Advertisement

  • Loading metrics

Open Access

Peer-reviewed

Research Article

Human cytomegalovirus blocks canonical TGFβ signaling during lytic infection to limit induction of type I interferons

  • Andrew H. Pham,

    Roles Investigation, Writing – review & editing

    Affiliation Vaccine and Gene Therapy Institute, Oregon Health and Science University, Beaverton, Oregon, United States of America

    ⨯
  • Jennifer Mitchell,

    Roles Investigation

    Affiliation Vaccine and Gene Therapy Institute, Oregon Health and Science University, Beaverton, Oregon, United States of America

    ⨯
  • Sara Botto,

    Roles Resources

    Affiliation Vaccine and Gene Therapy Institute, Oregon Health and Science University, Beaverton, Oregon, United States of America

    ⨯
  • Kara M. Pryke,

    Roles Resources

    Affiliation Vaccine and Gene Therapy Institute, Oregon Health and Science University, Beaverton, Oregon, United States of America

    ⨯
  • Victor R. DeFilippis,

    Roles Conceptualization, Resources, Writing – review & editing

    Affiliation Vaccine and Gene Therapy Institute, Oregon Health and Science University, Beaverton, Oregon, United States of America

    ⨯
  • Meaghan H. Hancock

    Roles Conceptualization, Formal analysis, Investigation, Methodology, Supervision, Writing – original draft, Writing – review & editing

    hancocme@ohsu.edu

    Affiliation Vaccine and Gene Therapy Institute, Oregon Health and Science University, Beaverton, Oregon, United States of America

    ⨯

Abstract

Human cytomegalovirus (HCMV) microRNAs (miRNAs) significantly rewire host signaling pathways to support the viral lifecycle and regulate host cell responses. Here we show that SMAD3 expression is regulated by HCMV miR-UL22A and contributes to the IRF7-mediated induction of type I IFNs and IFN-stimulated genes (ISGs) in human fibroblasts. Addition of exogenous TGFβ interferes with the replication of a miR-UL22A mutant virus in a SMAD3-dependent manner in wild type fibroblasts, but not in cells lacking IRF7, indicating that downregulation of SMAD3 expression to limit IFN induction is important for efficient lytic replication. These findings uncover a novel interplay between SMAD3 and innate immunity during HCMV infection and highlight the role of viral miRNAs in modulating these responses.

Author summary

Cells trigger the interferon (IFN) response to induce the expression of cellular genes that limit virus replication. In turn, viruses have evolved numerous countermeasures to avoid the effects of IFN signaling. Using a microRNA (miRNA) mutant virus we have uncovered a novel means of regulating the IFN response during human cytomegalovirus (HCMV) infection. Lytic HCMV infection induces the production of TGFβ, which binds to the TGFβ receptor and activates the receptor-associated SMAD SMAD3. SMAD3, together with IRF7, induces the expression of IFNβ and downstream IFN-stimulated genes in human fibroblasts. To counteract this, HCMV miR-UL22A, along with other HCMV gene products, directly targets SMAD3 for downregulation. Infection of fibroblasts with a miR-UL22A mutant virus results in enhanced type I IFN production in a SMAD3- and IRF7-dependent manner and the virus is impaired for growth in the presence of TGFβ, but only when both SMAD3 and IRF7 are present, highlighting the unique interaction between TGFβ and innate immune signaling.

Citation: Pham AH, Mitchell J, Botto S, Pryke KM, DeFilippis VR, Hancock MH (2021) Human cytomegalovirus blocks canonical TGFβ signaling during lytic infection to limit induction of type I interferons. PLoS Pathog 17(8): e1009380. https://doi.org/10.1371/journal.ppat.1009380

Editor: Erik K. Flemington, Tulane Health Sciences Center, UNITED STATES

Received: March 11, 2021; Accepted: August 3, 2021; Published: August 19, 2021

Copyright: © 2021 Pham 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 files.

Funding: A.H.P, J.M. and M.H.H are supported by National Institute of Health P01 AI127335 (https://www.niaid.nih.gov/). S.B., K.M.P. and V.R.D are supported by National Institute of Health R01 AI143660 (https://www.niaid.nih.gov/). The funders had no role in the 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.

Introduction

Human Cytomegalovirus (HCMV) has co-evolved with its host over millions of years, resulting in exquisite control of both the cellular environment and the viral lifecycle that is highly cell type-dependent. Successful viral gene expression depends on blocking the powerful innate antiviral responses induced by viral binding and entry to limit interferon (IFN) production and signaling that act to render cells less permissive to viral infection [1]. To evade the innate immune response, HCMV encodes numerous gene products that block the induction of intrinsic antiviral responses and the production of IFN and IFN stimulated genes (ISGs) (reviewed in [2,3]). Along with IFNs, viral infection results in the secretion of additional cellular and viral cytokines and chemokines, which act in an autocrine and paracrine fashion to alter the intracellular and extracellular environment [4]. For example, lytically infected cells [5–7] as well as latently-infected CD34+ hematopoietic progenitor cells (HPCs) [8–10], produce and secrete transforming growth factor beta (TGFβ) which causes myelosuppression and has significant implications for hematopoietic stem cell transplantation [9].

TGFβ is a powerful regulator of numerous cellular pathways and has important roles in inflammation, immune modulation and cellular differentiation. The specific transcriptional outcomes of TGFβ signaling depends on cell type, TGFβ concentration and presence of additional signaling regulators resulting in either transcriptional activation or repression of different subsets of cellular genes (reviewed in [11,12]). The complex and context-dependent outcome of canonical TGFβ signaling initiates with a relatively simple signal transduction pathway. TGFβ binding results in assembly of the receptor complex consisting of TGFβ receptors I and II. Receptor-associated SMADs (R-SMADs) SMAD2 and/or SMAD3 are then recruited and phosphorylated. Phosphorylation of R-SMADs enhances their interaction with the co-SMAD SMAD4 which results in shuttling of the R-SMAD/SMAD4 complex to the nucleus. SMADs regulate transcription by altering chromatin structure and generally have only weak affinity for the SMAD binding element (CAGAC) [13–15], therefore other DNA binding proteins are required for selective binding of the SMAD complexes to target elements. In addition to the choice of DNA binding partners, the recruitment of transcriptional coactivators, such as CBP/p300 [16] or corepressors, such as TGIF [17], SKI or SnoN [18] is also critical for determining the outcome of TGFβ signaling. Thus, the transcriptional outcome of canonical TGFβ signaling critically depends on the presence of a SMAD DNA binding cofactor as well as coactivators or corepressors, whose expression and localization are regulated by additional cellular signaling pathways in a context-dependent and cell type-specific manner [12].

Given the critical role that TGFβ plays in many cellular functions, activation and signaling by TGFβ is carefully regulated by the cell. One of the more recently studied means of regulation of the TGFβ signaling pathway is through the expression of cellular miRNAs [19]. miRNAs are small, ~22 nucleotide regulatory RNAs that post-transcriptionally regulate expression of genes through binding regions of complementarity in the targeted transcript. Sequence recognition occurs through the ‘seed region’ of miRNAs, nucleotides 2–8, and most commonly the 3’ UTR of the targeted gene, although binding to other regions of the transcript can also mediate regulation [20]. miRNAs are the mRNA recognition component of a larger multi-protein RNA-induced silencing complex (RISC) that recruits proteins that mediate translational repression and/or mRNA degradation to the targeted transcript [21]. Cellular miRNAs regulate the TGFβ signaling pathway at every level. Ligands, receptors, R-, inhibitory (I)- and co-SMADs and non-SMAD pathway components are all targets of cellular miRNAs [19]. By targeting components of the signaling pathway as well as downstream transcriptional targets, miRNAs regulate all aspects of the intricate negative and positive feedback networks of the TGFβ signaling pathway.

HCMV also encodes its own miRNAs [22], and we have previously demonstrated that HCMV miR-UL22A-5p and -3p block canonical TGFβ signaling in latently-infected CD34+ HPCs by decreasing expression of the critical R-SMAD SMAD3 [9]. Importantly, infection with a ΔmiR-UL22A mutant virus restored TGFβ signaling to levels observed in mock infected HPCs and the mutant virus was impaired for reactivation from latency due to a loss of viral genomes or viral genome-containing cells. The attenuation of canonical TGFβ signaling in CD34+ HPCs as well as maintenance of the viral genomes during latency was restored by expression of a SMAD3 shRNA from the ΔmiR-UL22A genome, indicating that targeting SMAD3 and the canonical TGFβ signaling pathway is essential to maintain the viral genome in latently infected cells [9]. The mechanism of TGFβ-mediated viral genome loss remains to be determined and the effect of TGFβ signaling on other stages of the HCMV lifecycle is also still unknown.

