Xie Q, Chapman M S
Department of Chemistry, Florida State University, Tallahassee 32306-3015, USA.
J Mol Biol. 1996 Dec 6;264(3):497-520. doi: 10.1006/jmbi.1996.0657.
The DNA-containing capsid of canine parvovirus (CPV) is analyzed following atomic refinement at 2.9 A resolution. The capsid contains 60 copies of the capsid protein related by icosahedral symmetry. The atomic model has been extended from the first residue (Gly37) of the unrefined 3.25 A structure towards the N terminus. The electron density shows that approximately 87% of the capsid proteins have N termini on the inside of the capsid, but for approximately 13%, the polypeptide starts on the outside and runs through one of the pores surrounding each 5-fold axis, explaining apparently conflicting antigenic data. Analysis of potential hydrogen bonds reveals approximately 50% more secondary structure than previously apparent. Most of the additional secondary structure are in the 71 and 221 residue-long loop insertions between beta-strands E and F and G and H, forming subunit-bridging sheets that likely add specificity to assembly interactions. Structural analysis of the extensive subunit interactions around the 3-fold axes shows that assembly is a multistep process with loops intertwining following initial contact. Estimated free energies of association suggest that the formation of 3 and 5-fold contacts likely takes precedence over 2-fold interactions. Energies for initial association into trimers or pentamers would be similar, but the intertwining of loops about the 3-fold axis adds an additional large activation barrier to dissociation. Analysis of the surfaces of the assembled capsid shows a surprising lack of basic amino acids that might have been expected to interact with the negatively charged phosphoribose backbone of the DNA. Instead, uncharged polar and van der Waal's interactions predominate in the packaging of single-stranded DNA into the capsid.
犬细小病毒（CPV）含DNA的衣壳在2.9埃分辨率下经原子精修后进行了分析。衣壳包含60个通过二十面体对称相关的衣壳蛋白拷贝。原子模型已从未精修的3.25埃结构的第一个残基（Gly37）向N端延伸。电子密度显示，约87%的衣壳蛋白N端位于衣壳内部，但约13%的多肽从外部开始，穿过围绕每个5重轴的孔之一，这解释了明显相互矛盾的抗原数据。对潜在氢键的分析揭示，二级结构比以前明显多出约50%。大部分额外的二级结构位于β链E和F以及G和H之间71和221个残基长的环插入中，形成亚基桥接片层，这可能会增加组装相互作用的特异性。对3重轴周围广泛亚基相互作用的结构分析表明，组装是一个多步骤过程，初始接触后环相互缠绕。估计的结合自由能表明，3重和5重接触的形成可能优先于2重相互作用。三聚体或五聚体初始结合的能量相似，但围绕3重轴的环的缠绕为解离增加了额外的大激活屏障。对组装后衣壳表面的分析显示，令人惊讶的是缺乏可能预期与DNA带负电荷的磷酸核糖骨架相互作用的碱性氨基酸。相反，在单链DNA包装到衣壳中的过程中，不带电荷的极性和范德华相互作用占主导。
