Cell Defense Against Viral Offense
The use of genetically modified mice deficient for the type I IFNAR or components of the IFN signaling pathway, such as STAT1, clearly establish the importance of type I IFN in the resistance to viral infection in vivo. Both IFN-a/p and STAT1 knockout mice are highly susceptible to viral infection and unable to establish an antiviral state. Similarly, the availability of genetically manipulated mice lacking either individual TLR receptors or cytoplasmic receptors has advanced the understanding of the cellular recognition of invading pathogens. However, it is still not completely clear what determines the specificity of the recognition and whether the receptor recognizes both the viral genome and replication intermediate. As shown with HSV-1, which is an effective IFN inducer, recognition may be complex. The unmethylated HSV-1 DNA genome is a very effective inducer of IFNa in human pDC, and this induction is dependent on TLR9 (Lund et al. 2003), while in human PBMCs, HSV-1 glycoprotein D alone can induce synthesis of IFNa. Another virus of the herpes group, mouse cytomegalovirus (MCMV), induces type I IFN through recognition of both TLR9 and TLR3 (Krug et al. 2004), but the recognition by CMV is through TLR2 (Compton et al. 2003).
dsRNA has been long considered the recognition entity for viral infection, and in the majority of the cells, RIG-I and MDA5 are important for the recognition of most RNA virus infections. However, not all viruses generate a significant amount of dsRNA intermediates and still are recognized by RIG-I (Pichlmair et al. 2006). The RIG-I pathway is induced by most of the viruses tested, whereas MDA5 is required for the response against picornaviruses (Kato et al. 2006). Recent observations indicate that this distinction is based on the specific recognition of the 5'-triphosphate viral RNA structure (Hornung et al. 2006). The uncapped 5'-triphosphate end of ssRNA and positive-strand RNA viruses or their transcripts is recognized by Rig I (Kato et al. 2006; Hornung et al. 2006). In contrast, the 5' end of picornavirus transcripts remains associated with the protein used as a primer and therefore uncapped 5' RNA is absent during picornavirus replication (Lee et al. 1977). RIG-I also plays a major role in the induction of the antiviral response to the hepatitis C virus (HCV) RNA
replicative intermediate in cultured hepatocytes (Sumpter et al. 2005) (see the chapter by Loo and Gale, this volume).
There are several indications that in the central nervous system (CNS), where TLR3 is expressed at high levels, the inflammatory response is initiated by TLR3. For instance, the inflammatory response initiated by TMEV and West Nile virus is dependent on TLR3 (Wang et al. 2004). The antiviral response to the ssRNA genomes of VSV and influenza virus in pDC was shown to depend on TLR7 (Lund et al. 2003) and there are some indications that the viral envelope can also be recognized by TLR4, which is a primary receptor for lipopolysaccharides (LPS). A TLR4-mediated antiviral response was induced by the fusion protein of RSV (Kurt-Jones et al. 2000) and by the envelope protein of MMTV (Burzyn et al. 2004). Altogether, these data indicate that both viral nucleic acids and glycoproteins are capable of generating the antiviral response and that the multiple patterns of recognition may enhance the antiviral response and its duration. However, in order to be able to replicate and establish infection, viruses develop various strategies for evading the innate immune response of the host, as is discussed by Haller in this volume.
Type I IFNs are not only essential for antiviral defense, but they also exert a number of immunoregulatory effects. They modulate the expression of major histocompatibility complex (MHC) antigens, and it is via this mechanism that IFN-a/ß increases susceptibility of vaccinia virus (VV), or lymphocytic choriomeningitis virus (LCMV) -infected fibroblasts to lysis by cytotoxic T lymphocytes (CTL) (Bukowski and Welsh 1985). IFNa/ß can also downregulate expression of IL-12 in human dendritic cells (DCs) and monocytes (Karp et al. 2000), stimulate expression of IFN-y in response to influenza virus infection (Sareneva et al. 1998), and induce expression of IL-15 (Durbin et al. 2000). IFNa has also multiple effects on the function of immune cells (Garcia-Sastre and Biron 2006) including enhancement of NK cell activity (Biron et al. 1999), activation of CD8+ T cells during the early steps of infection (Zhang et al. 1998), and protection of CD8+ T cells from antigen-induced cell death (Marrack et al. 1999). Human IFNa promotes the differentiation of dendritic cells (Santini et al. 2000), the upregulation of IFN-y expression, and stimulation of B cell differentiation in both a DC-dependent and -independent manner (Biron 2001; Santini et al. 2000). Recent data demonstrate that type I IFN can directly stimulate the B cell response during the early stages of influenza virus infection (Coro et al. 2006). IFNa has been also shown to induce the differentiation of human monocyte-derived DCs, which are able to induce Th1 polarization both in vitro and in vivo (Cella et al. 2000; Santini et al. 2000) and stimulate B cell proliferation and Ig class switching (Le Bon and Tough 2002). Thus, while in the past the IFN system was considered as only a part of the host innate immunity, recent data indicate that type I IFN has an important role in bridging innate and acquired immunity (Biron 2001).
