Virology 444 (2013) 317–328

Contents lists available at ScienceDirect

Virology

0042-68http://d

n Corrment ofsity of M

E-m(L. Bellmusolindteti@u

journal homepage: www.elsevier.com/locate/yviro

Hepatitis B virus (HBV) induces the expression of interleukin-8 thatin turn reduces HBV sensitivity to interferon-alpha

Teresa Pollicino a,c,n, Luigi Bellinghieri b, Agnese Restuccia b,c, Giuseppina Raffa b,c,Cristina Musolino b,c, Angela Alibrandi d, Diana Teti b, Giovanni Raimondo b,c

a Department of Pediatric, Gynecologic, Microbiological, and Biomedical Sciences, University of Messina, Messina, Italyb Department of Clinical and Experimental Medicine, University of Messina, Messina, Italyc Division of Clinical and Molecular Hepatology, University of Messina, Messina, Italyd Department of SEAM, University of Messina, Messina, Italy

a r t i c l e i n f o

Article history:Received 4 April 2013Returned to author for revisions2 May 2013Accepted 25 June 2013Available online 23 July 2013

Keywords:Hepatitis B virus chronic infectionInterleukin-8IL-8 promoterTranscription factorsHepatitis B X viral regulatory proteinSignal pathwaysChromatin immunoprecipitationChromatin modifying enzymesIL-8 siRNAInterferon-alpha

22 & 2013 The Authors. Published by Elseviex.doi.org/10.1016/j.virol.2013.06.028

esponding Author. Division of Clinical and MPediatric, Gynecologic, Microbiological, andessina, via Consolare Valeria, 98124 Messina,ail addresses: tpollicino@unime.it (T. Pollicinoinghieri), restuccia.a@gmail.com (A. Restucciaocristina@gmail.com (C. Musolino), aalibrandnime.it (D. Teti), raimondo@unime.it (G. Raim

a b s t r a c t

High levels of serum interleukin-8 (IL-8) have been detected in chronic hepatitis B (CHB) patients duringepisodes of hepatitis flares. We investigated whether hepatitis B virus (HBV) may directly induce IL-8production and whether IL-8 may antagonize interferon-alpha (IFN-α) antiviral activity against HBV. Weshowed that CHB patients had significantly higher IL-8 levels both in serum and in liver tissue thancontrols. In HBV-replicating HepG2 cells, IL-8 transcription was significantly activated. AP-1, C/EBP andNF-kB transcription factors were concurrently necessary for maximum IL-8 induction. Moreover, HBxviral protein was recruited onto the IL-8 promoter and this was paralleled by IL8-bound histonehyperacetylation and by active recruitment of transcriptional coactivators. Inhibition of IL-8 increases theantiviral activity of IFN-α against HBV. Our results indicate that HBV activates IL-8 gene expression bytargeting the epigenetic regulation of the IL-8 promoter and that IL-8 may contribute to reduce HBVsensitivity to IFN-α.

& 2013 The Authors. Published by Elsevier Inc. Open access under CC BY-NC-ND license.

Introduction

Hepatitis B virus (HBV) infection is a major health problem world-wide with estimates of nearly 400 million people currently infected(Dienstag, 2008). HBV infectionmay associate with a large spectrum ofclinical forms, ranging from very mild and asymptomatic clinicalpictures to the most severe liver diseases, including fulminanthepatitis, cirrhosis, and hepatocellular carcinoma (HCC) (Dienstag,2008).Chronic HBV infection is a dynamic condition in which theinteraction between virus and the host immune response influences

r Inc.

olecular Hepatology, Depart-Biomedical Sciences, Univer-Italy. Fax: +39 090 221 3594.), bellinghieri.luigi@email.it), graffa@unime.it (G. Raffa),i@unime.it (A. Alibrandi),ondo).

Open access under CC BY-NC-ND

the outcome of the disease. Its natural history is complex and variable.It includes different – not always sequential – phases that aredistinguishable on the basis of the presence in the serum of HBV “e”antigen (HBeAg) or its antibody (anti-HBe), of the HBV DNA serumlevels, of alanine aminotransferases (ALT) values and of the liverhistology (EASL, 2012). The anti-HBe positive phase may be character-ized by persistently low viremia levelso2000 UI/ml), normal ALT andminimal histological lesions (inactive chronic HBV carrier state) (EASL,2012) or by stable or fluctuating high levels of viral replication and ALTvalues, with active liver necroinflammation (HBeAg-negative chronichepatitis). Patients with HBeAg-negative chronic hepatitis B (CHB)may undergo recurrent, spontaneous “hepatic flares,” characterized bywide fluctuations of viremia levels, liver inflammation and a propen-sity to a rapid progression towards cirrhosis (EASL, 2012).

The HBV replication cycle is not directly cytotoxic to cells andthe pathogenesis of liver disease has conventionally been attrib-uted to cytolytic killing of infected hepatocytes by virus-specific Tcell response (Chisari and Ferrari, 1995; Perrillo, 2001). However,important data have shown that the high frequencies of HBV-specific CD8+ T cells in patients with chronic HBV infection areassociated with HBV control rather than with hepatic injury

license.

Table 1Characteristics of patients with HBV-related chronic liver disease (CLD) and control subjects.

Parameter Patients with HBV-related CLD Control subjectsa P value(n¼26) (n¼22)

Age (years), median (range) 41 (29–84) 42 (20–60) NS (0.987)Male sex, n (%) 21 (80.8) 13 (59) NS (0.150)ALT (IU/L), median (range) 100 (50–260) 18 (10–38) o 0.0001Intrahepatic IL–8 mRNA Rel.C. (fold), median (range) 0.71 (0.01–4.9) 0.06 (0.01–0.07) 0.003Serum IL–8 (pg/mL), median (range) 29.5 (7.7–278.3) 10.5 (2–5) o 0.0001

Abbreviations: ALT, alanine aminotransferases; Rel. C., Relative Concentration; HBV, hepatitis B virus.a None of the 22 control subjects had any sign of liver disease, seven of them were inactive HBV carriers and 15 were individuals negative for all HBV serum markers.

Fig. 1. Chronic hepatitis B (CHB) patients have significantly higher amounts of IL-8 mRNA in the liver and of IL-8 chemokine in the serum than control subjects.(A) Intrahepatic amounts of IL- 8 mRNA. (B) Serum IL-8 levels. Median relative amounts of IL-8 mRNA in the liver of CHB patients (n¼14) and of control subjects (n¼5) were0.71 (range: 0.01–4.9) and 0.06 (range: 0.01–0.079), respectively, whereas median levels of IL-8 chemokine in the serum of CHB patients (n¼26) and of control subjects(n¼22) were 29.5 pg/mL (range: 7.7–278.3 pg/mL) and 10.5 pg/mL (range: 2–20 pg/mL), respectively.

T. Pollicino et al. / Virology 444 (2013) 317–328318

(Bertoletti et al., 2010; Maini et al., 2000; Rehermann, 2003).In addition, it has been demonstrated that adoptive transfer ofHBsAg-specific CD8+ T cells in HBV transgenic mice results in therapid recruitment of HBV-non-specific bystander lymphocytes andthat liver damage becomes manifest when non-specific,chemokine-mediated recruitment of neutrophils, natural killer(NK) cells and activated T-cells occurs (Kakimi et al., 2001; Sitiaet al., 2004). Of interest, it has been recently shown that in chronicHBV infected patients flares of liver inflammation (either sponta-neous or induced by anti-viral withdrawal) are preceded by aparallel increase of interleukin-8 (IL-8) production and serum HBVDNA levels (Dunn et al., 2007; Tan et al., 2010). IL-8 may synergizewith interferon- alpha (IFN-α) to activate NK cells, (Dunn et al.,2007) and very recent data have shown that HBV-specific T cellsmaturing in the intrahepatic inflammatory environment canproduce IL-8, which in turn can contribute to the developmentof liver damage through the recruitment of granulocytes (Gehringet al., 2011; Zimmermann et al., 2011).

IL-8 is a CXC chemokine able to elicit granulocytes, NK cells and Tcell chemotaxis at the inflammatory site (Mukaida, 2003; Taub et al.,1996). It is a principal mediator of the inflammatory response to manyviruses and bacteria (Girard et al., 2002; Green et al., 2006; Hisatsuneet al., 2008; Polyak et al., 2001a, 2001b; Rajaiya et al., 2008; Venzaet al., 2007; Wagoner et al., 2007; Yu et al., 2011; Zheng et al., 2008)and it may interfere with the antiviral effect of IFN-α (Girard et al.,2002; Khabar et al., 1997a; Lee et al., 2011; Polyak et al., 2001a, 2001b).Previous data had demonstrated that the HBV-encoded regulatoryHBX protein is able to transactivate the IL-8 promoter (Mahe et al.,1991). However, the exact mechanisms through which HBV is able toinduce IL-8 gene expression remain mostly unknown.

