Exosomes originating from infection with the cytoplasmic single-stranded RNA virus Rift Valley fever virus (RVFV) protect recipient cells by inducing RIG-I mediated IFN-B response that leads to activation of autophagy

Abstract

Background: Although multiple studies have demonstrated a role for exosomes during virus infections, our understanding of the mechanisms by which exosome exchange regulates immune response during viral infections and affects viral pathogenesis is still in its infancy. In particular, very little is known for cytoplasmic single-stranded RNA viruses such as SARS-CoV-2 and Rift Valley fever virus (RVFV). We have used RVFV infection as a model for cytoplasmic single-stranded RNA viruses to address this gap in knowledge. RVFV is a highly pathogenic agent that causes RVF, a zoonotic disease for which no effective therapeutic or approved human vaccine exist.

Results: We show here that exosomes released from cells infected with RVFV (designated as EXi-RVFV) serve a protective role for the host and provide a mechanistic model for these effects. Our results show that treatment of both naïve immune cells (U937 monocytes) and naïve non-immune cells (HSAECs) with EXi-RVFV induces a strong RIG-I dependent activation of IFN-B. We also demonstrate that this strong anti-viral response leads to activation of autophagy in treated cells and correlates with resistance to subsequent viral infection. Since we have shown that viral RNA genome is associated with EXi-RVFV, RIG-I activation might be mediated by the presence of packaged viral RNA sequences.

Conclusions: Using RVFV infection as a model for cytoplasmic single-stranded RNA viruses, our results show a novel mechanism of host protection by exosomes released from infected cells (EXi) whereby the EXi activate RIG-I to induce IFN-dependent activation of autophagy in naïve recipient cells including monocytes. Because monocytes serve as reservoirs for RVFV replication, this EXi-RVFV-induced activation of autophagy in monocytes may work to slow down or halt viral dissemination in the infected organism. These findings offer novel mechanistic insights that may aid in future development of effective vaccines or therapeutics, and that may be applicable for a better molecular understanding of how exosome release regulates innate immune response to other cytoplasmic single-stranded RNA viruses.

Keywords: Autophagy; Exosome; IFN-B; Innate immune response; RIG-I; Rift Valley fever virus; SARS-CoV-2; Single-stranded RNA virus; Viral RNA; Virus infection.

Fig. 1

Characterization of Sucrose Gradient-Purified Exosomes. A Exosomes from both RVFV-infected cells and uninfected cells were harvested in parallel using the procedure describe in the Materials and Methods section. The sucrose density fractions were initially analyzed by western blot for the exosome marker CD63 (top). CD63-positive fractions with densities 1.08 g ml^-1 and 1.12 g ml^-1 were combined and further analyzed by western blot for additional exosome markers TSG101 (middle) and Flotillin-1 (bottom); three independent preparations were tested and are denoted on the figure as “Sample A”, “Sample B”, and “Sample C”. WCL refers to U937 whole cell lysate used as positive control. At least 25 biological replicates of this purification scheme have been performed and analyzed; the data presented here are representative of the findings. B Mean values ± SEM of plaque assay results from three biologically independent EXi-RVFV preparations are shown, quantifying the virions present in the starting material and following the ultracentrifugation spins and sucrose gradient separation. Plaque assays were performed for every single purified exosome preparation throughout our studies, totaling at least 15 biological replicates. A summary of plaque assay results for all the fractions is also presented. The red box highlights the two fractions that were combined to generate EXi-RVFV used in our studies (1.08 g ml-1 and 1.12 g ml-1 fractions). Circles with blue color in the middle represent exosomes and circles with yellow color in the middle represent virions. C TEM analysis of purified exosomes is shown. D ZetaView analyses of the mean diameter of the EXu and EXi-RVFV populations are presented. Three biological replicates were analyzed. E ZetaView measurements of the surface charge (Zeta Potential) for EXu and EXi-RVFV samples are presented. Mean values ± SEM from three biological replicates are shown. F ZetaView measurements of the concentration of EXu and EXi-RVFV samples that were matched for total protein content based on BCA are presented. Mean values ± SEM from three biological replicates are shown

