Hepatology highlights


  • Potential conflict of interest: Nothing to report.

Do Not Ask for Whom the Receptor Tolls: It Tolls-Like for NS3

Otsuka and coworkers in Tokyo perform a brilliant microdissection of the interactions between HCV-NS3 protein and the Toll-like 3 receptor (TLR-3) pathway. Their study builds on a prior study by Foy et al. (Science 2003) which demonstrated that NS3 protein inhibited the phosphorylation of interferon regulatory factor 3 (IRF-3), in turn inhibiting induction of IFN-β, a key component of the innate immune response. This virus-specific inhibit-ion of innate immunity was considered one of the early keys to establishing persistent infection. Otsuka et al. carry this observation further to more precisely define the mechanisms of NS3 inhibitory activity. As background, in the life cycle of RNA viruses, double-stranded RNA is generated and recognized by TLR-3 resulting in activation of IRF-3, a key regulator of type 1 interferons. To investigate these interactions at the molecular level, Otsuka et al. focused on TRIF, an adaptor molecule responsible for linking TRL-3 and IRF-3, and recently identified kinases, TANK-binding kinase-1 (TBK1) and IKKε, that have been shown to associate with both IRF-3 and TRIF. Using HCV replicon-harboring cells and interferon cured cells, they demonstrated that transfection with a TRIF expression plasmid induced IRF-3 and IFN-β promoter activation. They then showed that transient transfection with NS3 and NS3/4A inhibited TRIF-mediated induction of IFN-β promoter whereas NS2, NS4A, NS4B, NS5A and NS5B had no inhibitory effect. (See Fig.) By inserting a stop codon into NS-3 RNA such that the expression of NS-3 protein was blocked, they showed that the inhibitory effect was due to NS-3 protein and not NS-3 RNA. Next in coimmunoprecipitation experiments, they showed that NS-3 bound to the kinase TBK1 and to IKKε but they focused on TBK1 because it is ubiquitously expressed. It was further found using NS3-deleted expression plasmids that it is the helicase domain of NS3 that binds to TBK1 and not the protease domain. Having established that NS-3 binds to TBK1 and knowing that TBK1 binds to IRF-3, they then did critical experiments which showed that TBK1 binding to IRF-3 was inhibited in the presence of NS3 protein as was TRIF-related phosphorylation of IRF-3, the latter necessary for IRF dimerization and its downstream effect on IFN-β production. This is an important study that further builds on the concept that HCV proteins, especially, but not exclusively, NS3 inhibit the host innate and adaptive immune responses and foster viral persistence. While the bottom line of viral persistence may be diminished and dysfunctional cell-mediated immune responses, it is becoming increasingly clear that the virus plays a critical role in its own survival by its high mutability, quasispecies nature and the direct inhibitory effects of the proteins it expresses in abundance. This is a very clever virus that is clearly taking its Toll. (See HEPATOLOGY 2005;41:1004–1012)

Illustration 1.

