Interferon-lambda (IFN-λ) induces signal transduction and gene expression in human hepatocytes, but not in lymphocytes or monocytes

Interferon-alpha and interferon-lambda induce distinct functional responses in human hepatocytes, lymphocytes, and monocytes.

Abstract

This study compared the ability of IFN-α and IFN-λ to induce signal transduction and gene expression in primary human hepatocytes, PBLs, and monocytes. IFN-α drug products are widely used to treat chronic HCV infection; however, IFN-α therapy often induces hematologic toxicities as a result of the broad expression of IFNARs on many cell types, including most leukocytes. rIFN-λ1 is currently being tested as a potential alternative to IFN-α for treating chronic HCV. Although IFN-λ has been shown to be active on hepatoma cell lines, such as HepG2 and Huh-7, its ability to induce responses in primary human hepatocytes or leukocytes has not been examined. We found that IFN-λ induces activation of Jak/STAT signaling in mouse and human hepatocytes, and the ability of IFN-λ to induce STAT activation correlates with induction of numerous ISGs. Although the magnitude of ISG expression induced by IFN-λ in hepatocytes was generally lower than that induced by IFN-α, the repertoire of regulated genes was quite similar. Our findings demonstrate that although IFN-α and IFN-λ signal through distinct receptors, they induce expression of a common set of ISGs in hepatocytes. However, unlike IFN-α, IFN-λ did not induce STAT activation or ISG expression by purified lymphocytes or monocytes. This important functional difference may provide a clinical advantage for IFN-λ as a treatment for chronic HCV infection, as it is less likely to induce the leukopenias that are often associated with IFN-α therapy.

Introduction

IFNs are a diverse family of Class 2 cytokines that are characterized, in part, by their ability to induce antiviral activity in receptor-bearing target cells. IFNs are divided into three subtypes, denoted Types I, II, and III, based, in part, on their differential use of specific receptors to mediate signal transduction [1]. Type I IFNs include IFN-α, -β, -ε, -κ, and -ω, and all Type I IFNs signal through the IFNAR complex [2]. There is only one Type II IFN, known as IFN-γ, that signals through the IFN-γR complex [3]. The most recent addition to the IFN family, the Type III IFNs, includes three members: IFN-λ1, -λ2, and -λ3 [4, 5]. These cytokines are also known as IL-29, IL-28A, and IL-28B, respectively. The Type III IFNs signal through a distinct heterodimeric receptor complex composed of the ligand-binding chain, IFN-λR1 (also known as IL-28Rα), and the accessory chain, IL-10R2 [4–6].

One of the primary biological activities mediated by IFNs is their ability to induce antiviral activity in a wide variety of target cells. The antiviral activity of IFN is mediated via the action of multiple proteins that are encoded by a set of ISGs [7, 8]. The potent antiviral activity of the Type I IFNs, particularly IFN-α, has been exploited clinically, and there are several IFN-α drug products that are widely used to treat chronic viral infections, including HBV and HCV infections [9]. Type I IFNARs are broadly expressed on most cell types and tissues, including hepatocytes and most types of leukocytes. As a consequence, IFN-α can bind to many different cell types throughout the body and induces biological responses in multiple tissues. For example, in addition to liver cells, hematologic cell types, such as neutrophils and lymphocytes, express IFNARs, and IFN-α therapy often induces neutropenia and lymphopenia in a large percentage of patients with chronic HCV infection.

In 2003, we reported the discovery and characterization of the Type III IFNs, IFN-λ1, -λ2, and -λ3 [4]. The same three genes and corresponding proteins were codiscovered simultaneously by another group and designated IL-29, IL-28A, and IL-28B, respectively [5]. Both groups also identified and characterized the receptor for this subgroup of IFNs and showed that like Type I IFNs (IFN-α/β), the Type III IFNs induce antiviral protection in cells that express IFN-λRs. However, unlike the receptors for Type I IFNs, which are broadly expressed on virtually all cell types, IFN-λRs exhibit a more restricted tissue distribution [10, 11]. Several previous studies have shown that like IFN-α, IFN-λ inhibits replication of partial and full-length HCV replicons in the human hepatoma cell line, Huh7 [12–14]. However, relatively little is known regarding the biological effects of IFN-λ on primary human hepatocytes or leukocytes. We examined the ability of IFN-α and IFN-λ to activate the Jak/STAT signaling pathway in mouse and human hepatocytes. We also evaluated the ability of IFN-α and IFN-λ to regulate expression of various ISGs in human hepatocytes and compared the ability of IFN-α and IFN-λ to induce activation of STAT1 and STAT2 in hepatocytes, PBLs, and monocytes.