Here we show that canonical TGFβ signaling negatively affects lytic replication of HCMV and is counteracted by reduced SMAD3 expression mediated, in part, by miR-UL22A. We show that miR-UL22A targeting of SMAD3 is critical for viral replication in the presence of TGFβ through attenuating IRF7-mediated activation of type I IFNs and ISGs. This study uncovers a novel link between SMAD3 and IRF7 in the induction of type I IFNs, highlighting the crosstalk between TGFβ and innate immune signaling during HCMV infection.

Results

Canonical TGFβ signaling is impaired during lytic HCMV infection

Our work [9] and that of others [6–8,23,24] has shown that CMV-infected cells produce and secrete TGFβ. We hypothesized that HCMV protects the lytically infected cell from the effects of TGFβ signaling by manipulating components of the TGFβ signaling pathway, as we have observed during latent infection in CD34+ HPCs [9]. To test this hypothesis, we infected normal human dermal fibroblasts (NHDF) (Fig 1A and 1C) and primary human aortic endothelial cells (hAEC) (Fig 1B and 1D) with the wild type TB40/E strain of HCMV for 48 hours followed by overnight serum starvation and 4 hour stimulation with TGFβ. In contrast to mock-infected cells, which respond to TGFβ treatment by upregulating the TGFβ-responsive transcripts JunB and SERPINE, HCMV-infected fibroblasts and endothelial cells do not show an enhancement in transcript expression following TGFβ treatment (Fig 1A and 1B). We next harvested protein lysates from mock- and HCMV-infected cells treated with TGFβ and analyzed expression of total and phosphorylated levels of SMAD3. Levels of both total and phosphorylated SMAD3 are reduced in HCMV-infected fibroblasts (Fig 1C) and endothelial cells (Fig 1D) compared to mock-infected cells. These data indicate that lytic HCMV infection reduces the levels of SMAD3 protein which contributes to a block in canonical TGFβ signaling.

thumbnail
Download:
Fig 1. HCMV infection blocks canonical TGFβ signaling.

NHDF (A) of hAEC (B) were infected with WT TB40E for 48 hours, followed by overnight serum starvation and stimulation with recombinant TGFβ (100pg/mL) for 4 hours. RNA was isolated and subjected to qRT-PCR for JunB and SERPINE. Experiments were performed in triplicate. * p<0.05 by two tailed Student’s t test. NHDF (C) or hAEC (D) were infected as above and protein lysates subjected to immunoblotting using the indicated antibodies.

https://doi.org/10.1371/journal.ppat.1009380.g001

Mutation of miR-UL22A partially relieves the block to canonical TGFβ signaling observed during HCMV infection

We have previously identified SMAD3 as a target of the HCMV miRNAs miR-UL22A-5p and -3p, and have shown that miR-UL22A-mediated reduction of SMAD3 is required for genome maintenance during latent infection [9]. To investigate the contribution of miR-UL22A targeting of SMAD3 to the block in canonical TGFβ signaling observed during HCMV lytic infection, we assessed SMAD3 transcript levels after infection with WT and ΔmiR-UL22A virus. As shown in Fig 2A, we observed a significant decrease in SMAD3 transcript levels upon infection with WT HCMV at 3 (p = 0.05) and 6 (p<0.01) days post-infection (dpi), suggesting that the decrease in total and phosphorylated protein (Fig 1C) is due, at least in part, to a decrease in transcription of SMAD3. Moreover, we observed only a partial restoration in SMAD3 transcript levels after ΔmiR-UL22A infection, although significantly increased (p = 0.02) compared to WT-infected cells at 6 dpi. We next treated mock, WT or ΔmiR-UL22A virus-infected fibroblasts with TGFβ and assessed phospho- and total SMAD3 protein levels (Fig 2B). Cells infected with the miR-UL22A knockout virus showed enhanced SMAD3 phosphorylation and increased SMAD3 protein levels compared to WT-infected cells but did not reach the levels observed in mock-infected cells. The effect of miR-UL22A deletion on SMAD3 protein levels was quantified from 4 independent experiments (right panel, Fig 2B). We also assessed downstream transcriptional targets of TGFβ signaling after infection with WT or miR-UL22A knockout virus (Fig 2C). Infection with miR-UL22A knockout virus resulted in significantly increased JunB (p = 0.03) and SERPINE (p<0.01) transcript levels compared to WT infected cells but again infection with the ΔmiR-UL22A mutant virus did not restore transcript levels to that observed in mock-infected cells. Similar observations were made with endothelial cells (Fig 2D and 2E) infected with WT and miR-UL22A knockout viruses. These data indicate that miR-UL22A contributes to the blockade in TGFβ signaling observed during HCMV infection but, unlike in CD34+ HPCs where mutation of miR-UL22A completely restores TGFβ signaling, other viral factors contribute to the reduction in SMAD3 transcript and protein levels during lytic infection.

thumbnail
Download:
Fig 2. miR-UL22A targets SMAD3 during lytic infection.

(A) NHDF were infected for 3 or 6 days with the indicated virus and then RNA was isolated and subjected to qRT-PCR for SMAD3. Experiments were performed in triplicate. NHDF (B) or hAEC (D) were infected with the indicated viruses for 48 hours followed by overnight serum starvation and stimulation with TGFβ (100pg/mL) for 4 hours. Protein lysates were subjected to immunoblotting using the indicated antibodies. Western blot quantification of SMAD3 levels from 4 independent experiments in (B) was performed using ImageJ analysis software. NHDF (C) or hAEC (E) were infected as in (B) and RNA was isolated followed by qRT-PCR for JunB or SERPINE. Experiments were performed in triplicate. * p<0.05 by two tailed Student’s t test.

https://doi.org/10.1371/journal.ppat.1009380.g002

miR-UL22A downregulation of SMAD3 expression is important for viral replication in the presence of exogenous TGFβ

The data presented here indicate that HCMV encodes multiple mechanisms to reduce SMAD3 expression and block canonical TGFβ signaling suggesting an anti-viral role for TGFβ during lytic infection. With this in mind, we assessed the functional consequences of TGFβ signaling on viral replication using the ΔmiR-UL22A mutant virus which does not fully attenuate signaling through the canonical TGFβ signaling pathway (Fig 2). We also utilized a ΔmiR-UL22A mutant that expresses a SMAD3 shRNA in place of the miR-UL22A hairpin (ΔmiR-UL22A/SMAD3shRNA). We have previously determined that expression of a SMAD3 shRNA in place of miR-UL22A results in SMAD3 protein levels similar to WT infection [9]. Additionally, we observed that expression of an shRNA from the miR-UL22A locus restores TGFβ-responsive transcript levels to those observed during WT lytic infection (Fig 3A). We next performed multi-step growth analysis in fibroblasts infected with WT, ΔmiR-UL22A and ΔmiR-UL22A/SMAD3shRNA where exogenous TGFβ was added every 3 days. As shown in Fig 3B, replication of WT virus is unaffected by the addition of exogenous TGFβ. However, the miR-UL22A knockout virus, which shows an ~1 log growth defect upon low multiplicity infection, was further inhibited for growth in the presence of exogenous TGFβ. Expression of a SMAD3 shRNA in place of miR-UL22A restored growth of the mutant virus to WT levels and showed no growth defect upon TGFβ treatment. This data suggests that reducing SMAD3 protein levels and blocking the canonical TGFβ signaling pathway is important for efficient lytic viral replication.

thumbnail
Download:
Fig 3. Targeting SMAD3 for downregulation is necessary for efficient lytic replication.

