Potential conflict of interest: Nothing to report.
The anti-inflammatory and antiapoptotic heme degrading enzyme heme oxygenase-1 (HO-1) has been shown recently to interfere with replication of hepatitis C virus (HCV). We investigated the effect of HO-1 products carbon monoxide (CO), iron and biliverdin on HCV replication using the replicon cell lines Huh-5-15 and LucUbiNeo-ET, stably expressing HCV proteins NS3 through NS5B. Incubation of these cell lines in the presence of the CO donor methylene chloride transiently reduced HCV replication, whereas an increase of iron in cell culture by administration of FeCl3 or iron-saturated lactoferrin did not interfere with HCV replication. Likewise, depletion of iron by deferoxamine during induction of HO-1 by cobalt-protoporphyrin IX did not restore HCV replication. The most prominent effect was observed after incubation of replicon cell lines in the presence of biliverdin. Biliverdin seems to interfere with HCV replication–mediated oxidative stress by inducing expression of antiviral interferons, such as interferon alpha2 and alpha17. Conclusion: The antioxidant biliverdin reduces HCV replication in vitro by triggering the antiviral interferon response and might improve HCV therapy in the future. (HEPATOLOGY 2009.)
Hepatitis C virus (HCV) infection represents one of the leading causes of chronic hepatitis worldwide, resulting in cirrhosis, steatosis, and hepatocellular carcinoma.1 Treatment of HCV infection is predominantly performed by a combination of pegylated interferon (IFN) alpha2 with the nucleoside analog ribavirin. Antiviral activity also has been demonstrated for other alpha IFNs, such as IFN alpha17, which effectively suppresses HCV replication and was implicated for future therapeutic use.2 Novel therapies targeting viral replication are under investigation, and some of them have undergone clinical trials.3 Evaluation of drugs targeting HCV replication has been profoundly improved by the establishment of hepatoma cell lines containing stably replicating HCV RNAs and expressing HCV proteins, the so-called replicon system. The HCV replicon cell lines Huh-5-154 and LucUbiNeo-ET5 express nonstructural (NS) proteins NS3 through NS5B and are a useful tool to measure HCV replication in cell culture.
The heme-degrading enzyme heme oxygenase-1 (HO-1) exerts anti-inflammatory and antiapoptotic effects in vitro and in vivo. Induction or overexpression of HO-1 protects kidneys from acute ischemic failure6 or ischemia–reperfusion injury,7 cardiac xenografts from rejection,8 and livers from ischemia–reperfusion injury caused by either transplantation9 or hemorrhage/resuscitation,10 as well as from apoptotic damage.11 Degradation of heme by heme oxygenases results in the production of carbon monoxide (CO), free iron, and biliverdin. HO-1, in contrast to the isoforms HO-2 and HO-3, is inducible by various stimuli,12, 13 such as cobalt-protoporphyrin-IX (CoPP),14, 15 but also by hypoxia, which can be induced by, for example, high amounts of CO.16 Of the HO-1 products, CO and biliverdin seem to be the major mediators of protective HO-1 effects within the liver.17–19 CO application in vitro or in vivo can be achieved by special gas chambers, or by the use of CO donors, such as methylene chloride (MC).17, 19, 20 With respect to the third HO-1 product iron, various reports point to no or nonbeneficial effects within the liver.21, 22
Induction or overexpression of HO-1 has recently been shown to interfere with replication of human immunodeficiency virus (HIV),23 hepatitis B virus (HBV),24 and HCV.25, 26 We now investigated the effect of HO-1 products CO, biliverdin, and iron to interfere with HCV replication.
The replicon cell lines Huh-5-154 and LucUbiNeo-ET,5 as well as their parental cell line Huh-7, were cultured in Dulbecco's modified Eagle medium (Invitrogen GmbH, Karlsruhe, Germany) containing 10% fetal bovine serum and 1% penicillin/streptomycin. Medium for replicon cell lines also contained G418 (1% for Huh-5-15, 0.5% for LucUbiNeo-ET). For experimental procedures, 150,000 cells were seeded into 24-well plates and allowed to adhere overnight. Incubations were performed at concentrations indicated in the figures and figure legends. Experiments were performed at least in triplicate.
CoPP was purchased from Frontier Scientific Europe Ltd., Carnforth, Lancashire, UK). Methylene chloride (MC), lactoferrin, deferoxamine, and FeCl3 were purchased from Sigma Aldrich GmbH (Steinheim, Germany). Biliverdin was purchased from MP Biomedicals (Heidelberg, Germany).
Detection of Messenger RNA by Reverse Transcription Polymerase Chain Reaction.
