Gene Therapy (2003) 10, 1861–1873 & 2003 Nature Publishing Group All rights reserved 0969-7128/03 $25.00 www.nature.com/gt RESEARCH ARTICLE Exploiting activation of NFkBto deliver HCV-responsive expression of interferons a and g

AA Matskevich and DS Strayer Department of Pathology and Cell Biology, Jefferson Medical College, Philadelphia, PA, USA˙

Chronic with (HCV) may lead to and the magnitude of induction in response to HCV, were liver failure and . Current treatment greater in hepatocyte lines than in primary cultured hepato- for HCV includes high systemic doses of interferona (IFNa), cytes. However, even in the latter, supernatant IFNa which is effective in less than half of patients and may have concentrations achieved by this approach were similar to severe side effects. We designed conditional IFNa and IFNg therapeutic serum concentrations sought in systemic IFNa- expression constructs to be triggered by HCV-induced treated patients. In coculture studies, secreted IFNa acti- activation of NFkB, and delivered these using highly efficient vated its cognate response elements in untransduced cells, recombinant Tag-deleted SV40-derived vectors. NFkB acti- suggesting that its potential inhibitory effects on HCV may vates the HIV-1NL4-3 long terminal repeat (HIVLTR) as a not be limited to transduced cells. Although HCV replication , which accounts for the conditional transgene in culture is difficult to assess, HCV-induced IFNa production expression. hepatocyte lines and primary rat demonstrably reduced HCV . Conditional ex- hepatocytes (PRH) were transduced with SV[HIVLTR](IFN) pression of IFNs within the liver may represent an attractive vectors, and transfected with HCV cDNA. Production of approach to therapy of severe chronic HCV infection that human and murine IFNa and IFNg in cytosol and culture could avoid the side effects of systemic treatment regimens. supernatants was measured. HCV activated the HIVLTR to Gene Therapy (2003) 10, 1861–1873. doi:10.1038/ produce and secrete IFNs, and did so largely through the sj.gt.3302091 NFkB binding sites of the HIVLTR. Levels of IFNs secreted,

Keywords: hepatocytes; gene therapy; SV40 vectors

Introduction achieve therapeutic levels in the liver, unpleasant or toxic side effects, viral resistance to rhIFNa that is associated The interferons (IFN) are a family of proteins formally with certain HCV genotypes, and the impermanence divided into types I and II. The former (including IFNa of the therapeutic effect in many cases if therapy is and IFNb), of which there are several isoforms, are interrupted prematurely.3,4 produced by most cell types in response to virus An approach to treatment in which cDNAs for IFNs a infection and other stimuli. The latter (also called IFNg) and IFNg are delivered directly to the liver, that is the is mainly made by T lymphocytes and is important in therapeutic target, could theoretically magnify the local activating certain T-cell and NK responses. Although IFNg therapeutic effect without requiring high systemic levels is an effective antiviral and antineoplastic agent, its or repeated administration of the toxic and short-lived therapeutic effectiveness has been limited by the high proteins. Thus, gene delivery to the liver might poten- systemic toxicity of systemically administered recombi- tially be an adjunct to current therapies. For such liver- nant IFNg protein. directed gene therapy to be effective, a large number of Recombinant human IFNa (rhIFNa) is less toxic, and hepatocytes should be transduced, the DNA delivered is approved in the US for treatment of chronic hepatitis C must remain in these cells indefinitely, and, if possible, virus (HCV) infection. In combination with ribavirin, expression of IFN should respond to the presence of the rhIFNa is administered intensively to patients with offending virus but otherwise be quiescent. progressive liver damage due to chronic hepatitis C.1 For these studies, we used recombinant gene delivery This approach is effective in controlling the progression vectors derived from Tag-deleted simian virus-40 of chronic HCV infection in slightly less than half of (rSV40 s). For gene delivery to the liver, rSV40 vectors patients.2 Among the reasons for treatment failure are the have been shown to deliver enduring transgene expres- short half-life of rhIFNa in the blood, the necessity of sion to very high percentages of unselected hepatocytes, maintaining high blood levels of rhIFNa in order to whether resting or dividing, in vitro or in vivo.5,6 The goal of rendering transgene expression responsive to the presence of the offending HCV virus was Correspondence: Dr DS Strayer, Department of Pathology, Jefferson Medical College, 1020 Locust Street, Room 251, Philadelphia, PA 19107, approached by exploiting the reported ability of the USA HCV gene product NS5a to activate the cellular Received 04 October 2002; accepted 22 April 2003 transcription factor NFkB,7 as well as the known HCV-activated conditional expression vectors AA Matskevich and DS Strayer 1862 a

Pmel

IFN cDNA

VP1

HIV-1 LTR

VP3 ori

enh polyA VP2 ses SV40-EP

b

CMV- Full length HCV Termination Amp- IEP genomic cDNA signal resistance

pRC/CMV-HCV

Figure 1 (a) map of a generic SV[HIVLTR](IFN). In all of these constructs, the HIV-1 LTR from HIV-1NL4-3 was used as a promoter to drive expression of the IFN cDNA in question. The LTR is positioned just 30 (counter-clockwise) to the SV40 early promoter (SV40-EP). Since it overlaps the SV40 (ori), the SV40-EP must be retained intact in order to produce these vectors. To block transcription from it, four tandem signals (mpA) were inserted. The SV40 genes (VP1, VP2, and VP3) and regulatory sequences are included in these vectors (enh¼; ses¼SV40 encapsidation signal), but are not expressed. (b) Structure of plasmid pRC/CMV-HCV, containing full genomic cDNA form of HCV, strain 1b, under the control of immediate-early promoter (CMV-IEP).

responsiveness of the promoter activity of the HIV-1 long transfected HCV cDNA in activating HIVLTR promoter terminal repeat (LTR) to NFkB.8 Accordingly, we activity. We then evaluated the effectiveness of the engineered rSV40 vectors to carry the HIVLTR as a SV[HIVLTR](IFN) vectors in inhibiting HCV. These promoter, to drive expression of murine and human IFNa vectors were constructed as described previously9 and and IFNg as transgenes. The effectiveness of HCV in in Materials and methods. A map of a generic vector activating these constructs to produce and secrete the such as an SV[HIVLTR](IFN) vector, is shown in Figure several IFNs was tested, as was the ability of the IFNs so 1a. The HCV cDNA-carrying plasmid used in these generated to activate paracrine signaling through IFN- studies is shown in Figure 1b. signal response elements (ISRE) and the effect of IFN on This strategy relies upon the ability of transfected detectable HCV transcript. HCV cDNA to be transcribed upon transfection into liver cells. Normally, the HCV genome is transcribed as a single mRNA encoding a polyprotein that is subse- Results quently post-translationally cleaved into its several substituents.10 We therefore transfected HepG2 cells with HCV-induced expression of IFN delivered by the HCV cDNA construct and tested for its transcription SV[HIVLTR]IFN using nested RT-PCR, with both sets of PCR primers Our strategy for these studies entailed transducing cells chosen to amplify the NS5a region of this transcript, as with SV[HIVLTR](IFN) vectors, then transfecting them described in Materials and methods. We found that with HCV cDNA to examine first the effectiveness of the transfection of HCV cDNA results in its transcription,

Gene Therapy HCV-activated conditional expression vectors AA Matskevich and DS Strayer 1863 and documented that the observed PCR product was Mock SV[wtHIVLTR]IFNα + HCV derived from the transcribed HCV mRNA (Figure 2). SV[BUGT] alone SV[muHIVLTR]IFNα + HCV SV[wtHIVLTR]IFNα alone HCV alone Sequence analysis of this RT-PCR product was per- SV[muHIVLTR]IFNα alone SV[BUGT] + HCV formed (not shown), and documented that the ampli- a cand was in fact the appropriate part of the coding 70 sequence for NS5a. 60 In order to evaluate the potential therapeutic utility of rSV40-delivered, HIVLTR-driven IFN expression, we 50 tested the time course of responsiveness to HCV transfection of IFN production and secretion by cells 40 transduced with these constructs. Thus, HepG2 cells