In the present study we investigated IL-8 in the course of HBVinfection and for this purpose we quantified the IL-8 amounts inpaired serum and liver tissue samples of chronic HBV infected

individuals either CHB or inactive carriers. Moreover, by using acell-based HBV replication system that makes it possible torecapitulate the HBV replication cycle including the nuclear gen-eration of covalently closed-circular DNA (cccDNA) molecules(Pollicino et al., 2006), we studied the molecular mechanismsimplicated in the activation of IL-8 gene expression and verifiedwhether IL-8 might reduce HBV sensitivity to IFN-α.

Results

IL-8 amounts in liver tissues and sera of CHB patients, inactive HBVcarriers and HBV-neg subjects

Real-time PCR analysis on liver tissue specimens showed thatCHB patients had significantly higher median amounts of IL-8mRNA (0.71 [range: 0.01–4.9] versus 0.06 [range: 0.01–0.07];P¼0.003) compared with the subjects of the control group,including both inactive HBV carriers (IBC) and HBV-neg subjects(Table 1 and Fig. 1A). Moreover, among CHB patients intrahepaticIL-8 mRNA levels were significantly higher in HBeAg-negativepatients (1.7 [range: 0.5–4.7] versus 0.1 [range: 0.01–4.9];P¼0.028) than in HBeAg-positive ones. Both HBeAg-negative andHBeAg-positive CHB patients had much higher amounts of IL-8mRNA in the liver than controls (P¼0.004 and P¼0.015, respec-tively). Concentrations of circulating IL-8 chemokine, assayed by aquantitative ELISA kit, were also significantly increased in CHBpatients compared with controls (median, 29.5 pg/mL [range: 7.7–278.3 pg/mL] versus 10.5 pg/mL [range: 2–20 pg/mL], respectively;Po0.0001) (Table 1 and Fig. 1B) and this statistical significancewas maintained also when serum IL-8 amounts in CHB patientswere separately compared with the amounts in IBC (median,9.78 pg/mL; range, 4.2–19.9 pg/mL; Po0.0001) and in HBV-neg

Fig. 2. HBV replication is associated with significantly increased amounts of IL-8 transcripts in transfected HepG2 cells. (A) At 48 h after transfectionwith 0.5 mg of linear HBVgenomes (WTHBV-D: HBV of genotype D; WTHBV-A: HBV of genotype A) or 0.5 mg of pcDNA3 empty vector, total RNA was extracted from HepG2 cells and IL-8 mRNAquantification was performed by specific real-time PCR. (B) Transfection of increasing amounts of HBV DNA (WTHBV-A) induces a parallel increment of IL-8 transcriptproduction (up to 10-fold). (C) Southern blot analysis of HBV replicative intermediates. Lanes correspond to DNA extracted from cytoplasmic HBV core particles derived fromHepG2 cells that were transfected with increasing amounts of monomeric linear WT HBV genomes. Cells were harvested at 48 h post-transfection. Membranes were exposedto an X-ray film for autoradiography at �80 1C for 4 h. (D) Northern blot analysis of HBV transcripts isolated from HepG2 cells at 48 h after transfection with increasingamounts of monomeric linear HBV genomes. Membranes were exposed to an X-ray film for autoradiography at �80 1C for 24 h. Lanes correspond to pregenomic (pg), PreS1,and PreS2/S viral messenger RNA (mRNA). RNA loading was controlled by rehybridization with a 32P-labeled 18S specific probe. (E) Expression profile of IL-8 in HepG2 cellstransfected with increasing amounts of pcHBx, (F) with equal increasing amounts of the HBx-mutant viral genome, mutHBxHBV-A, or (G) cotransfected with increasingamounts of mutHBxHBV-A genome along with 0.2 mg of pcHBx. All experiments were performed in duplicate and the results are the average of three independentexperiments (mean7SD). Abbreviations: OC, open circular HBV DNA; DS, double-stranded HBV DNA; SS, single–stranded HBV DNA.

T. Pollicino et al. / Virology 444 (2013) 317–328 319

subjects (median, 11 pg/mL; range, 2–20 pg/mL; Po0.0001) of thecontrol group. Instead, no statistically significant difference wasfound between serum concentrations of IL-8 chemokine in HBeAg-negative and HBeAg-positive CHB patients (median, 31.5 versus28 pg/mL, respectively; P¼0.637). As expected, serum HBV DNAlevels were significantly higher in CHB patients (median,3.2�106 IU/mL; range, 2.1�104–1.7�108 IU/mL) compared withlevels detected in IBC (median, 6.6�102 IU/mL; range, 12–1.8�103 IU/mL; Po0.0001). Of note, a significant correlationwas found between serum IL-8 concentrations and HBV DNAamounts (r¼0.519; P¼0.002) in HBV infected individuals.

IL-8 expression Is upregulated in HBV-replicating hepatoma cell line

We have shown that transient transfection of plasmid freelinear HBV DNA into HepG2 or Huh7 cell lines makes it possible torecapitulate the HBV replication cycle, including the nucleargeneration of viable cccDNA (Pollicino et al., 2006). Thus, toinvestigate the role of HBV in the induction of IL-8 expression,HepG2 cells were transiently transfected either with 0.5 mg oflinear wild-type (WT) HBV DNAs of genotype D or genotype A(WTHBV-D and WTHBV-A, respectively) or with pcDNA3.1 emptyvector as a control at 24 h postplating. Total RNA was extracted at

48 h after transfection, and expression of IL-8 was analyzed byquantitative real-time PCR. As shown in Fig. 2A, the replication ofHBV (both of genotype D and A) in HepG2 cells was associatedwith a significant increase of IL-8 transcription (P¼0.038 andP¼0.046, respectively) when compared with the control cells.Moreover, when increasing amounts (0.1, 0.2, 0.5 and 1 mg) ofWTHBV-D or WTHBV-A were transfected, higher amounts of viralreplicative intermediates and transcripts were produced in a dose-dependent manner (Fig. 2C, D) and these events were paralleledby an increased production of IL-8 transcripts (Fig. 2B). In addition,to analyze the contribution of HBx viral regulatory protein in theinduction of IL-8 gene expression, HepG2 cells were transfectedwith the same increasing amounts of either the pcHBx recombi-nant vector or of the mutHBxHBV-A genome that does not expressHBx because of a stop codon for amino acid 8 of the HBx openreading frame (Belloni et al., 2009). Furthermore, to assess thepattern of IL-8 expression when the X protein is provided in transto the HBx-mutant HBV, HepG2 cells were cotransfected withmutHBxHBV-A genome and pcHBx vector. Real-time PCR quanti-fications showed that IL-8 mRNA amounts were increased up tothreefold in cells expressing HBx and up to fourfold in cellsreplicating the mutHBxHBV-A genome than in cells transfectedwith an empty vector (Fig. 2E, F). Thus, though IL-8 mRNA levels in

Fig. 3. The IL-8 promoter is activated by HBV. (A) HepG2 cells were cotransfected with 0.5 mg of a reporter plasmid containing 1.4 kb of the IL-8 promoter (wtIL-8 LUC) alongwith 0.1, 0.2, 0.5 or 1 mg of monomeric linear HBV genomes (both WTHBV-D and WTHBV-A were used), (B) 0.1, 0.2, 0.5 or 1 mg of monomeric linear mutHBxHBV-A or ofWTHBV-A genomes, (C) 0.1, 0.2, 0.5 or 1 mg of pcHBx expression vector, (D) 0.1, 0.2, 0.5 or 1 mg of mutHBxHBV-A and 0.2 mg of pcHBx. The basal activity of cells cotransfectedwith the wtIL-8 LUC reporter and empty vectors was set at1 (Control). Luciferase activities were determined 48 h after transfection, and the results are the average of threeindependent experiments carried out in duplicate. Means 7S.D. are presented.

T. Pollicino et al. / Virology 444 (2013) 317–328320

cells replicating the mutHBxHBV-A were lower compared toWTHBV-A transfected cells, they were higher compared to controlcells, indicating that, aside from HBx, other viral factors areimplicated in HBV-mediated IL-8 induction. Interestingly, inHepG2 cotransfected with mutHBxHBV-A genome and pcHBxvector IL-8 mRNA amounts were comparable to those detectedin WTHBV-A replicating cells (Fig. 2G).

Transactivation of the IL-8 promoter by HBV

The impact of HBV transcription/replication on IL-8 promoteractivity was investigated by using the HBV cccDNA-driven replica-tion system (Pollicino et al., 2006). A 1.4-kb IL-8 promoter-luciferase construct (wtIL-8 LUC) was transfected into HepG2 cellsalong with increasing amounts of WTHBV-A, WTHBV-D, ormutHBxHBV-A genomes or of pcHBx vector. Measurement of IL-8 promoter-mediated luciferase activity in HBV-replicating cellsshowed that both genotype A and D WT HBV genomes were ableto stimulate basal IL-8 promoter activity up to 100-fold in a dose-dependent manner (Fig. 3A), whereas transfection of equalamounts of mutHBxHBV-A virus led to significantly lower activa-tion of the IL-8 promoter (up to 30-fold; Po0.0001) (Fig. 3B). HBxexpression stimulated the basal IL-8 promoter activity up tothreefold when 0.2 mg of the pcHBx were transfected (Fig. 3C).Moreover, when the HBx-mutant virus was trans-complementedwith the wild type HBx gene, in co-transfection experiments, itsability to transactivate the IL-8 promoter was again largelyrestored (Fig. 3D).

