Negative-Strand RNA Viruses Maturation and Adaptive Immunity

The Journal of Immunology

TLR-Independent Induction of Dendritic Cell Maturation and Adaptive Immunity by Negative-Strand RNA Viruses1

Carolina B. Lo´pez,* Bruno Moltedo,* Lena Alexopoulou,2† Laura Bonifaz,‡ Richard A. Flavell,† and Thomas M. Moran3*

TLR signaling leads to dendritic cell (DC) maturation and immunity to diverse pathogens. The stimulation of TLRs by conserved viral structures is the only described mechanism leading to DC maturation after a virus infection. In this report, we demonstrate that mouse myeloid DCs mature normally after in vivo and in vitro infection with Sendai virus (SeV) in the absence of TLR3, 7, 8, or 9 signaling. DC maturation by SeV requires virus replication not necessary for TLR-mediated triggering. Moreover, DCs deficient in TLR signaling efficiently prime for Th1 immunity after infection with influenza or SeV, generating IFN-␥-producing T cells, CTLs and antiviral Abs. We have previously demonstrated that SeV induces DC maturation independently of the presence of type I IFN, which has been reported to mature DCs in a TLR-independent manner. The data presented here provide evidence for the existence of a novel intracellular pathway independent of TLR-mediated signaling responsible for live virus triggering of DC maturation and demonstrate its critical role in the onset of antiviral immunity. The revelation of this pathway should stimulate invigorating research into the mechanism for virus-induced DC maturation and immunity. The Journal of Immunology, 2004, 173: 6882–6889.

endritic cells (DCs)4 mediate the transition from innate TLR8 (7–9). Viral envelope proteins (e.g., the respiratory syncytial to adaptive immunity by undergoing a major shift in virus (RSV) fusion protein) have also been shown to trigger TLR D their maturation state that involves the up-regulation of activation (10–13). However, the lack of a common structure or costimulatory molecules, and the secretion of proinflammatory cy- PAMP within these proteins questions the role of these interactions tokines necessary for the differentiation of effector and memory T in the generic recognition of viruses. cells and B cells (1). The binding of pathogen-associated molec- Despite the triggering of DC maturation through the binding of ular patterns (PAMPs) to an assortment of receptors expressed on viral components to TLRs, the role of TLR signaling in eliciting the DCs, known as TLRs, constitute the primary mechanism de- antiviral adaptive immune responses is unknown. In humans, the scribed to trigger DC maturation (2). To date, more than 10 dif- virus-responsive TLR7, TLR8, and TLR9 are expressed in a re- ferent TLRs have been identified in vertebrates, with specificities stricted subset of DCs, the plasmacytoid DCs (pDCs), which are by guest on October 1, 2021. Copyright 2004 Pageant Media Ltd. for all major families of microbes (3). specialized in secreting large amounts of type I IFN after virus A role for TLRs in monitoring viral infections has been estab- recognition (14, 15). Controversy exists regarding the ability of lished based on the discovery of TLRs responsive to viral ele- these cells to stimulate naive T cells. Although CpG-matured ments. Double-stranded RNA (dsRNA), a viral PAMP generated pDCs are able to present Ag and stimulate naive T cells in vitro as during virus replication, binds and activates TLR3 and CpG motifs well as in vivo (16–18), evidence also exists suggesting that pDCs found in the DNA of HSV-activator TLR9 (5, 6). Moreover, sin- poorly present Ag and are unable to prime naive T cells (16, 19). gle-stranded viral RNA (ssRNA), representing the genomic mate- In addition, all virus-related TLRs can be stimulated by inactivated rial of many different viruses, triggers the activation of TLR7 and viruses or viral RNA surrogates (4, 7, 8), but live virus infection is

http://classic.jimmunol.org needed for the optimal induction of DC maturation and immunity *Department of Microbiology, Mount Sinai School of Medicine, New York, NY against ssRNA viruses (20, 21). These observations argue that the † 10029; Section of Immunobiology, Yale University School of Medicine and Howard predominant role for TLR signaling during virus infection might Hughes Medical Institute, New Haven, CT 06529; and ‡Laboratory of Cellular Phys- iology and Immunology, The Rockefeller University, New York, NY 10021 be to elicit innate immune responses rather than to generate adap- Received for publication July 13, 2004. Accepted for publication September 22, 2004. tive immunity. The costs of publication of this article were defrayed in part by the payment of page Thus, we set out to investigate whether the induction of DC charges. This article must therefore be hereby marked advertisement in accordance

Downloaded from maturation and adaptive immunity to viral infections could be trig- with 18 U.S.C. Section 1734 solely to indicate this fact. gered independently of TLR signaling. To do this, we took advan- 1 This work was supported by Grants 1R01AI41111 and 1PO1A148204 from the tage of the paramyxovirus Sendai virus (SeV) strain Cantell, a National Institute of Allergy and Infectious Diseases (to T.M.M.) and by funds granted by the Charles H. Revson Foundation (to C.B.L.). The statements made and particularly strong inducer of mouse DC maturation (22). We have views expressed, however, are solely the responsibility of the authors. previously demonstrated that the potent DC maturation-inducing 2 Current address: Centre d’Immunologie de Marseille-Luminy, Centre National de la activity of SeV is not dependent on the secretion of the antiviral Recherche Scientiﬁque-Institut National de la Sante´ et de la Recherche Me´dicale, Campus de Luminy, Case 906, 13288 Marseille Cedex 9, France. cytokine type I IFN, which has been related to the ampliﬁcation and preservation of DC maturation after infection with other vi- 3 Address correspondence and reprint requests to Dr. Thomas M. Moran, Department of Microbiology, Mount Sinai School of Medicine, One Gustave L. Levy Place, New ruses, such as Newcastle disease virus (21). SeV allowed us to York, NY 10029. E-mail address: [email protected] visualize the requirements for DC maturation after infection with 4 Abbreviations used in this paper: DC, dendritic cell; BMDC, bone marrow-derived live viruses independently of the contribution of type I IFN sig- DC; MOI, multiplicity of infection; PAMP, pathogen-associated molecular pattern; pDC, plasmacytoid DC; RSV, respiratory syncytial virus; SeV, Sendai virus; NP, naling. Our data demonstrate that TLR3 or the TLR adaptor pro- nucleoprotein. tein MyD88, necessary for the signaling of most TLRs, are not