Short RNAs Reach New Heights

RNA interference (RNAi) is the new buzzword in molecular circles. RNAi is a cellular mechanism that detects and destroys double-stranded RNA (dsRNA) and is a part of the host antiviral defense system. Short interfering RNA (siRNA) molecules are approximately 21-nucleotide dsRNA intermediates that serve to guide a unique RNAi silencing complex to the target RNA resulting in its degradation. It has been logical to presume that this natural silencing mechanism could be harnessed into a therapeutic paradigm, but translation from concept to therapeutic has been difficult. Major problems impairing therapeutic application are the instability of siRNAs due to nuclease digestion and the difficulty of delivering the molecule to its specific tissue target. Morrissey et al. have developed a method to stabilize siRNA and have tested its delivery and efficacy in a murine model of HBV replication. The authors were able to chemically modify their synthesized siRNAs by forming duplexes in which all 2′-OH groups were substituted. This increased siRNA half-life in the presence of nucleases from 1-5 minutes in the unmodified molecule to 15-18 hours in the chemically modified version. The investigators next showed these modified siRNAs could achieve effective gene silencing in an HBV cell culture system (transfected HepG2 cells). At the heart (liver) of this study were experiments to test the efficacy of gene silencing in vivo. They thus co-administered a chemically-modified HBV specific siRNA with an HBV vector via high pressure tail-vein injection into mice and, 3 days after injection, demonstrated a 3-log decrease in HBV DNA compared to control mice. They claim this to be the first demonstration of in vivo activity of a modified siRNA completely lacking′-OH residues. Subsequently, the modified siRNA was given intravenously thrice daily for two days beginning 72 hours after injection of the HBV vector. A dose dependent reduction in serum HBV DNA of approximately one log compared to controls was observed (P < .0006). The long, and mostly short of it is that the investigators were able to stabilize siRNAs to a therapeutically useful half-life, were able to administer the agent intravenously and were able to inhibit HBV DNA replication both in vitro and in vivo. This represents a proof of principle and, as the authors point out, efficacy was achieved despite the fact that less than 1% of the administered dose reached the liver. Thus, enhancements in tissue delivery could dramatically increase the antiviral effect. Nonetheless, we shouldn't be mouse-trapped into thinking that the hurdles of short interfering RNAs have been circumvented. Much work needs to be done before these small molecules can be tested for safety and efficacy in humans. Perhaps we should start with short subjects. (See HEPATOLOGY 2005;41:1349–1356)

MTHFR Genes, Cysteines and Steatotic Schemes

Hyperhomocysteinemia is not a word that rolls off one's tongue or that seems inherently interesting. However, Adinolfi and coworkers have given hyperhomocysteinemia a new spin that provides at least one mechanism for HCV-associated steatosis. Mato and Lu have written an excellent accompanying editorial titled “The Bad Thiol” that fully describes the biochemical pathways involved in these interactions, thankfully sparing me that need. The underlying supposition of the Adinolfi study is that hyperhomocysteinemia results in steatosis and that the C677T polymorphism in methylenetetrahydrofolate reductase (MTHFR) induces hyperhomocysteinemia (as I write this on Mother's Day, MTHFR has double meaning). MTHFR is an enzyme that is absolutely required for the methylation of homocysteine to methionine; a common polymorphism of the gene encoding MTHFR, the C677T transversion, inhibits this methylation and thus raises the level of homocysteine. The prevalence of this polymorphism in the general population is 12%-15%. Adinolfi et al. studied 116 patients with chronic hepatitis C (CHC) and first showed that CHC patients with steatosis >20% had higher HAI (P = .008) and fibrosis (P = .0001) scores than subjects with <20% or no steatosis. At the heart of the study, the mean values of plasma homocysteine were significantly higher in subjects with grade 3-4 steatosis than those with grades 1-2 or no steatosis (P = .0001). In a multivariate analysis, the grade of steatosis was independently associated with the level of homocysteinemia (OR = 7.1), liver fibrosis score (OR = 4.0), genotype 3 (OR = 4.6) and HAI score (OR = 3.8). Finally, MTHFR polymorphism correlated with the degree of steatosis, liver fibrosis and hyperhomocysteinemia; the serum level of homocysteine in those with the TT polymorphism was twice that of the common CC variant (P = .0001); among CHC patients with the CC genotype only 11% had a high degree of steatosis while in the CT and TT genotypes, high-grade steatosis (>20%) was present in 49% and 64%, respectively (P = .0001).The estimated relative risk for developing steatosis >20% was 6 times higher for the CT genotype and 20 times higher for the TT genotype of MTHFR. (See Fig.) Thus, in chronic hepatitis C, significant associations suggest a chain of events wherein a genetic variant of MTHFR causes a methylation transfer block that results in hyperhomocysteinemia. The latter leads to steatosis that in turn is associated with increased hepatic inflammation and fibrosis. A nice tight chain, but what is the basis for these associations? Adinolfi et al. suggest that homocysteine induces endoplasmic reticulum stress that activates genes responsible for cholesterol/triglyceride biosynthesis and uptake and also upregulates low density lipoprotein receptor gene. Although hepatic lipid export remains normal, more lipid is imported and synthesized than the export system can handle. Thus, HCV associated steatosis is a multifactorial event including host factors (BMI, insulin resistance, hypertriglycerinemia), viral factors (HCV genotype 3, HCV core and NS5 protein effects) and now the addition of another host factor, namely a genetic polymorphism that diminishes the ability to convert homocysteine to methionine. In addition to providing a new mechanism for steatosis and its injurious effects, this study opens the door to therapeutic approaches in hepatitis C designed to diminish serum and intrahepatic homocysteine levels. (See HEPATOLOGY 2005;41:995–1003)