To verify altered gene expression, RNA was transcribed into complementary DNA by using the Verso cDNA Kit (Thermo Fisher Scientific, Waltham, MA). Oligonucleotides for subsequent polymerase chain reaction (PCR) reactions were obtained from Metabion International AG (Martinsried, Germany). Oligonucleotide pairs for real-time reverse transcription (RT)-PCR are summarized in Table 1. Real-time RT-PCR was performed by using the CFX Real-Time system (BIO-RAD, Munich, Germany) and reagents from Abgene (Thermo Fisher Scientific, Germany). Reactions were performed in a 10-μL volume. To confirm amplification specificity, PCR products were subjected to melting curve analysis and gel electrophoresis.
Table 1. Oligonucleotide Sequences for Real Time RT-PCR
5′-CAT GGA GTC CTG TGG CAT CCA C-3′
5′-GTA ACG CAA CTA AGT CAT AGT CCG-3′
5′-GAG ATT GAG CGC AAC AAG GAG-3′
5′-CTG ACT GCG GGA GTC ATC TC-3′
5′-AGC ACA TGG GTG CTT GCT-3′
5′-CTC ATC AAA CGC CTC GTA CA-3′
5′-CTA TCA CCG GGT GCG TTA GT-3′
5′-TCC AGG TAT GGA GCC TTC TG-3′
5′-AAC AGT ATC CGG GTG TGC TC-3′
5′-AGC ATC GGT AGT GTC CAT CC-3′
5′-CCA TAG TTA CTC TCC AGG TGA GAT C-3′
5′-GTG TTT AGC TCC CCG TTC A-3′
5′-GCA AGT CAA GCT GCT CTG TG-3′
5′-GAT GGT TTC AGC CTT TTG GA-3′
5′-AGG AGT TTG ATG GCA ACC AG-3′
5′-CAT CAG GGG AGT CTC TTC CA-3′
5′-CAA GCT CAA GAG CCT CAT CC-3′
5′-TGG GCT GTG TTG AAA TGT GT-3′
5′-ACA GCT GAA AGC CTT TTG GA-3′
5′-GCA TTA AAG GCA GGA AGC AC-3′
5′-CAC TGA CAT CCC AGA CGA TG-3′
5′-GAT CAG GCT CTT CAG CTT GG-3′
5′-ATG ATG GAA AGC GAA CAA GG-3′
5′-GAG ATG ATG CCA TCC CGT AG-3′
5′-AGC CGT TTG GAA CAG AAA TG-3′
5′-AGC TGA ATG GCA GAT GGT CT-3′
Fifteen micrograms protein were fractionated by 12% sodium dodecyl sulfate polyacrylamide gel electrophoresis and blotted onto nitrocellulose membranes. Western blots were developed using an enhanced chemiluminescence system (Amersham, Freiburg, Germany) according to the manufacturer's instructions. Semiquantitative evaluation was performed using the VersaDoc Imaging System (BioRad Laboratories GmbH, Munich, Germany). Antibodies for western blots were rabbit anti-HO-1 (Stressgen Biomol, Hamburg, Germany), mouse anti-hepatitis C NS5 (MorphoSys UK Ltd., Oxford, UK), and mouse anti-glyceraldehyde 3-phosphate dehydrogenase (HyTest Ltd., Turku, Finland).
Luciferase activity of LucUbiNeo-ET replicon cells was measured using the Luciferase Assay System (Promega, Mannheim, Germany) and normalized to the protein content in the individual samples. For all luciferase assays shown, protein contents in lysates were comparable, indicating that incubations did not affect cell metabolism.
The results were analyzed using Student t test if two groups were compared and the Dunnett's test if more groups were tested against a control group. If variances were not homogeneous in the Student t test, the results were analyzed using the Welsh test. All data in this study are expressed as a mean ± standard error of the mean. P ≤ 0.05 was considered significant.
HO-1 Induction Interferes with HCV Replication.
HO-1 overexpression has recently been shown to interfere with HCV replication.25, 26 To define the impact of HO-1 on HCV replication more precisely, Huh-5-15 replicon cells and their parental cell line Huh-7 (Fig. 1), as well as LucUbiNeo-ET replicon cells (Fig. 2), were incubated in the presence of the HO-1 inducer CoPP. Measurement of HCV polyprotein expressions by real-time RT-PCR showed that HCV replication was dose-dependently impaired (Fig. 1A-D), whereas HO-1 was dose-dependently induced by CoPP (Fig. 1E). The response to HO-1 induction was found to be less prominent in the replicon cell line, compared with the parental cell line, whereas HO-1 expression in untreated Huh-5-15 cells was elevated compared with untreated Huh-7 cells (Fig. 1E). Induction of HO-1 in both cell lines increased the expression of ferritin (Fig. 1F), indicating bioactivity of HO-1. Similar effects of HO-1 on HCV replication were measured in the LucUbiNeo-ET replicon cell line by luciferase assay (Fig. 2A). Reduction of HCV replication by HO-1 induction was also detectable at the protein level. Incubation of Huh-5-15 replicon cells in the presence of 10 μg/mL CoPP resulted in decreased NS5-protein and increased HO-1-protein expression (Fig. 2B).