(A) NHDF were infected with the indicated viruses for 48 hours followed by overnight serum starvation and stimulation with recombinant TGFβ (100pg/mL) for 4 hours. RNA was isolated and subjected to qRT-PCR for JunB and SERPINE. Experiments were performed in triplicate. * p<0.05 by two tailed Student’s t test. (B) NHDF were infected at 0.01 PFU/mL for 2 hours. 100pg/mL TGFβ was added immediately after infection and every 3 days throughout the experiment. Samples were harvested at the indicated timepoints and titered on NHDF. Experiment was performed in duplicate.

https://doi.org/10.1371/journal.ppat.1009380.g003

miR-UL22A targeting of SMAD3 regulates expression of IFNβ and interferon stimulated genes

The negative effect of exogenous TGFβ on lytic replication of the ΔmiR-UL22A virus and the restoration of mutant virus growth upon downregulation of SMAD3 implicates SMAD3 protein in anti-viral responses. SMAD proteins themselves have very weak affinity for SMAD binding elements in the promoters of targeted genes, and most often interact with additional DNA binding cofactors to mediate their effects [11]. One cofactor known to interact with SMAD3 is interferon regulatory factor 7 (IRF7) which, along with SMAD3, has been shown to mediate activation of the IFNβ promoter [25]. Thus, we asked whether miR-UL22A, through its ability to target SMAD3, could affect the induction of IFNβ and ISGs in response to a variety of stimuli. As shown in Fig 4, transfection of miR-UL22A, SMAD3 siRNA or IRF7 siRNA significantly blocked the induction of IFNβ or the ISG RSAD2 (Viperin) in response to the cytosolic DNA stimuli ISD90 or UV-HCMV (Fig 4A and 4C) and RNA stimuli, including the single-stranded RNA virus Sendai virus and polyI:C (Fig 4B and 4D). We also assessed expression of SMAD3 and IRF7 following miR-UL22A and siRNA transfection (Fig 4E–4H) Interestingly, activation of pattern recognition receptors induced SMAD3 expression, which was regulated by miR-UL22A, as previously reported [9], and IRF7, suggesting that SMAD3 may behave as an ISG. Furthermore, IRF7 expression is regulated by miR-UL22A and SMAD3, suggesting an intriguing co-regulation. These data implicate both SMAD3 and IRF7 in IFNβ and ISG induction in fibroblasts and show that miR-UL22A, likely through targeting SMAD3, also interferes with IFNβ and ISG induction.

thumbnail
Download:
Fig 4. miR-UL22A and SMAD3 siRNA affect IFN and ISG induction in response to multiple stimuli.

(A-H) NHDF were transfected with negative control, miR-UL22A mimic or SMAD3 or IRF7 siRNA. 24 hours post-transfection cells were either: (A,C,E,G) infected with UV-inactivated HCMV or transfected with ISD90 for a further 24 hours or (B,D,F,H) infected with Sendai virus or transfected with polyI:C for a further 24 hours. After this time, RNA was isolated and qRT-PCR was performed for IFNβ, RSAD2, SMAD3 or IRF7. All experiments were performed in triplicate. * p<0.05 by two tailed Student’s t test.

https://doi.org/10.1371/journal.ppat.1009380.g004

We next wanted to validate that the effects of miR-UL22A and SMAD3 knockdown on IFN and ISG induction occurred through the well characterized STING and JAK/STAT mediated signaling pathways. To do this we tested the effects of miR-UL22A and SMAD3 or IRF7 siRNA expression on IFN and ISG induction in previously constructed telomerized human fibroblast (tHF) cell lines deficient for STING or the type I IFN receptor IFNAR (ΔSTING and ΔIFNAR) [26–28]. STING is a key regulator in innate immune signaling downstream of cGAS recognition of incoming viral DNA including the response to HCMV infection [29,30]. In WT tHF cells, expression of miR-UL22A, SMAD3 or IRF7 siRNAs reduced IFNα, IFNβ and RSAD2 induction in response to ISD90, in agreement with the data presented in Fig 4 in NHDF. In the absence of STING, IFNs and RSAD2 were not induced by ISD90 stimulation and miR-UL22A and SMAD3 or IRF7 siRNAs did not further affect expression (Fig 5A–5C), indicating that SMAD3, as a target of miR-UL22A, functions downstream of STING-mediated innate immune signaling. In ΔIFNAR cells, the initial induction of IFNβ is detected, but further amplification mediated by the IFNAR/JAK/STAT signaling pathway does not ensue (Fig 5D). We observed that miR-UL22A, SMAD3 and IRF7 siRNAs attenuated the (reduced) induction of IFNβ in ΔIFNAR cells (Fig 5D), indicating that they function in the initial induction of IFNβ that occurs prior to signal amplification through the JAK/STAT pathway. IFNα and RSAD2 expression was abrogated in ΔIFNAR cells and not further affected by expression of miR-UL22A or SMAD3 or IRF7 siRNAs (Fig 5E and 5F) consistent with their functions as ISGs induced by JAK/STAT signaling, in the case of RSAD2, and directly dependent on IRF7 expression, in the case of IFNα [31–33]. Thus, this data supports the hypothesis that miR-UL22A modulates SMAD3-mediated induction of IFNβ and downstream ISG induction. SMAD3 and IRF7 expression is reduced by transfection with the corresponding siRNAs in the tHF, ΔSTING and ΔIFNAR cell lines (S1 Fig).

thumbnail
Download:
Fig 5. SMAD3 participates in STING- and IFN receptor-mediated signaling.

tHF, ΔSTING (A-C) or ΔIFNAR (D-F) tHFs were transfected with negative control, miR-UL22A mimic or SMAD3 or IRF7 siRNA for 48 hours after which cells were additionally mock transfected or transfected with ISD90 for 16 hours. RNA was isolated and subjected to qRT-PCR for IFNβ, IFNα1 or RSAD2. All experiments were performed in triplicate. * p<0.05 by two tailed Student’s t test.

https://doi.org/10.1371/journal.ppat.1009380.g005

In order to determine if the effect of miR-UL22A expression and knockdown of SMAD3 on IFN and ISG induction required IRF7, we derived an IRF7 knockout tHF cell line using CRISPR/Cas9 genome editing [27,28,34]. Endogenous IRF7 protein levels are generally undetectable in many cell types, but its expression can be induced by IFN or other innate immune stimuli [33]. As shown in Fig 6A, IRF7 expression is induced by transfection of the parental tHF cells with ISD90, but IRF7 protein was undetectable in the IRF7 knockout tHF cell line. In addition, IRF7 plays a key role in the induction of IFNα [33,35], which we show is not induced in the IRF7 knockout cells following ISD90 treatment (Fig 6C), further validating the ΔIRF7 cell line. We next tested the effects of miR-UL22A and SMAD3 or IRF7 siRNA expression on IFN and ISG induction after treatment of ΔIRF7 tHF cells with ISD90. SMAD3 and IRF7 expression is reduced by transfection with the corresponding siRNAs in the tHF and ΔIRF7 cell lines (S1 Fig). In ΔIRF7 cells, type I IFN and ISG induction is abrogated, and miR-UL22A or siRNAs targeting SMAD3 and IRF7 had no additional effect on gene expression consistent with the hypothesis that miR-UL22A, through targeting SMAD3, affects IRF7-mediated induction of IFNβ and downstream ISGs.

thumbnail
Download:
Fig 6. SMAD3 participates in IRF7-mediated signaling.

(A) tHF or ΔIRF7 tHFs were transfected with ISD90 for 16 hours. Protein lysates were subjected to immunoblotting for IRF7 or GAPDH. (B-D) tHF or ΔIRF7 tHFs were transfected with negative control, miR-UL22A mimic or SMAD3 or IRF7 siRNA for 48 hours after which cells were additionally mock transfected or transfected with ISD90 for 16 hours. RNA was isolated and subjected to qRT-PCR for IFNβ, IFNα1 or RSAD2. All experiments were performed in triplicate. * p<0.05 by two tailed Student’s t test.

https://doi.org/10.1371/journal.ppat.1009380.g006

miR-UL22A targeting of SMAD3 limits induction of type I interferons and ISGs during lytic infection in an IRF7-dependent manner

