(C) PLOS One [1]. This unaltered content originally appeared in journals.plosone.org.
Licensed under Creative Commons Attribution (CC BY) license.
url:https://journals.plos.org/plosone/s/licenses-and-copyright

------------



A yellow fever virus NS4B inhibitor not only suppresses viral replication, but also enhances the virus activation of RIG-I-like receptor-mediated innate immune response

['Zhao Gao', 'Baruch S. Blumberg Institute', 'Doylestown', 'Pennsylvania', 'United States Of America', 'Xuexiang Zhang', 'Lin Zhang', 'Shuo Wu', 'Julia Ma', 'Fuxuan Wang']

Date: 2022-02 

Flavivirus infection of cells induces massive rearrangements of the endoplasmic reticulum (ER) membrane to form viral replication organelles (ROs) which segregates viral RNA replication intermediates from the cytoplasmic RNA sensors. Among other viral nonstructural (NS) proteins, available evidence suggests for a prominent role of NS4B, an ER membrane protein with multiple transmembrane domains, in the formation of ROs and the evasion of the innate immune response. We previously reported a benzodiazepine compound, BDAA, which specifically inhibited yellow fever virus (YFV) replication in cultured cells and in vivo in hamsters, with resistant mutation mapped to P219 of NS4B protein. In the following mechanistic studies, we found that BDAA specifically enhances YFV induced inflammatory cytokine response in association with the induction of dramatic structural alteration of ROs and exposure of double-stranded RNA (dsRNA) in virus-infected cells. Interestingly, the BDAA-enhanced cytokine response in YFV-infected cells is attenuated in RIG-I or MAD5 knockout cells and completely abolished in MAVS knockout cells. However, BDAA inhibited YFV replication at a similar extent in the parent cells and cells deficient of RIG-I, MDA5 or MAVS. These results thus provided multiple lines of biological evidence to support a model that BDAA interaction with NS4B may impair the integrity of YFV ROs, which not only inhibits viral RNA replication, but also promotes the release of viral RNA from ROs, which consequentially activates RIG-I and MDA5. Although the innate immune enhancement activity of BDAA is not required for its antiviral activity in cultured cells, its dual antiviral mechanism is unique among all the reported antiviral agents thus far and warrants further investigation in animal models in future.

Emergence and re-emergence of yellow fever (YF) caused by the yellow fever virus (YFV) infection have posed a global public health threat in previously non-epidemic as well as endemic regions. The approximately 30% of mortality rate makes the outbreaks particularly devastating. In addition to the vaccination campaign and mosquito controls, antiviral drugs are important components in the toolbox for combating YF outbreaks. However, only two nucleotide analogue drugs developed for the treatment of other RNA virus infections are currently repurposed for the treatment of YF with uncertain clinical efficacy. BDAA is a benzodiazepine compound discovered as a potent YFV-specific antiviral agent in our laboratory. The work reported herein further demonstrates that BDAA interaction with the YFV NS4B protein may impair the integrity of viral RNA replication organelles, which not only inhibits viral RNA replication, but also results in the leakage of viral RNA into the cytoplasm to activate RIG-I-like RNA receptors and enhances the innate antiviral immune response. The unprecedented antiviral mechanism of BDAA highlights the essential role of the NS4B protein in viral RNA replication and the evasion of host cellular innate immunity.

Funding: This study was supported by a grant from the National Institutes of Health of United States ( www.nih.gov ) R01AI134732 (JC). This study was supported by Hepatitis B Foundation and appropriation from the Commonwealth of Pennsylvania (JTG, JC). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

