Comparative Analysis of the Lambda-Interferons IL-28A and IL-29 regarding Their Transcriptome and Their Antiviral Properties against Hepatitis C Virus

Background Specific differences in signaling and antiviral properties between the different Lambda-interferons, a novel group of interferons composed of IL-28A, IL-28B and IL-29, are currently unknown. This is the first study comparatively investigating the transcriptome and the antiviral properties of the Lambda-interferons IL-28A and IL-29. Methodology/Principal Findings Expression studies were performed by microarray analysis, quantitative PCR (qPCR), reporter gene assays and immunoluminometric assays. Signaling was analyzed by Western blot. HCV replication was measured in Huh-7 cells expressing subgenomic HCV replicon. All hepatic cell lines investigated as well as primary hepatocytes expressed both IFN-λ receptor subunits IL-10R2 and IFN-λR1. Both, IL-28A and IL-29 activated STAT1 signaling. As revealed by microarray analysis, similar genes were induced by both cytokines in Huh-7 cells (IL-28A: 117 genes; IL-29: 111 genes), many of them playing a role in antiviral immunity. However, only IL-28A was able to significantly down-regulate gene expression (n = 272 down-regulated genes). Both cytokines significantly decreased HCV replication in Huh-7 cells. In comparison to liver biopsies of patients with non-viral liver disease, liver biopsies of patients with HCV showed significantly increased mRNA expression of IL-28A and IL-29. Moreover, IL-28A serum protein levels were elevated in HCV patients. In a murine model of viral hepatitis, IL-28 expression was significantly increased. Conclusions/Significance IL-28A and IL-29 are up-regulated in HCV patients and are similarly effective in inducing antiviral genes and inhibiting HCV replication. In contrast to IL-29, IL-28A is a potent gene repressor. Both IFN-λs may have therapeutic potential in the treatment of chronic HCV.

IFN-ls signal through a receptor complex comprised of IL-10R2 and a unique subunit, IFNl-R1. While IL-10R2 is widely expressed on a number of different cell types including hematopoietic cells, expression of the specific receptor IFNl-R1 is more restricted, e.g., it seems to be weakly expressed on leukocytes. As signaling through the type-I-interferon receptor, signaling through the IFN-l receptor results in the activation of signal transducer and activator of transcription (STAT)-1 and STAT2. Together with an accessory factor, IFN regulatory factor 9 (IRF-9; p48), STAT1 and STAT2 form the transcription factor IFN-stimulated gene factor-3 (ISGF3) which translocates to the nucleus to initiate the induction of target genes [1].
Like type I IFNs, IFN-ls are strongly induced by double stranded (ds) RNA or viral infection, suggesting common regulatory factors. In fact, it has recently been demonstrated that the IL29 gene, similar to the gene encoding IFN-b, is regulated by virus-activated IRF3 and IRF7. In contrast, IL28A and IL28B gene expression is mainly controlled by IRF7, similar to the gene encoding IFN-a [13].
Although the antiviral effects of IL-28A and IL-29 have been compared with IFN-a, IFN-b and IFN-c regarding their antiviral and gene-inducing activities [7,14,15,16,17], there are very limited data directly comparing signaling and antiviral properties of IL-28A and IL-29. Therefore, in this study, we directly compared these two cytokines regarding their signal transduction, target gene expression profiles, antiviral properties against HCV and their expression in different human liver diseases.