essential for the induction of costimulatory molecule expression or anti-CD86 or anti-CD80 Abs (BD Biosciences). Flow cytometry was per- the secretion of cytokines by mature DCs after in vitro or in vivo formed in a XLS-Coulter station (Beckman Coulter, Miami, FL). Data was infection and that their absence has no significant impact on the analyzed with Flowjo software. generation of antiviral Th1 immunity in mice. Mice infection, immunization, and Ab detection Mice were exposed for 30 min to ϳ900 influenza virus particles/min/cm3 Materials and Methods of air in an infection chamber (Glas-Col, Terre Haute, IN). For immuni- ϫ 5 7 Viruses and mice zations, mice were injected i.p. with 4 10 DCs or with 10 TCID50 of influenza or SeV. Detection of antiviral IgG1 and IgG2b Abs in mouse SeV Cantell and influenza A virus were grown, purified, and inactivated as serum was performed by capture ELISA as described (20). described (20, 22). RSV (A2 strain) was grown in mycoplasma-free Vero Ϫ/Ϫ cells. C57BL/6 and MHC class I mice were purchased from Taconic Spleen cell culture and cytotoxicity Farms (Germantown, NY). MyD88Ϫ/Ϫ mice were kindly provided by Dr. S. Akira (Osaka University, Osaka, Japan) and Dr. R. Steinman (Rock- Stimulator splenocytes from naive C57BL/6 or MyD88Ϫ/Ϫ mice were in- efeller University, New York, NY). Age- and sex-matched animals were fected with influenza or SeV at an MOI of 40 for 45 min at 37°C and used in all experiments. The animals were housed in pathogen-free con- cultured with spleen cells from immunized animals at a 1:10 stimulator to ditions and the experiments were performed according to institutionally responder ratio for SeV-infected cells and a 1:1 stimulator to responder approved protocols. ratio for influenza virus-infected cells (20). Effector cells from secondary in vitro cultures were mixed with 51Cr-labeled infected or mock-infected EL4 Bone marrow-derived DC (BMDC) preparation and treatment target cells as described (20). Supernatants were harvested and gamma radiation was measured. Killing of mock-infected EL4 cells was subtracted BMDCs were prepared as previously described (22) and infected with ei- from that observed with infected targets. ther influenza (multiplicity of infection, (MOI) of 10), SeV (MOI of 0.6– 10) or RSV (MOI of 2). Control BMDCs were mock infected or incubated Cytokine detection overnight with 10 ng/ml LPS (Sigma-Aldrich, St. Louis, MO) or 50 ␮g/ml poly(I:C) (Amersham Biosciences, Bucks, U.K.). For maturation studies Supernatants from DC and spleen cell cultures were analyzed by Capture the cells were incubated overnight at 1 ϫ 106 cells/ml in media containing ELISA for IL-6, IL-1␤, IL-12p40, TNF-␣, IL-4 (R&D Systems, Minne- 1% normal mouse serum. Removal of remaining extracellular virus parti- apolis, MN), and IFN-␣ (PBL Biomedical Laboratories, Piscataway, NJ) cles was performed as described using neutralizing anti-PR8 influenza following the manufacturer’s protocols. IFN-␥ was measured using Ab (PY102 and PY210) or anti-SeV Abs (obtained from Dr. A. Portner, St. pairs from BD Biosciences. Jude Children’s Hospital, Memphis, TN) (23). In vivo CTL assay and tetramer staining DC isolation from lymph nodes and staining Splenocytes from naive mice were pulsed at 4 ϫ 107 cells/ml with 20 ␮M

Lung-draining lymph nodes were harvested 48 h after mice infection. Tis- SeV nucleoprotein (NP)324–332 peptide or without peptide in PBS for 15 sues were incubated 20 min with collagenase (Liberase Blendzymes; min at room temperature. The cells were then labeled at 2 ϫ 107 cells/ml Roche, Indianapolis, IN). Washed single cell suspensions were incubated with different concentrations of CFSE (Molecular Probes, Eugene, OR; 2.5 for 4 h with brefeldin A (BD Biosciences, San Diego, CA). Cells were or 0.125 ␮M) at 37°C for 30 min. Labeling was stopped with an equal stainedwithallophycocyanin-conjugatedanti-CD11candFITC-labeledanti- volume of FBS for 1 min and 1 ϫ 107 cells of each peptide pulsed and CD86 Abs (BD Biosciences). Intracellular staining for IL-12 was per- unpulsed were mixed and injected i.v. into infected and uninfected syn- formed using the Cytoﬁx/Cytoperm kit from BD Biosciences and PE-con- genic mice. After 20 h from the injections, spleens were harvested and jugated anti-IL-12 Ab (BD Biosciences). Infected, mock-infected, or LPS- single cell suspensions were prepared and analyzed by ﬂow cytometry. Ϫ

by guest on October 1, 2021. Copyright 2004 Pageant Media Ltd. treated BMDCs were stained 24 h after treatment with FITC-conjugated Percentage of speciﬁc killing was calculated as described (24): 100 [(% http://classic.jimmunol.org Downloaded from

