Influenza A virus (IAV) triggers a contagious respiratory disease that produces considerable lethality. Although this lethality is likely due to an excessive host inflammatory response, the negative feedback mechanisms aimed at regulating such a response are unknown. In this study, we investigated the role of the eight “suppressor of cytokine signaling” (SOCS) regulatory proteins in IAV-triggered cytokine expression in human respiratory epithelial cells. SOCS1 to SOCS7, but not cytokine-inducible Src homology 2-containing protein (CIS), are constitutively expressed in these cells and only SOCS1 and SOCS3 expressions are up-regulated upon IAV challenge. Using distinct approaches affecting the expression and/or the function of the IFNαβ receptor (IFNAR)1, the viral sensors TLR3 and retinoic acid-inducible gene I (RIG-I) as well as the mitochondrial antiviral signaling protein (MAVS, a RIG-I signaling intermediate), we demonstrated that SOCS1 and SOCS3 up-regulation requires a TLR3-independent, RIG-I/MAVS/IFNAR1-dependent pathway. Importantly, by using vectors overexpressing SOCS1 and SOCS3 we revealed that while both molecules inhibit antiviral responses, they differentially modulate inflammatory signaling pathways.
Influenza A virus (IAV)3 is the etiological agent of a contagious acute respiratory disease. This virus is a major public health threat, killing >30,000 people annually in the United States of America alone, sickening millions, and inflicting substantial economic costs (1, 2). There is evidence that host immune responses contribute to the pathogenesis of seasonal human IAV infections (3). More importantly, aberrant and harmful host responses are believed to be critical in the pathogenesis of the highly pathogenic IAV strains, including the pandemic 1918 H1N1 virus and the more recent avian H5N1 virus (4, 5, 6, 7). Surprisingly, the host proteins involved in the negative feedback of IAV-triggered innate immune responses have yet to be characterized.
Bronchial epithelial cells are the primary targets of IAV and thus play an important role in the pathogenesis of this viral infection (8, 9). We and others have recently shown that IAV infection leads to the exposure in host cells of single-stranded genomic RNA bearing a 5′-triphosphate end that activates the RNA helicase retinoic acid-inducible gene-I (RIG-I) (10, 11, 12) and of the viral replicative intermediate double-stranded RNA (dsRNA) which activates TLR3 (12, 13) and critically contributes to viral pathology (14).
Although cytokines are required to control infection, it is now clear that their overproduction can lead to local and/or systemic pathology (15, 16). A number of mechanisms have been reported to down-regulate cytokine signaling to prevent the overaction of those mediators (17, 18). Among them, recent accumulating evidences suggest that the suppressors of cytokine signaling (SOCS) family is a pivotal negative regulator for cytokine signaling. This family consists of eight members: cytokine-inducible SH2 containing protein (CIS) and SOCS1–7 (17, 18).
Because cytokines appear to be important for IAV pathogenesis, there is an urgent need to identify the host mechanisms that regulate cytokine activities during this viral infection to allow for better intervention strategies. In this report, we demonstrate for the first time the critical role of SOCS1 and SOCS3 as negative regulators of IAV-mediated lung mucosal innate immune response.
Materials and Methods
Influenza A/Scotland/20/74 (H3N2) virus was prepared as indicated in Ref13 .
Short interfering RNA (siRNA)-mediated gene silencing
siRNA against human RIG-I (catalog no.M-012511-00), human IFNαβ receptor 1 (IFNAR1; catalog no.M-020209-00), human mitochondrial antiviral signaling protein (MAVS) (MAVS-1: 5′-UAGUUGAUCUCGCGGACGAdTdT-3′ (sense) and 5′-UCGUCCGCGAGAUCAACUAdTdT-3′ (antisense); MAVS-2: 5′-CCGUUUGCUGAAGACAAGAdTdT3′ (sense) and 5′-UCUUGUCUUCAGCAAACGGdTdT-3′ (antisense)) as well as a control siRNA (catalog no.D-001206-13) were obtained from Dharmacon Research. The human bronchial epithelial BEAS-2B cells were described previously (13). The cells were transiently transfected with siRNA at 100 nM or a mixture of MAVS-1 and MAVS-2 at 50 nM each, using 3 μl of TransIT-siQuest transfection reagent (Mirus Bio) per well according to the manufacturer’s instructions. Under these conditions, none of the siRNA used in this study transduced an IFN signaling pathway. Forty-eight hours after transfection the cells were infected with IAV at a multiplicity of infection (MOI) of 1.