Isolation of leukocytes, peripheral blood mononuclear cells (PBMC) and granulocytes
White blood cells were isolated from fresh human anticoagulated blood. For the isolation of total leukocytes, 5 ml of erythrocyte lysis buffer were added to 1 ml of blood. Following erythrocyte lysis and washing steps with PBS, the leukocytes were pelleted by centrifugation. For the isolation of PBMCs and granulocytes, a 6% dextran solution (molecular weight 250.000) was added to whole blood to precipitate the erythrocytes. The supernatant containing the white blood cells was treated with lysis buffer to remove any residual erythrocytes. Following washing steps, the cell suspension was layered onto a Ficoll-Hypaque density gradient and centrifuged at 4006g for 30 minutes to separate mononuclear cells from granulocytes.
Reverse transcriptase polymerase chain reaction (RT-PCR) and quantitative PCR Trizol reagent (Invitrogen, Karlsruhe, Germany) was used to isolate total cellular RNA. Reverse transcription of 2 mg RNA to cDNA was performed with Omniscript reverse transcriptase (Qiagen, Hilden, Germany). PCR cycling was run as follows: 40 cycles of denaturing at 95uC for 30 sec, annealing at 60uC for 30 sec, extension at 72uC for 30 sec. Real-time quantitative PCR was carried out using the Quantitect SYBR Green PCR Kit from Qiagen (Hilden, Germany) in an ABI Prism 7700 Sequence Detection System (Applied Biosystems, Darmstadt, Germany). Oligonucleotide primer pairs (MWG Biotech, Ebersberg, Germany) were designed according to the published sequences avoiding amplification of genomic DNA and are listed in Table 1.

Gel electrophoresis and immunoblotting
Cells were solubilized in lysis buffer consisting of 20 mM Tris-HCl (pH 7.4), 150 mM NaCl, 2 mM EDTA, 2 mM EGTA, 1% Nonidet P-40, 2 mM phenylmethylsulfonyl fluoride, a protease inhibitor cocktail (Roche, Mannheim, Germany) and phosphatase inhibitors (400 mM sodium orthovanadate and 4 mM NaF). Cell lysates were passed six times through a 21 G needle. After chilling on ice for 30 minutes, lysates were cleared by centrifugation for 20 minutes at 10,000 g. The Bradford method was used to quantify the protein concentration of each sample. Immunoblotting was performed as previously described [24].

Anti-HCV assay in Huh 5-2 cells
Huh 5-2 cells (''HCV-Huh-7'') were seeded in 6-well plates at a density of 2610 5 per well in complete DMEM. Following incubation for 24 hours at 37uC (5% CO 2 ), medium was replaced Table 1. Primers used for PCR and quantitative PCR. with 2 ml DMEM supplemented with IL-28A, IL-29 or IFN-a. After further incubation at 37uC for 72 hours, cell culture medium was removed and luciferase activity was determined using a Lumat LB9507 luminometer (Berthold, Freiburg, Germany) as described recently [21].

IL-28A immunoluminometric assay (ILMA)
For quantification of IL-28A in human serum samples, Human IL-28A/IFN-lambda 2 DuoSet (R&D Systems, Wiesbaden, Germany) was used to develop an IL-28A-specific immunoluminometric assay. The detection limit was 4.9 pg/ml. Signal detection was performed with a biotinylated detection antibody and incubation with neutravidin-HRP and the chemoluminescent substrate Femtoglow (Michigan Diagnostics, Troy, MI).

Murine cytomegalovirus (MCMV) infection in vivo
1610 6 plaque-forming units of MCMV of the Smith strain [25] in PBS were injected intravenously into C57/BL6 mice as described previously [8]. Control mice got an injection of PBS only. After 45 h, mice were euthanized by CO 2 inhalation, and the livers were collected and homogenized in Trizol reagent to isolate total RNA. The study was approved by the Animal Care and Use Committee of the State of Bavaria (Regierung von Oberbayern, approval ID 209.1/211-2531-18/03) according to the National Institutes of Health ''Guide for the Care and Use of Laboratory Animals''.

