RIG-I–Like Receptor Triggering by Dengue Virus Drives Dendritic Cell Immune Activation and TH1 Differentiation

RIG-I−like Receptor Triggering by Dengue Virus Drives Dendritic Cell Immune Activation and TH1 Differentiation

This information is current as Joris K. Sprokholt, Tanja M. Kaptein, John L. van Hamme, of September 27, 2021. Ronald J. Overmars, Sonja I. Gringhuis and Teunis B. H. Geijtenbeek J Immunol published online 15 May 2017 http://www.jimmunol.org/content/early/2017/05/13/jimmun

ol.1602121 Downloaded from

Supplementary http://www.jimmunol.org/content/suppl/2017/05/13/jimmunol.160212 Material 1.DCSupplemental http://www.jimmunol.org/

Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication by guest on September 27, 2021

*average

Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2017 by The American Association of Immunologists, Inc. All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. Published May 15, 2017, doi:10.4049/jimmunol.1602121 The Journal of Immunology

RIG-I–like Receptor Triggering by Dengue Virus Drives Dendritic Cell Immune Activation and TH1 Differentiation

Joris K. Sprokholt, Tanja M. Kaptein, John L. van Hamme, Ronald J. Overmars, Sonja I. Gringhuis, and Teunis B. H. Geijtenbeek

Dengue virus (DENV) causes 400 million infections annually and is one of several viruses that can cause viral hemorrhagic fever, which is characterized by uncontrolled immune activation resulting in high fever and internal bleeding. Although the underlying mechanisms are unknown, massive cytokine secretion is thought to be involved. Dendritic cells (DCs) are the main target cells of DENV, and we investigated their role in DENV-induced cytokine production and adaptive immune responses. DENV infection induced DC maturation and secretion of IL-1b, IL-6, and TNF. Inhibition of DENV RNA replication abrogated these responses. Notably, silencing of RNA sensors RIG-I or MDA5 abrogated DC maturation, as well as cytokine responses by DENV-infected DCs. DC maturation was

induced by type I IFN responses because inhibition of IFN-a/b receptor signaling abrogated DENV-induced DC maturation. Downloaded from Moreover, DENV infection of DCs resulted in CCL2, CCL3, and CCL4 expression, which was abrogated after RIG-I and MDA5

silencing. DCs play an essential role in TH cell differentiation, and we show that RIG-I and MDA5 triggering by DENV leads to TH1 polarization, which is characterized by high levels of IFN-g. Notably, cytokines IL-6, TNF, and IFN-g and chemokines CCL2, CCL3, and CCL4 have been associated with disease severity, endothelial dysfunction, and vasodilation. Therefore, we identiﬁed RIG-I and MDA5 as critical players in innate and adaptive immune responses against DENV, and targeting these receptors has the potential to

decrease hemorrhagic fever in patients. The Journal of Immunology, 2017, 198: 000–000. http://www.jimmunol.org/

engue virus (DENV) is a blood-borne pathogen that in- contributing to the pathogenesis of DENV (3). High serum levels fects almost 400 million people annually (1). Most pa- of IFN-g, IL-1b, IL-6, TNF, CCL2, CCL4, CCL5, CXCL8, and D tients experience mild disease, but dengue fever can CXCL10 have been detected in patients with DHF compared with progress to life-threatening dengue hemorrhagic fever (DHF) or dengue fever (4). In particular, IL-6, IL-8, and TNF are associated dengue shock syndrome (DSS), which are characterized by throm- with increased vascular permeability and plasma leakage (4). The bocytopenia and increased vascular permeability, resulting in cellular source of DENV-induced cytokines and chemokines is

plasma leakage and hemorrhages (2). With no specific antivirals unknown, but dendritic cells (DCs) and macrophages are the main by guest on September 27, 2021 available, treatment relies on fluid replacement and the use of target cells of DENV (5–8). DCs are especially susceptible to analgesics. Therefore, there is an urgent need to elucidate the DENV infection because of the expression of DENV entry re- factors that are involved in DSS and DHF to improve treatment. ceptor DC-SIGN (5, 9, 10). The underlying cause of DHF and DSS remains unclear, but it is DENV-infected DCs show increased expression of maturation believed that massive cytokine secretion is one of the major factors markers and produce cytokines IL-6, IL-10, and TNF and chemokines CCL2, CXCL8, CXCL10, and type I IFNs a and b (11, Department of Experimental Immunology, Academic Medical Center, University of 12). DCs express numerous pattern recognition receptors that Amsterdam, 1105 AZ Amsterdam, the Netherlands; and Amsterdam Infection & recognize conserved pathogen-associated molecular patterns to Immunity Institute, 1105 AZ Amsterdam, the Netherlands induce DC maturation and specific TH cell polarization. Several ORCID: 0000-0003-0765-9441 (J.L.v.H.). pattern recognition receptors have been implicated in DENV Received for publication December 19, 2016. Accepted for publication April 13, sensing, including TLRs, RIG-I–like receptors (RLRs), and 2017. C-type lectin receptors (CLRs) (5, 13–16). Activation of TLRs This work was supported by the Netherlands Organisation for Scientific Research and RLRs leads to cytokine responses via NF-kB, as well as (Grant NWO VICI 918.10.619) and the European Research Council (Advanced Grant 670424). induction of type I IFNs via IRFs. However, it is unknown which J.K.S. designed, performed, and interpreted most experiments and prepared the manu- sensor in DCs is responsible for DC maturation and cytokine script; T.M.K. performed dengue virus (DENV) infections, T cell differentiation, FACS production. analyses, and cDNA synthesis; J.L.v.H. performed DENV infections, T cell assays, and In this study, we identified that DENV RNA replication is es- cDNA synthesis; R.J.O. performed DENV infections and cDNA synthesis; S.I.G. helped to prepare the manuscript; and T.B.H.G. supervised all aspects of this study. sential for DENV-induced DC maturation and cytokine responses. Address correspondence and reprint requests to Prof. Teunis B. H. Geijtenbeek, Notably, the intracellular RLRs RIG-I and MDA5 sense DENV Department of Experimental Immunology, Academic Medical Center, University RNA replication products and induce innate signaling via adapter of Amsterdam, Meibergdreef 9, 1105 AZ Amsterdam, the Netherlands. E-mail address: protein MAVS to upregulate CD86 expression via type I IFN. [email protected] DENV-induced IL-1b, IL-6, and TNF secretion by DCs and in- The online version of this article contains supplemental material. duction of CCL2, CCL3, and CCL4 critically depended on RIG-I Abbreviations used in this article: CLR, C-type lectin receptor; DC, dendritic cell; DENV, dengue virus; DHF, dengue hemorrhagic fever; DSS, dengue shock syn- and MDA5. RLR activation by DENV was also essential for the drome; h.p.i., hour postinfection; ISG, IFN-stimulated gene; MOI, multiplicity of induction of adaptive immune responses because inhibiting RLR infection; NS, nonstructural protein; RLR, RIG-I–like receptor; ROS, reactive oxy- signaling in DCs abolished T 1 responses. These findings have the gen species; siRNA, short interfering RNA. H potential to enable novel therapeutic strategies that interfere with Copyright Ó 2017 by The American Association of Immunologists, Inc. 0022-1767/17/$30.00 RIG-I and MDA5 activation to limit DENV pathogenesis.