Real-time quantitative (q)RT-PCR analysis
qRT-PCR was performed using TaqMan gene expression assays specific for human SOCS1–7 and CIS (reference information are available on request) and TaqMan universal PCR master mix purchased from Applied Biosystems according to the manufacturer’s instructions. qRT-PCR was also conducted using SYBR Green PCR master mix (Applied Biosystems) or QuantiTect SYBR Green PCR master mix (Qiagen), according to the manufacturer’s instructions. Sequences of RIG-I, MAVS, IFNAR1, and β-actin primers are available on request. Duplicate cycle threshold (CT) values were analyzed by using the comparative CT (ΔΔCT) method (Applied Biosystems). The relative amount of target mRNA (2 −ΔCT) was obtained by normalizing to the endogenous β-actin reference in all experiments.
BEAS-2B cells were treated as previously described (19). An equal amount of protein (7 μg) was fractionated by SDS-PAGE, further electrotransferred, and probed by immunoblotting using specific Abs against SOCS1 (3 μg/ml; Zymed Laboratories), SOCS3 (2 μg/ml; Acris Antibodies), and β-actin (diluted 1/10,000; Sigma-Aldrich). Bound Abs were detected using ECL+ immunoblotting detection system (Amersham Biosciences).
Transfection of pulmonary epithelial cells and reporter gene assays
Stably transfected pZero-hTLR3 or control BEAS-2B cells were described previously (12). Concerning the reporter gene studies, BEAS-2B cells were transfected using FuGENE 6 (Roche Diagnostics) according to the manufacturer’s instructions. Cells were transiently cotransfected with 100 ng of an NF-κB- (generously provided by Dr. A. Israel, Pasteur Institute, Paris, France), an IL-8 promoter- (a generous gift from Prof. N. Mukaida, Kanazawa University, Kanazawa, Japan), an IFN-β promoter-, or an IRF-3-luciferase reporter plasmid (generously provided by Dr. J. Hiscott, McGill University, Montreal, Canada), 50 ng of pRSV-β-galactosidase to control DNA uptake, and 100 ng of a vector encoding a functional form of either SOCS1 (generously provided by Pr. A. Duschl, University of Salzburg, Salzburg, Austria) or SOCS3 (a generous gift of Dr. K. Ghosal, National Institute of Immunology, New Delhi, India) or the respective control empty vector. After 24 h, cells were infected or not infected (i.e., mock) with IAV (MOI = 1). Then, cells were processed as previously reported (12). Results are expressed as relative luciferase units (RLU) normalized with β-galactosidase activity.
Statistical differences between SOCS levels in mock-treated and IAV-infected cells or in pZero-hTLR3 relative to control BEAS-2B cells infected with IAV were tested using a one-way ANOVA followed by a Fisher test, with a threshold of p < 0.05.
Results and Discussion
IAV infection induces SOCS1 and SOCS3 up-regulation in human lung epithelial cells.
Increasing evidence suggests that SOCS proteins are important regulators for cytokine signaling (17, 18). However, no information is available concerning the constitutive SOCS expression in lung epithelial cells as well as their regulation following infection with IAV. This prompted us to perform a qRT-PCR analysis of the eight SOCS proteins in unstimulated human bronchial epithelial BEAS-2B cells. Fig. 1 A shows that SOCS1–7 are constitutively expressed whereas CIS remains under the detection limit. SOCS5 is the least expressed SOCS. Interestingly, constitutive expression of SOCS is rather heterogeneous, with the highest expressions for SOCS3, SOCS6, and SOCS1 relative to SOCS5.