MATERIALS AND METHODS

Reagents

Purified mouse rIFN-λ2 (Catalog No. 4635-ML) and human rIFN-λ1 (Catalog No. 1598-IL), -λ2 (Catalog No. 1587-IL), and -λ3 (Catalog No. 5259-IL) proteins were obtained from R&D Systems (Minneapolis, MN, USA). Purified mouse rIFN-α1 (Catalog No. 12,105-1) and human rIFN-α2a (Catalog No. 11,101-1) proteins were obtained from PBL Interferon Source (Piscataway, NJ, USA). Rabbit anti-phospho(Y)-STAT1 and anti-phospho(Y)-STAT2 antibodies were obtained from Santa Cruz Biotechnology (Santa Cruz, CA, USA).

Isolation and culture of mouse and human hepatocytes

The experiments with mouse hepatocytes were performed using cultures of primary murine hepatocytes derived from the livers of normal C57BL/6 mice. The cells were isolated and cultured as described previously [15]. The isolated mouse hepatocytes were cultured at 80–90% confluence in HepatoZYME-SFM media (Gibco, Life Technoloiges, Grand Island, NY, USA) on collagen-coated plates for 24 h prior to use. The experiments with human hepatocytes were performed using cultures of primary human hepatocytes derived from the livers of healthy, anonymous donors under IRB-approved protocol No. HSR10-008. Human hepatocytes were isolated by a two-step perfusion technique as described previously [16, 17]. After isolation, the cells were resuspended in Ham's F-12 medium containing 2% FBS and plated on collagen-coated plastic dishes at a density of 1 × 10⁵ cells/cm². The cells were cultured overnight at 37°C before use.

Isolation and culture of human lymphocytes and monocytes

Normal human PBLs and monocytes were obtained from healthy volunteers who provided written, informed consent to participate in this IRB-approved study for the collection of blood for in vitro research use. This protocol (No. 99-CC-0168) was approved by the NIH Office of Human Subjects Research (Bethesda, MD, USA). Lymphocytes and monocytes were isolated by countercurrent centrifugal elutriation in a Beckman JE-6B elutriator as described previously [18]. The complete medium used for culturing lymphocytes and monocytes consisted of RPMI-1640 medium (Gibco, Carlsbad, CA, USA), supplemented with 10% FCS (HyClone, Logan, UT, USA), 2 mM L-glutamine, and 50 μg/ml gentamycin. The elutriated lymphocytes and monocytes were cultured at 5 × 10⁶ cells/ml in complete medium in round-bottomed polypropylene tubes.

Northern blots

Total RNA was isolated from cultured hepatocytes using RNAzol B (Tel-Test, Friendswood, TX, USA) by the acid/guanidinium thiocyanate/phenol/chloroform extraction method, as described previously [19]. Equivalent amounts of RNA (10 μg/lane) were size-fractionated by electrophoresis in 1% agarose gels containing 0.66 M formaldehyde. The RNA was then transferred onto Nytran membranes and cross-linked by exposure to UV light. The membranes were then hybridized and washed according to standard procedures. The cDNA probes used to detect ISG expression have been described previously. Gel-purified insert DNA was radiolabeled by the random-primer method of Feinberg and Vogelstein [20].

qPCR arrays

We used pathway-focused RT² Profiler PCR arrays from SABiosciences (Frederick, MD, USA) to identify genes whose expression is induced by IFN-λ treatment in primary mouse and human hepatocytes. These arrays are low-density arrays of qPCR primers that amplify a defined set of 96 genes/array. Total cellular RNA was isolated and DNAse-treated to remove any residual genomic DNA using RT² qPCR Grade RNA isolation kits from SABiosciences, and 1 μg of the resulting total RNA was used as a template for synthesis of first-strand cDNAs using ReactionReady First Strand cDNA synthesis kits (SABiosciences). Thermal cycling using SuperArray RT² Real-Time SYBR Green PCR Master Mix with real-time detection by SYBR Green and 5-carboxy-carboxyl-X-rhodamine (ROX) dyes was performed on the Mx3000P instrument (Stratagene/Agilent Technologies, Santa Clara, CA, USA), per the kit manufacturer's instructions. Analysis of the PCR array gene expression data was carried out using the data analysis web portal provided by SABiosciences (http://www.sabiosciences.com/pcrarraydataanalysis.php).

