Fig 1.
NK cells enhance innate IFN-β response in cytomegalovirus-infected cells.
(a) HFFs were infected with GFP-HCMV-WT (AD169 [blue] or Merlin [orange], MOI = 0.2) for 2h and then maintained in medium for 18h. IFN-β mRNA levels were measured by qPCR. The value in the absence of virus was set to 1. Data is displayed as mean fold change ± SEM, n ≥ 3. (b-c) HFFs were infected with GFP-HCMV-WT (AD169 or Merlin, MOI = 0.2) for 2h and treated with NK cells (YT-Indy) at indicated E:T ratios for 18h. IFN-β mRNA levels were measured by qPCR. The value in the absence of NK cells was set to 1 to study the effect of NK cell treatment between the two virus strains. Data is displayed as mean fold change ± SEM, n ≥ 3. (d) HFFs were infected with GFP-HCMV-WT and co-cultured with NK cells at indicated E:T ratios for 18h. Fibroblasts were collected for immunoblotting for p-IRF3 and infection control (GFP). A representative blot is shown, n = 3. (e) HFFs were infected with GFP-HCMV-WT (AD169, MOI = 0.2) for 2h and co-cultured with NK cells (YT-Indy) at indicated E:T ratio for 18h. Supernatant was collected and tested for IFN-β protein by ELISA. Data is displayed as mean IFN-β concentration ± SD, n ≥ 3. (f-g) HFFs were infected with GFP-HCMV-WT (AD169 or Merlin, MOI = 0.2) for 2h and treated with primary NK cells at indicated E:T ratios for 18h. IFN-β mRNA levels were measured by qPCR. The value in the absence of primary NK cells was set to 1. Data is displayed as mean fold change ± SEM, n ≥ 3. (h) HeLa cells were transfected with cGAS and STING plasmids, together with pGL4.24 10 × ISRE reporter (100 ng), internal control hRL-EF1a (10 ng), and H2B-GFP (20 ng), in the presence or absence of pp71FL (410 ng), pp71FL expression is shown in the Western blot with loading control (GAPDH). 6h post-transfection, cells were treated with different amounts of NK cells (YT-Indy) for 18h. A dual luciferase reporter assay was used to measure ISRE activation. Relative luciferase activity, i.e., firefly/renilla, in the absence of pp71 and NK cells was set to 100% and data were corrected for effects of NK cells in the absence of pp71. Data represent the mean ± SD, n = 3. (i) HFFs were infected with GFP-HCMV-WT (AD169, MOI = 0.2) for 24h and co-cultured with NK cells (YT-Indy) at indicated E:T ratios for 18h. Fibroblasts were lysed in reducing Laemmli buffer and subjected to immunoblotting for pp71FL and loading control (GAPDH). A representative blot is shown, n = 3. One-way ANOVA (a-e) was used for statistical analysis, * = p < 0.05, ** = p < 0.01, *** = p < 0.001, **** = p < 0.0001. E:T, effector:target; HCMV, human cytomegalovirus.
Fig 2.
NK cells restore cGAS-STING signaling and IFN-β response via GrM-mediated cleavage of pp71 and signaling through TNFR1 and LTβR.
(a) HFFs were infected with GFP-HCMV-WT or GFP-HCMV-pp71L439A mutant (AD169, MOI = 0.2) for 2h. Viral load at indicated days post infection was measured by qPCR. Data represent the mean ± SD, n > 3. (b-c) HEK293T cells were transfected with pGL4.24 10 × ISRE (100 ng), hRL-EF1a (10 ng), c-Myc-cGAS (25 ng), Flag-STING (25 ng), H2B-GFP (30 ng), and increasing amounts of pp71L439A (b) or pp71-WT (c). 24h post-transfection, cells were lysed for dual luciferase reporter assay. Relative luciferase activity, i.e., firefly/renilla, is depicted relative to the condition without pp71 (100%). Lysates were subjected to immunoblotting, using antibodies against HA, p-IRF3, IRF3, and loading control (Hsp90). Data represent the mean ± SD, n = 3. (d) HFFs were infected with GFP-HCMV-WT (light blue) or GFP-HCMV-pp71L439A (dark blue) mutant (AD169, MOI = 0.2) and co-cultured with different amounts of NK cells (YT-Indy) for 18h. IFN-β mRNA was determined by qPCR. The value in the absence of NK cells was set to 1. Data represent the mean ± SEM, n = 3. (e) GrM-KO validation on flow cytometry and western blot with loading control (GAPDH). (f) HFFs were infected with GFP-HCMV-WT (AD169, MOI = 0.2) and co-cultured with different amounts of NK cells (WT (light blue) or GrM-KO (green) YT-Indy) for 18h. IFN-β mRNA was determined by qPCR. The value in the absence of virus and NK cells was set to 1. Data represent the mean ± SD, n = 3. (g) HFFs were infected with GFP-HCMV-WT (AD169, MOI = 0.2) and co-cultured with WT and GrM-KO NK cells (at E:T = 4:1) for 18h. IFN-β mRNA was determined by qPCR, WT NK cell IFN-β response was set at 100%. Data represent the mean ± SEM, n ≥ 3. (h) HFFs were infected with GFP-HCMV-WT (Merlin, MOI = 0.2) and co-cultured with WT and GrM-KO NK cells (at E:T = 8:1) for 18h. Cells were immunoblotted for pp71FL (and cleavage fragment) and loading control (GAPDH). Data shows representative blot, n = 3. (i) Perforin-KO validation on flow cytometry. (j) HFFs were infected with GFP-HCMV-WT (AD169, MOI = 0.2) and co-cultured with WT and GrM-KO NK cells (at E:T = 4:1) for 18h. IFN-β mRNA was determined by qPCR, WT NK cell IFN-β response was set at 100%. Data represent the mean ± SEM, n ≥ 3. (k) HFFs were infected with GFP-HCMV-WT (AD169, MOI = 0.2) and co-cultured with WT or Perforin-KO NK cells with or without 2h TNFR1-Fc and LTβR-Fc preincubation at E:T = 4:1 for 18h. IFN-β mRNA was determined by qPCR, WT NK cell IFN-β response without decoy proteins was set at 100%. Data represent the mean ± SEM, n ≥ 3. One-way ANOVA (b-d,f,k) or student t-test (g,j) was used for statistical analysis, * = p < 0.05, ** = p < 0.01, *** = p < 0.001, **** = p < 0.0001. Ctrl, control; E:T, effector:target; HCMV, human cytomegalovirus.