Because inhibition or enhancement of SOCS expression may play an important role in the regulation of innate immune responses of IAV-infected epithelial cells, we conducted a kinetic study of SOCS levels in mock-treated and IAV-infected BEAS-2B cells (Fig. 1,B). Cells were infected with an amount of virus previously shown to potently activate these cells (12, 13). As illustrated in Fig. 1,B, among all SOCS expressed in bronchial epithelial cells, only SOCS1 and SOCS3 are significantly up-regulated upon IAV infection (p < 0.0001). Increase in SOCS1 and SOCS3 mRNA expression was confirmed at the protein level (Fig. 1 C). CIS expression is not set off after IAV challenge (data not shown). Regardless, it is rather difficult at this stage to compare the expression and regulation profiles of SOCS molecules found here in respiratory epithelial cells to other cells types, although several studies did describe the regulated expression of some SOCS, mainly SOCS1 and SOCS3, in several tissues. For instance, an elevated expression of SOCS3 was observed in fibroblast cells expressing the hepatitis C virus core (20) or in an amnion cell line infected by the herpes simplex virus type 1 (21). Also, the immunosuppressive HIV 1 Nef protein induces SOCS1 and SOCS3 in B cells (22). However, our study is the first to our knowledge to give an extensive qualitative and quantitative expression pattern of the eight SOCS in a given cell type, especially in both a resting condition and after IAV challenge.
IAV-induced SOCS1 and SOCS3 up-regulation is TLR3-independent
We recently demonstrated that the host recognition of IAV and the subsequent cytokine responses require at least two major viral sensors, i.e., TLR3 and RIG-I (12, 13). To better understand the specific contribution of TLR3 vs RIG-I-dependent signaling in IAV-induced SOCS1 and SOCS3 up-regulation, we first examined the role of TLR3 using the previously described pZero-hTLR3 BEAS-2B cells that constitutively express a dominant negative, nonfunctional form of TLR3 (12). Interestingly, the kinetics of the SOCS1 and SOCS3 levels were statistically similar in both control and pZero-hTLR3 BEAS-2B cells infected with IAV (p > 0.05; Fig. 2 A), meaning that IAV induces SOCS1 and SOCS3 up-regulation in a TLR3-independent pathway.
A RIG-I/MAVS- and IFNAR1-dependent pathway mediates SOCS1 and SOCS3 up-regulation in IAV-infected bronchial epithelial cells.
The foregoing findings led us to further evaluate the role of RIG-I-dependent signaling in SOCS expression. To inhibit RIG-I signaling pathways, we knocked down RIG-I and its signaling intermediate MAVS by using specific siRNA. Fig. 2,B (left and central panels) indicates that the silencing condition of RIG-I and MAVS signaling was effective as assessed by qRT-PCR. Using these reagents, we noticeably reveal that SOCS1 and SOCS3 up-regulation by IAV is RIG-I and MAVS dependent (Fig. 2 C).
Many molecules can regulate SOCS expression, including hormones, colony-stimulating factors, and cytokines (17, 18). Among all mediators released by IAV-infected lung epithelial cells (12, 13), we focused on the central antiviral component type I IFN because it has a major impact on IAV pathogenesis and host immune response (23). We hypothesized that RIG-I-mediated expression of SOCS1 and SOCS3 could be secondary to a type I IFN positive feedback loop. To verify this assumption, we knocked down IFNAR1 expression by using specific siRNA. Using qRT-PCR, we first confirmed the effectiveness of our silencing reagents by showing that relative to control siRNA, IFNAR1 siRNA considerably inhibited IFNAR1 expression in both mock-treated and IAV-infected cells (Fig. 2,B, right panel). Importantly, IAV-induced SOCS1 and SOCS3 up-regulation was prevented by IFNAR1 silencing (Fig. 2 C). Altogether, our data reliably demonstrate that RIG-I/MAVS/IFNAR1-dependent antiviral signaling is a pivot of SOCS1 and SOCS3 up-regulation in IAV-infected bronchial epithelial cells. It is noteworthy that this regulatory pathway is likely a broad molecular mechanism that can occur in cells infected by other viral pathogens, e.g., Sendai virus, respiratory syncytial virus, vesicular stomatitis virus, Newcastle disease virus, and hepatitis C virus, all of which trigger a RIG-I-dependent type I IFN secretion (24, 25, 26).