Illustration 2.

CD8: Too Little, Too Late

The cell-mediated immune (CMI) response to HCV continues to be analyzed and dissected. Each study adds a new piece, although the full picture has yet to emerge. Many prior studies have compared CMI responses in chronically infected patients with responses in patients who recovered at some indefinite time in the past. These studies have been uniform in showing that HCV-specific CD4 and CD8 responses are impaired in chronically-infected patients. However, the retrospective nature of these studies has left open whether these diminished responses are the cause or the effect of the persistent carrier state. Cox and associates address this dilemma by identifying acute HCV infections in a population of continually at-risk intravenous drug users and entering them into long-term prospective follow-up. CD8 responses to overlapping peptides across the HCV genome were studied by IFN-γ ELISPOT in 23 acutely infected IDU. The key findings were as follows: (1) 83% developed detectable CD8 responses to a median of 4 HCV antigens (range 1–10); 4 subjects (17%) failed to recognize any HCV epitopes; (2) the major observation of this study is that the median number of epitopes recognized and the magnitude of the CD8 response early in HCV infection did not differ significantly between those who cleared infection (21%) and those who became persistent carriers (79%); chronic infection developed despite broadly reactive CD8 responses that were maintained for over a year; (3) CMI appeared approximately one month (range 29–50 days) after the onset of viremia and never before viremia. Thus, the emerging CMI response had to combat a state of established, high-level viral replication. Based on transgenic mouse and chimp studies, this late appearance of CMI vis-à-vis viremia, gives the virus an important survival advantage; (4) the vigor and the breadth of the CD8 response declined over time in both recovered and chronically-infected subjects, but the decline was much greater in those chronically infected; complete loss of recognition of one or more antigens occurred in 9 of 9 chronic carriers; (5) despite demonstrated viral quasispecies evolution, no new epitopes were recognized by CD8 cells after the first 6 months of chronic infection.

This prospective evaluation of CD8 responses during the evolution from acute to chronic HCV infection demonstrates a progressive loss in the strength and breadth of reactivity to HCV antigens in those who develop chronic infection, but importantly shows that the initial CD8 IFN-γ response was similar in those who recovered and those who became chronically infected; hence, the early CD8 response was not predictive of virological outcome (See Fig.). This study also showed that after 6 months, chronic carriers could no longer respond to new HCV epitopes. The temporal relationships shown in this study suggest that impaired CD8 responses are the consequence of chronic antigenic stimulation resulting in functional T-cell exhaustion rather than the cause of persistent infection. From this study and a mounting literature on CMI in HCV infection, we are left to conclude that viral persistence must have multiple underlying etiologies including, but not limited to, the suppressive effects of HCV antigens on the innate immune response, HCV-induced impaired dendritic cell function, early high-level viremia and late appearing effector T-cell response, diminished CD-4 T-cell help, escape mutation, perhaps inability to mount T-cell responses to new HCV strains as they appear over time and, the suppressive effects of T-regulatory cells.

Illustration 3.

