Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Type I interferon-dependent CCL4 is induced by a cGAS/STING pathway that bypasses viral inhibition and protects infected tissue, independent of viral burden

Type I interferon-dependent CCL4 is induced by a cGAS/STING pathway that bypasses viral... OPENACCESS Type I interferons (T1-IFN) are critical in the innate immune response, acting upon infected Citation: Parekh NJ, Krouse TE, Reider IE, Hobbs RP, Ward BM, Norbury CC (2019) Type I and uninfected cells to initiate an antiviral state by expressing genes that inhibit multiple interferon-dependent CCL4 is induced by a cGAS/ stages of the lifecycle of many viruses. T1-IFN triggers the production of Interferon-Stimu- STING pathway that bypasses viral inhibition and lated Genes (ISGs), activating an antiviral program that reduces virus replication. The protects infected tissue, independent of viral importance of the T1-IFN response is highlighted by the evolution of viral evasion strategies burden. PLoS Pathog 15(10): e1007778. https:// doi.org/10.1371/journal.ppat.1007778 to inhibit the production or action of T1-IFN in virus-infected cells. T1-IFN is produced via activation of pathogen sensors within infected cells, a process that is targeted by virus- Editor: Victor Robert DeFilippis, Oregon Health and Sciences University, UNITED STATES encoded immunomodulatory molecules. This is probably best exemplified by the prototypic poxvirus, Vaccinia virus (VACV), which uses at least 6 different mechanisms to completely Received: April 16, 2019 block the production of T1-IFN within infected cells in vitro. Yet, mice lacking aspects of T1- Accepted: September 16, 2019 IFN signaling are often more susceptible to infection with many viruses, including VACV, Published: October 11, 2019 than wild-type mice. How can these opposing findings be rationalized? The cytosolic DNA Copyright:© 2019 Parekh et al. This is an open sensor cGAS has been implicated in immunity to VACV, but has yet to be linked to the pro- access article distributed under the terms of the duction of T1-IFN in response to VACV infection. Indeed, there are two VACV-encoded pro- Creative Commons Attribution License, which teins that effectively prevent cGAS-mediated activation of T1-IFN. We find that the majority permits unrestricted use, distribution, and reproduction in any medium, provided the original of VACV-infected cells in vivo do not produce T1-IFN, but that a small subset of VACV- author and source are credited. infected cells in vivo utilize cGAS to sense VACV and produce T1-IFN to protect infected Data Availability Statement: All relevant data are mice. The protective effect of T1-IFN is not mediated via ISG-mediated control of virus repli- within the manuscript and its Supporting cation. Rather, T1-IFN drives increased expression of CCL4, which recruits inflammatory Information files. monocytes that constrain the VACV lesion in a virus replication-independent manner by lim- Funding: This work was supported by CCN - R21 iting spread within the tissue. Our findings have broad implications in our understanding of AI097787, R21 AI115230, R21 AI121876, R21 pathogen detection and viral evasion in vivo, and highlight a novel immune strategy to pro- AI133536, R21 AI140014 all from the National tect infected tissue. Institute of Allergy and Infectious Diseases. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. PLOS Pathogens | https://doi.org/10.1371/journal.ppat.1007778 October 11, 2019 1 / 25 Type I interferon-dependent CCL4 expression protects virus-infected tissue Competing interests: The authors have declared that no competing interests exist. Author summary The recognition of virus infection leads to a quick and robust antiviral response mediated by type I interferons (T1-IFN). Nearly all viruses have acquired genes that block the induction or action of T1-IFN in order to attain a replicative advantage. Some viruses thwart the T1-IFN response so thoroughly, that cells infected in vitro do not produce any T1-IFN. And yet, animal models with defects in T1-IFN signaling are more sensitive to infection with these viruses than their wild-type counterparts. In this study, we find evi- dence to explain these otherwise contradicting findings. We show that a small population of infected cells in vivo are able to utilize a pathogen-sensing pathway that is completely blocked in vitro. T1-IFN produced by these specialized cells protects mice by recruiting inflammatory monocytes that restrict the spread of virus within infected tissue, indepen- dent of viral burden. Our findings have a direct impact on our understanding of how viruses are detected in infected tissue, and present a novel strategy of the immune system to limit pathology at peripheral sites of infection. Introduction Type I interferons (T1-IFN) are a critical component of the response to many viruses [1]. First described in the 1950’s, cells infected with a virus produce T1-IFNs which induce an antiviral state in surrounding cells [2, 3]. All T1-IFNs signal through a common receptor, the T1-IFN receptor (IFNαR), and consist of a single IFN-β, about a dozen similar IFN-α, and IFN-ε,κ, z, δ, τ, ω, and υ [4]. T1-IFN production is triggered by recognition of viral proteins or nucleic acids by host Pattern Recognition Receptors (PRRs) [5]. PRRs can be cell surface or endosomal Toll-like receptors (TLRs) or cytosolic nucleic acid sensors. IFNαR signaling initiates a positive feedback loop that produces more T1-IFN and expression of the full breadth of IFN-α subtypes [6]. The antiviral state is the effect of hundreds of Interferon-Stimulated Genes (ISGs) driven by IFNαR signaling [7, 8], which function to inhibit virus entry, cell growth, and virus replica- tion. This is the major mechanism that controls virus replication and host tissue pathology early after infection. The importance of the T1-IFN pathway in antiviral defense is underscored by two observa- tions. First, animals lacking IFNαR are highly susceptible to virus infection [1, 9]. Second, the majority of animal viruses encode at least one, but usually many, immunomodulatory proteins that interfere with, and often completely block, the production or action of T1-IFN [10, 11]. If viruses have evolved to block T1-IFN so efficiently, why is T1-IFN so important in resistance to virus infection in vivo? Here, we compare the in vitro and in vivo responses to a virus that infects peripherally to answer this question. Viruses have evolved to evade the T1-IFN response using multiple strategies. First, many viruses can block the pathways that sense infection and trigger T1-IFN production by prevent- ing PRRs from accessing their ligands, or by blocking PRR activation or signaling [10, 11]. Viruses can also produce soluble cytokine-binding proteins that bind to, and prevent the action of, secreted T1-IFN upon neighboring cells, and can also encode proteins that inhibit signaling through IFNαR within infected cells, preventing ISG expression [11, 12]. However, despite all of these approaches deployed by viruses to subvert the host interferon response, the T1-IFN system remains an important effector arm of the innate immune response following virus infection. PLOS Pathogens | https://doi.org/10.1371/journal.ppat.1007778 October 11, 2019 2 / 25 Type I interferon-dependent CCL4 expression protects virus-infected tissue We have utilized a viral infection model which exemplifies the ability of viruses to evade the T1-IFN response, the prototypic poxvirus vector, Vaccinia virus (VACV). We find a complete blockade in production of T1-IFN by VACV-infected cells in vitro [13, 14]. VACV has been reported to block the action of multiple cell surface, endosomal or cytosolic PRRs, including the predominant cytosolic DNA sensor cyclic GMP-AMP synthase (cGAS) and downstream signal transduction by its adapter protein Stimulator of Interferon Genes (STING) [15–18]. Indeed, we find evidence of efficient action of multiple mechanisms of viral evasion in vivo. Nevertheless, T1-IFN is still produced by a subset of infected cells in vivo in a cGAS- and STING-dependent manner. The T1-IFN produced protects mice against tissue pathology caused by VACV infection. However, the protective effect of T1-IFN is not mediated by the classical mechanism of control of virus replication, as neither ISG induction nor VACV repli- cation are altered in the absence of T1-IFN signaling. Rather, the protective effect is mediated through T1-IFN-enhanced production of the chemokine CCL4, which is required for localiza- tion of inflammatory monocytes to the VACV lesion. Monocytes enter the VACV lesion, where they become non-productively infected, allowing them to soak up excess virions and prevent spread to neighboring cells. Therefore, in spite of all the mechanisms deployed against T1-IFN by viruses, the host stays ahead of pathogen evolution by deploying alternative strate- gies to both produce T1-IFN, and to mediate protective strategies downstream of T1-IFN signaling. Results T1-IFN is not expressed within infected cells in vitro, but T1-IFN expression is required in vivo The primary cell types infected with VACV after immunization in the skin are keratinocytes and recruited monocyte/macrophages [19]. We infected representative cell lines with VACV and measured the production of IFN-β. Neither a keratinocyte line (Fig 1A) nor a macrophage cell line (Fig 1B) produced T1-IFN, measured by mRNA or protein, following exposure to live VACV. However, both the keratinocyte and macrophage lines produced significant quantities of IFN-β mRNA and protein following exposure to controls, namely Vesicular stomatitis virus or heat-inactivated VACV, respectively (Fig 1A and 1B). Therefore, the expression of multiple (� 6) immune evasion proteins blocked T1-IFN production from cells present at the site of infection in the skin in response to live VACV [15, 16, 20]. In marked contrast, when we infected WT mice or mice lacking the T1-IFN receptor -/- (IFNαR ) with VACV intradermally in the ear [21, 22] there was a significant difference in the tissue response to the virus. This was apparent as early as 3 days post-infection, when swell- -/- ing at the infection site was greater in IFNαR mice (Fig 1C), and progressed to produce a larger VACV lesion (Fig 1D), greater tissue loss (Fig 1E) and overall tissue damage (lesion size -/- + tissue loss) (Fig 1F) in IFNαR mice. Fig 1G shows that VACV infection causes only a small hole in the tissue of wild-type mice, but the majority of the ear is lost following infection -/- of IFNαR mice. Therefore, T1-IFN signaling is clearly required to moderate tissue damage following dermal VACV infection in vivo. To measure T1-IFN production in vivo, we har- vested tissue from VACV-infected ears and assayed for induction of IFN-β and IFN-α mRNA. Even when utilizing stringent RNA purification procedures, no significant induction of IFN mRNA could be detected prior to 3 days post-infection (Fig 1H). However, expression of IFN- β and IFN-α mRNA could be detected at 3 days post-infection and peaked at 5 days post-infec- tion (Fig 1H). We also detected IFN-β protein by ELISA from ear tissue at the peak of mRNA production, 5 days after dermal infection (Fig 1I). Therefore, despite the blockade of T1-IFN production in infected cells by immune modulators in vitro, VACV induces T1-IFN in vivo, PLOS Pathogens | https://doi.org/10.1371/journal.ppat.1007778 October 11, 2019 3 / 25 Type I interferon-dependent CCL4 expression protects virus-infected tissue Fig 1. T1-IFN is not expressed within infected cells in vitro, but T1-IFN expression is required in vivo. (A,B) Expression of IFN-β from a murine keratinocyte (A) or macrophage (B) cell line measured by RT-qPCR 6 hpi (left) or ELISA 24 hpi (right). (C-F) Tissue pathology of the ear pinnae after intradermal infection with VACV; ear thickness (C), lesion size (D), tissue loss (E), or total tissue damage (F), n = 8 mice per group. (G) Representative pictures of ear pinnae following intradermal VACV infection. (H,I) Expression of T1-IFN in ear tissue of wild-type mice at the indicated time post-infection measured by RT-qPCR (H, n = 3 naïve, 4 VACV- infected mice per group), or 5 dpi by ELISA (I). https://doi.org/10.1371/journal.ppat.1007778.g001 and IFNαR signaling is required to prevent excessive tissue pathology following VACV infection. A positive feedback loop is required for T1-IFN production in response to VACV IFN-β and the different IFN-α subtypes have differential effects on signals transduced through the IFNαR [23]. To understand the profile of T1-IFN genes induced by VACV infection, we assayed a diverse cross-section of the IFNα family, based on a previous publication describing the antiviral activities of IFNα proteins [24], in the ear 5 days post-infection with VACV. IFN- β was strongly induced in all experiments (Fig 2A) but, although IFN-α2 is the predominant gene present in many arrays and multiplex assays, we found that it was not induced by VACV (Fig 2A). Rather, IFN-α4 and, to a lesser extent, IFN-α5 were the only IFN-α species induced by dermal VACV infection (Fig 2A). In most cells, robust T1-IFN production requires positive feedback signaling through IFNαR [25, 26]. We examined induction of T1-IFN in mice lacking PLOS Pathogens | https://doi.org/10.1371/journal.ppat.1007778 October 11, 2019 4 / 25 Type I interferon-dependent CCL4 expression protects virus-infected tissue Fig 2. A positive feedback loop is required for T1-IFN production in response to VACV. (A) Expression of T1-IFN subtypes in ear tissue of wild-type mice 5 dpi measured by RT-qPCR. (B-D) Expression of early T1-IFNs in ear tissue of mice 5 dpi measured by RT-qPCR. https://doi.org/10.1371/journal.ppat.1007778.g002 IFNαR signaling and found poor induction of both IFN-β and IFN-α4 mRNA in these mice, compared to wild-type controls (Fig 2B). IFNαR stimulation results in phosphorylation of the transcription factors Signal Transducer and Activator of Transcription (STAT) STAT1 and STAT2, which form a trimeric complex with IFN Regulatory Factor (IRF) IRF9, and are required to transduce T1-IFN signals [27, 28]. Both STAT1 (Fig 2C) and STAT2 (Fig 2D) were required for optimal induction of T1-IFN. Therefore, although virus-encoded proteins block T1-IFN from binding its receptor and the downstream activation and function of STAT proteins, a positive feedback loop is required for an efficient T1-IFN response following der- mal VACV infection. cGAS sensing of VACV-infected cells induces T1-IFN to limit local tissue pathology Induction of T1-IFN requires recognition of viral components by PRRs and signal transduction through adapter proteins. Cell surface and endosomal TLRs utilize either Myeloid Differentia- tion Primary Response 88 (MyD88) or TIR-domain-containing adapter-inducing interferon-β (TRIF) proteins to induce T1-IFN. Cytosolic nucleic acid sensors utilize the Mitochondrial Anti- viral Signaling protein (MAVS) or STING for signal transduction. Each of these pathways is reg- ularly targeted by a host of viruses, including poxviruses, which alter recognition by, or signaling downstream from, the MyD88 [29], TRIF [30], MAVS [20, 31] or STING [15, 16] pathways. We have previously found that MyD88 is required for optimal immune responses to dermal VACV infection [32], and others have found that MyD88 signaling is crucial in NK [33], T cell [34, 35] and Plasmacytoid DC [36] responses. However, we found that MyD88 was not required for the induction of IFN-β in response to VACV (Fig 3A). Although TRIF has been implicated in limiting VACV replication within infected macrophages [37], we found no defect in the induction of IFN-β in response to VACV in mice with a non-functional TRIF gene (Fig 3B). Deletion of the virus-encoded immunomodulatory protein E3L can rescue IFN production in vitro via recognition of dsRNA in a manner dependent on the PRR-adapter protein MAVS [20]. However, in vivo, MAVS-deficient mice were also able to effectively induce IFN-β in response to VACV (Fig 3C). VACV is a large dsDNA virus that carries out its lifecycle entirely in the cytoplasm of infected cells, and thus is likely sensed by cytosolic DNA sensors upstream of the adapter protein STING. Even though VACV efficiently blocks sensing via the STING pathway in vitro [15–17, 38], induction of IFN-β was heavily reliant on STING signaling (Fig 3D), indicating that recognition of VACV DNA was likely responsible for T1-IFN production. The almost total requirement for STING signaling was curious, as cells infected with VACV in vitro do not produce T1-IFN (Fig 1). Therefore, we considered the possibility that T1-IFN may only be produced by small numbers of uninfected cells, which are not subject to PLOS Pathogens | https://doi.org/10.1371/journal.ppat.1007778 October 11, 2019 5 / 25 Type I interferon-dependent CCL4 expression protects virus-infected tissue Fig 3. cGAS sensing of VACV-infected cells induces T1-IFN to limit local tissue pathology. (A-D) Expression of IFN-β in ear tissue of mice 5 dpi measured by RT-qPCR. (E) Cryosections of ear tissue from wild-type mice 5 dpi with VACV-OVA stained with Wheat-Germ Agglutinin (Grey) to visualize cell membranes and DAPI (Blue) for nuclei. RNA fluorescence in situ hybridization was used to detect viral RNA (OVA-Green) and cellular IFN-β (Red). (F,G) Expression of IFN-β in ear tissue of mice 5 dpi measured by RT-qPCR (F) or ELISA (G). (H-K) Tissue pathology of the ear pinnae after intradermal infection with VACV; ear thickness (H), lesion size (I), tissue loss (J), or total tissue damage (K), n = 9–10 mice per group. (L) Representative pictures of ear pinnae following intradermal VACV infection. https://doi.org/10.1371/journal.ppat.1007778.g003 the same strict immunomodulation as VACV-infected cells. We infected mice with a recombi- nant VACV expressing chicken egg ovalbumin (OVA), and on day 5 post-infection conducted in situ hybridization with probes targeting either OVA (to identify VACV-infected cells, in green) or IFN-β mRNA (in red). To allow accurate definition of cell boundaries, we stained with fluorescently-labeled wheat-germ agglutinin, which efficiently labeled all cell-surface membranes in the infected ear. In the majority of areas in which we could observe OVA mRNA there was little or no IFN-β mRNA visible (Fig 3E, left panel), raising the possibility that it was indeed uninfected cells that were the source of IFN-β. However, in multiple experi- ments we never observed IFN-β mRNA in cells that did not also display OVA mRNA (Fig 3E, right panel). Rather, it was a small, spatially distinct subpopulation of VACV-infected, OVA mRNA-expressing cells that produced IFN-β mRNA within the cytosol (Fig 3E). Therefore, the inhibitory effect of VACV upon STING signaling that appears complete in vitro, is bypassed in vivo, allowing T1-IFN production. Upstream of STING, cGAS has been implicated in the recognition of multiple virus infec- tions, including VACV [18, 39, 40]. However, in addition to its well established role in PLOS Pathogens | https://doi.org/10.1371/journal.ppat.1007778 October 11, 2019 6 / 25 Type I interferon-dependent CCL4 expression protects virus-infected tissue induction of T1-IFN production, cGAS can also activate NF-κB [41–43] and STAT6 [44, 45], both of which can impact the anti-poxvirus immune response [46–51]. We found that induc- tion of IFN-β mRNA and protein was impaired in cGAS-deficient mice to a similar extent to that observed in mice that lacked STING signaling (Fig 3F and 3G). To investigate a role for cGAS-induced T1-IFN during dermal VACV infection, we measured local tissue pathology in -/- -/- cGAS mice compared to wild-type mice after intradermal infection of the ear. cGAS mice displayed increased tissue thickness beginning by 3 days post-infection (Fig 3H), the time point at which we first measured induction of T1-IFN transcripts (Fig 1H). Similar to the phe- -/- -/- notype observed in IFNαR mice, cGAS mice also exhibited larger lesions, greater tissue loss and total tissue damage compared to wild-type mice (Fig 3I–3L). Therefore, cGAS-medi- ated production of T1-IFN is predominantly responsible for moderating tissue pathology fol- lowing dermal VACV infection. Thus, although cGAS-mediated induction of T1-IFN is completely blocked in vitro, the host uses this pathway to induce tissue protective T1-IFN in vivo, bypassing virus-mediated inhibitory pathways. T1-IFN does not control VACV replication in the skin Many ISGs have direct antiviral properties that can blunt propagation of virus within infected tissue [7]. Therefore, it was important to examine whether T1-IFN-mediated upregulation of antiviral ISGs in the ear was responsible for control of VACV replication, and thus whether a lack of ISG induction was responsible for the tissue pathology we observed after infection of -/- IFNαR mice. We examined the transcript levels of 81 prominent antiviral ISGs 5 days after dermal VACV infection by RT-qPCR array and found strong (>4-fold to ~300-fold) induction of a number of ISGs (Fig 4A). Surprisingly, we found that a small number of IFN-responsive transcripts were downregulated upon dermal VACV infection, including irf4, irf5 and ifi27 (Fig 4A). We also found robust induction of a number of antiviral ISGs including Ifit1, Ifi204, ISG15, OAS1, Ifih1, irf1, irf7 and irf8 (Fig 4A). However, we did not find that induction of any of the antiviral ISGs was dependent upon T1-IFN signaling (Fig 4B). This suggests that redun- dant pathway(s) exists for induction of these antiviral genes. In addition, it clearly demon- strates that upregulation or downregulation of these antiviral ISGs by T1-IFN signaling is not -/- -/- responsible for the tissue pathology phenotype we observed in IFNαR and cGAS mice. To directly address the role for T1-IFN in control of VACV replication in vivo, we com- -/- -/- pared VACV titer in the ears of cGAS and IFNαR mice to wild-type mice at 5 days post- -/- -/- infection, the peak of VACV replication [52]. Both IFNαR and cGAS mice displayed only a modest 2-fold increase in infectious VACV titer (Fig 4C) that, when compared to the 100,000 fold increase in titer in the 5 days since infection, was unlikely to explain the drastic change in pathology. VACV gene expression and infectious titer decrease after 5 days post- infection [19, 52], so it is possible that T1-IFN controls viral clearance rather than limiting -/- peak VACV titers. However, the tissue loss exhibited by IFNαR mice makes it difficult to accurately evaluate VACV titer at late times post-intradermal infection. In order to assess the -/- role of T1-IFN in clearance of VACV, we infected the ears of IFNαR and wild-type mice with VACV by scarification, an infectious methodology that results in a milder pathology. We found that the kinetics of VACV clearance was unaffected in the absence of T1-IFN signaling (Fig 4D), so the protective effect of T1-IFN during dermal VACV infection is independent of its direct antiviral characteristic. T1-IFN induces ccl4 to recruit protective