Fig. 2

EXi-RVFV Contain Viral RNA and Select Viral Proteins. A Total RNA extracts from the EXu and EXi-RVFV samples were subjected to RT-qPCR analysis using RVFV-specific primer pairs. Absolute quantitation of the samples was performed using the cycle threshold (Ct) value relative to the standard curve. Mean values ± SEM from three biological replicates are shown. B The results of mass spectrometry analysis of EXi-RVFV to identify associated viral proteins are shown. EXu samples were included as negative control. Mass spectral counts for each identified viral protein is reported. For each sample (EXu or EXi-RVFV), two biological replicates were analyzed. C EXi-RVFV samples were analyzed by western blot for the presence of the viral L, NSs, N, and G proteins. WLC corresponds to whole cell lysate preparations from RVFV-infected Vero cells that were included as positive control. Two biological repeat samples (1 and 2) were analyzed

Fig. 3

Pre-treatment of Naïve Recipient Human Lung Epithelial Cells with EXi-RVFV Significantly Reduces Viral Replication in the Cells during Subsequent Infection. A Naïve HSAECs were treated with either EXu or EXi-RVFV and cell viability was measured using acridine orange/propidium iodide (AO/PI) stain at 24 and 48 h. Mean values ± SEM from three biological replicates are shown. B Viral loads in naïve HSAECs that had been pretreated with either EXu or EXi-RVFV and subsequently infected with RVFV were compared based on genetic copy equivalent (RT-qPCR) measurements. Both the 6 h p.i. and the 24 h p.i. time points were analyzed. For each treatment condition and time point assayed, mean values ± SEM from three biological replicates are shown. **P ≤ 0.01; ***P ≤ 0.005; ****P ≤ 0.0005

Fig. 4

Pre-treatment of Naïve Recipient Human Monocytes with EXi-RVFV Significantly Reduces Viral Replication in the Cells during Subsequent Infection. A Naïve U937 cells were treated with either EXu or EXi-RVFV for 24 or 48 h, and cell counts and viability reads were subsequently performed using acridine orange/propidium iodide (AO/PI) staining. For each treatment condition and time point assayed, mean values ± SEM from three biological replicates are shown. B Plaque assays were performed to measure viral load in naïve U937 cells that had been pretreated with either EXu or EXi-RVFV and subsequently infected with RVFV. Both the 6 h p.i. and the 24 h pi. time points were analyzed. For each treatment condition and time point assayed, mean values ± SEM from three biological replicates are shown. C Plaque assays were performed to measure viral uptake by U937 cells post treatment with either EXu or EXi-RVFV for 24 h and subsequent infection with RVFV. Untreated cells were also included as control. At 1 h p.i., the cells were lysed and intracellular viral loads were determined by plaque assay. For cell lysis, three freeze–thaw cycles were performed in an ethanol-dry ice bath to release the intracellular virus, and quantitative lysis of the cells was verified through microscopic visualization of the sample. Mean values ± SEM from three biological replicates are shown. **P ≤ 0.01; ***P ≤ 0.005

Fig. 5

EXi-RVFV Induce Interferon-B in Naïve Recipient Cells. Naïve U937 cells (A), or naïve HSAECs (B), were either left untreated, or were treated with EXu or EXi-RVFV, and intracellular IFN-B expression was analyzed at 24 h post treatment by western blot analysis of whole cell lysates. Infection with the ΔNSs mutant derivative of the MP12 strain of RVFV was included as positive control, along with infection with the MP12 strain. For each treatment condition, mean values ± SEM from three biological replicates are shown. ***P ≤ 0.005; ****P ≤ 0.0005

Fig. 6

EXi-RVFV Depletion Eliminates Interferon-B Induction in Naïve Recipient Cells. A The sucrose density fractions containing EXi-RVFV, the most bottom fraction containing virus (density of 1.25 g ml⁻¹), and the working stock of MP12 strain (1.97 × 10⁷ pfu/ml), were analyzed by western blot for the exosome marker CD9. B EXi-RVFV samples were subjected to immunoprecipitation using magnetic Dynabeads coated with human anti-CD9 antibody. The efficacy of pull-down was analyzed by anti-CD9 immunoblot of the starting material (Input EXi-RVFV), the immunoprecipitated EXi-RVFV, and the EXi-depleted supernatant that was left behind (ΔEXi). Three biological repeats were performed. A typical western result is shown. C Naïve U937 cells were either left untreated, or were subjected to one of the following treatments: i) Poly(I:C); ii) EXu; iii) EXi-RVFV; iv) ΔEXi. The IFN-B expression levels in treated cells were then quantified by RT-qPCR analysis. For each treatment condition, mean value ± SEM from three biological replicates is shown (***P ≤ 0.005). D Anti-IFN-B immunoblot analysis was performed to compare IFN-B induction in naïve U937 cells treated with either EXi-RVFV or samples depleted of EXi-RVFV by anti-CD9 IP pulldown (ΔEXi). The intracellular IFN-B expression was analyzed at 24 h post treatment. Untreated cells, cells treated with EXu, or cells infected with either the MP12 strain or the ΔNSs mutant strain of RVFV, were also analyzed side by side. For each treatment condition, mean values ± SEM from three biological replicates are shown in the density bar graph at the bottom. ****P ≤ 0.0005. E Western blot analyses of the viral G protein and the viral N protein were performed for immunoprecipitated EXi-RVFV obtained by CD9 pulldown and the resulting supernatant depleted of exosomes (ΔEXi)