It is frustrating after all this time to lack a definitive answer to this central problem of HCV infection, but it is exciting to have an increasing array of tools and patient samples to unravel the pieces of this highly prevalent chronic infection. The persistence of this infection is matched only by the persistence of the investigators dedicated to unraveling its mysteries. (See HEPATOLOGY 2005;42:104–112)

CD4: IL-2 As Metaphor

The study of Semmo et al. is a nice complement to the above discussion of HCV persistence. In this study, IL-2 and IFN-γ secretion from PBMC and CD8-depleted PBMC were measured by ELISPOT in 23 patients with chronic HCV infection and 11 presumed recovered subjects. A recent study in HIV-infected patients showed that in the presence of high viral load, HIV-specific CD4 cells exist, but lack proliferative capacity and that upon antigenic stimulation these cells can produce IFN-γ, but not IL-2. When the antigen load was reduced by antiviral therapy, IL-2 secretion and cell proliferation were both restored. In this study, Semmo et al. showed that under antigenic stimulation, a positive IFN-γ response was found in 11/11 of HCV RNA negative subjects and in only 13/23 RNA positives (P = .01). Further, only 4/23 (17%) RNA+ patients had any IL-2 response compared to 7/11 (64%) RNA negatives (P = .01) and the magnitude of response was markedly diminished in those PCR-positive subjects who did respond (P = .01). When IFN-γ and IL-2 responses against the same HCV antigens were compared, the IFN-γ:IL-2 secretion ratio was 1:1 or 2:1 in recovered subjects and 10:1 in those chronically infected. In vitro experiments showed that the addition of exogenous IL-2 not only increased the magnitude of IFN-γ ELISPOT responses, but also induced responses to new epitopes.

In sum, among chronic carriers of HCV a significant proportion maintained IFN-γ–secreting cell populations, but few had CD4 cells secreting IL-2 and this could account for the low proliferative capacity of these cells. The addition of IL-2 in vitro was able to restore and enhance cellular responses. The authors explain their findings by suggesting that as T cells proliferate in response to antigen they lose IL-2 secretory capacity as they acquire effector functions. When antigenic stimulation is extreme or prolonged there is progressive loss of cytokine secretion before deletion occurs. If this were the case, reduction in viral load through treatment should restore CD4 cell responsiveness through a return to the IL-2–secreting state. Unfortunately, the authors did not include a population sequentially studied before and after successful antiviral therapy. Such a study is promised and we await the outcome with interest. Of additional interest would be the administration of IL-2 alone or in combination to patients who have failed conventional therapy, if toxicity issues could be circumvented.

As interesting as this observation is, it again begs the question of whether diminished CD4 IL-2 secretion plays a role in the pathogenesis of persistent HCV infection or is merely the consequence of the prolonged antigenic stimulation engendered by the chronic infection. It would be of great interest to do a comprehensive analysis of CD4 cells, including IL-2 secretion, in the acutely-infected and prospectively followed patients of Cox et al. noted previously. I would be willing to broker this deal for a small finder's fee. (See HEPATOLOGY 2005;41:1019–1028)

Son of GBV-B

The GB virus refuses to disappear. After being found to lack clinical relevance as a human hepatitis agent, it reemerged as an agent that inexplicably may ameliorate HIV infection. Now it surfaces again as a chimeric virus that will allow study of HCV. Of the 3 identified GB viruses, GBV-B is phylogenetically the closest to HCV sharing 27%–33% sequence homology in the protein coding regions. GBV-B is hepatotropic in tamarins and can cause chronic hepatitis in this species. Rijnbrand and associates produced a series of HCV-GBV-B chimeras using an infectious clone of GBV-B and substituting defined domains of the HCV 5′ NTR, including the IRES, into the GBV-B nucleotide backbone. Each of the chimeras were shown to bind ribosomes and to initiate translation of the RNA transcripts following transfection of primary tamarin hepatocytes, but only domain III of HCV NTR was able to substitute for the same domain in GBV-B to cause acute hepatitis when inoculated into naïve tamarin liver. Domain III plays a critical role in binding of the IRES to 40S ribosomal subunits. Domains I and II must contain critical virus-specific signals that are significantly divergent in HCV and GBV-B that they cannot be substituted. Even the domain III substitution required adaptive mutations outside the NTR in order to replicate efficiently in tamarin liver. This chimeric approach now provides a model to study the anti-viral potential of drugs targeting this important domain of the highly conserved HCV 5′NTR. (See HEPATOLOGY 2005;41:986–994)