monocytes to the site of infection In addition to the induction of antiviral ISGs, it is now becoming clear that T1-IFN can modu- late the innate and adaptive immune response, independent of any effects upon virus PLOS Pathogens | https://doi.org/10.1371/journal.ppat.1007778 October 11, 2019 7 / 25 Type I interferon-dependent CCL4 expression protects virus-infected tissue Fig 4. T1-IFN does not control VACV replication in the skin. (A) Expression of antiviral ISGs in ear tissue of wild- type mice 5 dpi compared to naïve controls measured using RT-qPCR array. (B) Expression of induced ISGs from (A) -/- in ear tissue of wild-type and IFNαR mice 5 dpi measured using RT-qPCR array. (C) VACV titer in ears of mice 5 days post-intradermal infection with VACV. Dotted line indicates initial inoculum. (D) VACV titer in ears of mice at the indicated day post-infection with VACV by scarification, n = 5 mice per group. https://doi.org/10.1371/journal.ppat.1007778.g004 PLOS Pathogens | https://doi.org/10.1371/journal.ppat.1007778 October 11, 2019 8 / 25 Type I interferon-dependent CCL4 expression protects virus-infected tissue replication [53, 54]. Therefore, we used a second qPCR array analysis to compare the change in expression of 84 cytokines and chemokines in the ear 5 days after dermal VACV infection. We found strong infection-mediated induction of 28 of 84 transcripts, with some that are known to be interferon-responsive, including the CXCR3 ligands cxcl9 and cxcl10, as well as IFN-γ (Fig 5A). To investigate which of these transcripts were dependent upon T1-IFN signal- -/- ing, we harvested mRNA from ear tissue of VACV-infected IFNαR mice and performed a -/- similar qPCR array analysis, comparing to uninfected IFNαR mice (Fig 5B). Somewhat sur- -/- prisingly, we found induction of many of the same genes in IFNαR mice upon VACV infec- tion that we observed after infection of wild-type mice. Although VACV strongly induced -/- many transcripts in both wild-type and IFNαR mice, we found that the expression of only one upregulated transcript was differentially modulated >2-fold in a statistically significant manner in the absence of T1-IFN signaling. The IFNαR-dependent gene was ccl4 (macrophage inflammatory protein-1β) (Fig 5C). We were able to confirm that expression of ccl4 in the ears of VACV-infected mice was dependent on both IFNαR signaling and cGAS expression using a targeted qPCR assay (Fig 5D). The cellular response to dermal VACV infection is detected at 3 days post-infection when inflammatory monocytes move into the ear, and numbers of these cells peak at 5 days post- infection [52], i.e. with similar kinetics to T1-IFN induction (Fig 1H). Recruitment of classical inflammatory monocytes occurs with similar kinetics to the induction of T1-IFN transcripts, and is followed by recruitment of a non-classical tissue-protective Ly6G cell population that primarily occurs after the peak of T1-IFN production [52]. Both classical inflammatory mono- cytes and the non-classical tissue-protective Ly6G cell population home to foci of VACV infection, although only classical inflammatory monocytes are infected within the lesion [19]. Local depletion of either inflammatory monocytes [55] or the tissue-protective Ly6G cells -/- -/- [52] produces tissue pathology similar to that observed in cGAS or IFNαR mice. However, depletion of T cells does not affect local pathology or systemic spread of virus following dermal VACV infection [52, 55]. Because the recruitment of classical inflammatory monocytes occurs concurrently with the induction of T1-IFN gene transcripts (Fig 1H), and because the onset of cell recruitment into the ear coincides with the onset of increased tissue pathology in mice lacking T1-IFN signaling (Fig 1C–1F), we sought to determine if T1-IFN-induced ccl4 plays a role in the recruitment of classical inflammatory monocytes, as well as the later recruitment of the Ly6G cell population. We assayed for the presence of chemokine receptors on the cell surface of inflammatory monocytes and tissue-protective Ly6G cells that traffic into VACV-infected ears. From our array data, we found a statistically significant reduction in the expression of ccl4, a large, but non-statistically significant change in expression of cxcl13 as well as a slight, significant change in the expression of cxcl10 (Fig 5C). Using flow cytometry, we assayed for cell-surface expres- sion of the receptors for those respective chemokines, as well as for the prototypic monocyte chemokine receptor, CCR2, the receptor for CCL2 that was upregulated upon VACV infection -/- + of both wild-type and IFNαR mice (Fig 5A and 5B). We found that the majority of CD11b hi - + LyC Ly6G CD64 inflammatory monocytes present in the ear of VACV-infected mice at 5 + + days post-infection were CCR5 , the receptor for CCL4, rather than CCR2 , or positive for the receptors for CXCL13 (CXCR5) or CXCL9 and CXCL10 (CXCR3) (Fig 5E). We next examined whether the deficiency in ccl4 expression by the absence of T1-IFN pre- vented recruitment of CCR5 inflammatory monocytes to the VACV-infected ear. We assayed -/- -/- myeloid cell infiltration in cGAS and IFNαR mice and found a defect in infiltration of inflammatory monocytes, but not Ly6G cells, into the infected ear compared to wild-type mice (Fig 5F). We observed a similar blockade in infiltration of inflammatory monocytes when we treated mice with the CCR5 antagonist, maraviroc (Fig 5G). A number of PLOS Pathogens | https://doi.org/10.1371/journal.ppat.1007778 October 11, 2019 9 / 25 Type I interferon-dependent CCL4 expression protects virus-infected tissue PLOS Pathogens | https://doi.org/10.1371/journal.ppat.1007778 October 11, 2019 10 / 25 Type I interferon-dependent CCL4 expression protects virus-infected tissue Fig 5. T1-IFN induces ccl4 to recruit protective monocytes to the site of infection. (A, B, C) Expression of cytokines and chemokines in ear tissue of mice 5 dpi measured using RT-qPCR array. Volcano plots demonstrate statistically significant points if above the p = 0.05 line. Gene expression changes in wild-type mice in response to VACV, with genes that are upregulated >2-fold by VACV infection in red in the upper right hand quadrant and genes that are downregulated > 2-fold by VACV infection in blue in the upper left hand quadrant (A). Gene -/- expression changes in IFNαR mice in response to VACV, with genes that are upregulated >2-fold by VACV infection in red in the upper right hand quadrant and genes that are downregulated > 2-fold by VACV infection in blue in the upper left hand quadrant (B). Gene -/- expression changes in VACV-infected wild-type mice compared to VACV-infected IFNαR mice with genes that are upregulated >2-fold in -/- VACV-infected IFNαR mice relative to wild-type VACV infected in red in the upper right hand quadrant and genes that are -/- downregulated > 2-fold in VACV-infected IFNαR mice relative to wild-type VACV infected in blue in the upper left hand quadrant (C). -/- Ccl4 (in green) is the only gene that is upregulated upon VACV infection but is downregulated in IFNαR mice relative to wild-type mice. n = 6 naïve, 9 VACV-infected mice per group. (D) Expression of ccl4 in ear tissue of VACV-infected mice 5 dpi measured by RT-qPCR. (E) Chemokine receptor expression on inflammatory monocytes and Ly6G cells recruited into VACV-infected ears 5 dpi. Representative flow cytometry plots (left) and proportion of cells expressing each receptor (right). (F) Myeloid cell populations in VACV-infected ears 5 dpi measured by flow cytometry. Representative flow cytometry plots (left) and absolute number of cells (right). (G) Recruitment of inflammatory monocytes to the ear of uninfected, VACV-infected mice 5 dpi, or VACV-infected mice treated systemically with the CCR5 antagonist Maraviroc, 5dpi. (H) Immunofluorescence microscopy of ears 5 dpi with VACV-NP-S-eGFP reveal monocyte localization (Ly6C-Red) in relation to VACV-infected cells (GFP-Green). (I) Percentage of GFP cells in the ear 5dpi with VACV Early-GFP or VACV Late-GFP measured by flow cytometry. (J) Percentage of GFP cells in the ear 5dpi with VACV-NP-S-eGFP measured by flow cytometry. https://doi.org/10.1371/journal.ppat.1007778.g005 inflammatory monocytes still moved into the infected ear, however, Ly6C inflammatory monocytes failed to localize to the VACV lesion in the absence of T1-IFN induction or signal- ing (Fig 5H). Therefore, cGAS-mediated induction of T1-IFN during dermal VACV infection is required for ccl4 expression and CCR5-dependent recruitment of inflammatory monocytes to viral lesions. Recruited inflammatory monocytes moderate pathology following VACV infection in a manner independent of control of virus replication and expression of antiviral ISGs. However, the mechanisms used by monocytes/macrophages to mediate protective antiviral responses remain poorly described at present. Following infection with some viruses, macrophage popu- lations are preferentially infected in order to facilitate initiation of an adaptive immune response [56–58] or otherwise restrict virus spread [55, 59, 60]. However, primary macrophage populations do not typically support VACV replication, restricting gene expression at a point prior to DNA replication [61]. Therefore, we sought to explore whether infected inflammatory monocytes restricted VACV replication in vivo using a similar mechanism. We infected wild- type mice with VACV that expresses GFP driven by either the early/late promoter p7.5 (Early- GFP) or the late promoter p11 (Late-GFP), which is active only after DNA replication, and assayed GFP expression in inflammatory monocytes in the ear 5 days post-infection. In CD45 cell populations there were equivalent numbers of cells expressing early and late promoter- driven GFP, indicating that there was no blockade in VACV replication within these cells (Fig 5I, left panel). However, a greater proportion of inflammatory monocytes expressed early pro- moter-driven GFP than GFP driven by the late p11 promoter (Fig 5I, right panel), indicative of a blockade of VACV replication in the inflammatory monocyte population. Therefore, inflammatory monocytes within the VACV lesion can potentially act as a “virus sponge,” becoming infected to spare nearby cells from infection. To examine whether this occurred, we -/- infected wild-type or IFNαR mice with VACV that expresses GFP that accumulates in the -/- nucleus of infected cells (VACV-NP-S-eGFP). IFNαR mice had a greater proportion of VACV-infected cells in the ear 5 days post-infection compared to wild-type mice (Fig 5J). Thus, inflammatory monocytes enter the VACV lesion, become non-productively infected, and prevent the infection of neighboring cells. These data support a role for T1-IFN-mediated restriction of virus spread and subsequent tissue damage via a novel immunomodulatory mechanism that does not require a direct effect upon virus replication via the expression of antiviral ISGs. PLOS Pathogens | https://doi.org/10.1371/journal.ppat.1007778 October 11, 2019 11 / 25 Type I interferon-dependent CCL4 expression protects virus-infected tissue ccl4 expression restores inflammatory monocyte recruitment and ameliorates tissue damage To examine the importance of T1-IFN-mediated ccl4 expression in the control of VACV- mediated tissue damage in the dermis, we constructed a recombinant VACV that expresses a construct driven by the relatively weak C6 promoter containing murine ccl4 and the red fluo- rescent protein mKate2 (VACV-CCL4) (Fig 6A). To ensure that any in vivo phenotype we observed was not due to differences in VACV replication we examined single step and multi- step growth curves for VACV-CCL4 vs wild-type VACV (Fig 6B and 6C). We found no differ- ence in growth of the recombinant VACV (Fig 6B and 6C). Upon infection of cells in vitro, the rVACV produced mKate (Fig 6D), as well as ccl4 mRNA (Fig 6E) and protein (Fig 6F). To examine whether VACV-driven expression of ccl4 could rescue recruitment of inflammatory -/- monocytes, we infected IFNαR mice with VACV-CCL4 or a control VACV that did not + hi - express ccl4. Five days after infection we assayed the recruitment of CD11b LyC Ly6G + + CD64 inflammatory monocytes or tissue protective Ly6G myeloid cells in the infected ear. VACV-expression of CCL4 increased the number and proportion of inflammatory monocytes in the ears of VACV-infected mice in the absence of T1-IFN signaling (Fig 6G) without alter- ing recruitment of tissue-protective Ly6G myeloid cells (Fig 6H). Further, and more impor- tantly, restoration of ccl4 expression was able to rescue localization of Ly6C inflammatory -/- monocytes to foci of virus infection in IFNαR mice (Fig 6I). To examine whether CCL4-me- diated inflammatory monocyte recruitment ameliorated the tissue damage we observed in -/- mice lacking IFNαR signaling, we infected IFNαR mice with VACV-CCL4 or a control VACV that did not express ccl4 and monitored multiple aspects of tissue damage as outlined above. We found a statistically significant reduction in ear swelling, lesion size and the com- bined criteria of total tissue damage (Fig 6J–6M). These data strongly support a requisite role for CCL4-mediated recruitment of inflammatory monocytes in moderating tissue damage, but not virus replication, following dermal VACV infection. Discussion The importance of T1-IFN in control of virus replication has led to the evolution of a myriad of evasion strategies employed by viruses across multiple families, aimed at allowing virus rep- lication and propagation to a new host. However, using a virus system in which multiple stages of T1-IFN induction and downstream signaling are inhibited, our data clearly indicate that, despite a lack of induction of T1-IFN within infected cells in vitro, T1-IFN is induced and acts to moderate tissue pathology following virus infection in vivo. Our data support a role for mul- tiple host mechanisms that allow control of virus-dependent pathology but are independent of control over virus replication. Importantly, these mechanisms occur in the face of multiple, effective evasion strategies deployed by the viral pathogen. Induction of T1-IFN following VACV infection in vivo occurred via the cytosolic DNA sen- sor, cGAS and its adapter protein STING. VACV completely blocks cGAS and STING-depen- dent induction of T1-IFN using multiple mechanisms in vitro, including expression of B2, F17, C4, C16 and C6 [15, 16, 17, 18]. However, production of both T1-IFN mRNA and pro- tein, as well as protection against enhanced tissue pathology following VACV infection, was dependent upon cGAS, but not upon other PRR signaling pathways. This contrasts with other studies, which have demonstrated that cGAS is required to protect against VACV challenge, but which have stopped short of demonstrating a role for cGAS in T1-IFN production follow- ing VACV infection [18, 39, 40]. The lack of a link between cGAS and T1-IFN production fol- lowing VACV infection in vivo likely stems from technical issues, and we developed a number of highly sensitive ex vivo approaches in order to be able to detect T1-IFN transcripts and PLOS Pathogens | https://doi.org/10.1371/journal.ppat.1007778 October 11, 2019 12 / 25 Type I interferon-dependent CCL4 expression protects virus-infected tissue Fig 6. ccl4 expression restores inflammatory monocyte recruitment and ameliorates tissue damage. (A) Schematic of rVACV expressing ccl4 and mKate2. (B) Single step growth curve of wild-type VACV and VACV-CCL4 in 143B cells. (C) Multi-step growth curve of wild-type VACV and VACV-CCL4 in 143B cells. (D) Detection of mKate2 expression in 143B cells 6 hpi assayed using flow cytometry. (E) Expression of murine ccl4 in 143B cells 6 hpi measured by RT-qPCR. (F) Secretion of murine -/- CCL4 from 143B cells 24 hpi measured using ELISA. (G) Monocyte recruitment into VACV-infected ears of IFNαR mice 5 dpi assayed by flow cytometry, n = 4 mice + -/- per group. (H) Ly6G cell recruitment into VACV-infected ears of IFNαR mice 5 dpi measured by flow cytometry, n = 4 mice per group. (I) Immunofluorescence -/- microscopy of IFNαR ears 5 dpi with VACV-CCL4 or VACV-mCherry for monocyte localization (Ly6C-Blue) in relation to VACV-infected cells (mKate/mCherry- -/- Red). (J-L) Tissue pathology of the ear pinnae of IFNαR mice after intradermal infection with VACV; ear thickness (J), lesion size (K), or total tissue damage (L), -/- n = 6–7 mice per group. (M) Representative pictures of ear pinnae of IFNαR mice following intradermal VACV infection. https://doi.org/10.1371/journal.ppat.1007778.g006 protein from VACV-infected tissue. Using these approaches, we clearly demonstrate that cGAS signaling is required for T1-IFN induction following VACV infection. We found that the majority of infected cells in vivo mirror our in vitro finding, and do not express T1-IFN mRNA, likely due to the action of multiple VACV immunomodulators. We anticipated that the induction of T1-IFN that we observed in vivo might be the result of the transfer of cGAS- produced cGAMP to neighboring uninfected cells [62, 63]. Once within neighboring cells, PLOS Pathogens | https://doi.org/10.1371/journal.ppat.1007778 October 11, 2019 13 / 25 Type I interferon-dependent CCL4 expression protects virus-infected tissue cGAMP could induce T1-IFN production via STING within cells that are not subject to immu- nomodulatory VACV proteins, allowing for efficient expression. However, all of the IFN-β mRNA we detected was present within a small population of VACV-infected cells that appeared functionally distinct from the majority of infected cells. Therefore, a specialized pop- ulation of cells may allow the host to subvert viral evasion mechanisms and express protective T1-IFN. Therapeutic targeting of such a population during infection or vaccination may pref- erentially allow for the generation of a strong T1-IFN response within a tissue. What would be the characteristic of an infected cell that can subvert the action of the viral B2, F17 and C6 proteins to allow for cGAS-dependent T1-IFN expression? T1-IFN expression begins by day 3 post-infection, a time point when inflammatory monocytes are the first recruited cells to accumulate at the site of VACV infection. However, induction of T1-IFN still occurred in the absence of inflammatory monocyte recruitment (S1 Fig). This was not unex- pected, as T1-IFN was required for effective inflammatory monocyte recruitment. We have previously described T1-IFN production by the tissue protective Ly6G population of recruited cells [52]. However, the Ly6G population does not begin to accumulate until day 5 post-infection and does not peak until day 8–10 post-infection [52] i.e. significantly after the peak of T1-IFN production. Therefore, Ly6G cells may contribute to T1-IFN production after the peak of VACV replication, but they do not play a role in CCL4-mediated recruitment of classical inflammatory monocytes, which precede them into the VACV lesion. Our data are indicative of a role for tissue-resident VACV-infected cells in cGAS-mediated T1-IFN produc- tion. These cells could be specialized populations of keratinocytes, dermal fibroblasts, Langer- hans cells, dermal DCs, tissue resident macrophages,γδ or other tissue-resident T cells, ILCs or even mast cells. Identification of the cells responsible for T1-IFN production following VACV infection will require extensive further studies. The mechanism by which infected cells can produce T1-IFN in a cGAS-dependent manner in vivo in the face of viral evasion strategies that efficiently block this pathway in vitro is partic- ularly interesting. The most effective way to bypass the blockade of cytosolic sensing within infected cells would be to inhibit expression of the viral proteins that target the cytosolic sen- sors. However, on a global level, the breadth of mechanisms employed by viruses to block cyto- solic sensing likely precludes this approach. Indeed, in the system examined here, both the cGAS/STING and MDA5/MAVS pathways are targeted by multiple VACV proteins including B2, F17, K13, C6, E3 and K3 [15–17, 64–66]. Therefore, it is more likely that multiple cGAS/ STING sensing and signaling pathways exist in specialized cells. In the case of the RNA-sens- ing pathway, MAVS signaling can occur from multiple sites within an infected cell, namely from the mitochondria or peroxisomal membranes [67]. These differential signaling sites pro- duce different functional outcomes [67], raising the possibility that they may be differentially sensitive to viral inhibition. This would be the case if viral proteins are targeted to the mito- chondria or ER, respectively, to inhibit the STING and MAVS signaling pathways. The possi- bility that these pathways may be differentially expressed, or perhaps enhanced, within specific cell populations is intriguing. Targeting of these specialized pathways may preferentially allow for the generation of a strong T1-IFN response during vaccination or therapeutic treatments. The majority of T1-IFNs produced during VACV infection were IFN-β and IFN-α4, each of which are produced within the first round of T1-IFN induction, without the need to first induce expression of further ISGs such as irf7. Localization of IFN-β exclusively within VACV-infected cells implies that nearby cells are not triggered to produce either IFN-β or IFN-αs that are downstream from IFNαR in the positive feedback loop of T1-IFN induction. This observation supports two conclusions. First, the soluble T1-IFN binding protein encoded by VACV, B18 [68], appears to be very effective at preventing the spread of T1-IFN induction to nearby cells. Second, the majority of both IFN-β and IFN-α4 induction within infected cells PLOS Pathogens | https://doi.org/10.1371/journal.ppat.1007778 October 11, 2019 14 / 25 Type I interferon-dependent CCL4 expression protects virus-infected tissue is dependent upon IFNαR, STAT1 and STAT2, despite the inhibitory effect of the viral VH1 protein upon STAT1 activation [69]. These findings indicate that some level of positive feed- back signaling through IFNαR in VACV-infected cells is required. B18 binds to the surface of uninfected cells and soaks up excess T1-IFN, preventing activation of these cells through IFNαR [68]. Therefore, B18 may not effectively bind to the surface of infected cells, perhaps due to virus-induced changes in the cell surface architecture and proteome. It was surprising that VACV replication was unaffected by T1-IFN signaling. However, upregulation of antiviral ISGs was independent of T1-IFN in the infected ear, despite the observation that T1-IFN expression occurred only within infected cells. Many of the ISGs induced by T1-IFN signaling are important mediators in restricting the ability of viruses to infect and replicate within cells, although many have not been studied in the context of a VACV infection. However, during infection with VACV, ISG15, which is strongly induced in a T1-IFN-independent manner in infected skin, controls mitochondrial function and limits virus replication within infected cells [70, 71]. The only logical explanation from these oppos- ing observations is that ISG expression is induced by redundant signaling pathways via the expression of other cytokines. It is possible that other families of interferon species may induce the expression of ISGs in VACV-infected skin. In addition, there is significant crosstalk between the NF-κB and T1-IFN signaling pathways [46], so it is possible that other pathways are activated to stimulate ISG induction. NF-κB signaling is also inhibited in VACV-infected cells [29, 72, 73], but our profiling of the cytokine and chemokine response in the infected ear demonstrates strong upregulation of a number of inflammatory cytokines. We were unable to examine the induction of cytokines, chemokines or ISGs within infected vs uninfected cells in situ, but it is likely that upregulation of these effector molecules occurs within uninfected cell populations. Irrespective of the mechanisms by which ISGs are induced, it is nonetheless clear that antiviral ISGs do not play a role in the protection of infected tissue by T1-IFN following VACV infection. Here, the action of T1-IFN is mediated through the CCL4-dependent recruitment of inflammatory monocytes, rather than through the direct inhibition of virus replication via the action of intracellular ISGs, a process that can be subverted by viral proteins within infected cells. In a number of viral infections, including the VACV system used here, tissue-resident macrophage populations in secondary lymphoid organs are required to prevent systemic dis- semination of viruses [55, 58–60, 74]. After infection with some viruses, infection of macro- phage populations is required in order to facilitate development of an effective adaptive immune response [57]. However, the role of recruited inflammatory monocyte populations after infection differs from that of macrophages resident in systemic organs [55]. Although inflammatory monocyte recruitment has been indirectly implicated in the response to virus infection [75–77], the mechanisms employed by recruited inflammatory monocyte popula- tions are unknown. Monocytes do not directly inhibit VACV replication within other infected cell populations, as there is no large change in VACV titer, despite the marked change in tissue pathology in the absence of T1-IFN signaling. Infectious VACV can be recovered from inflam- matory monocytes themselves when they are sorted from VACV-infected ears [19]. However, the titer recovered from inflammatory monocytes was less than 2 pfu per cell, and could repre- sent input levels of virus rather than progeny virions. Here, we show that monocyte recruit- ment to the VACV-infected ear was dependent upon T1-IFN-mediated induction of CCL4, although the requirement was incomplete. However, in the absence of CCL4 induction, recruitment to the VACV lesion was essentially completely ablated. As the majority of inflam- matory monocytes recruited to the VACV lesion become infected, it is likely that infection of these cells may play a role in the tissue-protective phenotype that they mediate. Here, we dem- onstrate that inflammatory monocytes are infected, but the blockade in DNA replication in PLOS Pathogens | https://doi.org/10.1371/journal.ppat.1007778 October 11, 2019 15 / 25 Type I interferon-dependent CCL4 expression protects virus-infected tissue the majority of cells prevents the production of infectious virions. Defective recruitment of -/- monocytes may account for the slight increase in VACV titer we recovered from cGAS and -/- IFNαR mice compared to wild-type mice, as we observed a similar increase in titer in mice in which we have used multiple methodologies to deplete monocytes at the site of VACV infec- tion, but which retain intact T1-IFN induction [55]. This further suggests that a lack of T1-IFN does not render ear tissue more permissive to VACV replication. We propose that inflammatory monocytes at the site of infection can function as a virus sponge, thereby limit- ing the proportion of ear tissue that is infected, and subsequently eliminated by the adaptive -/- immune response. Indeed, surrounding ear tissue from IFNαR mice had a greater propor- tion of VACV-infected cells compared to wild-type mice, but VACV titers were not substan- tially increased, supporting this model. It is also possible that inflammatory monocytes may facilitate a response to “wall-off” the infected tissue and promote local wound healing responses. In either way, inflammatory monocytes can moderate virus-dependent tissue pathology in a T1-IFN-dependent manner that is independent of the induction of classical antiviral ISGs or inhibition of virus replication. Taken as a whole, our findings provide a more complete picture of the way in which the host uses multiple mechanisms to bypass a myriad of viral evasion strategies, and thus still pro- duce and utilize T1-IFN in a protective antiviral capacity. Induction of T1-IFN requires infec- tion of specialized populations of tissue-resident cells that can subvert the viral blockade of cGAS/STING sensing. The infected cell population only produces “early” species of T1-IFN, likely due to the effects of soluble T1-IFN binding proteins and a blockade in STAT1 signaling within infected cells. However, the small amount of T1-IFN made is sufficient to induce a posi- tive feedback loop of T1-IFN within infected cells to stimulate downstream expression of ccl4 and subsequent recruitment of inflammatory monocytes. The inflammatory monocytes become the major infected cell population within the lesion, but are non-productively infected, effectively throwing themselves on the exploding grenade of the infection. Materials and methods Mice Mice were on the C57BL/6 background unless otherwise stated. Wild-type (catalog no. 00664), -/- gt/gt -/- cGAS (catalog no. 026554), STING (catalog no. 017537), MyD88 (catalog no. 009088), Lps/Lps -/- TRIF (catalog no. 005037), MAVS (catalog no. 008634 on the B6.129 background), -/- +/- STAT2 (catalog no. 023309), and STAT1 (catalog no. 012606, subsequently bred to pro- -/- duce homozygous STAT1 ) mice were purchased from the Jackson Laboratory and housed at -/- the Penn State College of Medicine. IFNαR (Jax catalog no. 32045) mice were a gift from Dr. Ziaur Rahman (Penn State College of Medicine). All mice were maintained in the specific- pathogen-free facility at the Penn State College of Medicine. In vivo experiments utilized 6- 12-week-old mice. All animals were treated in accordance with the National Institutes of Health and AAALAC International Regulations. The Penn State College of Medicine Institu- tional Animal Care and Use Committee approved all experiments. Ethics statement All animals were maintained in the specific-pathogen-free facility of the Hershey Medical Cen- ter and treated in accordance with the National Institutes of Health and AAALAC Interna- tional regulations. All animal experiments and procedures were approved in protocols 47731, 47033, 47597, 46111 and 44723 by the Penn State Hershey Institutional Animal Care and Use Committee (Animal Welfare Assurance # A3045-01) that follows the Office of Laboratory Ani- mal Welfare PHS Policy on Humane Care and Use of Laboratory Animals, 2015. PLOS Pathogens | https://doi.org/10.1371/journal.ppat.1007778 October 11, 2019 16 / 25 Type I interferon-dependent CCL4 expression protects virus-infected tissue Viruses and cell lines VACV (strain Western Reserve), or recombinant viruses (see below), were grown and purified as previously described [55]. Heat-inactivation of VACV was achieved by incubating aliquots of virus at 56˚C in a shaking incubator for 1 hour just prior to infection. To easily identify VACV-infected cells in vivo, mice were infected with a recombinant VACV expressing influ- enza nucleoprotein tagged with eGFP, in which fluorescent protein accumulates in the nucleus of infected cells (VACV-NP-S-eGFP) [74]. In order to identify virus-encoded RNA, we used VACV that expresses the ovalbumin gene from Gallus gallus (VACV-OVA) [78]. VSV was a kind gift from the late Dr. Leo Lefrancois and was grown and titered on BHK cells using stan- dard protocols. To identify infected cells in which abortive or productive infection was proceeding, we uti- lized VACV-eGFP-OVA that expresses eGFP coupled with OVA under the control of either the early/late p7.5 promoter (VACV Early-GFP) or the late promoter p11 (VACV Late-GFP). VACV-eGFP-OVA early and late viruses were prepared as follows. The plasmid pRB21 expressing the full-length vp37 VACV ORF with the p7.5 early/late promoter was a kind gift from Dr. Bernard Moss (Laboratory of Viral Diseases, MIAID, Bethesda, MD). The peGFP-C1 plasmid expressing OVA was a kind gift from Dr. Kenneth Rock (Department of Pathology, University of Massachusetts Medical School, Worcester, MA). pRB21 DNA was ligated with eGFP-OVA using T4 Ligase (Invitrogen). To make rVACV-eGFP-OVA with a late promoter, the p11 promoter was inserted in place of p7.5. All plasmid DNA was sequenced to ensure that the promoter, and eGFP-OVA sequences were correct. rVACV-eGFP-OVA-Early and rVACV-eGFP-OVA-Late were generated by transfecting plasmid DNA into BSC-1 cells infected with VACV-vRB21 at an MOI of 1 using the CellPhect Transfection Kit (GE Health- care). The resulting rVACVs were plaque-purified 3x prior to characterization. The resulting rVACV-eGFP-OVA-Early and rVACV-eGFP-OVA-Late produced green fluorescence upon infection of WT3 cells and sequencing revealed the presence of the correct promoter and OVA sequences in DNA purified from virions. To make a virus that expresses CCL4, a construct was ordered that contained 357 nucleo- tides of the 3’ end of the J4 open reading frame (ORF) followed by the early VACV promoter C6, the coding sequence for mKate2 and mouse CCL4 ORFs connected by an in-frame sequence for the P2A peptide. The CCL4 ORF was followed by a 3x VACV early transcription stop signal and 400 nucleotides of the 3’ end of the J5 ORF (IDT). These constructs were amplified by standard PCR and cloned into pCR2.1 and sequenced to ensure fidelity. Recom- binant viruses were generated using homologous recombination by infecting HeLa cells with VACV WR strain and transfecting in the constructs (VACV-CCL4). Recombinant viruses were isolated based on their ability to produce red fluorescent plaques. The isolated recombi- nants were plaque-purified 4x before being amplified to produce working stocks. The integrity of the inserts was verified by sequencing. For microscopy experiments utilizing VACV-CCL4, we used VACV mCherry Ub-SIINFEKL (VACV-mCherry) [79] as the control for expression of a red fluorescent protein without CCL4. The murine keratinocyte cell line 308 was originally derived from BALB/c mouse skin initi- ated with DMBA [80]. NR9456 is a murine macrophage cell line derived from differentiated bone marrow-derived cells immortalized by retroviral transformation [81] and was obtained through BEI Resources, NIAID, NIH. 143B cells are a thymidine kinase-deficient human oste- osarcoma cell line. 143B and NR9456 cells were grown at 37˚C in DMEM supplemented with 10% FBS, 20mM HEPES, 1.5 g/L NaHCO , 0.3 g/L L-glutamine, and 0.1 g/L penicillin and streptomycin. 308 cells were grown at 37˚C in DMEM containing 4.5 g/L glucose, L-glutamine PLOS Pathogens | https://doi.org/10.1371/journal.ppat.1007778 October 11, 2019 17 / 25 Type I interferon-dependent CCL4 expression protects virus-infected tissue and sodium pyruvate supplemented with 10% FBS, 0.1 g/L penicillin and streptomycin and an additional 0.3 g/L L-glutamine. VACV infection For intradermal VACV infection, 10 pfu VACV in a 10μl volume of Hank’s Balanced Salt Solution (HBSS)/0.1% bovine serum albumin (BSA) was injected into the middle of each ear pinnae [21]. Intradermal infection causes substantial tissue loss at late times post-infection, preventing accurate measurement of titers at those times. Therefore, to assess virus titer at late times post-infection, mice were infected by scarification of the ear pinnae. Briefly, a drop of HBSS/0.1%BSA containing 10 pfu VACV was placed on the center of the ear pinnae and a 28-gauge needle was used to scratch the ear through the droplet [82]. To produce a localized infection suitable for microscopy imaging, mice were infected by dipping a bifurcated needle into HBSS/0.1%BSA containing 2x10 pfu/ml VACV and then gently puncturing each ear pin- nae ten times [19]. To evaluate tissue pathology during VACV infection, ear thickness was measured daily using a 0.0001m dial micrometer (Mitutoyo). After the appearance of a visible lesion, the diameter of the lesion or of areas of tissue loss were measured [52]. 5 6 For infection of cells in vitro, 5e -1e cells were seeded onto 24-well plates 24 hours prior to infection. Cells were infected at an MOI of 10 in 250μl HBSS/0.1%BSA for 1 hour at 37˚C, inoculum was removed and replaced with 500μl of pre-warmed media and placed at 37˚C for either an additional 5 or 23 hours. To assess the abilities of wild-type or VACV-CCL4 to repli- cate in vitro, we performed single-step and multistep growth curves. Confluent monolayers of 143B cells were infected with 5 PFU/cell of the indicated virus diluted for single-step growth curves or with 0.05 PFU/cell for multistep growth curves. Two hours post-infection, the inocu- lum was replaced with fresh medium. At the indicated time points, cells were washed, har- vested and then freeze thawed three times and sonicated to release virus, and titers were determined as described below. VACV titer Virus replication was quantified by plaque assay as previously described [52, 55]. Infected ears were harvested at the times indicated and homogenized using a TissueLyser II (Qiagen), then centrifuged for 2 minutes at 12,000 x g. Supernatant was transferred into a fresh tube and virus titer was determined by plaque assay on 143B cells. RNA isolation and RT-qPCR For gene expression in vitro, cells were infected as described above. After 6 hours total infec- tion time, media was removed and cells were lysed with 500μl TRIzol reagent (Invitrogen), transferred to a 1.7ml tube, and placed at -20˚C until RNA extraction. For RNA isolation, 100μl chloroform was added to thawed tubes, briefly vortexed and centrifuged at 12,000 x g for 15 minutes at 4˚C. The upper aqueous phase was transferred into a clean tube, 1 volume of 70% EtOH was added, and RNA was purified using the RNeasy mini kit (Qiagen) with on-col- umn DNase-I treatment. For gene expression in vivo, ears were harvested from mice at the times indicated post- infection, flash-frozen in liquid NO and placed at -80˚C. For RNA isolation, ears were homogenized using a TissueLyser II (Qiagen) and RNA was purified from tissue lysate using the RNeasy Mini Kit (Qiagen) treated with DNase-I (Qiagen) in-solution. For qPCR using Taqman Gene Expression Assays (Applied Biosystems) or Universal Probe Library (Roche), cDNA was prepared using the Hi-Capacity cDNA Synthesis Kit (Applied Bio- 2 2 systems). For qPCR using RT PCR Profiler Arrays (Qiagen), cDNA was prepared using RT PLOS Pathogens | https://doi.org/10.1371/journal.ppat.1007778 October 11, 2019 18 / 25 Type I interferon-dependent CCL4 expression protects virus-infected tissue First Strand Kit (Qiagen). qPCR was carried out on a StepOnePlus (Applied Biosystems) with either FastStart Universal Probe Master Mix (Roche) or RT SYBR Green qPCR Master Mix (Qiagen). For Taqman and FastStart Universal Probe assays, changes in gene expression are Ct expressed as fold change using theΔΔ calculation method against naïve mice of the same genotype or mock-infected cells with gapdh as the housekeeping gene. For RT PCR Profiler Array data, changes in gene expression are displayed as mean fold change between groups of mice relative to a panel of “housekeeping” genes. Amplification of bulk IFN-α transcripts was detected using Universal Probe Library probe #76 (Roche) with forward primer ARSYtgtSt- gatgcaRcaggt and reverse primer ggWacacagtgatcctgtgg; and was compared to amplification of gapdh using Universal Probe Library probe #80 (Roche) with forward primer tgtccgtcgtggatct- gac and reverse primer cctgcttcaccaccttcttg. The oligos for IFNα (non-4) amplify sequences present within most of the genes within the IFNα subfamily except Ifna4. These primers are accepted to represent expression of the IRF7-mediated subset of IFNα genes. Primers were from Integrated DNA Technologies. ELISA For detection of protein from cells infected in vitro, cells were infected as described above. After the initial 1 hour of infection, the inoculum was removed and replaced with 500μl of pre- warmed media supplemented with 2% FBS and cells were incubated at 37˚C for an additional 23 hours. After 24 hours of total infection time, media was transferred into tubes containing phenylmethylsulfonyl fluoride (PMSF) (Sigma) and a Protease Inhibitor Cocktail (Sigma) and placed at -80˚C until the ELISA was performed. For detection of protein from VACV-infected ear tissue, ears were infected and harvested as described above. Frozen ears were homogenized in PBS containing PMSF (Sigma) and Pro- tease Inhibitor Cocktail (Sigma) using a TissueLyser II (Qiagen), centrifuged for 3 min at 12,000 x g and the lysate was transferred to a clean tube. IFN-β protein was detected using the mouse Hi-Sensitivity IFN-β ELISA kit (PBL Assay Science). CCL4 protein was detected using the mouse CCL4/MIP-1 beta Quantikine ELISA kit (R&D Systems). ELISAs were carried out according to manufacturer’s specifications. Flow cytometry For the detection of the mKate2 transgene, 2e 143B cells were infected in 5ml tubes at an MOI of 10 in 500μl HBSS/0.1%BSA, incubated in a 37˚C water bath for 1 hour, 3 ml of pre- warmed media was added, and tubes were incubated in a rotating incubator at 37˚C for an additional 5 hours. Cells were washed and samples were acquired on a LSRII flow cytometer (BD Biosciences), data were analyzed using FlowJo software (TreeStar). Ears were harvested and processed as previously described [52, 55]. Briefly, ears were sepa- rated into dorsal and ventral halves and digested in 1 mg/ml collagenase XI (Sigma) for 60 minutes at 37˚C and a single cell suspension made. Live cells were incubated in Fc block and 10% normal mouse serum (Gemini Bio-Products) at 4˚C, and then stained in that solution. Cells were stained with antibodies recognizing Ly6C (clone HK1.4), CD11b (clone M1/70), Ly6G (clone 1A8), CD45 (clone 30-F11), CD64 (clone X54-5/7.1), CD11c (clone N418), CCR5 (clone HM-CCR5), CXCR3 (clone CXCR3-173), CXCR5 (clone L138D7), Streptavidin from BioLegend, CD19 (clone 1D3), CD90.2 (clone 53–2.1), and NK1.1 (clone PK136) from BD Biosciences, and CCR2 (clone REA538) from Miltenyi Biotec. Samples were acquired on a LSRII flow cytometer (BD Biosciences), and data were analyzed using FlowJo software (TreeS- + - + hi tar). Inflammatory monocytes were gated as CD45 (CD19, CD90.2, NK1.1) CD11b Ly6C PLOS Pathogens | https://doi.org/10.1371/journal.ppat.1007778 October 11, 2019 19 / 25 Type I interferon-dependent CCL4 expression protects virus-infected tissue - - + + Ly6G CD11c CD64 [55]. Non-classical tissue-protective monocytes were gated as CD45 - hi + + (CD19, CD90.2, NK1.1) CD11b Ly6C Ly6G [52]. Fluorescent imaging Mice were infected with VACV in the center of each ear pinnae using a bifurcated needle. Five days post-infection, ears were harvested and embedded in Tissue-Tek OCT (Sakura Finetek), rapidly frozen by immersion in liquid nitrogen-cooled 2-methyl butane, and stored at -80˚C. Lateral cryostat sections (10 μm) were cut at -20˚C, mounted on glass slides and fixed for 30 min in 4% paraformaldehyde in PBS (pH 7.4). For both assays, three-dimensional images were collected on an Olympus IX81 fluorescent microscope, deconvolved and analyzed using Slide- book 6.0 software (Intelligent Imaging Innovations). For RNA fluorescence in situ hybridization, mice were infected with VACV-OVA. To visu- alize plasma membranes, tissue was stained with Wheat Germ Agglutinin conjugated to CF488 (1:1000, Biotium). Tissue was fixed again and RNA was visualized using the ViewRNA Cell Plus Assay (Invitrogen) with probes recognizing murine IFN-β and gallus ovalbumin (Affymetrix) according to manufacturer’s specifications with the omission of using a protease. Nuclei were stained with DAPI. For Immunofluorescence microscopy, mice were infected with VACV-NP-S-eGFP, VACV-CCL4, or VACV-mCherry. Tissue was stained with antibodies to Ly6C (clone HK1.4 1:2500, Abcam) and visualized using secondary donkey anti-rat Alexa 647 (1:500, Jackson Immunoresearch) or donkey anti-rat BV421 (1:100, Jackson Immunoresearch). CCR5 antagonism Maraviroc (Cayman Chemical) is a noncompetitive, slowly reversible small-molecule antago- nist of CCR5 that prevents signaling by all three ligands of CCR5 [i.e.,CCL5/RANTES, macro- phage inflammatory protein 1α (MIP-1α), andMIP-1β] and is used as a part of combination antiretroviral therapy for patients infected with CCR5-tropic HIV (i.e., a subtype that uses onlythe CCR5 co-receptor to enter cells) [83]. We treated with maraviroc (10 mg/kg, i.p.) for 4 days, starting the day of infection, until harvesting VACV infected ears 5 dpi to assay mono- cyte recruitment. Data analysis Data were analyzed and graphed using Excel (Microsoft), Prism (Graphpad), and RT Profiler Array web-based software (Qiagen). Data are expressed as mean ± standard error of the mean, with individual data points shown where applicable. RT PCR Profiler Array data are expressed as mean fold change vs. p-value. Means were compared using an unpaired, two-way student’s t-test or two-way ANOVA as applicable. Data represents at least three individual experiments. Significance between groups was determined by p-value <0.05, and is displayed � � � � � � � � � � as; = p<0.05, = p<0.01, = p<0.001, = p<0.0001. Supporting information S1 Fig. (A) Expression of T1-IFN subtypes IFN-β1 andα4 in ear tissue of wild-type (B6) vs -/- CCR2 mice, in which recruitment of inflammatory monocytes to the VACV infected ear is completely ablated (55). mRNA was purified and expression levels from uninfected mice, or at 5 dpi, measured by RT-qPCR. (TIFF) PLOS Pathogens | https://doi.org/10.1371/journal.ppat.1007778 October 11, 2019 20 / 25 Type I interferon-dependent CCL4 expression protects virus-infected tissue Author Contributions Conceptualization: Nikhil J. Parekh, Christopher C. Norbury. Data curation: Nikhil J. Parekh, Tracy E. Krouse, Irene E. Reider, Christopher C. Norbury. Formal analysis: Nikhil J. Parekh, Tracy E. Krouse, Irene E. Reider, Christopher C. Norbury. Funding acquisition: Christopher C. Norbury. Investigation: Nikhil J. Parekh, Tracy E. Krouse, Irene E. Reider, Christopher C. Norbury. Methodology: Nikhil J. Parekh, Tracy E. Krouse, Irene E. Reider, Brian M. Ward, Christopher C. Norbury. Project administration: Christopher C. Norbury. Resources: Ryan P. Hobbs, Brian M. Ward, Christopher C. Norbury. Software: Christopher C. Norbury. Validation: Christopher C. Norbury. Visualization: Christopher C. Norbury. Writing – original draft: Nikhil J. Parekh, Christopher C. Norbury. Writing – review & editing: Nikhil J. Parekh, Ryan P. Hobbs, Brian M. Ward, Christopher C. Norbury. References 1. van den Broek MF, Muller U, Huang S, Zinkernagel RM, Aguet M. Immune defence in mice lacking type I and/or type II interferon receptors. Immunol Rev. 1995; 148:5–18. https://doi.org/10.1111/j.1600-065x. 1995.tb00090.x PMID: 8825279. 