www.jimmunol.org/cgi/doi/10.4049/jimmunol.1602121 2 RLRs DRIVE DENV IMMUNE ACTIVATION

Materials and Methods rabbit (1:400, A-21206), or Alexa Fluor 647–conjugated anti-rabbit DC stimulation and RNA interference (1:400, A-21245; both from Thermo Fisher) in combination with FITC- conjugated anti–HLA-A2 (1:25, 551285), PE/Cy7-conjugated anti–HLA- Peripheral blood monocytes were isolated from buffy coats of healthy DR (1:200, 560651), PE-conjugated anti-CD80 (1:25, 557227), allophycocyanin- donors (Sanquin) by Lymphoprep (Axis-Shield) gradient, followed by conjugated anti-CD83 (1:25, 551073), FITC-conjugated anti-CD86 (1:25, Percoll (Amersham Biosciences) gradient steps. Monocytes were differ- 555657), or PE-conjugated anti-CD86 (1:50, 555658; all from BD Phar- entiated into immature DCs in the presence of 500 U/ml IL-4 and 800 U/ml mingen). Cells were analyzed on a FACSCanto II (BD Biosciences). GM-CSF (both from Invitrogen) for 6–7 d in RPMI 1640 supplemented with 10% FCS, 10 U/ml penicillin, 10 mg/ml streptomycin (all from ELISA Invitrogen), and 2 mM L-glutamine (Lonza). DC purity was determined by Cell culture supernatants were harvested 24 or 48 h after stimulation, and flow cytometry based on DC-SIGN, CD1c, CD11c, and HLA-DR ex- IL-1b, IL-6, IL-12p70, and TNF (eBioscience) were measured by ELISA, pression and the absence of CD14, CD19, CD3, and CD56; purity was according to the manufacturer’s instructions. routinely .95%. This study was done in accordance with the ethical guidelines of the Academic Medical Center. Real-time quantitative PCR DCs were stimulated with 10 ng/ml LPS (Sigma) or 2 mg/ml poly(I:C)/ LyoVec (InvivoGen), which is polyinosinic-polycytidylic acid complexed mRNA was isolated using an mRNA capture kit (Roche), and cDNA was with liposomes to transfect polyinosinic-polycytidylic acid into the cytoplasm synthesized with a reverse transcriptase kit (Promega). PCR amplifica- and stimulate RIG-I and MDA5. DENV replication inhibitor SDM25N tion was performed in the presence of SYBR Green using an ABI 7500 (10 mM; Tocris Bioscience), IFN-a/b capture protein B18R (eBioscience), Fast Real-Time PCR detection system (Applied Biosystems). Specific N-acetyl-L-cysteine (1–3 mM; Sigma), or catalase (100–500 U/ml; Sigma) primers were designed using Primer Express 2.0 (Applied Biosystems; was added simultaneously with DENV to DCs. Supplemental Table I). Expression of target genes was normalized to Ct (GAPDH) 2 Ct (target) DCs were transfected with 500 nM short interfering RNAs (siRNAs) GAPDH [Nt =2 ] and set at 1 in DENV-infected DCs using the Neon Transfection System (Thermo Fisher), according to the for each donor within one experiment. Downloaded from manufacturer’s instructions. In brief, DCs were washed with PBS, resus- pended in Buffer R (Thermo Fisher), and divided between different siR- Statistical analysis NAs. DCs were transfected with a single pulse of 1500 V for 20 ms, mixed Statistical analyses were performed using the Student t test for paired with complete RPMI 1640, and incubated for 48 h before stimulation. The observations. Statistical significance was set at p , 0.05. following SMARTpool siRNAs were used: MAVS (M-024237-02), RIG-I (M-012511-01), MDA5 (M-013041-00), and nontargeting siRNA (D-001206-13) as control (all from Dharmacon). Silencing was confirmed by real-time Results http://www.jimmunol.org/ PCR, flow cytometry, and immunoblotting (Supplemental Fig. 3). Abs DENV induces DC maturation and cytokine responses used for flow cytometry were anti–RIG-I (1:50, 3743), anti-MDA5 (1:50, 5321), and anti-MAVS (1:50, 3993; all from Cell Signaling Technology), in To investigate DC maturation and cytokine induction by DENV, we combination with PE-conjugated anti-rabbit (1:200, 711-116-152; Jackson infected DCs at an MOI of 1 or 3 for 24 and 48 h and measured the ImmunoResearch), and cells were analyzed on a FACSCanto II (BD Bio- expression of DENV nonstructural protein (NS) 3 (Supplemental sciences). Abs used for Western blot were similar as for flow cytometry but Fig. 1A). Nonstructural viral proteins are only produced during were used at a different dilution (all 1:1000 with the exception of anti–b- actin [1:500, sc-81178; Santa Cruz Biotechnology]). replication and, therefore, indicate active viral infection (18). NS3 was detected 24 h postinfection (h.p.i.), and DC infection T cell differentiation increased over time at an MOI of 1, whereas an MOI of 3 Naive CD4+ T cells were isolated from buffy coats of healthy blood donors resulted in maximal infection at 24 h.p.i. (Fig. 1A). A plateau of by guest on September 27, 2021 (Sanquin), using a human CD4+ T Cell Isolation Kit II (Miltenyi Biotec), infectionhasbeenshownpreviouslyinmonocyte-derivedDCs + by negative selection and subsequent depletion of CD45RO memory and CD1c+ DCs from human skin (19, 20). Next, we investi- T cells using PE-conjugated anti-CD45RO (80 mg/ml, R0843; Dako) and gated DC maturation upon infection by measuring the expres- anti-PE beads (Miltenyi Biotec). This study was approved by the Medical Ethics Review Committee of the Academic Medical Center. sion of DC maturation markers CD80, CD83, and CD86, as well DCs were silenced for indicated proteins and stimulated for 48 h, as as the expression of MHC class I (HLA-A2) and class II (HLA- indicated. DCs were combined with allogeneic naive CD4+ T cells (5,000 DR). Interestingly, DC maturation was observed at 48 h.p.i., but DCs:20,000 T cells) in the presence of 10 pg/ml Staphylococcus aureus en- notat24h.p.i.,evenatanMOIof3anddespiteactiveviral terotoxin B (Sigma). SDM25N (1 mM; Tocris Bioscience) was added to co- cultures of SDM25N-treated DCs to maintain inhibition of DENV replication. replication at this time point (Fig. 1B). DENV has been shown After 5 d, cells were further cultured in the presence of 10 U/ml IL-2 (Chiron). to induce apoptosis and interfere with DC activation and mat- Resting T cells were restimulated with 100 ng/ml PMA and 1 mg/ml ion- uration (11, 21, 22). Therefore, we analyzed the expression of omycin (both Sigma) for 6 h. For flow cytometry analysis of restimulated DC maturation markers in infected (DENV+) and bystander T cells, cells were fixed in 4% paraformaldehyde for 20 min, followed by (DENV2) DCs. Notably, maturation of the DENV2 population permeabilization in 0.1% saponin in PBS for 10 min. Cells were stained with + allophycocyanin-conjugated anti–IL-4 and FITC-conjugated anti–IFN-g (both is increased compared with the DENV population, indicating from BD Biosciences) before being analyzed on a FACSCanto II (BD that DENV limits DC maturation (Fig. 1C). Biosciences). We next investigated the induction of different cytokines in- Virus production and infection volved in fever (IL-1b, IL-6), TH skewing (IL-1b, IL-6, IL-12p70), or activation of endothelial cells (IL-6, TNF). Similar to DC C6/36 cells (80% confluent) were infected with DENV-2/16681 at a mul- maturation, DENV infection induced IL-1b, IL-6, and TNF at 48 tiplicity of infection (MOI) of 0.01 in HMEM supplemented with 2% FCS, 10 U/ml penicillin, 10 mg/ml streptomycin (all from Invitrogen), 2 mM h.p.i. but not at 24 h.p.i. (Fig. 1D). DENV infection did not induce L-glutamine, and 100 mM nonessential amino acids (both from Lonza). After IL-12p70, whereas DCs were able to produce IL-12p70 in re- 5–7 d, supernatant was harvested and cleared from cellular debris by cen- sponse to LPS (Fig. 1D). IL-12p70 is a heterodimeric protein trifugation and subsequent filtration using a 0.2-mM filter. Supernatant was consisting of subunits p40 and p35. Therefore, we measured the 2 aliquoted, snap-frozen in liquid nitrogen, and stored at 80˚C. Viral titers induction of both subunits at the mRNA level, as well as mRNA were determined as described previously (17). Supernatant of uninfected C6/36 cells cultured for 5–7 d was used for mock treatment of DCs. levels of IL-1b, IL-6, and TNF at 32 h.p.i. when transcription is DCs were infected with DENV at an MOI of 1, unless stated otherwise. maximal, as determined by time-course experiments (data not Infection was determined after 36–48 h by measuring the percentage of shown). Remarkably, DENV infection of DCs induced mRNA ex- + NS3 cells by flow cytometry. Cells were fixed in 4% paraformaldehyde pression of IL-1b, IL-6, IL-12p35, and TNF but not IL-12p40 for 15 min, followed by permeabilization in PBS supplemented with 0.1% saponin for 10 min. Cells were stained with anti-NS3 (1:800, (Fig. 1E). Our results strongly suggest that productive viral repli- SAB2700181; Sigma), followed by PE-conjugated anti-rabbit (1:200, 711- cation of DENV is required to induce DC maturation and cytokine 116-152; Jackson ImmunoResearch), Alexa Fluor 488–conjugated anti- responses. Interestingly, although maturation of DENV-infected The Journal of Immunology 3 Downloaded from http://www.jimmunol.org/ by guest on September 27, 2021

