Fig 1.
A loss-of-function RNAi screen uncovers many genes that significantly reduce the antiviral activity of ZAP when silenced.
(A) The experimental outline of the genome-wide siRNA screen is shown. T-REx-hZAP cells transfected with control or gene-specific siRNA were treated with doxycycline to induce ZAPS overexpression one day post-transfection and infected with SINV Toto1101/Luc two days post-transfection. Cell lysates were harvested for measurement of luciferase activity at 24 h post-infection (p.i.). Relative luciferase units represent the level of SINV replication. Cells treated with the control non-targeting (NT) pooled siRNA have low SINV replication while ZAP knockdown by ZC3HAV1-specific pooled siRNA rescues viral replication by 2 logs. The large dynamic range in which hypothetical hits (ZAP cofactors) were identified is plotted on the right side of the graph. (B) Pooled siRNAs targeting the entire human genome (Dharmacon) were tested in triplicate and genes with an average robust Z score of greater than 3 are plotted. ZC3HAV1 is highlighted in red while the top hits immediately following ZC3HAV1 are highlighted in blue.
Fig 2.
Secondary screen confirms the ZAP-dependent and -independent antiviral effects of hits.
A customized library of individual siRNAs (Ambion) targeting 102 hits from the primary screen was tested in triplicate in two different cell lines. Distribution of average Z scores for each of the 3 siRNAs targeting a candidate gene in the secondary screen is shown here. The screen was performed with (A) 6.25 nM or (B) 25 nM individual siRNAs in 293 cells induced to express ZAPS (T-REx-hZAP), and with (C) 6.25 nM or (D) 25 nM individual siRNAs in 293 cells induced to express the rZAPC88R dominant negative mutant (T-REx-rZAPC88R). Each dot represents the average Z score of an individual siRNA tested in triplicate. Silenced genes with an average Z score of >3 for at least 2 out of 3 siRNAs are identified and the siRNAs are labeled in color. (C and D) The average Z scores of TRIM25 and ZC3HAV1 are also plotted to indicate that TRIM25 does not have ZAP-independent antiviral effects.
Fig 3.
TRIM25 synergizes with ZAP to block SINV replication.
(A) Candidate genes that significantly increased SINV replication in the secondary screen when silenced by individual siRNAs at both concentrations were validated in a larger scale 24-well plate format. Triplicate wells of T-REx-hZAP cells were transfected with the indicated siRNA, induced to express ZAPS, and infected with SINV Toto1101/Luc at a MOI of 10. Each symbol represents the value obtained from a single well after 24 h of infection. White circles represent results using pooled siRNA controls that were either NT or ZC3HAV1-specific. The data is representative of 3 independent experiments. Asterisks indicate statistically significant differences (Student’s t-test, **, p<0.005; ***, p<0.0005; ****, p<0.0001). (B) Triplicate wells of T-REx-hZAP cells were transfected with the indicated siRNA, induced to express ZAPS, and infected with SINV Toto1101/Luc at a MOI of 10. Protein expression levels of TRIM25 and ZAP for the same transfections in a duplicate well were determined by immunoblotting. β-actin was used as a loading control. The data is representative of 4 independent experiments. The p-value from Student’s t-test is shown. (C) SINV replication in infected 293T cells in which ZC3HAV1 (left) or TRIM25 (middle) were silenced, and in ZC3HAV1-null 293T cells in which TRIM25 was silenced (right) is plotted. At 48 h post-transfection with siRNA, cells were infected with SINV Toto1101/Luc at a MOI of 0.01, and lysed at 6, 12, 24, and 40 h p.i. for measurement of luciferase activity. The data is representative of 3 independent experiments. Asterisks indicate statistically significant differences (two-way ANOVA, *, p<0.05; **, p<0.01; ****, p<0.0001).
Fig 4.
The SPRY domain of TRIM25 associates with ZAP.
(A) 293T cells were infected with the SINV strain Toto1101 (MOI = 10), and lysates were harvested at 0, 6, 12 and 24 h p.i. for co-immunoprecipitation with an anti-TRIM25 antibody and immunoblotting. The data is representative of 2 independent experiments. (B) Human ZAP short (S) and long (L) isoforms are shown, with the CCCH fingers, TIPARP, WWE and PARP-like domains indicated. Light green shading indicates sequences shared by the two isoforms whereas the hatched region containing the PARP-like domain is unique to the L isoform. WCL of ZC3HAV1-knockout 293T cells transfected with V5-tagged TRIM25 and/or ZAPS or ZAPL were used for co-immunoprecipitation with an anti-NZAP antibody and immunoblotting. (C) Full-length (FL) TRIM25 is shown, with the RING, B box, CCD, and SPRY domains indicated. WCL of ZC3HAV1-knockout 293T cells transfected with V5-tagged FL or truncated TRIM25 (RING, B box/CCD, SPRY only) and ZAPL were used for co-immunoprecipitation with an anti-NZAP antibody and immunoblotting. Different amounts of TRIM25 domain-expressing constructs were used for transfection in order to achieve similar RING, B box/CCD, and SPRY expression. The asterisk indicates a non-specific band. (B and C) The data is representative of 3 independent experiments.