FIGURE 1. TLR3Ϫ/Ϫ DCs mature normally after infection with SeV. C57BL/6 (wt) and TLR3Ϫ/Ϫ DCs (ko) were either infected with SeV or RSV or treated with poly(I:C) or LPS. After 24 h of culture, the cells were stained for CD86 or CD80 expression (ﬁlled histogram) and analyzed by ﬂow cytometry (a). Thin line represents isotype controls. Cytokines in the culture supernatants were analyzed after 24 h of culture by capture ELISA (b). 6884 TLR-INDEPENDENT INDUCTION OF DC MATURATION BY VIRUSES

FIGURE 2. MyD88Ϫ/Ϫ DCs mature normally after infection with SeV. C57BL/6 and MyD88Ϫ/Ϫ DCs were infected with either live or UV-inactivated SeV or RSV, or treatment with LPS. After 24 h of culture, the cells were stained for CD86 or CD80 expression (ﬁlled histogram) and analyzed by ﬂow cytometry (a). Thin line represents isotype controls. Cytokines in the culture supernatants were analyzed after 6 (c)or24(a, c, and d) h of culture by capture ELISA (b). Data are representative of more than three experiments.

peptide pulsed in infected mice/% unpulsed in infected mice)/(% peptide with wild-type DCs, mimicking the effect seen when TLR3-inde- pulsed in uninfected/% unpulsed in uninfected)] ϫ 100. To determine the pendent stimuli (LPS or RSV) are used. These results demonstrate percentage of T cells bearing the specific anti-SeV-NP peptide 324–332 that TLR3 stimulation is not necessary for the triggering of DC H-2Kb receptor, splenocytes or lung single cell suspensions were obtained by guest on October 1, 2021. Copyright 2004 Pageant Media Ltd. 10–12 days after infection with SeV and costained with anti-CD8 FITC Ab maturation after infection with SeV. b and PE-labeled H-2K tetramers carrying the SeV NP324–332 peptide (Na- tional Institutes of Health tetramer core facility). DC maturation after SeV infection is independent of MyD88- Statistical analysis mediated signaling Statistical analysis was performed by using t test, paired two sample In mice, TLR7, TLR8, and TLR9 are expressed in myeloid DCs, analysis. pivotal in the initiation of adaptive immunity (29). To test the requirement for TLR7, 8, or 9 stimulation in the induction of my- Results eloid DC maturation, we took advantage of mice deficient in http://classic.jimmunol.org TLR3 signaling is not required for DC maturation after MyD88, an adaptor protein essential for their signaling (7, 8). infection with SeV MyD88 binds to members of the IL-1R-associated protein kinases The first evidence for viral-induced TLR signaling was the dem- and signals further via the TNFR-associated factor-6 (7, 9). onstration that the synthetic dsRNA analog poly(I:C), as well as MyD88Ϫ/Ϫ DCs show normal up-regulation of CD80 and CD86 isolated viral dsRNA, could induce the expression of costimulatory following SeV infection (Fig. 2a). IL-12, IL-6, TNF, and IFN-␣ molecules on B cells and the secretion of cytokines from APCs secretion by infected DCs is also completely independent of

Downloaded from through TLR3 signaling (4). Since then, a number of seemingly MyD88 after SeV infection (Fig. 2b). Additionally, and in agree- contradictory studies have attempted to define the role of TLR3 in ment with previous reports (22), the induction of proinflammatory dsRNA-induced cell activation (8, 21, 25–27). Concurring with cytokine secretion by SeV-infected DCs depends on virus replica- reports by others (8, 21, 25), we observe that the up-regulation of tion as shown by the absence of cytokines in response to UV- costimulatory molecules and the secretion of IFN-␣ are indepen- inactivated virus (Fig. 2b). In contrast, the induction of cytokine dent of TLR3 signaling when poly(I:C) is used to treat DCs. In secretion by the TLR-activators RSV and LPS is completely abol- contrast, the secretion of IL-12 and TNF is impaired in the absence ished in MyD88Ϫ/Ϫ cells and RSV replication is not a requirement of this receptor (Fig. 1). for the stimulation of cytokine secretion by DCs (Fig. 2b). The To test the role of TLR3 in the triggering of DC maturation after independence of MyD88 signaling for the secretion of type I IFN live virus infection, DCs were infected with SeV. LPS and RSV and other proinflammatory cytokines and for the up-regulation of were used as controls for TLR3-independent induction of DC mat- costimulatory molecules, is observed as early as 6 h postinfection uration as they have been shown to trigger the secretion of cyto- (Fig. 2c and data not shown) and is independent of the dose of SeV kines through binding to TLR4 (10, 28). As shown in Fig. 1, nor- used (Fig. 2d and data not shown). These data demonstrate that mal costimulatory molecule up-regulation and cytokine secretion signaling by the adaptor protein MyD88 is not necessary for the is observed after SeV infection of TLR3Ϫ/Ϫ DCs when compared induction of DC maturation after infection with SeV. The Journal of Immunology 6885