FIGURE 1. DENV induces DC maturation and cytokine responses. DCs were mock treated or infected at an MOI of 1 or 3, and the percentage of NS3+ cells (A) or the expression of CD80, CD83, CD86, HLA-A2, and HLA-DR (B and C) was measured 24 or 48 h.p.i. by ﬂow cytometry. The DENV+ and DENV2 populations in (C) were gated as shown in Supplemental Fig. 1A. (D) DCs were similarly infected as in (A), and the production of IL-1b, IL-6, IL-12p70, and TNF was measured in the cell culture supernatant by ELISA. (E) DCs were infected at an MOI of 1, and mRNA expression of the indicated genes was measured 32 h.p.i. by real-time PCR, normalized to GAPDH, and set as 1 in DENV-infected samples. Data represent three (B and C) or four (A and D) donors from different experiments or are collated data (mean 6 SD) from at least four donors from different experiments (E) [mean 6 SD of duplicates in (A) and (D)]. **p , 0.01, Student t test.

DCs occurs, it is to a lesser degree than observed for bystander and TNF was dependent on DENV RNA replication (Fig. 2B, DCs, suggesting that DENV infection limits DC maturation. 2C). Reactive oxygen species (ROS) have been implicated in DENV-induced maturation (21). However, treatment of DCs with DENV RNA replication is essential for DC activation ROS inhibitors N-acetyl-L-cysteine or catalase did not decrease DENV viral particles consist of different proteins and contain DENV-induced DC maturation (Supplemental Fig. 2). These ssRNA that can activate TLR7 (15). During DENV RNA repli- data strongly suggest that DENV RNA replication is pivotal for cation, dsRNA intermediates are formed that can be sensed by DC activation. These data also explain the delayed response of RLRs and TLR3 (16, 23). To investigate whether DENV repli- DCs to DENV infection, because DENV RNA intermediates are cation is required for DC maturation and cytokine responses, we formed during replication and are not present in viral particles used the DENV replication inhibitor SDM25N, which efﬁciently that initiate infection. inhibits DENV RNA replication, as well as infection (24). We have focusedonCD86expressiontoinvestigatetheroleofDENV DENV-induced cytokine responses depend on RIG-I and MDA5 replication because this maturation marker was highly induced by DENV replication products are formed in the cytosol and can be DENV. DENV-induced upregulation of CD86 was completely sensed by the RLRs RIG-I and MDA5 (16, 23), and RLR activation abrogated when DENV replication was inhibited by SDM25N in DENV-infected DCs leads to T follicular helper cell polariza- (Fig. 2A). Also, DENV-induced expression of IL-1b, IL-6, IL-12p35, tion via IL-27 (data not shown). To investigate whether RIG-I and 4 RLRs DRIVE DENV IMMUNE ACTIVATION