SOCS1 and SOCS3 are critical negative feedback regulators of IAV-triggered innate immune response
As Fig. 1 shows that SOCS1 and SOCS3 are among the three most constitutively expressed SOCS in bronchial epithelial cells and are the only members of this family that are up-regulated by IAV, we speculated that SOCS1 and SOCS3 might play an active role in the modulation of IAV-induced innate immune response. We examined whether SOCS1 and SOCS3 could alter IRF-3- and/or NF-κB-dependent signaling by using specific luciferase reporter plasmids as well as vectors encoding functional SOCS1 or SOCS3 or the respective control vector. We also used plasmids reporting the activity of IFN-β and IL-8 promoters, two mediators with antiviral and proinflammatory effects, respectively. Fig. 3 indicates that increasing amounts of IAV (MOI = 0.1–1) strongly stimulate the activity of all gene reporters in BEAS-2B cells transfected with a control plasmid. Interestingly, IRF-3-dependent and IFN-β antiviral pathways are significantly reduced (>60%) in IAV-infected cells overexpressing SOCS1 (Fig. 3,A). By contrast, NF-κB- and IL-8-dependent proinflammatory pathways are ∼2-fold-higher in SOCS1-transfected cells than in control plasmid-transfected ones (Fig. 3,A). These opposite effects of SOCS1 on antiviral and proinflammatory responses are in fact consistent not only with the enhancement of type I IFN signaling in Socs1−/− mice infected with the Semliki Forest virus (27) but also with a recent study showing that SOCS1 overexpression in human keratinocytes stimulated with the synthetic dsRNA polyinosinic-polycytidylic acid enhances NF-κB activity, as revealed by the increased phosphorylation of IκBα, an upstream signaling intermediate (28). In striking contrast to the SOCS1 effect, NF-κB- and IL-8 proinflammatory activities are inhibited in cells transfected with SOCS3. Noteworthily, we found that SOCS3 overexpression similarly inhibits IFN-β expression, even though IRF-3-dependent signaling was not significantly reduced (Fig. 3,B). The expression of type I IFNs is strictly regulated by the activation of latent transcription factors, including NF-κB and IRF-3 (29). Thus, one could assume that the SOCS1 regulatory effect on IFN-β expression is a result of its activity on IRF-3, whereas the SOCS3 effect on the same gene is a consequence of NF-κB inhibition. The opposite effect of SOCS1 and SOCS3 on NF-κB signaling raises the question of the resultant NF-κB activity in IAV-infected bronchial epithelial cells. In fact, the different kinetics of SOCS1 and SOCS3 expression may help to answer this question. Indeed, whereas SOCS1 is highest at 13 h post-IAV infection and declines thereafter, SOCS3 increases until 18 h and only decreases after that time point (see Figs. 2,A and Fig. 1 B). As a result, this suggests a two-phase regulatory action of SOCS on NF-κB signaling triggered by IAV, starting by a potentiating effect of SOCS1 before an inhibitory action of SOCS3 as soon as the balance between SOCS1 and SOCS3 expression is at the advantage of the latter.
In summary, we establish for the first time that SOCS1 and SOCS3 are critical regulators of IAV-triggered innate immune responses through a RIG-I/MAVS/IFNAR1-dependent pathway. Despite the different transcriptional pathways targeted by each molecule, both SOCS act in concert to inhibit type I IFN antiviral signaling. This feedback loop might depend on three mechanisms (17, 18, 30): 1) a modulation of JAK activity; 2) a competition with specific STAT transcription factors for binding sites on IFNAR1; or 3) an ubiquitination and subsequent degradation by the ubiquitin-proteasome pathway of a complex including a SOCS molecule and a targeted TLR3 and/or RIG-I signaling component. Which mechanism or mechanisms run in IAV-infected bronchial epithelial cells is currently investigated. Importantly, our results may contribute to the design of molecules targeting SOCS1 and SOCS3 to restore appropriate innate immune defense (i.e., increasing the beneficial antiviral activity while decreasing the deleterious proinflammatory aspect) that ultimately clear IAV and lead to a favorable clinical outcome.
We are grateful to Prof. A. Duschl (University of Salzburg, Salzburg, Austria) and Dr. K. Ghosal (National Institute of Immunology, New Delhi, India) for providing the SOCS1 and SOCS3 expression vectors, respectively.
The authors have no financial conflict of interest.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
This study was partially funded by the Institut Pasteur through “Programme Transversal de Recherche” (PTR no.186). J.P. was financially supported by the French association “Vaincre la mucoviscidose”.
Abbreviations used in this paper: IAV, influenza A virus; CIS, cytokine-inducible Src homology 2-containing protein; IFNAR1, IFNαβ receptor 1; IRF-3, IFN regulatory factor 3; MAVS, mitochondrial antiviral signaling protein; MOI, multiplicity of infection; qRT-PCR, quantitative RT-PCR; RIG-I, retinoic acid-inducible gene-I; siRNA, short interfering RNA; SOCS, suppressor of cytokine signaling protein.