of July 6, 2015.This information is current as

Virus-Transgenic MiceB Virus Surface Antigen in Hepatitis B Overcoming T Cell Tolerance to the Hepatitis

Guidotti and Francis V. ChisariAlexander, Robert W. Chesnut, Kazuhiro Kakimi, Luca G. Alessandro D. Sette, Carla Oseroff, John Sidney, Jeff

http://www.jimmunol.org/content/166/2/1389doi: 10.4049/jimmunol.166.2.1389

2001; 166:1389-1397; ;J Immunol 

Referenceshttp://www.jimmunol.org/content/166/2/1389.full#ref-list-1

, 21 of which you can access for free at: cites 37 articlesThis article

Subscriptionshttp://jimmunol.org/subscriptions

is online at: The Journal of ImmunologyInformation about subscribing to

Permissionshttp://www.aai.org/ji/copyright.htmlSubmit copyright permission requests at:

Email Alertshttp://jimmunol.org/cgi/alerts/etocReceive free email-alerts when new articles cite this article. Sign up at:

Print ISSN: 0022-1767 Online ISSN: 1550-6606. Immunologists All rights reserved.Copyright © 2001 by The American Association of9650 Rockville Pike, Bethesda, MD 20814-3994.The American Association of Immunologists, Inc.,

is published twice each month byThe Journal of Immunology

by guest on July 6, 2015http://w

ww

.jimm

unol.org/D

ownloaded from

by guest on July 6, 2015

http://ww

w.jim

munol.org/

Dow

nloaded from

Overcoming T Cell Tolerance to the Hepatitis B Virus SurfaceAntigen in Hepatitis B Virus-Transgenic Mice1

Alessandro D. Sette,2* Carla Oseroff,* John Sidney,* Jeff Alexander,* Robert W. Chesnut,*Kazuhiro Kakimi, † Luca G. Guidotti, † and Francis V. Chisari†

The sequence of the hepatitis B virus (HBV) major envelope (Env) protein (ayw subtype) was scanned for the presence of H-2d,b

motifs. Following binding and immunogenicity testing, two new H-2d-restricted epitopes (Env.362 and Env.364) were identified.These epitopes induced CTLs capable of recognizing naturally processed HBV-Env, but were apparently generated with lowerefficiency than the previously defined dominant Env.28 epitope. Next, HBV-transgenic mice that express all of the HBV proteinsand produce fully infectious particles were immunized with a mixture of lipopeptides encompassing the Env.28, Env.362, andEnv.364 epitopes. Significant CTL responses were obtained, but they had no effect on viral replication in the liver, nor did theyinduce an inflammatory liver disease. However, in adoptive transfer experiments, CTL lines generated from the HBV-transgenicmice following immunization were able to inhibit viral replication in vivo without causing hepatitis. This is in contrast to CTL linesderived from nontransgenic mice that displayed both antiviral and cytopathic effects, presumably because they displayed higheravidity for the viral epitopes than the transgenic CTLs. These results suggest that T cell tolerance to HBV can be broken withappropriate immunization but the magnitude and characteristics of the resultant T cell response are significantly different fromthe response in HBV-naive individuals since their antiviral potential is stronger than their cytotoxic potential. This has obviousimplications for immunotherapy of chronic HBV infection. The Journal of Immunology,2001, 166: 1389–1397.

T he hepatitis B virus (HBV)3 is a noncytopathic DNA virusthat chronically infects;350 million people worldwide(1, 2). Adult onset infection usually results in self-limited

acute hepatitis followed by viral clearance, although up to 5% ofinfected adults become chronically infected (2). In contrast, neo-natal infection is seldom cleared and a majority of infected chil-dren develop persistent infection (2). Patients chronically infectedwith HBV are predisposed to developing cirrhosis of the liver andhepatocellular carcinoma (2).

CD81 T lymphocytes are an important component of host de-fense mechanisms responsible for HBV clearance in patients withacute hepatitis (3, 4). In particular, multispecific responses appearto be associated with spontaneous resolution of acute infection. Inthis respect, elicitation of responses directed against multipleepitopes may be of therapeutic interest (2). It has been proposedthat elimination of virus is due to CTL-mediated lysis of infectedhepatocytes and/or antiviral effects of CTL-derived cytokines,such as IFN-g and TNF-a (5, 6). Recent data demonstrate that, inacutely infected chimpanzees, HBV DNA is eliminated from theliver before the peak of T cell infiltration and the associated liver

disease (7). These effects coincide with the appearance of IFN-g inthe liver, suggesting that IFN-g-producing non-T cells, presum-ably NK and NKT cells, as well as T cells, play a key role in thecontrol of HBV infection. Thus, noncytopathic mechanisms thateliminate replicative HBV DNA intermediates from the cytoplasmand covalently closed circular DNA from the nucleus (7) appear tobe crucial for clearance of HBV from the hepatocyte. Therapeuticinduction of this kind of noncytopathic, antiviral activity in theliver of chronically infected patients would, therefore, lead to viralclearance, thereby preventing cirrhosis and hepatocellularcarcinoma.