FIGURE 2. DENV RNA replication is essential for DC maturation and cytokine responses. (A) DCs were infected with DENV at an MOI of 1 in the Downloaded from presence or absence of 10 mM SDM25N, and the expression of CD86 in the entire population of DCs was measured 48 h.p.i. by flow cytometry. DCs were infected similarly to (A), and the mRNA expression of the indicated genes was measured 32 h.p.i. by real-time PCR, normalized to GAPDH, and set as 1 in DENV-infected samples without SDM25N (B), or the secretion of the indicated proteins in the cell culture supernatant was measured by ELISA (C). Data are representative of at least four (C) or two (A) donors from different experiments or are collated data (mean 6 SD) from four donors from different experiments (B) [mean 6 SD of duplicates in (C)]. **p , 0.01, Student t test. http://www.jimmunol.org/ MDA5 are also involved in IL-1b, IL-6, and TNF responses, we MDA5 strongly decreased DENV-induced DC maturation (Fig. 5A), used RNA interference to silence RIG-I and MDA5 to investigate indicating that RLR activation is essential for DC–T cell interactions whether these RNA sensors are responsible for DENV-induced DC during DENV infection. We next investigated how RLRs drive DC activation. We have confirmed silencing at mRNA and protein maturation in response to infection. Because type I IFNs are known levels, as well as the specificity of the silencing, using different to induce DC maturation (30), we used rIFN-a/bR protein (B18R) RLR and TLR ligands. Moreover, silencing did not affect cell vi- to neutralize IFN-a/b secreted by DCs and, thereby, prevent the ability (Supplemental Fig. 3). Notably, silencing RIG-I and MDA5 induction of IFN-stimulated genes (ISGs) by IFN-a/bRsignaling. alone or together abrogated IL-1b, IL-6, and TNF expression upon Numerous ISGs have antiviral properties that limit viral replication, DENV infection (Fig. 3A, 3B). RIG-I and MDA5 triggering by and interfering with IFN-a/bR activation greatly enhanced DENV by guest on September 27, 2021 DENV also resulted in IL-12p35 expression (Fig. 3A). RLRs signal infection of DCs (Fig. 5B). B18R effectively inhibited the induction via the adapter molecule MAVS, which forms a scaffolding plat- of the ISGs MxA, IRF7, ISG15, OAS1, and ADAR1 by DENV form from which IRF and NF-kB activation is orchestrated (25). (Fig. 5C). Remarkably, inhibiting IFN-a/bR signaling abrogated DC Similar to RIG-I and MDA5, silencing MAVS abrogated DENV- maturationinresponsetoDENV,despite enhanced infection (Fig. 5D). induced cytokine expression (Fig. 3C, 3D). These data strongly These data indicate that RLR activation by DENV drives DC suggest that RIG-I and MDA5 are critical sensors of DENV RNA maturation and that this is mediated via type I IFN responses. replication that drive the cytokine responses of DCs in response to DENV-induced T skewing requires RIG-I and MDA5 DENV via the adapter protein MAVS. H activation RIG-I and MDA5 activation leads to chemokine responses IFN-g is associated with severe disease in dengue patients and In addition to elevated levels of cytokines that are associated with TH1 cells produce high levels of IFN-g. Therefore, we investi- disease severity, several chemokines can impact vascular integrity, gated DENV-induced TH cell polarization. First, we evaluated our including CCL2, CCL3, CCL4, and CXCL8 (4, 26). Therefore, we assay by coculturing LPS-stimulated DCs, either alone or in investigated whether DENV-infected DCs produced these che- combination with IFN-g or PGE2, with naive TH cells for 10– mokines in an RLR-dependent manner. DENV infection induced 12 d before measuring the percentage of IFN-g+ or IL-4+ cells, the expression of CCL2, CCL3, and CCL4, which was abrogated indicative of TH1orTH2 cells, respectively (Supplemental Fig. after silencing of MAVS or both RIG-I and MDA5 (Fig. 4). 1B). LPS-stimulated DCs induced mixed skewing of TH1 and TH2 CXCL8 expression was not induced (data not shown). These data cells, whereas LPS+IFN-g–stimulated DCs induced more TH1 indicate that the DENV-induced chemokine responses by DCs polarization and less TH2 differentiation compared to LPS (Fig. critically depend on RLR-MAVS activation. 6A, 6B). In contrast, coculturing naive TH cells with LPS+PGE2- stimulated DCs increased TH2 and decreased TH1 cell polarization RLR-dependent type I IFN responses induce DC maturation compared to LPS (Fig. 6A, 6B). Next, we examined TH polari- DC maturation is essential for inducing adaptive immune responses zation induced by DENV-infected DCs and the involvement of by providing costimulatory signals to T cells that strengthen TCR RIG-I and MDA5. DENV-infected DCs resulted in more TH1 cells signaling (27). CD86 expression can be induced by NF-kB, as and less TH2 cells compared to mock-treated DCs, and this effect well as IRFs, both of which are induced by RLRs (28, 29). was dependent on viral replication because SDM25N decreased Therefore, we investigated whether DENV-induced DC matura- TH1 polarization (Fig. 6C). Strikingly, silencing MAVS abrogated tion is dependent on RIG-I and MDA5 and compared the level of TH1 skewing by DENV-infected DCs (Fig. 6D). Moreover, direct CD86 expression induced by DENV with LPS. DENV induced stimulation of RIG-I and MDA5 using the chemical RLR li- CD86 comparable to LPS, and silencing MAVS or both RIG-I and gand poly(I:C)/LyoVec also resulted in strong TH1 polarization The Journal of Immunology 5 Downloaded from http://www.jimmunol.org/ by guest on September 27, 2021

FIGURE 3. DENV induces DC immune activation via RIG-I and MDA5. (A and C) DCs were mock treated or infected with DENV at an MOI of 1 after RIG-I, MDA5, or MAVS silencing, and the mRNA expression of the indicated genes was measured 32 h.p.i. by real-time PCR, normalized to GAPDH, and set as 1 in DENV-infected samples treated with control siRNA. (B and D) DCs were treated similarly to (A), and the production of the indicated proteins in the cell culture supernatant was measured by ELISA. Data are representative of at least two donors (B and D) from different experiments or collated data (mean 6 SD) from four (C)orthree(A) donors from different experiments [mean 6 SD of duplicates in (B)and(D)]. **p , 0.01, Student t test. n.d., not detected.