/mg protein, ± SEM 30 were transduced with SV[HIVLTR](IFN) constructs, then α transiently transfected with HCV cDNA. IFN production 20 was measured by ELISA in cell lysates and in culture supernatants. Although the specific IFN levels and pg mlFN 10 kinetics of responsiveness to HCV varied among the different IFNs, transfection of the HCV genome into b 0 transduced cells significantly increased IFN, both in- 80 tracellularly and secreted into culture media, as a 70 function of time (shown in Figure 3 for mouse IFNa) (CmIFNa). The peak production and secretion of IFN 60 induced by transient transfection with HCV cDNA 50 occurred 2–3 days post-transfection. 40

± SEM 30

HCV activation of HIVLTR occurs via NFkB binding /ml culture supernatant, sites in the HIVLTR α 20 It has been reported that HCV NS5a activates NFkB, 10 which, in turn, is known to activate the HIVLTR as a pg mlFN 0 promoter.7,8 In order to determine if NFkB was involved 1234 in the activation of IFN expression demonstrated above, Days post-transfection with HCV cDNA HepG2 cells were transduced with SV[wtHIVLTR](IFN) , or with SV[muHIVLTR](IFN) vectors. Compara- Figure 3 Effect of HCV on mIFNa production and secretion. HepG2 cells were transduced with 100 MOI of SV[wtHIVLTR]mIFNa, or with control vectors SV[muHIVLTR]mIFNa or SV(BUGT), or mock-transduced. They were then transfected with HCV cDNA 4 days later. IFN protein concentrations were measured by ELISA 1, 2, 3, and 4 days post- transfection. (a) mIFNa in cell lysates, expressed as pg/mg total protein. (b) mIFNa in culture supernatants, expressed as pg/ml culture super- natant. Data represent means of independent experiments repeated a Mock, DNase HCV, RNase HCV, RNase Mock, DNase Mock, RNase Mock, RNase HCV, DNase HCV, DNase 100 bp ladder 100 bp ladder minimum of three times. Standard error bars are shown. For days 2 and 3, the differences between SV[wtHIVLTR](IFN) þ HCV groups and all other groups were significant at the level of Po0.05 or better. 1st round, 2nd round, outer primers inner primers tive production of IFNs was assayed by ELISA in culture supernatants and in cell lysates (shown in Figure 4 for hIFNa). HCV activated wtHIVLTR to cause the cells to produce and secrete hIFNa. Production and secretion of hIFNa was comparable in cultures transduced with SV[wtHIVLTR](hIFNa) but without transfected HCV cDNA, in cultures transduced with SV[muHIVLTR](hIF- Na) with or without transfected HCV cDNA, and in untransduced cultures. Cultures that had been treated with SV[wtHIVLTR](hIFNa) þ HCV cDNA consistently produced greater amounts of IFN, and secreted it at much higher levels than did any of the control groups: at the minimum Po0.05 for all comparisons between SV[wtHIVLTR](hIFNa) þ HCV and all other groups. These data are further illustrated by the time course Figure 2 Detection of HCV transcripts by nested RT-PCR. HepG2 cells studies shown in Figure 3, which include cells trans- were transfected with pRC/CMV-HCV. Direct nested RT-PCR was duced with SV[muHIVLTR](hIFNa) in comparison to the performed on total RNA extracted from these cells 48 h post-transfection other treatment groups. These data also show that HCV as described in Materials and methods. To ascertain the accuracy of these a results, all samples were first treated with DNase; some were also treated did not activate significant hIFN secretion in cells with RNase before the reverse transcription step, as described in Materials receiving SV[muHIVLTR](IFN), in contrast to significant and methods. Results are representative of two independent RT-PCR secretion by cells receiving SV[wtHIVLTR](IFN) in studies. response to HCV.

Gene Therapy HCV-activated conditional expression vectors AA Matskevich and DS Strayer 1864 ab

60 60

50 50

40 40

30 30 /mg protein, ± SEM

α 20 20 SV[BUGT] + HCV SV[muHIVLTR]IFN + HCV SV[wtHIVLTR]IFN + HCV

SV[wtHIVLTR]IFN PROBE SV[muHIVLTR]IFN SV[BUGT] Mock + HCV /ml culture supernatant, ± SEM

10 α 10 pg hlFN 0 0 pg hlFN Mock Mock ) alone ) alone ) alone ) alone ) + HCV ) + HCV ) + HCV ) + HCV α α α α α α α α Mock + HCV Mock + HCV SV(BUGT) alone SV(BUGT) alone

SV(BUGT) + HCV SV(BUGT) + HCV mlFNα SV[wtHIVLTR](IFN SV[muHIVLTR](IFN SV[wtHIVLTR](IFN SV[wtHIVLTR](hIFN SV[muHIVLTR](IFN SV[muHIVLTR](hIFN SV[wtHIVLTR](hIFN SV[muHIVLTR](hIFN

Figure 4 Activation of wtHIVLTR by HCV induces production and β-actin secretion of hIFNa. HepG2 cells were transduced with 100 MOI of SV[wtHIVLTR]hIFNa, or with control vectors SV[muHIVLTR]hIFNa or SV(BUGT), or mock-transduced. They were then transfected with HCV cDNA 4 days later. HIFNa protein concentrations were measured by ELISA 48 h post-transfection: in cell lysates (a), shown as pg IFN/mg total protein; and in culture supernatants (b) shown as pg IFN/ml culture supernatant. Data represent means of independent experiments repeated a minimum of three times. Standard error bars are shown. IFN production and secretion levels for SV[HIVLTR](IFN) þ HCV groups were signifi- cantly different from all other groups (at the minimum Po0.05 for all such comparisons). SV[wtHIVLTR]IFN SV[muHIVLTR]IFN + HCV SV[BUGT]+HCV SV[muHIVLTR]IFN Mock + HCV SV[wtHIVLTR]IFN + HCV SV[BUGT] + HCV

Detection of IFN transcripts and proteins elicited by HCV in SV[wtHIVLTR](IFN)-transduced cells The ELISA data described above were further assessed γ by Northern and Western analyses, respectively, of cell mlFN lysates and culture supernatants. Thus, HepG2 cells were transduced with SV[wtHIVLTR](IFN), 7HCV, or with SV[muHIVLTR](IFN), þ HCV. Control cultures were transduced with SV(BUGT) (BUGT¼human bilirubin- UDP-glucuronysyl-transferase) as a control rSV40 vector, or were mock-transduced. RNA from these cultures was harvested after transfection with HCV cDNA, electro- β-actin phoresed, blotted, and hybridized with radiolabeled cDNA probes for the IFNs used under very stringent conditions (final wash: 0.1 Â SSC, 0.1% SDS, 551C). Figure Figure 5 HCV activation of HIV-1 LTR to yield mIFNg and mIFNa 5 shows representative results, using mIFNa and mIFNg expression. HepG2 cells were transduced with the several SV[HIVL- TR]IFN constructs shown, then transfected with HCV 4 days later, or as the transgenes; comparable results were obtained with mock-transfected. Total RNA was harvested from these cells 2 days after hIFNa and hIFNg. Blots were reprobed for b-actin as a transfection, electrophoresed, blotted to nylon membranes, and then loading control. The only cultures that demonstrated hybridized with the respective IFN cDNA probes, labeled with a-32P- mIFN production by this assay were those that had been dCTP. Results are representative of three Northern blot analyses performed both transduced with SV[wtHIVLTR](IFN) and trans- on three sets of cells transduced independently. They are also representative fected with HCV cDNA. Other cultures were negative, of comparable data seen with human IFNa and IFNg. save for small amounts of IFN mRNA in the group receiving SV[wtHIVLTR](IFN) without HCV challenge. but not challenged with HCV. The absence of detectable Culture supernatants and cell lysates were also IFN in cultures transduced with SV[muHIVLTR](IFN) analyzed by Western blotting, using antisera specific vectors underscores the importance of HCV activation of for the individual IFNs. Figure 6 shows the representa- HIVLTR via NFkB. However, the presence of cross- tive studies from cultures transduced with rSV40 vectors reactive proteins in all test groups may explain the carrying mIFNg as a transgene. In both cell lysates and apparent ‘background’ detection of ‘IFN’ in control culture supernatants, the 21 kDa mIFNg protein was groups in Figures 3 and 4 (and Figure 9). mainly detected in cultures that had been both trans- duced with SV[wtHIVLTR](mIFNg) and transfected with Frequency of IFN-producing cells HCV cDNA. Slight intracellular IFNg was observed in The transduction efficiency with rSV40 vectors in lysatesfromcells transducedwithSV[wtHIVLTR](mIFNg) unselected cultures is generally 498%.9 In order to