In order to assess the role of miR-UL22A regulation of SMAD3 and IRF7-mediated signaling in the context of HCMV infection, we next tested the induction of IFN transcripts following infection with WT and ΔmiR-UL22A mutant viruses in tHF and ΔIRF7 cells. As shown in Fig 7A and 7B, infection of tHF cells with the ΔmiR-UL22A virus resulted in increased IFNβ and IFNα transcript accumulation both in the absence and presence of IFN treatement compared WT- and ΔmiR-UL22A/SMAD3shRNA-infected cells. However, no significant induction of type I IFNs was observed when the viruses were used to infect ΔIRF7 cells (Fig 7C and 7D) We also measured IFN secretion following infection of tHF and ΔIRF7 cells with WT and miR-UL22A mutant viruses (Fig 7E–7H) and showed enhanced secretion of IFNβ and IFNα upon ΔmiR-UL22A mutant virus infection that is reduced in tHF cells infected with ΔmiR-UL22A/SMAD3shRNA virus and absent in ΔIRF7 cells. We then assessed the induction of additional ISGs following infection with the ΔmiR-UL22A mutant viruses and treatment with IFN. As shown in Fig 7I–7L, induction of IRF7 and RSAD2 is enhanced in tHF cells infected with ΔmiR-UL22A mutant virus compared to WT- and ΔmiR-UL22A/SMAD3shRNA-infected cells after IFN treatment but does not reach levels observed in IFN-treated, mock-infected cells. In contrast, IRF7 and RSAD2 were not induced by viral infection of ΔIRF7 cells. These data indicate that miR-UL22A, through targeting SMAD3, is important for reducing type I IFN and ISG production during lytic HCMV infection but also that other gene products are likely also involved in SMAD3 and IFN regulation. The data furthermore indicates that the induction of type I IFN and ISGs mediated by SMAD3 is via an IRF7-dependent pathway.

thumbnail
Download:
Fig 7. miR-UL22A targeting of SMAD3 regulates IFN and ISG induction in an IRF7-dependent manner.

(A-B; tHFs) and (C-D; ΔIRF7 cells) were infected with the indicated viruses for 48 hours followed by treatment with uIFN (1000U/mL) for 16 hours. RNA was isolated and subjected to qRT-PCR for IFNβ and IFNα1. (E-H) tHF and ΔIRF7 cells were infected as in (A) and supernatants were harvested after 72 hours. ELISAs were performed for (E,G) IFNβ and (F,H) pan-IFNα. (I-L) tHF and ΔIRF7 cells were infected as in (A) and treated with IFN after 24 hours of infection. RNA was isolated 48 hours post-infection and subjected to qRT-PCR for IRF7 and RSAD2. All experiments were performed in triplicate. * p<0.05 by two tailed Student’s t test.

https://doi.org/10.1371/journal.ppat.1009380.g007

TGFβ-mediated attenuation of lytic infection is mediated through SMAD3 and IRF7

Finally, in order to determine whether the TGFβ-mediated attenuation of replication of the ΔmiR-UL22A virus is due to the cross-talk between TGFβ and innate immune signaling, we analyzed the growth of WT and miR-UL22A mutant viruses in WT and ΔSTING or ΔIRF7 tHFs. Fig 8 demonstrates that while the ΔmiR-UL22A mutant virus shows reduced virus released into the supernatant upon TGFβ treatement in parental tHFs compared to WT and ΔmiR-UL22A/SMAD3shRNA viruses (Fig 8A), this defect was abrogated in cells lacking STING (Fig 8B), further supporting the role of innate immune signaling in attenuating replication of a ΔmiR-UL22A mutant virus. Moreover, the defect in ΔmiR-UL22A mutant virus growth is also abrogated in ΔIRF7 cells, (Fig 8C) indicating that IRF7 is directly involved in impeding ΔmiR-UL22A growth in response to TGFβ. Together, these data support the hypothesis that SMAD3 and IRF7 cooperate to induce IFN production during HCMV infection, which has negative effects on viral replication.

thumbnail
Download:
Fig 8. TGFβ impairs lytic replication in coordination with IRF7.

tHF (A), ΔSTING (B) or ΔIRF7 (C) tHFs were infected at 0.01 PFU/mL. TGFβ was added (100pg/mL) after initial infection and again at days 6 and 9. Supernatants were harvested and titered in NHDF. Experiments were performed in duplicate.

https://doi.org/10.1371/journal.ppat.1009380.g008

Discussion

In this study we show that SMAD3, a TGFβ receptor-associated SMAD, cooperates with IRF7 to induce type I IFN during HCMV infection. miR-UL22A, through downregulating SMAD3 expression, plays an important role in regulating IRF7-mediated IFN production and viral replication during lytic infection. Infection with virus lacking miR-UL22A results in enhanced IFN production and release, enhanced downstream ISG induction and inhibition of growth in the presence of exogenous TGFβ. However, if the ΔmiR-UL22A mutant expresses an shRNA targeting SMAD3, type I IFN and ISG induction as well as viral replication returns to WT levels, directly implicating the regulation of SMAD3 expression by miR-UL22A in modulation of IFN production and efficient viral replication. Importantly, this enhanced IFN and ISG induction in not observed when a ΔmiR-UL22A mutant is used to infect ΔIRF7 cells, indicating that the IFN induction occurs through a SMAD3- and IRF7-dependent pathway.

Herpesviruses manipulate the intrinsic and innate IFN signaling pathways to aid in their replication cycles, which is especially important during lytic infection, where efficient viral gene expression and new virion production is paramount. In order to induce the production of IFNs after recognition of viral infection by pattern recognition receptors, the transcription factors IRF3 and IRF7 are phosphorylated by the kinase TBK1, translocate to the nucleus and, along with c-Jun, ATF2, NFκB and CBP/p300, bind to the Ifnb promoter to induce its expression [36]. Autocrine IFN signaling through the IFN receptor then stimulates the production of more IRF7 and thus more type I IFN via positive feedback [35,37], establishing IRF7 as a ‘master regulator’ of IFN production [35]. In this study we show that SMAD3, a TGFβ receptor-associated SMAD, cooperates with IRF7 to induce type I IFNs during HCMV infection, highlighting a unique interconnection between TGFβ and IFN signaling.

An interaction between IRF7 and SMAD3 was first postulated due to the similarity of each transactivation domain and the fact that, upon phosphorylation, both proteins undergo structual rearrangements that promote complex formation [25]. SMAD3 was also shown to interact with IRF7 (but not IRF3) at the Ifnb promoter in mouse embryonic fibroblasts and this interaction was critical for Ifnb transcription [25]. Here we show that both IRF7 and SMAD3 are required for induction of IFN and ISG transcripts in response to a variety of PAMPs in human fibroblasts (Fig 4). The effect of miR-UL22A and SMAD3 siRNA on type I IFN and ISG induction was abrogated in the absence of the signaling adaptor STING (Fig 5A–5C), indicating that the SMAD3, like IRF7, functions as a component of the IFN-terminal innate immune signaling pathway. The hypothesis that SMAD3 and IRF7 work together in the initial induction of IFNβ is supported by the observations using ΔIFNAR cells, which are capable of the initial IFNβ production following infection but cannot amplify the response. Expression of miR-UL22A and SMAD3 or IRF7 siRNA blocks the initial induction of IFNβ in ΔIFNAR cells (Fig 5D). Furthermore, miR-UL22A and SMAD3 siRNA have no effect on the low levels of IFNβ transcription in IRF7 knockout cells (Fig 6B–6D), suggesting a cooperative function of SMAD3 and IRF7 in the induction of type I IFNs and downstream ISG induction in human fibroblasts.

The functional significance of SMAD3 targeting by miR-UL22A during lytic infection is underscored by the increased IFNα and IFNβ expression and secretion upon infection with a ΔmiR-UL22A mutant virus (Fig 7). This enhanced type I IFN response results in increased ISG induction in the absence of miR-UL22A expression, although not to levels observed in mock-infected cells indicating that likely other SMAD3-targeting gene products are involved in this process (Fig 7E and 7F). Critically, replacing the miR-UL22A hairpin locus with an shRNA targeting SMAD3 reduces type I IFN expression and secretion, along with ISG induction, to levels seen during WT infection. This indicates both the necessity and sufficiency of targeting SMAD3 for HCMV-mediated impairment of IFN. The induction of type I IFNs and the negative effect of TGFβ on replication of the ΔmiR-UL22A mutant virus is abrogated in cell lines lacking STING or IRF7 (Figs 7 and 8), further supporting the hypothesis that SMAD3 and IRF7 function together to limit replication of HCMV through the induction of a type I IFN response.