(Fig. 6E). These data indicate that DENV infection of DCs results in indicating that DENV viral particles do not directly activate DCs TH1 skewing and that this effect is mediated by RIG-I and MDA5. via cell surface receptors or endosomal TLRs. Instead, DENV RNA replication was essential for DC immune activation, and these re- Discussion sponses were mediated by the cytoplasmic RLRs RIG-I and MDA5. RNA viruses can induce viral hemorrhagic fever, and this effect is RIG-I and MDA5 activation leads to signaling cascades that thought to be mediated by immune activation; however, little is involve the adapter molecule MAVS. Activated RIG-I and MDA5 known about the underlying mechanisms. In this study, we show associate with MAVS, leading to the recruitment of IkB kinase– that DENV infection of DCs results in DC maturation and high related kinase IKKε and TANK-binding protein 1, which activate levels of cytokine secretion, chemokine expression, and TH1 po- IRF3 for nuclear translocation and IFN-b transcription (31). Our larization and that these responses depend on viral RNA replica- data indicate that RLR-induced IFN-b limits DENV infection in tion. Notably, we identiﬁed RIG-I and MDA5 as key sensors that DCs and is crucial for DC maturation by inducing IFN-a/bR mediate immune activation of DCs by DENV. Inhibiting RLR signaling. MAVS-induced type I IFN responses are initiated by activation or signaling abrogated DC maturation, cytokine secre- IRF3 and prolonged by IFN-a/bR-induced IRF7. Interestingly, tion, and chemokine expression in response to DENV. IRF7 is a transcription factor of CD86 (28), indicating that RLRs DENV viral particles contain multiple components that can might induce DC maturation via IFN-a/bR–dependent IRF7 ex- directly activate several receptors. The viral envelope proteins can pression during DENV infection. Because DC maturation is cru- interact with CLRs, whereas the viral genome can activate endo- cial for inducing adaptive immune responses, type I IFN is not somal TLRs that recognize ssRNA (5, 10, 15). During viral rep- only involved in innate suppression of DENV replication, it could lication, dsRNA intermediates are formed that can trigger also play a role in adaptive T cell responses. cytoplasmic RNA sensors (16, 23). We did not observe DC mat- Furthermore, the importance of IFN-a/bR signaling in DC uration or cytokine secretion during the ﬁrst 24 h of infection, maturation can explain the difference in maturation between in- 6 RLRs DRIVE DENV IMMUNE ACTIVATION

FIGURE 4. RIG-I and MDA5 activation by DENV induces chemokine expression. (A and B) DCs were infected with DENV after RIG-I, MDA5, or MAVS silencing, and the mRNA expression of the indicated genes was measured 32 h.p.i. by real-time PCR, normalized to GAPDH, and set as 1 in DENV-infected samples treated with control siRNA. Data are collated (mean 6 SD) from at least ﬁve donors from different experiments. *p , 0.05, **p , 0.01, Student t test.

fected and bystander DCs (11, 22). DENV is known to inhibit is tightly regulated and requires a multistage process to be func- Downloaded from IFN-a/bR signaling by targeting STAT2 for degradation by DENV tionally active. Priming signals induce pro–IL-1b transcription and protein NS5 (32, 33). Because NS5 is only expressed in infected translation, whereas secondary signals induce inflammasome acti- cells, and not bystander cells, this could explain why bystander cells vation to process pro–IL-1b into mature IL-1b by caspases. Nor- show increased maturation compared with infected cells. mally, pro–IL-1b synthesis and inflammasome activation are Another factor that is important for DC maturation by DENV is induced by separate signals, and only few receptors have the ca- oxidative stress. DENV infection of DCs induces oxidative stress, pacity to provide both signals (38). In human macrophages, CLR http://www.jimmunol.org/ resulting in ROS and enhanced DC maturation (21). Interestingly, CLEC5A is required for NLPR3 expression and subsequent pro- ROS is known to enhance RIG-I and MDA5 activation (34). cessing of pro–IL-1b by NLPR3 inflammasomes in response to Therefore, the production of ROS by infected DCs might be an DENV, but it requires priming by LPS for efficient IL-1b secretion antiviral strategy to limit viral replication via enhanced RIG-I and (39). In contrast, RIG-I can induce pro–IL-1b expression and as- MDA5 activation, although our results did not show that neu- sociate with inflammasome adapter molecule ASC to induce pro– tralizing ROS affects DC maturation. IL-1b processing by caspase 1 (35). We observed that DENV in- In parallel to IRF3-mediated responses, RLR activation leads to fection of DCs resulted in pro–IL-1b expression and that DCs did NF-kB activation via CARD9 and Bcl-10 (35). NF-kB activation not require priming signals for efficient IL-1b secretion. Notably, is essential for the expression of most cytokines, including IL-1b, the induction of pro–IL-1b, as well as the secretion of mature by guest on September 27, 2021 which potently induces fever (36, 37). Persistent high fever is one IL-1b, depended on RIG-I and MDA5. Because IL-1b is also a of the clinical symptoms of DENV, but the underlying mecha- potent inducer of vasodilation (40), inhibiting RLR-dependent nisms that lead to IL-1b secretion are unknown. IL-1b expression inflammasome activation has the potential to lower DHF.

FIGURE 5. RIG-I and MDA5 induce DC maturation via type I IFN. DCs were stimulated with LPS, mock treated, or infected with DENV at an MOI of 1 after RIG-I, MDA5, or MAVS silencing (A) or in the presence or absence of B18R (D), and the expression of CD86 in the entire population of DCs was measured 48 h.p.i. by ﬂow cytometry. Similar to (D) but MxA, IRF7, ISG15, OAS1, and ADAR1 expression was measured by real-time PCR, normalized to GAPDH, and set as 1 in DENV-infected samples without B18R (C), or the percentage of NS3+ cells was measured by ﬂow cytometry (B). Data are representative of at least four donors from different experiments (A, B, and D) or are collated data (mean 6 SD) from four donors from different experiments (C) [mean 6 SD of duplicates in (B)]. *p , 0.05, **p , 0.01, Student t test. The Journal of Immunology 7

FIGURE 6. RIG-I and MDA5 induce TH1 polarization. (A and B) Flow cytometry analysis of intracellular IFN-g or IL-4 expression in differentiated T cells after coculture of naive CD4+ T cells with DCs that were not stimulated (iDCs) or were stimulated with LPS alone or in combination with IFN-g or Downloaded from PGE2. Numbers in quadrants in (A) indicate the percentage of gated cells. DCs were mock treated or infected with DENV at an MOI of 1 in the presence or absence of SDM25N (C) or after MAVS silencing (D) and were cocultured with naive CD4+ T cells similar to (A); the percentage of IFN-g+ or IL-4+ T cells was measured by ﬂow cytometry. (E) Similar to (A), but DCs were left untreated or were stimulated with poly(I:C)/LyoVec before coculture with naive CD4+ T cells. Data are representative of at least three (A, B, and E) or two (C and D) independent experiments [mean 6 SD of duplicates in (B)–(E)]. iDCs, immature DCs; p(I:C)LV, poly(I:C)/LyoVec.