Subsequently, to determine the minimal IL-8 promoterdomains required for HBx and HBV-mediated transcriptionalactivation, a series of IL-8 reporter gene constructs containingmutated AP-1, C/EBP or NF-kB transcription factor binding siteswere utilized (Caristi et al., 2005; Venza et al., 2007). Inaccordance with previously reported results, we found thatHBx-induced transcriptional activation of the IL-8 promoter

requires intact C/EBP and NF-kB binding sites (Fig. 4A). In HBVreplicating cells, mutation in the AP-1, in the C/EBP or in theNF-kB binding sites reduced promoter activity by ∼50%, ∼40%and ∼46% respectively (Fig. 4B). These data indicate that allthree binding sites, NF-kB, C/EBP, and AP-1, are involved inHBV-induced IL-8 promoter activity.

To determine signal pathways implicated in HBV-mediatedtranscriptional activation of the IL-8 gene, HBV-replicating cellsand control cells were treated with inhibitors of PI3K/Akt(LY294002), p38 MAPK (SB203580), JNK (JNKi II), IkB kinase (Bay11-7082), Erk (PD98059), SRC tyrosine kinase (PP2) and AP-1(Tanshinone IIA) for 16 h. As shown in Fig. 4C, treatment of HBVreplicating-cells with Bay 11-7082 (10 μM) or Tanshinone IIA(25 μM) completely abolished the IL-8 promoter-mediated lucifer-ase activity, whereas treatment with LY294002 (30 μM), SB203580(25 μM) or JNKi II (25 μM) only partially inhibited IL-8 promoteractivation. PD98059 and PP2 treatments did not significantly affectHBV-mediated promoter activity (data not shown). Moreover, theELISA assay showed that Bay 11-7082 or Tanshinone IIA was ableto strongly reduce (up to 90% and 85%, respectively) HBV-mediated IL-8 production. (Fig. 4D). Treatments with LY294002,SB203580 or JNKi II reduced IL-8 production by 55%, 60% and 45%,respectively, (Fig. 4D). These last data confirm the importance ofNF-kB and AP1 pathways in the induction of IL-8 expression byHBV and indicate that PIK3/AKT, p38 MAPK and JNK signalingpathways also have a role in this induction.

The analysis of the expression of a selected number of genes byTaqMan Low-Density Arrays in HBV-replicating cells and controlcells confirmed the ability of HBV to significantly induce theproduction of IL-8 transcripts and, consistently, showed that viralreplication may induce the upregulation of numerous genes (as,NFKB1, IKKB, IL1B, TNFα, TNFRSF1A, TBK1 IRAK1, IRAK4, TRAF6,PIK3CB, PIK3R2, CREB, JUN, ATF4, MAPK11 and MAPK13) (Fig. 4E).involved – directly or indirectly – in the activation the IL-8promoter (Hoffmann et al., 2002).

Fig. 4. AP-1, C/EBP and NF-kB binding sites are all involved in HBV-induced IL-8 promoter activation. (A) Identification of the minimal domains required for HBx- and(B) HBV-induced IL-8 promoter transactivation. HepG2 cells were cotransfected with the indicated reporter plasmids (a series of mutated IL-8 reporter gene constructs) and0.2 mg of pcHBx expression vector or 0.5 mg of monomeric linear HBV genomes (WTHBV-A). At 48 h posttransfection, the luciferase assay was performed. The basal activity ofcells cotransfected with the wtIL-8 LUC reporter and empty vectors was set at1. The HBx- and HBV-induced fold activation for each IL-8 reporter plasmid is provided.(C) HepG2 cells were cotransfected with 0.5 mg of wtIL-8 LUC reporter and 0.5 mg of monomeric linear HBV genomes (WTHBV-A) or 0.5 mg of empty pcDNA3.1 vector andtreated with DMSO (Vehicle) or inhibitors of IkB kinase (Bay 11-7082: 10 μM), AP-1 (Tanshinone IIA: 25 μM), PI3K/Akt (LY294002: 30 μM), p38 MAPK (SB203580: 25 μM) andJNK (JNKi II; 25 μM). Cells cotransfected with 0.5 mg of wtIL-8 LUC reporter and 0.5 mg of WTHBV-A or 0.5 mg of empty vector and treated with DMSO (Vehicle) were used as anegative control (Control). At 16 h post-treatment, cells and supernatants were harvested to test luciferase activity by luciferase reporter assay and (D) IL-8 production byELISA test, respectively. All experiments were performed in duplicate and repeated at least three times. Means 7S.D. are presented. (E) Expression profiling of selected genesin HBV replicating HepG2 cells. Custom real-time PCR liquid arrays (TLDAs-Applied Biosystems) were loaded with 0.2 mg of cDNA obtained from total RNA extracted fromHepG2 cells at 48 h posttransfection with 0.5 mg of monomeric linear HBV genomes (WTHBV) or 0.5 mg empty pcDNA3.1 vector. Results are expressed as fold induction overthe basal level of expression of empty vector transfected cells that was set at 1. Experiments were performed in duplicate and the results are the average of threeindependent experiments (mean7SD).

T. Pollicino et al. / Virology 444 (2013) 317–328 321

The HBx viral protein is recruited onto the IL-8 promoter in livertissues of chronic HBV carriers and in HBV-replicating HepG2 cells

It is known that HBV may induce epigenetic changes in hostcells and HBx seems to play a central role in these epigeneticmodifications. (Herceg and Paliwal, 2011; Paschos and Allday,2010) Indeed, it has been shown that this protein may interactwith different components of epigenetic complexes and mayinduce a modulation of transcriptional activity of several targetgenes (Kew, 2011; Park et al., 2007).

To analyze the recruitment of chromatin-modifying enzymes ontothe IL-8 gene promoter, we applied the ChIP assay to HBV-replicatingcells and liver tissues from two CHB patients, two IBC and two HBV-neg subjects. Chromatin was immunoprecipitated using antibodies

that specifically recognize acetylated histone H4, the acetyltrans-ferases PCAF and p300, the histone deacetylase HDAC1 and HBx viralprotein. As shown in Fig. 5, histone H4 bound to the IL-8 promoterwas highly acetylated and the transcriptional coactivators PCAF andp300 as well as HBx were recruited onto the IL-8 promoter both inHepG2 cells replicating HBV and in liver tissues from CHB patientswith high viral loads. Interestingly, in IBC (subjects characterized by astate of persistently low HBV replication) the HDAC1 histone deace-tylases was recruited onto the IL-8 promoter. HADC1 recruitment inthese subjects was paralleled by the reduced binding to IL-8 of HBxviral protein, p300 and PCAF co-activators and by the presence ofvariable levels of IL8-bound histone acetylation. As expected, no HBxrecruitment onto the IL-8 promoter could be detected in HBV-negsubjects. Moreover, in these subjects the recruitment on the IL-8

Fig. 5. HBV modulates the epigenetic control of IL-8 promoter activity by affecting the recruitment of chromatin modifying enzymes. Chromatin from empty vectortransfected-cells, HBV-transfected HepG2 cells and liver tissues from two chronic hepatitis B (CHB) patients, two inactive HBV carriers (IBC) and two HBV-negative(HBV-neg) subjects, was immunoprecipitated with control IgG or specific anti-AcH4, anti-HDAC1, anti-p300, anti-PCAF, and anti-HBx antibodies and analyzed by PCR withIL-8 promoter selective primers. All experiments were repeated at least twice (upper panel). Signal intensity of bands was quantified with Quantity One 1-D AnalysisSoftware (BioRad Laboratories). Data are expressed as relative arbitrary units (lower panel).

T. Pollicino et al. / Virology 444 (2013) 317–328322

promoter of PCAF, and p300 acetyltransferase was highly reduced andthe binding of HDAC1 was associated with the hypoacetylation statusof H4 histone. Thus, in HBV infected patients the differential recruit-ment of coactivators and corepressors onto the IL-8 promoterappeared to be correlated with HBV replication activity.

Altogether, these results indicate that HBx viral regulatoryprotein produced during HBV replication in the infected hepato-cytes is recruited onto the IL-8 promoter and that its recruitmentis associated with IL-8 gene expression.

IL-8 impairs IFN alpha-inhibition of HBV replication

It has been shown that IL-8 may inhibit in vitro the antiviralaction of IFN-α (Khabar et al., 1997a) and an inverse correlation hasbeen found between serum IL8 levels and the sensitivity tointerferon treatment among patients infected with hepatitis Cvirus (Lee et al., 2011; Polyak et al., 2001b).