Sampling of human liver biopsy tissue, blood and serum samples including Ethics statement
Human liver biopsy tissue was obtained from patients undergoing diagnostic liver biopsy for medical reasons such as staging of chronic hepatitis C. The study was approved by the Ethics committee of the Ludwig-Maximilians-University Munich (Department of Medicine, Munich-Grosshadern) and adhered to the ethical principles for medical research involving human subjects of the Helsinki Declaration (http://www.wma.net/e/policy/b3.htm). All participating patients gave written, informed consent prior to liver biopsy sampling. A 3 mm long piece of the biopsy cylinder was immediately stored in Trizol reagent and RNA was isolated as described previously [26]. Human blood and serum samples were obtained after written informed consent from patients and controls and were stored at 280uC until further analysis.

Microarray analysis
After reaching 70% confluency, Huh-7 cells were incubated overnight with serum-reduced medium containing 1% FCS. The next day, cells were stimulated in triplicates with 100 ng/ml IL-28A, IL-29 or left unstimulated. RNA was isolated at the indicated time points using the RNeasy kit from Qiagen (Hilden, Germany). For the analysis of the cytokine-induced gene expression, Agilent Whole Human Genome Oligo Microarrays were used in combination with a One-Color based hybridization protocol. Signals on the microarrays were detected with the Agilent DNA Microarray Scanner. Differential gene expression was identified within the human cells by applying appropriate biostatistics to the data set. GeneSpring GX 10 analysis software (Agilent Technologies, Santa Clara, CA) was used to normalize and analyze the raw data. Cytokine-induced gene expression was calculated in comparison to unstimulated cells at the same time points. Welch's approximate t-test (''unpaired unequal variance'', parametric) was applied to the comparison of the different groups. Resulting pvalues were corrected for multiple testing using the algorithm of Benjamini and Hochberg [27]. Functional analysis (categories of biological processes, molecular functions and pathway categories) of induced and repressed genes was performed using the Panther software [28]. By comparing cytokine-regulated gene identification numbers (IDs) to the distribution of all gene IDs represented on the Whole Human Genome Oligo Microarray (Agilent Technologies), it was calculated whether a specific class is over-or underrepresented. P-values of p,10 25 (based on binomial test) were considered as a sign of manifest enrichment in the context of a Panther analysis for biological processes, molecular functions and pathway categories. All microarray data presented are MIAME compliant and the raw data have been deposited in a MIAME compliant database in MIAMExpress (available at www. ebi.ac.uk/microarray/, accession number E-MEXP-2861) as detailed on the MGED Society website http://www.mged.org/ Workgroups/MIAME/miame.html.

Statistical analysis
Statistical analysis was performed by using two-tailed Student's t-test. P levels,0.05 were considered as statistically significant. Standard errors of the mean (SEM) were calculated by dividing the standard deviation (SD) by the square root of the number of single data in the respective group.

Hepatic cells express the IFN-l receptor complex
In order to utilize a hepatic cell model to study the IFN-l ligand-receptor system, we first confirmed that the IFN-l receptor subunits IL-10R2 and IFNl-R1 are present in hepatic cells. RT-PCR analysis demonstrated IL-10R2 and IFNl-R1 mRNA expression in several human hepatic cancer derived cell lines (HepG2, Hep3B, Huh-7) as well as in HCV replicon expressing Huh-7 5-2 cells and Huh-7 cells cured from HCV by IFN-a and IFN-c (HCV-cured Huh-7) (Fig. 1A). Primary hepatocytes from two different donors also expressed mRNA for both IFN-l receptor subunits (Fig. 1A) while leukocytes show only low expression (Fig.1A). The prostate cancer cell line LNCaP was used as a negative control for IFNl-R1 expression. Quantitative PCR analyses including total leukocytes as well as peripheral blood mononuclear cells (PBMC) and granulocytes from four different donors revealed that leukocytes express only 5.561.8% of the level or IFN-lR1 in comparison to Huh-7 cells (Fig. 1B). Among the leukocytes, expression of IFN-lR1 was 1.960.5-fold higher in PBMC than in granulocytes (Fig. 1B). IL-10R2 mRNA expression was higher in leukocytes compared to liver cells (Fig. 1C), and granulocytes had a 1.760.1-fold higher IL-10R2 expression than PBMC.

