Interfering with interferons: Hepatitis C virus counters innate immunity

Hepatitis C virus (HCV) has evolved highly successful mechanisms for evading host immune responses. Approximately 3% of the global population is infected with HCV, and it is estimated that only one of five newly infected individuals will mount an immune response sufficient to clear infection (1). Chronic HCV infection leads to degenerative liver disease culminating in cirrhosis or hepatocellular carcinoma (2). Failure of host immunity is at least partially due to the ability of HCV to antagonize both the production of and signaling through type I IFNs. Host cells characteristically initiate IFN synthesis by recognizing a pathogen-associated molecular pattern (PAMP). In the case of viral infection, dsRNA is often produced as a byproduct of virus replication and can be detected by at least two distinct PAMP receptors. Extracellular dsRNA recognition occurs through Toll-like receptor 3 (TLR3), whereas intracellular dsRNA is detected by retinoic-acid inducible gene I (RIG-I) (3, 4). However, these responses fail during HCV infection. Until now, the precise mechanisms by which HCV thwarts the immune response have remained a mystery. In this issue of PNAS, Li et al. (5) demonstrate that the NS3/4A protease of HCV eliminates the IFN response by targeting a newly identified antiviral protein, mitochondrial antiviral signaling protein (MAVS; also named VISA, IPS-1, and CARDIF), which functions downstream of RIG-I, and thus explain one means by which HCV can establish a chronic infection (5).

Type I IFNs, which consist of IFN-α and -β, provide a “front line” of defense against invading viruses. Recently, progress has been made in understanding the events leading to type I IFN synthesis. RIG-I is a cytosolic protein that detects intracellular dsRNA by a C-terminal RNA helicase domain. Upon dsRNA binding, RIG-I presumably undergoes a conformational change that allows association and activation of an adaptor protein, the recently described MAVS protein, by means of “caspase activation and recruitment domains” (CARDs) present in both proteins (6–9). Surprisingly, MAVS resides at the mitochondria, and this localization is required for downstream signaling. MAVS activation causes phosphorylation and activation of IRF3 by one of two kinases, TBK1 or IKKε kinase. This process is essential for IRF3 to homodimerize and translocate from the cytoplasm to the nucleus. Activated IRF3 induces IFN-β transcription by cooperating with NF-κB, another transcription factor activated downstream of MAVS (Fig. 1). Additionally, IFN-β may be induced independently of this pathway through TLR3. Extracellular dsRNA, presumably originating from lysis of virally infected cells, binds to TLR3, which then signals through the Toll–IL-1 receptor adaptor protein TRIF (10). This signaling again leads to IRF3 and NF-κB stimulating IFN-β synthesis and involves RIP1, TBK1, and TRAF6. Whether MAVS is required downstream of TLR3 is unresolved (11, 12).

Fig. 1.

HCV protein NS3/4A eliminates antiviral signaling by cleavage of MAVS and TRIF. Production of dsRNA by HCV during infection should result in synthesis of IFN-β, which in turn leads to production of an array of antiviral proteins known as ISGs. PAMP receptors RIG-I and TLR3 signal through kinases IKKε and TBK1 to activate IRF3 and cause the phosphorylation and degradation of IκB, which allows nuclear translocation of NF-κB. HCV overcomes this response and establishes chronic infection through cleavage of MAVS and TRIF.

After PAMP receptors sound the alarm and announce the presence of a pathogen, how do type I IFNs counteract virus infection? Essentially all nucleated cells can produce type I IFNs that act in a paracrine and autocrine manner through specific IFN-α/β receptors. This process triggers the production of hundreds of IFN-stimulated genes (ISG) through activation of Janus kinase-signal transducers and activators of transcription (Jak-STAT). Several of the ISG proteins have been well characterized and directly inhibit critical steps in viral replication, although the antiviral functions of other ISGs remain unknown (13). In this way, IFN-α and -β create an inhospitable environment for replication in infected cells as well as in proximate cells that may soon become infected. Besides antagonizing replication, IFN-α and -β have an important influence on the impending adaptive immune response (14). Given these effects, it is not surprising that viruses have evolved numerous mechanisms for evading the IFN response.

HCV is remarkably proficient at antagonizing the type I IFN pathway (reviewed in ref. 15). Several viral proteins not only inhibit IFN production but also interfere with signaling through IFN-α/β receptors. NS5A reduces ISG expression, and Jak-STAT signaling downstream of the IFN-α/β receptor is inhibited by the HCV core protein. IFN production as a result of dsRNA binding by TLR3 is blocked by HCV protease NS3/4A, which cleaves the TLR3 adaptor TRIF between amino acids 372 and 373 (16) (Fig. 1). However, it was not previously known how HCV suppressed responses from intracellular dsRNA. Li et al. (5) have now answered this question by demonstrating that NS3/4A cleaves MAVS at Cys-508. After this cleavage event, MAVS diffuses away from the mitochondria and cannot activate IRF3 and NF-κB. The findings of Li et al. (5) corroborate and extend the findings recently published by Meylan et al. (7), which also describe cleavage of MAVS (referred to as CARDIF) by NS3/4A. In particular, Li et al. (5) localize NS3/4A at mitochondria, which raises additional interesting questions. Can a small amount of cytoplasmic NS3/4A protein eliminate signaling through TRIF, or does IFN inhibition occur primarily through MAVS cleavage? Also, how is the localization of NS3/4A regulated? Does it require cellular factors and chaperones for translocation to the mitochondria, and could abrogation of mitochondrial localization allow for IFN production?

Indeed, the relative contribution of each of HCV's evasion strategies for establishing persistent infection remains unclear, and recent advances in the field allow this question to be addressed. In particular, identification of mutations that confer resistance to cleavage by NS3/4A in both MAVS and TRIF and development of a robust HCV culture system may allow determination of exactly how HCV establishes persistent infection (17–19). Pharmacological disruption of NS3/4A by compounds such as BILN 2061 holds great promise for the treatment of HCV infection. Moreover, the observations made by Li et al. (5) and Meylan et al. (7) may provide a mechanistic explanation for the effectiveness of such pharmaceuticals in treatment of HCV (20). Given the influence of type I IFN on the adaptive immune response, it may be important to target NS3/4A during acute stages of infection.

Approximately 20% of HCV-infected individuals are capable of eliminating infection, and current evidence suggests that the ability to clear infection correlates with a robust type I IFN response. HCV infection of chimpanzees infrequently results in chronic infection, and ISGs are highly expressed during viral clearance (21). The finding by Li et al. (5) that a single amino acid substitution in MAVS is protective for cleavage by NS3/4A raises the intriguing possibility that polymorphisms in MAVS may exist and account for an individual's ability to clear infection. MAVS was discovered independently by several groups and represents a groundbreaking advance. Especially intriguing was the demonstration by Seth et al. (8) that mitochondrial localization was essential for MAVS function. Li et al. (5) now show that at least one virus finds MAVS in this location and incapacitates the protein. Together, these data establish a novel role for the mitochondria, historically studied mainly for roles in both respiration and programmed cell death. Might there be other proteins associated with mitochondria that cooperate with MAVS for activation of IRF3 and NF-κB in response to viral infection? These provocative findings indicate that it is probable. Several other important pathogens, including influenza and Ebola viruses, have been reported to block type I IFN production (22). It will be exciting to discover whether these viruses have evolved independent mechanisms to inhibit MAVS function or antagonize upstream or downstream events in the signaling cascade. Understanding these molecular interactions between host cells and pathogens may lead to novel therapeutics.