2. Isaacs A, Lindenmann J. Virus interference. I. The interferon. Proc R Soc Lond B Biol Sci. 1957; 147 (927):258–67. PMID: 13465720. 3. Nagano Y, Kojima Y, Sawai Y. [Immunity and interference in vaccinia; inhibition of skin infection by inac- tivated virus]. C R Seances Soc Biol Fil. 1954; 148(7–8):750–2. PMID: 13190824. 4. Pestka S, Krause CD, Walter MR. Interferons, interferon-like cytokines, and their receptors. Immunol Rev. 2004; 202:8–32. https://doi.org/10.1111/j.0105-2896.2004.00204.x PMID: 15546383. 5. Yan N, Chen ZJ. Intrinsic antiviral immunity. Nat Immunol. 2012; 13(3):214–22. https://doi.org/10.1038/ ni.2229 PMID: 22344284; PubMed Central PMCID: PMC3549670. 6. Fu XY, Schindler C, Improta T, Aebersold R, Darnell JE Jr., The proteins of ISGF-3, the interferon alpha-induced transcriptional activator, define a gene family involved in signal transduction. Proc Natl Acad Sci U S A. 1992; 89(16):7840–3. https://doi.org/10.1073/pnas.89.16.7840 PMID: 1502204; PubMed Central PMCID: PMC49807. 7. Schoggins JW, Rice CM. Interferon-stimulated genes and their antiviral effector functions. Curr Opin Virol. 2011; 1(6):519–25. https://doi.org/10.1016/j.coviro.2011.10.008 PMID: 22328912; PubMed Cen- tral PMCID: PMC3274382. 8. Schneider WM, Chevillotte MD, Rice CM. Interferon-stimulated genes: a complex web of host defenses. Annu Rev Immunol. 2014; 32:513–45. https://doi.org/10.1146/annurev-immunol-032713-120231 PMID: 24555472; PubMed Central PMCID: PMC4313732. 9. Sen GC. Viruses and interferons. Annu Rev Microbiol. 2001; 55:255–81. https://doi.org/10.1146/ annurev.micro.55.1.255 PMID: 11544356. 10. Hoffmann HH, Schneider WM, Rice CM. Interferons and viruses: an evolutionary arms race of molecu- lar interactions. Trends Immunol. 2015; 36(3):124–38. https://doi.org/10.1016/j.it.2015.01.004 PMID: 25704559; PubMed Central PMCID: PMC4384471. 11. Garcia-Sastre A. Ten Strategies of Interferon Evasion by Viruses. Cell Host Microbe. 2017; 22(2):176– 84. https://doi.org/10.1016/j.chom.2017.07.012 PubMed Central PMCID: PMC5576560. PMID: PLOS Pathogens | https://doi.org/10.1371/journal.ppat.1007778 October 11, 2019 21 / 25 Type I interferon-dependent CCL4 expression protects virus-infected tissue 12. Alcami A, Koszinowski UH. Viral mechanisms of immune evasion. 2000; 8(9):410–8. https://doi.org/10. 1016/s0966-842x(00)01830-8 PMID: 10989308 13. Cao H, Dai P, Wang W, Li H, Yuan J, Wang F, et al. Innate immune response of human plasmacytoid dendritic cells to poxvirus infection is subverted by vaccinia E3 via its Z-DNA/RNA binding domain. PLoS One. 2012; 7(5):e36823. https://doi.org/10.1371/journal.pone.0036823 PMID: 22606294; PubMed Central PMCID: PMC3351467. 14. Dai P, Wang W, Cao H, Avogadri F, Dai L, Drexler I, et al. Modified vaccinia virus Ankara triggers type I IFN production in murine conventional dendritic cells via a cGAS/STING-mediated cytosolic DNA-sens- ing pathway. PLoS Pathog. 2014; 10(4):e1003989. https://doi.org/10.1371/journal.ppat.1003989 PMID: 24743339; PubMed Central PMCID: PMC3990710. 15. Meade N, Furey C, Li H, Verma R, Chai Q, Rollins MG, et al. Poxviruses Evade Cytosolic Sensing through Disruption of an mTORC1-mTORC2 Regulatory Circuit. Cell. 2018; 174(5):1143–57 e17. https://doi.org/10.1016/j.cell.2018.06.053 PMID: 30078703; PubMed Central PMCID: PMC6172959. 16. Georgana I, Sumner RP, Towers GJ, Maluquer de Motes C. Virulent poxviruses inhibit DNA sensing by preventing STING activation. J Virol. 2018. https://doi.org/10.1128/JVI.02145-17 PMID: 29491158; PubMed Central PMCID: PMC5923072. 17. Scutts SR, Ember SW, Ren H, Ye C, Lovejoy CA, Mazzon M, et al. DNA-PK Is Targeted by Multiple Vaccinia Virus Proteins to Inhibit DNA Sensing. Cell Rep. 2018; 25(7):1953–65 e4. https://doi.org/10. 1016/j.celrep.2018.10.034 PMID: 30428360. 18. Eaglesham JB, Pan Y, Kupper TS, Kranzusch PJ. Viral and metazoan poxins are cGAMP-specific nucleases that restrict cGAS-STING signalling. Nature. 2019; 566(7743):259–63. https://doi.org/10. 1038/s41586-019-0928-6 PMID: 30728498. 19. Hickman HD, Reynoso GV, Ngudiankama BF, Rubin EJ, Magadan JG, Cush SS, et al. Anatomically restricted synergistic antiviral activities of innate and adaptive immune cells in the skin. Cell host & microbe. 2013; 13(2):155–68. Epub 2013/02/19. https://doi.org/10.1016/j.chom.2013.01.004 PMID: 23414756; PubMed Central PMCID: PMC3591514. 20. Deng L, Dai P, Parikh T, Cao H, Bhoj V, Sun Q, et al. Vaccinia virus subverts a mitochondrial antiviral signaling protein-dependent innate immune response in keratinocytes through its double-stranded RNA binding protein, E3. J Virol. 2008; 82(21):10735–46. https://doi.org/10.1128/JVI.01305-08 PMID: 18715932; PubMed Central PMCID: PMC2573174. 21. Tscharke DC, Smith GL. A model for vaccinia virus pathogenesis and immunity based on intradermal injection of mouse ear pinnae. J Gen Virol. 1999; 80 (Pt 10):2751–5. https://doi.org/10.1099/0022- 1317-80-10-2751 PMID: 10573171. 22. Tscharke DC, Reading PC, Smith GL. Dermal infection with vaccinia virus reveals roles for virus pro- teins not seen using other inoculation routes. J Gen Virol. 2002; 83(Pt 8):1977–86. https://doi.org/10. 1099/0022-1317-83-8-1977 PMID: 12124461. 23. Jaks E, Gavutis M, Uze G, Martal J, Piehler J. Differential receptor subunit affinities of type I interferons govern differential signal activation. J Mol Biol. 2007; 366(2):525–39. Epub 2006/12/19. https://doi.org/ 10.1016/j.jmb.2006.11.053 PMID: 17174979. 24. Li L, Sherry B. IFN-alpha expression and antiviral effects are subtype and cell type specific in the car- diac response to viral infection. Virology. 2010; 396(1):59–68. Epub 2009/11/10. https://doi.org/10. 1016/j.virol.2009.10.013 PMID: 19896686; PubMed Central PMCID: PMC2787694. 25. Marie I, Durbin JE, Levy DE. Differential viral induction of distinct interferon-alpha genes by positive feedback through interferon regulatory factor-7. EMBO J. 1998; 17(22):6660–9. https://doi.org/10.1093/ emboj/17.22.6660 PMID: 9822609; PubMed Central PMCID: PMC1171011. 26. Sato M, Hata N, Asagiri M, Nakaya T, Taniguchi T, Tanaka N. Positive feedback regulation of type I IFN genes by the IFN-inducible transcription factor IRF-7. FEBS Lett. 1998; 441(1):106–10. https://doi.org/ 10.1016/s0014-5793(98)01514-2 PMID: 9877175. 27. Horvath CM, Darnell JE. The state of the STATs: recent developments in the study of signal transduc- tion to the nucleus. Curr Opin Cell Biol. 1997; 9(2):233–9. https://doi.org/10.1016/s0955-0674(97) 80067-1 PMID: 9069254. 28. Levy DE, Marie IJ, Durbin JE. Induction and function of type I and III interferon in response to viral infec- tion. Curr Opin Virol. 2011; 1(6):476–86. https://doi.org/10.1016/j.coviro.2011.11.001 PMID: 22323926; PubMed Central PMCID: PMC3272644. 29. Bowie A, Kiss-Toth E, Symons JA, Smith GL, Dower SK, O’Neill LA. A46R and A52R from vaccinia virus are antagonists of host IL-1 and toll-like receptor signaling. Proc Natl Acad Sci U S A. 2000; 97 (18):10162–7. https://doi.org/10.1073/pnas.160027697 PMID: 10920188. 30. Stack J, Haga IR, Schroder M, Bartlett NW, Maloney G, Reading PC, et al. Vaccinia virus protein A46R targets multiple Toll-like-interleukin-1 receptor adaptors and contributes to virulence. J Exp Med. 2005; 201(6):1007–18. https://doi.org/10.1084/jem.20041442 PMID: 15767367. PLOS Pathogens | https://doi.org/10.1371/journal.ppat.1007778 October 11, 2019 22 / 25 Type I interferon-dependent CCL4 expression protects virus-infected tissue 31. Zhang P, Langland JO, Jacobs BL, Samuel CE. Protein kinase PKR-dependent activation of mitogen- activated protein kinases occurs through mitochondrial adapter IPS-1 and is antagonized by vaccinia virus E3L. J Virol. 2009; 83(11):5718–25. https://doi.org/10.1128/JVI.00224-09 PMID: 19321614; PubMed Central PMCID: PMC2681938. 32. Davies ML, Sei JJ, Siciliano NA, Xu RH, Roscoe F, Sigal LJ, et al. MyD88-dependent immunity to a nat- ural model of vaccinia virus infection does not involve Toll-like receptor 2. J Virol. 2014; 88(6):3557–67. https://doi.org/10.1128/JVI.02776-13 PMID: 24403581; PubMed Central PMCID: PMC3957935. 33. Martinez J, Huang X, Yang Y. Direct TLR2 signaling is critical for NK cell activation and function in response to vaccinia viral infection. PLoS pathogens. 2010; 6(3):e1000811. Epub 2010/03/20. https:// doi.org/10.1371/journal.ppat.1000811 PMID: 20300608; PubMed Central PMCID: PMC2837413. 34. Quigley M, Martinez J, Huang X, Yang Y. A critical role for direct TLR2-MyD88 signaling in CD8 T-cell clonal expansion and memory formation following vaccinia viral infection. Blood. 2009; 113(10):2256– 64. Epub 2008/10/25. https://doi.org/10.1182/blood-2008-03-148809 PMID: 18948575; PubMed Cen- tral PMCID: PMC2652371. 35. Zhao Y, De Trez C, Flynn R, Ware CF, Croft M, Salek-Ardakani S. The adaptor molecule MyD88 directly promotes CD8 T cell responses to vaccinia virus. Journal of immunology. 2009; 182(10):6278– 86. Epub 2009/05/06. https://doi.org/10.4049/jimmunol.0803682 PMID: 19414781; PubMed Central PMCID: PMC2712123. 36. Martinez J, Huang X, Yang Y. Toll-like receptor 8-mediated activation of murine plasmacytoid dendritic cells by vaccinia viral DNA. Proceedings of the National Academy of Sciences of the United States of America. 2010; 107(14):6442–7. Epub 2010/03/24. https://doi.org/10.1073/pnas.0913291107 PMID: 20308556; PubMed Central PMCID: PMC2851984. 37. Hoebe K, Du X, Georgel P, Janssen E, Tabeta K, Kim SO, et al. Identification of Lps2 as a key trans- ducer of MyD88-independent TIR signalling. Nature. 2003; 424(6950):743–8. https://doi.org/10.1038/ nature01889 PMID: 12872135. 38. Meade N, King M, Munger J, Walsh D. mTOR Dysregulation by Vaccinia Virus F17 Controls Multiple Processes with Varying Roles in Infection. J Virol. 2019;93(15). https://doi.org/10.1128/JVI.00784-19 PMID: 31118254. 39. Schoggins JW, MacDuff DA, Imanaka N, Gainey MD, Shrestha B, Eitson JL, et al. Pan-viral specificity of IFN-induced genes reveals new roles for cGAS in innate immunity. Nature. 2014; 505(7485):691–5. https://doi.org/10.1038/nature12862 PMID: 24284630; PubMed Central PMCID: PMC4077721. 40. Lian H, Wei J, Zang R, Ye W, Yang Q, Zhang XN, et al. ZCCHC3 is a co-sensor of cGAS for dsDNA rec- ognition in innate immune response. Nat Commun. 2018; 9(1):3349. https://doi.org/10.1038/s41467- 018-05559-w PMID: 30135424; PubMed Central PMCID: PMC6105683. 41. Abe T, Barber GN. Cytosolic-DNA-mediated, STING-dependent proinflammatory gene induction necessitates canonical NF-kappaB activation through TBK1. J Virol. 2014; 88(10):5328–41. https://doi. org/10.1128/JVI.00037-14 PMID: 24600004; PubMed Central PMCID: PMC4019140. 42. Fang R, Wang C, Jiang Q, Lv M, Gao P, Yu X, et al. NEMO-IKKbeta Are Essential for IRF3 and NF-kap- paB Activation in the cGAS-STING Pathway. J Immunol. 2017; 199(9):3222–33. https://doi.org/10. 4049/jimmunol.1700699 PMID: 28939760. 43. Cai X, Chiu YH, Chen ZJ. The cGAS-cGAMP-STING pathway of cytosolic DNA sensing and signaling. Mol Cell. 2014; 54(2):289–96. https://doi.org/10.1016/j.molcel.2014.03.040 PMID: 24766893. 44. Chen H, Sun H, You F, Sun W, Zhou X, Chen L, et al. Activation of STAT6 by STING is critical for antivi- ral innate immunity. Cell. 2011; 147(2):436–46. Epub 2011/10/18. https://doi.org/10.1016/j.cell.2011. 09.022 PMID: 22000020. 45. Bhat N, Fitzgerald KA. Recognition of cytosolic DNA by cGAS and other STING-dependent sensors. Eur J Immunol. 2014; 44(3):634–40. https://doi.org/10.1002/eji.201344127 PMID: 24356864; PubMed Central PMCID: PMC4621431. 46. Rubio D, Xu RH, Remakus S, Krouse TE, Truckenmiller ME, Thapa RJ, et al. Cross-talk between the Type I interferon and Nuclear Factor Kappa B pathways rescues resistance to a viral disease. Cell Host Microbe. 2013;In press. 47. Howell MD, Gao P, Kim BE, Lesley LJ, Streib JE, Taylor PA, et al. The signal transducer and activator of transcription 6 gene (STAT6) increases the propensity of patients with atopic dermatitis toward dis- seminated viral skin infections. J Allergy Clin Immunol. 2011; 128(5):1006–14. https://doi.org/10.1016/j. jaci.2011.06.003 PMID: 21762972; PubMed Central PMCID: PMC3205328. 48. Di Pilato M, Mejias-Perez E, Sorzano COS, Esteban M. Distinct Roles of Vaccinia Virus NF-kappaB Inhibitor Proteins A52, B15, and K7 in the Immune Response. J Virol. 2017;91(13). https://doi.org/10. 1128/JVI.00575-17 PMID: 28424281; PubMed Central PMCID: PMC5469261. PLOS Pathogens | https://doi.org/10.1371/journal.ppat.1007778 October 11, 2019 23 / 25 Type I interferon-dependent CCL4 expression protects virus-infected tissue 49. Howell MD, Gallo RL, Boguniewicz M, Jones JF, Wong C, Streib JE, et al. Cytokine milieu of atopic der- matitis skin subverts the innate immune response to vaccinia virus. Immunity. 2006; 24(3):341–8. Epub 2006/03/21. https://doi.org/10.1016/j.immuni.2006.02.006 PMID: 16546102. 50. Mahalingam S, Karupiah G, Takeda K, Akira S, Matthaei KI, Foster PS. Enhanced resistance in STAT6-deficient mice to infection with ectromelia virus. Proc Natl Acad Sci U S A. 2001; 98(12):6812– 7. https://doi.org/10.1073/pnas.111151098 PMID: 11371617. 51. Ember SW, Ren H, Ferguson BJ, Smith GL. Vaccinia virus protein C4 inhibits NF-kappaB activation and promotes virus virulence. The Journal of general virology. 2012; 93(Pt 10):2098–108. Epub 2012/ 07/14. https://doi.org/10.1099/vir.0.045070-0 PMID: 22791606; PubMed Central PMCID: PMC3541790. 52. Fischer MA, Davies ML, Reider IE, Heipertz EL, Epler MR, Sei JJ, et al. CD11b(+), Ly6G(+) cells pro- duce type I interferon and exhibit tissue protective properties following peripheral virus infection. PLoS Pathog. 2011; 7(11):e1002374. https://doi.org/10.1371/journal.ppat.1002374 PMID: 22102816; PubMed Central PMCID: PMC3213107. 53. Gale M Jr., Understanding the Expanding Roles of Interferon and Cytokines in Health and Disease. J Interferon Cytokine Res. 2018; 38(5):195–6. https://doi.org/10.1089/jir.2018.29007-mgj PMID: 54. Gale M Jr., Looking Forward in Interferon and Cytokine Research: Defining Effector Genes of Interferon and Cytokine Actions. J Interferon Cytokine Res. 2018; 38(6):235. https://doi.org/10.1089/jir.2018. 29008.mgj PMID: 29920132. 55. Davies ML, Parekh NJ, Kaminsky LW, Soni C, Reider IE, Krouse TE, et al. A systemic macrophage response is required to contain a peripheral poxvirus infection. PLoS Pathog. 2017; 13(6):e1006435. https://doi.org/10.1371/journal.ppat.1006435 PMID: 28614386; PubMed Central PMCID: PMC5484545. 56. Iannacone M, Moseman EA, Tonti E, Bosurgi L, Junt T, Henrickson SE, et al. Subcapsular sinus macro- phages prevent CNS invasion on peripheral infection with a neurotropic virus. Nature. 2010; 465 (7301):1079–83. Epub 2010/06/26. https://doi.org/10.1038/nature09118 PMID: 20577213; PubMed Central PMCID: PMC2892812. 57. Junt T, Moseman EA, Iannacone M, Massberg S, Lang PA, Boes M, et al. Subcapsular sinus macro- phages in lymph nodes clear lymph-borne viruses and present them to antiviral B cells. Nature. 2007; 450(7166):110–4. Epub 2007/10/16. https://doi.org/10.1038/nature06287 PMID: 17934446. 58. Moseman EA, Iannacone M, Bosurgi L, Tonti E, Chevrier N, Tumanov A, et al. B cell maintenance of subcapsular sinus macrophages protects against a fatal viral infection independent of adaptive immu- nity. Immunity. 2012; 36(3):415–26. https://doi.org/10.1016/j.immuni.2012.01.013 PMID: 22386268; PubMed Central PMCID: PMC3359130. 59. Frederico B, Chao B, Lawler C, May JS, Stevenson PG. Subcapsular sinus macrophages limit acute gammaherpesvirus dissemination. J Gen Virol. 2015; 96(8):2314–27. https://doi.org/10.1099/vir.0. 000140 PMID: 25872742. 60. Farrell HE, Davis-Poynter N, Bruce K, Lawler C, Dolken L, Mach M, et al. Lymph Node Macrophages Restrict Murine Cytomegalovirus Dissemination. J Virol. 2015; 89(14):7147–58. https://doi.org/10.1128/ JVI.00480-15 PMID: 25926638; PubMed Central PMCID: PMC4473555. 61. Broder CC, Kennedy PE, Michaels F, Berger EA. Expression of foreign genes in cultured human pri- mary macrophages using recombinant vaccinia virus vectors. Gene. 1994; 142(2):167–74. https://doi. org/10.1016/0378-1119(94)90257-7 PMID: 8194748. 62. Ablasser A, Schmid-Burgk JL, Hemmerling I, Horvath GL, Schmidt T, Latz E, et al. Cell intrinsic immu- nity spreads to bystander cells via the intercellular transfer of cGAMP. Nature. 2013; 503(7477):530–4. Epub 2013/10/01. https://doi.org/10.1038/nature12640 PMID: 24077100. 63. Xu S, Ducroux A, Ponnurangam A, Vieyres G, Franz S, Musken M, et al. cGAS-Mediated Innate Immu- nity Spreads Intercellularly through HIV-1 Env-Induced Membrane Fusion Sites. Cell Host Microbe. 2016; 20(4):443–57. https://doi.org/10.1016/j.chom.2016.09.003 PMID: 27736643. 64. Rice AD, Turner PC, Embury JE, Moldawer LL, Baker HV, Moyer RW. Roles of vaccinia virus genes E3L and K3L and host genes PKR and RNase L during intratracheal infection of C57BL/6 mice. J Virol. 2011; 85(1):550–67. https://doi.org/10.1128/JVI.00254-10 PMID: 20943971; PubMed Central PMCID: PMC3014211. 65. Rivas C, Gil J, Melkova Z, Esteban M, Diaz-Guerra M. Vaccinia virus E3L protein is an inhibitor of the interferon (i.f.n.)-induced 2-5A synthetase enzyme. Virology. 1998; 243(2):406–14. https://doi.org/10. 1006/viro.1998.9072 PMID: 9568039. 66. Chang HW, Watson JC, Jacobs BL. The E3L gene of vaccinia virus encodes an inhibitor of the inter- feron-induced, double-stranded RNA-dependent protein kinase. Proc Natl Acad Sci U S A. 1992; 89 PLOS Pathogens | https://doi.org/10.1371/journal.ppat.1007778 October 11, 2019 24 / 25 Type I interferon-dependent CCL4 expression protects virus-infected tissue (11):4825–9. https://doi.org/10.1073/pnas.89.11.4825 PMID: 1350676; PubMed Central PMCID: PMC49180. 67. Odendall C, Dixit E, Stavru F, Bierne H, Franz KM, Durbin AF, et al. Diverse intracellular pathogens acti- vate type III interferon expression from peroxisomes. Nat Immunol. 2014; 15(8):717–26. https://doi.org/ 10.1038/ni.2915 PMID: 24952503; PubMed Central PMCID: PMC4106986. 68. Alcami A, Symons JA, Smith GL. The vaccinia virus soluble alpha/beta interferon (IFN) receptor binds to the cell surface and protects cells from the antiviral effects of IFN. J Virol. 2000; 74(23):11230–9. https://doi.org/10.1128/jvi.74.23.11230-11239.2000 PMID: 11070021. 69. Najarro P, Traktman P, Lewis JA. Vaccinia virus blocks gamma interferon signal transduction: viral VH1 phosphatase reverses Stat1 activation. J Virol. 2001; 75(7):3185–96. https://doi.org/10.1128/JVI.75.7. 3185-3196.2001 PMID: 11238845; PubMed Central PMCID: PMC114112. 70. Baldanta S, Fernandez-Escobar M, Acin-Perez R, Albert M, Camafeita E, Jorge I, et al. ISG15 governs mitochondrial function in macrophages following vaccinia virus infection. PLoS Pathog. 2017; 13(10): e1006651. https://doi.org/10.1371/journal.ppat.1006651 PMID: 29077752; PubMed Central PMCID: PMC5659798. 71. Guerra S, Caceres A, Knobeloch KP, Horak I, Esteban M. Vaccinia virus E3 protein prevents the antivi- ral action of ISG15. PLoS Pathog. 2008; 4(7):e1000096. https://doi.org/10.1371/journal.ppat.1000096 PMID: 18604270; PubMed Central PMCID: PMC2434199. 72. Haga IR, Bowie AG. Evasion of innate immunity by vaccinia virus. Parasitology. 2005; 130 Suppl:S11– 25. https://doi.org/10.1017/S0031182005008127 PMID: 16281988. 73. Harte MT, Haga IR, Maloney G, Gray P, Reading PC, Bartlett NW, et al. The poxvirus protein A52R tar- gets Toll-like receptor signaling complexes to suppress host defense. J Exp Med. 2003; 197(3):343–51. https://doi.org/10.1084/jem.20021652 PMID: 12566418. 74. Norbury CC, Malide D, Gibbs JS, Bennink JR, Yewdell JW. Visualizing priming of virus-specific CD8+ T cells by infected dendritic cells in vivo. Nat Immunol. 2002; 3(3):265–71. https://doi.org/10.1038/ni762 PMID: 11828323. 75. Lehmann MH, Kastenmuller W, Kandemir JD, Brandt F, Suezer Y, Sutter G. Modified vaccinia virus ankara triggers chemotaxis of monocytes and early respiratory immigration of leukocytes by induction of CCL2 expression. J Virol. 2009; 83(6):2540–52. https://doi.org/10.1128/JVI.01884-08 PMID: 19129447; PubMed Central PMCID: PMC2648269. 76. Lehmann MH, Torres-Dominguez LE, Price PJ, Brandmuller C, Kirschning CJ, Sutter G. CCL2 expres- sion is mediated by type I IFN receptor and recruits NK and T cells to the lung during MVA infection. J Leukoc Biol. 2016; 99(6):1057–64. https://doi.org/10.1189/jlb.4MA0815-376RR PMID: 26992431. 77. Bonduelle O, Duffy D, Verrier B, Combadiere C, Combadiere B. Cutting edge: Protective effect of CX3CR1+ dendritic cells in a vaccinia virus pulmonary infection model. Journal of immunology. 2012; 188(3):952–6. Epub 2012/01/06. https://doi.org/10.4049/jimmunol.1004164 PMID: 22219332. 78. Restifo NP, Bacik I, Irvine KR, Yewdell JW, McCabe BJ, Anderson RW, et al. Antigen processing in vivo and the elicitation of primary CTL responses. J Immunol. 1995; 154(9):4414–22. PMID: 7722298. 79. Lev A, Dimberu P, Das SR, Maynard JC, Nicchitta CV, Bennink JR, et al. Efficient cross-priming of anti- viral CD8+ T cells by antigen donor cells is GRP94 independent. J Immunol. 2009; 183(7):4205–10. https://doi.org/10.4049/jimmunol.0901828 PMID: 19752220; PubMed Central PMCID: PMC2749969. 80. Strickland JE, Greenhalgh DA, Koceva-Chyla A, Hennings H, Restrepo C, Balaschak M, et al. Develop- ment of murine epidermal cell lines which contain an activated rasHa oncogene and form papillomas in skin grafts on athymic nude mouse hosts. Cancer Res. 1988; 48(1):165–9. PMID: 3121168. 81. Kim J, Huh J, Hwang M, Kwon EH, Jung DJ, Brinkmann MM, et al. Acidic amino acid residues in the jux- tamembrane region of the nucleotide-sensing TLRs are important for UNC93B1 binding and signaling. J Immunol. 2013; 190(10):5287–95. https://doi.org/10.4049/jimmunol.1202767 PMID: 23585677. 82. Tian T, Jin MQ, Dubin K, King SL, Hoetzenecker W, Murphy GF, et al. IL-1R Type 1-Deficient Mice Demonstrate an Impaired Host Immune Response against Cutaneous Vaccinia Virus Infection. J Immu- nol. 2017; 198(11):4341–51. https://doi.org/10.4049/jimmunol.1500106 PMID: 28468973; PubMed Central PMCID: PMC5506850. 83. Ochoa-Callejero L, Perez-Martinez L, Rubio-Mediavilla S, Oteo JA, Martinez A, Blanco JR. Maraviroc, a CCR5 antagonist, prevents development of hepatocellular carcinoma in a mouse model. PLoS One. 2013; 8(1):e53992. Epub 2013/01/18. https://doi.org/10.1371/journal.pone.0053992 PMID: 23326556; PubMed Central PMCID: PMC3541191. PLOS Pathogens | https://doi.org/10.1371/journal.ppat.1007778 October 11, 2019 25 / 25 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png PLoS Pathogens Public Library of Science (PLoS) Journal

Type I interferon-dependent CCL4 is induced by a cGAS/STING pathway that bypasses viral inhibition and protects infected tissue, independent of viral burden

Loading next page...