Regulation of IRF7 expression and function is utilized by α- and γ-herpesviruses as a means to dampen the IFN response during infection. Kaposi’s Sarcoma-associated Herpesvirus (KSHV) and Herpes Simplex virus (HSV) encode proteins that sequester or degrade IRF7 [38–45]. In contrast, during Epstein Barr virus (EBV) latent infection, IRF7 expression is stimulated by latent membrane protein-1 (LMP-1) [46–49] which in turn regulates expression of the EBNA1 Q promoter [50] and LMP-1 itself. However, during reactivation, EBV IE proteins BZLF1 [51] and LF2 [52] bind and repress IRF7 activity and BRLF1 downregulates IRF7 expression and IFNβ production [53]. Thus, along with the data presented here for HCMV, it is clear that targeting IRF7 expression and/or function during lytic infection is a common theme amongst the herpesvirus family.

The role of TGFβ signaling in herpesvirus infection is highly relevant yet complex, with differing effects on lytic and latency stages of the lifecycle. TGFβ treatment of EBV latently infected cells can induce reactivation via SMAD binding to the BZLF1 promoter [54]. Thus, EBV encodes factors that directly or indirectly interfere with components of the TGFβ signaling pathway [55–57]. Additionally, EBV proteins upregulate the cellular miRNA miR-146a which directly targets SMAD4 [58,59]. Likewise, KSHV proteins induce the expression of the cellular miR-17-92 family, which targets SMAD2 [60]. KSHV also encodes proteins [61–63] and miRNAs [64,65] that target components of the TGFβ signaling pathway. In contrast, infection of human mononuclear cells with HSV-1 induces TGFβ production [66] and use of conditional TGFβ knockout mouse models suggests that TGFβ signaling, while dampening the innate immune response, enhances HSV-1 latency [67].

Similar to HSV-1, the HCMV major immediate early proteins can activate the TGFβ promoter, which occurs indirectly through IE2 transactivation of the cellular immediate early transcription factor EGR-1, but was also shown to require additional viral factors during infection [6,7]. Recent studies implicate miR-US5-2 targeting of the transcription repressor NAB1 as one of the possible additional mechanisms [9]. Furthermore, HCMV IE1 and IE2 induce the expression of MMP-2 in renal tubular epithelial cells, which may result in enhanced activation of latent TGFβ in the extracellular matrix of HCMV-infected cells and contribute to the fibrosis observed during transplantation [68]. Thus, CMV infection induces the production of TGFβ by multiple mechanisms but encodes multiple factors to block the canonical TGFβ signaling pathway (Figs 1–3).

While removal of the pre-miR-UL22A sequence from the viral genome results in enhanced SMAD3 protein levels and expression of classical downstream TGFβ transcriptional targets during lytic infection, responses do not return to levels observed in mock-infected cells (Fig 2B–2E), suggesting that the virus uses additional mechanisms (utilizing viral proteins and/or additional miRNAs or long non-coding RNAs) to alter SMAD3 mRNA and protein levels and inhibit the TGFβ signaling pathway. Possible additional mechanisms utilized by HCMV to manipulate the TGFβ signaling pathway remain to be identified, but could include inhibiting SMAD3 transcription initiation, affecting the expression or stability of additional members of the signaling pathway or induction of negative pathway regulators, such as the I-SMADs. Interestingly, HCMV blocks the signaling pathway of the related TGFβ family member activin by directly targeting the activin receptor (ACVR1B) using miR-UL148D to prevent the production and release of IL6 in monocytes [69].

The interplay between type I IFN and TGFβ signaling during HCMV infection highlights the complex interconnection of signaling pathways that are regulated by viral proteins and non-coding RNAs. An inability to downregulate SMAD3 during latent infection results in a loss of viral genomes from the infected cells [9] and whether this phenotype is due to enhanced IFN signaling remains to be determined. Future studies will explore how this novel interplay between the TGFβ and IFN signaling pathways is important for HCMV latency in hematopoietic progenitor cells.

Materials and methods

Cell lines

Normal human dermal fibroblasts (NHDF), human foreskin fibroblasts (stably transduced with constitutively expressed human telomerase reverse transcriptase and the IRF/IFN-responsive pGreenFire-ISRE lentivector; tHF) [27] and the tHF ΔSTING [26], ΔIFNAR [28] and ΔIRF7 cell lines were cultured in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 5% heat-inactivated fetal bovine serum (FBS; HyClone), 100 units/ml penicillin, and 100ug/ml streptomycin (ThermoFisher). Human aortic endothelial cells (AEC) (CC-2535; Lonza) were cultured in EBM-2 basal medium with EGM-2 SingleQuotsTM supplement excluding Heparin (Lonza), as well as 10% FBS, penicillin, and streptomycin. All cells were maintained at 37°C and 5% CO2. Recombinant human TGFβ and Universal Type I IFN was obtained from R&D Systems.

HCMV Constructs and infections

HCMV used in this study include BAC-generated WT TB40/E expressing GFP from the SV40 promoter [70], a TB40/E mutant virus lacking the pre-miR-UL22A sequence or a TB40/E mutant virus with the pre-miR-UL22A sequence replaced by a SMAD3 shRNA generated by galK-mediated recombination [9]. Unintended off-target alterations in the genome sequence were ruled our by next-generation sequencing (NGS) of the complete genome using an Illumina MiSeq sequencing platform. All virus stocks were propagated and titered on NHDFs. Fibroblasts were infected with HCMV at three plaque-forming units (PFU)/cell and hAEC were infected with HCMV at five PFU/cell for 2 hours at 37°C. After this time, the inoculum was removed and replaced with fresh medium and samples were harvested as appropriate for each experiment. For experiments involving TGFβ stimulation, cells were infected as above for 48 hours followed by serum starvation overnight. The next day cells were treated with 100pg/mL TGFβ for 4 hours. Multi-step growth curves were performed in duplicate using NHDF, tHF or derivitives infected with 0.01 PFU/cell in DMEM containing 1% FBS and recombinant TGFβ (100pg/mL) was added immediately after infection and every 3 days thereafter. UV inactivation of HCMV was performed using the Spectrolinker XL-1000 (Spectronics Corporation) by exposing virus resuspended in 200 μl for 30 s at 600 μJ three times sequentially [71]. Sendai virus (SeV) was obtained from Charles River Laboratories and used at 160 hemagglutination units (HAU)/mL.

Transfections

NHDF or tHF cells seeded in 12 well plates were transfected with 40uM siRNAs (SMAD3 and IRF7; Thermo Fisher Scientific) or miRNA mimics (custom designed; IDT) per well using Lipofectamine RNAiMax (Thermo Fisher Scientific) according to the manufacturer’s instructions. 33ug/mL ISD90 (IDT) and 16.5ug/mL polyI:C (Invivogen) was transfected into cells using Lipofectamine 3000 according to the manufacturer’s instructions.

qRT-PCR

Total RNA was isolated from transfected or infected cells using the Trizol RNA isolation method. cDNA was prepared using 1000ng of total RNA and random hexamer primers. Samples were incubated at 16°C for 30 minutes, 42°C for 30 minutes and 85°C for 5 minutes. Real-time PCR (Taqman) was used to analyze cDNA levels in transfected or infected samples. An ABI StepOnePlus Real Time PCR machine was used with the following program for 40 cycles: 95°C for 15 sec and 60°C for one minute. Relative expression was determined using the ΔΔCt method using 18S as the standard control. JunB, SERPINE, SMAD3, IFNA1, IFNB1, RSAD2, IRF7 and 18S primer/probe sets were obtained from Thermo Fisher Scientific.

Immunoblotting

Protein extracts were run on an 8% SDS-PAGE, transferred to Immobilon-P Transfer Membranes (Milipore Corp., Bedford, MA), and visualized by chemiluminescence with specific antibodies: phosho-SMAD3 (Abcam), total SMAD3 (Abcam), IE86 (mAb 810; Millipore), IRF7 (Santa Cruz) and GAPDH (Abcam). Relative intensity of bands detected by western blotting was quantitated using ImageJ software.

ELISAs

Supernatants harvested from infected cells were centrifuged at maximum speed for 30sec to remove cell debris and stored at −80°C prior to cytokine measurements. IFNα was quantified using the pan-IFNα ELISA kit (Stem Cell Technologies). IFNβ was quantified using human IFN-beta Quantikine ELISA kit (R&D Systems). All measurements were made following the manufacturers’ protocols.

