Schlafen 11 Restricts Flavivirus Replication

Abstract

Schlafen 11 (Slfn11) is an interferon-stimulated gene that controls the synthesis of proteins by regulating tRNA abundance. Likely through this mechanism, Slfn11 has previously been shown to impair human immunodeficiency virus type 1 (HIV-1) infection and the expression of codon-biased open reading frames. Because replication of positive-sense single-stranded RNA [(+)ssRNA] viruses requires the immediate translation of the incoming viral genome, whereas negative-sense single-stranded RNA [(-)ssRNA] viruses carry at infection an RNA replicase that makes multiple translation-competent copies of the incoming viral genome, we reasoned that (+)ssRNA viruses will be more sensitive to the effect of Slfn11 on protein synthesis than (-)ssRNA viruses. To evaluate this hypothesis, we tested the effects of Slfn11 on the replication of a panel of ssRNA viruses in the human glioblastoma cell line A172, which naturally expresses Slfn11. Depletion of Slfn11 significantly increased the replication of (+)ssRNA viruses from the Flavivirus genus, including West Nile virus (WNV), dengue virus (DENV), and Zika virus (ZIKV), but had no significant effect on the replication of the (-)ssRNA viruses vesicular stomatitis virus (VSV) (Rhabdoviridae family) and Rift Valley fever virus (RVFV) (Phenuiviridae family). Quantification of the ratio of genome-containing viral particles to PFU indicated that Slfn11 impairs WNV infectivity. Intriguingly, Slfn11 prevented WNV-induced downregulation of a subset of tRNAs implicated in the translation of 11.8% of the viral polyprotein. Low-abundance tRNAs might promote optimal protein folding and enhance viral infectivity, as previously reported. In summary, this study demonstrates that Slfn11 restricts flavivirus replication by impairing viral infectivity.IMPORTANCE We provide evidence that the cellular protein Schlafen 11 (Slfn11) impairs replication of flaviviruses, including West Nile virus (WNV), dengue virus (DENV), and Zika virus (ZIKV). However, replication of single-stranded negative RNA viruses was not affected. Specifically, Slfn11 decreases the infectivity of WNV potentially by preventing virus-induced modifications of the host tRNA repertoire that could lead to enhanced viral protein folding. Furthermore, we demonstrate that Slfn11 is not the limiting factor of this novel broad antiviral pathway.

Keywords: Schlafen 11; flavivirus; virus restriction factors.

FIG 1

Kinetics of WNV replication, type I interferon production, and Slfn11 expression in A172 cells. (a) WNV replication in A172 cells. Cells were infected with WNV at an MOI of 0.1, and viral replication was measured by titration of the cell supernatant in a plaque assay at different times postinfection. Titers were determined in triplicate experiments. Data represent the means ± standard errors of the means from three independent experiments. The x axis is not to scale. NI, not infected. (b) Expression of Slfn11 in WNV-infected A172 cells. Cells were lysed at different times postinfection, and Slfn11 and α-tubulin (loading control) were detected with specific antibodies by immunoblotting. Results are representative of data from 2 independent infection experiments. (c) Kinetics of IFN-α and IFN-β (all subtypes) production in WNV-infected A172 cells. The culture supernatant was collected at different times postinfection, and type I IFN was quantified by an ELISA. Data represent the means ± standard errors of the means from three independent experiments. (d). Effect of IFN-α1 on Slfn11 expression. HeLa, A172, and HEK293T cells were treated with 5,000 U/ml of IFN-α1 for 24 h, and the expression levels of the type I IFN-stimulated genes Slfn11 and tetherin were evaluated by immunoblotting.

FIG 2

Effect of Slfn11 on WNV replication. (a) Immunoblot analysis of the expression of Slfn11 in A172 cells stably expressing shRNAs directed against Slfn11 (A172-KD) or scrambled (A172-SCR) RNA sequences and A172-KD cells engineered to reexpress Slfn11 (A172-BC). α-Tubulin was detected as a loading control. (b and c) WNV replication in A172-derived cells. A172-SCR, A172-KD, and A172-BC cells were infected with WNV (MOI of 0.1 [b] and MOI of 1 [c]), and viral replication was determined by quantification of the viral titer in the cell supernatant at different hours postinfection by a plaque assay. Statistically significant differences were calculated with repeated-measures ANOVA and Tukey-Kramer post hoc tests, and they are indicated with asterisks. Mean values and standard deviations indicate the variability of the viral titer found in triplicate plaque assays of samples from 8 independent infection experiments performed on different days with different viral preparations. x axes are not to scale. (d) Expression of WNV E in cells infected at an MOI of 0.1, evaluated by immunoblotting 40 h after infection. α-Tubulin was detected as a loading control. These results are representative of data from 3 independent infections. (e) Expression of WNV E (red) as detected by indirect immunofluorescence analysis of cells infected at an MOI of 1 at 48 h postinfection. Nuclei were labeled with Hoechst dye (blue).