 
/lp/public-library-of-science-plos-journal/type-i-interferon-dependent-ccl4-is-induced-by-a-cgas-sting-pathway-uFy2p7K0NN

References (87)

Publisher
Public Library of Science (PLoS) Journal
Copyright
Copyright: © 2019 Parekh et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Data Availability: All relevant data are within the manuscript and its Supporting Information files. Funding: This work was supported by CCN - R21 AI097787, R21 AI115230, R21 AI121876, R21 AI133536, R21 AI140014 all from the National Institute of Allergy and Infectious Diseases. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Competing interests: The authors have declared that no competing interests exist.
ISSN
1553-7366
eISSN
1553-7374
DOI
10.1371/journal.ppat.1007778
Publisher site
See Article on Publisher Site

Abstract

OPENACCESS Type I interferons (T1-IFN) are critical in the innate immune response, acting upon infected Citation: Parekh NJ, Krouse TE, Reider IE, Hobbs RP, Ward BM, Norbury CC (2019) Type I and uninfected cells to initiate an antiviral state by expressing genes that inhibit multiple interferon-dependent CCL4 is induced by a cGAS/ stages of the lifecycle of many viruses. T1-IFN triggers the production of Interferon-Stimu- STING pathway that bypasses viral inhibition and lated Genes (ISGs), activating an antiviral program that reduces virus replication. The protects infected tissue, independent of viral importance of the T1-IFN response is highlighted by the evolution of viral evasion strategies burden. PLoS Pathog 15(10): e1007778. https:// doi.org/10.1371/journal.ppat.1007778 to inhibit the production or action of T1-IFN in virus-infected cells. T1-IFN is produced via activation of pathogen sensors within infected cells, a process that is targeted by virus- Editor: Victor Robert DeFilippis, Oregon Health and Sciences University, UNITED STATES encoded immunomodulatory molecules. This is probably best exemplified by the prototypic poxvirus, Vaccinia virus (VACV), which uses at least 6 different mechanisms to completely Received: April 16, 2019 block the production of T1-IFN within infected cells in vitro. Yet, mice lacking aspects of T1- Accepted: September 16, 2019 IFN signaling are often more susceptible to infection with many viruses, including VACV, Published: October 11, 2019 than wild-type mice. How can these opposing findings be rationalized? The cytosolic DNA Copyright:© 2019 Parekh et al. This is an open sensor cGAS has been implicated in immunity to VACV, but has yet to be linked to the pro- access article distributed under the terms of the duction of T1-IFN in response to VACV infection. Indeed, there are two VACV-encoded pro- Creative Commons Attribution License, which teins that effectively prevent cGAS-mediated activation of T1-IFN. We find that the majority permits unrestricted use, distribution, and reproduction in any medium, provided the original of VACV-infected cells in vivo do not produce T1-IFN, but that a small subset of VACV- author and source are credited. infected cells in vivo utilize cGAS to sense VACV and produce T1-IFN to protect infected Data Availability Statement: All relevant data are mice. The protective effect of T1-IFN is not mediated via ISG-mediated control of virus repli- within the manuscript and its Supporting cation. Rather, T1-IFN drives increased expression of CCL4, which recruits inflammatory Information files. monocytes that constrain the VACV lesion in a virus replication-independent manner by lim- Funding: This work was supported by CCN - R21 iting spread within the tissue. Our findings have broad implications in our understanding of AI097787, R21 AI115230, R21 AI121876, R21 pathogen detection and viral evasion in vivo, and highlight a novel immune strategy to pro- AI133536, R21 AI140014 all from the National tect infected tissue. Institute of Allergy and Infectious Diseases. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. PLOS Pathogens | https://doi.org/10.1371/journal.ppat.1007778 October 11, 2019 1 / 25 Type I interferon-dependent CCL4 expression protects virus-infected tissue Competing interests: The authors have declared that no competing interests exist. Author summary The recognition of virus infection leads to a quick and robust antiviral response mediated by type I interferons (T1-IFN). Nearly all viruses have acquired genes that block the induction or action of T1-IFN in order to attain a replicative advantage. Some viruses thwart the T1-IFN response so thoroughly, that cells infected in vitro do not produce any T1-IFN. And yet, animal models with defects in T1-IFN signaling are more sensitive to infection with these viruses than their wild-type counterparts. In this study, we find evi- dence to explain these otherwise contradicting findings. We show that a small population of infected cells in vivo are able to utilize a pathogen-sensing pathway that is completely blocked in vitro. T1-IFN produced by these specialized cells protects mice by recruiting inflammatory monocytes that restrict the spread of virus within infected tissue, indepen- dent of viral burden. Our findings have a direct impact on our understanding of how viruses are detected in infected tissue, and present a novel strategy of the immune system to limit pathology at peripheral sites of infection. Introduction Type I interferons (T1-IFN) are a critical component of the response to many viruses [1]. First described in the 1950’s, cells infected with a virus produce T1-IFNs which induce an antiviral state in surrounding cells [2, 3]. All T1-IFNs signal through a common receptor, the T1-IFN receptor (IFNαR), and consist of a single IFN-β, about a dozen similar IFN-α, and IFN-ε,κ, z, δ, τ, ω, and υ [4]. T1-IFN production is triggered by recognition of viral proteins or nucleic acids by host Pattern Recognition Receptors (PRRs) [5]. PRRs can be cell surface or endosomal Toll-like receptors (TLRs) or cytosolic nucleic acid sensors. IFNαR signaling initiates a positive feedback loop that produces more T1-IFN and expression of the full breadth of IFN-α subtypes [6]. The antiviral state is the effect of hundreds of Interferon-Stimulated Genes (ISGs) driven by IFNαR signaling [7, 8], which function to inhibit virus entry, cell growth, and virus replica- tion. This is the major mechanism that controls virus replication and host tissue pathology early after infection. The importance of the T1-IFN pathway in antiviral defense is underscored by two observa- tions. First, animals lacking IFNαR are highly susceptible to virus infection [1, 9]. Second, the majority of animal viruses encode at least one, but usually many, immunomodulatory proteins that interfere with, and often completely block, the production or action of T1-IFN [10, 11]. If viruses have evolved to block T1-IFN so efficiently, why is T1-IFN so important in resistance to virus infection in vivo? Here, we compare the in vitro and in vivo responses to a virus that infects peripherally to answer this question. Viruses have evolved to evade the T1-IFN response using multiple strategies. First, many viruses can block the pathways that sense infection and trigger T1-IFN production by prevent- ing PRRs from accessing their ligands, or by blocking PRR activation or signaling [10, 11]. Viruses can also produce soluble cytokine-binding proteins that bind to, and prevent the action of, secreted T1-IFN upon neighboring cells, and can also encode proteins that inhibit signaling through IFNαR within infected cells, preventing ISG expression [11, 12]. However, despite all of these approaches deployed by viruses to subvert the host interferon response, the T1-IFN system remains an important effector arm of the innate immune response following virus infection. PLOS Pathogens | https://doi.org/10.1371/journal.ppat.1007778 October 11, 2019 2 / 25 Type I interferon-dependent CCL4 expression protects virus-infected tissue We have utilized a viral infection model which exemplifies the ability of viruses to evade the T1-IFN response, the prototypic poxvirus vector, Vaccinia virus (VACV). We find a complete blockade in production of T1-IFN by VACV-infected cells in vitro [13, 14]. VACV has been reported to block the action of multiple cell surface, endosomal or cytosolic PRRs, including the predominant cytosolic DNA sensor cyclic GMP-AMP synthase (cGAS) and downstream signal transduction by its adapter protein Stimulator of Interferon Genes (STING) [15–18]. Indeed, we find evidence of efficient action of multiple mechanisms of viral evasion in vivo. Nevertheless, T1-IFN is still produced by a subset of infected cells in vivo in a cGAS- and STING-dependent manner. The T1-IFN produced protects mice against tissue pathology caused by VACV infection. However, the protective effect of T1-IFN is not mediated by the classical mechanism of control of virus replication, as neither ISG induction nor VACV repli- cation are altered in the absence of T1-IFN signaling. Rather, the protective effect is mediated through T1-IFN-enhanced production of the chemokine CCL4, which is required for localiza- tion of inflammatory monocytes to the VACV lesion. Monocytes enter the VACV lesion, where they become non-productively infected, allowing them to soak up excess virions and prevent spread to neighboring cells. Therefore, in spite of all the mechanisms deployed against T1-IFN by viruses, the host stays ahead of pathogen evolution by deploying alternative strate- gies to both produce T1-IFN, and to mediate protective strategies downstream of T1-IFN signaling. Results T1-IFN is not expressed within infected cells in vitro, but T1-IFN expression is required in vivo The primary cell types infected with VACV after immunization in the skin are keratinocytes and recruited monocyte/macrophages [19]. We infected representative cell lines with VACV and measured the production of IFN-β. Neither a keratinocyte line (Fig 1A) nor a macrophage cell line (Fig 1B) produced T1-IFN, measured by mRNA or protein, following exposure to live VACV. However, both the keratinocyte and macrophage lines produced significant quantities of IFN-β mRNA and protein following exposure to controls, namely Vesicular stomatitis virus or heat-inactivated VACV, respectively (Fig 1A and 1B). Therefore, the expression of multiple (� 6) immune evasion proteins blocked T1-IFN production from cells present at the site of infection in the skin in response to live VACV [15, 16, 20]. In marked contrast, when we infected WT mice or mice lacking the T1-IFN receptor -/- (IFNαR ) with VACV intradermally in the ear [21, 22] there was a significant difference in the tissue response to the virus. This was apparent as early as 3 days post-infection, when swell- -/- ing at the infection site was greater in IFNαR mice (Fig 1C), and progressed to produce a larger VACV lesion (Fig 1D), greater tissue loss (Fig 1E) and overall tissue damage (lesion size -/- + tissue loss) (Fig 1F) in IFNαR mice. Fig 1G shows that VACV infection causes only a small hole in the tissue of wild-type mice, but the majority of the ear is lost following infection -/- of IFNαR mice. Therefore, T1-IFN signaling is clearly required to moderate tissue damage following dermal VACV infection in vivo. To measure T1-IFN production in vivo, we har- vested tissue from VACV-infected ears and assayed for induction of IFN-β and IFN-α mRNA. Even when utilizing stringent RNA purification procedures, no significant induction of IFN mRNA could be detected prior to 3 days post-infection (Fig 1H). However, expression of IFN- β and IFN-α mRNA could be detected at 3 days post-infection and peaked at 5 days post-infec- tion (Fig 1H). We also detected IFN-β protein by ELISA from ear tissue at the peak of mRNA production, 5 days after dermal infection (Fig 1I). Therefore, despite the blockade of T1-IFN production in infected cells by immune modulators in vitro, VACV induces T1-IFN in vivo, PLOS Pathogens | https://doi.org/10.1371/journal.ppat.1007778 October 11, 2019 3 / 25 Type I interferon-dependent CCL4 expression protects virus-infected tissue Fig 1. T1-IFN is not expressed within infected cells in vitro, but T1-IFN expression is required in vivo. (A,B) Expression of IFN-β from a murine keratinocyte (A) or macrophage (B) cell line measured by RT-qPCR 6 hpi (left) or ELISA 24 hpi (right). (C-F) Tissue pathology of the ear pinnae after intradermal infection with VACV; ear thickness (C), lesion size (D), tissue loss (E), or total tissue damage (F), n = 8 mice per group. (G) Representative pictures of ear pinnae following intradermal VACV infection. (H,I) Expression of T1-IFN in ear tissue of wild-type mice at the indicated time post-infection measured by RT-qPCR (H, n = 3 naïve, 4 VACV- infected mice per group), or 5 dpi by ELISA (I). https://doi.org/10.1371/journal.ppat.1007778.g001 and IFNαR signaling is required to prevent excessive tissue pathology following VACV infection. A positive feedback loop is required for T1-IFN production in response to VACV IFN-β and the different IFN-α subtypes have differential effects on signals transduced through the IFNαR [23]. To understand the profile of T1-IFN genes induced by VACV infection, we assayed a diverse cross-section of the IFNα family, based on a previous publication describing the antiviral activities of IFNα proteins [24], in the ear 5 days post-infection with VACV. IFN- β was strongly induced in all experiments (Fig 2A) but, although IFN-α2 is the predominant gene present in many arrays and multiplex assays, we found that it was not induced by VACV (Fig 2A). Rather, IFN-α4 and, to a lesser extent, IFN-α5 were the only IFN-α species induced by dermal VACV infection (Fig 2A). In most cells, robust T1-IFN production requires positive feedback signaling through IFNαR [25, 26]. We examined induction of T1-IFN in mice lacking PLOS Pathogens | https://doi.org/10.1371/journal.ppat.1007778 October 11, 2019 4 / 25 Type I interferon-dependent CCL4 expression protects virus-infected tissue Fig 2. A positive feedback loop is required for T1-IFN production in response to VACV. (A) Expression of T1-IFN subtypes in ear tissue of wild-type mice 5 dpi measured by RT-qPCR. (B-D) Expression of early T1-IFNs in ear tissue of mice 5 dpi measured by RT-qPCR. https://doi.org/10.1371/journal.ppat.1007778.g002 IFNαR signaling and found poor induction of both IFN-β and IFN-α4 mRNA in these mice, compared to wild-type controls (Fig 2B). IFNαR stimulation results in phosphorylation of the transcription factors Signal Transducer and Activator of Transcription (STAT) STAT1 and STAT2, which form a trimeric complex with IFN Regulatory Factor (IRF) IRF9, and are required to transduce T1-IFN signals [27, 28]. Both STAT1 (Fig 2C) and STAT2 (Fig 2D) were required for optimal induction of T1-IFN. Therefore, although virus-encoded proteins block T1-IFN from binding its receptor and the downstream activation and function of STAT proteins, a positive feedback loop is required for an efficient T1-IFN response following der- mal VACV infection. cGAS sensing of VACV-infected cells induces T1-IFN to limit local tissue pathology Induction of T1-IFN requires recognition of viral components by PRRs and signal transduction through adapter proteins. Cell surface and endosomal TLRs utilize either Myeloid Differentia- tion Primary Response 88 (MyD88) or TIR-domain-containing adapter-inducing interferon-β (TRIF) proteins to induce T1-IFN. Cytosolic nucleic acid sensors utilize the Mitochondrial Anti- viral Signaling protein (MAVS) or STING for signal transduction. Each of these pathways is reg- ularly targeted by a host of viruses, including poxviruses, which alter recognition by, or signaling downstream from, the MyD88 [29], TRIF [30], MAVS [20, 31] or STING [15, 16] pathways. We have previously found that MyD88 is required for optimal immune responses to dermal VACV infection [32], and others have found that MyD88 signaling is crucial in NK [33], T cell [34, 35] and Plasmacytoid DC [36] responses. However, we found that MyD88 was not required for the induction of IFN-β in response to VACV (Fig 3A). Although TRIF has been implicated in limiting VACV replication within infected macrophages [37], we found no defect in the induction of IFN-β in response to VACV in mice with a non-functional TRIF gene (Fig 3B). Deletion of the virus-encoded immunomodulatory protein E3L can rescue IFN production in vitro via recognition of dsRNA in a manner dependent on the PRR-adapter protein MAVS [20]. However, in vivo, MAVS-deficient mice were also able to effectively induce IFN-β in response to VACV (Fig 3C). VACV is a large dsDNA virus that carries out its lifecycle entirely in the cytoplasm of infected cells, and thus is likely sensed by cytosolic DNA sensors upstream of the adapter protein STING. Even though VACV efficiently blocks sensing via the STING pathway in vitro [15–17, 38], induction of IFN-β was heavily reliant on STING signaling (Fig 3D), indicating that recognition of VACV DNA was likely responsible for T1-IFN production. The almost total requirement for STING signaling was curious, as cells infected with VACV in vitro do not produce T1-IFN (Fig 1). Therefore, we considered the possibility that T1-IFN may only be produced by small numbers of uninfected cells, which are not subject to PLOS Pathogens | https://doi.org/10.1371/journal.ppat.1007778 October 11, 2019 5 / 25 Type I interferon-dependent CCL4 expression protects virus-infected tissue Fig 3. cGAS sensing of VACV-infected cells induces T1-IFN to limit local tissue pathology. (A-D) Expression of IFN-β in ear tissue of mice 5 dpi measured by RT-qPCR. (E) Cryosections of ear tissue from wild-type mice 5 dpi with VACV-OVA stained with Wheat-Germ Agglutinin (Grey) to visualize cell membranes and DAPI (Blue) for nuclei. RNA fluorescence in situ hybridization was used to detect viral RNA (OVA-Green) and cellular IFN-β (Red). (F,G) Expression of IFN-β in ear tissue of mice 5 dpi measured by RT-qPCR (F) or ELISA (G). (H-K) Tissue pathology of the ear pinnae after intradermal infection with VACV; ear thickness (H), lesion size (I), tissue loss (J), or total tissue damage (K), n = 9–10 mice per group. (L) Representative pictures of ear pinnae following intradermal VACV infection. https://doi.org/10.1371/journal.ppat.1007778.g003 the same strict immunomodulation as VACV-infected cells. We infected mice with a recombi- nant VACV expressing chicken egg ovalbumin (OVA), and on day 5 post-infection conducted in situ hybridization with probes targeting either OVA (to identify VACV-infected cells, in green) or IFN-β mRNA (in red). To allow accurate definition of cell boundaries, we stained with fluorescently-labeled wheat-germ agglutinin, which efficiently labeled all cell-surface membranes in the infected ear. In the majority of areas in which we could observe OVA mRNA there was little or no IFN-β mRNA visible (Fig 3E, left panel), raising the possibility that it was indeed uninfected cells that were the source of IFN-β. However, in multiple experi- ments we never observed IFN-β mRNA in cells that did not also display OVA mRNA (Fig 3E, right panel). Rather, it was a small, spatially distinct subpopulation of VACV-infected, OVA mRNA-expressing cells that produced IFN-β mRNA within the cytosol (Fig 3E). Therefore, the inhibitory effect of VACV upon STING signaling that appears complete in vitro, is bypassed in vivo, allowing T1-IFN production. Upstream of STING, cGAS has been implicated in the recognition of multiple virus infec- tions, including VACV [18, 39, 40]. However, in addition to its well established role in PLOS Pathogens | https://doi.org/10.1371/journal.ppat.1007778 October 11, 2019 6 / 25 Type I interferon-dependent CCL4 expression protects virus-infected tissue induction of T1-IFN production, cGAS can also activate NF-κB [41–43] and STAT6 [44, 45], both of which can impact the anti-poxvirus immune response [46–51]. We found that induc- tion of IFN-β mRNA and protein was impaired in cGAS-deficient mice to a similar extent to that observed in mice that lacked STING signaling (Fig 3F and 3G). To investigate a role for cGAS-induced T1-IFN during dermal VACV infection, we measured local tissue pathology in -/- -/- cGAS mice compared to wild-type mice after intradermal infection of the ear. cGAS mice displayed increased tissue thickness beginning by 3 days post-infection (Fig 3H), the time point at which we first measured induction of T1-IFN transcripts (Fig 1H). Similar to the phe- -/- -/- notype observed in IFNαR mice, cGAS mice also exhibited larger lesions, greater tissue loss and total tissue damage compared to wild-type mice (Fig 3I–3L). Therefore, cGAS-medi- ated production of T1-IFN is predominantly responsible for moderating tissue pathology fol- lowing dermal VACV infection. Thus, although cGAS-mediated induction of T1-IFN is completely blocked in vitro, the host uses this pathway to induce tissue protective T1-IFN in vivo, bypassing virus-mediated inhibitory pathways. T1-IFN does not control VACV replication in the skin Many ISGs have direct antiviral properties that can blunt propagation of virus within infected tissue [7]. Therefore, it was important to examine whether T1-IFN-mediated upregulation of antiviral ISGs in the ear was responsible for control of VACV replication, and thus whether a lack of ISG induction was responsible for the tissue pathology we observed after infection of -/- IFNαR mice. We examined the transcript levels of 81 prominent antiviral ISGs 5 days after dermal VACV infection by RT-qPCR array and found strong (>4-fold to ~300-fold) induction of a number of ISGs (Fig 4A). Surprisingly, we found that a small number of IFN-responsive transcripts were downregulated upon dermal VACV infection, including irf4, irf5 and ifi27 (Fig 4A). We also found robust induction of a number of antiviral ISGs including Ifit1, Ifi204, ISG15, OAS1, Ifih1, irf1, irf7 and irf8 (Fig 4A). However, we did not find that induction of any of the antiviral ISGs was dependent upon T1-IFN signaling (Fig 4B). This suggests that redun- dant pathway(s) exists for induction of these antiviral genes. In addition, it clearly demon- strates that upregulation or downregulation of these antiviral ISGs by T1-IFN signaling is not -/- -/- responsible for the tissue pathology phenotype we observed in IFNαR and cGAS mice. To directly address the role for T1-IFN in control of VACV replication in vivo, we com- -/- -/- pared VACV titer in the ears of cGAS and IFNαR mice to wild-type mice at 5 days post- -/- -/- infection, the peak of VACV replication [52]. Both IFNαR and cGAS mice displayed only a modest 2-fold increase in infectious VACV titer (Fig 4C) that, when compared to the 100,000 fold increase in titer in the 5 days since infection, was unlikely to explain the drastic change in pathology. VACV gene expression and infectious titer decrease after 5 days post- infection [19, 52], so it is possible that T1-IFN controls viral clearance rather than limiting -/- peak VACV titers. However, the tissue loss exhibited by IFNαR mice makes it difficult to accurately evaluate VACV titer at late times post-intradermal infection. In order to assess the -/- role of T1-IFN