Positive and Negative Role of Type I IFN in Bacterial Infection
Although synthesis of type I IFN was originally associated with viral infection, the production of type I IFN is also induced as an immediate innate response to bacterial infection, where interferon has been shown to modulate an innate antibacterial response (Bogdan et al. 2004). Like viruses, bacteria can be recognized by membrane-bound receptors and cytoplasmic receptors. The binding of a specific ligand to a given TLR recruits specific adaptors and initiates cellular signaling pathways, leading to the activation of the IRFs and the NFk B family of transcription factors. LPS present on Gram-negative bacteria is recognized by TLR4 and initiates the association of TLR4, either with TRIF and TRAM adaptor proteins, leading to the activation of IRF-3 and IRF-7 or with MyD88 adaptor, which leads to NFk B activation. Several Gram-negative bacteria, such as Salmonella typhimurium, Shi-gella flexneri and Escherichia spp., stimulate type I IFN synthesis after the invasion of the cell (Bogdan et al. 2004). The unmethylated bacterial DNA is recognized by endosomally expressed TLR9. Binding of dsDNA to TLR9 occurs in the endosomal compartment and results in the activation of IRF-5 and IRF-7. However, there is also cytoplasmic recognition of B-form DNA, which occurs in the cytosol and results in the activation of IRF-3. Thus an intracellular Gram-positive bacterium that has a cytoplasmic life cycle phase, such as Listeria monocytogenes, probably activates the IFN response via the cytosolic DNA recognition pathway (O'Connell et al. 2005; Stetson and Medzhitov 2006). Bacterial flagellin is a TLR5 ligand, which mediates signaling through MyD88, resulting in the activation of NFkB factors and the induction of inflammatory cytokines (Hayashi et al. 2001). Whether TLR5 also activates IRFs and type I IFN has not been yet determined.
Type I IFN can also modulate the outcome of bacterial infection. It is important to realize, however, that interferon can be both protective and detrimental to the host. Type I IFN inhibits intracellular replication of Legionella pneumophila (Schiavoni et al. 2004) and contributes to the clearance of pathogens in Leishmania infection (Diefenbach et al. 1998). Type I IFN also increases resistance against Gram-positive bacteria such as Streptococcus pneu-moniae and Bacillus anthracis (Gold et al. 2004; Weigent et al. 1986). In contrast, during Listeria infection, type I IFN synthesis increases the susceptibility of lymphocytes to infection (Carrero et al. 2006). IFN treatment also reduces host resistance to L. monocytogenes infection and has a negative impact on the survival of infected mice (Auerbuch et al. 2004).
Taken together, these data indicate that the effect of type I IFN on bacterial infection is complex. On one hand, it contributes to the clearance of pathogens; on the other, it can have harmful effects on the host.
Super IFN Producers: pDC
Although type I IFN can be produced essentially by any infected cell, most infected cells produce low levels of IFN that can act in an autocrine manner, or protect only cells localized in close proximity to the focus of infection. In human PBMCs, there is a rare type of cells, designated as natural interferon-producing cells, that produce very high levels of IFNa in response to viral infection and therefore can generate a systemic response (Fitzgeral-Bocarsly et al. 1988). Further characterization of these cells revealed that these cells are a CD123 and CD4+CD11c+Lin- subset of DCs referred to as plasmacytoid DCs (Siegal et al. 1999). Later, a pDC subset was also identified in mice; however, murine pDCs do not express CD123, but can be defined as CD11b-CD11c low B220+ cells that also express Ly6C (Colonna et al. 2004). PDCs differ from the monocyte derived DCs (mDCs), not only by their phenotype but also by their migration pattern (Penna et al. 2002). PDCs are recruited to the site of inflammation, where they are activated, while immature mDCs in peripheral tissues migrate after maturation to lymphatic tissues (Jahnsen et al. 2002). mDCs and pDCs express a distinct set of TLRs and therefore recognize different pathogens. TLRs expressed in pDCs are those associated with recognition of viral or bacterial DNA and viral RNA, namely TLR7/8 and TLR9. The induction of the antiviral response is dependent on the co-adaptor MyD88 (Fig. 2). mDCs express relatively high levels of TLR3 and low levels of TLR4, the induction of the IFN response is through the adaptor TRIF, and it is MyD88-independent (see the chapter by Severa and Fitzgerald, this volume).