(NP324–332), in vivo, 7 days after the infection (Fig. 4a). However, less efficient killing is observed in MyD88Ϫ/Ϫ mice consistent with a reduction in the number of CD8ϩ cells bearing SeV-NP-specific TCR, as detected by staining with NP-peptide bearing tetramers (Fig. 4b). Wild-type and MyD88Ϫ/Ϫ mice develop equivalent amounts of total virus-specific IgG Abs (data not shown) with a slight bias toward the Th2-isotype IgG1 in MyD88Ϫ/Ϫ mice (Fig. 4c). Our laboratory has a well-characterized in vivo model for influenza infection (31). We took advantage of this model to study the role of MyD88 signaling in the development of virus-specific humoral and memory cellular responses. After infection with aerosolized live influenza virus, known to induce potent Th1 immunity (31, 32), control and MyD88Ϫ/Ϫ mice develop equivalent amounts of total virus-specific IgG Abs (data not shown) with a slight bias toward the Th2-isotype IgG1 in MyD88Ϫ/Ϫ mice (Fig. 4d), similar to that observed in SeV-infected animals (Fig. 4c). Moreover, splenocytes from MyD88Ϫ/Ϫ and wild-type mice restimulated in vitro with influenza virus 14 days after the infection, secrete com- parable levels of the Th1 cytokine IFN-␥ together with a remark- able production of the Th2 cytokine IL-4 only by MyD88Ϫ/Ϫ cells (Fig. 4e). A small reduction in the clearance of both viruses is observed in MyD88Ϫ/Ϫ animals when compared with wild-type mice 7 days postinfection (data not shown), but both groups of animals show complete recovery from infection as measured by recovery of lost weight and virus clearance from the lungs 10 days postinfection (data not shown). Interestingly, despite the secretion of IL-4 and the Th2-biased Ab production, normal development of memory CTL precursors is observed in MyD88Ϫ/Ϫ mice (Fig. 4f). Thus, although Ϫ/Ϫ FIGURE 3. Normal maturation of MyD88Ϫ/Ϫ DCs after SeV infection MyD88 mice show minor differences in the in vivo response to in vivo. C57BL/6 (wt) and MyD88Ϫ/Ϫ mice were infected intranasally with both, SeV and influenza virus, a normal immune adaptive response is

10 ID50 of SeV. Two days later, the lung-draining lymph nodes were col- generated that leads to complete recovery from infection. lected and stained for FACS analysis. CD11cϩ cells were gated. Cells from mock-infected (gray line) or infected (black line) animals were stained for Virus-infected DCs deficient in MyD88 signaling induce CD86 (a). Isotype control is represented in the filled histogram. Cells were polarized Th1 immune response in mice also intracellularly stained for IL-12p40/p70 (b). The percentage of IL- by guest on October 1, 2021. Copyright 2004 Pageant Media Ltd. MyD88 is a required element in the signaling through IL-1 and 12p40/p70 expression over the basal level presented on noninfected cells is IL-18 receptors, both members of the Toll/IL-1R containing re- shown in c. ceptor family (33, 34). Therefore, the less efficient development of CD8ϩ cells bearing SeV-NP-specific TCR (Fig. 4b) and a nonpolarized memory immune response in MyD88Ϫ/Ϫ mice may result DCs mature normally after SeV infection in MyD88Ϫ/Ϫ mice from a failure of these cytokines to promote the differentiation of To corroborate our in vitro observations, we analyzed the induc- CD4ϩ T cells into a Th1 phenotype (35). To unambiguously de- tion of DC maturation, in vivo, after intranasal infection of termine the role of MyD88 in the initiation of antiviral immunity, Ϫ Ϫ MyD88 / mice with SeV. Two days after infection with influ- its function in APCs must be distinguished from its effect in T http://classic.jimmunol.org enza or SeV, mature DCs with up-regulated costimulatory mole- cells. We immunized mice by adoptively transferring ex vivo in- cules and producing proinflammatory cytokines, can be obtained fected DCs to distinguish the effect of MyD88 on these cells from from the lymph nodes of infected animals (Ref. 30 and data not that on T cells. DCs, which are abortively infected by influenza shown). As shown in Fig. 3a, mature DCs with up-regulated CD86 and SeV, have been shown to efficiently induce protective antiviral can be seen in the lung-draining lymph nodes of wild-type and immunity (23). BMDCs were infected with influenza or SeV and Ϫ Ϫ MyD88 / mice 2 days after infection with SeV. Moreover, a injected i.p. into wild-type animals. Both IgG1 and IgG2b Abs are Ϫ/Ϫ Downloaded from similar increase in the expression of IL-12 p40/p70 is observed in produced when control or MyD88 -infected DCs are used for Ϫ Ϫ DCs from wild-type and MyD88 / mice upon virus infection immunization (Fig. 5a). No evidence for a Th2 (IgG1)-biased re- despite a difference in the basal level of expression of this cytokine sponse is seen when immunizing with infected MyD88Ϫ/Ϫ DCs as among these animals (Fig. 3, b and c). These results demonstrate observed in directly immunized MyD88Ϫ/Ϫ mice (Fig. 4a). In ad- that following SeV infection, DCs equivalently matured can be dition, high levels of IFN-␥ were secreted by splenocytes from distinguished in the lung-draining lymph nodes of wild-type and animals immunized with infected wild-type or MyD88Ϫ/Ϫ DCs Ϫ Ϫ MyD88 / mice. (Fig. 5b), corresponding with the development of CTLs (Fig. 5c). We were unable to detect IL-4 production in any of the cultures. Ϫ/Ϫ ␥ MyD88 mice expand IFN- secreting cells and CTLs after Similarly, adoptively transferred TLR3Ϫ/Ϫ DCs efficiently induced virus infection antiviral CTLs when infected with influenza or SeV (data not To analyze the role of MyD88 signaling in the development of shown). antiviral immunity, we infected mice with SeV and analyzed the To validate this model, we analyzed the contribution of cross- primary generation of CTLs, as well as the development of anti- presentation to the immune response generated by this method. SeV Abs. Both wild-type and MyD88Ϫ/Ϫ mice expand CTLs that This phenomenon occurs when Ag is transferred from dead cells to eliminate target cells carrying the immunodominant SeV peptide endogenous cells which present antigenic peptides on their own 6886 TLR-INDEPENDENT INDUCTION OF DC MATURATION BY VIRUSES