Construction of IRF7 KO fibroblasts

Genome editing using lentivirus-mediated delivery of CRISPR-Cas9 components was performed generally as described previously [27,28,34]. Breifly, a 20 nucleotide guide RNA (gRNA) sequence targeting the IRF7 protein-coding region was inserted into the lentiCRISPRv2 vector (Addgene; catalog #52961). The IRF7 guide sequences used was: TACACCTTGTGCGGGTCGGC. tHF cells stably transduced with the IFN-responsive pGreenFire-ISRE lentivector (System Biosciences) were further stably transduced with the lentiCRISPRv2 vector, selected using puromycin at 3ug/mL and IRF7 knockdown was confirmed by western blotting.

Statistical analysis

The Student’s two tailed t test (Microsoft Excel software) was used to determine p values. Results were considered significant at a probability (p) < 0.05.

Supporting information

S1 Fig. SMAD3 and IRF7 siRNA reduce target transcript expression in ΔSTING, ΔIFNAR and ΔIRF7 cell lines.

tHF, ΔSTING (A-B), ΔIFNAR (C-D) or ΔIRF7 (E-F) cell lines were transfected with negative control, miR-UL22A mimic or SMAD3 or IRF7 siRNA for 48 hours after which cells were additionally mock transfected or transfected with ISD90 for 16 hours. RNA was isolated and subjected to qRT-PCR for SMAD3 or IRF7. All experiments were performed in triplicate. * p<0.05 by two tailed Student’s t test.

https://doi.org/10.1371/journal.ppat.1009380.s001

(EPS)

Acknowledgments

The authors would like to acknowledge members of the Nelson and DeFilippis labs for helpful discussions.

