Characterization of protein-protein interactions critical for poliovirus replication: analysis of 3AB and VPg binding to the RNA-dependent RNA polymerase

Abstract

Two critical interactions within the poliovirus RNA replication complex are those of the RNA-dependent RNA polymerase 3D with the viral proteins 3AB and VPg. 3AB is a membrane-binding protein responsible for the localization of the polymerase to the membranous vesicles at which replication occurs. VPg (a peptide comprising the 3B region of 3AB) is the 22-residue soluble product of 3AB cleavage and serves as the protein primer for RNA replication. The detailed interactions of these proteins with the RNA-dependent RNA polymerase 3D were analyzed to elucidate the precise roles of 3AB and VPg in the viral RNA replication complex. Using a membrane-based pull-down assay, we have identified a binding "hot-spot" spanning residues 100 to 104 in the 3B (VPg) region of 3AB which plays a critical role in mediating the interaction of 3AB with the polymerase. Isothermal titration calorimetry shows that the interaction of VPg with 3D is enthalpically driven, with a dissociation constant of 11 microM. Mutational analyses of VPg indicate that a subset of the residues important for 3AB-3D binding are also important for VPg-3D binding. Two residues in particular, P14 and R17, were shown to be absolutely critical for the binding interaction. This work provides the direct characterization of two binding interactions critical for the replication of this important class of viruses and identifies a conserved polymerase binding sequence responsible for targeting the polymerase.

FIG. 1.

Sequence alignment of 3A (A) and VPg (B) from members of the enterovirus and rhinovirus genera of the Picornaviridae family. Completely conserved (100% identity) residues are highlighted in blue, and highly conserved (>75% identity) residues are highlighted in green. Filled circles indicate residues of the 3A region which were mutated and tested by the polymerase recruitment assay. Open circles indicate residues of VPg which were mutated and tested by ITC analysis. Below, an open box indicates the folded region of 3A, and the closed box indicates the 22-residue hydrophobic region. (C) Cartoon illustrating the proposed topology model for 3AB binding to membranous vesicles in infected cells.

FIG. 2.

Effects of point mutations in the 3A region of membrane-bound 3AB on binding to the 3D polymerase. (A) Ribbon diagram of 3A-N viewed from the “top.” Hydrophobic residues in the dimer interface and other structurally critical residues mutated for this study are labeled, and side chain carbons are displayed. (B) Ribbon diagram of the structure of the soluble, N-terminal region of 3AB (3A-N) viewed from the “side.” Charged surface residues mutated for this study are labeled, and side chain carbons are displayed. (C) Results of the polymerase recruitment assay using membranes containing a 3AB construct containing the indicated mutation. Bars indicate the amount of 3D bound by the membranes corrected for the amount bound to control membranes and normalized to the amount bound to wild-type 3AB. The assays were performed in triplicate, and error bars indicate the standard error of the mean.

FIG. 3.

Effects of point mutations in the 3B region of membrane-bound 3AB on binding to the 3D polymerase. (A) Results of the polymerase recruitment assay using membranes containing wild-type 3A or 3AB with the indicated 3B residue mutated to alanine. Note that the alanine residues at positions 89 and 106 were not tested. The residues making up the PBS are labeled. Bars and error bars are as described in the legend of Fig. 2. (B) Sample of the 3D region of an sodium dodecyl sulfate-polyacrylamide gel electrophoresis gel used for quantifying the polymerase recruitment assay.

FIG. 4.

Uridylylation of E. coli-derived VPg. Electrophoretic gel indicating uridylylation of VPg which was either obtained by peptide synthesis (lanes 1 and 2 and lanes 5 and 6) or expressed as a fusion protein in E. coli (lanes 3 and 4 and lanes 7 and 8). The reactions are performed using either a poly(A) RNA template (lanes 1 to 4) or the 2C^CRE RNA hairpin template (lanes 5 to 8). Uridylylated VPg peptides are labeled. Larger-molecular-weight bands in the synthesized peptide lanes (lanes 1 and 2 and lanes 5 and 6) represent peptide heterogeneity not present in the E. coli-expressed peptides.

FIG. 5.

ITC analysis of wild-type VPg binding to 3D. (Top) Thermogram displaying the change in heat required by the instrument to maintain an isothermal condition between the sample cell and reference cell. Each peak occurs upon the addition of an aliquot of VPg to the 3D solution in the sample cell. (Bottom) Points displaying the integration of the thermogram peaks indicate the heat released per mole of VPg during each injection. The line connecting the points is the best fit to the data of a two-state binding model using a nonlinear least squares algorithm to solve for the stoichiometry (N), the dissociation constant (K_d), and the change in enthalpy (ΔH) for the reaction. The change in entropy, ΔS, is calculated from these values.

FIG. 6.

ITC analysis of VPg mutants. Representative ITC data illustrate the effects on 3D-binding of the three classes of VPg mutations. (A) VPg mutant R17A does not bind to 3D, displaying little or no heat released upon addition to 3D. (B) VPg mutant T15A binds to 3D with a somewhat reduced affinity, displaying smaller peaks and a shallower curve than wild-type VPg but still allowing for accurate determination of thermodynamic binding parameters. (C) VPg mutant Y3A binds to 3D much like wild-type VPg, displaying similar thermograms and allowing for accurate determination of thermodynamic binding parameters.

FIG. 7.

Comparison of 3AB and VPg binding to the 3D polymerase. The sequences of 3AB and VPg are shown with their corresponding residue numbers, and the polymerase binding sequence of 3AB is indicated. A lowercase h indicates the homoserine residue remaining at the C terminus of VPg due to CNBr cleavage. Residues not tested for their effects on 3D-binding are shown in black, residues tested but not displaying a significant effect on binding are shown in blue, residues tested and displaying a moderate effect on 3D-binding are shown in orange, and residues tested and displaying a large effect on 3D-binding are shown in red and underlined.