Modulation of translation initiation efficiency in classical swine fever virus

Abstract

Modulation of translation initiation efficiency on classical swine fever virus (CSFV) RNA can be achieved by targeted mutations within the internal ribosome entry site (IRES). In this study, cDNAs corresponding to the wild-type (wt) or mutant forms of the IRES of CSFV strain Paderborn were amplified and inserted into dicistronic reporter plasmids encoding Fluc and Rluc under the control of a T7 promoter. The mutations were within domains II, IIId(1), and IIIf of the IRES. The plasmids were transfected into baby hamster kidney (BHK) cells infected with recombinant vaccinia virus vTF7-3, which expresses the T7 RNA polymerase. IRES mutants with different levels of IRES activity were identified and then introduced by homologous recombination into bacterial artificial chromosomes (BACs) containing CSFV Paderborn cDNA downstream of a T7 promoter. From the wt and mutant BACs, full-length CSFV RNA transcripts were produced in vitro and electroporated into porcine PK15 cells. Rescued mutant viruses were obtained from RNAs that contained mutations within domain IIIf which retained more than 75% of the wt translation efficiency. Sequencing of cDNA generated from these rescued viruses verified the maintenance of the introduced changes within the IRES. The growth characteristics of each rescued mutant virus were compared to those of the wt virus. It was shown that viable mutant viruses with reduced translation initiation efficiency can be designed and generated and that viruses containing mutations within domain IIIf of the IRES have reduced growth in cell culture compared to the wt virus.

Fig 1

Structure of the dicistronic reporter plasmids. The wt (nt 47 to 427) IRES or mutant derivatives from the CSFV Paderborn strain were inserted downstream from the coding region of firefly luciferase sequence (Fluc) and upstream from the Renilla luciferase (Rluc) sequence. Both coding regions were placed under the transcriptional control of a T7 promoter. The configuration of the reporter plasmids allows the translation of Fluc by a cap-dependent mechanism, whereas the translation of Rluc is governed by the CSFV IRES or the mutant derivatives.

Fig 2

Generation of full-length CSFV cDNAs containing mutations in the IRES element. Construction of pBeloPader10 was described previously (41). (A) Nucleotides 99 to 377 of the CSFV full-length cDNA were replaced by a cassette conferring streptomycin sensitivity and neomycin resistance (rpsL/neo cassette), giving rise to a new construct, pBeloPader_rpsL. (B) Fragments corresponding to nt 47 to 427, containing the mutated IRES element, were recombined within E. coli into pBeloPader_rpsL, replacing the rpsL/neo cassette, thus restoring streptomycin resistance and neomycin sensitivity to the cells.

Fig 3

Secondary structure of the CSFV strain Paderborn IRES. The model is based on a previous published model of the CSFV strain Alfort IRES (11). The names of the different domains are indicated. Ovals mark the domains that were targeted for mutagenesis in this study. The initiation codon is indicated in bold and underlined.

Fig 4

The role of the CSFV domain II in IRES activity. The full-length 5′ UTR or IRES elements were introduced into dicistronic plasmids that express Fluc and Rluc. The Fluc activity was used to normalize the data for differences in transfection efficiency. Data are presented as normalized Rluc activities relative to those of the wt IRES (nt 47 to 427) (% wt). (A) Comparison of translation efficiencies of the full-length 5′ UTR and nt 47 to 427 (wt IRES) from dicistronic reporter plasmids in vTF7-3-infected BHK cells. (B) Secondary structures of the wt and truncated mutants within domain II. (C) Comparison of translation efficiencies of the wt IRES element and the mutants containing truncated domain II sequences within the IRES. Data in panels A and C are presented as means ± SD (n ≥ 3). N.S, not significant; *, P < 0.05; **, P < 0.01.

Fig 5

The role of domain IIId₁ in CSFV IRES activity. (A) Predicted secondary structures of the wt and mutant domain IIId₁ elements. (B) Comparison of translation efficiencies of the wt IRES and mutant IRES elements with modifications within domain IIId₁. The IRES elements were inserted into a dicistronic reporter plasmid and assayed as described for Fig. 4. Data are represented as means ± SD (n ≥ 3). **, P < 0.01; ***, P < 0.001.

Fig 6

The roles of stem 1a and stem2 of domain IIIf in IRES activity. (A) Secondary structure of the wt and mutant domain IIIf (pseudoknot) elements. (B) Comparison of translation efficiencies of the wt IRES element and IRES elements mutated within stem 1a and stem 2 within domain IIIf. The IRES elements were inserted into a dicistronic reporter plasmid and assayed as described for Fig. 4. Data are represented as means ± SD (n ≥ 3). N.S, not significant; *, P < 0.05; **, P < 0.01; ***, P < 0.001.

Fig 7

Growth characteristics of wt and mutant CSFVs in cells. (A) wt Pader10- and PaderS2m5_IIId₁4-infected PK15 cells stained with NS3-specific antibody (green) and DAPI (blue) 72 h postinfection. Pictures were taken at ×10 magnification. (B) The growth of the rescued viruses wt Pader10, PaderS2m5, PaderS2m6, and PaderS2m7 in PK15 cells was evaluated using one-step growth curves. Virus titers were determined from harvests prepared at 3, 12, 24, 48, and 72 h postinfection. (C) Viral RNA accumulation from PK15 extracts infected with the rescued viruses wt Pader10, PaderS2m5, PaderS2m6 and paderS2m7 was determined by quantitative RT-PCR at 3, 12, 24, 48, and 72 h postinfection. Data in panels B and C are represented as means + SD (n = 3).