We recently reported the discovery of a benzodiazepine compound, BDAA, with potent antiviral effect against YFV specifically, but not other flaviviruses, in cell cultures and in vivo via oral dosing in hamsters. BDAA-resistant mutation had been mapped to proline 219 (P219) in the nonstructural 4B (NS4B) protein. The substitution of the NS4B P219 with serine confers YFV resistance to BDAA [ 13 ]. Furthermore, we have demonstrated that BDAA targets a post entry event and does not affect YFV polyprotein translation and processing [ 14 ] and there is a synergistic effect between BDAA and sofosbuvir in inhibiting YFV replication [ 15 ]. In the continuing efforts toward understanding the mode of action of BDAA, we found that in addition to inhibiting YFV replication, BDAA also specifically enhances the YFV-induced inflammatory cytokine response in infected cells in a RIG-I-like receptor (RLR)-dependent manner. To our knowledge, the dual antiviral and innate immune enhancing mechanisms are unprecedented and unique to BDAA, but not other NS4B-targeting antiviral agents against YFV and other flaviviruses.

The live-attenuated YF vaccine 17D, which was developed in 1936 by serial passages of the virus in chicken embryos, remains to be the best control measurement against YFV infection. However, the limitation in manufacture of the vaccine has caused a shortage in its supply in response to re-emerging outbreaks recently [ 3 ]. While new vaccines have been under development, combination of vaccination and therapeutic interventions should be paramount to prevent and manage the future YF outbreaks [ 4 ]. In a study performed during the 2018 outbreak of YF in Brazil, a 36% fatality rate was observed, and the high viral load was found to be a key determinant of disease severity, suggesting that an effective antiviral drug against YFV can be anticipated to significantly improve the clinical outcome of YF [ 5 ]. Antiviral therapy may also help control the outbreak by administration to potentially exposed individuals as a prophylaxis since the protection from vaccine will take weeks to become effective [ 6 – 8 ]. This approach is also important when YF vaccine is in short supply during unexpected outbreaks. Furthermore, for travelers to the endemic areas with potential risk of safely receiving YFV vaccine, prophylaxis antiviral drug can serve as an alternative prevention method [ 9 ]. However, the YF antiviral therapeutic candidates are currently limited to repurposed nucleoside analogs originally developed for hepatitis C virus (HCV) infection, among which BCX4430 (Galidesivir) [ 10 ] and Sofosbuvir [ 11 ] were found to inhibit YFV replication in vitro and in vivo in animal models. A phase 1 clinical trial to evaluate the safety, pharmacokinetics and antiviral effects in patients with YF or COVID-19 is currently ongoing via intravenous injection of BCX4430 (NCT03891420). Whereas, Sofosbuvir has been previously used to treat YFV-infected patients as compassionate use in two patients with acute liver failure [ 12 ]. Apparently, bona fide antiviral drugs against YFV are urgently needed for the control of YF outbreaks.

Yellow fever (YF), a disease caused by infection of the yellow fever virus (YFV), was once considered as the most dangerous infectious disease with high fatality rate in the beginning of the 19th century. Subsequently, with the availability of a highly effective live-attenuated vaccine together with control of mosquito vectors, YF cases have been significantly reduced and the outbreaks have been mainly limited to the tropical and subtropical forested regions of Africa and South America. However, the re-emergence of YF outbreaks in previously low-risk regions such as Angola, Democratic Republic of Congo, and Brazil since 2016, indicates that actions must be taken to deal with the changing epidemiology of YF. Furthermore, the recent report of imported YFV-infected individual into Asia as well as the evidence that Asian-Pacific Aedes mosquitoes are competent vectors for YFV raise a concern for YF outbreaks in previously non-endemic and unvaccinated areas [ 1 , 2 ]. Accordingly, new strategies in vaccination, therapeutics, and public health policies must be implemented to eliminate future global YFV threats.

