Co-folding and RNA activation of poliovirus 3Cpro polyprotein precursors

Figures

• 

  Figure 1. Poliovirus P3 polyprotein region and structure.A, overview of the P3 region with lengths of the fully processed 3A, 3B, 3C, and 3D proteins indicated. Sequence details show the complete 3A with its consensus secondary structure prediction from the Phyre2 server (38). The two short soluble domain helices are highlighted in green and the predicted membrane binding region is in red. B, structures of different segments of the P3 proteins with PDB codes indicated. NMR studies of the 3A soluble region show residues 15 to 45 folding into a pair of short helices that drive homodimerization via the formation of a compact four-helix bundle. The crystal structure of 3CD is composed of independent 3C and 3D domains that are essentially identical to those of the individual structures of 3C and 3D.

• 

  Figure 2. Oligomerization states of P3 proteins.A, size exclusion chromatography of P3 proteins in the absence and presence of DDM detergent. In the absence of detergent proteins containing 3Amem elute entirely at the void volume as large soluble aggregates, but these are disrupted by the addition of DDM to yield distinct micelle-associated species. The non-membrane-associated proteins 3AsolBC and 3C elute according to their monomeric molecular weights in both the presence and absence of DDM. B, single-molecule TIRF microscopy visualization of protein aggregation states via two-color labeling where dimer or large complexes appear yellow while monomers are either red or green. In the absence of detergent, 3ABC forms large aggregates, 3AmemBC forms smaller multimers, and 3AsolBC is mostly monomeric. Adding 0.5 mM DDM detergent solubilizes the larger aggregates into mostly dimers and monomers.

• 

  Figure 3. Circular dichroism spectra of P3 region proteins.A, spectra of 3A, 3AB, and 3ABC, obtained from fusion proteins with an N-terminal ubiquitin whose CD signal (solid grey line) has already been subtracted. B, spectra for 3D, 3CD, 3BCD, 3BC, and 3C proteins that do not require ubiquitin for solubility. C, comparison of the spectrum from native 3C with difference spectra reflecting 3C in the context of 3CD and 3ABC polyproteins. A typical β-sheet type spectrum with a significant ≈195 nm peak and 215 nm minimum is only observed in the 3ABC context, while damped βII type spectra are observed for 3C alone and in the 3CD and 3BC (panel B) contexts. D, a range of CD spectrum shapes are observed for 3C proteases from poliovirus, coxsackievirus A21, coxsackievirus B3, and enterovirus 71.

• 

  Figure 4. Solubility and structure of deletions within 3A region of 3ABC.A, complete sequences of the 3A region from all the 3ABC proteins generated. The bacterial expression levels and solubility of each is listed on a Low/Medium/High scale with “–” indicating no protein was observed. B, CD analysis of deletion mutants to examine sequence elements needed for co-folding to stabilize the 3C conformation. Left; deletion of either the 3Asol or 3Amem regions will trigger the damped CD spectrum for the 3C region. Middle; deletion of 6 and 10 N-terminal residues do not greatly alter the structure while larger deletions could not be expressed and purified. Right; partial deletions within the 3A membrane binding region result primarily in loss of helix signal, with greater effects arising from deleting the first and last halves than deleting the middle of the region. Notably, the ≈195 nm peak remains strong, implying that the 3C region is unaffected by the partial deletions within 3Amem, unlike the complete 3Amem deletion in the left panel.

• 

  Figure 5. Polyprotein junction site cleavage by 3C and 3ABC proteases.A, details of enzyme (E) and substrate (S) constructs, including 3ABCpro mutations made to prevent self-cleavage (red, underlined) and the short N-terminal tag on 3Cpro to separate it from 3C product on gels. A series of substrates were made wherein 14-residue segments containing the various P2 and P3 region polyprotein cleavage sites were placed between two ubiquitin molecules, the C-terminal of which was labeled with ATTO-532 to provide a fluorescence-based cleavage assay. The gel shows reaction endpoint samples of all five junctions stained by Coomassie to visualize all species or imaged by fluorescence to visualize only the substrate and the C-terminal product. Indicators show enzyme (E), substrate (S), and product (P) bands. B, 3Cpro and 3ABCpro cleavage of the five junction sites using 1 μM enzyme and 5 μM substrates. Gels and corresponding plots show that cleavage does not differ significantly in the absence (solid lines) and presence (dotted lines) of 1.5 μM cre26 RNA. Experiments without RNA were replicated with longer reaction times, as shown by matching line colors in the plots at right.

• 

  Figure 6. RNA activation of polyprotein cleavage by 3Cproand 3ABCpro.A, sequences of the polyprotein substrate that contains both the 3A:3B and 3B:3C cleavage sites with full-length 3BC but only part of 3A (Δ marks deletion of 3Amem region and the little “a” in 3aBC denotes a partial 3A protein). The substrate has two cysteine fluorophore labeling sites(F) to allow for visualization of the products formed during the proteolysis assay. The gel shows the two purified enzymes, the substrate, and a partially cleaved substrate stained with Coomassie and imaged by fluorescence; note Coomassie does not stain the small 3aB and 3a product bands. B, fluorescence images of SDS-PAGE resolved protease reaction products with 3C and 3ABC enzymes in the absence and presence of 1.5 μM cre RNA. Bands for substrate (S) and products 3C (P1), Ub-3aB (P2), and Ub-3a (P3) are indicated, and note the 3C labeling efficiency is ≈20% that of the Ub, making the 3C product band appear less intense. C, progress curves for the buildup of 3B:3C site cleavage in reactions containing increasing amounts of cre26 and polyA15 RNA (per color gradient). D, quadratic binding isotherm curve fits of single exponential kobs values…