RT-PCR profiling

Verification of ISG expression was performed by end-point RT-PCR expression profiling using standard methods after first-strand cDNA synthesis from 1 μg purified cellular RNA as described above. Kits containing lyophilized primers for 11 ISGs and a housekeeping gene (MultiGene-12 RT-PCR profiling kit, SABiosciences) were used for gene expression analysis in ReactionReady HotStart Taq polymerase reagent using a PTC-100 thermal cycler with a heated lid (MJ Research, Watertown, MA, USA), per the kit manufacturer's protocol. The PCR products were visualized by standard techniques using agarose gel electrophoresis.

qRT-PCR

Verification and quantification of changes in gene expression identified using PCR arrays were carried out by qRT-PCR analyses of individual ISGs. Total cellular RNA was isolated and treated with DNase, and 1 μg of the purified RNA from each treatment group was reverse-transcribed to generate first-strand cDNA, as described above. Specific primer assays for selected ISGs were obtained from SABiosciences and analyzed on the Mx3000P system. PCR amplification was performed using the SABioscience master mix in 25-μl reactions as described above. Changes in gene expression levels were analyzed using MxPro software v.4.10 (Stratagene/Agilent), and the results are expressed as the mean-fold increase relative to the control gene expression levels normalized to the housekeeping gene, GAPDH. Graphing and statistical analysis of qPCR results were performed using Prism 5.0 (Graph Pad Software, San Diego, CA, USA). The data were analyzed using a two-tailed Student's t-test. Values represent the mean ± sd of triplicate determinations. P values <0.05 were considered statistically significant. All experiments were repeated at least three times with similar results.

Western blots

STAT activation was examined by measuring the levels of pY-STAT1 and -STAT2 following treatment with IFN-α or IFN-λ for 30 min at 37°C. The levels of pY-STAT1 and -STAT2 were measured by Western blotting as described previously [21]. After cytokine treatment, the cells were washed three times with Dulbecco's PBS, and whole-cell lysates were prepared. Total STAT1 or STAT2 protein was immunoprecipitated using rabbit anti-STAT1 or -STAT2 antibodies, respectively. Immunoprecipitated proteins were resolved by electrophoresis on 8% SDS-PAGE gels (Invitrogen, Carlsbad, CA, USA) and then transferred to PVDF membranes. The levels of pY-STAT1 and -STAT2 were measured by ECL using rabbit antiphospho-Tyr⁷⁰¹-STAT1 (Cat. #9171L; Cell Signaling Technology, Beverly, MA, USA) and rabbit antiphospho-Tyr⁶⁸⁹-STAT2 (Cat. #07-224; Millipore, Billerica, MA, USA) antibodies, respectively.

EMSAs

Nuclear protein extracts were prepared from cytokine-treated cells by using a modification [22] of the original method described by Dignam et al. [23]. A dsISRE oligonucleotide probe, based on a DNA sequence present in the promoter of the human ISG15 gene, was used as a probe in the gel-shift assays [24]. Binding reactions were performed as described previously [21, 22]. A portion of each binding-reaction mixture (8 μl/sample) was electrophoresed on nondenaturing, 6% polyacrylamide gels (Invitrogen) by using 0.25 Tris-borate-ethylenediamine-N,N, N′,N′-tetraacetic acid buffer (22 mM Tris-HCl, pH 8.0/22 mM borate/0.5 mM ethylenediamine-N,N, N′,N′-tetraacetic acid). The gels then were dried and visualized by autoradiography.

RESULTS

IFN-λ induces expression of ISGs in primary mouse hepatocytes

We reported previously that murine rIFN-λ induces Jak/STAT signaling and antiviral protection in mouse keratinocytes but not in fibroblasts, splenocytes, or bone marrow-derived macrophages [25]. To determine if murine hepatocytes can respond to IFN-λ, we prepared cultures of freshly isolated mouse hepatocytes and treated these cells with murine rIFN-λ2 (100 ng/mL) for 24 h. We then prepared total RNA extracts from these cells and measured expression of ISGs by qRT-PCR using 96-well pathway-focused RT² Profiler PCR arrays. The fold-increase values for expression of individual ISGs was determined by comparing gene expression levels of the IFN-λ-treated cells with that of the control nontreated cells. As shown in Table 1, IFN-λ induced significant increases in the expression levels of many ISGs in murine hepatocytes. These ISGs included chemokine genes, such as Cxcl9 (monokine induced by IFN-γ), Cxcl10 (IP-10), and Cxcl11 (IFN-inducible T-cell α-chemoattractant), as well as antiviral genes, such as Mx1 (IFN-induced GTP-binding protein), Oas1a, and Isg15. IFN-λ treatment also up-regulated expression of the Stat1 and Stat2 genes in mouse hepatocytes. Increased expression of Stat1 and Stat2 may serve to amplify the response to IFN-λ, as the proteins encoded by these genes are essential components of ISGF3 transcription factor complexes.

Table 1. Changes in ISG Expression Levels in Murine Hepatocytes after Treatment with IFN-λ for 24 h at 37°C.

