STAT Activation during Viral Infection In Vivo: Where's the Interferon?

Abstract

In this issue of Cell Host and Microbe, O'Gorman et al. identify a key role for early TLR2-mediated IL-6 production and STAT3 activation in generating protective immunity against poxviruses. These findings highlight the importance of early inflammatory cytokine production in antiviral defense and have implications for enhancing vaccination efficacy.

Cytokines and interferons (IFNs) play an important role in host defense against viruses. In the current paradigm (Fig. 1A), IFNs and other cytokines contribute to induction of an antiviral state and coordinate antiviral immunity. Initial viral infection activates various conserved and broadly expressed recognition receptors including TLRs, RLRs and DNA sensors (Takeuchi and Akira, 2010). Viral recognition leads to production of type I IFNs (IFN-α and IFN-β) and various cytokines. Type I IFNs play a key role in the earliest stages of host defense by inducing expression of antiviral genes that confer a cell autonomous antiviral state that suppresses viral replication and makes uninfected cells refractory to infection. Virus-induced type I and type II IFNs, along with inflammatory cytokines, coordinate the innate immune response to viruses during the first several days of infection and subsequently play a key role in orchestrating a specific acquired immune response, which involves T and B cell differentiation and induction of protective cytotoxic lymphocytes and neutralizing antibodies. Prior to the current study (O'Gorman, 2010), the role of cytokines other than IFNs induced during the early phase of anti-viral immunity (typically 1 hour to several days after viral infection) was not well understood.

Figure 1. Role of cytokines and STATs in antiviral responses.

(A) Viral infection of cells leads to recognition and signaling by endosomal and cytoplasmic receptors such as TLRs and RLRs, with downstream induction of IFN and cytokine production. Type I IFNs activate their cognate IFNAR receptor in an autocrine, paracrine and possibly systemic manner. IFNAR signaling leads to activation of the STAT1-STAT2-IRF9-containing ISGF3 complex that binds to ISRE DNA elements and induces expression of IFN target genes that are important for an antiviral state and innate amd acquired immune responses. The role of other cytokines induced in the earliest stages of viral infection was not well appreciated.

(B) IFNs were previously shown to play an important role in host defense against vaccinia virus. The current study shows that vaccinia also activates TLR2, which typically resides in plasma membranes, leading to production of IL-6 and downstream STAT3 activation. TLR2 and early IL-6-STAT3 signaling are important for generating protective antibody-mediated immunity against vaccinia virus in mice.

Measurements of IFN/cytokine production and genetic approaches utilizing deletion of cytokine and IFN genes have taught us much about the role of these molecules in antiviral responses. However, these approaches do not yield insight into the key questions of which cells respond to cytokines in vivo, the time course of cytokine action, the functional importance of cytokine signaling in particular cells at specific time points, and whether cellular responses to cytokines are regulated in parallel with the well established regulation of cytokine production. These important questions are addressed for mouse poxvirus infection in the current study using a flow cytometric approach developed by Gary Nolan's lab to study cytokine-mediated activation of Jak-STAT signaling at the single cell level in vivo (Krutzik et al., 2005). In this approach, splenic cells are isolated after viral infection of mice and STAT tyrosine phosphorylation, a marker of activation, is quantitated in various cell types, including DCs, T cells, B cells and granulocytes. This validated method allows an unprecedented ability to follow cellular responses to cytokines in vivo.

Understanding the determinants of host defense to poxviruses has important clinical implications. The human smallpox virus (variola) does not induce effective immune responses and represents one of the most dangerous human pathogens; similarly ectromelia (mousepox) does not induce protective immunity in certain mouse strains. In contrast, humans and mice mount effective immune responses against the vaccinia poxvirus, leading to protective immunity that is mainly mediated by neutralizing antibodies. The effective immune response against vaccinia and the cross-protection it offers against smallpox has been exploited to achieve one of the major advances of modern medicine, the eradication of smallpox by immunization with vaccinia. Thus, understanding the mechanistic basis for differential immune responses against vaccinia vs. smallpox and ectromelia can lead to approaches to strengthen host defense against poxviruses and to design improved vaccination strategies.

TLR2 and IL-6/STAT3

In their study, O'Gorman et al. first examined STAT tyrosine phosphorylation in splenic T cells, B cells, DCs and granulocytes in vivo 1 hour after intravenous (IV) injection of various TLR ligands. Strikingly, IV TLR ligand injection induced very rapid (readily detectable after 1 hour) tyrosine phosphorylation of predominantly STAT1 and STAT3 in more than half of splenic DCs and CD4+ and CD8+ T cells, with lesser activation of STAT4, STAT5 and STAT6 at this time point, and substantially less STAT activation in B cells and granulocytes. Overall, STAT1 activation was primarily dependent on type I IFNs, while STAT3 activation was primarily dependent on IL-6. These results establish that cytokine responses occur very rapidly in vivo, but only in certain cell populations; it is not clear why B cells and granulocytes that express receptors for IFNs and IL-6 did not manifest a detectable cytokine response. Another important point to note is that the early burst of splenic cytokine production in response to IV TLR ligand injection is substantial enough to activate STAT signaling in the majority of DCs and T cells. Thus, splenic T cells are broadly exposed to a pulse of Jak-STAT signaling independent of their antigen specificity; this signaling pulse may play a role in conditioning T cells for subsequent antigen-specific responses (discussed below).

