【作者】 徐锋；

【导师】 余龙； 寿天德；

【作者基本信息】 复旦大学 ， 遗传学， 2008， 博士

【摘要】 蛋白质磷酸化修饰是细胞内最广泛、最重要的信息调控方式。这个过程能够影响细胞的生长、分化、代谢、分裂、迁移,乃至生物体的肌肉收缩、免疫、学习和记忆。蛋白激酶是最大的、也是最重要的蛋白质家族之一。在真核细胞内,大约有2%的基因编码激酶蛋白,有约30%的蛋白至少有一个位点被磷酸化。蛋白质激酶能够转移ATP的γ磷酸根到底物蛋白的丝氨酸、苏氨酸和酪氨酸等。在真核生物,几乎所有的激酶都具有这样的特征:(1)具有结构保守的催化域双叶结构;(2) ATP和底物多肽结合在两叶之间的腔沟中;(3)具有活性的激酶其催化域呈现为“闭合的”构象。已有的研究表明激酶的底物结合与磷酸根转移之间存在耦合关系,而蛋白质底物结合又分为两个结合位点,一个位于活性位点区,即通常的共有序列(consensus sequence)多肽所在的位置;另一个在活性位点之外的区域,通常位于催化域的羧基末端。那么,蛋白质底物结合区与活性位点之间是如何耦合的昵?这个问题至今没有解决。根据功能上或结构上互相耦合的氨基酸位点,会引起它们共进化的这个朴素思想,我们采用了统计耦合分析、互信息分析和残基相关分析等方法,对丝氨酸/苏氨酸激酶催化域家族的序列进行了分析,随后我们又对酪氨酸激酶催化域家族进行了类似的分析,并对两个家族做了结构比对比较,最后,我们使用分子动力学模拟方法,对环腺苷酸依赖蛋白激酶催化亚基(PKAc)与其多肽底物之间相互作用进行了分子动力学模拟和氨基酸网络分析。通过上述研究,我们得到了以下结论:(1)我们在丝氨酸/苏氨酸激酶催化域家族发现了两个不同的共进化氨基酸网络:θ网络和γ网络。在三维结构上,θ网络连接激酶催化域ATP结合区与底物结合区,根据已有的实验证据和这些位点的分布,我们认为θ网络可能通过构象改变的方式,介导底物结合与活性位点区之间的耦合。另外,θ网络可能也参与决定了激酶的底物特异性。γ网络主要分布在底物结合区的一侧,连接激活环与催化域羧基端,根据文献资料,我们认为γ网络是在激酶催化作用发生之前,起支撑底物结合区与激活环的作用,在催化作用完成之后,γ网络可能介导底物的释放。对PKA催化亚基与其多肽底物相互作用的分子动力学模拟研究及氨基酸网络分析给我们的推测提供了一定的理论支持。这两个共进化网络对于整个激酶家族具有一定的普适性。(2)对丝氨酸/苏氨酸激酶和酪氨酸激酶进行催化域结构比对和序列分析,我们定量地、系统地说明了这两个家族在催化域序列上的异同,我们发现这两个家族在催化域序列的差异主要在催化环、P+1环以及位点158、238和273上,这些区域主要与底物结合和催化有密切的关系。这些结果有助于我们对这两个家族底物特异性差异的深入理解。(3)在方法应用上,我们在Dijkstra最短路径算法的基础上,设计完成了氨基酸网络分析程序。在统计耦合分析方法应用上,我们也有一些创新的结论和应用。这两个共进化网络的发现有助于理解蛋白激酶的催化机制和这个蛋白家族的进化与起源。由于激酶的功能异常会引起许多疾病,我们的研究也有助于基于激酶靶标的药物设计等。更多还原

【Abstract】 Protein phosphorylation is the most widespread and important type of post-translational modification used in cellular regulation. It effects every basic cellular process, including metabolism, growth, differentiation, division, motility, and muscle contraction, immunity, learning and memory. It has been estimated that 30% of all cellular proteins are phosphorylated on at least one residue in a typical eukaryotic cell. Therefore, protein kinases (PKs) are one of the largest protein families, comprising～2% of all eukaryotic genes. Protein kinases catalyse the transfer of theγ-phosphate from ATP to specific amino acids in proteins; in eukaryotes, these are usually Ser, Thr and Tyr residues.In eukaryotes, almost all PKs have a structurally well conserved two-lobe structure of the catalytic domains. The peptide substrate is held in the groove between the two lobes. When PKs are in the active forms, the catalytic domains of these PKs have a similar ’closed’ conformation. In fact, protein substrate binding can be divided into two components: binding interactions at the active site and binding interactions at a site distal to the active site, usually at the C-terminal of catalytic subunit. Previous studies suggested that there are communication pathways between the active and distal binding sites in PKs. This coupling pathway remains to be determined.How can we identify this coupling pathway in PKs? We used sequence-based statistical methods for estimating covariant residues in the multiple sequence alignments (MSA) of catalytic subunit families of Serine/Threonine and Tyrosine PKs, respectively. These methods include the Statistical Coupling, Residue Correlated, and Mutual Information analyses. The basis of these statistical methods is that the coupling of two sites in a protein, whether for structural or functional reasons, should cause those two positions to co-evolve. And then we performed the structural alignment for these two familes. At last, we made molecular dynamic simulations and residue network analysis for catalytic domain of cAMP-dependent protein kinase (PKAc) with and without its peptide substrate. Based on these studies, we got the following conclusions on catalytic subunit of PK:1) We identified two distinct co-evolving networks (i.e. the 9-shaped and y-shaped networks) in the catalytic subunits family of Serine/Threonine PKs (Ser/Thr PKc family) by using three statistical analysis methods. Theθ-shaped network links the protein substrate binding and active sites, which might participate in the coupling between substrate binding and catalysis.θ-shaped network also participates in determinants of the substrate specificity of PKs. Theγ-shaped network is mainly located the one side of substrate binding region, linking the activation loop and protein substrate binding region. It might play important role in supporting substrate binding region and the activation loop before catalysis, and mediating product releasing after catalysis. Our studies of molecular dynamics simulations and residue network analysis for interactions between PKAc and its peptide substrate provide some support for our speculation on the function of these two co-evolving networks2) We showed both differences and similaries between the sequences of Ser/Thr PKc and TyrKc families by using structural alignment and sequence analysis. These results are helpful to extensively understand the difference of substrate specificity for these two families.3) We designed, implemented and tested the new programs for residue network analysis based on the Dijkstra’s algorithm of the shortest pathway. In addition, we made some new conclusions on the application of Statistical Coupling Analysis method.更多还原