HCV IRES domain IIb affects the configuration of coding RNA in the 40S subunit's decoding groove

Abstract

Hepatitis C virus (HCV) uses a structured internal ribosome entry site (IRES) RNA to recruit the translation machinery to the viral RNA and begin protein synthesis without the ribosomal scanning process required for canonical translation initiation. Different IRES structural domains are used in this process, which begins with direct binding of the 40S ribosomal subunit to the IRES RNA and involves specific manipulation of the translational machinery. We have found that upon initial 40S subunit binding, the stem-loop domain of the IRES that contains the start codon unwinds and adopts a stable configuration within the subunit's decoding groove. This configuration depends on the sequence and structure of a different stem-loop domain (domain IIb) located far from the start codon in sequence, but spatially proximal in the IRES•40S complex. Mutation of domain IIb results in misconfiguration of the HCV RNA in the decoding groove that includes changes in the placement of the AUG start codon, and a substantial decrease in the ability of the IRES to initiate translation. Our results show that two distal regions of the IRES are structurally communicating at the initial step of 40S subunit binding and suggest that this is an important step in driving protein synthesis.

FIGURE 1.

SHAPE analysis of wild-type HCV IRES RNA in the unbound and 40S-bound forms. (A) Schematic of HCV IRES translation initiation where the IRES first binds the 40S subunit (orange), then eukaryotic initiation factor (eIF) 3 (green), and the ternary complex (eIF2-GTP-Met-tRNA_i^Met, blue), followed by GTP hydrolysis and eIF release. The 60S subunit (red) then joins to form an 80S ribosome. Removal of domain II (ΔdII) blocks the formation of 80S ribosomes. (B) Representative SHAPE analysis gel of full-length HCV IRES RNA in the free form (left) and 40S subunit-bound form (right). Lanes containing free RNA or bound to the 40S subunit are labeled, reactions lanes are marked as NMIA (two concentrations of NMIA are shown), control lanes are marked as DMSO (DMSO added, no NMIA) and RT (no DMSO or NMIA added). Lanes 1–4,9–12 are the sequencing reactions. Reference nucleotide numbers are bulleted on the left and the parts of the gel that correspond to different IRES secondary structural elements are indicated by gray bars to the right. The location of the start codon AUG is indicated with red arrowheads. (C) Secondary structure of the full-length HCV IRES RNA, where decreases in NMIA modification upon 40S subunit binding are designated with gray, and regions with increases in modification are in blue. Structural elements are labeled. The nucleotides 3′ of domain IV (faded gray) are not visible in this analysis, as this is where the primer anneals for reverse-transcription. (D) Quantitated, normalized, and background-corrected modification data from two independent SHAPE probing experiments, with error bars representing one standard deviation from the mean of both experiments. The experiment was repeated three additional times (data not shown), and the replicates validate the quantitated data shown here. The degree of modification is on the y-axis; each nucleotide is on the x-axis with the start codon AUG colored red and the location of domain IV indicated. Red bars indicate free IRES; gray bars are 40S bound IRES. Gray regions indicate a decrease in modification upon 40S subunit binding and cyan regions indicate an increase in SHAPE modification upon the addition of the 40S subunit.

FIGURE 2.