in clearance of VACV, we infected the ears of IFNαR and wild-type mice with VACV by scarification, an infectious methodology that results in a milder pathology. We found that the kinetics of VACV clearance was unaffected in the absence of T1-IFN signaling (Fig 4D), so the protective effect of T1-IFN during dermal VACV infection is independent of its direct antiviral characteristic. T1-IFN induces ccl4 to recruit protective monocytes to the site of infection In addition to the induction of antiviral ISGs, it is now becoming clear that T1-IFN can modu- late the innate and adaptive immune response, independent of any effects upon virus PLOS Pathogens | https://doi.org/10.1371/journal.ppat.1007778 October 11, 2019 7 / 25 Type I interferon-dependent CCL4 expression protects virus-infected tissue Fig 4. T1-IFN does not control VACV replication in the skin. (A) Expression of antiviral ISGs in ear tissue of wild- type mice 5 dpi compared to naïve controls measured using RT-qPCR array. (B) Expression of induced ISGs from (A) -/- in ear tissue of wild-type and IFNαR mice 5 dpi measured using RT-qPCR array. (C) VACV titer in ears of mice 5 days post-intradermal infection with VACV. Dotted line indicates initial inoculum. (D) VACV titer in ears of mice at the indicated day post-infection with VACV by scarification, n = 5 mice per group. https://doi.org/10.1371/journal.ppat.1007778.g004 PLOS Pathogens | https://doi.org/10.1371/journal.ppat.1007778 October 11, 2019 8 / 25 Type I interferon-dependent CCL4 expression protects virus-infected tissue replication [53, 54]. Therefore, we used a second qPCR array analysis to compare the change in expression of 84 cytokines and chemokines in the ear 5 days after dermal VACV infection. We found strong infection-mediated induction of 28 of 84 transcripts, with some that are known to be interferon-responsive, including the CXCR3 ligands cxcl9 and cxcl10, as well as IFN-γ (Fig 5A). To investigate which of these transcripts were dependent upon T1-IFN signal- -/- ing, we harvested mRNA from ear tissue of VACV-infected IFNαR mice and performed a -/- similar qPCR array analysis, comparing to uninfected IFNαR mice (Fig 5B). Somewhat sur- -/- prisingly, we found induction of many of the same genes in IFNαR mice upon VACV infec- tion that we observed after infection of wild-type mice. Although VACV strongly induced -/- many transcripts in both wild-type and IFNαR mice, we found that the expression of only one upregulated transcript was differentially modulated >2-fold in a statistically significant manner in the absence of T1-IFN signaling. The IFNαR-dependent gene was ccl4 (macrophage inflammatory protein-1β) (Fig 5C). We were able to confirm that expression of ccl4 in the ears of VACV-infected mice was dependent on both IFNαR signaling and cGAS expression using a targeted qPCR assay (Fig 5D). The cellular response to dermal VACV infection is detected at 3 days post-infection when inflammatory monocytes move into the ear, and numbers of these cells peak at 5 days post- infection [52], i.e. with similar kinetics to T1-IFN induction (Fig 1H). Recruitment of classical inflammatory monocytes occurs with similar kinetics to the induction of T1-IFN transcripts, and is followed by recruitment of a non-classical tissue-protective Ly6G cell population that primarily occurs after the peak of T1-IFN production [52]. Both classical inflammatory mono- cytes and the non-classical tissue-protective Ly6G cell population home to foci of VACV infection, although only classical inflammatory monocytes are infected within the lesion [19]. Local depletion of either inflammatory monocytes [55] or the tissue-protective Ly6G cells -/- -/- [52] produces tissue pathology similar to that observed in cGAS or IFNαR mice. However, depletion of T cells does not affect local pathology or systemic spread of virus following dermal VACV infection [52, 55]. Because the recruitment of classical inflammatory monocytes occurs concurrently with the induction of T1-IFN gene transcripts (Fig 1H), and because the onset of cell recruitment into the ear coincides with the onset of increased tissue pathology in mice lacking T1-IFN signaling (Fig 1C–1F), we sought to determine if T1-IFN-induced ccl4 plays a role in the recruitment of classical inflammatory monocytes, as well as the later recruitment of the Ly6G cell population. We assayed for the presence of chemokine receptors on the cell surface of inflammatory monocytes and tissue-protective Ly6G cells that traffic into VACV-infected ears. From our array data, we found a statistically significant reduction in the expression of ccl4, a large, but non-statistically significant change in expression of cxcl13 as well as a slight, significant change in the expression of cxcl10 (Fig 5C). Using flow cytometry, we assayed for cell-surface expres- sion of the receptors for those respective chemokines, as well as for the prototypic monocyte chemokine receptor, CCR2, the receptor for CCL2 that was upregulated upon VACV infection -/- + of both wild-type and IFNαR mice (Fig 5A and 5B). We found that the majority of CD11b hi - + LyC Ly6G CD64 inflammatory monocytes present in the ear of VACV-infected mice at 5 + + days post-infection were CCR5 , the receptor for CCL4, rather than CCR2 , or positive for the receptors for CXCL13 (CXCR5) or CXCL9 and CXCL10 (CXCR3) (Fig 5E). We next examined whether the deficiency in ccl4 expression by the absence of T1-IFN pre- vented recruitment of CCR5 inflammatory monocytes to the VACV-infected ear. We assayed -/- -/- myeloid cell infiltration in cGAS and IFNαR mice and found a defect in infiltration of inflammatory monocytes, but not Ly6G cells, into the infected ear compared to wild-type mice (Fig 5F). We observed a similar blockade in infiltration of inflammatory monocytes when we treated mice with the CCR5 antagonist, maraviroc (Fig 5G). A number of PLOS Pathogens | https://doi.org/10.1371/journal.ppat.1007778 October 11, 2019 9 / 25 Type I interferon-dependent CCL4 expression protects virus-infected tissue PLOS Pathogens | https://doi.org/10.1371/journal.ppat.1007778 October 11, 2019 10 / 25 Type I interferon-dependent CCL4 expression protects virus-infected tissue Fig 5. T1-IFN induces ccl4 to recruit protective monocytes to the site of infection. (A, B, C) Expression of cytokines and chemokines in ear tissue of mice 5 dpi measured using RT-qPCR array. Volcano plots demonstrate statistically significant points if above the p = 0.05 line. Gene expression changes in wild-type mice in response to VACV, with genes that are upregulated >2-fold by VACV infection in red in the upper right hand quadrant and genes that are downregulated > 2-fold by VACV infection in blue in the upper left hand quadrant (A). Gene -/- expression changes in IFNαR mice in response to VACV, with genes that are upregulated >2-fold by VACV infection in red in the upper right hand quadrant and genes that are downregulated > 2-fold by VACV infection in blue in the upper left hand quadrant (B). Gene -/- expression changes in VACV-infected wild-type mice compared to VACV-infected IFNαR mice with genes that are upregulated >2-fold in -/- VACV-infected IFNαR mice relative to wild-type VACV infected in red in the upper right hand quadrant and genes that are -/- downregulated > 2-fold in VACV-infected IFNαR mice relative to wild-type VACV infected in blue in the upper left hand quadrant (C). -/- Ccl4 (in green) is the only gene that is upregulated upon VACV infection but is downregulated in IFNαR mice relative to wild-type mice. n = 6 naïve, 9 VACV-infected mice per group. (D) Expression of ccl4 in ear tissue of VACV-infected mice 5 dpi measured by RT-qPCR. (E) Chemokine receptor expression on inflammatory monocytes and Ly6G cells recruited into VACV-infected ears 5 dpi. Representative flow cytometry plots (left) and proportion of cells expressing each receptor (right). (F) Myeloid cell populations in VACV-infected ears 5 dpi measured by flow cytometry. Representative flow cytometry plots (left) and absolute number of cells (right). (G) Recruitment of inflammatory monocytes to the ear of uninfected, VACV-infected mice 5 dpi, or VACV-infected mice treated systemically with the CCR5 antagonist Maraviroc, 5dpi. (H) Immunofluorescence microscopy of ears 5 dpi with VACV-NP-S-eGFP reveal monocyte localization (Ly6C-Red) in relation to VACV-infected cells (GFP-Green). (I) Percentage of GFP cells in the ear 5dpi with VACV Early-GFP or VACV Late-GFP measured by flow cytometry. (J) Percentage of GFP cells in the ear 5dpi with VACV-NP-S-eGFP measured by flow cytometry. https://doi.org/10.1371/journal.ppat.1007778.g005 inflammatory monocytes still moved into the infected ear, however, Ly6C inflammatory monocytes failed to localize to the VACV lesion in the absence of T1-IFN induction or signal- ing (Fig 5H). Therefore, cGAS-mediated induction of T1-IFN during dermal VACV infection is required for ccl4 expression and CCR5-dependent recruitment of inflammatory monocytes to viral lesions. Recruited inflammatory monocytes moderate pathology following VACV infection in a manner independent of control of virus replication and expression of antiviral ISGs. However, the mechanisms used by monocytes/macrophages to mediate protective antiviral responses remain poorly described at present. Following infection with some viruses, macrophage popu- lations are preferentially infected in order to facilitate initiation of an adaptive immune response [56–58] or otherwise restrict virus spread [55, 59, 60]. However, primary macrophage populations do not typically support VACV replication, restricting gene expression at a point prior to DNA replication [61]. Therefore, we sought to explore whether infected inflammatory monocytes restricted VACV replication in vivo using a similar mechanism. We infected wild- type mice with VACV that expresses GFP driven by either the early/late promoter p7.5 (Early- GFP) or the late promoter p11 (Late-GFP), which is active only after DNA replication, and assayed GFP expression in inflammatory monocytes in the ear 5 days post-infection. In CD45 cell populations there were equivalent numbers of cells expressing early and late promoter- driven GFP, indicating that there was no blockade in VACV replication within these cells (Fig 5I, left panel). However, a greater proportion of inflammatory monocytes expressed early pro- moter-driven GFP than GFP driven by the late p11 promoter (Fig 5I, right panel), indicative of a blockade of VACV replication in the inflammatory monocyte population. Therefore, inflammatory monocytes within the VACV lesion can potentially act as a “virus sponge,” becoming infected to spare nearby cells from infection. To examine whether this occurred, we -/- infected wild-type or IFNαR mice with VACV that expresses GFP that accumulates in the -/- nucleus of infected cells (VACV-NP-S-eGFP). IFNαR mice had a greater proportion of VACV-infected cells in the ear 5 days post-infection compared to wild-type mice (Fig 5J). Thus, inflammatory monocytes enter the VACV lesion, become non-productively infected, and prevent the infection of neighboring cells. These data support a role for T1-IFN-mediated restriction of virus spread and subsequent tissue damage via a novel immunomodulatory mechanism that does not require a direct effect upon virus replication via the expression of antiviral ISGs. PLOS Pathogens | https://doi.org/10.1371/journal.ppat.1007778 October 11, 2019 11 / 25 Type I interferon-dependent CCL4 expression protects virus-infected tissue ccl4 expression restores inflammatory monocyte recruitment and ameliorates tissue damage To examine the importance of T1-IFN-mediated ccl4 expression in the control of VACV- mediated tissue damage in the dermis, we constructed a recombinant VACV that expresses a construct driven by the relatively weak C6 promoter containing murine ccl4 and the red fluo- rescent protein mKate2 (VACV-CCL4) (Fig 6A). To ensure that any in vivo phenotype we observed was not due to differences in VACV replication we examined single step and multi- step growth curves for VACV-CCL4 vs wild-type VACV (Fig 6B and 6C). We found no differ- ence in growth of the recombinant VACV (Fig 6B and 6C). Upon infection of cells in vitro, the rVACV produced mKate (Fig 6D), as well as ccl4 mRNA (Fig 6E) and protein (Fig 6F). To examine whether VACV-driven expression of ccl4 could rescue recruitment of inflammatory -/- monocytes, we infected IFNαR mice with VACV-CCL4 or a control VACV that did not + hi - express ccl4. Five days after infection we assayed the recruitment of CD11b LyC Ly6G + + CD64 inflammatory monocytes or tissue protective Ly6G myeloid cells in the infected ear. VACV-expression of CCL4 increased the number and proportion of inflammatory monocytes in the ears of VACV-infected mice in the absence of T1-IFN signaling (Fig 6G) without alter- ing recruitment of tissue-protective Ly6G myeloid cells (Fig 6H). Further, and more impor- tantly, restoration of ccl4 expression was able to rescue localization of Ly6C inflammatory -/- monocytes to foci of virus infection in IFNαR mice (Fig 6I). To examine whether CCL4-me- diated inflammatory monocyte recruitment ameliorated the tissue damage we observed in -/- mice lacking IFNαR signaling, we infected IFNαR mice with VACV-CCL4 or a control VACV that did not express ccl4 and monitored multiple aspects of tissue damage as outlined above. We found a statistically significant reduction in ear swelling, lesion size and the com- bined criteria of total tissue damage (Fig 6J–6M). These data strongly support a requisite role for CCL4-mediated recruitment of inflammatory monocytes in moderating tissue damage, but not virus replication, following dermal VACV infection. Discussion The importance of T1-IFN in control of virus replication has led to the evolution of a myriad of evasion strategies employed by viruses across multiple families, aimed at allowing virus rep- lication and propagation to a new host. However, using a virus system in which multiple stages of T1-IFN induction and downstream signaling are inhibited, our data clearly indicate that, despite a lack of induction of T1-IFN within infected cells in vitro, T1-IFN is induced and acts to moderate tissue pathology following virus infection in vivo. Our data support a role for mul- tiple host mechanisms that allow control of virus-dependent pathology but are independent of control over virus replication. Importantly, these mechanisms occur in the face of multiple, effective evasion strategies deployed by the viral pathogen. Induction of T1-IFN following VACV infection in vivo occurred via the cytosolic DNA sen- sor, cGAS and its adapter protein STING. VACV completely blocks cGAS and STING-depen- dent induction of T1-IFN using multiple mechanisms in vitro, including expression of B2, F17, C4, C16 and C6 [15, 16, 17, 18]. However, production of both T1-IFN mRNA and pro- tein, as well as protection against enhanced tissue pathology following VACV infection, was dependent upon cGAS, but not upon other PRR signaling pathways. This contrasts with other studies, which have demonstrated that cGAS is required to protect against VACV challenge, but which have stopped short of demonstrating a role for cGAS in T1-IFN production follow- ing VACV infection [18, 39, 40]. The lack of a link between cGAS and T1-IFN production fol- lowing VACV infection in vivo likely stems from technical issues, and we developed a number of highly sensitive ex vivo approaches in order to be able to detect T1-IFN transcripts and PLOS Pathogens | https://doi.org/10.1371/journal.ppat.1007778 October 11, 2019 12 / 25 Type I interferon-dependent CCL4 expression protects virus-infected tissue Fig 6. ccl4 expression restores inflammatory monocyte recruitment and ameliorates tissue damage. (A) Schematic of rVACV expressing ccl4 and mKate2. (B) Single step growth curve of wild-type VACV and VACV-CCL4 in 143B cells. (C) Multi-step growth curve of wild-type VACV and VACV-CCL4 in 143B cells. (D) Detection of mKate2 expression in 143B cells 6 hpi assayed using flow cytometry. (E) Expression of murine ccl4 in 143B cells 6 hpi measured by RT-qPCR. (F) Secretion of murine -/- CCL4 from 143B cells 24 hpi measured using ELISA. (G) Monocyte recruitment into VACV-infected ears of IFNαR mice 5 dpi assayed by flow cytometry, n = 4 mice + -/- per group. (H) Ly6G cell recruitment into VACV-infected ears of IFNαR mice 5 dpi measured by flow cytometry, n = 4 mice per group. (I) Immunofluorescence -/- microscopy of IFNαR ears 5 dpi with VACV-CCL4 or VACV-mCherry for monocyte localization (Ly6C-Blue) in relation to VACV-infected cells (mKate/mCherry- -/- Red). (J-L) Tissue pathology of the ear pinnae of IFNαR mice after intradermal infection with VACV; ear thickness (J), lesion size (K), or total tissue damage (L), -/- n = 6–7 mice per group. (M) Representative pictures of ear pinnae of IFNαR mice following intradermal VACV infection. https://doi.org/10.1371/journal.ppat.1007778.g006 protein from VACV-infected tissue. Using these approaches, we clearly demonstrate that cGAS signaling is required for T1-IFN induction following VACV infection. We found that the majority of infected cells in vivo mirror our in vitro finding, and do not express T1-IFN mRNA, likely due to the action of multiple VACV immunomodulators. We anticipated that the induction of T1-IFN that we observed in vivo might be the result of the transfer of cGAS- produced cGAMP to neighboring uninfected cells [62, 63]. Once within neighboring cells, PLOS Pathogens | https://doi.org/10.1371/journal.ppat.1007778 October 11, 2019 13 / 25 Type I interferon-dependent CCL4 expression protects virus-infected tissue cGAMP could induce T1-IFN production via STING within cells that are not subject to immu- nomodulatory VACV proteins, allowing for efficient expression. However, all of the IFN-β mRNA we detected was present within a small population of VACV-infected cells that appeared functionally distinct from the majority of infected cells. Therefore, a specialized pop- ulation of cells may allow the host to subvert viral evasion mechanisms and express protective T1-IFN. Therapeutic targeting of such a population during infection or vaccination may pref- erentially allow for the generation of a strong T1-IFN response within a tissue. What would be the characteristic of an infected cell that can subvert the action of the viral B2, F17 and C6 proteins to allow for cGAS-dependent T1-IFN expression? T1-IFN expression begins by day 3 post-infection, a time point when inflammatory monocytes are the first recruited cells to accumulate at the site of VACV infection. However, induction of T1-IFN still occurred in the absence of inflammatory monocyte recruitment (S1 Fig). This was not unex- pected, as T1-IFN was required for effective inflammatory monocyte recruitment. We have previously described T1-IFN production by the tissue protective Ly6G population of recruited cells [52]. However, the Ly6G population does not begin to accumulate until day 5 post-infection and does not peak until day 8–10 post-infection [52] i.e. significantly after the peak of T1-IFN production. Therefore, Ly6G cells may contribute to T1-IFN production after the peak of VACV replication, but they do not play a role in CCL4-mediated recruitment of classical inflammatory monocytes, which precede them into the VACV lesion. Our data are indicative of a role for tissue-resident VACV-infected cells in cGAS-mediated T1-IFN produc- tion. These cells could be specialized populations of keratinocytes, dermal fibroblasts, Langer- hans cells, dermal DCs, tissue resident macrophages,γδ or other tissue-resident T cells, ILCs or even mast cells. Identification of the cells responsible for T1-IFN production following VACV infection will require extensive further studies. The mechanism by which infected cells can produce T1-IFN in a cGAS-dependent manner in vivo in the face of viral evasion strategies that efficiently block this pathway in vitro is partic- ularly interesting. The most effective way to bypass the blockade of cytosolic sensing within infected cells would be to inhibit expression of the viral proteins that target the cytosolic sen- sors. However, on a global level, the breadth of mechanisms employed by viruses to block cyto- solic sensing likely precludes this approach. Indeed, in the system examined here, both the cGAS/STING and MDA5/MAVS pathways are targeted by multiple VACV proteins including B2, F17, K13, C6, E3 and K3 [15–17, 64–66]. Therefore, it is more likely that multiple cGAS/ STING sensing and signaling pathways exist in specialized cells. In the case of the RNA-sens- ing pathway, MAVS signaling can occur from multiple sites within an infected cell, namely from the mitochondria or peroxisomal membranes [67]. These differential signaling sites pro- duce different functional outcomes [67], raising the possibility that they may be differentially sensitive to viral inhibition. This would be the case if viral proteins are targeted to the mito- chondria or ER, respectively, to inhibit the STING and MAVS signaling pathways. The possi- bility that these pathways may be differentially expressed, or perhaps enhanced, within specific cell populations is intriguing. Targeting of these specialized pathways may preferentially allow for the generation of a strong T1-IFN response during vaccination or therapeutic treatments. The majority of T1-IFNs produced during VACV infection were IFN-β and IFN-α4, each of which are produced within the first round of T1-IFN induction, without the need to first induce expression of further ISGs such as irf7. Localization of IFN-β exclusively within VACV-infected cells implies that nearby cells are not triggered to produce either IFN-β or IFN-αs that are downstream from IFNαR in the positive feedback loop of T1-IFN induction. This observation supports two conclusions. First, the soluble T1-IFN binding protein encoded by VACV, B18 [68], appears to be very effective at preventing the spread of T1-IFN induction to nearby cells. Second, the majority of both IFN-β and IFN-α4 induction within infected cells PLOS Pathogens | https://doi.org/10.1371/journal.ppat.1007778 October 11, 2019 14 / 25 Type I interferon-dependent CCL4 expression protects virus-infected tissue is dependent upon IFNαR, STAT1 and STAT2, despite the inhibitory effect of the viral VH1 protein upon STAT1 activation [69]. These findings indicate that some level of positive feed- back signaling through IFNαR in VACV-infected cells is required. B18 binds to the surface of uninfected cells and soaks up excess T1-IFN, preventing activation of these cells through IFNαR [68]. Therefore, B18 may not effectively bind to the surface of infected cells, perhaps due to virus-induced changes in the cell surface architecture and proteome. It was surprising that VACV replication was unaffected by T1-IFN signaling. However, upregulation of antiviral ISGs was independent of T1-IFN in the infected ear, despite the observation that T1-IFN expression occurred only within infected cells. Many of the ISGs induced by T1-IFN signaling are important mediators in restricting