IFN-ls induce STAT1 but not STAT3 phosphorylation
Previous studies in other cell systems reported activation of STAT signaling by IFN-ls [1,14]. Therefore, we investigated the influence of IL-28A and IL-29 on phosphorylation levels of STAT1 and STAT3 in naïve Huh-7 cells. As demonstrated in Fig. 2A, both cytokines activated STAT1 whereas IL-29 had a stronger effect than IL-28A. On the other hand, STAT3 was not activated by IL-28A and stimulation with IL-29 had only minor effects (Fig. 2B).
IL-28A and IL-29 induce expression of similar genes in Huh-7 cells but differ in their gene down-regulating abilities Next, we analyzed by microarray experiments the IFN-linduced gene expression in hepatic cells. Naïve Huh-7 cells were stimulated for 6 hours with 100 ng/ml IL-28A or IL-29, while controls were left unstimulated for the same time interval. Altogether, a total number of 389 genes were influenced equal to or more than two-fold by IL-28A (117 genes up-regulated, 272 down-regulated) and a total number of 115 genes by IL-29 (111 genes up-regulated, 4 down-regulated), respectively (p,0.01 without correction for multiple testing) ( Table 2). When applying more stringent criteria (adjusted p-values [adj. -p],0.05; corrected for multiple testing by the Benjamini and Hochberg algorithm [27]), a total of 154 genes was significantly regulated by IL-28A (65 genes up-regulated, 89 down-regulated) and only 3 genes were significantly regulated by IL-29 (all up-regulated). The top 20 hits of up-regulated genes were identical for both cytokines although in a slightly different order ( Table 3). The gene that was most induced by both cytokines was MX1 which was increased 167.0and 183.4-fold by IL-28A and IL-29, respectively. Furthermore, IL-28A stimulation resulted in an at least two-fold decreased expression of a multitude of genes (272 genes for p,0.01 without multiple testing adjustment; 89 genes for adj.p,0.05) with up to 9.3-fold reduced levels (Table 4).
We then analyzed the induced and repressed genes for specific enrichment of defined biological processes and molecular functions using the Panther database [28]. The genes activated by IL-28A and IL-29 comprised genes of immunity and defense Amongst the downregulated genes following IL-28A stimulation, those involved in the biological process of mRNA transcription regulation were especially enriched (p = 1.2610 26 ; Table 5; Fig. S2A). Additionally, analysis revealed a downregulation in the molecular function class of nucleic acid binding proteins (p = 1.3610 27 ) with its subclass ''other DNA-binding proteins'' (p = 2.5610 29 ; Table 5; Fig. S2B). Moreover, homeobox transcription factors, a subclass of the transcription factor group, were significantly enriched (p = 2.3610 27 ; Table 5 and Fig. S2B).
In contrast, IL-29 reduced gene expression of only four genes more than two-fold (p,0.01; data not shown). However, when corrected for multiple testing, none of these regulations remained significant (adj.-p.0.05; data not shown).
In addition, we analyzed gene expression after three hours of stimulation with IL-28A or IL-29, respectively. The overall number of genes regulated more than two-fold was much lower than after 6 hours (IL-28A: 16 genes up-regulated, 0 down-