In addition to cytokines, DENV infection of DCs resulted in are less efficient at inducing TH1 responses. We observed upregu- http://www.jimmunol.org/ robust chemokine expression of CCL2, CCL3, CCL4, and CCL5. lation of costimulatory molecules and IFN responses upon DENV CCL2 and CCL4 are strong chemoattractants for monocytes, infection, and these factors are known to promote TH1 differenti- whereas CCL3 potently attracts neutrophils to sites of inflamma- ation. Because Chase et al. (46) observed some T cell polarization, tion. Early during DENV infection, monocytes are recruited to the although to a lesser extent, it is possible that differences in assays skin where they differentiate into DCs and, thereby, increase the can explain the apparent disparity between our studies. pool of susceptible cells for DENV (41). This could lead to a self- Interestingly, only RNA viruses can induce viral hemorrhagic sustained loop in which DENV-infected DCs attract monocytes fever, which could indicate that RNA sensors are key players in that differentiate into DCs, become infected, and recruit more viral hemorrhagic fever. TNF directly activates endothelial cells DCs. We show that DENV-induced CCL2 and CCL4 expression in and is critical for DENV-induced vascular leakage in animal by guest on September 27, 2021 DCs is under the control of RIG-I and MDA5. Therefore, inter- models, probably in concert with other factors (47, 48). We show fering with RIG-I and MDA5 function could prevent the recruit- that DENV stimulates TNF production by DCs to a much greater ment of susceptible cells to the site of infection. degree than does LPS. Notably, DENV-induced TNF responses Notably, our data strongly suggest that RIG-I and MDA-5 have depended critically on RIG-I and MDA5. Numerous other RNA partial nonredundant roles, because silencing of RIG-I or MDA-5 alone viruses that cause viral hemorrhagic fever activate RIG-I and/or already decreased cytokine responses, DC maturation, and chemokine MDA5, and RLR triggering could be a general mechanism un- expression. This is in accordance with previous studies in which RIG-I derlying viral hemorrhagic fever (49, 50). or MDA5 deficiency results in lower ISG induction in DENV-infected In conclusion, we have identified RIG-I and MDA5 as key cells, and double deficiency of RIG-I and MDA5 further decreases sensors of DENV infection in DCs and as critical players in DC- ISG expression (23). The nonredundancy could be explained by the dependent innate and adaptive immune responses against DENV. fact that RIG-I and MDA5 bind different RNA structures (31, 42). Many of the cytokines and chemokines produced by DCs in re- Furthermore, MAVS signaling depends on the cellular location and sponse to DENV are associated with disease severity and endo- size of the signaling complex, and activation of RIG-I and MDA5 thelial dysfunction. Hence, molecules targeting RIG-I and MDA5 could be required for maximum induction of MAVS signaling (43). signaling components have the potential to be used as therapeutic Studies investigating RLR function in adaptive immune re- agents for DHF, DSS, and possibly other viruses. sponses are scarce, but RLR activation can inhibit TH1 polarization in response to TLR ligands in mice (44). IL-12p70 plays a vital Acknowledgments role in TH1 responses, and RLR-dependent IRF3 activation blocks We thank K. Luhn (University of Oxford, Oxford, U.K.) for kindly provid- IL-12p40 mRNA expression and, thereby, prevents IL-12p70 ing DENV-2/16681, C. Verweij (VU University Medical Center, Amster- formation (44). We also did not observe IL-12p40 induction by dam, the Netherlands) for providing IFN-b and MxA primers, and DENV, in contrast to IL-12p35, which resulted in undetectable M. van Hemert (Leiden University Medical Center, Leiden, the Netherlands) levels of mature IL-12p70. However, our data show that DENV for kindly providing C6/36 cells. infection of DCs did result in TH1 formation, and this was dependent on the RLR-adapted protein MAVS. Also, direct stimu- Disclosures lation of RIG-I and MDA5 with chemical ligands resulted in TH1 The authors have no financial conflicts of interest. polarization. Other factors in addition to IL-12p70 have the potential to induce TH1 polarization, including IFN-b (45). Our data show that DENV induces functional type I IFN responses in DCs, References 1. Bhatt, S., P. W. Gething, O. J. Brady, J. P. Messina, A. W. Farlow, C. L. Moyes, and RLR-dependent IFN-b could be driving TH1 differentiation J. M. Drake, J. S. Brownstein, A. G. Hoen, O. Sankoh, et al. 2013. The global by DENV-infected DCs. Chase et al. (46) suggest that infected DCs distribution and burden of dengue. Nature 496: 504–507. 8 RLRs DRIVE DENV IMMUNE ACTIVATION