By the use of the cell-based HBV replication system, we verifiedwhether IL-8 might weaken the action of IFN-α against HBV. Thus,HBV replication was examined in HepG2 cells co-transfected withHBV genomes and IL-8 siRNA (or siRNA-control) and then exposed

or not to IFN-α treatment. Results from the ELISA assay ascertainedthat IL-8 protein released in the supernatants of transfected cellswas significantly reduced by treatment with siRNA-IL-8 but notaffected by siRNA-control (Fig. 6A).

Results from both Southern and Northern blotting showed thatsiRNA-directed inhibition of IL-8 expression improved the efficacyof IFN-α treatment against HBV (Fig. 6D and E, respectively).Indeed, HBV/IL8-siRNA co-transfected HepG2 cells showed astronger reduction of both HBV replicative intermediates(Fig. 6D) and viral transcripts (Fig. 6E) compared to cellsco-transfected with HBV genomes and siRNA-control, after IFN-αtreatment. Interestingly, even in the absence of IFN-α treatmentthe suppression of IL-8 in HBV/IL8-siRNA co-transfected cellsinduced a reduction of the steady-state amounts of capsid asso-ciated HBV DNA and of all the viral transcripts (Fig. 6B and C), thussuggesting that IL-8 is able to induce HBV replication. To furtherconfirm these results, viral transcription/replication was evaluatedin HBV transfected-cells pretreated or not pretreated with recom-binant human IL-8 (rIL-8) and then exposed or not exposed to IFN-α treatment. As shown in Fig. 6F and G, when HBV replicating-cellswere pretreated with rIL-8 (10 ng/mL), a significant reduction of

Fig. 6. IL-8 weakens IFN-α antiviral action against HBV and induces viral transcription/replication in HBV-replicating HepG2 cells. (A) Determination of the effectiveness and specificityof siRNA-IL-8 on IL-8 gene expression. HepG2 cells were transfected with empty pcDNA3.1 vector monomeric or with linear HBV genomes (WTHBV-A) alone or along with siRNA-control or with siRNA-IL-8. IL-8 protein in culture supernatants was detected by ELISA. (B) Analysis of the effect of siRNA-IL-8 on HBV replication by Southern blotting of HBV replicativeintermediates at 48 h post transfection. (C) Analysis of the effect of siRNA-IL-8 on HBV transcription by Northern blotting of viral mRNAs from HepG2 cells 48 h after transfection withWTHBV-A. (D) Southern blot analysis of HBV replicative intermediates from HepG2 cells transfected with WTHBV-A alone or co-transfected with WTHBV-A and IL-8 siRNA (or siRNA-control) and then exposed or not to IFN-α IFN‐apha treatment (1000 U/ml) for 48 h. (E) Northern blot analysis of HBV transcripts extracted from HepG2 cells transfected withWTHBV-A alone or co-transfected with WTHBV-A and IL-8 siRNA (or siRNA-control) and then exposed or not to IFN-α IFN-alpha treatment (1000 U/ml) for 48 h. (F) Southern blot analysis ofHBV replicative intermediates from HBV transfected-cells pretreated or not pretreated with recombinant human IL-8 (rIL-8) and exposed or not exposed to IFN-α treatment.(G) Northern blot analysis of HBV transcripts isolated from HBV transfected-cells pretreated or not pretreated with recombinant human IL-8 (rIL-8) and exposed or not exposed to IFN-α treatment. (H) Southern blot analysis of HBV replicative intermediates from HBV transfected-cells treated with increasing amount of rIL-8 (10-30 ng/ml). (I) Northern blot analysis ofHBV transcripts from HBV transfected-cells treated with equal increasing amount of rIL-8. In all panels densitometric quantification of HBV replicative intermediates or of HBVtranscripts is shown. Signal intensity of the single-strand (SS) band underneath the linear double-stranded (DS) HBV DNA band or of pregenomic RNA (pgRNA) band was quantifiedwith Quantity One 1-D Analysis Software (BioRad Laboratories). Data are expressed as relative arbitrary units (mean7SD) from three independent experiments. Abbreviations: OC,open circular HBV DNA; DS, double-stranded HBV DNA; SS, single-stranded HBV DNA.

T. Pollicino et al. / Virology 444 (2013) 317–328 323

T. Pollicino et al. / Virology 444 (2013) 317–328324

IFN-α efficacy was consistently observed. Moreover, treatment ofHBV transfected cells with rIL-8 in the absence of IFN-αwas able toinduce a dose dependent increase of HBV replication and tran-scription (Fig. 6H and I).

Discussion

The IL-8 chemokine is an important mediator of the innateimmunity with well-defined immunomodulatory effects on T-cellfunction and inflammatory response (Mukaida, 2003; Taub et al.,1996). Recent evidence has suggested that this chemokine mayplay an important role in the immunopathogenesis of HBV infec-tion (Bertoletti et al., 2010). High levels of serum IL-8 have beendetected in CHB patients with active liver inflammation as well asin patients with acute hepatitis B (but not in patients with otheracute viral infection) (Dunn et al., 2007; Gehring et al., 2011; Tanet al., 2010; Zimmermann et al., 2011). Moreover, it has beenrecently demonstrated that during the episodes of HBV reactiva-tion, the increase of viremia levels is paralleled by an increment ofserum IL-8 amounts and that these events precede the onset ofhepatic flare (Dunn et al., 2007; Tan et al., 2010).

In the current study, IL-8 was quantified in paired serum andliver tissue samples from chronically HBV-infected and HBV-negative subjects. Through this analysis, we found that in CHBpatients the presence of significantly higher serum concentrationsof IL-8 chemokine was paralleled by an upregulation of IL-8 geneexpression in the paired liver specimens compared to controls,suggesting that in CHB patients the high circulating amounts of IL-8 are likely derived from the liver. We also found that HBeAg-negative patients had significantly higher levels of IL-8 transcriptsin the liver than HBeAg-positive patients, and this observation isconsistent with the fact that HBeAg-negative patients are muchmore susceptible to flares of liver inflammation associated withrapid changes in HBV viremia levels than HBeAg-positive patients(EASL, 2012). Our data are also in accordance with the body ofevidence indicating that HBeAg-positive infection is usually asso-ciated with lower levels of pro-inflammatory cytokines and ingeneral with a more efficient inhibition of the host's innateimmune response (Chen et al., 2004; Visvanathan et al., 2007;Walsh and Locarnini, 2012).

By using a plasmid-free HBV transfection approach (Pollicinoet al., 2006), we were able to demonstrate that IL-8 genetranscription is specifically induced in response to HBV replicationand that the increased amounts of IL-8 directly correlate with thelevels of HBV transcription/replication. In accordance with thesein vitro results, we also found a significant correlation between IL-8 chemokine concentrations and HBV viremia levels in the HBVinfected patients. Altogether, these data indicate that active HBVreplication is able to directly induce IL-8 production and providenew evidence in support of an important role of the magnitude ofviral replication in the virus–host interplay, particularly in theactivation of the innate immune response (Ait-Goughoulte et al.,2010; Bertoletti et al., 2010). It has been shown that the X proteinof HBV is able to transactivate IL-8 gene expression through NF-kBand C/EBP-like cis-elements (Mahe et al., 1991). By the use of ourHBV replication system, we demonstrated that binding sites forNF-kB, AP-1, and C/EBP are required for maximum HBV-inducedIL-8 promoter activity as well as for maximal IL-8 protein produc-tion. In this context, our finding that in cells replicating HBx-defective HBV the activation of IL-8 gene promoter was signifi-cantly reduced, but not abolished, suggests that other HBV geneproducts might also be involved in the regulation of IL-8 geneexpression. This hypothesis is supported by available data showingthat in addition to HBx other viral proteins, such as the large andmiddle envelope proteins or the core and HBe proteins, have the

ability to deregulate the cellular transcription program (Ait-Goughoulte et al., 2010; Hildt et al., 2002; Locarnini et al., 2005).Our results showing that CHB patients infected with HBeAg-negative HBV had higher intrahepatic levels of IL-8 compared topatients infected with HBeAg-positive viral strains further confirmour hypothesis. Thus, the control of IL-8 gene expression by HBVappears to be complex and multifaceted with the involvement ofdifferent signaling pathways, as also indicated by the resultsobtained from the PCR arrays showing that HBV upregulates theexpression of a number of transcription factors and proteinkinases, which in principle have the ability to modulate NF-kB orAP-1 activity. In addition, in our HBV replication system theupregulation of IL-1 and TNF-alpha proinflammatory cytokines –

which are known to strongly induce IL-8 production (Hoffmannet al., 2002) – indicates that different mechanisms synergize toinduce IL-8 overexpression during active HBV replication.