Gene symbol	RefSeq	Fold increase	Gene description
IFIT3	NM_010501	291.03	IFN-induced protein with tetratricopeptide repeats 3
CXCL9	NM_008599	90.35	Chemokine (C-X-C motif) ligand 9
IFIT1	NM_008331	86.52	IFN-induced protein with tetratricopeptide repeats 1
IFI44	NM_133871	76.37	IFN-induced protein 44
MX1	NM_010846	58.69	Myxovirus (influenza virus) resistance 1
ISG15	NM_015783	56.49	IFN-stimulated gene 15 ubiquitin-like modifier
CXCL11	NM_019494	39.88	Chemokine (C-X-C motif) ligand 11
CXCL10	NM_021274	39.33	Chemokine (C-X-C motif) ligand 10
IFI204	NM_008329	32.79	IFN-activated gene 204
IFIH1	NM_027835	24.68	IFN induced with helicase C domain 1
IL13RA2	NM_008356	24.50	IL-13R, α2
CXCL13	NM_018866	15.11	Chemokine (C-X-C motif) ligand 13
FcGR1	NM_010186	13.00	FcR, IgG, high-affinity I
PTPRC	NM_011210	11.39	Protein tyrosine phosphatase, receptor type, C
OAS1A	NM_145211	9.19	2′-5′-Oligoadenylate synthetase 1A
STAT1	NM_009283	8.94	STAT1
GDF5	NM_008109	8.21	Growth differentiation factor 5
CCL4	NM_013652	6.53	Chemokine (C-C motif) ligand 4
CCL9	NM_011338	6.31	Chemokine (C-C motif) ligand 9
CCR9	NM_009913	6.01	Chemokine (C-C motif) receptor 9
CCL7	NM_013654	5.23	Chemokine (C-C motif) ligand 7
IRGM	NM_008326	5.12	Immunity-related GTPase family, M
STAT2	NM_019963	4.99	STAT2
CCL17	NM_011332	4.72	Chemokine (C-C motif) ligand 17
IFIT2	NM_008332	4.68	IFN-induced protein with tetratricopeptide repeats 2
IL13	NM_008355	4.49	IL-13
MMP3	NM_010809	4.29	Matrix metallopeptidase 3
CXCL5	NM_009141	4.25	Chemokine (C-X-C motif) ligand 5
IRF7	NM_016850	4.18	IFN regulatory factor 7
SOCS4	NM_080843	4.14	Suppressor of cytokine signaling 4
IFI35	NM_027320	4.10	IFN-induced protein 35
IL1A	NM_010554	4.02	IL-1α

We confirmed the induction of selected ISGs by IFN-λ in mouse hepatocytes by standard RT-PCR. As shown in Supplemental Fig. 1, treatment with IFN-α or IFN-λ2 induced or increased expression of a number of ISGs in mouse hepatocytes, including Cxcl10, Ifit1 (ISG56), Mx1, and Mx2. IFN-α and IFN-λ also amplified the basal expression of several other ISGs, including Ifi35, Irf7, Irf9, and Isg15. Expression of these ISGs was sustained in IFN-α- and IFN-λ-treated cells for at least 24 h.

Induction of ISG expression by IFN-λ in murine hepatocytes is STAT1-dependent and correlates with activation of ISGF3

To examine the molecular basis for induction of ISG expression by IFN-λ, we compared the ability of IFN-α, IFN-λ, and IFN-γ to induce ISGF3 activity in primary mouse hepatocytes. We treated mouse hepatocytes with murine rIFN-α (Type I IFN), IFN-λ2 (Type III IFN), IFN-γ (Type II IFN), or IL-22 for 30 min at 37°C and then prepared nuclear protein extracts. The nuclear extracts were assayed for ISGF3 activity by EMSA using an ISRE oligonucleotide probe that binds ISGF3 with high affinity [24]. As shown in Fig. 1A, IFN-α and IFN-λ induced ISGF3 activity in murine hepatocytes. IFN-γ also induced weak ISGF3 activity in hepatocytes. In contrast, although murine hepatocytes express IL-22Rs [26], IL-22 did not induce ISGF3 activity in these cells. The ability of IFN-α and IFN-λ to induce ISGF3 activity in mouse hepatocytes correlated with their ability to induce ISG expression, as determined by Northern blot analysis. As shown in Fig. 1B, treatment with IFN-α or IFN-λ markedly up-regulated expression of all five of the ISGs that we examined: Mx1, Oas1g, Irf7, Ddx58 (RIG-I), and Dhx58 (LGP2).

Figure 1. Induction of ISG expression by IFN-α or IFN-λ in murine hepatocytes correlates with activation of ISGF3.