2. Jaenisch, T., D. T. Tam, N. T. Kieu, T. Van Ngoc, N. T. Nam, N. Van Kinh, 25. Seth, R. B., L. Sun, C.-K. Ea, and Z. J. Chen. 2005. Identification and charac- S. Yacoub, N. Chanpheaktra, V. Kumar, L. L. See, et al. 2016. Clinical evaluation terization of MAVS, a mitochondrial antiviral signaling protein that activates of dengue and identification of risk factors for severe disease: protocol for a NF-kappaB and IRF 3. Cell 122: 669–682. multicentre study in 8 countries. BMC Infect. Dis. 16: 120–131. 26. Srikiatkhachorn, A., and J. F. Kelley. 2014. Endothelial cells in dengue hem- 3. Tisoncik, J. R., M. J. Korth, C. P. Simmons, J. Farrar, T. R. Martin, and orrhagic fever. [Published erratum appears in 2015 Antiviral Res. 114: 47] An- M. G. Katze. 2012. Into the eye of the cytokine storm. Microbiol. Mol. Biol. Rev. tiviral Res. 109: 160–170. 76: 16–32. 27. Geijtenbeek, T. B. H., and S. I. Gringhuis. 2016. C-type lectin receptors in the 4. John, D. V., Y.-S. Lin, and G. C. Perng. 2015. Biomarkers of severe dengue control of T helper cell differentiation. Nat. Rev. Immunol. 16: 433–448. disease - a review. J. Biomed. Sci. 22: 83–90. 28. Lim, W., K. Gee, S. Mishra, and A. Kumar. 2005. Regulation of B7.1 costim- 5. Tassaneetrithep, B., T. H. Burgess, A. Granelli-Piperno, C. Trumpfheller, ulatory molecule is mediated by the IFN regulatory factor-7 through the acti- J. Finke, W. Sun, M. A. Eller, K. Pattanapanyasat, S. Sarasombath, D. L. Birx, vation of JNK in lipopolysaccharide-stimulated human monocytic cells. et al. 2003. DC-SIGN (CD209) mediates dengue virus infection of human J. Immunol. 175: 5690–5700. dendritic cells. J. Exp. Med. 197: 823–829. 29. Zou, G. M., and W. Y. Hu. 2005. LIGHT regulates CD86 expression on dendritic 6. Pham, A. M., R. A. Langlois, and B. R. TenOever. 2012. Replication in cells of cells through NF-kappaB, but not JNK/AP-1 signal transduction pathway. hematopoietic origin is necessary for Dengue virus dissemination. PLoS Pathog. J. Cell. Physiol. 205: 437–443. 8: e1002465. 30. Simmons, D. P., P. A. Wearsch, D. H. Canaday, H. J. Meyerson, Y. C. Liu, 7. Marovich, M., G. Grouard-Vogel, M. Louder, M. Eller, W. Sun, S. J. Wu, Y. Wang, W. H. Boom, and C. V. Harding. 2012. Type I IFN drives a distinctive R. Putvatana, G. Murphy, B. Tassaneetrithep, T. Burgess, et al. 2001. Human dendritic cell maturation phenotype that allows continued class II MHC syn- dendritic cells as targets of dengue virus infection. J. Investig. Dermatol. Symp. thesis and antigen processing. J. Immunol. 188: 3116–3126. Proc. 6: 219–224. 31. Takeuchi, O., and S. Akira. 2008. MDA5/RIG-I and virus recognition. Curr. 8. Miller, J. L., B. J. M. de Wet, L. Martinez-Pomares, C. M. Radcliffe, Opin. Immunol. 20: 17–22. R. A. Dwek, P. M. Rudd, and S. Gordon. 2008. The mannose receptor mediates 32. Mazzon, M., M. Jones, A. Davidson, B. Chain, and M. Jacobs. 2009. Dengue dengue virus infection of macrophages. [Published erratum appears in 2008 virus NS5 inhibits interferon-alpha signaling by blocking signal transducer and PLoS Pathog. 4] PLoS Pathog. 4: e17. activator of transcription 2 phosphorylation. J. Infect. Dis. 200: 1261–1270. 9. Sun, P., S. Fernandez, M. A. Marovich, D. R. Palmer, C. M. Celluzzi, 33. Morrison, J., M. Laurent-Rolle, A. M. Maestre, R. Rajsbaum, G. Pisanelli, K. Boonnak, Z. Liang, H. Subramanian, K. R. Porter, W. Sun, and T. H. Burgess. V. Simon, L. C. Mulder, A. Fernandez-Sesma, and A. Garcıá-Sastre. 2013. Downloaded from 2009. Functional characterization of ex vivo blood myeloid and plasmacytoid Dengue virus co-opts UBR4 to degrade STAT2 and antagonize type I interferon dendritic cells after infection with dengue virus. Virology 383: 207–215. signaling. PLoS Pathog. 9: e1003265. 10. Lozach, P.-Y., L. Burleigh, I. Staropoli, E. Navarro-Sanchez, J. Harriague, 34. Tal, M. C., M. Sasai, H. K. Lee, B. Yordy, G. S. Shadel, and A. Iwasaki. 2009. J.-L. Virelizier, F. A. Rey, P. Despre`s, F. Arenzana-Seisdedos, and A. Amara. 2005. Absence of autophagy results in reactive oxygen species-dependent amplifica- Dendritic cell-specific intercellular adhesion molecule 3-grabbing non- tion of RLR signaling. Proc. Natl. Acad. Sci. USA 106: 2770–2775. integrin (DC-SIGN)-mediated enhancement of dengue virus infection is 35. Poeck, H., M. Bscheider, O. Gross, K. Finger, S. Roth, M. Rebsamen, independent of DC-SIGN internalization signals. J. Biol. Chem. 280: 23698– N. Hannesschla¨ger, M. Schlee, S. Rothenfusser, W. Barchet, et al. 2010. Rec-

23708. ognition of RNA virus by RIG-I results in activation of CARD9 and inflam- http://www.jimmunol.org/ 11. Palmer, D. R., P. Sun, C. Celluzzi, J. Bisbing, S. Pang, W. Sun, M. A. Marovich, masome signaling for interleukin 1 beta production. Nat. Immunol. 11: 63–69. and T. Burgess. 2005. Differential effects of dengue virus on infected and by- 36. Dinarello, C. A. 2011. Interleukin-1 in the pathogenesis and treatment of in- stander dendritic cells. J. Virol. 79: 2432–2439. flammatory diseases. Blood 117: 3720–3732. 12. Alayli, F., and F. Scholle. 2016. Dengue virus NS1 enhances viral replication and 37. Netea, M. G., B. J. Kullberg, and J. W. Van der Meer. 2000. Circulating cyto- pro-inflammatory cytokine production in human dendritic cells. Virology 496: kines as mediators of fever. Clin. Infect. Dis. 31(Suppl. 5): S178–S184. 227–236. 38. Gringhuis, S. I., T. M. Kaptein, B. A. Wevers, B. Theelen, M. van der Vlist, 13. Tsai, Y.-T., S.-Y. Chang, C.-N. Lee, and C.-L. Kao. 2009. Human TLR3 rec- T. Boekhout, and T. B. Geijtenbeek. 2012. Dectin-1 is an extracellular pathogen ognizes dengue virus and modulates viral replication in vitro. Cell. Microbiol. sensor for the induction and processing of IL-1b via a noncanonical caspase-8 11: 604–615. inflammasome. Nat. Immunol. 13: 246–254. 14. da Conceiçaõ,T.M.,N.M.Rust,A.C.E.R.Berbel,N.B.Martins,C.A.do 39. Wu, M. F., S. T. Chen, A. H. Yang, W. W. Lin, Y. L. Lin, N. J. Chen, I. S. Tsai, Nascimento Santos, A. T. Da Poian, and L. B. de Arruda. 2013. Essential role L. Li, and S. L. Hsieh. 2013. CLEC5A is critical for dengue virus-induced