Gene Therapy HCV-activated conditional expression vectors AA Matskevich and DS Strayer 1865 transcript in cultures transfected with HCV cDNA. By this technique, approximately 10% of HepG2 cells in cultures transfected with HCV cDNA expressed HCV mRNA (Figure 7c). We also asked if prior transduction with an rSV40 vector altered the transfection efficiency of transduced Culture cells. That is, we asked whether the differences in SV[BUGT] + HCV

Markers supernatants SV[BUGT] Mock + HCV SV[wtHIVLTR]IFN+HCV SV[muHIVLTR]IFN SV[wtHIVLTR]IFN SV[muHIVLTR]IFN+HCV IFN expression among the various study and control Antibody groups might represent differences in the success probe used of the secondary transfection step that followed trans- duction. Accordingly, HepG2 cells were transduced with SV[wtHIVLTR](IFN), SV[muHIVLTR](IFN) or SV(BUGT), or mock-transduced. They were then trans- fected with pCMVluc, respectively carrying enhanced green fluorescent protein and luciferase as reporter mlFNγ genes. There was no significant difference in luciferase 21 kDa activity (Figure 8) between any of the experimental groups (P40.1), indicating that transduction did not produce significant differences in transfection efficiency. These data were confirmed by quantitation of fluorescent cells on cotransfection with pT7egfp (data not shown).

Effectiveness of SV[HIVLTR](IFN) activation by HCV SV[muHIVLTR]FN +HCV Mock + HCV SV[wtHIVLTR]FN+ HCV Markers SV[wtHIVLTR]FN SV[BUGT] + HCV SV[muHIVLTR]FN Cell lysates in primary hepatocytes In the light of these results achieved using a hepato- Antibody probe used blastoma cell line, we asked if SV[HIVLTR}(IFN) vectors would show comparable responsiveness to HCV in primary hepatocytes. Accordingly, PRH were cultured from normal rat livers, and transduced with the IFN- carrying vectors. Cultures were then ‘challenged’ by transfecting HCV cDNA, and intracellular and secreted IFN concentrations were measured by ELISA (shown in mlFNγ, 21 kDa Figure 9 for mIFNa). Although levels of IFN made by primary hepatocytes in response to HCV were lower than were seen using HepG2 cells, similar patterns of responsiveness to HCV β-actin were evident. That is, liver cells that had been trans- duced with SV[wtHIVLTR](mIFNa) and transfected with Figure 6 Production and secretion of mIFNg, induced by HCV in HCV produced and secreted more mIFNa than did SV[HIVLTR](mIFNg)-transduced HepG2 cells. HepG2 cells were trans- mock-treated cultures, cultures transduced similarly but duced with the SV[HIVLTR]IFN constructs shown, then transfected with not transfected with HIV, or cultures transduced with HCV 4 days later. Total proteins from cell lysate or from medium were SV[muHIVLTR](mIFNa). The relative magnitudes of electrophoresed, immunoblotted, then probed with monoclonal antibodies against mouse IFNg as described in Materials and methods. Following these differences are comparable to those seen in the visualization of mouse IFNg, blots of cell lysates were stripped and experiments described above. Differences between reprobed with antibody to b-actin, as a loading control. These results are groups receiving SV[muHIVLTR](mIFNa) þ HCV and representative of two independent immunoblotting experiments carried out all other groups were highly significant: at the least, with these reagents, and of similar studies carried out with rSV40 vectors Po0.05 for all such comparisons. carrying cDNAs for the other IFNs studied. The efficiency of transfection of these primary hepa- tocytes was about 5%, as observed using pCMV-eGFP plasmid (Figure 10a). Furthermore, the morphology of determine the efficiency of induction of IFN expression these cells was not appreciably altered by transduction þ in these cultures, HepG2 cells were transduced with transfection (Figure 10b). SV[HIVLTR](IFN) vectors, then transfected with HCV. Control cultures were transduced, but not transfected, or were treated with a control rSV40 vector. Cells were then Secreted, HIVLTR-driven IFNa activates signaling in a immunostained for the IFN whose cDNA was delivered paracrine fashion in untransduced cells to the cells. Approximately 10% of cultured cells An important part of the antiviral activity of IFNs is their demonstrably expressed the IFN proteins delivered ability to activate cellular signaling pathways, which (Figure 7a, b), that is transfection with HCV cDNA alter transcription of a number of genes. At the same activated transduced IFN expression in approximately time, since all hepatocytes in vivo would probably not be 10% of HepG2 cells. transduced with the IFN-carrying vector, it was impor- Owing to the high transduction efficiency of rSV40 s, tant to establish whether IFN produced by cells that were we attempted to correlate this percentage with the both transduced with the SV[HIVLTR](IFN) vector and percentage of cells effectively transfected with HCV infected with HCV would secrete enough IFN to exert cDNA. For this, we used in situ RT-PCR to detect HCV paracrine – possibly protective – effects on neighboring

Gene Therapy HCV-activated conditional expression vectors AA Matskevich and DS Strayer 1866 a hIFNα

Mock SV(BUGT) SV[muHIV- SV[muHIV- SV[wtHIV- SV[wtHIV- +HCV + HCV LTR](IFN) LTR](IFN) LTR](IFN) LTR](IFN) b + HCV + HCV mIFNγ

c Mock+RNase Mock+DNase HCV+RNase HCV+DNase

Figure 7 Assessment of transduction efficiency for hIFNa and mIFNg expression induced by HCV. (a) and (b) Immunochemical analysis. HepG2 cells were transduced with SV[wtHIVLTR]IFN or SV(BUGT), then transfected with HCV and 48 h later stained with mouse anti-hIFNa, followed by ALEXA- conjugated rabbit anti-mouse IgG (a), or with mouse anti-mouse IFNg, followed by ALEXA-conjugated rabbit anti-mouse IgG (b). Results are representative of two independent studies performed on two different sets of independently transduced and transfected cells. (c) Analysis by in situ RT-PCR. HepG2 cells were transfected with HCV. Direct in situ RT-PCR was performed 48 h post-transfection as described in Materials and methods. Results are representative of two independent in situ RT-PCR studies.

cells that might not have been transduced with the vector. That is, is there a possibility that SV[HIVL- TR](IFN)-transduced, HCV-infected cells could protect neighboring, untransduced cells from HCV? Two differ- ent types of studies were used to test this possibility. In one series of studies, primary hepatocytes or HepG2 cells were transduced with SV[HIVLTR](mIFNa), then chal- lenged with transfected HCV DNA. In parallel, similar cells were transfected with pISRE-Luc, which carries the IFN-signal response element, driving the reporter gene luciferase. On the fourth day after HCV challenge, the pISRE-Luc-transfected cells were added to the cultures and the luciferase activity that was stimulated by the IFNa made by the transduced cells was measured. In these studies (Figure 11a), luciferase activity in response to IFNa made by cells that were transduced and Figure 8 Assessment of HCV transfection efficiency in transduced cells. challenged was substantially higher than in mock- HepG2 cells were transduced with one of the several vectors indicated, then transduced cultures. This increase was consistent and 2 days later were transfected with pCMVluc. Luciferase activity in cell statistically significant (minimum significance for all lysates was measured thereafter, as described in Materials and methods, and is expressed as a percentage of the luciferase activity seen in cells that such comparisons Po0.05). were mock-transduced and pCMVluc-transfected, 7s.e.m. from two An additional series of studies tested culture super- independent experiments. All groups were statistically indistinguishable natants from transduced, HCV-challenged primary he- (P40.1). patocytes or HepG2 cells. These supernatants were

Gene Therapy HCV-activated conditional expression vectors AA Matskevich and DS Strayer 1867 a b a b 20 25 + HCV + HCV 18 16 20 14 12 15 10