(A) Cultures of freshly isolated murine hepatocytes were incubated with medium alone (control) or murine rIFN-α, IFN-λ2, IFN-γ, or IL-22 for 30 min at 37°C. At the end of this incubation period, nuclear protein extracts were prepared and analyzed by EMSA for ISGF3 activity using a radiolabeled ISRE probe. (B) Identical cultures of murine hepatocytes were treated with IFN-α (10 ng/mL) or IFN-λ (100 ng/mL) for 6 h at 37°C. At the end of this incubation period, RNA extracts were prepared and analyzed by Northern blotting using radiolabeled probes specific for ISGs Mx1, Oas1g, Irf7, Ddx58 (Rig-I), and Dhx58 (Lgp2) and the housekeeping gene, Gapdh. (C) Primary cultures of freshly isolated murine hepatocytes were prepared from control WT mice and STAT1 knockout mice. The cells were incubated with medium alone (control), IFN-α, IFN-β, IFN-γ, or IFN-λ for 6 h at 37°C. At the end of this incubation period, RNA extracts were prepared and analyzed by Northern blotting using radiolabeled probes specific for several ISGs, including Mx1, Oas1g, Irf7, Ddx58 (Rig-I), and Dhx58 (Lgp2) and the housekeeping gene, Gapdh.

Signaling through Type I (IFN-α/β) or Type III (IFN-λ) IFNRs catalyzes formation of ISGF3 complexes. Once formed, ISGF3 translocates to the nucleus and binds to ISRE elements in the promoters of IFN-responsive genes. We compared the ability of IFN-α, -β, -γ, and -λ to induce ISG expression in hepatocytes derived from control (WT) mice and STAT1 knockout mice [27, 28]. As shown in Fig. 1C, IFN-α, -β, -γ, and -λ up-regulated expression of the five ISGs that we examined: Mx1, Oas1g, Irf7, Ddx58 (RIG-I), and Dhx58 (LGP2). The ability of IFNs to induce expression of the same five genes was reduced markedly or abrogated in hepatocytes derived from STAT1 knockout mice. These findings demonstrate that STAT1 is essential for induction of hepatocyte ISG expression by IFN-λ as well as by Type I (IFN-α/β) and Type II (IFN-γ) IFNs.

IFN-λ induces expression of ISGs in primary human hepatocytes

To determine if the ability of IFN-λ to induce ISG expression in murine hepatocytes is also demonstrable in human hepatocytes, we treated cultures of primary human hepatocytes with human rIFN-α or rIFN-λ1 (IL-29) for 24 h. We then prepared total RNA extracts and measured expression of multiple ISGs using 96-well, pathway-focused RT² Profiler PCR arrays. The fold-increase in expression of individual ISGs was determined by comparing the gene expression levels of the IFN-λ1-treated cells with that of the control nontreated cells. As shown in Table 2, IFN-α and IFN-λ induced significant increases in the expression levels of many ISGs in human hepatocytes. These genes included a number of classical ISGs, such as MX1, MX2, OAS1, OAS2, ISG15, and ISG20. Similar to the response of murine hepatocytes, IFN-λ also up-regulated expression of the STAT1, STAT2, and IRF9 (ISGF3γ) genes in primary human hepatocytes. The levels of ISG expression induced by IFN-α in hepatocytes were generally greater than the levels induced by IFN-λ, despite the fact that the cells were treated with equivalent concentrations of each cytokine.

Table 2. Changes in ISG Expression Levels in Human Hepatocytes after Treatment with IFN-λ1 or IFN-α2a for 24 h.

Gene symbol	RefSeq	Fold increase		Gene description
		IFN-λ1	IFN-α2a
IFIT1	NM_001548	64.53	241.52	IFN-induced protein with tetratricopeptide repeats 1
CXCL10	NM_001565	54.27	795.65	Chemokine (C-X-C motif) ligand 10
MX2	NM_002463	29.90	107.34	Myxovirus (influenza virus) resistance 2 (mouse)
ISG15	NM_005101	29.49	121.60	IFN-stimulated gene 15 ubiquitin-like modifier
IFITM1	NM_003641	25.67	89.02	IFN-induced transmembrane protein 1 (9–27)
IFIT3	NM_001549	17.53	65.62	IFN-induced protein with tetratricopeptide repeats 3
IFI6	NM_002038	16.36	28.36	IFN, α-inducible protein 6
MX1	NM_002462	12.23	17.83	Myxovirus (influenza virus) resistance 1
OAS1	NM_002534	9.79	20.76	2′,5′-Oligoadenylate synthetase 1, 40/46 kDa
IFIH1	NM_022168	8.89	24.18	IFN induced with helicase C domain 1
OAS2	NM_002535	7.53	12.60	2′,5′-Oligoadenylate synthetase 2, 69/71 kDa
IFI16	NM_005531	7.42	18.58	IFN, γ-inducible protein 16
STAT1	NM_007315	7.42	14.68	STAT 91 kDa
ISG20	NM_002201	7.17	25.92	IFN-stimulated exonuclease gene 20 kDa
IRF7	NM_001572	6.24	12.26	IFN regulatory factor 7
IFI27	NM_005532	6.03	22.88	IFN, α-inducible protein 27
EIF2AK2	NM_002759	5.28	7.14	Eukaryotic translation initiation factor 2-α kinase 2
HLA-DOA	NM_002119	4.41	14.99	MHC, Class II, DO α
TNFSF10	NM_003810	4.32	19.51	TNF (ligand) superfamily, member 10
TAP1	NM_000593	4.03	7.60	Transporter 1, ATP-binding cassette, subfamily B (MDR/TAP)
IFNA4	NM_021068	3.90	5.76	Interferon, α 4
STAT2	NM_005419	3.79	9.68	STAT2, 113 kDa
CD70	NM_001252	3.14	5.30	CD70 molecule
NMI	NM_004688	2.81	5.68	N-myc (and STAT) interactor
BST2	NM_004335	2.61	7.04	Bone marrow stromal cell antigen 2
SAMSN1	NM_022136	2.55	5.56	Sterile α motif domain, Src homology 3 domain and nuclear localization signals 1
GBP1	NM_002053	2.35	4.49	Guanylate-binding protein 1, IFN-inducible, 67 kDa
ARL5B	NM_178815	2.28	3.70	ADP-ribosylation factor-like 5B
IRF9	NM_006084	2.25	3.50	IFN regulatory factor 9
HLA-F	NM_018950	2.07	7.39	MHC Class I, F