Validation of the microarray data by quantitative PCR and luciferase assay
In the next set of experiments, we verified the IFN-l induced gene expression in naïve Huh-7 cells by quantitative RT-PCR in an independent set of RNA samples at different time points. In this analysis, we included the most strongly induced genes for both cytokines (MX1; Table 3) and additionally analyzed expression of OAS1, coding for another important antiviral protein (29,59-OAS) regulated by IL-28A and IL-29 (32.2-and 41.0-fold, respectively; Table 3).
The results are depicted in Figure 3 and show a significant increase in OAS1 mRNA expression of more than 2800-fold following IL-28A and more than 5000-fold by IL-29 stimulation after 24 hours (Fig. 3A). MX1 mRNA expression was induced up to 95-fold by IL-28A and 78-fold by IL-29 (Fig. 3B). While MX1 reached maximal expression levels after 9 hours, OAS1 was induced most strongly after 24 hours (Fig. 3A, B).
To analyze possible additive/synergistic effects of IL-28A and IL-29, Huh-7 cells were stimulated with either cytokine alone (at a concentration of 100 ng/ml) or with a combination of both cytokines together (50 ng/ml each and 100 ng/ml each). As shown in Figure 3C, no significant difference was observed between the different treatments.
We next aimed to confirm our data in another hepatic cell line (HepG2) using a luciferase promoter assay as an additional experimental approach to determine the influence of IFN-ls on the transcriptional regulation of these two antiviral genes. Promoter activity of a human -970 nt OAS1 promoter-luciferase construct and of a 2553/+10 human MX1 promoter-luciferase construct were examined, following incubation of HepG2 cells with 100 ng/ml IL-28A and IL-29. OAS1 promoter activity was significantly stimulated 5.1-fold and 14.3-fold, respectively, above baseline by treatment with IL-28A and IL-29 for 6 hrs ( Fig. 3D; p,0.0005). MX1 promoter activity was increased 2.4-fold and 4.8-fold by IL-28A and IL-29, respectively ( Fig. 3D; p,0.05).

IFN-ls decrease HCV replication in vitro
To investigate whether the activation of genes encoding the antiviral proteins 29,59-OAS and MX1 results in antiviral activity in vitro, we analyzed the effect of IL-28A and IL-29 on the HCV replication rate in HCV replicon expressing Huh-7 cells. In these experiments, both IL-28A and IL-29 (100 ng/ml) significantly decreased HCV replication in Huh-7 cells by 84.8% and 87.7%, Table 3. Overview of the 20 genes whose expression was most strongly induced (p,0.01) by IL-28A and IL-29 in Huh-7 cells after 6 hours of stimulation.  Fig. 4). At a cytokine concentration of 10 ng/ml, IL-29 had significantly stronger inhibitory effects on HCV replication (reduction of 83.4%) than IL-28A (reduction of 70.1%; p,0.005).

IL-28A and IL-29 mRNA and protein expression is increased in the liver tissue and serum of patients with HCV infection
Having shown that IFN-ls inhibit HCV replication, we next analyzed IFN-l expression in viral infection in vivo. First, we Table 4. Overview of the 20 genes whose expression was most strongly repressed (p,0.01) by IL-28A in Huh-7 cells after 6 hours of stimulation.  Table 5. Functional classification of IL-28A and IL-29-induced and repressed genes regarding the categories of biological processes and molecular functions applying the Panther software [28].   Fig. 5A) and was detectable in 100% (9/9) of HCV biopsies but in only 58% (11/19) of biopsies from non-HCV hepatitis (p,0.05; Fig. 5B). Similarly, IL-29 mRNA expression was highest in HCV patients (26.4-fold vs. non-HCV liver diseases, p,0.05; Fig. 5A) and was detectable in 78% (7/9) of HCV biopsies but in only 63% (12/19) of biopsies from non-HCV hepatitis (Fig. 5B). In all biopsies, IL-28 and IL-29 mRNA expression correlated highly with each other (r = 0.892). We then measured IL-28A serum protein concentration using an immunoluminometric assay (ILMA) in another group of liver disease patients, each comprising 15 patients with HCV or HBV infection, primary sclerosing cholangitis (PSC), and 24 patients with primary biliary cirrhosis (PBC) as well as 15 controls. IL-28A serum protein levels were significantly higher in patients with viral infection (mean concentrations of 52.2 pg/ml in HCV and 46.3 pg/ml in HBV patients) in comparison to non-viral liver diseases such as PBC (mean concentration 14.9 pg/ml; p,0.01 vs. HCV/HBV), PSC (mean concentration 21.2 pg/ml; p,0.05 vs. HCV/HBV) or a control group (mean concentration 23.6 pg/ml; p,0.05 vs. HCV/HBV) (Fig. 5C).