By the use of ChIP assay, we found that active HBV replicationwas associated – both in vivo and in vitro – with a hyperacetyla-tion status of IL8 promoter and with the recruitment on thispromoter of PCAF and p300 acetyltransferases as well as of theHBx viral protein. Of interest, the presence of the class I deacety-lase HDAC1 on the IL-8 promoter was associated with a low HBVreplication state in vivo indicating that HBV may exert an epige-netic control of IL-8 gene by favouring the recruitment of chro-matin modifying enzymes onto the IL-8 promoter. The HBxregulatory protein might have an important role in this controlsince it has been shown that it is able to interact directly with theCREB-binding protein (CBP)/p300 to synergistically enhance itsactivity and to modify chromatin dynamics of target genes (Cougotet al., 2007). Altogether our results suggest that HBV mightactively contribute to the constitution of the “enhanceosome”-likestructure known to control IL-8 gene expression (Hoffmann et al.,2002).

Cytokines and chemokines released in response to virus infec-tions have the main role of recruiting inflammatory cells, ofconstraining virus replication and spread, and of inducing adaptiveimmunity (Biron, 2001; Guidotti and Chisari, 2000; Redpath et al.,2001). However, when the production of chemokines (like IL-8) inthe context of viral infections is continuous it may becomeharmful to the host (Luster, 1998; Mukaida, 2003). The existenceof a kind of vicious circle between IL-8 production and several viralinfections has been described (Alcorn et al., 2001; Girard et al.,2002; Khabar et al., 1997a; Khabar et al., 1997b; Murayama et al.,1994; Wagoner et al., 2007). It has been shown that humancytomegalovirus (CMV) can induce IL-8 gene transcription invarious types of cells and that IL-8 may enhance human CMVreplication (Murayama et al., 1994). In the present study, we foundthat IL-8 is able to induce HBV replication in a dose dependentmanner and that the induction of IL-8 gene expression by HBV isalso dose-dependent. These results suggest the presence of anamplifying loop between IL-8 production and HBV activity, whichin part might explain both the amplification of liver damage thatoccurs during the phase of immune reactivation of HBV and thedifficulties in finding a temporal correlation between immunolo-gical parameters and the kinetics of HBV viremia in the course ofhepatic flares (Bertoletti et al., 2010).

Accumulating evidence indicates that IL-8 induced by virusesmay also contribute to counteract IFN-α antiviral action (Girardet al., 2002; Lee et al., 2011; Polyak et al., 2001a, 2001b). Ourresults showing that the specific inhibition of IL-8 increases thepotency of IFN-α against HBV in vitro and that the addition ofrecombinant human IL-8 almost totally rescues HBV replicationduring IFN challenge strongly suggest that IL-8 expression inducedby HBV can impair the ability of endogenous IFN-α to inhibit earlystages of viral replication, thus favoring viral persistence, and canalso contribute to the poor response to IFN-α treatment.

T. Pollicino et al. / Virology 444 (2013) 317–328 325

This hypothesis is consistent with a great deal of evidenceindicating that viruses have evolved strategies to turn chemokineactivity to their own advantage (Alcami, 2003; Finlay andMcFadden, 2006; Kotwal et al., 2012). In that respect, the generalinflammatory microenvironment induced by HBV can be consid-ered an effective stratagem for manipulating the host's immuneresponse to achieve viral persistence and the establishmentof CHB.

In conclusion, HBV is able to directly induce the production ofthe IL-8 chemokine which in turn may contribute to the persis-tence of the infection, to the pathogenesis of viral related liverdamage as well as to the mechanisms of resistance to IFN.Consequently, inhibition of IL-8 production or activity could beinstrumental to the suppression of viral activity, to the reduction/block of liver inflammation and to the improvement of IFN-αantiviral action and it might open novel avenues for the develop-ment of effective immunotherapeutic approaches against HBV.

Materials and methods

Study Patients

We studied serum samples from 48 subjects who were seen inthe outpatient service of the Liver Unit at the Messina UniversityHospital (Table 1). Twenty-six of the 48 subjects (23 men andthree women; mean age, 45717 years) had a histologically provenHBV-related chronic liver disease whereas 22 (16 men and sixwomen; mean age, 37714.6 years) had no clinical, biochemicaland ultrasound signs of liver disease (control group). Of these 22subjects, seven were inactive HBV carriers (IBC) –with persistentlyserum HBV-DNA levelso2000 IU/mL and normal ALT (o40 U/L)over a 10-year median period of follow-up – and 15 wereindividuals negative for all HBV serum markers and without anysign of liver disease (HBV-neg subjects).

Eight of the 26 CHB patients were HBV “e” antigen (HBeAg)positive and 18 were HBeAg-negative and positive for the corre-sponding antibody (anti-HBe). All the seven inactive HBV carrierswere HBeAg-negative/anti-HBe positive.

Frozen liver biopsies were available for molecular analysesfrom 14 CHB patients (seven HBeAg-positive and seven anti-HBeAg positive). In addition, frozen liver tissues were also avail-able from five subjects of the control group, two inactive HBVcarriers and three HBV-neg individuals who underwent surgeryfor benign liver tumor or metastasis from extra-hepatic tumors.Liver samples from the control subjects were obtained fromunaffected areas of liver resections. All the subjects were negativefor hepatitis delta virus, hepatitis C virus and human immunode-ficiency virus serum markers and all CHB patients were treatment-naïve.

The study protocol was performed according to the principlesof the Declaration of Helsinki. The collection of blood and tissuesamples for research was approved by the ethics committee of theUniversity Hospital of Messina, Messina, Italy. Written informedconsent was obtained from all patients.

Chemicals and Antibodies

Phosphatidylinositol 3′ -kinase (PI3K)-Akt inhibitor LY294002,p38 kinase inhibitor SB203580, c-Jun N-terminal kinase (JNK)inhibitor JNKi II, mitogen-activated protein/ERK kinase/extracellu-lar signal-regulated kinase (MEK) inhibitor PD98059, SRC tyrosinekinase inhibitor PP2 and nuclear factor of kappa light polypeptidegene enhancer in B-cell (NFkB) inhibitor BAY 11-7082 were fromCalbiochem-Novabiochem Corp (San Diego, CA, USA). The activatorprotein-1 (AP-1) inhibitor Tanshinone IIA was from Enzo Life

Sciences (Farmingdale, NY USA). Recombinant human CXCL8/IL-8(208-IL) was from R&D SYSTEMS (Minneapolis, USA) and recom-binant human IFN-α2b was from Schering-Plough Corporation(Kenilworth, NJ, USA). Rabbit polyclonal antibody against AcH4(IgG recognizing histone H4, which is tetra-acetylated at lysines 6,9,13, and 17) (06-598) and rabbit polyclonal antibody againstHDAC1 (06-720) were from Upstate (Waltham, MA USA), rabbitpolyclonal antibody against p300 (sc-584) and rabbit polyclonalantibody against PCAF (sc-8999) were from Santa Cruz Biotech-nology (Santa Cruz, CA. USA), mouse monoclonal antibody againstHBx (MAI-081) was from Affinity Bioreagent (Rockford, IL USA).

Plasmid Constructs and full-length HBV DNA Genomes

Plasmid constructs containing the full-length wild-type (wtIL-8LUC) or truncated IL-8 promoters [IL-8 AP-1 mutant (IL-8-mAP1),IL-8 C/EBP mutant (IL-8-mC/EBP), IL-8 NF-kB mutant (IL-8-mNF-kB), and IL-8 double mutant for C/EBP and NF-kB (IL-8-C/EBP-NFkB)] controlling the expression of the luciferase gene weregenerated as described previously (Caristi et al., 2005; Venzaet al., 2007).

Monomeric linear full-length wild type (WT) and HBx mutantHBV genomes of genotype A (WTHBV-A and mutHBxHBV-A,respectively) were released from the pCR.HBV.A.EcoRI (Pollicinoet al., 2006) and pCR.mtHBx.A.EcoRI (Belloni et al., 2009) plasmidsusing EcorRI-PvuI restriction enzymes (New England BiolabsGmbH, Frankfurt, Germany). Monomeric linear full-length WTHBV genome of genotype D (WTHBV-D) was released from pUC.HBV.D plasmid using SapI restriction enzyme (New EnglandBiolabs) (Pollicino et al., 2006).

The X coding region of HBV genotype A was amplified by PCRand the resulting 465 bp fragment (positions 1374–1838 on theHBV map) was cloned into pcDNA3.1 HA cloning vector (Invitro-gen, Milan, Italy) (pcHBx-HA). Cloned HBx coding region wasverified by sequencing.

IL-8 ELISA

Patients sera and supernatants of transfected cells were testedfor IL-8 using the BD OptEIA human IL-8 ELISA kit II (BDBiosciences, San Jose, CA USA) where 100 μl of patient serum, cellculture supernatant or standard dilutions were analyzed accordingto the manufacturer's protocol. This assay has a lower limit ofdetection of 0.8 pg/ml.

