Rational attenuation of RNA viruses with zinc finger antiviral protein

Figures

• 

  Fig. 1. CpG dinucleotide numbers and HIV-1 replication. a, Schematic representation of HIV-1 genome containing an EGFP reporter in place of the Nef ORF and introduced restriction sites BstEII and ClaI in the 5′ portion of the env gene. b, ZAP proteins detected by western blotting in unmanipulated and functionally ZAP-deficient human MT4 T cells. The CRISPR lesion results in the expression of truncated non-functional ZAP proteins indicated by asterisks. c, Replication of a collection of HIV-1 mutants containing between 0 and 23 CpG dinucleotides, as well as WT and CG-43 in ZAP-expressing or ZAP-deficient cells. At each day after initial infection, a small sample of cells was collected and the percentage of GFP-positive cells was measured by flow cytometry. d, Percentage of infected cells at day 4 post initial infection across all mutant viruses. Mean ± s.d. from 3 independent experiments; two-way ANOVA for the presence of ZAP (column factor) P < 0.0001, number of CpG (row factor) P < 0.0001. Šídák’s multiple comparisons test was used to calculate adjusted P values between ZAP+/+ and ZAP−/− groups, comparisons of virus mutants with more than 3 CpG display P(adj) < 0.05. Source data

• 

  Fig. 2. Spacing and composition between CpG dinucleotides affect HIV-1 replication. a, Viral replication in cells infected with mutants of HIV-1 GFP reporter viruses containing 15 additional CpG dinucleotides with a mean of 6, 11, 14, 32 or 40 nucleotides between each CpG dinucleotide. Each day, the percentage of infected cells was measured by flow cytometry. b, Summary of the percentage of GFP-positive cells at day 4 post initial infection. Mean ± s.d. from 3 independent experiments; two-way ANOVA for the presence of ZAP (column factor) P < 0.0001, spacing (row factor) P = 0.002. Šídák’s multiple comparisons test was used to calculate adjusted P values between ZAP+/+ and ZAP−/− groups, comparisons of virus mutants with spacing between CpGs ≥14 nt display P(adj) < 0.0001. c, CLIP-seq analysis of ZAP binding to HIV-1 RNA: ZAP-deficient and TRIM25-deficient 293T cells were transfected with a plasmid encoding ZAP-L as well as HIV-1 proviral plasmids with 0 CpG dinucleotides (black line), or 15 CpG dinucleotides with average spacing of 11 nucleotides (red line) or 32 nucleotides (blue line) in the BstEII-ClaI interval. CLIP reads that mapped to the BstEII-ClaI interval are plotted as the…

• 

  Fig. 3. ZAP can impose selective pressure to deplete CpG dinucleotides. a, Replication, monitored by flow cytometry, of the CG-43 HIV-1 mutant in unmanipulated MT4 cells in 3 replicate cultures (Rep1–3). b, Replication of the CG-15 HIV-1 mutant in MT4 cells in 3 replicate cultures. Supernatants were collected, filtered and used to infect a new culture of MT4 cells when the percentage of infected cells exceeded 80%, as indicated by arrows. c, Replication of the CG-15 HIV-1 mutant with re-introduced mutations (G6574A or G6565A) in ZAP+/+ and ZAP−/− cells. Source data

• 

  Fig. 4. Genomic recoding sensitizes EV-A71 to ZAP. a, CpG content of example virus genomes; observed/expected ratios (based on mononucleotide composition) for HIV-1, Sindbis virus and several picornaviruses. b, Schematic diagram of reporter EV-A71 containing the NanoLuc luciferase gene (blue) and a 2A cleavage site. P1, P2 and P3 indicate the primary processed proteins derived from the EV-A71 polyprotein. The recoded region (dashed line box) is approximately 1 kb and encodes a portion of the polyprotein located in P2 and P3. c, Summary of the recoding modifications introduced in the EV-A71 genome. d–f, Replication of reporter EV-A71 mutants in ZAP+/+ or ZAP−/− HeLa cells. NanoLuc luciferase activity was measured every 12 h (d). Summary of NanoLuc luciferase levels at 60 h post infection, mean ± s.d. of 3 independent experiments (e). Viral RNA levels measured at 60 h post infection, mean ± s.d. of 3 independent experiments (f). g, Replication of EV-A71(CG-48)/A+ in ZAP+/+ or ZAP−/− HeLa cells over 4 passages. At 4 d post initial infection, supernatants were collected, filtered through a 0.22 µm filter and used to infect a fresh culture of cells. Source data

• 

  Fig. 5. Recoded EV-A71 has ZAP-dependent attenuation and elicits protective antibodies in mice. a, Clinical score following infection of 1-day-old ZAP+/+ and ZAP−/− mice with mEV-A71 WT or mEV-A71/CG-48/A+ (mean ± s.d.; n = 11–26 mice per group; P values calculated using two-way ANOVA; NS, non-significant). b, Examples of characteristic pathology developed in infected mice following mEV-A71 infection, including one-limb and two-limb paralysis. c, Probability of ZAP+/+ and ZAP−/− mouse survival (%) following infection with mEV-A71 WT or mEV-A71/CG-48/A+ (n = 11–36 mice per group; P value calculated using Mantel-Cox test). d, One-day-old mice (n = 4–6 per group) were infected with indicated virus; 6-days-post-infection mice were sacrificed and muscles from both posterior limbs were collected, homogenized and total RNA was extracted. EV-A71-specific RNA was quantified by qPCR. Dashed line indicates limit of qPCR detection. e, Schematic representation of experimental design. ZAP+/+Ifnar−/− female mice previously inoculated with mEV-A71/CG-48/A+ (or mock-infected females) were mated with ZAP+/+Ifnar−/− male mice. Resulting offspring was challenged at 1 d of age with mEV-A71 WT. f,…