/mg protein, ± S.E.M 8 10 α 6 4

/ml culture supernatatants, ± S.E.M 5 α pg mlFN 2 0 0 pg mlFN Mock Mock SV(BUGT) SV(BUGT) Mock + HCV Mock + HCV SV(BUGT) + HCV SV(BUGT) + HCV SV[wtHIVLTR](IFN) SV[wtHIVLTR](IFN) SV[muHIVLTR](IFN) SV[muHIVLTR](IFN) V[muHIVLTR](IFN) + HCV V[muHIVLTR](IFN) + HCV Figure 10 Transfection efficiency and morphology of PRH. (a) PRH cells SV[wtHIVLTR](IFN) + HCV SV[wtHIVLTR](IFN) + HCV S S were transfected with pCMV-eGFP. Direct visualization was performed Figure 9 Effect of HCV on mIFNa production and secretion in PRH 48 h post-transfection. PRH before (upper picture) and after (lower transduced with SV[HIVLTR]mIFNa. Normal PRH cultures were treated picture) transduction with SV[wtHIVLTR]IFN and subsequent transfec- with SV[HIVLTR](mIFNa), then transfected with HCV cDNA 4 days tion with HCV cDNA. (b) Morphology of PRH cells before (above) and later. At 2 days after transfection, cell lysates (a) and culture supernatants after (below) transduction with SV[wtHIVLTR](hIFNa). (b) were assayed for mIFNa protein, expressed as pg IFN/mg protein (a) or as pg IFN/ml culture supernatant (b). Data represent means of independent experiments repeated a minimum of three times. Standard error bars are shown. IFNa production (a) and secretion (b) for cells potential approach to HCV therapy using gene delivery receiving SV[HIVLTR](IFN) þ HCV were significantly different form all to provide expression of IFNa and IFNg, in order to other groups: at the minimum, Po0.02 (a) and Po0.05 (b) for all such protect hepatocytes from HCV infection and replication. comparisons. IFNa and IFNg exert antiviral activities via several different mechanisms. One of their principal effects is to added to separate cultures of pISRE-Luc-transfected cells inhibit protein synthesis. Thus, IFNa is produced in of the same type. Again, similar patterns of paracrine response to virus entry into the cell, possibly as a result activity were seen: supernatants from cells that were of the activation of protein kinase R (PKR) by double- SV[HIVLTR](mIFNa)-transduced and HCV-challenged stranded viral RNAs (dsRNAs).11–13 The expression of induced much more luciferase activity than did super- one IFN, induced in this fashion, may upregulate natants from mock-transduced cells and consistently expression of other IFNs via receptor-mediated signaling 14,15 0 more luciferase activity than supernatants from cells that pathways. Viral dsRNAs also activate 5 -oligoadeny- were transduced but not challenged (Figure 11b) late synthetases (OAS), which in turn activate an enzyme 16 (Po0.05). that degrades 18S and 28S ribosomes, RNase L. Activated PKR phosphorylates and so inactivates eucar- SV[HIVLTR[(IFN)-delivered IFNa production yotic translation initiation factor 2 (eIF2a).17,18 Both decreases HCV transcript activation of RNase L and inactivation of eIF2a inhibit HCV does not easily replicate in cultured cells, so virus replication by decreasing production of viral (and analysis of potential effectiveness of SV[HIVLTR](IFN)- cellular) proteins. delivered IFN production was approached using in situ It has been reported that one or more signal transduc- RT-PCR analysis to visualize cells in which the trans- tion pathways, beginning with viral dsRNA, activate a b fected HCV had produced transcripts. Thus, transcription of IFN and IFN , which in turn signal their abilities to inhibit virus infection and replication via HepG2 cells were transduced with SV[wtHIVLTR](hIF- 19–22 Na), or mock-transduced. They were then challenged a cognate receptor. This receptor, on binding these type I IFNs, activates the Jak-STAT signaling cascade, to with HCV by transfection of HCV cDNA, or mock- 23–25 transfected. In situ RT-PCR was performed to visualize upregulate transcription of an array of cellular genes. cells producing HCV mRNA. We found by in situ RT- These IFNs have important effects on the immune PCR that numbers of cells positive for HCV transcript in system: they alter expression, activation, and localization cultures transduced with SV[wtHIVLTR](hIFNa) and of many cellular proteins, including MHC-I and -II, and a then challenged with HCV were much less than in TNF , that activate and execute antiviral T and NK cell 26 a cultures that had been mock-transduced and then responses. IFN also stimulates Th1 responses to such transfected with HCV DNA (Figure 12). Th1-tropic cytokines such as IL-12, IL-15, and IL-18, and mediates T-cell responses to IL-12-induced secretion of IFNg.26,27 Discussion IFNg is generally made and secreted by activated T and NK lymphocytes.28–30 It mediates many immune The development of new strategies to manage chronic functions, and is important in activating cytotoxic hepatitis C remains a major goal for the treatment of responses against virus-infected cells. IFNg also pos- HCV-infected individuals. In this report, we describe a sesses antiviral activities that are mechanistically similar

Gene Therapy HCV-activated conditional expression vectors AA Matskevich and DS Strayer 1868 a viral, neoplastic, and other diseases. rhIFNg, tried in this PRH: supernatants PRH: coculture 300 250 mode, was far too toxic to be useful when administered + HCV + HCV systemically.33 rhIFNa is used effectively not only as 250 200 therapy for viral illnesses such as hepatitis C, but also for several malignancies, particularly certain types of leuke- 200 150 mia, myeloma, and some solid tumors that are otherwise 150 resistant to drug therapies, such as malignant melanoma 34 100 and renal cell carcinoma. Although systemically ad- 100 ministered rhIFNa is less toxic than systemic rhIFNg,

50 treatment regimens using either have significant draw- 50 backs in addition to toxicity. These include the need to

relative luciferase activity (%), ± S.E.M. 0 maintain high blood concentrations of the recombinant 0 protein in order to provide adequate therapeutic levels at Mock Mock disease sites, and uneven distribution of the proteins outside of the circulatory system.20 Mock + HCV Mock + HCV Gene delivery could conceivably address these limita-

SV(BUGT) + HCV tions, including the toxic side effects (especially of IFNg) SV(BUGT) + HCV SV[wtHIVLTR](IFN) SV[wtHIVLTR](IFN) SV[muHIVLTR](IFN) SV[muHIVLTR](IFN) by allowing localized gene delivery to the liver, and by making IFN expression responsive to the presence of the SV[wtHIVLTR](IFN) + HCV SV[wtHIVLTR](IFN) + HCV SV[muHIVLTR](IFN) + HCV SV[muHIVLTR](IFN) + HCV offending virus. Thus we sought to provide the potential b for permanent, HCV-responsive IFN expression at the 200 site where the IFN is needed, without having to maintain 500 HepG2 cells: supernatants HepG2 cells: coculture + HCV + HCV comparable levels throughout the body. Gene delivery would be accomplished using rSV40 vectors, which are 400 150 very efficient in permanently transducing hepatocytes in vitro and the liver in vivo. 300 We also report the use of HIV-1 LTR as an HCV- 100 responsive promoter, to drive expression of IFNs.6,35 The 200 potential utility of HIVLTR as a promoter that could be activated by HCV was first tested using plasmid 50 100 cotransfection with HCV cDNA þ reporter plasmids incorporating HIVLTR driving either luciferase or

zrelative luciferase activity (%), ± S.E.M. chloramphenicol acetyl transferase. These studies (not 0 0 shown) suggested that transducing with SV[HIVL- TR](IFN), followed by transient transfection with HCV Mock Mock cDNA, might demonstrate effective gene delivery and Mock + HCV Mock + HCV IFN induction by HCV. Studies reported here repeatedly demonstrated that rSV40-delivered, HIVLTR-driven SV(BUGT) + HCV SV(BUGT) + HCV SV[wtHIVLTR](IFN) SV[wtHIVLTR](IFN) SV[muHIVLTR](IFN) SV[muHIVLTR](IFN) IFNs (IFNa and IFNg) are produced and secreted in response to the presence of HCV. SV[wtHIVLTR](IFN) + HCV SV[wtHIVLTR](IFN) + HCV SV[muHIVLTR](IFN) + HCV SV[muHIVLTR](IFN) + HCV The incremental production and secretion of IFNs, as Figure 11 Paracrine IFNa-induced biological responses in nontransduced shown by ELISA, was generally 2- to 5-fold, compared to cells. PRH (a) and HepG2 cells (b) were transduced with 100 MOI of mock-transduced cells, or to cells that were transduced SV[wtHIVLTR]mIFNa, SV(BUGT), SV[muHIVLTR](mIFNa), or mock- with SV[HIVLTR](IFN) but not transfected with HCV. transduced. After 4 days, the cultures were transfected with HCV. In Relative levels of IFN were comparable in all ELISAs, parallel, the same cell types were transfected with pISRE-Luc. (left panels) whether the transgenes were human or murine, and Supernatants from the transduced, HCV-transfected cell cultures were a g added to separate cultures of pISRE-Luc-transfected cells. Luciferase whether they were IFN or IFN . The levels of IFN (eg activity induced by IFNRE was measured on day 6 in a standard IFNa) secreted in these studies are comparable to luminometric assay. (right panels) In coculture studies, pISRE-Luc- therapeutic levels for chronic HCV infection. The value transfected cells were added to cultures of transduced, HCV-transfected of the ELISAs for these studies was their adaptation to cells. Luciferase activities were measured as indicated. Data represent testing the large numbers of samples generated, and means of independent experiments repeated a minimum of three times. their reproducibility and comparability to externally Standard error bars are shown. Differences between SV[wtHIVL- TR](IFN) þ HCV and all other groups were significant at least at supplied quantitation standards. However, they carried Po0.05 (a) and Po0.02 (b). the potential for detecting immunologically cross-reac- tive proteins besides the IFNs being tested. This possibility is underscored by the Western blotting to those described for IFNa and IFNb. IFNg receptor, like studies, which demonstrated some cross-reactivity with the receptor for type I IFNs, signals via the Jak-STAT other proteins. Our Northern analyses, which were pathway to activate transcription of an array of performed under conditions of very high stringency, genes.31,32 Also like type I IFNs, IFNg activation of showed that virtually the only detectable IFN expression transcription may occur via transcription factor binding occurred in cells that had been both transduced with to ISRE.19 SV[wtHIVLTR](IFN) and challenged by transfection with Owing to their potent effects on the immune system HCV cDNA. and on virus , as well as inhibitory effects on HIVLTR-driven is controlled by cell proliferation, IFNs have been tested as therapy for numerous cell regulatory proteins that interact with cis-