Cell lines, Transfections, Treatments and luciferase reporter Assays

The HepG2 cell line was maintained in Dulbecco's modifiedEagle's medium with 10% fetal bovine serum (Life Technologies,Grand Island, NY USA), 1% penicillin/streptomycin, and 1% gluta-mine (Sigma-Aldrich S.r.l., Milan, Italy). For reporter assays, semi-confluent cells in 60-mm-diameter Petri dishes were transientlytransfected with 0.5 mg of monomeric linear wild-type HBVgenomes (WTHBV-A, mutHBxHBV-A or WTHBV-D) or 0.2 mg thepcHBX-HA expression vector and 0.5 mg of different pIL-8-LUCconstructs using the FuGENE transfection reagent (Roche, India-napolis, IN). Luciferase activity was determined 48 h later. Allexperiments were performed in duplicate and repeated at leastthree times. For transfection control, cells were cotransfected withthe plasmid pRL-TK (Clontech, Mountain View, CA), whichexpresses Renilla luciferase under the control of the thymidinekinase promoter. Luciferase levels in cell lysates were determinedby using the Dual-Luciferase Reporter Assay System (PromegaCorp, Madison, WI). The total amount of transfected DNA was keptconstant by adding pcDNA3.1 vector.

T. Pollicino et al. / Virology 444 (2013) 317–328326

For the kinase and AP-1 inhibition assay, drugs were added tocell culture medium 24 h post-transfection. The carrier dimethylsulfoxide (DMSO) was added as control. The concentrations ofdrugs used are as follows: SB203580 (25 μM), LY294002 (30 μM),JNKi II (25 μM), Bay 11-7082 (10 μM) PD98059 (30 μM), PP2(20 μM) and Tanshinone IIA (25 μM). At 16 h post-treatment, cellsand supernatants were harvested to test luciferase activity byluciferase reporter assay and IL-8 production by ELISA test,respectively.

For treatments with IFN-α, HepG2 cells were seeded in a 6-wellplate at a density of 2�105/well and transfected 24 h later with0.5 μg of monomeric linear WTHBV-A genomes or cotransfectedwith 0.5 μg of WTHBV-A genomes and 40 mM of IL-8 smallinterfering RNA (IL-8 siRNA: sc-39631, Santa Cruz Biotechnology)by using Lipofectamine 2000 (Invitrogen). IFN-αwas used at a finalconcentration of 1000 IU/ml and was added directly to the culturemedium starting at 3 h after transfection. HepG2 cells transfectedwith WTHBV-A genomes were also exposed to recombinanthuman IL-8 (rIL-8) that was used either alone or in combinationwith IFN-α at a final concentration of 10 ng/ml. Recombinant rIL-8was added to the culture medium at1 h after transfection. Cellculture medium plus IFN-α and/or rIL-8 was changed 1 day aftertransfection and cells were harvested at 48 h posttransfection. Indose-response experiments, rIL-8 was used at concentrations of10, 20 and 30 ng/ml. Transfection experiments included 0.5 μg ofpcGFP expression vector to assess transfection efficiency (range:40–50%).

Purification of core particles associated with HBV DNA from HBV-replicating cells

To purify HBV DNA from intracellular core particles, transfectedcells were washed once with ice-cold PBS and lysed in 50 mMTris–HCl, pH 7.4, 1 mmol EDTA, and 1% NP-40. Nuclei were pelletedby centrifugation for 1 min at 10,000 g. The supernatant wasadjusted to 100 mMMgCl2 and treated with 100 mg/ml of DNaseI for 30 min at 37 1C. The reaction was stopped by adding EDTA to afinal concentration of 25 mM. Protein was digested with 0.5 mg/mlproteinase K and 1% SDS for 2 h at 50 1C. Nucleic acids werepurified by phenol/chloroform (1:1) extraction and ethanol pre-cipitation adding glycogen and examined by Southern blot follow-ing standard procedures as previously described (Pollicino et al.,2007).

HBV RNAs and cellular mRNA analysis

Total RNA was extracted from liver tissue specimens andHepG2 cells at 48 h post-transfection with the TRIzol reagent (LifeTechnologies) as recommended by the manufacturer. RNA con-centration was measured using an ND-1000 spectrophotometer(NanoDrop Technologies, Wilmington, DE) at 260 nm. RNA qualityand quantity were monitored by ethidium bromide staining andby UV absorbance.

Northern blot analysis was performed following standardprocedures as previously described (Pollicino et al., 2007). Radio-active probes were prepared by random priming protocol, usingeither full-length HBV DNA or 18S cDNA templates and 32P α-dCTP(Amersham).

For quantification of IL8 mRNA in liver specimens and trans-fected HepG2 cells, 5 μg of extracted RNA for each sample weretreated with RQ1 RNase-free DNase (Promega, Madison, WI) for1 h at 37 1C, and used as a template for first-strand cDNA synthesiswith Transcriptor First Strand cDNA Synthesys kit (Roche, India-napolis, USA) and oligo(dT) primers according to manufacturer'sprotocol. Five ml of each cDNA were then quantified by real-timePCR analysis using the following IL-8 mRNA specific primers and

probes: forward primer 5′-ATGACTTCCAAGCTGGCCGTG-3′, reverseprimer 5′-TTTGATAAATTTGGGGTGGAAAGG

TT-3′. FRET hybridization probes 5′-AGCCTTCCTGATTTCTG-CAGCTCTGTGTGAAG-FL-3′, 5′-LC640-AGTTTTGCCAAGGAGTGCTAAA-GAACTTAGATG-PH-3′. The LightCycler h-G6PDH housekeeping GeneSet (Roche) was used to normalize the RNA samples. Real-time PCR bythe use of the “utility channel” of COBAS TaqMan 48 was performed asfollows: 95 1C for 10 min then 60 cycles of 95 1C for 30 s, 57 1C for20 s, 72 1C for 20 s. The plasmid pCRII-IL-8, containing one copy of theIL-8 coding sequence (nt 1 to 1551) was used as standard for IL-8cDNA quantification. The plasmid was digested with EcoRI (NewEngland Biolabs) and the IL-8 sequence was gel purified using theWizard SV Gel and PCR Clean-Up System (Promega, Madison, WIUSA). The concentration of purified IL-8 DNA was determined withthe ND-1000 spectrophotometer and the corresponding copy numberwas calculated. A series of 10-fold dilutions of the plasmid pCRII-IL-8was used as standard for IL-8 cDNA quantification.

For gene expression analysis by TaqMan Low-Density Arrays(TLDAs) (Applied Biosystems, Foster City, CA), total RNA fromtransfected HepG2 cells was extracted using the RNeasy Mini kit(Qiagen) according to the manufacturer's instruction. RNA (5 μg)was reverse transcribed using the Transcriptor First Strand cDNASynthesys kit and 200 ng of each cDNAwere loaded in double on acustomized TLDA with 18S RNA used as control. TLDAs were runon an ABI 7900HT Sequence detection system and, real-time PCRdata were collected and analyzed with the SDS2.2 software(Applied Biosystems)

ChIP assays

Forty-eight hours after transfection with linear HBV monomers,HepG2 cells were fixed in 1% formaldehyde, resuspended in 1 mlof ChIP lysis buffer (50 mM Tris–HCl, pH 8.0, 0.5% NP40, 1 mMEDTA, and 100 mM NaCl) and incubated 10 min at 4 1C. The lysatewas centrifuged to pellet the nuclei. Isolated cross-linked nucleiwere extracted with a 20 mM Tris–HCl, pH 8.0, 3 mM MgCl2,20 mM KCl buffer containing protease inhibitors, pelleted bymicrocentrifugation, and lysed by incubation in SDS lysis buffer(1% SDS, 10 mM EDTA, 50 mM Tris–HCl, pH 8.1) containingprotease inhibitors. The resulting chromatin solution was soni-cated for 10 pulses of 45 s at 80% power to generate 300- to 1000-bp DNA fragments using a Sonopuls HD 2070 Sonicator (Bandelinelectronic GmbH & Co. KG, Berlin, Germany). After microcentrifu-gation, the supernatant was diluted 1:10 in a dilution buffer (0.01%SDS, 1.1% Triton X-100, 1.2 mM EDTA, 16.7 mM Tris–HCl, pH 8.1,167 mM NaCl, containing protease inhibitors), precleared withblocked Protein G Plus (Pierce, Rockford, IL USA), and divided intoaliquots. The chromatin was then subjected to immunoprecipita-tion for 14–16 h at 4 1C using antibodies specific to AcH4, HDAC1,p300, PCAF and HBx. Immunoprecipitation with non-specificimmunoglobulins (Santa Cruz Biotechnology) was included ineach experiment as a negative control. After the reverse cross-linking step, immunoprecipitated chromatin was treated with20 U Plasmid-safe DNase (Epicenter, Madison, WI) for 1 h at37 1C to degrade contaminating HBV open circular species andthen subjected to PCR amplification using the HBV P23 and HBVP24 primers specific for the HBV cccDNA and encompassing theprecore/core promoter region of the viral genome (Pollicino et al.,2004).