We confirmed the ability of IFN-λ1 to induce expression of selected ISGs in human hepatocytes by standard RT-PCR. Cultures of primary human hepatocytes were treated with IFN-α, IFN-λ1, or both for 4 or 16 h, and then ISG expression levels were evaluated by RT-PCR. As shown in Fig. 2, treatment of human hepatocytes with IFN-α or IFN-λ1 markedly up-regulated expression of multiple ISGs, including CXCL10 (IP-10), IFIT1 (ISG56), IRF7, IRF9, MX1, and MX2. The increased expression of some of these genes (e.g., IRF7 and IRF9) was maximal at the early time-point (4 h), whereas the expression levels for several other ISGs (e.g., IFI35, IFIT1, and MX2) were higher at the later time-point (16 h). Cotreatment with IFN-α plus IFN-λ1 did not significantly increase ISG expression above the levels induced by either cytokine alone.

Figure 2. IFN-α and IFN-λ induce ISG expression by primary human hepatocytes.

Primary cultures of freshly isolated human hepatocytes were treated with human rIFN-α2a or IFN-λ1 for 4 or 16 h at 37°C. At the end of the incubation period, RNA extracts were prepared, and the expression levels of various ISGs were measured by standard RT-PCR using primers specific for each gene.

To determine if the induction of ISG expression by IFN-λ1 might be mediated indirectly via induction of Type I IFNs (IFN-α or -β), we treated cultures of primary human hepatocytes with IFN-α or IFN-λ1 in the presence or absence of a neutralizing mAb (MMHAR-2) that blocks signaling through IFNARs [29] and then measured expression of six well-known ISGs by qRT-PCR. These included IRF7, PKR (eukaryotic translation initiation factor 2-α kinase 2), MX1, OAS1, IFI44, and IFIT1. As shown in Fig. 3, the anti-IFNAR2 chain mAb blocked induction of expression of all six ISGs by IFN-α but did not inhibit induction of expression by IFN-λ1. Therefore, the ability of IFN-λ to induce ISG expression in hepatocytes was not mediated indirectly through induction of Type I IFNs.

Figure 3. Induction of ISG expression by IFN-λ is not mediated indirectly via Type I IFNs (IFN-α or -β).

Cultures of primary human hepatocytes were treated with IFN-α2a or IFN-λ1 in the presence or absence of a neutralizing mAb (MMHAR-2) that blocks signaling through the Type I IFNR complex [29]. After incubation for 18 h at 37°C, total RNA extracts were prepared, and the expression levels of several representative ISGs were measured by qPCR. The gene expression levels in control (nontreated) and anti-IFNAR2, antibody-treated samples were compared for significance of difference, and the corresponding P values are shown. Nonsignificant differences are denoted as “ns”. *P < 0.05, **P < 0.005.

IFN-λ activates the JAK/STAT signaling pathway in human hepatocytes but not in lymphocytes or monocytes

There are three human IFN-λ genes, denoted IL29, IL28A, and IL28B, which encode the three distinct but highly related proteins: IFN-λ1, IFN-λ2, and IFN-λ3 [4, 5]. To compare the ability of the three IFN-λ variants to induce signal transduction in hepatocytes, we treated cultures of primary human hepatocytes with human rIFN-λ1, -λ2, or -λ3 for 30 min at 37°C. We then prepared whole cell lysates and measured the activation of STAT1 and STAT2 by Western blotting with antibodies that specifically detect pY-STAT1 and -STAT2. As shown in Fig. 4A, all three IFN-λ proteins induced activation of STAT1 and STAT2 in hepatocytes. The magnitude of STAT activation induced by IFN-λ1 and -λ3 was slightly greater than that induced by IFN-λ2. These findings are consistent with recent studies that demonstrated comparable bioactivity of the three human IFN-λ isoforms when they were assayed for antiviral activity using various HCV replicon models in vitro [30, 31]. Our findings are also consistent with a related report by Dellgren et al. [32], which showed that IFN-λ2 is less potent than IFN-λ1 or -λ3, as measured by its ability to induce antiviral activity in HepG2 cells.