FIGURE 4. MyD88Ϫ/Ϫ mice generate a nonpolarized immune response against influenza and SeVs. C57BL/6 (wt) and MyD88Ϫ/Ϫ (ko) mice were infected intranasally (i.n.) with SeV (a–c) or with aerosolized influenza virus (d–f). Six days after SeV infection, peptide pulsed and unpulsed CFSE-labeled cells were injected into the animals. Histograms show the number of CFSElow unpulsed or CSFEhigh peptide-pulsed cells 20 h after injection (a). Results are representative of two experiments. Average percentage of specific killing from two experiments is also shown. Ten days after infection lymphocytes by guest on October 1, 2021. Copyright 2004 Pageant Media Ltd. from the lung were stained for CD8ϩ cells bearing the SeV NP peptide-specific TCR using PE-labeled tetramers (b). Two weeks after virus infection, anti-SeV (c) and influenza virus (d) Abs were measured from the mouse sera (1/5000 and 1/1000 serum dilution shown for SeV and influenza virus, p Ͻ 0.006. Splenocytes from infected mice were restimulated in vitro (e and f). Cytokine secretion was determined by capture ELISA after ,ء ,(respectively 3 days of culture (e). CTL generation was determined after 5 days of culture by a 51Cr release assay (f).

MHC class I molecules and could bias the interpretation of results mans, the expression of these TLRs is restricted among the DC obtained when DCs are used for immunization (36). Control or populations to pDCs, and most studies analyzing their role in cell Ϫ Ϫ MHC class I / DCs were infected with SeV and injected into activation by viruses have only focused on this DC subpopulation Ϫ/Ϫ http://classic.jimmunol.org C57BL/6 mice. As shown in Fig. 6, a and b, MHC class I DCs (7, 8). In mouse, TLR7 and TLR8 are also expressed in myeloid are very inefﬁcient in generating CTL only partially eliminating DCs, the primary cells responsible for the induction of adaptive target cells carrying the immunodominant SeV peptide NP324–332. immunity (29). As signaling through these receptors has been ϩ This correlates with more than a 70% reduction of CD8 cells shown to be dependent on MyD88, our results using MyD88Ϫ/Ϫ bearing a TCR speciﬁc for the SeV NP324–332 peptide as detected BMDCs demonstrate that in mice TLR7 and TLR8 are not needed by staining with NP324–332-conjugated tetramers (Fig. 6c). These for maturation of myeloid DCs in response to SeV. The failure of

Downloaded from data demonstrate that at least 50% of the CTLs generated after DC inactivated SeV to trigger DC maturation strengthens the argument immunization are the result of direct presentation of viral Ags by that TLR7 and TLR8, which have been shown to be stimulated by the injected DCs, and allowed us to demonstrate that DCs deﬁcient inactivated virus, are not necessary for myeloid DC maturation. in TLR3 or MyD88 can induce normal antiviral adaptive immunity TLR3 that responds to the virus replication intermediary dsRNA, is in mice. also expressed on myeloid DCs. This receptor signals by MyD88- Discussion dependent and -independent pathways forcing us to investigate its role DC maturation, necessary for the transition from innate to adaptive in the induction of DC maturation by SeV separately from the MyD88 immunity, can be triggered by TLR signaling after recognition of dependent TLRs. Our results demonstrate that TLR3 signaling is not microbial structures. Because a number of TLRs can signal in re- needed for the induction of myeloid DC maturation by SeV and sponse to viruses, we sought to determine whether the virus-trig- agrees with accumulating data arguing that TLR3 signaling is not gered TLR signaling is essential for the maturation of myeloid DC, necessary for the induction of DC maturation and immunity by other Ϫ/Ϫ and the subsequent induction of antiviral immunity. viruses (9, 21, 37). Furthermore, DCs from MyD88 and wild-type TLR7 and TLR8 are receptors for ssRNA, the genetic material mice showed identical up-regulation of costimulatory molecules and of many viruses including those used in this study (7–9). In hu- an increase in IL-12 production following in vivo infection with SeV, The Journal of Immunology 6887

experiments adoptively transferring virus-infected MyD88Ϫ/Ϫ DCs into wild-type mice. No Th2 bias was observed under these conditions and MyD88Ϫ/Ϫ DCs induced cytokines, Ab and CTL generation are indistinguishable from the animals infused with virus-infected wild-type DCs. Studies using MHC IϪ/Ϫ DCs vali- dated the model system by showing that a signiﬁcant part of the response observed following adoptive transfer is elicited by the injected cells and not by cross-priming or in vivo uptake of virus or viral proteins. The independence of MyD88 signaling for the normal development of antiviral immunity concurs with the reported normal antiviral immunity in children carrying a mutation in IL-1R-associated protein kinase-4, a crucial downstream element in the signaling pathway that uses the adaptor MyD88 (38). Altogether, the data presented argue for a TLR-independent pathway for myeloid DC maturation that is dependent on virus replication. The cellular machinery responsible for the induction of myeloid DC maturation after virus infection remains elusive. Al- though the dsRNA-dependent protein kinase R has been shown to mediate these events upon transfection of poly(I:C), we and others have demonstrated that live virus-induced maturation is independent of protein kinase R (21, 22), arguing that the pathway acti- vated by replicating virus may be separate or at least more com- plex than what is triggered by dsRNA alone. Further studies are required to determine the role of other cellular proteins, such as the TANK (TNFR-associated factor-associated NF-␬B kinase) binding kinase I, the inhibitor of NF-␬B kinase ⑀ (39, 40), NOD (nu- cleotide-binding oligomerization domain) (41), and the recently described RNA helicase retinoic acid inducible gene I (42), which are known to be triggered by dsRNA, in the induction of DC maturation after virus infection. Additionally, type I IFN produced by all virus-infected cells as a means to contain viral replication and spreading has been shown to be required for optimal DC maturation after infection with New- castle disease virus (21). We have previously shown that SeV-