Fig 5.
CRISPR targeting of TRIM25 leads to increased virus replication and both the RING and CCD domains of TRIM25 are required for ZAP activation.
(A) Wild type (clone E) and TRIM25lo ZC3HAV1-knockout 293T cells (clones D and F) were transfected with empty vector or vector expressing ZAPS or ZAPL and infected with Toto1101/Luc (MOI = 0.01) 2 days post-transfection. (B) TRIM25lo ZC3HAV1-knockout 293T cells (clones D and F) were reconstituted with expression of FL or truncated TRIM25 (ΔRING, ΔCCD) and/or ZAPS or ZAPL, and infected with Toto1101/Luc (MOI = 10) 2 days post-transfection. (A and B) The data is representative of 2 independent experiments performed on both clones D and F. Cell lysates were harvested for measurement of luciferase activity at 24 h p.i. Relative luciferase units represent the level of SINV replication. Asterisks indicate statistically significant differences (Student’s t-test, *, p<0.05; **, p<0.005; ***, p<0.0005).
Fig 6.
Both ZAPS and ZAPL are ubiquitinated by TRIM25.
Cells were lysed in denaturing conditions to ensure pulldown of ZAP only and not ZAP-associated proteins. (A) WCL of 293T cells transfected with vector expressing HA-tagged ubiquitin (Ub), and mock infected or infected with the SINV Toto1101 strain (MOI = 1) for 18 hours were used for immunoprecipitation of endogenous ZAP with an anti-ZAP antibody and immunoblotting. The level of HA-tagged Ub in the ZAP pulldown is shown. The data is representative of 3 independent experiments. (B) WCL of ZC3HAV1-knockout 293T cells transfected with vectors expressing HA-tagged Ub, and ZAPS or ZAPL were used for immunoprecipitation of overexpressed ZAP with an anti-ZAP antibody and immunoblotting. The level of HA-tagged Ub in the ZAP pulldown is shown. The data is representative of 3 independent experiments. (C) WCL of wild type and TRIM25lo ZC3HAV1-knockout 293T cells transfected with vectors expressing HA-tagged Ub, and ZAPS or ZAPL were used for immunoprecipitation with an anti-ZAP antibody and immunoblotting. The data is representative of 2 independent experiments performed on both clones D and F. Only data for clone D is shown here. (D) WCL of ZC3HAV1-knockout 293T cells transfected with vectors expressing ZAPS or ZAPL, and/or V5-tagged TRIM25 were used for immunoprecipitation with an anti-ZAP antibody and immunoblotting. The level of endogenous Ub in the ZAP pulldown is shown. The data is representative of 3 independent experiments. (E) WCL of ZC3HAV1-knockout 293T cells transfected with vector expressing HA-tagged Ub, ZAPS or ZAPL, and/or V5-tagged TRIM25 were used for immunoprecipitation with an anti-ZAP antibody and immunoblotting. The level of HA-tagged Ub in the ZAP pulldown is shown. The data is representative of 2 independent experiments. (F) WCL of ZC3HAV1-knockout 293T cells transfected with vector expressing HA-tagged wild type (WT), K48 or K63 Ub, ZAPS or ZAPL, and/or V5-tagged TRIM25 were used for immunoprecipitation with an anti-ZAP antibody and immunoblotting. The level of HA-tagged WT or mutant Ub in the ZAP pulldown is shown. The data is representative of 2 independent experiments.
Fig 7.
ZAP synergizes with TRIM25 to block SINV translation.
T-REx-hZAP cells were transfected with TRIM25-targeting or NT pool siRNA, induced for ZAPS expression, and infected with a temperature-sensitive SINV that expresses luciferase and is unable to replicate at 40°C (Toto1101/Luc:ts6; MOI = 10). Following 1 hour of adsorption with virus at 37°C, cells were moved to 40°C, washed, and lysed at 0, 2, 4 and 6 h p.i. for measurement of (A) viral RNA by RT-qPCR, and (B) luciferase activity, which represents translation of the incoming viral genome. The data is representative of 2 independent experiments.