Fig. 7

EXi-RVFV Induction of Interferon-B Is Mediated by RIG-I Activation and Is Impervious to RNase Treatment of the Exosomes. A U937 cells were transfected with RIG-I siRNA (Santa Cruz) and RIG-I expression was measured at 24 h p.i. by western blot analysis. Cells that were either left untransfected or transfected with control scrambled siRNA (Santa Cruz) were also included as controls. B U937 cells were transfected with either a scrambled control siRNA (Santa Cruz) or with RIG-I siRNA (Santa Cruz) 24 h prior to any further treatment. Subsequently, the cells were split to generate the following treatment conditions: (i) Uninfected (control); (ii) Infection with the MP12-L-V5, NSs-Flag strain of RVFV; (iii) Infection with the ΔNSs derivative of MP12 strain (positive control); (iv) Treatment with EXu without any infection; (v) Treatment with EXi-RVFV without any infection. Whole cell lysates were prepared 24 h post treatment or post infection, and analyzed by western blot for IFN-B expression. C Naïve U937 cells were either left untreated or were treated with EXu, or EXi-RVFV, or RNase-treated EXI-RVFV (EXi-RVFV + RNase), and intracellular IFN-B expression was analyzed at 24 h post treatment by western blot analysis of whole cell lysates. For each treatment condition, mean values ± SEM from three biological replicates are shown. ***P ≤ 0.005; ****P ≤ 0.0005

Fig. 8

Treatment of Naïve Recipient Human Monocytes with EXi-RVFV Activates Autophagy through the Induction of IFN-B. A Naïve U937 cells were either left untreated, or were treated with EXi-RVFV or EXu for 24 h, and whole cell lysates matched for total protein levels were subjected to western blot analysis for LC3. Mean values ± SEM for three biological replicates is shown. B Immunofluorescence microscopy analysis of LC3 expression was performed for U937 cells that were either left untreated, or were treated with EXi-RVFV or EXu for 24 h, followed by LC3 antibody staining. Green signal represents LC3 and blue signal corresponds to DAPI staining of the nuclei. C Naïve U937 cells were either left untreated, or were treated with EXi-RVFV or EXu for 24 h, and whole cell lysates matched for total protein levels were subjected to western blot analysis to measure p62 levels. Mean values ± SEM from two biological replicates is shown. D Naïve U937 cell were pre-treated with anti-IFNAR2 antibody for 4 h and subsequently were treated with EXi-RVFV for 24 h. Naïve U937 cells without anti-IFNAR1/2 antibody treatment but treated with either EXi-RVFV or EXu were also included to serve as control conditions. Western blot was performed for LC3-II analysis. Mean values ± SEM from two biological replicates are shown. *P ≤ 0.03; **P ≤ 0.01; ***P ≤ 0.005; ****P ≤ 0.0005

Fig. 9

Proposed Model of Exosomal Function During RVFV Infection. EXi-RVFV released by infected cells travel to local and distant target naïve cells. Following entry, viral RNA genome carried by EXi-RVFV are released first, and specific viral RNA sequence(s) that form appropriate dsRNA panhandle structures induce RIG-I. The activation of RIG-I leads to induction of IFN-B that is subsequently released to elicit both autocrine and paracrine IFN responses that involves activation of autophagy. Increased autophagy leads to encapsulation of virus particles that could reach and enter the cell at a later stage, eliminating them through autolysosomes to result in a significant decrease in viral replication, and therefore propagation