References

  1. 1. Ashley CL, Abendroth A, McSharry BP, Slobedman B. Interferon-Independent Innate Responses to Cytomegalovirus. Front Immunol. 2019;10:2751. Epub 2020/01/11. pmid:31921100; PubMed Central PMCID: PMC6917592.
  2. 2. Goodwin CM, Ciesla JH, Munger J. Who’s Driving? Human Cytomegalovirus, Interferon, and NFkappaB Signaling. Viruses. 2018;10(9). Epub 2018/08/24. pmid:30134546; PubMed Central PMCID: PMC6163874.
  3. 3. Amsler L, Verweij M, DeFilippis VR. The tiers and dimensions of evasion of the type I interferon response by human cytomegalovirus. J Mol Biol. 2013;425(24):4857–71. Epub 2013/09/10. pmid:24013068; PubMed Central PMCID: PMC3864659.
  4. 4. Chinta P, Garcia EC, Tajuddin KH, Akhidenor N, Davis A, Faure L, et al. Control of Cytokines in Latent Cytomegalovirus Infection. Pathogens. 2020;9(10). Epub 2020/10/25. pmid:33096622; PubMed Central PMCID: PMC7589642.
  5. 5. Kwon YJ, Kim DJ, Kim JH, Park CG, Cha CY, Hwang ES. Human cytomegalovirus (HCMV) infection in osteosarcoma cell line suppresses GM-CSF production by induction of TGF-beta. Microbiol Immunol. 2004;48(3):195–9. Epub 2004/03/20. pmid:15031532.
  6. 6. Yoo YD, Chiou CJ, Choi KS, Yi Y, Michelson S, Kim S, et al. The IE2 regulatory protein of human cytomegalovirus induces expression of the human transforming growth factor beta1 gene through an Egr-1 binding site. J Virol. 1996;70(10):7062–70. Epub 1996/10/01. pmid:8794351; PubMed Central PMCID: PMC190757.
  7. 7. Michelson S, Alcami J, Kim SJ, Danielpour D, Bachelerie F, Picard L, et al. Human cytomegalovirus infection induces transcription and secretion of transforming growth factor beta 1. J Virol. 1994;68(9):5730–7. Epub 1994/09/01. pmid:8057454; PubMed Central PMCID: PMC236976.
  8. 8. Mason GM, Poole E, Sissons JG, Wills MR, Sinclair JH. Human cytomegalovirus latency alters the cellular secretome, inducing cluster of differentiation (CD)4+ T-cell migration and suppression of effector function. Proc Natl Acad Sci U S A. 2012;109(36):14538–43. Epub 2012/07/25. pmid:22826250; PubMed Central PMCID: PMC3437838.
  9. 9. Hancock MH, Crawford LB, Pham AH, Mitchell J, Struthers HM, Yurochko AD, et al. Human Cytomegalovirus miRNAs Regulate TGF-beta to Mediate Myelosuppression while Maintaining Viral Latency in CD34(+) Hematopoietic Progenitor Cells. Cell Host Microbe. 2020;27(1):104–14 e4. Epub 2019/12/24. pmid:31866424; PubMed Central PMCID: PMC6952548.
  10. 10. Kim JH, Collins-McMillen D, Buehler JC, Goodrum FD, Yurochko AD. Human Cytomegalovirus Requires Epidermal Growth Factor Receptor Signaling To Enter and Initiate the Early Steps in the Establishment of Latency in CD34(+) Human Progenitor Cells. J Virol. 2017;91(5). Epub 2016/12/16. pmid:27974567; PubMed Central PMCID: PMC5309964.
  11. 11. Batlle E, Massague J. Transforming Growth Factor-beta Signaling in Immunity and Cancer. Immunity. 2019;50(4):924–40. Epub 2019/04/18. pmid:30995507; PubMed Central PMCID: PMC7507121.
  12. 12. David CJ, Massague J. Contextual determinants of TGFbeta action in development, immunity and cancer. Nat Rev Mol Cell Biol. 2018;19(7):419–35. Epub 2018/04/13. pmid:29643418; PubMed Central PMCID: PMC7457231.
  13. 13. Shi Y, Wang YF, Jayaraman L, Yang H, Massague J, Pavletich NP. Crystal structure of a Smad MH1 domain bound to DNA: insights on DNA binding in TGF-beta signaling. Cell. 1998;94(5):585–94. Epub 1998/09/19. pmid:9741623.
  14. 14. BabuRajendran N, Palasingam P, Narasimhan K, Sun W, Prabhakar S, Jauch R, et al. Structure of Smad1 MH1/DNA complex reveals distinctive rearrangements of BMP and TGF-beta effectors. Nucleic Acids Res. 2010;38(10):3477–88. Epub 2010/02/12. pmid:20147459; PubMed Central PMCID: PMC2879523.
  15. 15. Martin-Malpartida P, Batet M, Kaczmarska Z, Freier R, Gomes T, Aragon E, et al. Structural basis for genome wide recognition of 5-bp GC motifs by SMAD transcription factors. Nature communications. 2017;8(1):2070. Epub 2017/12/14. pmid:29234012; PubMed Central PMCID: PMC5727232.
  16. 16. Feng XH, Zhang Y, Wu RY, Derynck R. The tumor suppressor Smad4/DPC4 and transcriptional adaptor CBP/p300 are coactivators for smad3 in TGF-beta-induced transcriptional activation. Genes Dev. 1998;12(14):2153–63. Epub 1998/07/25. pmid:9679060; PubMed Central PMCID: PMC317015.
  17. 17. Wotton D, Lo RS, Lee S, Massague J. A Smad transcriptional corepressor. Cell. 1999;97(1):29–39. Epub 1999/04/13. pmid:10199400.
  18. 18. Luo K. Ski and SnoN: negative regulators of TGF-beta signaling. Curr Opin Genet Dev. 2004;14(1):65–70. Epub 2004/04/28. pmid:15108807.
  19. 19. Suzuki HI. MicroRNA Control of TGF-beta Signaling. International journal of molecular sciences. 2018;19(7). Epub 2018/07/01. pmid:29958433; PubMed Central PMCID: PMC6073626.
  20. 20. Grey F, Tirabassi R, Meyers H, Wu G, McWeeney S, Hook L, et al. A viral microRNA down-regulates multiple cell cycle genes through mRNA 5’UTRs. PLoS Pathog. 2010;6(6):e1000967. Epub 2010/06/30. pmid:20585629; PubMed Central PMCID: PMC2891821.
  21. 21. Bartel DP. Metazoan MicroRNAs. Cell. 2018;173(1):20–51. Epub 2018/03/24. pmid:29570994; PubMed Central PMCID: PMC6091663.
  22. 22. Stark TJ, Arnold JD, Spector DH, Yeo GW. High-resolution profiling and analysis of viral and host small RNAs during human cytomegalovirus infection. J Virol. 2012;86(1):226–35. Epub 2011/10/21. pmid:22013051; PubMed Central PMCID: PMC3255895.
  23. 23. Haagmans BL, Teerds KJ, van den Eijnden-van Raaij AJ, Horzinek MC, Schijns VE. Transforming growth factor beta production during rat cytomegalovirus infection. J Gen Virol. 1997;78 (Pt 1):205–13. Epub 1997/01/01. pmid:9010305.
  24. 24. Helantera I, Loginov R, Koskinen P, Tornroth T, Gronhagen-Riska C, Lautenschlager I. Persistent cytomegalovirus infection is associated with increased expression of TGF-beta1, PDGF-AA and ICAM-1 and arterial intimal thickening in kidney allografts. Nephrol Dial Transplant. 2005;20(4):790–6. Epub 2005/02/18. pmid:15716293.
  25. 25. Qing J, Liu C, Choy L, Wu RY, Pagano JS, Derynck R. Transforming growth factor beta/Smad3 signaling regulates IRF-7 function and transcriptional activation of the beta interferon promoter. Mol Cell Biol. 2004;24(3):1411–25. Epub 2004/01/20. pmid:14729983; PubMed Central PMCID: PMC321430.
  26. 26. Abraham J, Botto S, Mizuno N, Pryke K, Gall B, Boehm D, et al. Characterization of a Novel Compound That Stimulates STING-Mediated Innate Immune Activity in an Allele-Specific Manner. Front Immunol. 2020;11:1430. Epub 2020/08/01. pmid:32733475; PubMed Central PMCID: PMC7360819.
  27. 27. Pryke KM, Abraham J, Sali TM, Gall BJ, Archer I, Liu A, et al. A Novel Agonist of the TRIF Pathway Induces a Cellular State Refractory to Replication of Zika, Chikungunya, and Dengue Viruses. mBio. 2017;8(3). Epub 2017/05/04. pmid:28465426; PubMed Central PMCID: PMC5414005.
  28. 28. Gall B, Pryke K, Abraham J, Mizuno N, Botto S, Sali TM, et al. Emerging Alphaviruses Are Sensitive to Cellular States Induced by a Novel Small-Molecule Agonist of the STING Pathway. J Virol. 2018;92(6). Epub 2017/12/22. pmid:29263267; PubMed Central PMCID: PMC5827377.
  29. 29. Paijo J, Doring M, Spanier J, Grabski E, Nooruzzaman M, Schmidt T, et al. cGAS Senses Human Cytomegalovirus and Induces Type I Interferon Responses in Human Monocyte-Derived Cells. PLoS Pathog. 2016;12(4):e1005546. Epub 2016/04/09. pmid:27058035; PubMed Central PMCID: PMC4825940.
  30. 30. Lio CW, McDonald B, Takahashi M, Dhanwani R, Sharma N, Huang J, et al. cGAS-STING Signaling Regulates Initial Innate Control of Cytomegalovirus Infection. J Virol. 2016;90(17):7789–97. Epub 2016/06/24. pmid:27334590; PubMed Central PMCID: PMC4988162.
  31. 31. 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. Epub 1999/01/07. pmid:9877175.
  32. 32. Sato M, Suemori H, Hata N, Asagiri M, Ogasawara K, Nakao K, et al. Distinct and essential roles of transcription factors IRF-3 and IRF-7 in response to viruses for IFN-alpha/beta gene induction. Immunity. 2000;13(4):539–48. Epub 2000/11/09. pmid:11070172.
  33. 33. 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. Epub 1998/11/21. pmid:9822609; PubMed Central PMCID: PMC1171011.
  34. 34. Sali TM, Pryke KM, Abraham J, Liu A, Archer I, Broeckel R, et al. Characterization of a Novel Human-Specific STING Agonist that Elicits Antiviral Activity Against Emerging Alphaviruses. PLoS Pathog. 2015;11(12):e1005324. Epub 2015/12/10. pmid:26646986; PubMed Central PMCID: PMC4672893.
  35. 35. Honda K, Yanai H, Negishi H, Asagiri M, Sato M, Mizutani T, et al. IRF-7 is the master regulator of type-I interferon-dependent immune responses. Nature. 2005;434(7034):772–7. Epub 2005/04/01. pmid:15800576.
  36. 36. Fitzgerald KA, McWhirter SM, Faia KL, Rowe DC, Latz E, Golenbock DT, et al. IKKepsilon and TBK1 are essential components of the IRF3 signaling pathway. Nat Immunol. 2003;4(5):491–6. Epub 2003/04/15. pmid:12692549.
  37. 37. Au WC, Moore PA, LaFleur DW, Tombal B, Pitha PM. Characterization of the interferon regulatory factor-7 and its potential role in the transcription activation of interferon A genes. J Biol Chem. 1998;273(44):29210–7. Epub 1998/10/24. pmid:9786932.