Gene Therapy HCV-activated conditional expression vectors AA Matskevich and DS Strayer 1869

Mock SV[muHIV- SV[wtHIV- rhlFNα +HCV LTR](IFNα) + HCV LTR](IFNα) + HCV + HCV Figure 12 SV[HIVLTR](hIFNa) decreases detectable HCV transcripts. HepG2 cells were transduced with SV(BUGT), or with SV[wtHIVLTR]hIFNa at MOI¼100, and 4 days later were transfected with HCV cDNA. One group was transfected with HCV cDNA and 24 h later treated with exogenous hIFNa at 50 pg/ml. Cells expressing HCV transcripts were visualized by in situ RT-PCR, incorporating fluorescent nucleotides into the HCV RT-PCR product, as described in Materials and methods. These data are representative of two independent experiments. acting sequences located in the HIV-1LTR. Among the increases in its HCV-independent transcriptional multiple regulatory elements of the HIV-1 LTR, the kB activation. enhancer, which contains two copies of kB elements at Although NFkB has been shown to be activated by a nucleotides 104–81, is considered the main inducible cis- large number of stimuli, our data suggest relatively acting element,36 and therefore our study was mainly specific responses to HCV. Transcriptional activation of focused on NFkB-mediated activation. Deletion or site- the HIV-1 LTR via NFkB has also been described for directed mutations of this regulatory sequence may other viruses, such as type-1, affect LTR transcriptional activation induced by T-cell cytomegalovirus, human T-cell leukemia virus, and activation stimuli and by HIV Tat protein.36–38 It has been human herpesvirus 8.44–47 Therefore, in HIV carriers, widely demonstrated that the kB enhancer element binds disease progression may be directly influenced by and responds to NFkB/Rel family of transcription coinfection with multiple viruses. Results presented here factors, which are induced by a number of stimuli such thus suggest that there may be complex interactions as mitogens, cytokines, and specific T-cell activators.39 between HCV and HIV-1 replicative cycles, and that The core promoter region of the HIV-1 LTR also contains HCV may be a cofactor for HIV disease progression. three tandem Sp1-binding sites located upstream of the Thus, the identification of HIVLTR as an HCV-responsive TATA box. These Sp1-binding sites are necessary for promoter is a potentially important advance in consider- basal and Tat-induced transcriptional activity.40 Other ing gene therapy of hepatitis C, but may also offer insight transcription factors such as AP-1, or ATF-1 can activate into the effects of the frequently observed coinfection of HIV replication deleted in these Sp1 sites.41,42 AIDS patients with HCV. Members of the NF-AT family of transcription factors It should be noted in this context that SV[HIVL- also bind an overlapping but distinct sequence at the kB TR](IFN) strongly inhibits HIV.9 At the same time, since enhancer: NF-AT2 (also called NF-ATc) cooperates with not all hepatocytes would be transduced by the rSV40 NFkB and Tat in HIV-1 LTR transcriptional activation.43 vector, it was important to understand whether IFNs These data may also help to explain why wtHIVLTR and produced by cells that were both transduced with rSV40 muHIVLTR could be active in the absence of the direct and infected with HCV were likely to protect cells that NFkB activator (HCV) in some situations. were not transduced, should the latter come into contact The responsiveness of IFN to HCV was expected to with HCV. Thus, we tested IFNs made by transduced reflect the ability of HCV NS5a to activate NFkB,7 which and transfected cells for their ability to elicit detectable in turn would upregulate HIVLTR as a promoter.8 We signaling via IFN receptors in naive neighboring cells. did not directly address the mechanism of HCV Levels of ISRE activation by IFNs produced by SV[HIVL- activation of NFkB, but we found that mutating NFkB TR](IFN)-transduced, HCV-transfected cells were con- binding sites in the HIVLTR eliminated HCV respon- sistently 25–50% greater than levels seen in control cells. siveness in IFN production and secretion. Detectable IFN Since ISRE are activated by a variety of cellular made by cells treated with SV[muHIVLTR](IFN) þ HCV transcription factors, many of which are not specific for was comparable to levels seen in negative control cells IFN-signaled responses, it is not surprising that back- and supernatants. Thus, these studies support the ground ISRE activities seen in these studies were high. responsiveness of HIVLTR promoter activity to HCV, Thus, although such studies do not establish that naive via NFkB. cells would be protected, they suggest that the rSV40- Although the transcriptional activity of the HIVLTR delivered IFN response to HCV has paracrine effects on was strongly, consistently and significantly activated by neighboring cells, via ISRE. HCV in all studies, some leakiness of the wt HIVLTR was It is difficult to study HCV challenge in tissue culture observed. This may reflect HCV activation of other (or, even, in experimental animals). For this reason, we transcription factors, as indicated above. Thus, addi- delivered HCV to HepG2 human hepatoblastoma cells tional optimization of the HIVLTR as a promoter is and to PRH by transfecting HCV cDNA. The effect we desirable, and is underway, to avoid such low level sought was production of HCV proteins, in particular