SHAPE analysis of ΔdII HCV IRES RNA in the unbound and 40S-bound forms. (A) Representative SHAPE analysis gel of WT full-length and ΔdII HCV IRES RNAs in the free (unbound) form and 40S subunit-bound form. Lanes containing free RNA or bound to the 40S subunit are labeled, reaction lanes are marked as NMIA, control lanes are marked as DMSO (DMSO added, no NMIA) and RT (no DMSO or NMIA added). Lanes 1–4 contain the sequencing reactions. Reference nucleotide numbers are bulleted on the left and the parts of the gel that correspond to different IRES secondary structural elements are indicated by gray bars to the right. The location of the start codon AUG is indicated with red arrowheads. (B) Secondary structure of the ΔdII HCV IRES RNA where decreases in NMIA modification due to 40S subunit binding are designated with gray and regions with increases in modification upon 40S subunit binding are in blue. Structural elements are labeled. The nucleotides 3′ of domain IV (faded gray) are not visible in this analysis, as this is where the primer anneals for reverse-transcription. The part of the ΔdII IRES RNA that shows a different modification than WT full-length IRES, when bound to the 40S subunit, is indicated. (C) Quantitated, normalized, and background-corrected modification data from two independent SHAPE probing experiments, with error bars representing one standard deviation from the mean of both. The experiment was repeated three additional times (data not shown), and the replicates validate the quantitated data shown here. The degree of modification is on the y-axis, each nucleotide is on the x-axis, with the start codon AUG colored red and the location of domain IV indicated. Red bars indicate full-length IRES, blue bars are ΔdII IRES. The top graph contains the modifications for each RNA in the free form; the bottom graph contains a comparison of modifications on each RNA in the 40S subunit-bound form. The dashed box indicates regions that change in their modification pattern when domain II is removed.

FIGURE 3.

SHAPE analysis of dIIb mutants. (A) Model of the placement of domain II and the start codon in the P-site built from the cryo-EM reconstruction of the HCV IRES RNA bound to the 40S subunit (Spahn et al. 2001) (pale transparent yellow and magenta surface), the crystal structure of the 40S subunit from Tetrahymena thermophila (white ribbons) (Rabl et al. 2011), the NMR structure of domain II of the HCV IRES RNA (magenta) (Lukavsky et al. 2003), and the crystal structure of the Thermus thermophilus 70S ribosome with tRNAs and mRNA bound (Selmer et al. 2006) (the mRNA from this structure is shown in blue with the AUG in the P-site in red). This model has little or no steric clash and matches all structural information very well. The model shows the proximity of the tip of domain II (dIIb) to the decoding groove and to the RNA surrounding the AUG. The cryo-EM reconstruction (Spahn et al. 2001) of the full-length IRES (magenta) bound to the 40S subunit (yellow) is inset for reference. DIIb is indicated, and the region where nucleotide G82 is located is indicated by the red box, and the decoding groove and approximate location of the AUG are labeled. (B) Secondary structure cartoon of the HCV IRES with dIIb indicated. Mutants used in this analysis are shown. (C) Representative SHAPE analysis gel of the WT full-length IRES and each of the dIIb mutants. Reactions with or without bound 40S subunit (+/−40S) are in lanes 6–21, the sequencing ladder is in lanes 1–4, and the reverse transcription (RT) control is in lane 5. Reference nucleotide numbers are bulleted on the left and secondary structural domains indicated by gray bars on the right of the gel. The location of the start codon AUG is indicated with red arrowheads. (D) Quantitated, normalized, and background-corrected modification data from two independent SHAPE probing experiments, with the error bars representing one standard deviation from the mean of both. The experiment was repeated two additional times (data not shown), and the replicates validate the quantitated data shown here. The degree of modification is on the y-axis, each nucleotide is on the x-axis with the start codon AUG colored red and the location of domain IV indicated. Red bars indicate WT IRES, bars in different shades of blue indicate the dIIb mutant IRESs. The top graph contains the modifications for each RNA in the free form; the bottom graph contains a comparison of modifications on each RNA in the 40S subunit-bound form. The dashed box indicates regions that change in their modification pattern (relative to WT) when dIIb is mutated.

FIGURE 4.

In vitro translation initiation assays of WT and mutant IRES RNAs. (A) Diagram of the uncapped monocistronic reporter construct used in these assays. (B) Denaturing gel of each IRES containing reporter construct showing that all RNAs are homogenous in species and concentration. (C) Level of Photinus luciferase after 90 min for WT, ΔdII, and each dIIb mutant. Error bars represent one standard error from the mean of three independent, triplicate experiments. (D) Comparison of the degradation rates of internally radiolabeled WT and dIIb mutant reporter mRNAs in RRL over the course of the experiment. (E) Graph of the degradation rates of WT and dIIb mutant reporter RNAs in RRL over a 90-min time course. Error bars represent one standard deviation from the mean of three experiments.