Figure 4. IFN-λ activates JAK/STAT signaling in human hepatocytes but not in PBLs or monocytes.

(A) Comparative analyses of the ability of the three isoforms of IFN-λ to induce activation of STAT1 and STAT2 in primary human hepatocytes. Cultures of freshly isolated human hepatocytes were incubated with medium alone [control (Cont)], IFN-λ1, -λ2, or -λ3 (500, 50, or 5 ng/mL each) for 30 min at 37°C. At the end of the incubation period, whole cell protein extracts were prepared. The levels of pY-STAT1 and -STAT2 were measured by Western blotting using antibodies specific for pY-STAT1 and -STAT2. (B) Primary cultures of human hepatocytes, human PBLs, or human monocytes were treated with IFN-α2a or IFN-λ1 (100 ng/mL) for 30 min at 37°C. The levels of pY-STAT1 and -STAT2 were measured by Western blotting as in A.

IFNARs are present on virtually all somatic cells, including most types of leukocytes [1, 2]. To determine if human leukocytes can also respond to IFN-λ, we prepared cultures of purified human PBLs and monocytes by countercurrent centrifugal elutriation, as described previously [18], and examined the ability of these cells to respond to treatment with IFN-α or IFN-λ1. As shown in Fig. 4B, IFN-α induced activation of STAT1 and STAT2 in lymphocytes and monocytes, as well as in hepatocytes. In contrast, IFN-λ induced activation of STAT1 and STAT2 in hepatocytes but not in lymphocytes or monocytes.

DISCUSSION

Although several previous reports have shown that human hepatoma cell lines, such as HepG2 and Huh7, can respond to IFN-λ [12–14], this is the first comprehensive study to evaluate the ability of primary human hepatocytes to respond to this cytokine. We found that IFN-λ, like IFN-α, is able to induce ISGF3 activity and ISG expression in mouse and human hepatocytes. However, the magnitude of ISGF3 activity and ISG expression induced by IFN-λ in mouse and human hepatocytes was consistently lower than the magnitude induced by IFN-α. These differences may be a result of differences in the relative levels of IFN-λRs (IL-28R) versus IFNARs on hepatocytes. Although the magnitude of ISG expression induced by IFN-λ was generally lower than that induced by IFN-α in hepatocytes, the genes that are up-regulated by IFN-λ were largely the same as those induced by IFN-α. These findings are consistent with findings from a related comparative analysis of ISG expression induced by IFN-α versus IFN-λ in Huh-7 hepatoma cells [14].

We found that IFN-λ, like IFN-α, induces ISGF3 activity and expression of multiple ISGs in mouse and human hepatocytes. The induction of ISG expression [e.g., Mx1, Oas1g, Irf7, Ddx58 (RIG-I), and Dhx58 (LGP2)] by Type I IFNs (IFN-α and -β) or Type III IFNs (IFN-λ) was markedly diminished in STAT1-deficient hepatocytes. These findings are consistent with the fact that STAT1 is required for formation of ISGF3 transcription factor complexes in all cell types. Furthermore, the activation of hepatocytes by IFN-λ was not mediated indirectly via production and activity of Type I IFNs, as treatment with an inhibitory anti-IFNAR2 mAb blocked induction of ISG expression by IFN-α but did not inhibit induction of ISGs by IFN-λ in human hepatocytes. Our findings demonstrate that Type I (IFN-α or -β) and Type III (IFN-λ) IFNs mediate signal transduction through distinct receptor complexes in mouse and human hepatocytes but activate the same downstream intracellular signaling pathway. Although they signal through distinct receptor complexes, IFN-α and IFN-λ induce expression of a common set of ISGs in hepatocytes.

So far, there have only been a few studies to evaluate the effects of IFN-λ in animal models. With the use of a murine model of HSV-2 viral infection, Ank et al. [33] found that treatment with rIFN-λ in vivo significantly reduced HSV-2 viral titers in the livers of infected mice. The activity of IFN-λ was not dependent on functional Type I IFNRs, as the antiviral effects were also fully demonstrable in IFNAR1 knockout mice. These findings are consistent with our results with human hepatocytes, which showed that blocking signaling through Type I IFNRs with an anti-IFNAR2 mAb did not inhibit responses to IFN-λ. A recent report showed that parenteral administration of pegylated-IFN-λ1 protein to cynomolgus monkeys induced significant increases in ISG expression in the livers of these animals but did not induce activation of STAT1 or ISG expression by peripheral blood leukocytes [34].