Gene Therapy HCV-activated conditional expression vectors AA Matskevich and DS Strayer 1870 NS5a. As attempts to demonstrate NS5a or other HCV Materials and methods proteins by immunostaining or Western blotting using commercial antibodies gave equivocal results (data not Cell lines shown), we tested HCV transcription by in situ RT-PCR. The human hepatoblastoma cell line, HepG2, was Cells adherent to coated slides were used to visualize pathogen and PPLO-free, and was maintained in those cells in which HCV cDNA transfection elicited an Dulbecco’s modified Eagle’s medium (DMEM) þ 2mM HCV RNA. We were able determine that between 5 and L-glutamine, 2 mM nonessential amino acids, 1 mM 10% of cells in transfected cultures expressed HCV sodium pyruvate, streptomycin (100 mg/ml), penicillin transcript detectably. In cultures transduced with (200 U/ml), and 10%(v/v) bovine calf serum (NCS, SV[HIVLTR](IFN), and then transfected with HCV Hyclone). cDNA, the percentage of cells expressing the IFN protein Freshly isolated PRH were grown in William’s as detected by immunostaining was comparable (E5– medium E, supplemented with 10% (v/v) bovine calf 10%). Thus, the transduction efficiency approached serum, streptomycin (100 mg/ml), penicillin (200 U/ml), 100%, and the limiting factor in determining how many gentamicin (10 mg/ml), 0.015 U insulin/ml, and 2 mM cells produced IFN was the transfection efficiency of dexamethasone. HCV cDNA. All these cells were cultivated on rat type-1 collagen- The amounts of IFN produced and secreted should be treated plates. judged in this context: HepG2 cells secreted about 50 pg COS-7 cells, used to package rSV40 vectors, were IFN/day and primary hepatocytes about 20 pg IFN/day. maintained in DMEM as described above. Lower production of IFN in primary cells may reflect differences in levels, or accessibility to activation, of Plasmids NFkB in transformed versus primary cells in response to A cDNA form of HCV, strain 1b, complete virus genome HCV. Lower transfection efficiency for PRH than for as pRC/CMV-HCV, was the kind gift of Dr Mark HepG2 cells may also reflect the fact that about 20% of Feitelson (Department of Pathology, TJU) (Figure 1). hepatocytes are naturally binucleate, which might affect rSV40 vectors were made by cloning the transgene of transduction by rSV40 vectors.48 The levels of IFN interest (7specific promoter) into a plasmid-carried measured reflected IFN secretion by only 5–10% of cells modified SV40 genome.52,53 The production of in the culture – those cells that were also transfected with SV[wtHIVLTR](IFNa) has been reported in the context HCV. Had delivery of HCV been more effective, much of inhibition of HIV replication in lymphocytes.9 The more IFN secretion would probably have been observed. other SV[HIVLTR](IFN) constructs were made similarly. Inhibition of HCV transcription was demonstrated by Thus, the IFN cDNAs and the wt or mutant (doubly the semiquantitative approach of comparing the num- mutated at its NFkB binding sites, so as to destroy NFkB bers of cells making detectable HCV transcript as binding (muHIVLTR)) HIVLTRs were cloned into visualized by in situ RT-PCR. These studies showed that pT7(DD) mpa. This plasmid carries an SV40 genome in the IFNa delivered by SV[HIVLTR](mIFNa), and elicited which the Tag gene was replaced by a polylinker by HCV, decreased detectable HCV transcript in PRH. upstream of the SV40 polyadenylation site. Transcription Although this is not a precise simulation of HCV from the SV40 early promoter (which overlaps the ori, infection in people, it provides a reasonable assessment and so cannot be deleted) is blocked by multiple tandem of the efficacy of the IFN expressed in those hepatocytes polyadenylation signals. All structures were verified by in inhibiting HCV, since HCV does not replicate well in automated DNA sequencing (PE Applied Biosystems, cultured cells. Inc., Kimmel Cancer Center, TJU). Recombinant gene transfer vectors derived from SV40 virus (rSV40) are not subject to many of the problems rSV40 vectors that have limited gene delivery using other vector Construction of recombinant SV40 derivative viruses systems. rSV40s are made at a very high titers and infect (rSV40) for gene transfer has been described previously.54 – and so transduce – almost all nucleated cell types very Briefly, rSV40 genomes were excised from the modified efficiently, regardless of lineage or whether they are pT7blue (Novagen) carrier plasmid, gel purified, recir- resting or dividing; they integrate and are not susceptible cularized, and transfected into COS-7 cells. These cells to transgene silencing.35 They also elicit no detectable supply all packaging functions in trans. Replication- immune response by normal animals and so can be used incompetent vectors are isolated from COS-7 cell lysates, to deliver multiple transgenes over time and in purified by ultracentrifugation and titered by in situ PCR, sequence.49–51 as described. Typical infectious titers for vectors pre- The ability of rSV40 vectors to deliver long-term pared in this manner are between 1012 and 2 Â 1013 transgene expression to mostly quiescent liver cells has infectious units (IU)/ml.41 The rSV40 vectors used for been illustrated in murine models, in vivo: immuno- these studies were named according to the promoter chemical staining carried out 7 weeks after administra- (wt or mutant HIVLTR, the IFN type and its species tion of SV(BUGT) showed that X60% of hepatocytes of origin). Thus, SV[wtHIVLTR](hIFNa) carries wt continued to express the BUGT transgene.6 Expression HIVLTR þ human IFNa, the vector carrying mutant continued undiminished for X18 months.51 HIVLTR þ mouse IFNa is SV[muHIVLTR](mIFNg), etc. IFNa has proven to be therapeutically efficacious in Generically, these vectors are referred to as SV[HIVL- many cases of HCV infection. Our data suggest that TR]IFN (Figure 1). Control viruses for these studies have SV[HIVLTR](IFNa), perhaps in combination with been reported: SV(HBS) and SV(BUGT), respectively SV[HIVLTR](IFNg), may deliver IFN expression that carrying cDNAs for virus surface responds to HCV and so may be useful in the treatment (HBsAg) and human bilirubin-UDP-glucuronysyl-trans- of HCV infection. ferase.

Gene Therapy HCV-activated conditional expression vectors AA Matskevich and DS Strayer 1871 Detection of HCV RNA by nested RT-PCR box) plus five repeats of ISRE cis-enhancer element Primers from the NS5A coding region of the HCV (Stratagene, La Jolla, CA, USA), was purchased. genome were selected for RT-PCR55:50-TGGATG All cloned plasmids were purified using the Qiaquick GAGTGCGGTTGCACAGGTA as the outer sense strand plasmid kit (Qiagen, Hilden, Germany). Nucleotide primer; 50-TCTTTCTCCGTGGAGGTGGTATTGC as the sequencing of constructed plasmids was confirmed by outer antisense strand primer; 50-CAGGTACGCTCC using an autosequencer (PE Applied Biosystems) and the GGCGTGCA as the inner sense strand primer; standard dye termination. 50-GGGGCCTTGGTAGGTGGCAA as the inner antisense primer. The expected PCR products are 617 bp for the Transduction and transfection experiments outer set and 571 bp for the inner set. For transduction with SV40-derived virus, HepG2 or The specificity of the transfection with HCV was PRH were treated once at an MOI of 100, as described. confirmed by nested PCR. The PCR was performed on After 5 days, cells were transfected with 7.5 mg HCV- the extracted RNA using the Rneasy Mini Kit (Qiagen, cDNA-containing plasmid (Lipofectamine 2000, Invitro- MA, USA) utilizing outer set of primers for the first gen), according to the manufacturer’s instructions, and round. The reaction mixture contained RT buffer then were cultured for 2–3 days. (Invitrogen single step RT PCR kit) (50 mM Tris-HCl, Culture medium and cells were collected 48 h post- 75 mM KCl, 1.5 mM MgCl2) containing 10 mM dithio- transfection. Cells were lysed on ice in 300 ml of lysis threitol, 200 U of Superscript II buffer (50 mM TRIS, pH 7.4/150 mM NaCl/1% Nonidet (Invitrogen), 40 U of RNase inhibitor (RNAsin; Pharma- P-40/1% sodium deoxycholate/1 Â protease inhibitor cia), 300 mM each dGTP, dATP, dCTP, 5 U of platinum Taq mix (Sigma)/0.5 mM sodium orthovanadate/0.02% so- polymerase (Invitrogene), and 0.2 mM of each primer. The dium azide) for 5 min. Cell extracts was analyzed for extracted RNA (5 ml) was added to the mix. As a control, luciferase activity and IFN expression. RNA was treated with 40 U of DNase (Promega) or 10 U of RNase One (Promega). The PCR was performed over Luciferase activity 1 25 cycles, the first one consisting of 30 min at 58 C, Luciferase activity was measured using the Steady-Glo 1 1 followed by 25 cycles of 1 min at 94 C, 1 min at 60 C, and Luciferase Assay System (Promega), according to the 1 1 1 min at 72 C, and finally 7 min at 72 C. The first round manufacturer’s instructions. of product (5 ml) was added to 45 ml of PCR mix containing 1 Â buffer, 300 mM each dNTP, 2.5 U of Determination of IFN by ELISA platinum Taq polymerase, and 0.2 mM of inner set of primers. Thermocycling was performed over 25 cycles, Expression of all the IFNs was measured using their each consisting of 1 min at 941C, 1 min at 601C, and 1 min respective IFN ELISA kits (Antigenix America). This kit at 721C. detects IFN using a sandwich immunoassay. In this assay, plates are coated with monoclonal antibodies against the particular IFN. Binding to that antibody was Detection of HCV RNA by in situ RT-PCR detected using an anti-secondary antibody conjugated to HepG2 cells were spread on chamber, fixed in sterile horseradish peroxidase (HRP). Tetramethyl-benzidine saline, resuspended in ice-cold 10% buffered formalde- was used as a substrate to quantitate the antigen– hyde solution and kept for 2 h at 371C and treated with antibody complex by subsequent color development. Proteinase K at 371C for 10 min, washed again and treated with 40 U of DNase (Invitrogen) for 4 h, at 371C. Western blot analysis As a control, cells were treated with 40 U of RNase. 106 HepG2 cells either SV[HIVLTR]IFN-transduced, Permeabilized HepG2 cells were then washed in PBS and mock-transduced or expressing HBsAg or BUGT and treated by RT buffer (Invitrogen single step RT PCR kit) subsequently transfected with HCV (if mentioned) were