Fig 3.
pp71FL and pp71S inhibit while pp71L activates IFN-β production.
(a) Schematic overview of GrM-mediated cleavage of pp71FL into pp71L and pp71S. (b) Schematic overview of the ISRE and IFN-β promoter reporter constructs. (c) HeLa cells were transfected with ISRE reporter (100 ng), hRL-EF1a (10 ng), H2B-GFP (30 ng), and indicated plasmids. The relative luciferase activity in the condition of both cGAS and STING was defined as 100%. Data represent the mean ± SD, n = 3. (d-f,k) HeLa cells were transfected with ISRE reporter (100 ng), hRL-EF1a (10 ng), cGAS (25 ng), STING (25 ng), H2B-GFP (30 ng), and increasing amounts of pp71FL, pp71L and/or pp71S. Faint pp71S bands are a result of quick proteasomal degradation (see Fig 4f). Data represent the mean ± SD, n = 3. (g) HEK293T cells were transfected with IFN-β reporter (100 ng), hRL-EF1a (10 ng), H2B-GFP (30 ng) and indicated plasmids. The relative luciferase activity in the condition of both cGAS and STING was defined as 100%. Data represent the mean ± SD, n = 3. (h-j) HEK293T cells were transfected with IFN-β reporter (100 ng), hRL-EF1a (10 ng), cGAS (25 ng), STING (25 ng), H2B-GFP (30 ng), and increasing amounts of pp71FL, pp71L or pp71S. Faint pp71S bands are a result of quick proteasomal degradation (see Fig 4f). Data represent the mean ± SD, n = 3. Dual luciferase reporter assays were performed 24h after transfection. Lysates used in the luciferase reporter assay were subjected to immunoblotting using an anti-HA antibody. One-way ANOVA (c-k) was used for statistical analysis, * = p < 0.05, ** = p < 0.01, *** = p < 0.001, **** = p < 0.0001. FF, firefly; MinP, minimal promoter; pp71FL, pp71 full length; pp71L, pp71 long cleavage fragment; pp71S, pp71 short cleavage fragment.
Fig 4.
pp71L reverses pp71FL function while pp71S is rapidly degraded by the proteasome.
(a) HEK293T cells were transfected with Flag-STING and HA-pp71FL/L/S for 24h. Cell lysates were incubated with anti-HA Sepharose beads and analyzed by immunoblotting with the indicated antibodies. (b) HEK293T cells were transfected with cGAS (400 ng), STING (400 ng), and pp71FL/L (400 ng) for 24h. Samples were subjected to immunoblotting using antibodies against pp71FL/L, flag-STING, p-IRF3, p-TBK-1, and loading control (HSP90). Densometry analysis was performed for p-IRF3 and p-TBK-1 in ImageJ and corrected for loading control. Data shows a representative immunoblot, n = 3 (c-e) HeLa cells were transfected with pGL4.24 10 × ISRE reporter (100 ng), hRL-EF1a (10 ng), cGAS (25 ng), STING (25 ng), H2B-GFP (30 ng), pp71FL (205 ng), and increasing amounts of pp71L and/or pp71S. Dual luciferase reporter assays were performed 24h after transfection. Lysates used in the luciferase reporter assay were subjected to immunoblotting using an anti-HA antibody. Data represent the mean ± SD, n = 3. (f) HEK293T cells were transfected with pp71FL, pp71L, pp71S (each 410 ng) and H2B-GFP (30 ng). 24h post-transfection, CHX (20 µM) was added for indicated time points. Samples were subjected to immunoblotting using antibodies against HA and loading/infection control (GFP). Data shows a representative immunoblot, n = 3 (g) HEK293T cells were transfected with pp71S and H2B-GFP. 24h after transfection, proteasome inhibitor MG132 (20 µM) and CHX (20 µM) were added to the cells for 4h. Cells were subjected to immunoblotting using anti-HA, loading/infection control (GFP), and ubiquitin antibodies. Data shows a representative immunoblot, n = 3. One-way ANOVA (c-e) was used for statistical analysis, * = p < 0.05, ** = p < 0.01, *** = p < 0.001, **** = p < 0.0001. Ctrl, control; pp71FL, pp71 full length; pp71L, pp71 long cleavage fragment; pp71S, pp71 short cleavage fragment.
Fig 5.
Graphical Abstract.