【作者】 柳慧慧；

【导师】 牛立文； 滕脉坤；

【作者基本信息】 中国科学技术大学 ， 结构生物学， 2014， 博士

【摘要】 CK2是一种保守的蛋白激酶,它广泛分布于真核生物中,通过磷酸化蛋白底物的丝氨酸/苏氨酸/酪氨酸在很多生物学过程中发挥作用。CK2不同于其它蛋白激酶之处在于它具有组成型活性,并且既可以利用ATP又可以利用GTP作为磷酰基供体。CK2是一个多亚基复合物,它包括α-亚基(催化亚基)和β-亚基(调节亚基),前者包含蛋白激酶的催化结构域。我们解析了重组scCK2α与不同核苷酸辅底物(GMPPNP,ATP和AMPPN)与金属离子(Mg2+或Mn2+)的晶体结构。scCK2α的整体结构和同源蛋白相似,包括两个结构域及两个结构域之间的辅底物结合位点。然而,scCK2α的晶体结构呈现出与其同源蛋白不同的三个特征。首先,scCK2α在序列上的独特之处在于它包含了一段38个残基的插入序列,功能目前尚不清楚。根据结构分析并结合之前的研究,我们推测它可能参与保持其活性中心的开放构象,同时它并不参与和两个调节亚基的结合。其次,scCK2α中的两对残基Lys45-Glu53和Arg48-Glu53的相互作用使得Lys5o采取了一种特殊的构象,使其可以直接固定辅底物的γ-磷酸基团,这样就使得“必需二价阳离子”显得不那么必需。我们推断scCK2α可能存在多种“核苷酸-二价阳离子”结合方式,这明显不同于在活性中心有两个二价阳离子结合的zmCK2α或hCK2α。最后,scCK2α-AMPPN结构中G1u53的构象变化破坏了它与Lys45和Arg48的结合,使得辅底物结合口袋变大。这种构象变化暗示了一个关于ADP/GDP释放的路径,因为在CK2a中,催化三联体由"DFG"变成了"DWG",其中的Trp的NE1原子与Leu212的O原子形成氢键,这就使得由原来的"DFG翻转”促成ADP的释放变得不太可能。巧合的是,在辅底物结合部位我们观察到了两个残基Arg161和Lys197捕捉到了两个S042-离子,对这两个残基进行突变,突变体结合和利用ATP/GTP的效率都大大降低,这与我们关于潜在的释放路径的推测相符。虽然CK2各亚基间的亲和力很高,但是目前的研究表明CK2只是一个结合较强的瞬时复合物,各个亚基在细胞内也是独立运动的。hCK2的全酶形式α2β2的晶体结构表明,两个β-亚基形成一个同二聚体,两个α-亚基各自结合在它的两侧形成一个“蝴蝶”形状。我们的实验表明,酿酒酵母CK2的四个亚基相互作用方式似乎与hCK2不同,而且核苷酸与催化亚基scCK2α的结合可能会调节它与β-亚基的结合。目前仅知道CK2β和CK2β’对于组装成全酶必不可少,也不可替代,但具体组装方式还有待进一步的实验验证。CK2不仅能做为一个全酶存在并发挥功能,其亚基还可与其它蛋白质组装成不同的复合物,只不过相比之下CK2亚基间的结合更强。当细胞被紫外线照射受损时,hFACT可以结合到hCK2上,改变后者的构象和底物特异性,使其在细胞内特异地磷酸化p53的Ser392,从而稳定p53的四聚体结构以重启基因转录或复制。在酿酒酵母中,有报道指出CK2可能是通过CHD1间接地与Spt16-Pob3结合,也有报道提到CK2可以和Spt16-Pob3直接结合,但是并没有给出直接的实验证据。FACT在转录和复制过程中都发挥了重要的作用,人源FACT (hFACT)包括Spt16和SSRP1两个亚基,但是在酿酒酵母中FACT (yFACT)却包含三个亚基：Spt16, Pob3,以及一个类HMGl蛋白——Nhp6A/Nhp6B。人源Spt16和酵母Spt16同源,具有相似的长度,Pob3缺少相对于SSRP1的C端部分,后者包含一个HMG1结构域。我们克隆并纯化了scCK2的四个亚基,Spt16、Pob3和Nhp6A/Nhp6B,并证明了它们确实存在相互作用(Pob3和CK2β、CK2β’),但是CK2与FACT的结合方式及意义还需要进一步探究。与此同时我们还得到了Pob3-M和Pob3(1-112aa)的晶体结构,后者可以结合dsDNA,但是它在整个yFACT中发挥的功能还不清楚。更多还原