Results

BDAA treatment enhances YFV-induced IFN-β expression at sub-optimal antiviral concentrations 293/IFNβLuc is a HEK293-derived reporter cell line that expresses a firefly luciferase under the control of a human IFN-β promoter. Infection of this cell line with RNA viruses, including dengue virus, YFV, Sendai virus (SeV) and encephalomyocarditis virus (EMCV), activates the reporter gene expression that quantitatively correlates with the levels of virus replication and progeny virus production, and can be inhibited in a dose-dependent manner by known antiviral compounds [16]. Interestingly, during study of BDAA’s antiviral activity in this cell line, we fortuitously discovered that treatment of BDAA, starting at 1 h post infection (hpi) for 48 h, enhanced the reporter gene expression in cells infected with YFV, but not SeV or EMCV, at concentrations approximate to its EC 50 value (Fig 1A–1C) [13]. This finding was further validated by demonstrating that treatment of YFV-infected, but not mock-infected 293/IFN-βLuc cells with BDAA using the same experimental schedule enhanced the IFN-β mRNA expression at the similar range of concentrations (Fig 1D and 1E). Moreover, treatment of YFV-infected HEK293 cells with BDAA or its three enantiomers enhanced IFN-β mRNA expression, exclusively at concentrations close to their respective EC 50 values (S1 Fig). Interestingly, it appears that BDAA and its enantiomers specifically enhanced YFV-induced IFN-β response only at sub-optimal antiviral concentrations and as anticipated, near complete inhibition of viral RNA replication at higher, but non-cytotoxic, concentrations prevented the activation of IFN-β response (Figs 1D and S1). However, treatment of 293/IFNβLuc cells with YFV NS4B inhibitors CGG-4088 or CGG-3394 [17], NS5 polymerase inhibitor Sofosbuvir [11] or an inhibitor of host endoplasmic reticulum α-glucosidases (IHVR-19029) [18–20] inhibited viral RNA replication and IFN-β mRNA expression in a concentration-dependent manner, but none of these compounds enhanced IFN-β mRNA expression at any concentration tested (S2 Fig). These results clearly indicate that BDAA uniquely enhances IFN-β gene expression induced by YFV, but not other viruses, at sub-optimal antiviral concentrations under the defined treatment schedule. PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 1. BDAA treatment specifically enhances YFV-induced IFN-β expression. (A to C) Effect on reporter expression in 293/IFNβLuc cells expressing luciferase under IFN-β promotor control. 293/IFNβLuc cells were seeded on 96-well plate and infected with YFV at MOI of 0.01 (A), Sendai virus (Sev) at 40 HAU (B), or encephalomyocarditis virus (EMCV) at MOI of 0.01 (C) for 1 h, followed by treatment with indicated concentration of BDAA. Luciferase activity was detected at 48 hpi and expressed as fold of induction relative to that in uninfected cells. Values represent average and standard deviation (STDV) from 3 independent experiments. (D to E) Effect on YFV RNA and IFN-β mRNA expression in 293/IFNβLuc cells. 293/IFNβLuc cells were either infected with YFV at MOI of 0.01 (D), or mock infected (E), for 1 h followed by treatment with indicated concentration of BDAA. Total cellular RNA was extracted at 48 hpi to detect YFV RNA and/or IFN-β mRNA by qRT-PCR. YFV RNA was expressed as percentage of YFV-infected and untreated control (D). IFN-β mRNA was expressed as fold of induction relative to that in uninfected cells (D) or untreated cells (E). BDAA concentrations for 50% inhibition of YFV RNA and 50% enhancement of IFN-β mRNA were calculated based on corresponding dose-responsive curves using GraphPad Prism 7. BDAA concentration for 50% inhibition of cell viability was determined by MTT assay (Sigma). Values represent average and standard deviation from 3 independent experiments. * indicates P<0.05, ** indicates P<0.01, ***indicates P<0.001 (IFN-β enhancement relative to no treatment control). https://doi.org/10.1371/journal.ppat.1010271.g001