• 

  Extended Data Fig. 1. HIV-1 platform for analysis of CpG effects and ZAP-deficient MT4 cells. (a) Details of the insertion of BstEII and ClaI restriction sites at nucleotide positions 6325 and 6771, respectively, in the HIV-1 genome. (b) Summary of the number of CpG dinucleotides present in the HIV-1 reporter virus genome and in the BstEII-ClaI interval in env. (c) Western blot analysis of 293T cells cotransfected with HIV-1 (WT) or HIV-1 (CG-43) and plasmids expressing WT ZAP or the ZAP deletion mutants (Δ3, Δ24 and Δ191) recovered from the CRISPR/Cas9 edited MT4 cells. Blots are representative of 1 experiment. (d) Infectious virus yield from 293T cells cotransfected with HIV-1 (WT) or HIV-1 (CG-43) and plasmids expressing WT ZAP or the ZAP deletion mutants (Δ3, Δ24 and Δ191) recovered from the CRISPR/Cas9 edited MT4 cells. (e) Viral replication in ZAP-expressing and ZAP-knockout MT4 cells infected with wildtype (WT) or CG-43 mutant or a mutant containing the BstEII and ClaI restriction sites with 0 CpGs in the intervening region (CG-0). Each day, the percentage of infected cells was measured by flow cytometry. (f) Summary of the percentage of GFP-positive cells at day 4 post initial infection. Mean…

• 

  Extended Data Fig. 2. Mononucleotide composition of HIV-1 mutants. (a) Summary of mononucleotide composition changes relative to the wildtype HIV-1 in the BstEII-ClaI interval for CG-0/A+/U+/C+/G+ and CG-14/A+/U+/C+/G+ mutants. (b, c) Replication of CG-0/A+/U+/C+/G+ mutant viruses in ZAP-positive and ZAP-negative MT4 cells. The percentage of GFP positive cells was measured daily (b) and the mean and SD of 3 independent experiments measured at day 4 post infection (c). Source data

• 

  Extended Data Fig. 3. Effect of genome position in ZAP antiviral activity. (a) Diagram of the location of the mutated sequence in the env gene. Rev response element (RRE) is located between nucleotides 1560 and 1800. Position coordinates for each mutant is indicated by the location of the first and last CpG dinucleotides. (b) Replication of a collection of HIV-1 mutants (CG-15(110) through CG-15(1330)) each containing 15 CpG dinucleotides at different positions in the env gene as indicated in (a). Numbers in parentheses indicate the nucleotide position within env of the first introduced CpG. Percentage of infected cells was monitored by flow cytometry. (c) Summary of percentage of infected cells 4 days post infection, mean and s.d. of three experiments is plotted. (d) Summary of overall A and U content in the mutated regions of the mutants depicted in (a) (plus 10 nucleotides 5′ and 3′ of the first and last CpG, respectively) and mean number of nucleotides between each CpG. nt, nucleotide. (e) Replication of CG-15(889) recoded to contain high levels of adenine (CG-15(889)/A+). Cells were infected and infection was monitored as in (a). Source data

• 

  Extended Data Fig. 4. Replication and monitoring of CpG reversion during passage of CpG enriched HIV-1. (a) Replication of the CG-43 HIV-1 mutant in ZAP-deficient MT4 cells in three replicate cultures (Rep1-3). Percentage of infected cells was monitored by flow cytometry. (b, c) Sequence analysis of CG-15 HIV-1 mutant during replication. Unique point mutations in each replicate (b) and the total number of CpGs in each replicate at each passage (c) are indicated. Source data

• 

  Extended Data Fig. 5. Replication and monitoring of CpG reversion during passage of CpG enriched EV-A71. (a) EV-A71 RNA levels measured at the end of passage #5, 60 h post-infection in three replicate cultures (mean and s.d. of 3 replicates is plotted). (b) PCR amplification of the modified region of EV-A71/CG-48/A+ in three replicates at the end of a long-term passage in ZAP+/+ and ZAP−/− cells (amplified region is approximately 1 kb). (c–e) Replication of EV-A71 wildtype and mutant reporter viruses in ZAP+/+ and ZAP−/− HeLa cells. NanoLuc luciferase activity was measured every 12 h (c), or quantification of viral RNA at day 4 (d), mean and s.d. plotted; data from 3 replicates. Alternatively replication of EV-A71 wildtype and mutant viruses monitored by TCID50 measurement at each timepoint (e); 3 replicates shown. Source data

• 

  Extended Data Fig. 6. Generation of a ZAP knockout mouse line and ZAP-dependent attenuation of EV-A71 lacking the NanoLuc reporter. (a) Schematic diagram of the ZC3HAV1 locus in the Mus musculus genome. Sequence chromatogram of CRISPR target site PCR product amplified from homozygous ZAP−/− mice. A frameshift mutation in the essential N-terminal RNA binding domain via the insertion of two nucleotides is indicated. (b) Western blot analysis of mouse peripheral blood mononuclear cells (PBMCs) isolated from ZAP+/+IFNAR−/− and ZAP−/−IFNAR−/− mouse lines. (c) Replication of EV71 WT and EV71/CG48/A+ in ZAP+/+ and ZAP−/− HeLa cells, monitored by TCID50 measurement at each timepoint; 3 replicates are shown. Source data

• 

  Extended Data Fig. 7. Clinical score in individual mEV71/CG48/A+infected ZAP+/+ and in ZAP−/− mice. (a, b) Clinical scores of individual mice following infection of one-day old ZAP+/+ (a) and ZAP−/− (b) mice with mEV-A71/CG-48/A+. Source data