Iwona Minor

The structure of coxsackievirus B3 at 3.5 A resolution.

BACKGROUND Group B coxsackieviruses (CVBs) are etiologic agents of a number of human diseases that range in severity from asymptomatic to lethal infections. They are small, single-stranded RNA icosahedral viruses that belong to the enterovirus genus of the picornavirus family. Structural studies were initiated in light of the information available on the cellular receptors for this virus and to assist in the design of antiviral capsid-binding compounds for the CVBs. RESULTS The structure of coxsackievirus B3 (CVB3) has been solved to a resolution of 3.5 A. The beta-sandwich structure of the viral capsid proteins VP1, VP2 and VP3 is conserved between CVB3 and other picornaviruses. Structural differences between CVB3 and other enteroviruses and rhinoviruses are located primarily on the viral surface. The hydrophobic pocket of the VP1 beta-sandwich is occupied by a pocket factor, modeled as a C16 fatty acid. An additional study has shown that the pocket factor can be displaced by an antiviral compound. Myristate was observed covalently linked to the N terminus of VP4. Density consistent with the presence of ions was observed on the icosahedral threefold and fivefold axes. CONCLUSIONS The canyon and twofold depression, major surface depressions, are predicted to be the primary and secondary receptor-binding sites on CVB3, respectively. Neutralizing immunogenic sites are predicted to lie on the extreme surfaces of the capsid at sites that lack amino acid sequence conservation among the CVBs. The ions located on the icosahedral threefold and fivefold axes together with the pocket factor may contribute to the pH stability of the coxsackieviruses.

Structure determination of coxsackievirus B3 to 3.5 A resolution.

The crystal structure of coxsackievirus B3 (CVB3) has been determined to 3.5 A resolution. The icosahedral CVB3 particles crystallize in the monoclinic space group, P2(1), (a = 574.6, b = 302.1, c = 521.6 A, beta = 107.7 degrees ) with two virions in the asymmetric unit giving 120-fold non-crystallographic redundancy. The crystals diffracted to 2.7 A resolution and the X-ray data set was 55% complete to 3.0,4, resolution. Systematically weak reflections and the self-rotation function established pseudo R32 symmetry with each particle sitting on a 32 special position. This constrained the orientation and position of each particle in the monoclinic cell to near face-centered positions and allowed for a total of six possible monoclinic space-group settings. Correct interpretation of the high-resolution (3.0-3.2 A) self-rotation function was instrumental in determining the deviations from R32 orientations of the virus particles in the unit cell. Accurate particle orientations permitted the correct assignment of the crystal space-group setting amongst the six ambiguous possibilities and for the correct determination of particle positions. Real-space electron-density averaging and phase refinement, using human rhinovius 14 (HRV14) as an initial phasing model, have been carried out to 3.5 A resolution. The initial structural model has been built and refined to 3.5 A resolution using X-PLOR.