Switch from cap- to factorless IRES-dependent 0 and +1 frame translation during cellular stress and dicistrovirus infection

Abstract

Internal ribosome entry sites (IRES) are utilized by a subset of cellular and viral mRNAs to initiate translation during cellular stress and virus infection when canonical cap-dependent translation is compromised. The intergenic region (IGR) IRES of the Dicistroviridae uses a streamlined mechanism in which it can directly recruit the ribosome in the absence of initiation factors and initiates translation using a non-AUG codon. A subset of IGR IRESs including that from the honey bee viruses can also direct translation of an overlapping +1 frame gene. In this study, we systematically examined cellular conditions that lead to IGR IRES-mediated 0 and +1 frame translation in Drosophila S2 cells. Towards this, a novel bicistronic reporter that exploits the 2A "stop-go" peptide was developed to allow the detection of IRES-mediated translation in vivo. Both 0 and +1 frame translation by the IGR IRES are stimulated under a number of cellular stresses and in S2 cells infected by cricket paralysis virus, demonstrating a switch from cap-dependent to IRES-dependent translation. The regulation of the IGR IRES mechanism ensures that both 0 frame viral structural proteins and +1 frame ORFx protein are optimally expressed during virus infection.

Figure 1. Secondary structures of the CrPV (Type I) and IAPV (Type II) IGR IRESs.

(A) Distinct IRESs direct translation of nonstructural (ORF1) and structural (ORF2) polyproteins. (B) Schematic of the IRESs showing pseudoknots, PKI, PKII, and PKIII, stem loops SLIII, SLIV, SLV, and SLVI, and loop L1.1. The UAA stop codon of the IAPV ORF1 is shown in bold. The overlapping +1 frame ORFx is shown. Translation of IAPV ORFx is mediated by a U-G base pair adjacent to PKI (dashed lines).

Figure 2. Construction of a T2A-containing +1 frame IRES bicistronic reporter construct.

(A) The T2A sequence (dark grey) is inserted between an NdeI restriction site (boxed and italicized) and the Firefly luciferase (FLuc) gene. The arrow indicates the ‘self-cleavage’ or ‘stop-go’ site. A mutation within the T2A peptide (D to E), which inactivates T2A ‘self-cleavage’ activity is shown. (B) Bicistronic reporter constructs containing the IAPV IGR IRES and the ORFx region fused in the +1 frame with the FLuc gene. The T2A coding sequence (in grey) is inserted between the ORFx and FLuc. T2A-minus (left) and T2A-containing (right) bicistronic reporter constructs are shown. (C) T2A-minus and T2A-containing +1 frame bicistronic constructs were incubated in Sf21 extracts for 120 minutes in the presence of [³⁵S]-methionine and analyzed by SDS-PAGE and autoradiography. (D) In vivo translation in S2 cells. In vitro transcribed 5′ capped bicistronic reporter RNAs were transfected into Drosophila S2 cells. Cells were harvested at 6 hours, lysed and luciferase activity was measured. The white and black boxes represent RLuc and FLuc luciferase expression, respectively, indicative of cap-dependent and IRES-dependent translation. Relative luciferase activities (RLA), the quantitation of FLuc and RLuc enzymatic activity, and the relative ratios of FLuc/RLuc are normalized to that observed with the +1 frame T2A-containing reporter RNA. Shown are averages from at least three independent experiments (± s.d.).

Figure 3. IRES-mediated translation in Drosophila S2 cells.

(A) In vitro transcribed 5′ capped bicistronic RNAs containing wild-type or mutant CrPV and IAPV IGR IRESs were transfected into Drosophila S2 cells. Cells are harvested at 6 hours after transfection, lysed and luciferase activity was measured. To confirm CrPV IGR IRES dependent activity, reporter RNA containing a CrPV double mutant (ΔPKI/ΔPKIII) bearing both ΔPKI (CC6214-5GG) and ΔPKIII (CAC6148-50GUG) mutations and an empty construct, which denotes a bicistronic RNA that does not have an IRES were assayed. FLuc, RLuc, and the ratio of FLuc/RLuc are normalized to that observed with reporter RNAs containing the CrPV IGR IRES and IAPV IGR IRES, respectively. (B) Comparison of different dicistrovirus IGR IRES translation in S2 cells. FLuc, RLuc, and the ratio of FLuc/RLuc are normalized to translation of the reporter RNA containing the IAPV IGR IRES. Shown are averages from at least three independent experiments (± s.d.).

Figure 4. IAPV IRES-mediated +1 frame translation using the T2A-containing reporter construct.

(A) Schematic of mutations used within the PKI domain of the IAPV IGR IRES. (B) Translational activity of reporter constructs containing wild-type or mutant IAPV IGR IRESs in Sf21 extracts. Bicistronic reporter constructs were incubated in Sf21 extracts in the presence of [³⁵S]-methionine. (below) Quantitation of the FLuc/RLuc ratio (below) is normalized to the wild type ratio. (C) Translational activity in Drosophila S2 cells. In vitro transcribed capped T2A-containing reporter RNAs were transfected into S2 cells. Cells were harvested at 6 hours after transfection and luciferase activities were measured. (D) Comparison of 0 and +1 frame IAPV IGR IRES-mediated translation of T2A-containing reporter RNAs transfected in S2 cells. Luciferase activities were quantitated 6 hours after transfection. For C) and D), luciferase activities are shown as a ratio of FLuc/RLuc and as individual FLuc and RLuc activities. Shown are averages from at least three independent experiments (± s.d.).

Figure 5. IGR IRES-mediated translation during cellular stress.

CrPV and IAPV IGR IRES-mediated 0 and +1 frame translation was monitored in S2 cells treated with DTT (4 mM) (A–C), pateamine A (D), or 4E1RCat (8 µM) (E). Bicistronic reporter RNAs containing the CrPV or IAPV IGR IRES were transfected in S2 cells. After one hour transfection, cells were treated alone or with the drug for another 5 hours. (C) Relative ratio of IAPV IGR-mediated 0 and +1 frame translation of T2A-containing reporter RNAs transfected in DTT-treated S2 cells. Except in (D), shown are averages from at least three independent experiments (± s.d.).

Figure 6. CrPV IGR IRES-mediated translation in CrPV-infected S2 cells.

Bicistronic reporter RNAs containing the wild-type or mutant (ΔPKI) CrPV IGR IRES were transfected into mock or CrPV-infected (inf) S2 cells (MOI 25) at 0.5 hour post infection. (A) Cells were harvested at 6 hours after transfection (shown are averages from at least three independent experiments ± s.d.) or (B) at the indicated times and luciferase activities were measured. The RLuc and FLuc activities are normalized to that at 6 hours after transfection in mock-infected cells.

Figure 7. IAPV IGR IRES-mediated 0 and +1 frame translation in CrPV-infected S2 cells.

(A, B) Bicistronic reporter RNAs containing the wild-type or mutant (ΔPKI) IAPV IGR IRES were transfected into mock or CrPV-infected (inf) S2 cells (MOI 25) at 0.5 hour post infection. (A, B) Cells were harvested at 6 hours post transfection or (C, D) at the indicated times and luciferase activities were measured. Reporter RNAs monitoring (C) 0 frame or (D) +1 frame translation are shown. The RLuc and FLuc activities are normalized to that at 6 hours after transfection in mock-infected cells. (E) Quantitation of the relative ratio of +1/0 frame IAPV IGR IRES translation in mock- or CrPV-infected cells. The ratios are normalized to that of the 1 hour after transfection in mock cells. Shown are averages from at least three independent experiments ± s.d.