The IFNa subtypes and their relative level of expression induced in pDCs appear to be virus specific. While HSV-1 induced approximately 10- to 100fold higher levels of IFNa in pDCs than in mDCs, the difference in the relative levels of IFNa induced by Sendai virus in pDCs and mDCs was much smaller. Since HSV-1 is recognized by TLR9 in pDCs, but by RIG-I in other cell types (Melchjorsen et al. 2005), the above observation indicates that the antiviral response induced by TLR9 is much stronger. Furthermore, the subtypes of IFNa induced in Sendai virus and HSV-1-infected cells were distinct (Iza-guirre et al. 2003). The difference in the profile of IFNa subtypes expressed in pDC upon stimulation of TLR9 and TLR7 has not been yet determined. Both human and mouse pDCs also express high levels of IRF-8. While IRF-8 plays a critical role in pDC development its role in the activation of type I IFN is not yet clear (Tamura et al. 2005).
Dendritic cells, fibroblasts
Plasmacytoid dendritic cells dsRNA
Fig. 2 The distinct difference in the induction of the antiviral response in pDCs and other cell types. In fibroblast and conventional DCs, ds viral RNA and viral transcripts are recognized by cytoplasmic RNA helicase RIG I (or MDA5) or TLR3. The TLR3 and RIG I pathways are mediated by cofactor TRIF or MAV, respectively. Both of these pathways activate TBK1 and IKKe and consequently IRF-3 and IRF-7. In pDCs, the antiviral pathway is mediated either by TLR7, which recognizes ss viral RNA, or TLR9 recognizing the unmethylated viral DNA. The activation of the respective TLRs leads to an assembly of multicomponent complex containing MyD88, IRF-7, IRAK 1, and TRAF-6 and activation of con-stitutively expressed IRF-7 and IRF-5
Several factors may contribute to the high production of IFNa in pDCs. In cells that constitutively express only IRF-3, an autocrine IFN0 feedback is required for an efficient production of IFNa (Marie et al. 1998; Prakash et al. 2005). However, in pDCs, which express relatively high levels of IRF-7, this autocrine feedback is not required (Dai et al. 2004; Izaguirre et al. 2003). The degradation of IRF-7 also seems to be attenuated in pDCs (Prakash et al. 2005). However, it has recently been suggested that the main reason for the high IFN production in pDCs is the distinct intracellular localization of TLR and TLR/ MyD88 complexes in pDCs and other cell types. While in cells other than pDCs the TLR9-MyD88-TLR7 complex is rapidly translocated to lysosomes and degraded, in pDCs it is retained in the endosomal compartment for a longer period of time (Honda et al. 2005).
The specific impact of pDCs on innate and acquired immunity in vivo is virus-dependent. In MCMV or VSV infection, pDCs are the major producer of type I IFN (Dalod et al. 2003). In contrast, pDCs do not contribute to type I IFN synthesis in mice infected with LCMV or West Nile virus (Colonna et al. 2004; Dalod et al. 2003), and the cells producing type I IFN in either one of these viral infections have not yet been identified. Compared to other type I IFN-producing cells, pDCs have two unique functions: they can rapidly produce high levels of type I IFN and the induction of the antiviral response does not require direct viral infection. Since pDCs can respond to noninfectious viral particles or viral nucleic acid, the induction of the antiviral response in these cells is not subjected to viral mimicry (Hengel et al. 2005). The downside of this property is that the ability of pDCs to respond to exogenous nucleic acids or nucleic acid-protein complexes can result in the unregulated production of IFNa and inflammatory cytokines, such as that associated with autoimmune and inflammatory diseases (see the chapter by Crow, this volume).
Was this article helpful?