FIG 3

Effect of Slfn11 on viral replication. A172-SCR, A172-KD, and A172-BC cells were infected with dengue virus (DENV) (MOI of 0.1) (a), Zika virus (ZIKV) (MOI of 0.1) (b), vesicular stomatitis virus (VSV) (MOI of 0.1) (c), and Rift Valley fever virus (RVFV) (MOI of 0.1) (d). Viral replication was determined by quantification of the viral titer in the cell supernatant at different hours postinfection by a plaque assay. Statistically significant differences are indicated with asterisks and were calculated as described above. Mean values and standard deviations for each graphic represent the variability of the viral titer found in triplicate plaque assays of samples from 3 independent infection experiments performed on different days with different viral preparations. x axes are not to scale.

FIG 4

Mutagenesis analysis of antiviral activity of Slfn11. (a) Immunoblot analysis of the expression of Slfn11 in A172-derived cells. A172-KD cells were engineered to express the N or C terminus of Slfn11. A172-SCR, -KD, and -BC cells were used as controls. Different anti-Slfn11 antibodies were employed to identify the mutant proteins. α-Tubulin was detected as a loading control. (b) A172-BC, A172-N-term, and A172-C-term cells were infected with WNV (MOI of 0.1), and viral replication was determined by quantification of the viral titer in the cell supernatant at different hours postinfection. Statistically significant differences were calculated as described in the Fig. 2 legend and are indicated with asterisks. Mean values and standard deviations represent the variability of the viral titer found in triplicate plaque assays of samples from 3 independent infection experiments. The x axis is not to scale. (c) Cellular distribution of full-length Slfn11 and deletion mutants. A172-BC, A172-N-term, and A172-C-term cells were fixed/permeabilized and stained with the anti-Slfn11 antibodies used in panel a. Cell nuclei were identified with Hoechst staining (blue staining).

FIG 5

Effect of WNV infection on the tRNA pool of Slfn11-deficient and control cells. (a) Expression levels of IFN-β (all subtypes) in WNV-infected A172 cells. The culture supernatant was collected at 8 h and 48 h postinfection, and type I IFN was quantified by an ELISA. Data represent the means ± standard errors of the means from three independent infection experiments. (b) Heat map of fold changes of tRNAs up- or downregulated beyond the significance threshold. Each cell represents the tRNA ratio of infected/noninfected cells for each of the three cell lines evaluated. tRNAs that are increased (red cells), decreased (green cells), or unchanged (black cells) are indicated. The cutoff was set at a −1.89- to 1.89-fold change.

FIG 6

Evaluation of the antiviral activity of Slfn11 in HEK293T, HeLa, and BHK-21 cells. (a) Slfn11 expression in HEK293T, HEK293, HeLa, and BHK-21 parental and derived cell lines detected by immunoblot analysis. WT, wild type. (b to e) WNV replication in HEK293T^Slfn11 (b), HEK293-KD (c), HeLa^Slfn11 (d), BHK-21^Slfn11 (e), and parental cell lines. Parental (open diamonds) and Slfn11-expressing derivative (filled triangles) cell lines as well as HEK293 and HEK293-KD cells were infected with WNV (MOI of 0.1), and viral replication was determined by quantification of the viral titer in the cell supernatant at different hours postinfection by a plaque assay. Statistically significant differences were calculated as described in the Fig. 2 legend and are indicated with asterisks. Mean values and standard deviations represent the variability of the viral titers found in triplicate plaque assays of samples from 3 independent infection experiments. (f) HIV-1 infection of cells expressing or not expressing Slfn11. Cells were infected with replication-defective HIV-1. Mean values and standard deviations represent the variability of HIV-1 p24 levels found in 3 independent infection experiments. Statistical analysis was performed by a t test (HEK293 and HeLa cells) and ANOVA with a Tukey honestly significant difference (HSD) post hoc test (A172). ** indicates a P value of <0.01. (g) Cellular distribution of Slfn11 in HEK293T^Slfn11 cells. Cells were fixed/permeabilized and stained with an anti-Slfn11 antibody by indirect immunofluorescence (red). Nuclei were labeled with Hoechst dye (blue).