(50 mM Tris-HCl, 75 mM KCl, 3 mM MgCl2) containing washed in PBS. Cells were lysed on ice in 300 ml of lysis 10 mM dithiothreitol, 200 U of Superscript II Reverse buffer (50 mM TRIS, pH7.4/150 mM NaCl/1% nonidet P- transcriptase (Invitrogene), 40 U of RNase inhibitor 40/1% sodium deoxycholate/1 Â protease inhibitor mix (RNAsin; Pharmacia), 1 mM each dGTP, dATP, dCTP (Sigma)/0.5 mM sodium orthovanadate/0.02% sodium and 1 mM of dUTP-FITC, platinum Taq polymerase and azide) for 5 min. Cell culture medium was concentrated 50 pmol of the outer oligonucleotide primers of HCV. two times using Centricon-100 concentrators (Amicon, The tubes were incubated at 581C for 45 min and at 941C Bedford, MA, USA). for 5 min to inactivate residual RT activity. The RT step Equal amounts of protein were separated by 4–20% was followed by first round PCR: 35 cycles of 1 min each gradient Tris-glycine SDS/PAGE (Bio-Rad) and trans- at 94, 58 and 721C, with the final extension done at 721C ferred to a PVDF-Plus membrane (Osmonics, Minneton- for 10 min. At the end of RT-PCR, slides were incubated ka, MN, USA). After overnight blocking (PBS (pH 7.4)/ in PBS protected from light until the coverslips were 0.1% Tween-20 containing 5% nondry milk), blots were removed. Slides were washed in 0.1 Â SSC at 371C for incubated with monoclonal anti-IFN antibody (Antigenix 15 min, then analyzed under epifluorescence microscope America) (1:400) in PBS (pH 7.4)/0.1% Tween-20 contain- for FITC emission. ing 5% nondry milk) for 1 h at room temperature. Goat anti-mouse IgG conjugated to HRP (Pierce) (1:10 000) was added, and blots were incubated for 1 h at room IFN pathway reporter plasmid temperature. Each antibody incubation was followed by To elucidate the activation of IFN-inducible Elements three extensive washes in PBS (pH 7.4)/0.1% Tween-20. (ISRE) reporter plasmid ISRE-Luc, containing luciferase Detection was by chemiluminescence (ECL Plus, Amer- reporter gene driven by a basic promoter element (TATA sham) according to the manufacturer’s instruction.

Gene Therapy HCV-activated conditional expression vectors AA Matskevich and DS Strayer 1872 Northern blot analysis vectors were made. The current work was supported by Total RNA from 106 HepG2 cells either SV[HIVLTR]IFN- NIH Grants AI41399, AI48244, and RR13156. transduced, mock- or SV-BUGT-transduced and subse- quently transfected with HCV (if mentioned) was extracted using the RNA easy Mini Kit (Qiagen, CA, USA). Samples of 15 mg of total RNA were electrophor- References esed in a 1% agarose/formaldehyde gel, transferred on a nylon filter (Nytran Super Charge, Schleicher and 1 EASL International Consensus Conference on hepatitis C. Paris Schuell), UV cross-linked with a Stratalinker oven February 26–28, 1999; consensus statement. J Hepatol 1999; 30: (Stratagene) and baked for 2 h at 801C in a vacuum 956–961. oven. Filters were prehybridized in 50% formamide, 2 McHutchinson JG et al. Interferon alfa-2b alone or in 5  SSPE, 20 mg/ml denatured salmon sperm DNA, 5  combination with ribavirin as initial treatment for chronic hepatitis C. N Engl J Med 1998; 339: 1485–1492. Denhardt’s solution, 0.1% SDS at 421C for 8 h, then 3 Lauer GM, Walker BD. Medical progress: hepatitis C virus hybridized under the same conditions with an IFN 32 infection. N Engl J Med 2001; 345: 41–52. cDNA probe that had been labeled with a P-dCTP using 4 Gale Jr M, Katze MG. Molecular mechanisms of interferon a random priming labeling kit (Gibco BRL). Hybridiza- resistance mediated by viral-directed inhibition of PKR, the tion was performed at 421Cin2 SSC overnight. After interferon-induced protein kinase. Pharmocol Ther 1998; 78: hybridization, filters were washed under high stringency 29–46. conditions in 0.1  SSC, 0.1% SDS at 37–551C for 30 min 5 Zern M et al. A novel SV40-based vector successfully transduces and signals were visualized using a phosphoimager and expresses an a1-antitrypsin ribozyme in a human hepatoma- (Molecular Dynamics Storm 840). To assess loading of derived cell line. Gene Therapy 1999; 6: 114–120. various lanes, Northern blots were stripped and re- 6 Sauter BV et al. Gene transfer to the liver using a replication- probed with a radiolabeled cDNA for human b-actin. deficient recombinant SV40 vector results in long-term amelioration of jaundice in Gunn rats. Gastroenterology 2000; 119 Immunostaining : 1348–1357. 5 7 Gong G, Waris G, Tanveer R, Siddiqui A. Human hepatitis C HepG2 cells (2  10 ) HepG2 cells were grown on 12-well virus NS5A protein alters intracellular calcium levels, induces plates coated with collagen. At 5 days after transduction oxidative stress, and activates STAT-3 and NFkB. Proc Natl Acad with either SV[HIVLTR]IFN or SV(BUGT) and subse- Sci 2001; 98: 9599–9604. quently transfection with HCV (if mentioned), cells were 8 Li Y, Mak G, Franza Jr BR. In vitro study of functional fixed with 2% paraformaldehyde for 1 h at room involvement of Sp1, NFkB/Rel, and AP1 in phorbol 12- temperature, permeabilized with PBS/0.1%NP-40 for myristate 13-acetate-mediated HIV-1 long terminal repeat 10 min, blocked overnight at 41C with the normal serum, activation. J Biol Chem 1994; 269: 30616–30619. corresponding to the animal source of the secondary 9 Cordelier P, Calarota S, Strayer DS. Trans-activated IFNa2 antibody, and immunostained with anti-IFN antibodies delivered to T cells by SV40 inhibits early stages in the HIV-1 for 1 h at 371C. After extensive washes in PBS, replicative cycle. J Hematother Stem Cell Res 2002; 11: 817–828. corresponding FITC-conjugated secondary antibodies 10 Lindenbach BD, Rice CM. Flaviviridae: the viruses and their (donkey-anti-rabbit IgG; Molecular Probes) were used replication. In: Knipe DM, Howley PM (eds). Fields’ Virology. for 1 h at 371C. Slides were extensively washed, Lippincott, Williams & Wilkins: Philadelphia, 2001. mounted, and photographed at  400 magnification. 11 Iordanov MS, Wong J, Bell JC, Magun BE. Activation of NFkBby double-stranded RNA (dsRNA) in the absence of protein kinase R and RNase L demonstrates the existence of two separate Statistical analysis dsRNA-triggered antiviral programs. Mol Cell Biol 2001; 21: IFN levels and luciferase activity in HepG2 and PRH 61–72. cells from various treatment groups and controls at 12 Chu WM et al. JNK2 and IKkb are required for activating the various time points were compared by the Student’s t- innate response to viral infection. Immunity 1999; 11: 721–731. test or analysis of variance. Levels of statistical signifi- 13 Zamanian-Daryoush M, Mogensen TH, DiDonato JA, Williams cance between the test group, on the one hand, and the BR. NFkB activation by double-stranded-RNA-activated protein several control groups, on the other, varied considerably. kinase (PKR) is mediated through NFkB-inducing kinase and In each figure, the minimum level of statistical signifi- IkB kinase. Mol Cell Biol 2000; 20: 1278–1290. cance is mentioned (usually, Po0.05), but often compar- 14 Malmgaard L, Salazar-Mather TP, Lewis CA, Biron CA. isons between test and control groups typically reached Promotion of a/b interferon induction during in vivo viral much higher levels of significance (eg Po0.001 or better). infection through a/b interferon receptor/STAT1 system- dependent and -independent pathways. J Virol 2002; 76: 4520–4525. 15 Mogensen TH, Paludan SR. Molecular pathways in virus- Acknowledgements induced cytokine production. Microbiol Mol Biol Rev 2001; 65: 131–150. We greatly appreciate the input of Drs J Roy Chowdhury, 16 Iordanov MS et al. Activation of p38 mitogen-activated protein Pierre Cordelier, Mark Feitelson, Geetha Jayan, and kinase and c-Jun NH2-terminal kinase by double-stranded RNA Aleem Siddiqui. The assistance of Dr Jan Hoek’s lab and encephalomyocarditis virus: involvement of RNase L, was helpful in establishing primary hepatocyte cultures. protein kinase R, and alternative pathways. Mol Cell Biol 2000; Dr Mark Feitelson (TJU) kindly supplied us with pRC/ 20: 617–627. CMV-HCV; Dr Henry Wu (Immune Response Corp) 17 Williams BRG. PKR; a sentinel kinase for cellular stress. provided us with the cDNA for hIFNa 2b; Dr Janet S Oncogene 1999; 18: 6112–6120. Butel (Baylor College of Medicine) gave us the plasmids 18 De Haro C, Mendez R, Santoyo J. The eIF-2a kinases and the carrying the original wtSV40 genome, from which rSV40 control of protein synthesis. FASEB J 1996; 10: 1378–1387.

