Evolutionary selection of pestivirus variants with altered or no microRNA dependency

Figures

• 

  Figure 1.. Importance of the BVDV 3′ UTR miRNA seed sites. (A) Schematic representation of the BVDV genome drawn to scale. The single ORF is flanked by 5′ and 3′ UTRs. The secondary structure of the NADL strain genomic 3′ end was predicted using Mfold and illustrated using VARNA. SLI and SLII are depicted and the let-7 seed (S1) and the miR-17 seed (S2) sites are indicated. Only S2 miR-17 binding was assumed for this structure. (B) Viral replication of BVDV wild-type and its seed site mutants (S1p3,4 and S2p3,4) after electroporation of MDBK cells measured in GE by RT-qPCR on intracellular RNA normalized to housekeeping gene RPS11 and the 4 h time point. Pol(−), non-replicating control. (C) Virus production after infection of MDBK cells with BVDV wild-type and seed site mutants was quantified by the PFU assay. (D) Viral replication of BVDV wild-type or BVDV-S2p3,4 by RT-qPCR on intracellular RNA normalized to RPS11 after infection at MOI = 0.005. Trans-complementation with 10 nM miRNA mimics is indicated. (E, F) Viral replication of BVDV wild-type, BVDV-S2p3,4, BVDV-S2p3,4,14–16 and BVDV-S2p3,4,G12576–7 after electroporation of MDBK cells measured by RT-qPCR on intracellular RNA normalized…

• 

  Figure 2.. Evolutionary selection of alternative miRNA seed sites by randomization of S2. (A) Evolutionary selection strategy and outcome for BVDV S2 randomization experiment. Mock- or LNA-17-treated MDBK cells were electroporated with a pool of S2 randomized genomes. Shown are the genomes selected 24 h after passage of supernatant to naïve cells. (B) AGO-CLIP-derived miRNA abundance profile of MDBK cells. Data from (22). (C) Population structure over time after electroporation of BVDV S2 randomized pool in mock cells. RNA from the input pool and RNA isolated after 6 and 12 h (intracellular) and 24 h (extracellular) were deep sequenced and analyzed for miRNA seed site abundance. The fraction of each selected sequence is shown based on the S2 6mer. (D) Population structure of subdominant seed sites. Shown are data from (C) after removing genomes containing the miR-17 site. The top selected seeds are color coded according to (B). (E) Prediction of BVDV S2 auxiliary pairing energies of the top 50 miRNAs expressed in MDBK cells relative to their abundance levels. Binding energies were predicted using IntaRNA with a fixed seed region and unpaired nt 10–11, such that only auxiliary pairing was…

• 

  Figure 3.. Evolutionary selection of alternative S1 and S2 sequences by randomization. Population structure over time after electroporation of MDBK cells with an S2 randomized pool in context of an miR-17 S1 site (A) or S1 randomized pools in context of an S2 miR-17 site (B) or S2 let-7 site (C). RNA from the input pool and RNA isolated at time points indicated by dotted lines were deep sequenced and analyzed for miRNA seed site abundance. Intracellular RNA was used for 6 and 12 h time points and extracellular RNA was used for 24 h and later time points. The fraction of each selected sequence is shown based on the seed site 6mer. Schematics of BVDV genomes are depicted above each graph. (D) Summary of the top 10 selected 6mer seed sites and their abundance for BVDV-let-7/7N, BVDV-miR-17/7N, BVDV-7N/miR-17 and BVDV-7N/let-7 experiments at 24 h (S2 randomizations) or 120 h (S1 randomizations) after electroporation in MDBK cells. Frequencies of sequences with only 1 nt difference from miR-17 or let-7 sites at any core 6mer position were summed and reported as miR-17- or let-7-like. The 6mer nucleotide sequence is shown whenever no corresponding miRNA seed site was identified.

