Structural and molecular basis for Cardiovirus 2A protein as a viral gene expression switch

Abstract

Programmed -1 ribosomal frameshifting (PRF) in cardioviruses is activated by the 2A protein, a multi-functional virulence factor that also inhibits cap-dependent translational initiation. Here we present the X-ray crystal structure of 2A and show that it selectively binds to a pseudoknot-like conformation of the PRF stimulatory RNA element in the viral genome. Using optical tweezers, we demonstrate that 2A stabilises this RNA element, likely explaining the increase in PRF efficiency in the presence of 2A. Next, we demonstrate a strong interaction between 2A and the small ribosomal subunit and present a cryo-EM structure of 2A bound to initiated 70S ribosomes. Multiple copies of 2A bind to the 16S rRNA where they may compete for binding with initiation and elongation factors. Together, these results define the structural basis for RNA recognition by 2A, show how 2A-mediated stabilisation of an RNA pseudoknot promotes PRF, and reveal how 2A accumulation may shut down translation during virus infection.

Fig. 1. 2A adopts a highly basic RNA-binding fold with intrinsic flexibility.

a SDS-PAGE analysis of EMCV 2A (Coomassie). Representative gel from five independent purifications. b SEC-MALS analysis of 2A. The differential refractive index is shown across the elution profile (blue) and weight-averaged molar masses of the indicated peaks are listed. c Topological diagram of “beta-shell” fold: a curved central sheet comprising seven antiparallel beta strands, supported by two helices. d Crystal structure of EMCV 2A in three orthogonal views. N- and C- termini are indicated. <Inset> Electrostatic surface potential calculated at pH 7.4, coloured between +3 (blue) and −3 (red) kT/e⁻. e Four molecules of 2A are present in the asymmetric unit of the crystal, arranged as two pairs of disulfide-linked dimers (spheres). f Superposition of the four NCS-related 2A chains in e reveals regions of conformational flexibility. The width of the cartoon is proportional to atomic B-factor. <Insets > Close-up view of surface loops exhibiting the greatest variation per molecule. Flexible sidechains are shown as sticks, and the Cα backbone deviation is indicated in Å. The positions of two sulfate ions from the crystallisation buffer are indicated with spheres. Source data are provided as a Source Data file.

Fig. 2. 2A binds to a minimal 47 nt element in the viral RNA.

a–b Sequences and schematic diagrams of the EMCV 1–6 constructs used to assay 2A binding. c EMSA analyses showing that removal of the 5′ extension (blue) disables 2A binding. d Microscale thermophoresis (MST) was used to quantify the interactions observed in c. All measurements were repeated as two independent experiments and error bars represent the standard deviation from the mean. RNA concentration ranges between 60 pM–20 µM (for EMCV 1) and 150 pM–5 µM (for EMCV 2–6). Source data are provided as a Source Data file.

Fig. 3. Conformations of EMCV frameshifting RNA and effect of 2A on RNA unwinding.

a <Upper> Schematic diagram illustrating the optical tweezer experiments. RNA is hybridized to ssDNA handles and immobilised on beads. These are used to exert pulling force on the RNA with a focused laser beam. <Lower> Primary sequence of the construct used in optical tweezer experiments, colour coded as in Fig. 2. The location of the cytosine triplet (wild-type, WT) and point mutation (CUC) is indicated. b Predicted conformations of the RNA construct in a. The number of nucleotides involved in each folded structure is indicated. Also see Supplementary Table 6. c Representative force-distance curves of the unfolding (pink) and refolding (blue) transitions of the wild-type (WT) CCC RNA element. d Representative force-distance curves of the unfolding (pink) and refolding (blue) transitions of the wild-type (WT) CCC RNA element in the presence of 300 nM 2A protein. e Global analysis of all the unfolding force trajectories. Number of individual measurements are WT = 117, WT + 2A = 104, CUC = 85, CUC + 2A = 109. Data (black line) are presented as mean values ± SD error bars. f Global analysis of all refolding force trajectories. Number of individual measurements are WT = 111, WT + 2A = 89, CUC = 74, CUC + 2A = 97. Data (black line) are presented as mean values  ± SD error bars. Source data are provided as a Source Data file.