IL-28 mRNA expression is increased in the liver of murine cytomegalovirus (MCMV)-infected mice
Given the current lack of a simple murine model for HCV infection [29] and in order to analyze if the up-regulation of IFNls in vivo can be found in viral liver disease other than HCV and HBV, we studied IFN-l expression in murine cytomegalovirus (MCMV) infection, an established model of murine viral hepatitis [30,31]. Given that no IL-29 gene is known in mice, we solely determined IL-28 mRNA expression levels 45 hours after infection. Compared to non-infected mice (n = 4), IL-28 mRNA expression was 2.7-fold higher in MCMV-infected mice (n = 10; p,0.005; Fig. 6).

Discussion
This study represents the first detailed comparative investigation of IL-28A-and IL-29-mediated biological activities and gene expression patterns in viral hepatitis and non-viral liver disease in vivo. We demonstrate that the IFN-l receptor complex is functionally expressed in liver cells while IFN-lR1 is expressed only at low levels in leukocytes. Both IL-28A and IL-29 are able to induce STAT1 phosphorylation in hepatic cells with IL-29 showing slightly stronger effects. Our microarray analysis revealed activation of mostly identical genes by IL-28A and IL-29. Among them were numerous genes involved in interferon-mediated immunity and antiviral defense, such as MX1 [32], OAS1, OAS3 and OASL [33], BST2 (tetherin) [34], inhibitors of protein synthesis such as PKR [35], IFIT1 and IFIT2 [36], antiproliferative genes like IFITM1 [37] and the Bcl-2-related proapoptotic gene APOL6 [38], TAP1 (involved in antigen processing and presentation) [39], as well as other interferon-stimulated genes such as ISG15 (ubiquitin-like modifier) [33] or IFI6 (6-16 protein) [40]. A recent study analyzing the effect of IL-29 in HepG2 cells revealed 35 upregulated genes following IL-29 stimulation [41]. Among these, 27 were found in our analysis of IL-28A as well as IL-29 confirming the similarity between these two cytokines.
The analogy of IL-28A and IL-29 is further supported by the fact that we observed no synergistic or additive effect of both cytokines concerning the induction of gene expression. This suggests common signaling pathways and functions of these cytokines. Moreover, the dose of IFN-l used in this study (100 ng/ ml) seems to be a saturation concentration for this cytokine.
IFN-ls also increased mRNA expression of STAT1, STAT2 and IRF9 whose protein products together form the ISGF3 transcription factor complex characteristic for IFN type I and type III signaling. ISGF3 in turn activates a number of IFN-stimulated genes (ISG), thereby contributing to the antiviral response. Similarly, a recent study showed that IL-29 stimulation also leads to increased levels of total STAT1 and STAT2 protein and hence to a prolonged induction of target genes [14].
Three additional genes with relevant immunological functions which were up-regulated by IFN-ls included TLR3, MDA5 (IFIH1) and RIG-I (DDX58). The proteins belong to the pattern recognition receptors (PRRs) of the innate immune system and bind specifically extracellular derived dsRNA (TLR3) and cytoplasmic viral RNA (MDA5, RIG-I) [42]. It has recently been demonstrated that TLR3 ligands mediate an antiviral state against HCV in hepatic cells [43] and induce antiviral activity of IFN-ls [44]. RIG-I is likewise important for the antiviral state in HCV infection [45]. On the other hand, HCV is able to inhibit several PRR pathways [46], suggesting that up-regulation of PRR mRNA expression by IL-28A and IL-29 might counteract this HCVmediated effect. In our study, IL-28A and IL-29 up-regulated a after 9 hours of stimulation as determined by qPCR and normalized to untreated cells. (C) IL-28A and IL-29 do not act synergistically. OAS1 and MX1 gene induction by IL-28A and IL-29 alone (100 ng/ml) was not significantly different from treatment with both cytokines together (50 or 100 ng/ml each) in Huh-7 cells. (D) Reporter gene assays and OAS1 luciferase construct revealed activation of the OAS1 