We observed subtle differences in the relative potencies of human rIFN-λ1, -λ2, and -λ3 when we compared their ability to induce activation of STAT1 and STAT2 in primary human hepatocytes (Fig. 4). These findings are consistent with a related report by others that showed that IFN-λ2 is less potent than IFN-λ1 or -λ3, as measured by its ability to induce antiviral activity in HepG2 cells [32]. Similar to what we observed with primary human hepatocytes, these authors also found that human rIFN-α2 is significantly more potent than human rIFN-λ1, -λ2, or -λ3 based on a comparison of the ability of these proteins to induce antiviral activity in HepG2 cells or Huh-7 cells.

The subset of genes that are up-regulated by IFN-α and IFN-λ is largely distinct from that induced by Type II IFN (IFN-γ). Previous studies showed that the gene expression profiles induced by Type I (IFN-α) versus Type II (IFN-γ) IFNs do not display significant overlap [35, 36]. In contrast, we observed that the gene expression profiles induced by Type I (IFN-α) and Type III (IFN-λ) IFNs are quite similar. However, we did observe significant differences in the relative levels of ISG expression induced by IFN-α versus IFN-λ in mouse and human hepatocytes. The overall magnitude of ISG expression induced by IFN-λ in hepatocytes was generally lower than that induced by IFN-α, even when the cells were treated with equivalent concentrations of each type of IFN. A related comparative analysis of the ability of IFN-α and -λ to induce ISG expression in Huh7.5 cells showed that the induction of ISG expression by IFN-λ was slower but more sustained compared with the more rapid and more transient ISG expression induced by IFN-α [14].

Our findings regarding the effects of IFN-λ on primary human hepatocytes are consistent with previous studies that showed that IFN-λ can inhibit replication of partial and full-length HCV replicons in Huh7 hepatoma cells [12–14]. The ability of Type I and III IFNs to mediate antiviral activity is a composite function of the activity of multiple gene products [7, 8]. As IFN-λ can induce antiviral activity in vitro similar to IFN-α, a pegylated form of human rIFN-λ1 (IL-29) is now being tested clinically as a potential therapeutic alternative to IFN-α for the treatment of chronic HCV infection [37, 38]. The results derived from an initial Phase I clinical trial support the hypothesis that IFN-λ can induce antiviral activity comparable with that induced by IFN-α but without many of the undesirable hematologic side-effects that are commonly associated with IFN-α therapy [37].

Several recent studies by others examined the potential cross-regulatory effects of IFN-α and IFN-λ on liver cells. Makowska et al. [39] showed that pretreating normal mice with IFN-α, -β or -λ induced a state of refractoriness to a secondary challenge with IFN-α in vivo. This inhibitory effect was specific for IFN-α, as responsiveness to IFN-β or IFN-λ was not inhibited by pretreatment with Type I or Type III IFN. In a related study by François-Newton et al. [40], the authors showed that priming human hepatocytes with Type I or Type III IFN in vitro suppresses a subsequent secondary response to IFN-α but not to IFN-β. These authors suggested that this state of cellular desensitization may be a result of impaired formation of binding sites for IFN-α. Together, these studies indicate that pretreatment with IFN-α or -λ inhibits responsiveness to a secondary challenge with IFN-α but not IFN-λ. These findings may have important clinical relevance, as they predict that patients who are treated chronically with rIFN-λ will be less likely to become refractory to repeated administration of this cytokine.

Our finding that IFN-λ induces STAT activation and ISG expression in hepatocytes but not in PBLs or monocytes highlights an important functional difference between IFN-λ and IFN-α. However, the inability of IFN-λ to induce Jak/STAT signaling by peripheral blood lymphocytes or monocytes does not exclude the possibility that there may be a small subset of leukocytes that can respond to this cytokine. In fact, we [41] and others [42] have shown recently that IFN-λ can induce functional responses in human plasmacytoid DCs. However, the inability of IFN-λ to activate most types of leukocytes, including lymphocytes and monocytes, may provide a significant clinical advantage for this cytokine as a potential therapeutic agent for chronic HCV infection, as it will be less likely to induce the hematologic toxicities, such as lymphopenia and monocytopenia, that are frequently associated with the clinical use of IFN-α drug products.

AUTHORSHIP

H.D., F.S., and O.P. performed the experiments and discussed the results. H.D., B.G., and R.P.D. designed the experiments, analyzed the data, and wrote the manuscript. B.G. and R.P.D. conceived of the study and reviewed the manuscript.