Both the inhibition of viral replication and the enhancement of IFN-β expression in YFV-infected cells by BDAA rely on its specific interaction with NS4B protein We previously reported that the antiviral activity of BDAA depends on its specific interaction with YFV NS4B protein since the substitution of the NS4B residue P219 with several different amino acids conferred significant resistance to BDAA’s antiviral effect [13]. In order to determine whether BDAA enhancement of the IFN-β mRNA expression in YFV-infected cells also depends on its specific interaction with the NS4B protein, we compared the effects of BDAA treatment on the IFN-β expression in HEK293 cells infected with wild-type or BDAA-resistant YFVs. As shown in Fig 2, while BDAA enhanced the IFN-β mRNA expression in cells infected by all YFV strains, the concentrations required for significant enhancement of the cytokine response were always much higher in cells infected with BDAA-resistant YFVs, at suboptimal antiviral concentrations that are in accordance with the higher EC 50 values of BDAA against the respective mutant YFVs. In marked contrast, treatment of YFV-infected HEK293 cells with CCG-4088, an anti-YFV compound with a resistant mutation mapped to residue K128 of NS4B transmembrane domain 3 (pTMD3) [17], dose-dependently inhibited YFV replication, but did not enhance IFN-β expression at any tested concentration (S2 Fig panel B). Hence, these results provide strong genetic evidence suggesting that both the inhibition of YFV replication and the enhancement of IFN-β expression by BDAA in YFV-infected cells require its specific interaction with NS4B at a structural motif including residue P219. PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 2. BDAA enhances IFN-β in cells infected with wild-type and BDAA resistant YFV at different concentrations. HEK293 cells were infected with wild-type YFV (A), or YFV NS4B P219S (B), P219A (C), P219T (D) mutant virus for 1 h, followed by treatment with indicated concentration of BDAA. Total cellular RNA was extracted at 48 hpi to detect YFV RNA and IFN-β mRNA by qRT-PCR. YFV RNA was expressed as percentage of untreated control. IFN-β mRNA was expressed as fold of induction relative to that in uninfected cells. Values represent average and standard deviation from 4 independent experiments. * indicates P<0.05, ** indicates P<0.01, ***indicates P<0.001 (IFN-β enhancement relative to no treatment control). EC 50 values were calculated using GraphPad Prism 7. https://doi.org/10.1371/journal.ppat.1010271.g002

BDAA enhanced a broad range of cytokine response in YFV-infected cells upon the onset of viral RNA replication The results presented above indicate that BDAA significantly enhances YFV-induced IFN-β expression only at concentrations that partially inhibit viral RNA replication. More potent inhibition of viral replication at higher concentrations will profoundly reduce viral RNA, the ligand of cytoplasmic RNA sensors, and consequentially prevented YFV activation of IFN-β response. This interpretation of the results predicts that BDAA-enhanced IFN-β mRNA expression depends on the onset of viral RNA replication in infected cells. Therefore, we hypothesized that a short term BDAA treatment of YFV-infected cells to avoid significant reduction of YFV RNA may enhance