the ability of viruses to infect and replicate within cells, although many have not been studied in the context of a VACV infection. However, during infection with VACV, ISG15, which is strongly induced in a T1-IFN-independent manner in infected skin, controls mitochondrial function and limits virus replication within infected cells [70, 71]. The only logical explanation from these oppos- ing observations is that ISG expression is induced by redundant signaling pathways via the expression of other cytokines. It is possible that other families of interferon species may induce the expression of ISGs in VACV-infected skin. In addition, there is significant crosstalk between the NF-κB and T1-IFN signaling pathways [46], so it is possible that other pathways are activated to stimulate ISG induction. NF-κB signaling is also inhibited in VACV-infected cells [29, 72, 73], but our profiling of the cytokine and chemokine response in the infected ear demonstrates strong upregulation of a number of inflammatory cytokines. We were unable to examine the induction of cytokines, chemokines or ISGs within infected vs uninfected cells in situ, but it is likely that upregulation of these effector molecules occurs within uninfected cell populations. Irrespective of the mechanisms by which ISGs are induced, it is nonetheless clear that antiviral ISGs do not play a role in the protection of infected tissue by T1-IFN following VACV infection. Here, the action of T1-IFN is mediated through the CCL4-dependent recruitment of inflammatory monocytes, rather than through the direct inhibition of virus replication via the action of intracellular ISGs, a process that can be subverted by viral proteins within infected cells. In a number of viral infections, including the VACV system used here, tissue-resident macrophage populations in secondary lymphoid organs are required to prevent systemic dis- semination of viruses [55, 58–60, 74]. After infection with some viruses, infection of macro- phage populations is required in order to facilitate development of an effective adaptive immune response [57]. However, the role of recruited inflammatory monocyte populations after infection differs from that of macrophages resident in systemic organs [55]. Although inflammatory monocyte recruitment has been indirectly implicated in the response to virus infection [75–77], the mechanisms employed by recruited inflammatory monocyte popula- tions are unknown. Monocytes do not directly inhibit VACV replication within other infected cell populations, as there is no large change in VACV titer, despite the marked change in tissue pathology in the absence of T1-IFN signaling. Infectious VACV can be recovered from inflam- matory monocytes themselves when they are sorted from VACV-infected ears [19]. However, the titer recovered from inflammatory monocytes was less than 2 pfu per cell, and could repre- sent input levels of virus rather than progeny virions. Here, we show that monocyte recruit- ment to the VACV-infected ear was dependent upon T1-IFN-mediated induction of CCL4, although the requirement was incomplete. However, in the absence of CCL4 induction, recruitment to the VACV lesion was essentially completely ablated. As the majority of inflam- matory monocytes recruited to the VACV lesion become infected, it is likely that infection of these cells may play a role in the tissue-protective phenotype that they mediate. Here, we dem- onstrate that inflammatory monocytes are infected, but the blockade in DNA replication in PLOS Pathogens | https://doi.org/10.1371/journal.ppat.1007778 October 11, 2019 15 / 25 Type I interferon-dependent CCL4 expression protects virus-infected tissue the majority of cells prevents the production of infectious virions. Defective recruitment of -/- monocytes may account for the slight increase in VACV titer we recovered from cGAS and -/- IFNαR mice compared to wild-type mice, as we observed a similar increase in titer in mice in which we have used multiple methodologies to deplete monocytes at the site of VACV infec- tion, but which retain intact T1-IFN induction [55]. This further suggests that a lack of T1-IFN does not render ear tissue more permissive to VACV replication. We propose that inflammatory monocytes at the site of infection can function as a virus sponge, thereby limit- ing the proportion of ear tissue that is infected, and subsequently eliminated by the adaptive -/- immune response. Indeed, surrounding ear tissue from IFNαR mice had a greater propor- tion of VACV-infected cells compared to wild-type mice, but VACV titers were not substan- tially increased, supporting this model. It is also possible that inflammatory monocytes may facilitate a response to “wall-off” the infected tissue and promote local wound healing responses. In either way, inflammatory monocytes can moderate virus-dependent tissue pathology in a T1-IFN-dependent manner that is independent of the induction of classical antiviral ISGs or inhibition of virus replication. Taken as a whole, our findings provide a more complete picture of the way in which the host uses multiple mechanisms to bypass a myriad of viral evasion strategies, and thus still pro- duce and utilize T1-IFN in a protective antiviral capacity. Induction of T1-IFN requires infec- tion of specialized populations of tissue-resident cells that can subvert the viral blockade of cGAS/STING sensing. The infected cell population only produces “early” species of T1-IFN, likely due to the effects of soluble T1-IFN binding proteins and a blockade in STAT1 signaling within infected cells. However, the small amount of T1-IFN made is sufficient to induce a posi- tive feedback loop of T1-IFN within infected cells to stimulate downstream expression of ccl4 and subsequent recruitment of inflammatory monocytes. The inflammatory monocytes become the major infected cell population within the lesion, but are non-productively infected, effectively throwing themselves on the exploding grenade of the infection. Materials and methods Mice Mice were on the C57BL/6 background unless otherwise stated. Wild-type (catalog no. 00664), -/- gt/gt -/- cGAS (catalog no. 026554), STING (catalog no. 017537), MyD88 (catalog no. 009088), Lps/Lps -/- TRIF (catalog no. 005037), MAVS (catalog no. 008634 on the B6.129 background), -/- +/- STAT2 (catalog no. 023309), and STAT1 (catalog no. 012606, subsequently bred to pro- -/- duce homozygous STAT1 ) mice were purchased from the Jackson Laboratory and housed at -/- the Penn State College of Medicine. IFNαR (Jax catalog no. 32045) mice were a gift from Dr. Ziaur Rahman (Penn State College of Medicine). All mice were maintained in the specific- pathogen-free facility at the Penn State College of Medicine. In vivo experiments utilized 6- 12-week-old mice. All animals were treated in accordance with the National Institutes of Health and AAALAC International Regulations. The Penn State College of Medicine Institu- tional Animal Care and Use Committee approved all experiments. Ethics statement All animals were maintained in the specific-pathogen-free facility of the Hershey Medical Cen- ter and treated in accordance with the National Institutes of Health and AAALAC Interna- tional regulations. All animal experiments and procedures were approved in protocols 47731, 47033, 47597, 46111 and 44723 by the Penn State Hershey Institutional Animal Care and Use Committee (Animal Welfare Assurance # A3045-01) that follows the Office of Laboratory Ani- mal Welfare PHS Policy on Humane Care and Use of Laboratory Animals, 2015. PLOS Pathogens | https://doi.org/10.1371/journal.ppat.1007778 October 11, 2019 16 / 25 Type I interferon-dependent CCL4 expression protects virus-infected tissue Viruses and cell lines VACV (strain Western Reserve), or recombinant viruses (see below), were grown and purified as previously described [55]. Heat-inactivation of VACV was achieved by incubating aliquots of virus at 56˚C in a shaking incubator for 1 hour just prior to infection. To easily identify VACV-infected cells in vivo, mice were infected with a recombinant VACV expressing influ- enza nucleoprotein tagged with eGFP, in which fluorescent protein accumulates in the nucleus of infected cells (VACV-NP-S-eGFP) [74]. In order to identify virus-encoded RNA, we used VACV that expresses the ovalbumin gene from Gallus gallus (VACV-OVA) [78]. VSV was a kind gift from the late Dr. Leo Lefrancois and was grown and titered on BHK cells using stan- dard protocols. To identify infected cells in which abortive or productive infection was proceeding, we uti- lized VACV-eGFP-OVA that expresses eGFP coupled with OVA under the control of either the early/late p7.5 promoter (VACV Early-GFP) or the late promoter p11 (VACV Late-GFP). VACV-eGFP-OVA early and late viruses were prepared as follows. The plasmid pRB21 expressing the full-length vp37 VACV ORF with the p7.5 early/late promoter was a kind gift from Dr. Bernard Moss (Laboratory of Viral Diseases, MIAID, Bethesda, MD). The peGFP-C1 plasmid expressing OVA was a kind gift from Dr. Kenneth Rock (Department of Pathology, University of Massachusetts Medical School, Worcester, MA). pRB21 DNA was ligated with eGFP-OVA using T4 Ligase (Invitrogen). To make rVACV-eGFP-OVA with a late promoter, the p11 promoter was inserted in place of p7.5. All plasmid DNA was sequenced to ensure that the promoter, and eGFP-OVA sequences were correct. rVACV-eGFP-OVA-Early and rVACV-eGFP-OVA-Late were generated by transfecting plasmid DNA into BSC-1 cells infected with VACV-vRB21 at an MOI of 1 using the CellPhect Transfection Kit (GE Health- care). The resulting rVACVs were plaque-purified 3x prior to characterization. The resulting rVACV-eGFP-OVA-Early and rVACV-eGFP-OVA-Late produced green fluorescence upon infection of WT3 cells and sequencing revealed the presence of the correct promoter and OVA sequences in DNA purified from virions. To make a virus that expresses CCL4, a construct was ordered that contained 357 nucleo- tides of the 3’ end of the J4 open reading frame (ORF) followed by the early VACV promoter C6, the coding sequence for mKate2 and mouse CCL4 ORFs connected by an in-frame sequence for the P2A peptide. The CCL4 ORF was followed by a 3x VACV early transcription stop signal and 400 nucleotides of the 3’ end of the J5 ORF (IDT). These constructs were amplified by standard PCR and cloned into pCR2.1 and sequenced to ensure fidelity. Recom- binant viruses were generated using homologous recombination by infecting HeLa cells with VACV WR strain and transfecting in the constructs (VACV-CCL4). Recombinant viruses were isolated based on their ability to produce red fluorescent plaques. The isolated recombi- nants were plaque-purified 4x before being amplified to produce working stocks. The integrity of the inserts was verified by sequencing. For microscopy experiments utilizing VACV-CCL4, we used VACV mCherry Ub-SIINFEKL (VACV-mCherry) [79] as the control for expression of a red fluorescent protein without CCL4. The murine keratinocyte cell line 308 was originally derived from BALB/c mouse skin initi- ated with DMBA [80]. NR9456 is a murine macrophage cell line derived from differentiated bone marrow-derived cells immortalized by retroviral transformation [81] and was obtained through BEI Resources, NIAID, NIH. 143B cells are a thymidine kinase-deficient human oste- osarcoma cell line. 143B and NR9456 cells were grown at 37˚C in DMEM supplemented with 10% FBS, 20mM HEPES, 1.5 g/L NaHCO , 0.3 g/L L-glutamine, and 0.1 g/L penicillin and streptomycin. 308 cells were grown at 37˚C in DMEM containing 4.5 g/L glucose, L-glutamine PLOS Pathogens | https://doi.org/10.1371/journal.ppat.1007778 October 11, 2019 17 / 25 Type I interferon-dependent CCL4 expression protects virus-infected tissue and sodium pyruvate supplemented with 10% FBS, 0.1 g/L penicillin and streptomycin and an additional 0.3 g/L L-glutamine. VACV infection For intradermal VACV infection, 10 pfu VACV in a 10μl volume of Hank’s Balanced Salt Solution (HBSS)/0.1% bovine serum albumin (BSA) was injected into the middle of each ear pinnae [21]. Intradermal infection causes substantial tissue loss at late times post-infection, preventing accurate measurement of titers at those times. Therefore, to assess virus titer at late times post-infection, mice were infected by scarification of the ear pinnae. Briefly, a drop of HBSS/0.1%BSA containing 10 pfu VACV was placed on the center of the ear pinnae and a 28-gauge needle was used to scratch the ear through the droplet [82]. To produce a localized infection suitable for microscopy imaging, mice were infected by dipping a bifurcated needle into HBSS/0.1%BSA containing 2x10 pfu/ml VACV and then gently puncturing each ear pin- nae ten times [19]. To evaluate tissue pathology during VACV infection, ear thickness was measured daily using a 0.0001m dial micrometer (Mitutoyo). After the appearance of a visible lesion, the diameter of the lesion or of areas of tissue loss were measured [52]. 5 6 For infection of cells in vitro, 5e -1e cells were seeded onto 24-well plates 24 hours prior to infection. Cells were infected at an MOI of 10 in 250μl HBSS/0.1%BSA for 1 hour at 37˚C, inoculum was removed and replaced with 500μl of pre-warmed media and placed at 37˚C for either an additional 5 or 23 hours. To assess the abilities of wild-type or VACV-CCL4 to repli- cate in vitro, we performed single-step and multistep growth curves. Confluent monolayers of 143B cells were infected with 5 PFU/cell of the indicated virus diluted for single-step growth curves or with 0.05 PFU/cell for multistep growth curves. Two hours post-infection, the inocu- lum was replaced with fresh medium. At the indicated time points, cells were washed, har- vested and then freeze thawed three times and sonicated to release virus, and titers were determined as described below. VACV titer Virus replication was quantified by plaque assay as previously described [52, 55]. Infected ears were harvested at the times indicated and homogenized using a TissueLyser II (Qiagen), then centrifuged for 2 minutes at 12,000 x g. Supernatant was transferred into a fresh tube and virus titer was determined by plaque assay on 143B cells. RNA isolation and RT-qPCR For gene expression in vitro, cells were infected as described above. After 6 hours total infec- tion time, media was removed and cells were lysed with 500μl TRIzol reagent (Invitrogen), transferred to a 1.7ml tube, and placed at -20˚C until RNA extraction. For RNA isolation, 100μl chloroform was added to thawed tubes, briefly vortexed and centrifuged at 12,000 x g for 15 minutes at 4˚C. The upper aqueous phase was transferred into a clean tube, 1 volume of 70% EtOH was added, and RNA was purified using the RNeasy mini kit (Qiagen) with on-col- umn DNase-I treatment. For gene expression in vivo, ears were harvested from mice at the times indicated post- infection, flash-frozen in liquid NO and placed at -80˚C. For RNA isolation, ears were homogenized using a TissueLyser II (Qiagen) and RNA was purified from tissue lysate using the RNeasy Mini Kit (Qiagen) treated with DNase-I (Qiagen) in-solution. For qPCR using Taqman Gene Expression Assays (Applied Biosystems) or Universal Probe Library (Roche), cDNA was prepared using the Hi-Capacity cDNA Synthesis Kit (Applied Bio- 2 2 systems). For qPCR using RT PCR Profiler Arrays (Qiagen), cDNA was prepared using RT PLOS Pathogens | https://doi.org/10.1371/journal.ppat.1007778 October 11, 2019 18 / 25 Type I interferon-dependent CCL4 expression protects virus-infected tissue First Strand Kit (Qiagen). qPCR was carried out on a StepOnePlus (Applied Biosystems) with either FastStart Universal Probe Master Mix (Roche) or RT SYBR Green qPCR Master Mix (Qiagen). For Taqman and FastStart Universal Probe assays, changes in gene expression are Ct expressed as fold change using theΔΔ calculation method against naïve mice of the same genotype or mock-infected cells with gapdh as the housekeeping gene. For RT PCR Profiler Array data, changes in gene expression are displayed as mean fold change between groups of mice relative to a panel of “housekeeping” genes. Amplification of bulk IFN-α transcripts was detected using Universal Probe Library probe #76 (Roche) with forward primer ARSYtgtSt- gatgcaRcaggt and reverse primer ggWacacagtgatcctgtgg; and was compared to amplification of gapdh using Universal Probe Library probe #80 (Roche) with forward primer tgtccgtcgtggatct- gac and reverse primer cctgcttcaccaccttcttg. The oligos for IFNα (non-4) amplify sequences present within most of the genes within the IFNα subfamily except Ifna4. These primers are accepted to represent expression of the IRF7-mediated subset of IFNα genes. Primers were from Integrated DNA Technologies. ELISA For detection of protein from cells infected in vitro, cells were infected as described above. After the initial 1 hour of infection, the inoculum was removed and replaced with 500μl of pre- warmed media supplemented with 2% FBS and cells were incubated at 37˚C for an additional 23 hours. After 24 hours of total infection time, media was transferred into tubes containing phenylmethylsulfonyl fluoride (PMSF) (Sigma) and a Protease Inhibitor Cocktail (Sigma) and placed at -80˚C until the ELISA was performed. For detection of protein from VACV-infected ear tissue, ears were infected and harvested as described above. Frozen ears were homogenized in PBS containing PMSF (Sigma) and Pro- tease Inhibitor Cocktail (Sigma) using a TissueLyser II (Qiagen), centrifuged for 3 min at 12,000 x g and the lysate was transferred to a clean tube. IFN-β protein was detected using the mouse Hi-Sensitivity IFN-β ELISA kit (PBL Assay Science). CCL4 protein was detected using the mouse CCL4/MIP-1 beta Quantikine ELISA kit (R&D Systems). ELISAs were carried out according to manufacturer’s specifications. Flow cytometry For the detection of the mKate2 transgene, 2e 143B cells were infected in 5ml tubes at an MOI of 10 in 500μl HBSS/0.1%BSA, incubated in a 37˚C water bath for 1 hour, 3 ml of pre- warmed media was added, and tubes were incubated in a rotating incubator at 37˚C for an additional 5 hours. Cells were washed and samples were acquired on a LSRII flow cytometer (BD Biosciences), data were analyzed using FlowJo software (TreeStar). Ears were harvested and processed as previously described [52, 55]. Briefly, ears were sepa- rated into dorsal and ventral halves and digested in 1 mg/ml collagenase XI (Sigma) for 60 minutes at 37˚C and a single cell suspension made. Live cells were incubated in Fc block and 10% normal mouse serum (Gemini Bio-Products) at 4˚C, and then stained in that solution. Cells were stained with antibodies recognizing Ly6C (clone HK1.4), CD11b (clone M1/70), Ly6G (clone 1A8), CD45 (clone 30-F11), CD64 (clone X54-5/7.1), CD11c (clone N418), CCR5 (clone HM-CCR5), CXCR3 (clone CXCR3-173), CXCR5 (clone L138D7), Streptavidin from BioLegend, CD19 (clone 1D3), CD90.2 (clone 53–2.1), and NK1.1 (clone PK136) from BD Biosciences, and CCR2 (clone REA538) from Miltenyi Biotec. Samples were acquired on a LSRII flow cytometer (BD Biosciences), and data were analyzed using FlowJo software (TreeS- + - + hi tar). Inflammatory monocytes were gated as CD45 (CD19, CD90.2, NK1.1) CD11b Ly6C PLOS Pathogens | https://doi.org/10.1371/journal.ppat.1007778 October 11, 2019 19 / 25 Type I interferon-dependent CCL4 expression protects virus-infected tissue - - + + Ly6G CD11c CD64 [55]. Non-classical tissue-protective monocytes were gated as CD45 - hi + + (CD19, CD90.2, NK1.1) CD11b Ly6C Ly6G [52]. Fluorescent imaging Mice were infected with VACV in the center of each ear pinnae using a bifurcated needle. Five days post-infection, ears were harvested and embedded in Tissue-Tek OCT (Sakura Finetek), rapidly frozen by immersion in liquid nitrogen-cooled 2-methyl butane, and stored at -80˚C. Lateral cryostat sections (10 μm) were cut at -20˚C, mounted on glass slides and fixed for 30 min in 4% paraformaldehyde in PBS (pH 7.4). For both assays, three-dimensional images were collected on an Olympus IX81 fluorescent microscope, deconvolved and analyzed using Slide- book 6.0 software (Intelligent Imaging Innovations). For RNA fluorescence in situ hybridization, mice were infected with VACV-OVA. To visu- alize plasma membranes, tissue was stained with Wheat Germ Agglutinin conjugated to CF488 (1:1000, Biotium). Tissue was fixed again and RNA was visualized using the ViewRNA Cell Plus Assay (Invitrogen) with probes recognizing murine IFN-β and gallus ovalbumin (Affymetrix) according to manufacturer’s specifications with the omission of using a protease. Nuclei were stained with DAPI. For Immunofluorescence microscopy, mice were infected with VACV-NP-S-eGFP, VACV-CCL4, or VACV-mCherry. Tissue was stained with antibodies to Ly6C (clone HK1.4 1:2500, Abcam) and visualized using secondary donkey anti-rat Alexa 647 (1:500, Jackson Immunoresearch) or donkey anti-rat BV421 (1:100, Jackson Immunoresearch). CCR5 antagonism Maraviroc (Cayman Chemical) is a noncompetitive, slowly reversible small-molecule antago- nist of CCR5 that prevents signaling by all three ligands of CCR5 [i.e.,CCL5/RANTES, macro- phage inflammatory protein 1α (MIP-1α), andMIP-1β] and is used as a part of combination antiretroviral therapy for patients infected with CCR5-tropic HIV (i.e., a subtype that uses onlythe CCR5 co-receptor to enter cells) [83]. We treated with maraviroc (10 mg/kg, i.p.) for 4 days, starting the day of infection, until harvesting VACV infected ears 5 dpi to assay mono- cyte recruitment. Data analysis Data were analyzed and graphed using Excel (Microsoft), Prism (Graphpad), and RT Profiler Array web-based software (Qiagen). Data are expressed as mean ± standard error of the mean, with individual data points shown where applicable. RT PCR Profiler Array data are expressed as mean fold change vs. p-value. Means were compared using an unpaired, two-way student’s t-test or two-way ANOVA as applicable. Data represents at least three individual experiments. Significance between groups was determined by p-value <0.05, and is displayed � � � � � � � � � � as; = p<0.05, = p<0.01, = p<0.001, = p<0.0001. Supporting information S1 Fig. (A) Expression of T1-IFN subtypes IFN-β1 andα4 in ear tissue of wild-type (B6) vs -/- CCR2 mice, in which recruitment of inflammatory monocytes to the VACV infected ear is completely ablated (55). mRNA was purified and expression levels from uninfected mice, or at 5 dpi, measured by RT-qPCR. (TIFF) PLOS Pathogens | https://doi.org/10.1371/journal.ppat.1007778 October 11, 2019 20 / 25 Type I interferon-dependent CCL4 expression protects virus-infected tissue Author Contributions Conceptualization: Nikhil J. Parekh, Christopher C. Norbury. Data curation: Nikhil J. Parekh, Tracy E. Krouse, Irene E. Reider, Christopher C. Norbury. Formal analysis: Nikhil J. Parekh, Tracy E. Krouse, Irene E. Reider, Christopher C. Norbury. Funding acquisition: Christopher C. Norbury. Investigation: Nikhil J. Parekh, Tracy E. Krouse, Irene E. Reider, Christopher C. Norbury. Methodology: Nikhil J. Parekh, Tracy E. Krouse, Irene E. Reider, Brian M. Ward, Christopher C. Norbury. Project administration: Christopher C. Norbury. Resources: Ryan P. Hobbs, Brian M. Ward, Christopher C. Norbury. Software: Christopher C. Norbury. Validation: Christopher C. Norbury. Visualization: Christopher C. Norbury. Writing – original draft: Nikhil J. Parekh, Christopher C. Norbury. Writing – review & editing: Nikhil J. Parekh, Ryan P. Hobbs, Brian M. Ward, Christopher C. Norbury. References 1. van den Broek MF, Muller U, Huang S, Zinkernagel RM, Aguet M. Immune defence in mice lacking type I and/or type II interferon receptors. Immunol Rev. 1995; 148:5–18. https://doi.org/10.1111/j.1600-065x. 1995.tb00090.x PMID: 8825279. 