  38. 38. Yu Y, Wang SE, Hayward GS. The KSHV immediate-early transcription factor RTA encodes ubiquitin E3 ligase activity that targets IRF7 for proteosome-mediated degradation. Immunity. 2005;22(1):59–70. Epub 2005/01/25. pmid:15664159.
  39. 39. Zhu FX, King SM, Smith EJ, Levy DE, Yuan Y. A Kaposi’s sarcoma-associated herpesviral protein inhibits virus-mediated induction of type I interferon by blocking IRF-7 phosphorylation and nuclear accumulation. Proc Natl Acad Sci U S A. 2002;99(8):5573–8. Epub 2002/04/12. pmid:11943871; PubMed Central PMCID: PMC122811.
  40. 40. Zhu FX, Sathish N, Yuan Y. Antagonism of host antiviral responses by Kaposi’s sarcoma-associated herpesvirus tegument protein ORF45. PLoS One. 2010;5(5):e10573. Epub 2010/05/21. pmid:20485504; PubMed Central PMCID: PMC2868026.
  41. 41. Joo CH, Shin YC, Gack M, Wu L, Levy D, Jung JU. Inhibition of interferon regulatory factor 7 (IRF7)-mediated interferon signal transduction by the Kaposi’s sarcoma-associated herpesvirus viral IRF homolog vIRF3. J Virol. 2007;81(15):8282–92. Epub 2007/05/25. pmid:17522209; PubMed Central PMCID: PMC1951281.
  42. 42. Lubyova B, Pitha PM. Characterization of a novel human herpesvirus 8-encoded protein, vIRF-3, that shows homology to viral and cellular interferon regulatory factors. J Virol. 2000;74(17):8194–201. Epub 2000/08/10. pmid:10933732; PubMed Central PMCID: PMC112355.
  43. 43. Hwang SW, Kim D, Jung JU, Lee HR. KSHV-encoded viral interferon regulatory factor 4 (vIRF4) interacts with IRF7 and inhibits interferon alpha production. Biochem Biophys Res Commun. 2017;486(3):700–5. Epub 2017/03/28. pmid:28342865; PubMed Central PMCID: PMC5490377.
  44. 44. Shahnazaryan D, Khalil R, Wynne C, Jefferies CA, Ni Gabhann-Dromgoole J, Murphy CC. Herpes simplex virus 1 targets IRF7 via ICP0 to limit type I IFN induction. Sci Rep. 2020;10(1):22216. Epub 2020/12/19. pmid:33335135; PubMed Central PMCID: PMC7747705.
  45. 45. Lin R, Noyce RS, Collins SE, Everett RD, Mossman KL. The herpes simplex virus ICP0 RING finger domain inhibits IRF3- and IRF7-mediated activation of interferon-stimulated genes. J Virol. 2004;78(4):1675–84. Epub 2004/01/30. pmid:14747533; PubMed Central PMCID: PMC369457.
  46. 46. Zhang L, Pagano JS. Interferon regulatory factor 7 is induced by Epstein-Barr virus latent membrane protein 1. J Virol. 2000;74(3):1061–8. Epub 2000/01/11. pmid:10627515; PubMed Central PMCID: PMC111439.
  47. 47. Ning S, Campos AD, Darnay BG, Bentz GL, Pagano JS. TRAF6 and the three C-terminal lysine sites on IRF7 are required for its ubiquitination-mediated activation by the tumor necrosis factor receptor family member latent membrane protein 1. Mol Cell Biol. 2008;28(20):6536–46. Epub 2008/08/20. pmid:18710948; PubMed Central PMCID: PMC2577435.
  48. 48. Huye LE, Ning S, Kelliher M, Pagano JS. Interferon regulatory factor 7 is activated by a viral oncoprotein through RIP-dependent ubiquitination. Mol Cell Biol. 2007;27(8):2910–8. Epub 2007/02/14. pmid:17296724; PubMed Central PMCID: PMC1899925.
  49. 49. Zhang L, Pagano JS. Interferon regulatory factor 7 mediates activation of Tap-2 by Epstein-Barr virus latent membrane protein 1. J Virol. 2001;75(1):341–50. Epub 2000/12/19. pmid:11119603; PubMed Central PMCID: PMC113927.
  50. 50. Zhang L, Pagano JS. IRF-7, a new interferon regulatory factor associated with Epstein-Barr virus latency. Mol Cell Biol. 1997;17(10):5748–57. Epub 1997/10/07. pmid:9315633; PubMed Central PMCID: PMC232423.
  51. 51. Hahn AM, Huye LE, Ning S, Webster-Cyriaque J, Pagano JS. Interferon regulatory factor 7 is negatively regulated by the Epstein-Barr virus immediate-early gene, BZLF-1. J Virol. 2005;79(15):10040–52. Epub 2005/07/15. pmid:16014964; PubMed Central PMCID: PMC1181586.
  52. 52. Wu L, Fossum E, Joo CH, Inn KS, Shin YC, Johannsen E, et al. Epstein-Barr virus LF2: an antagonist to type I interferon. J Virol. 2009;83(2):1140–6. Epub 2008/11/07. pmid:18987133; PubMed Central PMCID: PMC2612359.
  53. 53. Bentz GL, Liu R, Hahn AM, Shackelford J, Pagano JS. Epstein-Barr virus BRLF1 inhibits transcription of IRF3 and IRF7 and suppresses induction of interferon-beta. Virology. 2010;402(1):121–8. Epub 2010/04/13. pmid:20381110; PubMed Central PMCID: PMC2871977.
  54. 54. Iempridee T, Das S, Xu I, Mertz JE. Transforming growth factor beta-induced reactivation of Epstein-Barr virus involves multiple Smad-binding elements cooperatively activating expression of the latent-lytic switch BZLF1 gene. J Virol. 2011;85(15):7836–48. Epub 2011/05/20. pmid:21593157; PubMed Central PMCID: PMC3147924.
  55. 55. Prokova V, Mosialos G, Kardassis D. Inhibition of transforming growth factor beta signaling and Smad-dependent activation of transcription by the Latent Membrane Protein 1 of Epstein-Barr virus. J Biol Chem. 2002;277(11):9342–50. Epub 2002/01/10. pmid:11781310.
  56. 56. Mori N, Morishita M, Tsukazaki T, Yamamoto N. Repression of Smad-dependent transforming growth factor-beta signaling by Epstein-Barr virus latent membrane protein 1 through nuclear factor-kappaB. Int J Cancer. 2003;105(5):661–8. Epub 2003/05/13. pmid:12740915.
  57. 57. Wood VH, O’Neil JD, Wei W, Stewart SE, Dawson CW, Young LS. Epstein-Barr virus-encoded EBNA1 regulates cellular gene transcription and modulates the STAT1 and TGFbeta signaling pathways. Oncogene. 2007;26(28):4135–47. Epub 2007/05/09. pmid:17486072.
  58. 58. Kim DH, Chang MS, Yoon CJ, Middeldorp JM, Martinez OM, Byeon SJ, et al. Epstein-Barr virus BARF1-induced NFkappaB/miR-146a/SMAD4 alterations in stomach cancer cells. Oncotarget. 2016;7(50):82213–27. Epub 2016/07/21. pmid:27438138; PubMed Central PMCID: PMC5347686.
  59. 59. Motsch N, Pfuhl T, Mrazek J, Barth S, Grasser FA. Epstein-Barr virus-encoded latent membrane protein 1 (LMP1) induces the expression of the cellular microRNA miR-146a. RNA Biol. 2007;4(3):131–7. Epub 2008/03/19. pmid:18347435.
  60. 60. Choi HS, Jain V, Krueger B, Marshall V, Kim CH, Shisler JL, et al. Kaposi’s Sarcoma-Associated Herpesvirus (KSHV) Induces the Oncogenic miR-17-92 Cluster and Down-Regulates TGF-beta Signaling. PLoS Pathog. 2015;11(11):e1005255. Epub 2015/11/07. pmid:26545119; PubMed Central PMCID: PMC4636184.
  61. 61. Di Bartolo DL, Cannon M, Liu YF, Renne R, Chadburn A, Boshoff C, et al. KSHV LANA inhibits TGF-beta signaling through epigenetic silencing of the TGF-beta type II receptor. Blood. 2008;111(9):4731–40. Epub 2008/01/18. pmid:18199825; PubMed Central PMCID: PMC2343603.
  62. 62. Seo T, Park J, Choe J. Kaposi’s sarcoma-associated herpesvirus viral IFN regulatory factor 1 inhibits transforming growth factor-beta signaling. Cancer Res. 2005;65(5):1738–47. Epub 2005/03/09. pmid:15753369.
  63. 63. Tomita M, Choe J, Tsukazaki T, Mori N. The Kaposi’s sarcoma-associated herpesvirus K-bZIP protein represses transforming growth factor beta signaling through interaction with CREB-binding protein. Oncogene. 2004;23(50):8272–81. Epub 2004/10/07. pmid:15467747.
  64. 64. Lei X, Zhu Y, Jones T, Bai Z, Huang Y, Gao SJ. A Kaposi’s sarcoma-associated herpesvirus microRNA and its variants target the transforming growth factor beta pathway to promote cell survival. J Virol. 2012;86(21):11698–711. Epub 2012/08/24. pmid:22915806; PubMed Central PMCID: PMC3486299.
  65. 65. Samols MA, Skalsky RL, Maldonado AM, Riva A, Lopez MC, Baker HV, et al. Identification of cellular genes targeted by KSHV-encoded microRNAs. PLoS Pathog. 2007;3(5):e65. Epub 2007/05/16. pmid:17500590; PubMed Central PMCID: PMC1876501.
  66. 66. Mendez-Samperio P, Hernandez M, Ayala HE. Induction of transforming growth factor-beta 1 production in human cells by herpes simplex virus. J Interferon Cytokine Res. 2000;20(3):273–80. Epub 2000/04/13. pmid:10762074.
  67. 67. Allen SJ, Mott KR, Wechsler SL, Flavell RA, Town T, Ghiasi H. Adaptive and innate transforming growth factor beta signaling impact herpes simplex virus 1 latency and reactivation. J Virol. 2011;85(21):11448–56. Epub 2011/09/02. pmid:21880769; PubMed Central PMCID: PMC3194985.
  68. 68. Shimamura M, Murphy-Ullrich JE, Britt WJ. Human cytomegalovirus induces TGF-beta1 activation in renal tubular epithelial cells after epithelial-to-mesenchymal transition. PLoS Pathog. 2010;6(11):e1001170. Epub 2010/11/17. pmid:21079788; PubMed Central PMCID: PMC2973835.
  69. 69. Lau B, Poole E, Krishna B, Sellart I, Wills MR, Murphy E, et al. The Expression of Human Cytomegalovirus MicroRNA MiR-UL148D during Latent Infection in Primary Myeloid Cells Inhibits Activin A-triggered Secretion of IL-6. Sci Rep. 2016;6:31205. Epub 2016/08/06. pmid:27491954; PubMed Central PMCID: PMC4974560.
  70. 70. Umashankar M, Petrucelli A, Cicchini L, Caposio P, Kreklywich CN, Rak M, et al. A novel human cytomegalovirus locus modulates cell type-specific outcomes of infection. PLoS Pathog. 2011;7(12):e1002444. Epub 2012/01/14. pmid:22241980; PubMed Central PMCID: PMC3248471.
  71. 71. Botto S, Abraham J, Mizuno N, Pryke K, Gall B, Landais I, et al. Human Cytomegalovirus Immediate Early 86-kDa Protein Blocks Transcription and Induces Degradation of the Immature Interleukin-1beta Protein during Virion-Mediated Activation of the AIM2 Inflammasome. mBio. 2019;10(1). Epub 2019/02/14. pmid:30755509; PubMed Central PMCID: PMC6372796.