Impact of HO-1 Products Carbon Monoxide and Iron on HCV Replication.
To determine the downstream mediator or mediators of HO-1 responsible for its effects on HCV replication, we first incubated Huh-5-15 or LucUbiNeo-ET replicon cells in the presence of the CO donor MC. Incubation was able to reduce HCV replication, as detected by luciferase reporter assay dose-dependently (Fig. 3A). HO-1 expression in cells incubated in the presence of 100 mM MC for 6 hours was slightly increased, as measured by real-time RT-PCR, but induction was not significant (untreated: 1.011 ± 0.05235, n = 12; MC treated: 1.175 ± 0.1212, n = 10; P = 0.2012). The effect of MC on HCV replication was transient; it was no longer detectable at 24 hours after the start of incubation (Fig. 3B).
Induction of HO-1 results in a release of iron, which in turn induces ferritin (Fig. 1F) and also might contribute to a reduction of HCV replication. We therefore tried to reverse the inhibitory effect of CoPP on HCV replication by co-incubating cells with the iron-trapping agent deferoxamine (Fig. 3C, D). Our results show that, whereas CoPP incubation reduced HCV replication, co-incubation with deferoxamine did not reverse this effect, as measured by real-time RT-PCR for NS5B expression (Fig. 3C) or luciferase reporter assay (Fig. 3D). We also tried to mimic an iron effect by adding iron III chloride solution (FeCl3) or iron-saturated lactoferrin to LucUbiNeo-ET replicon cells. Our results show that neither addition of FeCl3 (Fig. 3E) nor addition of lactoferrin (Fig. 3F) reduced HCV replication, as measured by luciferase assay.
Impact of the HO-1 Product Biliverdin on HCV Replication.
Incubation of LucUbiNeo-ET replicon cells in the presence of biliverdin dose-dependently interfered with HCV replication, as detected by luciferase assay (Fig. 4A). The same result was obtained in Huh-5-15 cells by real-time RT-PCR (Fig. 4B) and western blot (Fig. 4C) for the HCV nonstructural protein NS5. The effect of biliverdin was not attributable to overall induction of HO-1 (Fig. 4B, C). To exclude unspecific effects of the green biliverdin solution on luciferase activity, we used a reporter construct constitutively expressing casein kinase 2 beta subunit as an unspecific control. Biliverdin incubation did not change luciferase activity of this construct, confirming a specific effect on HCV replication (Fig. 4D).
Biliverdin Inhibits HCV Replication by Induction of Antiviral Interferons.
HCV core protein and nonstructural protein NS5A have been shown to induce oxidative stress,26, 27 a trigger of viral replication.28 In fact, we found that incubation of LucUbiNeo-ET replicon cells in the presence of H2O2 dose dependently resulted in modification of HCV replication, which was increased by low concentrations and reduced by higher concentrations (Fig. 5A). Oxidative stress, for example, caused by viral replication, down-regulates expression of antiviral factors.29, 30 We found that biliverdin, which, like bilirubin, is able to reduce oxidative stress,31 induced expression of antiviral interferons (Fig. 5B). We measured endogenous interferon alpha2 and alpha17 expression (Fig. 5B), as well as expression of interferon-dependent antiviral genes, such as 2′,5′-oligoadenylate synthetase (OAS) 1 and OAS2, or protein kinase R (PKR), but not OAS3, as well as heme-regulated eIF2alpha kinase (Fig. 5C). Therefore, biliverdin seems to interfere with HCV replication by restoring the expression of antiviral interferons, which are reduced during viral infection.