Fig. 4. 2A binds directly to eukaryotic and prokaryotic ribosomes.

a MST binding curves and apparent K_D values using unlabelled 40S subunits at a concentration range of 20 pM–0.4 μM. All measurements were repeated as two independent experiments and error bars represent the standard deviation from the mean. 2A binds with high affinity to the small ribosomal subunit. b As in a with 60S subunits. Error bars as above. c Binding curve and apparent K_D values using unlabelled 30S subunits at a concentration range of 30 pM–1 μM. Error bars as above. 2A shows a strong interaction with the prokaryotic small subunit. d As in c with 50S subunits at a concentration range of 27 pM–0.9 μM. e Binding curves and reported K_D values for 2A-70S IC interactions. Error bars as above. f Same as e, with 2A and vacant 70S. Source data are provided as a Source Data file.

Fig. 5. 2A binds to the 70S ribosome via interactions with the 16S rRNA.

a Cryo-EM analysis of a complex formed between initiated E. coli 70S ribosomes and EMCV 2A. Images (×75,000) were recorded on a Titan Krios microscope. Representative micrograph from dataset of 5730 images. b Cryo-EM map at 2.7Å resolution after focused classification and refinement. Three copies of 2A (orange, red, yellow) are bound to the 16S rRNA of the small (30S) subunit (blue ribbon). c Close-up view of the 2A-binding site. Ribbon diagrams of 2A (coloured as above) and ribosomal RNA (purple) are shown. Protein N- and C- termini are labelled. d Superposition of the three copies of 2A reveals a common RNA-binding surface with conformational flexibility. Residues involved in rRNA binding are labelled and shown as sticks.

Fig. 6. The ‘arginine loop’ plays a central role in RNA recognition.

a–c Details of rRNA recognition by 2A. For each copy of 2A, selected residues involved in interactions are labelled and shown as sticks. <Insets> View of the rRNA surface bound by each copy of 2A. The rRNA helices are colour-coded and labelled. The 2A contact surface is shown as a coloured mesh (orange, red and yellow, respectively). d–f, Close-up view of interactions between the 2A ‘arginine loop’ residues (R95, R97 and R100) and the rRNA backbone (sticks) for each copy of 2A (orange, red, yellow). Polar or electrostatic contacts are indicated by a green dashed line.

Fig. 7. 2A binding may clash with translational GTPases.

a Ribbon diagram of initiated 70S-mRNA-tRNA^fMet-2A complex. Ribosome sites are labelled A, P and E. The initiator tRNA^fMet (dark green), mRNA (light green), and 2A (orange, red, yellow) are shown in two orthogonal views. b Comparison of 70S-2A complex to 70S pre-translocation complex with EF-G (4V7D [10.2210/pdb4v7d/pdb]). 2A binding would clash (red wedges) with EF-G binding. c Comparison of 70S-2A complex to 70S complex with EF-Tu (5WE6 [10.2210/pdb5we6/pdb]). 2A binding would clash (red wedges) with EF-Tu binding.

Fig. 8. Molecular basis for 2A-induced reprogramming of gene expression.

The PRF stimulatory RNA element is predicted to form either stem-loop or pseudoknot conformations. As 2A accumulates during EMCV infection, it selectively binds to and stabilises a pseudoknot-like conformation of the PRF stimulatory element, thereby enabling PRF, producing trans-frame product 2B* and downregulating the expression of enzymatic viral proteins later in infection. 2A also binds directly to the small ribosomal subunit at the translational GTPase factor binding site, progressively inhibiting both initiation and elongation as it accumulates. This may contribute to the shutdown of host cell translation during lytic infection.