promoter 5.1-fold (IL-28A) and 14.3-fold (IL-29), respectively, and the MX1 promoter 2.4-fold (IL-28A) and 4.8-fold (IL-29), respectively, in HepG2 cells following stimulation for 6 hours with 100 ng/ml IL-28A or IL-29, respectively. IFN-a was used as a positive control. Baseline reporter gene activity was set as 1. IL-28A/IL-29 induced reporter gene activity in all other groups was calculated as -fold increase in comparison to this control group. *p,0.05; **p,0.0005 vs. control. doi:10.1371/journal.pone.0015200.g003 nearly identical gene transcription program which also resembles that of IFN-a [41]. However, it has recently been demonstrated that neither IFN-a nor IL-29 are able to down-regulate gene expression in hepatic cells [16,17,41]. In concordance with these previous studies [16,17,41], we measured no significant downregulation of genes following IL-29 stimulation.
Therefore, it is of great interest that in contrast IL-28A significantly reduced the expression of 89 genes more than 2-fold (adj.-p,0.05) in our experiments. This number was even higher than the number of induced genes (65, adj.-p,0.05). Many of these genes code for DNA-binding proteins and are involved in the transcriptional regulation.
The activation of antiviral proteins by IL-28A and IL-29 tempted us to investigate the effect of IFN-ls on the replication rate of HCV in an in vitro system expressing HCV replicons. In these experiments, we demonstrated that both, IL-28A and IL-29 at a dose of 100 ng/ml reduce significantly the replication rate of HCV with the same efficacy and comparable to IFN-a. However, at a concentration of 10 ng/ml, IL-29 is 20% more effective in inhibiting HCV replication than IL-28A (p,0.005). Currently, a pegylated form of IL-29 is tested in a phase 1b clinical study in HCV patients [47]. Preliminary results indicate that it is effective in reducing viral load without typical side effects seen with IFN-a [47] which may be related to the more restricted expression of the IFN-l receptor subunit IFN-lR1 compared to the IFN-a receptor subunits IFNAR1 and IFNAR2.
A recent study demonstrated that IL-28B appears to be the most potent IFN-l cytokine, at least in EMCV infection [48]. However, in VSV infection, IL-28B did not show any effect [48]. Moreover, there were also considerable differences in specific activities between the same cytokines derived from different sources [48] indicating that the production and preparation methods are crucial variables. Interestingly, several recent publications describe an association between single nucleotide polymorphisms (SNPs) in the IL28B gene region and the clearance of HCV infection, either naturally occurring [49] or induced by treatment with a combination therapy of IFN-a and ribavirin [50,51,52]. Some of these SNPs are located in the IL28B gene itself while others are situated upstream or downstream of IL28B [16,17,41,49]. The functional consequences of the SNPs in the IL28B gene region are not yet clear and need further investigations. Given that the IL28A and IL28B genes lie in close proximity on chromosome 19q13.13, it is possible that some of these SNPs influence regulatory elements of both IL28A and IL28B [51]. This is supported by data demonstrating lower IL-28A/B mRNA expression in whole blood and PMBCs, respectively, in minor allele carriers and nonresponders to IFN-a therapy [51,52]. As IL28B mRNA is 98% identical to IL28A mRNA and can hardly be distinguished from the latter by PCR analysis, it cannot be excluded that IL-28A also plays a major role in HCV viral clearance.
Furthermore, in our study, we measured increased expression of IL-28 and IL-29 mRNA in the liver of patients infected with HCV in comparison to non-viral liver disease. IFN-l mRNA was detectable in only 60.5% of the biopsies of non-viral liver disease but in 88.9% of the livers with HCV infection suggesting an essential role of HCV in the regulation of IFN-l gene transcription. Further studies need to determine if the IFN-l upregulation has a significant influence on the clinical presentation and the outcome of HCV infection. It may be hypothesized that patients with higher intrinsic IFN-l expression show a better HCV clearance. This hypothesis is supported by the fact that lower IL-28A/B mRNA expression has been observed in non-responders to IFN-a therapy [51,52]. Moreover, HCV patients with low endogenous IFN-l expression might benefit more from a novel treatment with pegylated IL-29 than those with high IFN-l levels.