HO-1 has profound antiviral effects on HBV, HCV and HIV, although the replication machinery of these viruses is quite different. In case of HBV inhibition, HO-1 is able to reduce stability of HBV core protein and thus block refill of nuclear HBV covalently closed circular DNA.24 The contribution of HO-1 products to inhibition of HBV replication is the subject of our ongoing investigations. In the case of HIV, HO-1 has been shown to reduce, but not completely block, viral entry and also to interfere with nonspecified post-entry events in viral replication.23 A connection between HO-1 and HCV replication is implicated by the observation that HCV increases basal HO-1 expression32 but interferes with HO-1 induction, rendering cells more susceptible to cytotoxicity.33 This effect has been attributed to HCV core protein33; however, nonstructural HCV proteins also might be involved, because we observed that HO-1 induction by CoPP was less prominent in HCV replicon cells expressing NS3 to NS5, compared with the parental cell line Huh-7 (Fig. 1E). Basic HO-1 expression in Huh-5-15 cells was found to be elevated but not significantly higher than in Huh-7 cells (Fig. 1E), which might be a cellular defense mechanism against oxidative stress induced by viral infection.29
Recently, HO-1 effects on HCV replication have been demonstrated by induction25 or overexpression of the enzyme,26 where the underlying mechanism has been attributed to modulation of oxidative cellular stress.26 In previous HO-1–related work, we have been able to connect HO-1 effects in the liver to one, or a combination, of its products.11, 17 This was also the aim of the current study. Our results show that predominantly the HO-1 product biliverdin mediated anti-HCV activity, whereas CO had moderate effects, and the third HO-1 product iron in our hands did not significantly interfere with HCV replication, which is in contrast to a previous report showing that iron inhibits HCV replication by inactivating NS5B.34
With respect to CO effects on HCV replication (Fig. 3A, B), we only found transient reduction of replication, which was detectable at 6 hours after the onset of experiments but was no longer detectable after 24 hours. The mechanism of CO-induced transient repression of HCV replication remains elusive and might be partially attributable to a slight induction of HO-1 expression.
Biliverdin has been shown to reduce replication of HIV35 and human herpes virus type 6,36 whereas it did not interfere with replication of human herpesvirus type 1 or cytomegalovirus.36 It has been speculated that biliverdin might interfere with cellular processes specific for replication of certain viruses,36 and, in case of HIV, also directly inactivate viral particles.35 Conversely, oxidative stress seems to trigger viral replication,29 a process that might be a target of the antioxidant biliverdin. In fact, it has been shown that HCV induces oxidative stress27, 28 and that oxidative stress interferes with antiviral gene expression.30 We also found that moderate oxidative stress in replicon cells triggered HCV replication (Fig. 5A). Biliverdin incubation induced expression of antiviral alpha interferons, for example, interferon alpha2 and alpha17 (Fig. 5B). Downstream effects of interferon alpha treatment, such as expression of PKR or OAS, were also enhanced by biliverdin (Fig. 5C), underlining its antiviral effect. It has been shown that phosphorylation of translation initiation factor eIF2alpha by interferon-inducible protein kinase PKR is able to suppress HCV replication in JFH1-infected Huh-7 cells, further elucidating the mechanism of IFN alpha–induced inhibition of HCV replication.37 Likewise, under conditions of heme deficiency, heat shock, or oxidative stress, heme-regulated eIF2alpha kinase is able to phosphorylate eIF2alpha and inhibit translation.38 In fact, expression of both kinases was found to be increased after biliverdin incubation (Fig. 5C). These findings do not exclude an additional and yet unknown effect of biliverdin on HCV replication that is independent of reduction in oxidative stress.
Recently, it has been shown that oxidative stress also might interfere with HCV replication,39 as we observed for higher H2O2 concentrations (Fig. 5A). This effect has been attributed to modulation of the MEK-ERK1/2 signaling pathway,40 and also might be a result of reduced cellular viability and proliferation, because oxidative stress is involved in regulation of both hepatocyte apoptosis and proliferation.40 In fact, modulation of oxidative stress has been proposed as a therapy concept for HCV. A clinical trial showed that patients might benefit from a combination treatment with IFNalpha and the anti-oxidant N-acetyl-cystein in comparison with IFN alpha treatment alone,41 although others did not observe this effect.42 Interestingly, only limited data are available concerning the effects of a single treatment with anti-oxidants on HCV replication.
Biliverdin is reduced to bilirubin by biliverdin reductase, and it is not clear whether an elevation of biliverdin/bilirubin would be beneficial. Hepatitis A virus infection is often associated with high levels of plasma bilirubin. Interestingly, it has been reported that super-infection of HCV patients with hepatitis A virus frequently results in a decrease of HCV replication.43 Recently, a case report described complete clearance of HCV ribonucleic acid for at least 4 years in a chronic HCV patient superinfected with hepatitis A virus. During the onset of disease, an increase of interferon gamma as well as bilirubin levels of 24 mg/dL were observed.44
In the past, we investigated effects of biliverdin on acute hepatitis in mice.11, 17 In these experiments, mice were treated with biliverdin for 24 hours, and they seemed to tolerate this treatment very well. With respect to future therapy, it will now be important to investigate long-term or dose-dependent toxic effects of biliverdin.
The perfect technical assistance of Elena Tasika and Christine Loscher is gratefully acknowledged.