of RIG-I in the activation of endothelial cells by dengue virus. Virology 435: inflammasome activation in human macrophages. Blood 121: 95–106. by guest on September 27, 2021 281–292. 40. Hottz, E. D., J. F. Lopes, C. Freitas, R. Valls-de-Souza, M. F. Oliveira, 15. Chen, S.-T., Y.-L. Lin, M.-T. Huang, M.-F. Wu, S.-C. Cheng, H.-Y. Lei, C.-K. Lee, M. T. Bozza, A. T. Da Poian, A. S. Weyrich, G. A. Zimmerman, F. A. Bozza, and T.-W. Chiou, C.-H. Wong, and S.-L. Hsieh. 2008. CLEC5A is critical for dengue- P. T. Bozza. 2013. Platelets mediate increased endothelium permeability in virus-induced lethal disease. Nature 453: 672–676. dengue through NLRP3-inflammasome activation. Blood 122: 3405–3414. 16. Nasirudeen, A. M., H. H. Wong, P. Thien, S. Xu, K.-P. Lam, and D. X. Liu. 2011. 41. Schmid, M. A., and E. Harris. 2014. Monocyte recruitment to the dermis and RIG-I, MDA5 and TLR3 synergistically play an important role in restriction of differentiation to dendritic cells increases the targets for dengue virus replication. dengue virus infection. PLoS Negl. Trop. Dis. 5: e926. PLoS Pathog. 10: e1004541. 17. Lambeth, C. R., L. J. White, R. E. Johnston, and A. M. De Silva. 2005. Flow 42. Pichlmair, A., O. Schulz, C. P. Tan, T. I. Nasland, P. Liljestrom, F. Weber, and cytometry-based assay for titrating dengue virus. J. Clin. Microbiol. 43: C. Reis e Sousa. 2006. RIG-I–mediated antiviral responses to single-stranded 3267–3272. RNA bearing 59-phosphates. Science 314: 997–1001. 18. Alcaraz-Estrada, S. L., M. Yocupicio-Monroy, and R. M. del Angel. 2010. In- 43. Hou, F., L. Sun, H. Zheng, B. Skaug, Q. X. Jiang, and Z. J. Chen. 2011. MAVS sights into dengue virus genome replication. Future Virol. 5: 575–592. forms functional prion-like aggregates to activate and propagate antiviral innate 19. Cerny, D., M. Haniffa, A. Shin, P. Bigliardi, B. K. Tan, B. Lee, M. Poidinger, immune response. Cell 146: 448–461. E. Y. Tan, F. Ginhoux, and K. Fink. 2014. Selective susceptibility of human 44. Negishi, H., H. Yanai, A. Nakajima, R. Koshiba, K. Atarashi, A. Matsuda, skin antigen presenting cells to productive dengue virus infection. PLoS K. Matsuki, S. Miki, T. Doi, A. Aderem, et al. 2012. Cross-interference of RLR Pathog. 10: e1004548. and TLR signaling pathways modulates antibacterial T cell responses. Nat. 20. Flipse, J., S. Torres, M. Diosa-Toro, H. van der Ende-Metselaar, J. Herrera- Immunol. 13: 659–666. Rodriguez, S. Urcuqui-Inchima, A. Huckriede, I. A. Rodenhuis-Zybert, and 45. Rodriguez-Madoz, J. R., D. Bernal-Rubio, D. Kaminski, K. Boyd, and J. M. Smit. 2016. Dengue tropism for macrophages and dendritic cells: the host A. Fernandez-Sesma. 2010. Dengue virus inhibits the production of type I in- cell effect. J. Gen. Virol. 97: 1531–1536. terferon in primary human dendritic cells. J. Virol. 84: 4845–4850. 21. Olagnier, D., S. Peri, C. Steel, N. van Montfoort, C. Chiang, V. Beljanski, 46. Chase, A. J., F. A. Medina, and J. L. Munõz-Jordań. 2011. Impairment of CD4+ M. Slifker, Z. He, C. N. Nichols, R. Lin, et al. 2014. Cellular oxidative stress T cell polarization by dengue virus-infected dendritic cells. J. Infect. Dis. 203: response controls the antiviral and apoptotic programs in dengue virus-infected 1763–1774. dendritic cells. PLoS Pathog. 10: e1004566. 47. Phanthanawiboon, S., K. Limkittikul, Y. Sakai, N. Takakura, M. Saijo, and 22. Nightingale, Z. D., C. Patkar, and A. L. Rothman. 2008. Viral replication and T. Kurosu. 2016. Acute systemic infection with dengue virus leads to vascular paracrine effects result in distinct, functional responses of dendritic cells fol- leakage and death through tumor necrosis factor-a and Tie2/angiopoietin signaling lowing infection with dengue 2 virus. J. Leukoc. Biol. 84: 1028–1038. in mice lacking type I and II interferon receptors. PLoS One 11: e0148564. 23. Loo, Y.-M., J. Fornek, N. Crochet, G. Bajwa, O. Perwitasari, L. Martinez- 48. Shresta, S., K. L. Sharar, D. M. Prigozhin, P. R. Beatty, and E. Harris. 2006. Sobrido, S. Akira, M. A. Gill, A. Garcıá-Sastre, M. G. Katze, and M. Gale, Jr. Murine model for dengue virus-induced lethal disease with increased vascular 2008. Distinct RIG-I and MDA5 signaling by RNA viruses in innate immunity. permeability. J. Virol. 80: 10208–10217. J. Virol. 82: 335–345. 49. Zinzula, L., and E. Tramontano. 2013. Strategies of highly pathogenic RNA 24. van Cleef, K. W. R., G. J. Overheul, M. C. Thomassen, S. J. F. Kaptein, viruses to block dsRNA detection by RIG-I-like receptors: hide, mask, hit. An- A. D. Davidson, M. Jacobs, J. Neyts, F. J. M. van Kuppeveld, and R. P. van Rij. tiviral Res. 100: 615–635. 2013. Identification of a new dengue virus inhibitor that targets the viral NS4B 50. Kell, A. M., and M. Gale, Jr. 2015. RIG-I in RNA virus recognition. Virology protein and restricts genomic RNA replication. Antiviral Res. 99: 165–171. 479–480: 110–121.