Similar to TLR ligands, intravenous injection of the immunizing vaccinia virus induced STAT3 tyrosine phosphorylation (pSTAT3) in splenic DCs and T cells, with an early peak 1 hour after injection and a lower amplitude second phase of activation observed at 7 hours. In contrast to TLR ligands, vaccinia induced unexpectedly low pSTAT1, and this was more dependent on IFN-γ than type I IFNs. The early induction of pSTAT3 was dependent on IL-6, consistent with the results obtained using TLR ligands. Surprisingly, induction of IL-6 and activation of STAT3 by vaccinia were dependent on TLR2, which previously has been implicated mainly in recognition of bacterial lipopetides. An important role for TLR2 and IL-6 in the control of vaccinia infection was corroborated by experiments showing increased viral burden and decreased clearance associated with decreased serum neutralizing antibodies in TLR2- or IL-6-deficient mice. Further support for the importance of IL-6 and STAT3 activation in the control of poxvirus infection was provided by experiments showing that lethal infection by the pathogenic ectromelia in Balb/C mice was associated with minimal STAT3 activation.

Collectively, the above described experiments establish two unexpected and important points for host defense against poxviruses that extend our understanding of antiviral immunity (Fig. 1B). First is the importance of rapid IL-6 and thus inflammatory cytokine production for successful host defense against poxviruses. Second is the key role of TLR2 in antiviral immunity. An emerging and unexpected role for TLR2 in detection of viruses (previously thought to be accomplished by nucleic acid-specific receptors) is supported by a recent report showing that monocytic TLR2 mediates type I IFN production in response to viral stimulation (Barbalat et al., 2009). Viral ligands that are recognized by TLR2 are not known, and recognition may occur via TLR2-associated coreceptors such as integrins or scavenger receptors.

Another important issue is whether and how an early burst of STAT3 signaling that peaks 1 hour post infection can contribute to effective antiviral immunity. Direct testing of the importance of early signaling by selectively ablating only the first hour of IL-6-STAT3 signaling can not be achieved using currently available experimental approaches. Instead, the authors used a gain of function approach, and, remarkably, showed that just one injection of IL-6 at the time of vaccinia infection can rescue antiviral immunity in TLR2-deficient mice. Even more impressive is that one injection of theTLR2 ligand PAM3CSK4 at the time of ectromelia infection was more effective at reducing viral burden than therapeutic vaccination. Surprisingly, the early burst of TLR2-mediated IL-6-STAT3 signaling did not affect the innate response to vaccinia, but instead was important for subsequent generation of neutralizing antibodies. It is possible that an early wave of IL-6 signaling can condition later stages of T cell differentiation and helper function, possibly through the induction of IL-21 or other cytokines (Dienz et al., 2009). Overall, these results potentially have profound significance for designing approaches to enhance the efficacy of antiviral vaccines by co-administration of cytokines or TLR ligands.

Where's the IFN Response?

An important point to consider is related to ‘the dog that did not bark’ – the authors did not detect a type I IFN response (as assessed by STAT1 activation) in the cells examined, despite confirming the importance of type I IFNs for host defense against poxviruses. One potential explanation is that type I IFN signaling was below the threshold of detection of the flow cytometric assay that was used or occurred only at later time points. This issue could be addressed by extending the analysis to include additional cell types, later time points, analysis of pSTAT2 (preferentially induced by type I IFNs) and perhaps most importantly by measuring expression of IFN-inducible genes. Type I IFNs can robustly induce gene expression at low or even undetectable levels of signaling, and analysis of IFN target gene expression (‘IFN signature’) in vivo can provide an important approach complementary to in vivo signaling studies, as can in vivo imaging approaches. A more interesting explanation for low pSTAT1 in these experiments is that IFN signaling is regulated in vivo, either by virus-encoded products (as partially investigated in the current study) or by host cell-expressed viral recognition receptors or other immunoreceptors that cross-regulate IFN receptor signaling (Ivashkiv, 2008; Wang et al., 2010). Such regulation of IFN signaling during viral or bacterial infection in vivo has been previously described (Hotson et al., 2009; Nguyen et al., 2002). Suppression of IFN signaling by viral products would represent a pathogenic mechanism to evade antiviral responses; modulation of IFN signaling by host cell receptors would represent a fine-tuning of IFN responses (Fig. 1A) to maintain host defense while avoiding toxicity associated with high IFN expression. Thus, the lack of detection of an IFN response represents both a limitation and strength of the experimental approach used. The limitation is that important host defense mechanisms may not be detected; the strength, that this method allows investigation of the regulation of cytokine signaling in vivo.

Conclusions

The current study used in vivo signaling analysis to discover an important role for TLR2, IL-6 and STAT3 signaling in immunity against poxviruses. This work opens new areas of investigation in early cytokine production and STAT signaling in innate and acquired antiviral immune responses, and supports the development of vaccine adjuvants that boost early cytokine action in order to enhance vaccine efficacy.