foundation for the future modification of terpene synthase, we research them by mutation analysis and theoretical calculation. The three-dimension structure of α-pinene synthase obtained by homology modeling is the basis of our research. We dock the substrate intermediates into the homology model, and mutated the amino acid site around the center of the active pocket centered on the docked substrate intermediate. We found that after mutating F482 to other non-aromatic neutral amino acids, and the main product of α-pinene synthase changed from α-pinene to sabinene, indicating that the function of phenylalanine at this site is essential and is related to carbocation rearrangement. After the mutation at F482 to other aromatic amino acids, α-pinene synthase lost most of its catalytic activity, which demonstrated that 482-site phenylalanine is significately related to α-pinene synthase producing alpha-pinene specificly. In addition, we also found that the I335 site can also change the products distribution and produce sabinene after mutating isoleucine to a less sterically hindered amino acid, which means that the steric hindrance at site-335 of the activie pocket in the α-pinene synthase changes products distribution. At the same time, we also found that when the I335 site was mutated to a polar amino acid, the catalytic activity decreased significantly, which indicated that the polarity of the site affected the catalytic activity of α-pinene synthase. Besides, through the molecular dynamics simulation of α-pinene synthase, we found that simply calculating the volume of the active pocket can’t predict the catalytic activity or other catalytic properties of the enzyme, suggesting that the steric hindrance of several sites playing a more important role. However, the results of the calculation of the amino acid positions van der Waals and coulomb potential are not well fitted with the experimental data of the wild-type and mutant, indicating that it is more accurate need to be applied for predicting complex catalytic processes as α-pinene synthase catalysis. Although molecular dynamics simulation has yet to be optimized in predicting terpene synthase catalysis, they have played a role in explaining the results of mutations to some extent. During the interpretation of the mutation results at the Y573 site in 4S-limonene, molecular dynamics simulations revealed a hydrogen bond between Y573 and D496, and a tendency for Y573 to converge toward helical LPP, which may contribute to the isomerization of LPP by the function of D496 and Y573. Of course, the results of catalytic experiments using NPP as a substrate show that Y573 is not the only amino acid involving in the isomerization step for LPP. Although the simulation resluts do not fundamentally reveal the function of Y573 in catalysis, it also provides some guidances and evidences for exploring terpene catalytic mechanism. The effects of F482 and I335 in products distribution and catalytic activity of α-pinene synthase were demonstrated by site-directed mutagenesis and theoretical calculations. The complexity of monoterpene synthase catalysis is also revealed sidely by fitting results of molecular dynamics simulation and experimental results. In addition, molecular dynamics simulation also explained to some extent the catalytic effect of the Y573 site in 4S-limonene synthase. Our research provides theoretical guidance for the modification of terpene synthase.