the cytokine response in a broader range of BDAA concentration, including concentrations much higher than its EC 50 , only after the onset of viral RNA replication. Accordingly, HEK293 cells were infected with YFV at a MOI of 10 to ensure more than 99% of the cells were infected and then treated with BDAA, starting at the indicated time points post infection for a period of 2 h. As shown in Fig 3A, treatment of the YFV-infected HEK293 cells with 3 μM BDAA, starting at any time after the onset of viral RNA replication at 8 hpi, significantly enhanced IFN-β mRNA expression, along with significant yet modest inhibition of YFV replication. Specifically, less than 50% reductions in YFV RNA were observed over the 2 h treatment period with 3 μM BDAA initiated at any indicated time after the onset of viral RNA replication. These results thus indicate that BDAA rapidly enhances YFV-induced IFN-β expression upon the onset of viral RNA replication. As anticipated, we further demonstrated that BDAA treatment not only enhanced the expression of IFN-β, but also the expression of many other inflammatory cytokines, chemokines and interferon stimulated genes (ISGs) in HEK293 (Fig 3B) as well as HepG2 and Huh7 cells infected by YFV (S3 Fig). Moreover, an RNAseq analysis with a cutoff of false discovery rate of <0.2 and a fold change of >1.5 further demonstrated that the BDAA treatment of YFV-infected HEK293 cells at 18 hpi for 2 h increased the expression of a total of 39 cellular genes, 33 of which are inflammatory cytokines, chemokines or ISGs, and the remaining are mostly genes related to NFκB, TNF-α and MAPK signal transduction (Fig 3C and S1 Table). These results thus suggest that BDAA uniquely interacts with the YFV NS4B protein to trigger an enhanced activation of inflammatory cytokine responses. PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 3. BDAA treatment enhances a broad range of YFV-induced cytokines, chemokines and ISGs. (A) Upper panel. Experimental design. HEK293 cells were infected with YFV at MOI of 10 for 1 h, followed by treatment with 3μM of BDAA for 2 h, starting at various time post infection. Total cellular RNA was extracted at the end of the 2 h treatment (indicated by arrow heads). Lower panel YFV RNA and IFN-β mRNA were detected by qRT-PCR and expressed as fold relative to that in uninfected cells. Values represent average and standard deviation from 4 independent experiments. * indicates P<0.05, ***indicates P<0.001 (IFN-β enhancement relative to DMSO mock treatment control). (B) HEK293 cells were either uninfected or infected with YFV at MOI of 10 for 1 h. At 18 hpi, cells were either mock treated with DMSO or treated with 3μM of BDAA for 2 h. Total cellular RNA was extracted at the end of the 2 h treatment (20 hpi). YFV RNA and indicated cytokine, chemokine or ISG mRNAs were detected by qRT-PCR and expressed as fold relative to that in uninfected cells. Numbers and bars represent average and standard deviation from 3 independent experiments. * indicates P<0.05, ** indicates P<0.01, ***indicates P<0.001. (C) Gene expression profile change across all samples was analyzed by RNAseq. The heat map listed up- or down-regulated genes with fold change greater than 1.5 and false discovery rate of less than 20%. https://doi.org/10.1371/journal.ppat.1010271.g003