FIGURE 5.

Domain IIb mutants interact with the 40S subunit properly. (A) RNase T1 footprinting gel with G82 cleavage indicated by the arrowhead. Lanes with and without bound 40S subunit are shown. A RNase T1 ladder made under denaturing conditions and a hydrolysis ladder are marked as T1 and OH, respectively. BKG lane is a no-RNase T1 control. (B) Quantified and normalized G82 bands from the gel in A show that the G82 of the ΔapexC mutant RNA is protected to the same degree as is WT. (C) Quantitated, normalized, and background-corrected modification data from two independent SHAPE probing experiments of the WT and dIIb mutant IRESs in the 40S subunit-bound form with an additional repeat (data not shown), validating the quantitated data shown here. Error bars represent one standard deviation from the mean of both independent experiments. The degree of modification is on the y-axis; each nucleotide is on the x-axis. Red bars indicate WT IRES; bars in different shades of blue indicate the dIIb mutant IRESs. The part of the gel corresponding to the tip of domain IIb (dIIb) is indicated.

FIGURE 6.

Investigation of possible base-pairing between domain IIb and domain IV. (A) Comparison of the sequences at the apical loop of domain II (dIIb) and the RNA in and around the AUG start codon (underlined) for 11 IRES RNAs proposed to have HCV-like secondary structures (Pestova and Hellen 1999; Hellen and de Breyne 2007). Boxes around parts of each sequence indicate the proposed location of the dIIb apical loop. Sequences were aligned based on the proposed position in each secondary structure. The possible base-pairing interaction in the HCV IRES is indicated with bars above the aligned sequences. In each sequence, red indicates stretches of two or more bases that could potentially pair (identified by visual examination). Uppercase red letters could form Watson–Crick pairs; lowercase could form G•U pairs. The pairing sequences, number of potential pairs, and location of these pairs is not conserved. (B) Model from Figure 3A with only the tip of domain II and the position of the mRNA and AUG in the P-site shown. The location of bases that could pair is shown with various colors; pairing partners have matching colors. The distance between potential pairing bases is shown. The cryo-EM reconstruction (Spahn et al. 2001) of the full-length IRES (magenta) bound to the 40S subunit (yellow) is inset for reference. The part of the cryo-EM shown in the model is boxed in red. (C) Cartoon of HCV IRES secondary structure with mutations made in dIIb and dIV highlighted in red, and the putative base-pairing interaction indicated by lines connecting the two domains and yellow shading. AUG is indicated at the apex of dIV. (D) Quantitated, normalized, and background-corrected modification data from two independent SHAPE probing experiments of the WT IRES RNA and the mutants shown in C, all bound to the 40S subunit, with error bars representing one standard deviation from the mean of both. Note that the additional repeat for both the unbound and the IRES•40S bound state (data not shown) validates the quantitated data shown here. The degree of modification is on the y-axis; each nucleotide is on the x-axis, with the start codon AUG colored red and the location of domain IV indicated. Orange bars indicate WT IRES; bars in different shades of purple indicate mutants. The top graph contains the modifications for each RNA in the dIIb apical loop region; the bottom graph contains a comparison of modifications on each RNA in the dIV region. The start codon AUG is indicated in red.

FIGURE 7.

Toeprinting analysis of WT and mutant IRES RNAs. (A) Toeprinting gel of all mutants in the free form and bound to the 40S subunit. The relevant part of this gel is boxed and expanded to the right. The location of the A in the AUG is indicated as the +1 position, and stops at the +9–11 position and at location +15/16 are indicated by arrows to the right of the expanded gel region. (B) Quantitated, background-corrected, and graphed toeprints of IRES•40S from the gel in A. Location +1 (AUG) is indicated by the black line, the +16 toeprint by the blue, and the +9–11 by the gray line. The location of IRES secondary structural regions is shown beneath the graphs.