by guest on October 1, 2021. Copyright 2004 Pageant Media Ltd. induced maturation of myeloid DC is independent of exogenous type I IFN signaling (22). In addition, SeV efficiently blocks the IFN signaling pathway in infected cells (43). We have noted that viruses that strongly activate the type I IFN synthesis machinery, such as SeV strain Cantell, are also the best inducers of DC maturation (22) suggesting the existence of shared components in the induction pathways of type I IFN production and DC maturation. The difference in the need for exogenous type I IFN signaling for the complete maturation of DCs after different virus infections may http://classic.jimmunol.org FIGURE 5. Influenza and SeV-infected MyD88Ϫ/Ϫ DCs induce a po- depend on the strength of the viral activation signal. In light of larized Th1 immune response in mice. C57BL/6 mice were immunized i.p. these observations, DCs may require, in some instances, exoge- with ex vivo-infected C57BL/6 (f) or MyD88Ϫ/Ϫ (p) DCs. Control nous sources of type I IFN for optimal maturation. Type I IFN C57BL/6 (wt) mice were directly immunized with influenza (Flu) or SeV. released after TLR signaling not reliant on virus replication and Two weeks after virus infection anti-influenza or SeV Abs were measured therefore not inhibitable by type I IFN antagonists, may fulfill this from the mouse sera (1/1000 serum dilution shown) (a) and splenocytes requirement. This would suggest a novel physiologically relevant were restimulated in vitro. IFN-␥ secretion was determined by capture Downloaded from role for pDCs as endogenous adjuvants. ELISA after 3 days of culture (b). CTL generation was determined after 5 Despite our observations, a growing number of examples im- days of culture by a 51Cr release assay (c). plicate TLRs in the recognition of diverse viruses, including the binding of the RSV F protein to TLR4 (10) and the measles virus H protein to TLR2 (11). Interestingly, even though these interac- corroborating the dispensable role of MyD88 signaling in the induc- tions result in activation of APCs and the secretion of cytokines, tion of DC maturation after infection with SeV observed in vitro. they do not always result in the development of efficient antiviral To complement these studies, we show that the development of immunity. This is well exemplified by the ineffective immune re- an efficient antiviral adaptive immune response is independent of sponse generated against RSV (44, 45). Moreover, no defined TLR3 or MyD88-dependent activation. Although a slight Th2 bias common structure seems to be involved in the binding of these in the antiviral immune response is observed in MyD88Ϫ/Ϫ mice, viruses to TLRs leaving open the possibility that these obligatory probably resulting from a deficiency in IL-1 and IL-18 signaling, intracellular pathogens escape or conveniently use the scrutiny of viral clearance, the production of IFN-␥ by T cells, and cytotoxic TLRs. Interestingly, RSV induces a MyD88-dependent secretion T cell expansion are not significantly different from control ani- of cytokines by DCs but fails to up-regulate costimulatory mole- mals. The normal functioning of MyD88Ϫ/Ϫ DCs is supported by cules (Fig. 3). This unusual TLR-mediated maturation of DCs by 6888 TLR-INDEPENDENT INDUCTION OF DC MATURATION BY VIRUSES

FIGURE 6. MHC class I Ϫ/Ϫ DCs infected with SeV poorly induce CTLs. C57BL/6 mice were immunized i.p. with ex vivo-infected C57BL/6 (wt) or MHC class IϪ/Ϫ (MHC I ko) DCs 2 wk after infection, peptide pulsed and unpulsed CFSE-labeled cells were injected into the animals. Histograms show number of CFSElow unpulsed or CSFEhigh peptide-pulsed cells 20 h after injection (a). Average percentage of speciﬁc killing from two experiments is shown (b). Twelve days after infection splenocytes were stained for CD8ϩ cells bearing the SeV NP peptide-speciﬁc TCR using PE-labeled tetramers (c). Results are representative of two experiments.