The up-regulation of IFN-l gene expression has also been described for other viral infections [1,2,8,15,53]. In contrast, the study of Mihm et al. shows that IFN-l expression in the liver is similar in non-viral liver disease and HCV infection [54]. This difference to our study might be due to their small sample size in the non-viral liver disease group (8 samples) which included one outlier. However, there was a higher expression in HCV infection in comparison to healthy control liver tissue [54]. Moreover, they found higher IFN-l mRNA expression in PBMC of HCV patients in comparison to healthy controls [54]. To our knowledge, IL-28A protein concentration has not been measured previously in the serum of HCV patients. Therefore, we developed an immunoluminometric assay which detected significantly higher IL-28A protein expression levels in the serum of HCV-or HBV-infected patients in comparison to healthy controls, but also in comparison to non-viral liver disease such as PBC. This suggests that IL-28A does not only have a ''local'' liver-specific role in the antiviral defense but also modulates the systemic antiviral immune response against HCV.
In further studies, it will be of great interest if the IL-28Bmediated gene expression and repression in hepatic cells resembles the pattern of IL-28A or IL-29 or is even different to both origin (HUO; n = 9) as determined by quantitative PCR. IL-29 expression was 26.4-fold higher in HCV vs. non-HCV biopsies (*p,0.05 vs. any other group). (B) Detailed analysis of each single biopsy reveals expression of total IL-28 in all 9 out of 9 HCV patients ( = number marked by an asterisk) while in 8 out of 19 non-HCV biopsies, IL-28 could not be detected after 40 PCR cycles ({ = IL-28 not detectable; p,0.05). IL-29 was expressed in 7 out of 9 HCV biopsies and in 12 out of 19 non-HCV biopsies (# = IL-29 not detectable). (C) Analysis of IL-28A serum levels in liver disease patients and controls by an IL-28A-specific ILMA demonstrated significantly higher IL-28A protein expression in the sera of HBV and HCV patients in comparison to controls or PBC patients (*p,0.05; n = 15 in each group [except PBC: n = 24]). doi:10.1371/journal.pone.0015200.g005 cytokines. Further investigations should elucidate if the different abilities of IL-28A and IL-29 to repress gene transcription have functional consequences in HCV infection in vivo as the replicon system represents only one single aspect of HCV life cycle. In addition, it will be of interest if these differences have practical impact in other viral infections in vivo, but also for the treatment of other diseases such as cancer. Given their ability to inhibit proliferation and to induce apoptosis [8,55], IFN-ls have been also been discussed as a future cancer treatment option.
In summary, we have shown that both IL-28A and IL-29 induce expression of antiviral proteins, inhibit HCV replication and are up-regulated during viral infection with no major differences. However, in contrast to IL-29, IL-28A has the capacity to repress gene expression. Both cytokines are promising candidates for the treatment of HCV infection with likely low side effects on leukocytes. Nevertheless, further studies are needed to clarify which of the three IFN-l cytokines is the most potent with the least amount of side effects.  Figure S1A). IL-29 did not down-regulate genes significantly (data not shown). (B) The molecular functions of IL-28A down-regulated genes comprise mainly of nucleic acid binding proteins and of homeobox transcription factors (for color chart legend, see Figure S1B). (TIF)