Both RIG-I and MDA5 play a role in YFV activation of inflammatory cytokine response and are enhanced by BDAA treatment in HEK293 cells RIG-I like receptors (RLRs), including retinoic-acid-inducible protein I (RIG-I) and melanoma differentiation antigen 5 (MDA5), are the primary viral RNA sensors for the activation of the innate immune response in flavivirus-infected cells [32, 33]. Particularly, the essential and nonredundant roles of RIG-I and MDA5 in the detection and control of West Nile virus infection by recognizing 5’ triphosphate and double-stranded RNAs, respectively, have been well documented [34, 35]. Both RIG-I and MDA5 are involved in responses to dengue virus serotype-2 (DENV-2) infection in mouse embryonic fibroblasts [36] and Zika virus infection in human trophoblasts [37]. Concerning the role of RLRs in YFV infection, it was reported that replication of YFV vaccine strain YF-17D in human pDCs and pDC-like cell lines, and YFV reference strain Asibi in human hepatoma cells stimulated type I IFN production through activation of RIG-I [38, 39]. To determine the roles of RLR pathway in YFV induction and BDAA enhancement of cytokine response, we created HEK293-derived cell lines deficient in the expression of RIG-I, MDA5 or MAVS, the central adaptor for RLR signaling, by CRISPR/Cas9 gene editing technology. The knockout of specific gene expression in each cell line was confirmed by Western blot assays (Fig 7A) and sequencing analyses of the chromosomal DNA at the respective guide RNA binding sites (Fig 7B). The functionality of RLR signaling in parental HEK293 and the three knockout cell lines were also characterized. As shown in Fig 7C, consistent with previous reports, while the induction of IFN-β mRNA expression by high molecular weight poly(I:C) transfection was almost completely abolished in MDA5 knockout cells, SeV-induced IFN-β expression was more significantly compromised in RIG-I knockout cells [33, 40]. Also as expected, MAVS knockout profoundly abolished IFN-β induction by both types of the RNA ligands (Fig 7C). These results indicate that all the knockout cells with deficiency in RLR signaling displayed the expected phenotypes in responding to well-characterized RLR ligands. To determine the role of each of the three genes in YFV replication and induction of inflammatory cytokine response, parental HEK293 and each of the three knockout cell lines were infected with YFV and infected cells were harvested at the indicated time points post infection for qRT-PCR quantification of intracellular viral RNA and IFN-β mRNA (S6 Fig). The levels of YFV RNA in each of the RLR pathway component knockout cell lines were only slightly higher (less than 6-fold) than that in parental HEK293 cells (S6 Fig, panel A). However, induction of IFN-β mRNA expression was drastically reduced in both RIG-I and MDA5 knockout cells and even more profoundly reduced in MAVS knockout cells up to 36 to 48 h post infection. However, after 48 h post infection, IFN-β mRNA was significantly induced in all the three knockout cell lines infected by YFV (S6 Fig, panel B). Because the later surge of IFN-β mRNA induction occurred not only in RIG-I or MDA5 knockout cells, but also in MAVS knockout cells, the results strongly indicate the activation of innate immune pathways other than RLRs in the later stage of YFV infection in these cell lines. It is rather interesting that the robust activation of IFN response in HEK293 cells only modestly inhibited YFV replication at the condition of a low (0.1) MOI infection. In fact, similar observations that RLR silencing did not significantly affect viral replication in cell cultures had also been reported for other flaviviruses as reviewed by Valadão et al. in 2016 [41]. PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 7. Characterization of HEK293 cells with RIG-I, MDA5 or MAVS knockout. (A) Expression of RIG-I, MDA5 or MAVS in wildtype and knockout cell lines were detected by western blot assay, using β-actin as a loading control. (B) Analysis of sequence flanking the corresponding sgRNA targeting site in RIG-I, MDA5 or MAVS KO cell lines. Dotted line represents base deletion and red letter represents base insertion. (C) Functional analysis of KO cell lines. Parental HEK293 or derived cell lines with indicated gene KO were seeded in 24-well plate and either infected with SeV or transfected with high molecular weight (HMW) poly (I:C). Total RNA was extracted at 24 h later and IFN-β mRNA was measured by qRT-PCR and expressed as fold of induction (average ± STDV, n = 3). (D) Parental HEK293 or derived cell lines with indicated gene KO were seeded in 24-well plate and infected with YFV at MOI of 10. At 18 hpi, the cells were treated with DMSO or 3μM of BDAA for 2h. Total RNA was extracted at 20 hpi. IFN-β mRNA was measured by qRT-PCR and expressed as fold of induction (average ± STDV, n = 3). n.s. indicates not significant, ** indicates P<0.01, ***indicates P<0.001 compared to DMSO mock treatment controls. https://doi.org/10.1371/journal.ppat.1010271.g007 Finally, we examined the pathways responsible for the BDAA enhanced IFN-β production in YFV-infected cells. As shown in Fig 7D (blue bars), in both RIG-I and MDA5 knockout cells infected with YFV, BDAA treatment still enhanced IFN-β mRNA production. However, both the YFV-induced and BDAA-enhanced IFN-β production were almost completely abolished in MAVS knockout cells. These results clearly indicate that at least in the early time post YFV infection (18 hpi), both RIG-I and MDA5 play important roles in the induction of inflammatory cytokine response and BDAA treatment enhanced the cytokine response by both RNA sensors. These results imply that the RNA ligands responsible for the YFV-induced innate immune response may share the same molecular features with those responsible for BDAA enhanced cytokine responses, presumably both are various types of viral RNA, including dsRNA, derived from viral replication complexes.

[END]

[1] Url: https://journals.plos.org/plospathogens/article?id=10.1371/journal.ppat.1010271

(C) Plos One. "Accelerating the publication of peer-reviewed science."
Licensed under Creative Commons Attribution (CC BY 4.0)
URL: https://creativecommons.org/licenses/by/4.0/


via Magical.Fish Gopher News Feeds:
gopher://magical.fish/1/feeds/news/plosone/