RSV could be fundamental for the generation of the impaired 7. Heil, F., H. Hemmi, H. Hochrein, F. Ampenberger, C. Kirschning, S. Akira, adaptive immune response against this virus that allows multiple G. Lipford, H. Wagner, and S. Bauer. 2004. Species-specific recognition of single-stranded RNA via Toll-like receptor 7 and 8. Science 303:1526. re-infections during childhood. 8. Diebold, S. S., T. Kaisho, H. Hemmi, S. Akira, and E. S. C. Reis. 2004. Innate by guest on October 1, 2021. Copyright 2004 Pageant Media Ltd. In conclusion, we have demonstrated that SeV can induce effi- antiviral responses by means of TLR7-mediated recognition of single-stranded RNA. Science 303:1529. cient DC maturation and immunity in a TLR3- and MyD88-inde- 9. Lund, J. M., L. Alexopoulou, A. Sato, M. Karow, N. C. Adams, N. W. Gale, pendent manner. This pathway, which requires live virus, implies A. Iwasaki, and R. A. Flavell. 2004. Recognition of single-stranded RNA viruses a novel mechanism for virus detection and initiation of an antiviral by Toll-like receptor 7. Proc. Natl. Acad. Sci. USA 101:5598. 10. Kurt-Jones, E. A., L. Popova, L. Kwinn, L. M. Haynes, L. P. Jones, R. A. Tripp, immune response. Its further characterization should allow a better E. E. Walsh, M. W. Freeman, D. T. Golenbock, L. J. Anderson, and understanding of the mechanisms of viral antagonism that interfere R. W. Finberg. 2000. Pattern recognition receptors TLR4 and CD14 mediate with the generation of immunity, as well as the identification of response to respiratory syncytial virus. Nat. Immunol. 1:398. 11. Bieback, K., E. Lien, I. M. Klagge, E. Avota, J. Schneider-Schaulies, molecular targets amenable to manipulation for the control of an- W. P. Duprex, H. Wagner, C. J. Kirschning, V. Ter Meulen, and tiviral immune responses and the improvement of vaccination S. Schneider-Schaulies. 2002. Hemagglutinin protein of wild-type measles virus http://classic.jimmunol.org strategies. activates Toll-like receptor 2 signaling. J. Virol. 76:8729. 12. Rassa, J. C., J. L. Meyers, Y. Zhang, R. Kudaravalli, and S. R. Ross. 2002. Murine retroviruses activate B cells via interaction with Toll-like receptor 4. Acknowledgments Proc. Natl. Acad. Sci. USA 99:2281. 13. Compton, T., E. A. Kurt-Jones, K. W. Boehme, J. Belko, E. Latz, We acknowledge Dr. Adolfo Garcia-Sastre for critically reviewing this D. T. Golenbock, and R. W. Finberg. 2003. Human cytomegalovirus activates manuscript, Mount Sinai Hybridoma Shared Research Facility (New York, inflammatory cytokine responses via CD14 and Toll-like receptor 2. J. Virol. NY) for preparation of Abs to Sendai virus, and Luis Munõz and 77:4588. Downloaded from Sharon Czelusniak for invaluable technical support. 14. Siegal, F. P., N. Kadowaki, M. Shodell, P. A. Fitzgerald-Bocarsly, K. Shah, S. Ho, S. Antonenko, and Y. J. Liu. 1999. The nature of the principal type 1 interferon-producing cells in human blood. Science 284:1835. References 15. Cella, M., D. Jarrossay, F. Facchetti, O. Alebardi, H. Nakajima, A. Lanzavecchia, 1. Banchereau, J., F. Briere, C. Caux, J. Davoust, S. Lebecque, Y. J. Liu, and M. Colonna. 1999. Plasmacytoid monocytes migrate to inflamed lymph B. Pulendran, and K. Palucka. 2000. Immunobiology of dendritic cells. Annu. nodes and produce large amounts of type I interferon. Nat. Med. 5:919. Rev. Immunol. 18:767. 16. Dalod, M., T. Hamilton, R. Salomon, T. P. Salazar-Mather, S. C. Henry, 2. Kaisho, T., and S. Akira. 2003. Regulation of dendritic cell function through J. D. Hamilton, and C. A. Biron. 2003. Dendritic cell responses to early murine Toll-like receptors. Curr. Mol. Med. 3:373. cytomegalovirus infection: subset functional specialization and differential reg- 3. Qureshi, S. T., and R. Medzhitov. 2003. Toll-like receptors and their role in ulation by interferon ␣/␤. J. Exp. Med. 197:885. experimental models of microbial infection. Genes Immun. 4:87. 17. Brawand, P., D. R. Fitzpatrick, B. W. Greenfield, K. Brasel, C. R. Maliszewski, 4. Alexopoulou, L., A. C. Holt, R. Medzhitov, and R. A. Flavell. 2001. Recognition and T. De Smedt. 2002. Murine plasmacytoid pre-dendritic cells generated from of double-stranded RNA and activation of NF-␬B by Toll-like receptor 3. Nature Flt3 ligand-supplemented bone marrow cultures are immature APCs. J. Immunol. 413:732. 169:6711. 5. Krug, A., G. D. Luker, W. Barchet, D. A. Leib, S. Akira, and M. Colonna. 2004. 18. Salio, M., M. J. Palmowski, A. Atzberger, I. F. Hermans, and V. Cerundolo. Herpes simplex virus type 1 (HSV-1) activates murine natural interferon-produc- 2004. CpG-matured murine plasmacytoid dendritic cells are capable of in vivo ing cells (IPC) through Toll-like receptor 9. Blood 103:1433. priming of functional CD8 T cell responses to endogenous but not exogenous 6. Lund, J., A. Sato, S. Akira, R. Medzhitov, and A. Iwasaki. 2003. Toll-like re- antigens. J. Exp. Med. 199:567. ceptor 9-mediated recognition of herpes simplex virus-2 by plasmacytoid den- 19. Krug, A., R. Veeraswamy, A. Pekosz, O. Kanagawa, E. R. Unanue, M. Colonna, dritic cells. J. Exp. Med. 198:513. and M. Cella. 2003. Interferon-producing cells fail to induce proliferation of The Journal of Immunology 6889