To perform ChIP assays on liver tissues from chronically HBV-infected patients, liquid nitrogen frozen biopsy specimens (from0.5 to 1.0 cm) were lysed at 4 1C in 500 μL homogenization buffer(50 mM Tris–HCl, pH 8.0, 1 mM EDTA, 0.2% NP-40, 150 mM NaCl)with a TissueRupter instrument (Qiagen, Hilden, Germany). Nucleiwere isolated by centrifugation and then fixed in 1% formaldehydefor 15 min at 4 1C. After sonication and dilution, the chromatin was

T. Pollicino et al. / Virology 444 (2013) 317–328 327

subjected to immunoprecipitation for 14–16 h at 4 1C using anti-bodies specific to acH4, HDAC1, p300, PCAF and HBx. Immunopre-cipitations with nonspecific immunoglobulins (Santa CruzBiotechnology Inc.) were included in each experiment as anegative control. After the reverse cross-linking step, immunopre-cipitated chromatin was analyzed by PCR amplification usingcccDNA-specific primers.

Statistical analysis

Continuous variables were reported as median (range) andcategorical variables as frequencies (percentages). A nonpara-metric approach was used to examine variables showing theabsence of a normal distribution, as verified by the Kolmogorov–Smirnov test. The interdependence between numerical variableswas performed by the use of the Spearman Rank correlation,whereas the Mann–Whitney test was applied to perform compar-isons of continuously distributed variables between two indepen-dent groups. For in vitro experiments the non parametric ANOVAwas applied and the Kruskall Wallis test was used. Bar graphs wereplotted to show the mean7standard deviation (SD) of at leastthree independent experiments. P values o0.05 were consideredas statistically significant. Statistical analysis was performed withSPSS version 12.0 software package (SPSS Inc, Chicago, IL).

Acknowledgment

This study was supported by Associazione Italiana per laRicerca sul Cancro (AIRC) grant (IG 12022).

References

Ait-Goughoulte, M., Lucifora, J., Zoulim, F., Durantel, D., 2010. Innate antiviralimmune responses to hepatitis B virus. Viruses 2 (7), 1394–1410.

Alcami, A., 2003. Viral mimicry of cytokines, chemokines and their receptors. Nat.Rev. Immunol. 3 (1), 36–50.

Alcorn, M.J., Booth, J.L., Coggeshall, K.M., Metcalf, J.P., 2001. Adenovirus type7 induces interleukin-8 production via activation of extracellular regulatedkinase 1/2. J. Virol. 75 (14), 6450–6459.

Belloni, L., Pollicino, T., De Nicola, F., Guerrieri, F., Raffa, G., Fanciulli, M., Raimondo, G.,Levrero, M., 2009. Nuclear HBx binds the HBV minichromosome and modifies theepigenetic regulation of cccDNA function. Proc. Natl. Acad. Sci. U. S. A. 106 (47),19975–19979.

Bertoletti, A., Maini, M.K., Ferrari, C., 2010. The host-pathogen interaction duringHBV infection: immunological controversies. Antivir. Ther. 15 (Suppl 3), 15–24.

Biron, C.A., 2001. Interferons alpha and beta as immune regulators–a new look.Immunity. 14 (6), 661–664.

Caristi, S., Piraino, G., Cucinotta, M., Valenti, A., Loddo, S., Teti, D., 2005. Prosta-glandin E2 induces interleukin-8 gene transcription by activating C/EBPhomologous protein in human T lymphocytes. J. Biol. Chem. 280 (15),14433–14442.

Chen, M.T., Billaud, J.N., Sallberg, M., Guidotti, L.G., Chisari, F.V., Jones, J., Hughes, J.,Milich, D.R., 2004. A function of the hepatitis B virus precore protein is toregulate the immune response to the core antigen. Proc. Natl. Acad. Sci. U. S. A.101 (41), 14913–14918.

Chisari, F.V., Ferrari, C., 1995. Hepatitis B virus immunopathogenesis. Annu. Rev.Immunol. 13, 29–60.

Cougot, D., Wu, Y., Cairo, S., Caramel, J., Renard, C.A., Levy, L., Buendia, M.A.,Neuveut, C., 2007. The hepatitis B virus X protein functionally interacts withCREB-binding protein/p300 in the regulation of CREB-mediated transcription. J.Biol. Chem. 282 (7), 4277–4287.

Dienstag, J.L., 2008. Hepatitis B virus infection. N. Engl. J. Med. 359 (14), 1486–1500.Dunn, C., Brunetto, M., Reynolds, G., Christophides, T., Kennedy, P.T., Lampertico, P.,

Das, A., Borrow, P., Williams, K., Humphreys, E., Afford, S., Adams, D.H.,Bertoletti, A., Maini, M.K., Lopes, A.R., 2007. Cytokines induced during chronichepatitis B virus infection promote a pathway for NK cell-mediated liverdamage. J. Exp. Med. 204 (3), 667–680.

EASL, 2012. EASL clinical practice guidelines: management of chronic hepatitis Bvirus infection. J. Hepatol. 57 (1), 167–185.

Finlay, B.B., McFadden, G., 2006. Anti-immunology: evasion of the host immunesystem by bacterial and viral pathogens. Cell 124 (4), 767–782.

Gehring, A.J., Koh, S., Chia, A., Paramasivam, K., Chew, V.S., Ho, Z.Z., Lee, K.H., Maini,M.K., Madhavan, K., Lim, S.G., Bertoletti, A., 2011. Licensing virus-specific T cellsto secrete the neutrophil attracting chemokine CXCL-8 during hepatitis B virusinfection. PLoS One 6 (8), e23330.

Girard, S., Shalhoub, P., Lescure, P., Sabile, A., Misek, D.E., Hanash, S., Brechot, C.,Beretta, L., 2002. An altered cellular response to interferon and up-regulation ofinterleukin-8 induced by the hepatitis C viral protein NS5A uncovered bymicroarray analysis. Virology. 295 (2), 272–283.

Green, J., Khabar, K.S., Koo, B.C., Williams, B.R., Polyak, S.J., 2006. Stability of CXCL-8and related AU-rich mRNAs in the context of hepatitis C virus replicationin vitro. J. Infect. Dis. 193 (6), 802–811.

Guidotti, L.G., Chisari, F.V., 2000. Cytokine-mediated control of viral infections.Virology 273 (2), 221–227.

Herceg, Z., Paliwal, A., 2011. Epigenetic mechanisms in hepatocellular carcinoma:how environmental factors influence the epigenome. Mutat. Res. 727 (3),55–61.

Hildt, E., Munz, B., Saher, G., Reifenberg, K., Hofschneider, P.H., 2002. The PreS2activator MHBs(t) of hepatitis B virus activates c-raf-1/Erk2 signaling intransgenic mice. EMBO J. 21 (4), 525–535.

Hisatsune, J., Nakayama, M., Isomoto, H., Kurazono, H., Mukaida, N., Mukhopadhyay, A.K., Azuma, T., Yamaoka, Y., Sap, J., Yamasaki, E., Yahiro, K., Moss, J., Hirayama, T.,2008. Molecular characterization of Helicobacter pylori VacA induction of IL-8 inU937 cells reveals a prominent role for p38MAPK in activating transcription factor-2, cAMP response element binding protein, and NF-kappaB activation. J. Immunol.180 (7), 5017–5027.

Hoffmann, E., Dittrich-Breiholz, O., Holtmann, H., Kracht, M., 2002. Multiple controlof interleukin-8 gene expression. J. Leukocyte Biol. 72 (5), 847–855.

Kakimi, K., Lane, T.E., Wieland, S., Asensio, V.C., Campbell, I.L., Chisari, F.V., Guidotti,L.G., 2001. Blocking chemokine responsive to gamma-2/interferon (IFN)-gamma inducible protein and monokine induced by IFN-gamma activityin vivo reduces the pathogenetic but not the antiviral potential of hepatitis Bvirus-specific cytotoxic T lymphocytes. J. Exp. Med. 194 (12), 1755–1766.

Kew, M.C., 2011. Hepatitis B virus x protein in the pathogenesis of hepatitis B virus-induced hepatocellular carcinoma. J. Gastroenterol. Hepatol. 26 (Suppl 1),144–152.

Khabar, K.S., Al-Zoghaibi, F., Al-Ahdal, M.N., Murayama, T., Dhalla, M., Mukaida, N.,Taha, M., Al-Sedairy, S.T., Siddiqui, Y., Kessie, G., Matsushima, K., 1997a. Thealpha chemokine, interleukin 8, inhibits the antiviral action of interferon alpha.J. Exp. Med. 186 (7), 1077–1085.