【Abstract】 CK2is a ubiquitous and conserved protein kinase in eucaryotic organisms, and important in many biological processes. It is unique by keeping constitutive activity and using both ATP and GTP as phosphor donor. CK2is a heterotetramer like a "butterfly" consisting of catalytic and regulatory subunits. In this study, we present the crystal structures of recombinant scCKα with GMPPNP, ATP and AMPPN complexed with either Mg2+or Mn2+as the coordinated divalent cations. The overall structure of scCK2α shows high similarity with its homologous proteins by consisting of two domains and that co-substrate lies in the cleft between them. Whereas, three characteristic features distinct scCK2a from its homolog. Firstly, scCK2a is unique by containing an insertion region that contributes to keep the constitutive activity conformation of scCK2a catalytic site. Secondly, interaction of Lys45-Glu53and Arg48-Glu53in scC2a leads Lys50to adopt a unique conformation able to stabilize the y-phosphate of co-substrate, which makes the existence of the "essential divalent cation" not that essential. The multiple nucleotide-divalent cation binding modes of the active site of scCK2a is apparently different from the two-divalent cation ion occupied the catalytic active site of zmCK2a or hCK2a. Finally, conformation change of Glu53of scCK2a-AMPPN breaks its interaction with Lys45and Arg48, as a result, the co-substrate-binding pocket becomes more open. This could suggest a clue of a possible ADP/GDP release pathway, because the NE1of Trp in "DWG motif" of CK2α forms a hydrogen bond with the O atom of Leu212, which seems making the ADP release process by means of "DFG-in flip to DFG-out" model in most of eukaryotic protein kinase impossible.Previous researches provided that CK2was a strong transient complex and its subunits could move independently though high affinity (4nM of Kd) among its subunits.The crystal structure of hCK2isoform α2β2demonstrated that the two β-subunits interacted with each other to form a dimer, and the two catalytic subunits bind to the opposite side. Our results displayed that interaction between the subunits of scCK2may be different from hCK2isoform α2β2, though more experimental verification is required.The subunits of CK2could combine with other proteins to form complex with weak affinity compared with inner CK2. Interestingly, this interaction is mediated by regulatory subunits.hSpt16and SSRP1could bind to CK2induced by UV irradiation and this direct interaction results in conformation change of CK2, which could then specifically target p53over other substrates. In the case of Saccharomyces cerevisiae, properly the CK2interact indirectly with Spt16-Pob3through CHD1. There are also reports referring to that CK2associated with Spt16-Pob3, but they did not provide experimental evidence.FACT is a histone chaperone with roles in transcription, replication, and repair. It could promote and subsequently restrict access to DNA as a result of dynamic nucleosome reorganization. Human FACT is a stable heterodimer comprised of hSpt16and SSRP1proteins. Whereas in S. cerevisiae, yFACT consists of Spt16/Cdc68, pob3and Nhp6which is a HMGl-like protein.The yeast and human Spt16proteins are structurally related throughout their lengths. However, the Pob3protein does not contain sequences related to the C-terminal portion of SSRP1, which contains a high-mobility group (HMG) box domain. We cloned and purified the subunits of CK2and yFACT of S. cerevisiae. Our result confirmed that CK2could interact with FACT though their function is still unclear. Meanwhile, we obtained the crystal structure of Pob3-M and Pob3(1-112aa), but Pob3-M does not participate in the interaction of Pob3and CK2. EMSA assay demonstrated that Pob3(1-112aa) could bound dsDNA, which need more studies to determine the significance of its DNA-binding function.更多还原