Gene Therapy HCV-activated conditional expression vectors AA Matskevich and DS Strayer 1873 19 Boehm U, Groot M, Howard JC, Klamp T. Cellular responses to virus replication yet are not absolutely required for viral growth. interferon-g. Annu Rev Immunol 1997; 15: 749–794. J Virol 1997; 71: 5495–5504. 20 Corssmit EPM, De Metz J, Saurwein HP, Romijn JA. Biologic 39 Ghosh S, May MJ, Koop EB. NFkB and Rel proteins: responses to IFN-a administration in . J Interferon evolutionarily conserved mediators of immune responses. Cytokine Res 2000; 20: 1039–1047. Annu Rev Immunol 1998; 16: 225–240. 21 Domanski P et al. A region of the b subunit of the interferon a 40 Berkhout B, Jeang KT. Functional roles for the TATA promoter receptor different from box 1 interacts with Jak1 and is sufficient and enhancers in basal and Tat-induced expression of the to activate the Jak-Stat pathway and induce an antiviral state. human immunodeficiency virus type 1 long terminal repeat. J Biol Chem 1997; 272: 26388–26393. J Virol 1992; 66: 139–149. 22 Barnes BJ, Moore PA, Pitha PM. Virus-specific activation of a 41 Parrott C et al. Variable role of the long terminal repeat Sp1- novel interferon regulatory factor, IRF-5, results in the induction binding sites in human immunodeficiency virus replication in T of distinct interferon a genes. J Biol Chem 2001; 276: 23382–23390. lymphocytes. J Virol 1991; 65: 1414–1419. 23 Decker T. Double-stranded RNA and interferon-a induce 42 Jones KA. Tat and the HIV-1 promoter. Curr Opin Cell Biol 1993; transcription through different molecular mechanisms. 5: 461–468. J Interferon Res 1992; 12: 445–448. 43 Kinoshita S et al. The T cell activation factor NF-ATc positively 24 Harcourt JL, Offerman MK. Multiple signaling cascades are regulates HIV-1 replication and gene expression in T cells. differntially involved in gene induction by double stranded Immunity 1997; 6: 235–240. RNA in interferon-a-primed cells. Eur J Biochem 2001; 268: 44 Schafer SL, Vlach J, Pitha PM. Cooperation between Herpes 1373–1381. simplex virus type 1-encoded ICP0 and Tat to support 25 Melen K, Keskinen P, Lehtonen A, Julkuen I. Interferon-induced transcription of immunodeficiency virus type 1 long terminal gene expression and signaling in human hepatoma cell lines. repeat in vivo can occur in the absence of the TAT binding site. J Hepatol 2000; 33: 764–772. J Virol 1996; 70: 6937–6946. 26 Biron CA. Initial and innate responses to viral infections-pattern 45 Dal Monte P et al. TAR and Sp1-independent transactivation of setting in immunity or disease. Curr Opin Microbiol 1999; 2: HIV long terminal repeat by the Tat protein in the presence of 374–381. human cytomegalovirus IE1/IE2. AIDS 1997; 11: 297–303. 27 Sareneva T, Julkunen I, Matikainen S. IFN-a and IL-I2 induce 46 Cheng H, Tarnok J, Parks WP. Human immunodeficiency virus IL-I8 receptor gene expression in human NK and T cells. type 1 genome activation induced by human T-cell leukemia J Immunol 2000; 165: 1933–1938. virus type 1 Tax protein is through cooperation of NFkB and Tat. 28 Mosmann TR, Coffman RL. TH1 and TH2 cells: different J Virol 2001; 72: 13427–13432. patterns of lymphokine secretion lead to different functional 47 Huang L-M et al. Reciprocal regulatory interaction between properties. Annu Rev Immunol 1989; 7: 145–173. human herpes virus 8 and human immunodeficiency virus type 29 Perussia B. Lymphokine-activated killer cells, natural killer cells 1. J Biol Chem 2001; 276: 13427–13432. and cytokines. Curr Opin Immunol 1991; 3: 49–55. 48 Engellman GL, Richardson A, Katz A, Fierer JA. Age-related 30 Sad S, Marcotte R, Mosmann TR. Cytokine-induced changes in isolated rat hepatocytes. Comparison of size, differentiation of precursor mouse CD8+ T cells into cytotoxic morphology, binucleation, and protein content. Mech Ageing CD8+ T cells secreting Th1 or T2 cytokines. Immunity 1995; 2: Dev 1981; 16: 385–395. 271–279. 49 McKee HJ, Strayer DS. Immune responses against SIV envelope 31 Valente G et al. Distribution of interferon-g receptor in human glycoprotein, using recombinant SV40 as a vaccine delivery tissues. Eur J Immunol 1992; 22: 2403–2412. vector. Vaccine 2002; 20: 3613–3625. 32 Schindler C, Darnell Jr JE. Transcriptional responses to 50 Kondo R, Feitelson MA, Strayer DS. Use of SV40 to immunize polypeptide ligands: the JAK-STAT pathway. Annu Rev Biochem against hepatitis B surface antigen: implications for the use of 1995; 64: 621–651. SV40 for gene transduction and its use as an immunizing agent. 33 Mannering GJ, Deloria LB. The pharmacology and toxicology of Gene Therapy 1998; 5: 575–582. the interferons: an overview. Annu Rev Pharmacol Toxicol 1986; 26: 51 Strayer DS, Zern MA, Chowdhury JR. What can SV40-derived 455–515. vectors do for gene therapy? Curr Opin Mol Ther 2002; 4: 313–323. 34 Brassard DL, Grace MJ, Bordens RW. Interferon-a as an 52 Strayer DS. SV40-based gene transfer vectors: turning an immunotherapeutic protein. J Leukoc Biol 2002; 71: 565–581. adversary into a friend. Curr Opin Mol Ther 2000; 2: 570–578. 35 Strayer DS et al. Durability of transgene expression and vector 53 Strayer DS. Gene Therapy using SV40-derived vectors: what integration: recombinant SV40-derived gene therapy vectors. does the future hold? J Cell Physiol 1999; 181: 375–384. Mol Ther 2002; 6: 227–237. 54 Strayer DS et al. Generation of recombinant SV40 vectors 36 Jones KA, Peterlin BM. Control of RNA initiation and elongation for gene transfer. In: Raptis L (ed) SV40 Protocols, vol. 165, at the HIV-1 promoter. Annu Rev Biochem 1994; 63: 717–743. Methods in Molecular Biology Walker JM (ed). Humana Press: 37 Alcami J et al. Absolute dependence on kappa-B responsive Totowa, NJ, 2001. elements for initiation and Tat-mediated amplification of HIV 55 Watanabe H et al. Number and position of mutations in the transcription in blood CD4 T lymphocytes. EMBO J 1995; 14: interferon (IFN) sensitivity-determining region of the gene for 1552–1560. nonstructural protein 5A correlate with IFN efficacy in hepatitis 38 Chen BK, Feinberg MB, Baltimore D. The kB sites in the human C virus genotype 1b infection. J Infect Dis 2001; 183: immunodeficiency virus type 1 long terminal repeat enhance 1195–1203.

Gene Therapy