• 

  Figure 4.. miRNA binding and dependence for BVDV variants with alternative seed sites. (A) Viral replication of BVDV-2×let-7, BVDV-swap and BVDV-S1p3,4 and BVDV-S2p3,4 mutants after electroporation of MBDK cells measured by RT-qPCR on intracellular RNA normalized to RPS11 and the 4 h time point. Data for BVDV wild-type and BVDV-pol(−) are the same as in Figure 1B. (B) AGO/miRNA binding map across the genomes of BVDV seed site mutants normalized to library read depth. MDBK cells were infected (MOI = 0.1, except for BVDV-let-7p3,4/let-7 with MOI = 0.025) with the indicated mutants and, after 48 h, were cross-linked and processed for CLIP. Genome schematic with cellular Jiv90 insertion shown in cyan is drawn to scale below. (C–F) miRNA-specific binding maps and inhibitor experiments for BVDV seed site mutants. Left panels show miRNA-specific chimera-derived binding maps across the BVDV 3′ UTR seed site region. Schematics of BVDV genomes are depicted above each graph. The binding site locations for let-7 and miR-17 are shown across the graphs in purple and green shaded bars, respectively. Specific miRNA binding is color coded according to legend in Figure 2B. Right panels show infectious virus…

• 

  Figure 5.. Infection kinetics and translation efficiencies of BVDV seed site mutants. AGO/miRNA binding maps across the genome (A) and miRNA-specific chimera-derived binding profiles across the 3′ UTR seed site regions (B) are shown for BVDV-UUUGUAG/let-7 and BVDV-2×miR-17. MDBK cells were infected (MOI = 1) with the indicated mutants and after 24 h cells were cross-linked and processed for CLIP. Annotation as in Figure 4. (C) Infectious virus production (TCID50/ml) after infection (MOI = 0.01) of MDBK cells with third passage stocks of BVDV seed site mutants. (D) Virus production as in (C) with third passage stock of a culture adapted population of viruses that came from a BVDV-let-7/let-7p3,4 clone propagated for 12 days (BVDV-let-7/let-7p3,4-day12) in MDBK cells. LLOQ: lower limit of quantification. Error bars represent SD of the mean. (E) BVDV IRES-driven translation measured by bidirectional reporters with and without modification of the S1 and S2 miRNA seed sites. Translation levels from all reporters were normalized to the wild-type and differences to this were compared by ANOVA analysis. Corrections were done according to Dunnett’s test; ns, not significant; *P ≤ 0.05; **P ≤ 0.01;…

• 

  Figure 6.. BVDV modifications enabling miRNA-independent replication. (A) Schematic of the BVDV wild-type genomic 3′ end depicting the location of the AG and poly-U insertions. (B) Infectious virus production (TCID50/ml) after infection (MOI = 0.1) with BVDV-let-7/let-7p3,4-day12 supernatant of LNA-treated MDBK cells. Schematic of the representative BVDV genome is given in (C). miRNA-specific chimera-derived binding profiles across the 3′ UTR seed site region (C) and AGO/miRNA binding maps across the genome (D) for the engineered BVDV-independent mutant, BVDV-ind. MDBK cells were infected (MOI = 1), and after 24 h cells were cross-linked and processed for CLIP. Annotation as in Figure 4. (E) Viral replication of BVDV-ind in AGO-depleted MDBK cells after infection (MOI = 0.01) measured by RT-qPCR on intracellular RNA normalized to RPS11 and the 4 h time point. (F) Infectious virus production (TCID50/ml) in MDBK cells after electroporation with AG insertion variants. (G) Viral replication of BVDV-poly-U variants after electroporation of MDBK cells measured by RT-qPCR on intracellular RNA normalized to RPS11 and the 4 h time point. (H) Viral replication of BVDV-let-7/let-7p3,4+AG insertion…

• 

  Figure 7.. Putative model for miRNA-mediated structure switching. Predicted structure alternatives for BVDV wild-type with and without miR-17 binding (top) and BVDV-ind (bottom) are shown including folding energies (kcal/mol). Only S2 binding was assumed. miR-17 seed sites are drawn in green, let-7 in purple, p3,4 mutations in gray and 3′ UTR insertions in orange. Arrows represent the balance between structural states and the difference in energy (ΔΔG) is given.