2. Isaacs A, Lindenmann J. Virus interference. I. The interferon. Proc R Soc Lond B Biol Sci. 1957; 147 (927):258–67. PMID: 13465720. 3. Nagano Y, Kojima Y, Sawai Y. [Immunity and interference in vaccinia; inhibition of skin infection by inac- tivated virus]. C R Seances Soc Biol Fil. 1954; 148(7–8):750–2. PMID: 13190824. 4. Pestka S, Krause CD, Walter MR. Interferons, interferon-like cytokines, and their receptors. Immunol Rev. 2004; 202:8–32. https://doi.org/10.1111/j.0105-2896.2004.00204.x PMID: 15546383. 5. Yan N, Chen ZJ. Intrinsic antiviral immunity. Nat Immunol. 2012; 13(3):214–22. https://doi.org/10.1038/ ni.2229 PMID: 22344284; PubMed Central PMCID: PMC3549670. 6. Fu XY, Schindler C, Improta T, Aebersold R, Darnell JE Jr., The proteins of ISGF-3, the interferon alpha-induced transcriptional activator, define a gene family involved in signal transduction. Proc Natl Acad Sci U S A. 1992; 89(16):7840–3. https://doi.org/10.1073/pnas.89.16.7840 PMID: 1502204; PubMed Central PMCID: PMC49807. 7. Schoggins JW, Rice CM. Interferon-stimulated genes and their antiviral effector functions. Curr Opin Virol. 2011; 1(6):519–25. https://doi.org/10.1016/j.coviro.2011.10.008 PMID: 22328912; PubMed Cen- tral PMCID: PMC3274382. 8. Schneider WM, Chevillotte MD, Rice CM. Interferon-stimulated genes: a complex web of host defenses. Annu Rev Immunol. 2014; 32:513–45. https://doi.org/10.1146/annurev-immunol-032713-120231 PMID: 24555472; PubMed Central PMCID: PMC4313732. 9. Sen GC. Viruses and interferons. Annu Rev Microbiol. 2001; 55:255–81. https://doi.org/10.1146/ annurev.micro.55.1.255 PMID: 11544356. 10. Hoffmann HH, Schneider WM, Rice CM. Interferons and viruses: an evolutionary arms race of molecu- lar interactions. Trends Immunol. 2015; 36(3):124–38. https://doi.org/10.1016/j.it.2015.01.004 PMID: 25704559; PubMed Central PMCID: PMC4384471. 11. Garcia-Sastre A. Ten Strategies of Interferon Evasion by Viruses. Cell Host Microbe. 2017; 22(2):176– 84. https://doi.org/10.1016/j.chom.2017.07.012 PubMed Central PMCID: PMC5576560. PMID: PLOS Pathogens | https://doi.org/10.1371/journal.ppat.1007778 October 11, 2019 21 / 25 Type I interferon-dependent CCL4 expression protects virus-infected tissue 12. Alcami A, Koszinowski UH. Viral mechanisms of immune evasion. 2000; 8(9):410–8. https://doi.org/10. 1016/s0966-842x(00)01830-8 PMID: 10989308 13. Cao H, Dai P, Wang W, Li H, Yuan J, Wang F, et al. Innate immune response of human plasmacytoid dendritic cells to poxvirus infection is subverted by vaccinia E3 via its Z-DNA/RNA binding domain. PLoS One. 2012; 7(5):e36823. https://doi.org/10.1371/journal.pone.0036823 PMID: 22606294; PubMed Central PMCID: PMC3351467. 14. Dai P, Wang W, Cao H, Avogadri F, Dai L, Drexler I, et al. Modified vaccinia virus Ankara triggers type I IFN production in murine conventional dendritic cells via a cGAS/STING-mediated cytosolic DNA-sens- ing pathway. PLoS Pathog. 2014; 10(4):e1003989. https://doi.org/10.1371/journal.ppat.1003989 PMID: 24743339; PubMed Central PMCID: PMC3990710. 15. Meade N, Furey C, Li H, Verma R, Chai Q, Rollins MG, et al. Poxviruses Evade Cytosolic Sensing through Disruption of an mTORC1-mTORC2 Regulatory Circuit. Cell. 2018; 174(5):1143–57 e17. https://doi.org/10.1016/j.cell.2018.06.053 PMID: 30078703; PubMed Central PMCID: PMC6172959. 16. Georgana I, Sumner RP, Towers GJ, Maluquer de Motes C. Virulent poxviruses inhibit DNA sensing by preventing STING activation. J Virol. 2018. https://doi.org/10.1128/JVI.02145-17 PMID: 29491158; PubMed Central PMCID: PMC5923072. 17. Scutts SR, Ember SW, Ren H, Ye C, Lovejoy CA, Mazzon M, et al. DNA-PK Is Targeted by Multiple Vaccinia Virus Proteins to Inhibit DNA Sensing. Cell Rep. 2018; 25(7):1953–65 e4. https://doi.org/10. 1016/j.celrep.2018.10.034 PMID: 30428360. 18. Eaglesham JB, Pan Y, Kupper TS, Kranzusch PJ. Viral and metazoan poxins are cGAMP-specific nucleases that restrict cGAS-STING signalling. Nature. 2019; 566(7743):259–63. https://doi.org/10. 1038/s41586-019-0928-6 PMID: 30728498. 19. Hickman HD, Reynoso GV, Ngudiankama BF, Rubin EJ, Magadan JG, Cush SS, et al. Anatomically restricted synergistic antiviral activities of innate and adaptive immune cells in the skin. Cell host & microbe. 2013; 13(2):155–68. Epub 2013/02/19. https://doi.org/10.1016/j.chom.2013.01.004 PMID: 23414756; PubMed Central PMCID: PMC3591514. 20. Deng L, Dai P, Parikh T, Cao H, Bhoj V, Sun Q, et al. Vaccinia virus subverts a mitochondrial antiviral signaling protein-dependent innate immune response in keratinocytes through its double-stranded RNA binding protein, E3. J Virol. 2008; 82(21):10735–46. https://doi.org/10.1128/JVI.01305-08 PMID: 18715932; PubMed Central PMCID: PMC2573174. 21. Tscharke DC, Smith GL. A model for vaccinia virus pathogenesis and immunity based on intradermal injection of mouse ear pinnae. J Gen Virol. 1999; 80 (Pt 10):2751–5. https://doi.org/10.1099/0022- 1317-80-10-2751 PMID: 10573171. 22. Tscharke DC, Reading PC, Smith GL. Dermal infection with vaccinia virus reveals roles for virus pro- teins not seen using other inoculation routes. J Gen Virol. 2002; 83(Pt 8):1977–86. https://doi.org/10. 1099/0022-1317-83-8-1977 PMID: 12124461. 23. Jaks E, Gavutis M, Uze G, Martal J, Piehler J. Differential receptor subunit affinities of type I interferons govern differential signal activation. J Mol Biol. 2007; 366(2):525–39. Epub 2006/12/19. https://doi.org/ 10.1016/j.jmb.2006.11.053 PMID: 17174979. 24. Li L, Sherry B. IFN-alpha expression and antiviral effects are subtype and cell type specific in the car- diac response to viral infection. Virology. 2010; 396(1):59–68. Epub 2009/11/10. https://doi.org/10. 1016/j.virol.2009.10.013 PMID: 19896686; PubMed Central PMCID: PMC2787694. 25. Marie I, Durbin JE, Levy DE. Differential viral induction of distinct interferon-alpha genes by positive feedback through interferon regulatory factor-7. EMBO J. 1998; 17(22):6660–9. https://doi.org/10.1093/ emboj/17.22.6660 PMID: 9822609; PubMed Central PMCID: PMC1171011. 26. Sato M, Hata N, Asagiri M, Nakaya T, Taniguchi T, Tanaka N. Positive feedback regulation of type I IFN genes by the IFN-inducible transcription factor IRF-7. FEBS Lett. 1998; 441(1):106–10. https://doi.org/ 10.1016/s0014-5793(98)01514-2 PMID: 9877175. 27. Horvath CM, Darnell JE. The state of the STATs: recent developments in the study of signal transduc- tion to the nucleus. Curr Opin Cell Biol. 1997; 9(2):233–9. https://doi.org/10.1016/s0955-0674(97) 80067-1 PMID: 9069254. 28. Levy DE, Marie IJ, Durbin JE. Induction and function of type I and III interferon in response to viral infec- tion. Curr Opin Virol. 2011; 1(6):476–86. https://doi.org/10.1016/j.coviro.2011.11.001 PMID: 22323926; PubMed Central PMCID: PMC3272644. 29. Bowie A, Kiss-Toth E, Symons JA, Smith GL, Dower SK, O’Neill LA. A46R and A52R from vaccinia virus are antagonists of host IL-1 and toll-like receptor signaling. Proc Natl Acad Sci U S A. 2000; 97 (18):10162–7. https://doi.org/10.1073/pnas.160027697 PMID: 10920188. 30. Stack J, Haga IR, Schroder M, Bartlett NW, Maloney G, Reading PC, et al. Vaccinia virus protein A46R targets multiple Toll-like-interleukin-1 receptor adaptors and contributes to virulence. J Exp Med. 2005; 201(6):1007–18. https://doi.org/10.1084/jem.20041442 PMID: 15767367. PLOS Pathogens | https://doi.org/10.1371/journal.ppat.1007778 October 11, 2019 22 / 25 Type I interferon-dependent CCL4 expression protects virus-infected tissue 31. Zhang P, Langland JO, Jacobs BL, Samuel CE. Protein kinase PKR-dependent activation of mitogen- activated protein kinases occurs through mitochondrial adapter IPS-1 and is antagonized by vaccinia virus E3L. J Virol. 2009; 83(11):5718–25. https://doi.org/10.1128/JVI.00224-09 PMID: 19321614; PubMed Central PMCID: PMC2681938. 32. Davies ML, Sei JJ, Siciliano NA, Xu RH, Roscoe F, Sigal LJ, et al. MyD88-dependent immunity to a nat- ural model of vaccinia virus infection does not involve Toll-like receptor 2. J Virol. 2014; 88(6):3557–67. https://doi.org/10.1128/JVI.02776-13 PMID: 24403581; PubMed Central PMCID: PMC3957935. 33. Martinez J, Huang X, Yang Y. Direct TLR2 signaling is critical for NK cell activation and function in response to vaccinia viral infection. PLoS pathogens. 2010; 6(3):e1000811. Epub 2010/03/20. https:// doi.org/10.1371/journal.ppat.1000811 PMID: 20300608; PubMed Central PMCID: PMC2837413. 34. Quigley M, Martinez J, Huang X, Yang Y. A critical role for direct TLR2-MyD88 signaling in CD8 T-cell clonal expansion and memory formation following vaccinia viral infection. Blood. 2009; 113(10):2256– 64. Epub 2008/10/25. https://doi.org/10.1182/blood-2008-03-148809 PMID: 18948575; PubMed Cen- tral PMCID: PMC2652371. 35. Zhao Y, De Trez C, Flynn R, Ware CF, Croft M, Salek-Ardakani S. The adaptor molecule MyD88 directly promotes CD8 T cell responses to vaccinia virus. Journal of immunology. 2009; 182(10):6278– 86. Epub 2009/05/06. https://doi.org/10.4049/jimmunol.0803682 PMID: 19414781; PubMed Central PMCID: PMC2712123. 36. Martinez J, Huang X, Yang Y. Toll-like receptor 8-mediated activation of murine plasmacytoid dendritic cells by vaccinia viral DNA. Proceedings of the National Academy of Sciences of the United States of America. 2010; 107(14):6442–7. Epub 2010/03/24. https://doi.org/10.1073/pnas.0913291107 PMID: 20308556; PubMed Central PMCID: PMC2851984. 37. Hoebe K, Du X, Georgel P, Janssen E, Tabeta K, Kim SO, et al. Identification of Lps2 as a key trans- ducer of MyD88-independent TIR signalling. Nature. 2003; 424(6950):743–8. https://doi.org/10.1038/ nature01889 PMID: 12872135. 38. Meade N, King M, Munger J, Walsh D. mTOR Dysregulation by Vaccinia Virus F17 Controls Multiple Processes with Varying Roles in Infection. J Virol. 2019;93(15). https://doi.org/10.1128/JVI.00784-19 PMID: 31118254. 39. Schoggins JW, MacDuff DA, Imanaka N, Gainey MD, Shrestha B, Eitson JL, et al. Pan-viral specificity of IFN-induced genes reveals new roles for cGAS in innate immunity. Nature. 2014; 505(7485):691–5. https://doi.org/10.1038/nature12862 PMID: 24284630; PubMed Central PMCID: PMC4077721. 40. Lian H, Wei J, Zang R, Ye W, Yang Q, Zhang XN, et al. ZCCHC3 is a co-sensor of cGAS for dsDNA rec- ognition in innate immune response. Nat Commun. 2018; 9(1):3349. https://doi.org/10.1038/s41467- 018-05559-w PMID: 30135424; PubMed Central PMCID: PMC6105683. 41. Abe T, Barber GN. Cytosolic-DNA-mediated, STING-dependent proinflammatory gene induction necessitates canonical NF-kappaB activation through TBK1. J Virol. 2014; 88(10):5328–41. https://doi. org/10.1128/JVI.00037-14 PMID: 24600004; PubMed Central PMCID: PMC4019140. 42. Fang R, Wang C, Jiang Q, Lv M, Gao P, Yu X, et al. NEMO-IKKbeta Are Essential for IRF3 and NF-kap- paB Activation in the cGAS-STING Pathway. J Immunol. 2017; 199(9):3222–33. https://doi.org/10. 4049/jimmunol.1700699 PMID: 28939760. 43. Cai X, Chiu YH, Chen ZJ. The cGAS-cGAMP-STING pathway of cytosolic DNA sensing and signaling. Mol Cell. 2014; 54(2):289–96. https://doi.org/10.1016/j.molcel.2014.03.040 PMID: 24766893. 44. Chen H, Sun H, You F, Sun W, Zhou X, Chen L, et al. Activation of STAT6 by STING is critical for antivi- ral innate immunity. Cell. 2011; 147(2):436–46. Epub 2011/10/18. https://doi.org/10.1016/j.cell.2011. 09.022 PMID: 22000020. 45. Bhat N, Fitzgerald KA. Recognition of cytosolic DNA by cGAS and other STING-dependent sensors. Eur J Immunol. 2014; 44(3):634–40. https://doi.org/10.1002/eji.201344127 PMID: 24356864; PubMed Central PMCID: PMC4621431. 46. Rubio D, Xu RH, Remakus S, Krouse TE, Truckenmiller ME, Thapa RJ, et al. Cross-talk between the Type I interferon and Nuclear Factor Kappa B pathways rescues resistance to a viral disease. Cell Host Microbe. 2013;In press. 47. Howell MD, Gao P, Kim BE, Lesley LJ, Streib JE, Taylor PA, et al. The signal transducer and activator of transcription 6 gene (STAT6) increases the propensity of patients with atopic dermatitis toward dis- seminated viral skin infections. J Allergy Clin Immunol. 2011; 128(5):1006–14. https://doi.org/10.1016/j. jaci.2011.06.003 PMID: 21762972; PubMed Central PMCID: PMC3205328. 48. Di Pilato M, Mejias-Perez E, Sorzano COS, Esteban M. Distinct Roles of Vaccinia Virus NF-kappaB Inhibitor Proteins A52, B15, and K7 in the Immune Response. J Virol. 2017;91(13). https://doi.org/10. 1128/JVI.00575-17 PMID: 28424281; PubMed Central PMCID: PMC5469261. PLOS Pathogens | https://doi.org/10.1371/journal.ppat.1007778 October 11, 2019 23 / 25 Type I interferon-dependent CCL4 expression protects virus-infected tissue 49. Howell MD, Gallo RL, Boguniewicz M, Jones JF, Wong C, Streib JE, et al. Cytokine milieu of atopic der- matitis skin subverts the innate immune response to vaccinia virus. Immunity. 2006; 24(3):341–8. Epub 2006/03/21. https://doi.org/10.1016/j.immuni.2006.02.006 PMID: 16546102. 50. Mahalingam S, Karupiah G, Takeda K, Akira S, Matthaei KI, Foster PS. Enhanced resistance in STAT6-deficient mice to infection with ectromelia virus. Proc Natl Acad Sci U S A. 2001; 98(12):6812– 7. https://doi.org/10.1073/pnas.111151098 PMID: 11371617. 51. Ember SW, Ren H, Ferguson BJ, Smith GL. Vaccinia virus protein C4 inhibits NF-kappaB activation and promotes virus virulence. The Journal of general virology. 2012; 93(Pt 10):2098–108. Epub 2012/ 07/14. https://doi.org/10.1099/vir.0.045070-0 PMID: 22791606; PubMed Central PMCID: PMC3541790. 52. Fischer MA, Davies ML, Reider IE, Heipertz EL, Epler MR, Sei JJ, et al. CD11b(+), Ly6G(+) cells pro- duce type I interferon and exhibit tissue protective properties following peripheral virus infection. PLoS Pathog. 2011; 7(11):e1002374. https://doi.org/10.1371/journal.ppat.1002374 PMID: 22102816; PubMed Central PMCID: PMC3213107. 53. Gale M Jr., Understanding the Expanding Roles of Interferon and Cytokines in Health and Disease. J Interferon Cytokine Res. 2018; 38(5):195–6. https://doi.org/10.1089/jir.2018.29007-mgj PMID: 54. Gale M Jr., Looking Forward in Interferon and Cytokine Research: Defining Effector Genes of Interferon and Cytokine Actions. J Interferon Cytokine Res. 2018; 38(6):235. https://doi.org/10.1089/jir.2018. 29008.mgj PMID: 29920132. 55. Davies ML, Parekh NJ, Kaminsky LW, Soni C, Reider IE, Krouse TE, et al. A systemic macrophage response is required to contain a peripheral poxvirus infection. PLoS Pathog. 2017; 13(6):e1006435. https://doi.org/10.1371/journal.ppat.1006435 PMID: 28614386; PubMed Central PMCID: PMC5484545. 56. Iannacone M, Moseman EA, Tonti E, Bosurgi L, Junt T, Henrickson SE, et al. Subcapsular sinus macro- phages prevent CNS invasion on peripheral infection with a neurotropic virus. Nature. 2010; 465 (7301):1079–83. Epub 2010/06/26. https://doi.org/10.1038/nature09118 PMID: 20577213; PubMed Central PMCID: PMC2892812. 57. Junt T, Moseman EA, Iannacone M, Massberg S, Lang PA, Boes M, et al. Subcapsular sinus macro- phages in lymph nodes clear lymph-borne viruses and present them to antiviral B cells. Nature. 2007; 450(7166):110–4. Epub 2007/10/16. https://doi.org/10.1038/nature06287 PMID: 17934446. 58. Moseman EA, Iannacone M, Bosurgi L, Tonti E, Chevrier N, Tumanov A, et al. B cell maintenance of subcapsular sinus macrophages protects against a fatal viral infection independent of adaptive immu- nity. Immunity. 2012; 36(3):415–26. https://doi.org/10.1016/j.immuni.2012.01.013 PMID: 22386268; PubMed Central PMCID: PMC3359130. 59. Frederico B, Chao B, Lawler C, May JS, Stevenson PG. Subcapsular sinus macrophages limit acute gammaherpesvirus dissemination. J Gen Virol. 2015; 96(8):2314–27. https://doi.org/10.1099/vir.0. 000140 PMID: 25872742. 60. Farrell HE, Davis-Poynter N, Bruce K, Lawler C, Dolken L, Mach M, et al. Lymph Node Macrophages Restrict Murine Cytomegalovirus Dissemination. J Virol. 2015; 89(14):7147–58. https://doi.org/10.1128/ JVI.00480-15 PMID: 25926638; PubMed Central PMCID: PMC4473555. 61. Broder CC, Kennedy PE, Michaels F, Berger EA. Expression of foreign genes in cultured human pri- mary macrophages using recombinant vaccinia virus vectors. Gene. 1994; 142(2):167–74. https://doi. org/10.1016/0378-1119(94)90257-7 PMID: 8194748. 62. Ablasser A, Schmid-Burgk JL, Hemmerling I, Horvath GL, Schmidt T, Latz E, et al. Cell intrinsic immu- nity spreads to bystander cells via the intercellular transfer of cGAMP. Nature. 2013; 503(7477):530–4. Epub 2013/10/01. https://doi.org/10.1038/nature12640 PMID: 24077100. 63. Xu S, Ducroux A, Ponnurangam A, Vieyres G, Franz S, Musken M, et al. cGAS-Mediated Innate Immu- nity Spreads Intercellularly through HIV-1 Env-Induced Membrane Fusion Sites. Cell Host Microbe. 2016; 20(4):443–57. https://doi.org/10.1016/j.chom.2016.09.003 PMID: 27736643. 64. Rice AD, Turner PC, Embury JE, Moldawer LL, Baker HV, Moyer RW. Roles of vaccinia virus genes E3L and K3L and host genes PKR and RNase L during intratracheal infection of C57BL/6 mice. J Virol. 2011; 85(1):550–67. https://doi.org/10.1128/JVI.00254-10 PMID: 20943971; PubMed Central PMCID: PMC3014211. 65. Rivas C, Gil J, Melkova Z, Esteban M, Diaz-Guerra M. Vaccinia virus E3L protein is an inhibitor of the interferon (i.f.n.)-induced 2-5A synthetase enzyme. Virology. 1998; 243(2):406–14. https://doi.org/10. 1006/viro.1998.9072 PMID: 9568039. 66. Chang HW, Watson JC, Jacobs BL. The E3L gene of vaccinia virus encodes an inhibitor of the inter- feron-induced, double-stranded RNA-dependent protein kinase. Proc Natl Acad Sci U S A. 1992; 89 PLOS Pathogens | https://doi.org/10.1371/journal.ppat.1007778 October 11, 2019 24 / 25 Type I interferon-dependent CCL4 expression protects virus-infected tissue (11):4825–9. https://doi.org/10.1073/pnas.89.11.4825 PMID: 1350676; PubMed Central PMCID: PMC49180. 67. Odendall C, Dixit E, Stavru F, Bierne H, Franz KM, Durbin AF, et al. Diverse intracellular pathogens acti- vate type III interferon expression from peroxisomes. Nat Immunol. 2014; 15(8):717–26. https://doi.org/ 10.1038/ni.2915 PMID: 24952503; PubMed Central PMCID: PMC4106986. 68. Alcami A, Symons JA, Smith GL. The vaccinia virus soluble alpha/beta interferon (IFN) receptor binds to the cell surface and protects cells from the antiviral effects of IFN. J Virol. 2000; 74(23):11230–9. https://doi.org/10.1128/jvi.74.23.11230-11239.2000 PMID: 11070021. 69. Najarro P, Traktman P, Lewis JA. Vaccinia virus blocks gamma interferon signal transduction: viral VH1 phosphatase reverses Stat1 activation. J Virol. 2001; 75(7):3185–96. https://doi.org/10.1128/JVI.75.7. 3185-3196.2001 PMID: 11238845; PubMed Central PMCID: PMC114112. 70. Baldanta S, Fernandez-Escobar M, Acin-Perez R, Albert M, Camafeita E, Jorge I, et al. ISG15 governs mitochondrial function in macrophages following vaccinia virus infection. PLoS Pathog. 2017; 13(10): e1006651. https://doi.org/10.1371/journal.ppat.1006651 PMID: 29077752; PubMed Central PMCID: PMC5659798. 71. Guerra S, Caceres A, Knobeloch KP, Horak I, Esteban M. Vaccinia virus E3 protein prevents the antivi- ral action of ISG15. PLoS Pathog. 2008; 4(7):e1000096. https://doi.org/10.1371/journal.ppat.1000096 PMID: 18604270; PubMed Central PMCID: PMC2434199. 72. Haga IR, Bowie AG. Evasion of innate immunity by vaccinia virus. Parasitology. 2005; 130 Suppl:S11– 25. https://doi.org/10.1017/S0031182005008127 PMID: 16281988. 73. Harte MT, Haga IR, Maloney G, Gray P, Reading PC, Bartlett NW, et al. The poxvirus protein A52R tar- gets Toll-like receptor signaling complexes to suppress host defense. J Exp Med. 2003; 197(3):343–51. https://doi.org/10.1084/jem.20021652 PMID: 12566418. 74. Norbury CC, Malide D, Gibbs JS, Bennink JR, Yewdell JW. Visualizing priming of virus-specific CD8+ T cells by infected dendritic cells in vivo. Nat Immunol. 2002; 3(3):265–71. https://doi.org/10.1038/ni762 PMID: 11828323. 75. Lehmann MH, Kastenmuller W, Kandemir JD, Brandt F, Suezer Y, Sutter G. Modified vaccinia virus ankara triggers chemotaxis of monocytes and early respiratory immigration of leukocytes by induction of CCL2 expression. J Virol. 2009; 83(6):2540–52. https://doi.org/10.1128/JVI.01884-08 PMID: 19129447; PubMed Central PMCID: PMC2648269. 76. Lehmann MH, Torres-Dominguez LE, Price PJ, Brandmuller C, Kirschning CJ, Sutter G. CCL2 expres- sion is mediated by type I IFN receptor and recruits NK and T cells to the lung during MVA infection. J Leukoc Biol. 2016; 99(6):1057–64. https://doi.org/10.1189/jlb.4MA0815-376RR PMID: 26992431. 77. Bonduelle O, Duffy D, Verrier B, Combadiere C, Combadiere B. Cutting edge: Protective effect of CX3CR1+ dendritic cells in a vaccinia virus pulmonary infection model. Journal of immunology. 2012; 188(3):952–6. Epub 2012/01/06. https://doi.org/10.4049/jimmunol.1004164 PMID: 22219332. 78. Restifo NP, Bacik I, Irvine KR, Yewdell JW, McCabe BJ, Anderson RW, et al. Antigen processing in vivo and the elicitation of primary CTL responses. J Immunol. 1995; 154(9):4414–22. PMID: 7722298. 79. Lev A, Dimberu P, Das SR, Maynard JC, Nicchitta CV, Bennink JR, et al. Efficient cross-priming of anti- viral CD8+ T cells by antigen donor cells is GRP94 independent. J Immunol. 2009; 183(7):4205–10. https://doi.org/10.4049/jimmunol.0901828 PMID: 19752220; PubMed Central PMCID: PMC2749969. 80. Strickland JE, Greenhalgh DA, Koceva-Chyla A, Hennings H, Restrepo C, Balaschak M, et al. Develop- ment of murine epidermal cell lines which contain an activated rasHa oncogene and form papillomas in skin grafts on athymic nude mouse hosts. Cancer Res. 1988; 48(1):165–9. PMID: 3121168. 81. Kim J, Huh J, Hwang M, Kwon EH, Jung DJ, Brinkmann MM, et al. Acidic amino acid residues in the jux- tamembrane region of the nucleotide-sensing TLRs are important for UNC93B1 binding and signaling. J Immunol. 2013; 190(10):5287–95. https://doi.org/10.4049/jimmunol.1202767 PMID: 23585677. 82. Tian T, Jin MQ, Dubin K, King SL, Hoetzenecker W, Murphy GF, et al. IL-1R Type 1-Deficient Mice Demonstrate an Impaired Host Immune Response against Cutaneous Vaccinia Virus Infection. J Immu- nol. 2017; 198(11):4341–51. https://doi.org/10.4049/jimmunol.1500106 PMID: 28468973; PubMed Central PMCID: PMC5506850. 83. Ochoa-Callejero L, Perez-Martinez L, Rubio-Mediavilla S, Oteo JA, Martinez A, Blanco JR. Maraviroc, a CCR5 antagonist, prevents development of hepatocellular carcinoma in a mouse model. PLoS One. 2013; 8(1):e53992. Epub 2013/01/18. https://doi.org/10.1371/journal.pone.0053992 PMID: 23326556; PubMed Central PMCID: PMC3541191. PLOS Pathogens | https://doi.org/10.1371/journal.ppat.1007778 October 11, 2019 25 / 25

Journal

PLoS PathogensPublic Library of Science (PLoS) Journal

Published: Oct 11, 2019

There are no references for this article.