Khabar, K.S., Al-Zoghaibi, F., Murayama, T., Matsushima, K., Mukaida, N., Siddiqui, Y.,Dhalla, M., Al-Ahdal, M.N., 1997b. Interleukin-8 selectively enhances cytopathiceffect (CPE) induced by positive-strand RNA viruses in the human WISH cellline. Biochem. Biophys. Res. Commun. 235 (3), 774–778.

Kotwal, G.J., Hatch, S., Marshall, W.L., 2012. Viral infection: an evolving insight intothe signal transduction pathways responsible for the innate immune response.Adv. Virol. 2012, 131457.

Lee, C.M., Yen, Y.H., Hung, C.H., Lu, S.N., Wang, J.H., Wang, J.C., Chen, C.H., Kee, K.M.,Hu, T.H., Changchien, C.S., 2011. Liver interleukin-8 messenger RNA expressionand interferon sensitivity-determining region mutations relate to treatmentresponse in hepatitis C 1b. Antivir. Ther. 16 (6), 825–832.

Locarnini, S., Shaw, T., Dean, J., Colledge, D., Thompson, A., Li, K., Lemon, S.M., Lau, G.G., Beard, M.R., 2005. Cellular response to conditional expression of thehepatitis B virus precore and core proteins in cultured hepatoma (Huh-7) cells.J. Clin Virol. 32 (2), 113–121.

Luster, A.D, 1998. Chemokines–chemotactic cytokines that mediate inflammation.N. Engl. J. Med. 338 (7), 436–445.

Mahe, Y., Mukaida, N., Kuno, K., Akiyama, M., Ikeda, N., Matsushima, K., Murakami,S., 1991. Hepatitis B virus X protein transactivates human interleukin-8 genethrough acting on nuclear factor kB and CCAAT/enhancer-binding protein-likecis-elements. J. Biol. Chem. 266 (21), 13759–13763.

Maini, M.K., Boni, C., Lee, C.K., Larrubia, J.R., Reignat, S., Ogg, G.S., King, A.S., Herberg,J., Gilson, R., Alisa, A., Williams, R., Vergani, D., Naoumov, N.V., Ferrari, C.,Bertoletti, A., 2000. The role of virus-specific CD8(+) cells in liver damage andviral control during persistent hepatitis B virus infection. J. Exp. Med. 191 (8),1269–1280.

Mukaida, N., 2003. Pathophysiological roles of interleukin-8/CXCL8 in pulmonarydiseases. Am. J. Physiol. Lung. Cell Mol. Physiol. 284 (4), L566–577.

Murayama, T., Kuno, K., Jisaki, F., Obuchi, M., Sakamuro, D., Furukawa, T., Mukaida,N., Matsushima, K., 1994. Enhancement human cytomegalovirus replication ina human lung fibroblast cell line by interleukin-8. J. Virol. 68 (11), 7582–7585.

Park, I.Y., Sohn, B.H., Yu, E., Suh, D.J., Chung, Y.H., Lee, J.H., Surzycki, S.J., Lee, Y.I.,2007. Aberrant epigenetic modifications in hepatocarcinogenesis induced byhepatitis B virus X protein. Gastroenterology 132 (4), 1476–1494.

Paschos, K., Allday, M.J., 2010. Epigenetic reprogramming of host genes in viral andmicrobial pathogenesis. Trends Microbiol. 18 (10), 439–447.

Perrillo, R.P., 2001. Acute flares in chronic hepatitis B: the natural and unnaturalhistory of an immunologically mediated liver disease. Gastroenterology 120 (4),1009–1022.

Pollicino, T., Belloni, L., Raffa, G., Pediconi, N., Squadrito, G., Raimondo, G., Levrero,M., 2006. Hepatitis B virus replication is regulated by the acetylation status ofhepatitis B virus cccDNA-bound H3 and H4 histones. Gastroenterology 130 (3),823–837.

Pollicino, T., Raffa, G., Costantino, L., Lisa, A., Campello, C., Squadrito, G., Levrero, M.,Raimondo, G., 2007. Molecular and functional analysis of occult hepatitis Bvirus isolates from patients with hepatocellular carcinoma. Hepatology 45 (2),277–285.

Pollicino, T., Squadrito, G., Cerenzia, G., Cacciola, I., Raffa, G., Craxi, A., Farinati, F.,Missale, G., Smedile, A., Tiribelli, C., Villa, E., Raimondo, G., 2004. Hepatitis Bvirus maintains its pro-oncogenic properties in the case of occult HBV infection.Gastroenterology 126 (1), 102–110.

T. Pollicino et al. / Virology 444 (2013) 317–328328

Polyak, S.J., Khabar, K.S., Paschal, D.M., Ezelle, H.J., Duverlie, G., Barber, G.N., Levy, D.E.,Mukaida, N., Gretch, D.R., 2001a. Hepatitis C virus nonstructural 5 A protein inducesinterleukin-8, leading to partial inhibition of the interferon-induced antiviralresponse. J. Virol. 75 (13), 6095–6106.

Polyak, S.J., Khabar, K.S., Rezeiq, M., Gretch, D.R., 2001b. Elevated levels ofinterleukin-8 in serum are associated with hepatitis C virus infection andresistance to interferon therapy. J. Virol. 75 (13), 6209–6211.

Rajaiya, J., Xiao, J., Rajala, R.V., Chodosh, J., 2008. Human adenovirus type 19infection of corneal cells induces p38 MAPK-dependent interleukin-8 expres-sion. Virol. J. 5, 17.

Redpath, S., Angulo, A., Gascoigne, N.R., Ghazal, P., 2001. Immune checkpoints inviral latency. Annu. Rev. Microbiol. 55, 531–560.

Rehermann, B., 2003. Immune responses in hepatitis B virus infection. Semin. LiverDis. 23 (1), 21–38.

Sitia, G., Isogawa, M., Iannacone, M., Campbell, I.L., Chisari, F.V., Guidotti, L.G., 2004.MMPs are required for recruitment of antigen-nonspecific mononuclear cellsinto the liver by CTLs. J. Clin. Invest. 113 (8), 1158–1167.

Tan, A.T., Koh, S., Goh, W., Zhe, H.Y., Gehring, A.J., Lim, S.G., Bertoletti, A., 2010. Alongitudinal analysis of innate and adaptive immune profile during hepaticflares in chronic hepatitis B. J. Hepatol. 52 (3), 330–339.

Taub, D.D., Anver, M., Oppenheim, J.J., Longo, D.L., Murphy, W.J., 1996. T lymphocyterecruitment by interleukin-8 (IL-8). IL-8-induced degranulation of neutrophilsreleases potent chemoattractants for human T lymphocytes both in vitro andin vivo. J. Clin. Invest. 97 (8), 1931–1941.

Venza, I., Cucinotta, M., Caristi, S., Mancuso, G., Teti, D., 2007. Transcriptionalregulation of IL-8 by Staphylococcus aureus in human conjunctival cellsinvolves activation of AP-1. Invest. Ophthalmol. Vis. Sci. 48 (1), 270–276.

Visvanathan, K., Skinner, N.A., Thompson, A.J., Riordan, S.M., Sozzi, V., Edwards, R.,Rodgers, S., Kurtovic, J., Chang, J., Lewin, S., Desmond, P., Locarnini, S, 2007.Regulation of Toll-like receptor-2 expression in chronic hepatitis B by theprecore protein. Hepatology 45 (1), 102–110.

Wagoner, J., Austin, M., Green, J., Imaizumi, T., Casola, A., Brasier, A., Khabar, K.S.,Wakita, T., Gale Jr., M., Polyak, S.J., 2007. Regulation of CXCL-8 (interleukin-8)induction by double-stranded RNA signaling pathways during hepatitis C virusinfection. J. Virol. 81 (1), 309–318.

Walsh, R., Locarnini, S., 2012. Hepatitis B precore protein: pathogenic potential andtherapeutic promise. Yonsei Med. J. 53 (5), 875–885.

Yu, Y., Gong, R., Mu, Y., Chen, Y., Zhu, C., Sun, Z., Chen, M., Liu, Y., Zhu, Y., Wu, J., 2011.Hepatitis B virus induces a novel inflammation network involving threeinflammatory factors, IL-29, IL-8, and cyclooxygenase-2. J. Immunol. 187 (9),4844–4860.

Zheng, J.C., Huang, Y., Tang, K., Cui, M., Niemann, D., Lopez, A., Morgello, S., Chen, S.,2008. HIV-1-infected and/or immune-activated macrophages regulate astro-cyte CXCL8 production through IL-1beta and TNF-alpha: involvement ofmitogen-activated protein kinases and protein kinase R. J. Neuroimmunol.200 (1-2), 100–110.

Zimmermann, H.W., Seidler, S., Gassler, N., Nattermann, J., Luedde, T., Trautwein, C.,Tacke, F., 2011. Interleukin-8 is activated in patients with chronic liver diseasesand associated with hepatic macrophage accumulation in human liver fibrosis.PLoS One 6 (6), e21381.