naive T cells but can promote expansion and T helper 1 differentiation of antigen- 32. Doherty, P. C., D. J. Topham, R. A. Tripp, R. D. Cardin, J. W. Brooks, and experienced unpolarized T cells. J. Exp. Med. 197:899. P. G. Stevenson. 1997. Effector CD4ϩ and CD8ϩ T-cell mechanisms in the 20. Lopez, C. B., A. Fernandez-Sesma, J. L. Schulman, and T. M. Moran. 2001. control of respiratory virus infections. Immunol. Rev. 159:105. Myeloid dendritic cells stimulate both Th1 and Th2 immune responses depending 33. Von Stebut, E., J. M. Ehrchen, Y. Belkaid, S. L. Kostka, K. Molle, J. Knop, on the nature of the antigen. J. Interferon Cytokine Res. 21:763. C. Sunderkotter, and M. C. Udey. 2003. Interleukin 1␣ promotes Th1 differen- 21. Honda, K., S. Sakaguchi, C. Nakajima, A. Watanabe, H. Yanai, M. Matsumoto, tiation and inhibits disease progression in Leishmania major-susceptible BALB/c T. Ohteki, T. Kaisho, A. Takaoka, S. Akira, T. Seya, and T. Taniguchi. 2003. mice. J. Exp. Med. 198:191. ␣ ␤ Selective contribution of IFN- / signaling to the maturation of dendritic cells 34. Adachi, O., T. Kawai, K. Takeda, M. Matsumoto, H. Tsutsui, M. Sakagami, induced by double-stranded RNA or viral infection. Proc. Natl. Acad. Sci. USA K. Nakanishi, and S. Akira. 1998. Targeted disruption of the MyD88 gene results 100:10872. in loss of IL-1- and IL-18-mediated function. Immunity 9:143. 22. Lopez, C. B., A. Garcia-Sastre, B. R. Williams, and T. M. Moran. 2003. Type I 35. Nakanishi, K., T. Yoshimoto, H. Tsutsui, and H. Okamura. 2001. Interleukin-18 interferon induction pathway, but not released interferon, participates in the mat- regulates both Th1 and Th2 responses. Annu. Rev. Immunol. 19:423. uration of dendritic cells induced by negative-strand RNA viruses. J. Infect. Dis. 187:1126. 36. Albert, M. L., B. Sauter, and N. Bhardwaj. 1998. Dendritic cells acquire antigen 23. Lopez, C. B., A. Fernandez-Sesma, S. M. Czelusniak, J. L. Schulman, and from apoptotic cells and induce class I-restricted CTLs. Nature 392:86. T. M. Moran. 2000. A mouse model for immunization with ex-vivo virus infected 37. Edelmann, K. H., S. Richardson-Burns, L. Alexopoulou, K. L. Tyler, dendritic cells. Cell Immunol. 206:107. R. A. Flavell, and M. B. Oldstone. 2004. Does Toll-like receptor 3 play a bio- 24. Barber, D. L., E. J. Wherry, and R. Ahmed. 2003. Cutting edge: rapid in vivo logical role in virus infections? Virology 322:231. killing by memory CD8 T cells. J. Immunol. 171:27. 38. Picard, C., A. Puel, M. Bonnet, C. L. Ku, J. Bustamante, K. Yang, C. Soudais, 25. Hoebe, K., E. M. Janssen, S. O. Kim, L. Alexopoulou, R. A. Flavell, J. Han, and S. Dupuis, J. Feinberg, C. Fieschi, et al. 2003. Pyogenic bacterial infections in B. Beutler. 2003. Upregulation of costimulatory molecules induced by lipopoly- humans with IRAK-4 deficiency. Science 299:2076. saccharide and double-stranded RNA occurs by Trif-dependent and Trif-inde- 39. Sharma, S., B. R. tenOever, N. Grandvaux, G. P. Zhou, R. Lin, and J. Hiscott. pendent pathways. Nat. Immunol. 4:1223. 2003. Triggering the interferon antiviral response through an IKK-related path- 26. Hoebe, K., X. Du, P. Georgel, E. Janssen, K. Tabeta, S. O. Kim, J. Goode, P. Lin, way. Science 300:1148. N. Mann, S. Mudd, et al. 2003. Identification of LPS2 as a key transducer of 40. Fitzgerald, K. A., S. M. McWhirter, K. L. Faia, D. C. Rowe, E. Latz, MyD88-independent TIR signalling. Nature 424:743. D. T. Golenbock, A. J. Coyle, S. M. Liao, and T. Maniatis. 2003. IKK⑀ and TBK1 27. Yamamoto, M., S. Sato, H. Hemmi, K. Hoshino, T. Kaisho, H. Sanjo, are essential components of the IRF3 signaling pathway. Nat. Immunol. 4:491. O. Takeuchi, M. Sugiyama, M. Okabe, K. Takeda, and S. Akira. 2003. Role of 41. Inohara, N., and G. Nunez. 2003. NODs: intracellular proteins involved in in- adaptor TRIF in the MyD88-independent Toll-like receptor signaling pathway. flammation and apoptosis. Nat. Rev. Immunol. 3:371. Science 301:640. 42. Yoneyama, M., M. Kikuchi, T. Natsukawa, N. Shinobu, T. Imaizumi, 28. Kawai, T., O. Adachi, T. Ogawa, K. Takeda, and S. Akira. 1999. Unresponsive- M. Miyagishi, K. Taira, S. Akira, and T. Fujita. 2004. The RNA helicase RIG-I ness of MyD88-deficient mice to endotoxin. Immunity 11:115. has an essential function in double-stranded RNA-induced innate antiviral re- 29. Edwards, A. D., S. S. Diebold, E. M. Slack, H. Tomizawa, H. Hemmi, T. Kaisho, sponses. Nat. Immunol 5:730. S. Akira, and C. Reis e Sousa. 2003. Toll-like receptor expression in murine DC subsets: lack of TLR7 expression by CD8␣ϩ DC correlates with unresponsive- 43. Garcin, D., J. Curran, M. Itoh, and D. Kolakofsky. 2001. Longer and shorter ness to imidazoquinolines. Eur. J. Immunol. 33:827. forms of Sendai virus C proteins play different roles in modulating the cellular 30. Brimnes, M. K., L. Bonifaz, R. M. Steinman, and T. M. Moran. 2003. Influenza antiviral response. J. Virol. 75:6800. virus-induced dendritic cell maturation is associated with the induction of strong 44. Kim, H. W., J. G. Canchola, C. D. Brandt, G. Pyles, R. M. Chanock, K. Jensen, T cell immunity to a coadministered, normally nonimmunogenic protein. J. Exp. and R. H. Parrott. 1969. Respiratory syncytial virus disease in infants despite Med. 198:133. prior administration of antigenic inactivated vaccine. Am. J. Epidemiol. 89:422. 31. Moran, T. M., H. Isobe, A. Fernandez-Sesma, and J. L. Schulman. 1996. Inter- 45. Chang, J., and T. J. Braciale. 2002. Respiratory syncytial virus infection sup- leukin-4 causes delayed virus clearance in influenza virus- infected mice. J. Virol. presses lung CD8ϩ T-cell effector activity and peripheral CD8ϩ T-cell memory 70:5230. in the respiratory tract. Nat. Med. 8:54. by guest on October 1, 2021. Copyright 2004 Pageant Media Ltd. http://classic.jimmunol.org Downloaded from