Unfortunately, chronic HBV infection, like many other chronicviral diseases and cancers (2), is associated with T cell hypore-sponsiveness or tolerance. The exact molecular mechanisms asso-ciated with this phenomenon are not entirely clear but negativeselection, peripheral anergy, and imbalances in lymphokine pro-duction all appear to contribute to maintaining the hyporesponsivestate of the host chronically exposed to viral or cancer Ags (2).

Transgenic mice of the 1.3.32 lineage contain the completeHBV genome, express all HBV gene products, and replicate HBVin their hepatocytes similar to chronically infected patients, butthey do not develop chronic hepatitis because they are immuno-logically tolerant to the viral Ags (5, 8). The availability of thesemice provides a model system to evaluate immunotherapeuticstrategies to break tolerance and terminate persistent HBVinfection.

Thus far, several approaches to overcome CTL tolerance inHBV-transgenic mice have been investigated. In previous studies,five different lineages of HBV-transgenic mice, including lineage1.3.32, were immunized with either plasmid DNA encoding wholeHBV Ags or activated dendritic cells. DNA immunization inducedAb responses in some of the lineages, but no CTL responses (8).Dendritic cell immunization generated CTLs but they had no effecton viral replication and they did not cause hepatitis (8). Theseresults are in contrast with a previous study which had indicated

*Epimmune, Inc., San Diego, CA 92121; and†Department of Molecular and Exper-imental Medicine, The Scripps Research Institute, La Jolla, CA 92121

Received for publication June 27, 2000. Accepted for publication October 24, 2000.

The costs of publication of this article were defrayed in part by the payment of pagecharges. This article must therefore be hereby markedadvertisementin accordancewith 18 U.S.C. Section 1734 solely to indicate this fact.1 This work was supported in part with federal funds from the National Institute forAllergy and Infectious Diseases, National Institutes of Health Grants R37 CA40489(to F.V.C.) and RO1 AI40696 (to L.G.G.), Small Business Innovation Research PhaseII Grant AI-32847-03, and Contracts NO1-1A-45241 and NO1-AI-95362.2 Address correspondence and reprint requests to Dr. Alessandro Sette, Epimmune,Inc, 5820 Nancy Ridge Drive, Suite 100, San Diego, CA 92121. E-mail address:asette@epimmune.com3 Abbreviations used in this paper: HBV, hepatitis B virus; Env, envelope; HTL,helper T lymphocyte; sALT, serum alanine aminotransferase; LCMV, lymphocyticchoriomeningitis virus.

Copyright © 2001 by The American Association of Immunologists 0022-1767/01/$02.00

by guest on July 6, 2015http://w

ww

.jimm

unol.org/D

ownloaded from

that tolerance could be broken by DNA immunization in a singlelineage of transgenic mice that contain a subgenomic fragment ofHBV and express the major envelope (Env) protein within hepa-tocytes (9). In these animals, DNA immunization was associatedwith a reduction of HBV RNA expression in hepatocytes (9) andthis effect was shown to be dependent on IFN-g. The immunolog-ical, virological, and clinical relevance of those results is uncer-tain; however, the transgene in the particular lineage of mice usedin those experiments is susceptible to genomic imprinting and si-lencing of transgene expression by methylation (10). Therefore,these events may have been triggered by immunization but do notoccur during infection.

In the present study, we used the complete HBV genome trans-genic mouse model to evaluate the relative immunogenicity of apool of HBV-specific lipopeptides with optimized helper T lym-phocyte (HTL) function for their capacity to break T cell toleranceand to down-regulate HBV gene expression and replication.

Materials and MethodsPeptides and lipopeptides

Peptides were synthesized according to standard F-moc solid-phase syn-thesis methods (11). In HTL-CTL peptide constructs, the HTL epitope wasplaced at the amino terminus of the CTL epitope. In the case of the li-popeptide synthesis, each amino acid coupling was followed by a cappingcycle with 10% acetic anhydride: dimethylformamide (v/v) to facilitatesubsequent purification. Lipopeptides were prepared by coupling pre-formed symmetrical anhydride of palmitic acid to the amino terminus ofthe resin-bound KSS-elongated peptide. The peptides and the protectinggroups on the amino acids were cleaved using trifluoroacetic acid,ethanedithiol, water (9.5:2.5:2.5, v/v/v), or trifluoroacetic acid, thioanisole,phenol, ethanedithiol, water (10 ml:0.5 ml:0.75 g:0.25 ml:0.5 ml), depend-ing upon the specific protecting group present on the various amino acids.

Unlipidated peptides were purified using conventional HPLC methods.Lipopeptides were purified by repetitive washes with 50% acetic acid:water (v/v), followed by two washes with H2O and lyophilization. Li-popeptides were characterized by reverse phase-HPLC and mass spectrom-etry. Reverse phase-HPLC was performed using Waters analytical HPLC(polymer reverse phase 4.63 150 mm, 5mM, 300 Å, column; PolymerLaboratories, Shropshire, U.K.) at a column temperature of 80°C and de-tection at 214 nm. The mobile phase buffers were A ( water with 0.1%trifluoroacetic acid) and B (acetonitrile with 0.1% trifluoroacetic acid), andthe elution was effected with a gradient of 20–100% B over 30 min at aflow rate of 1 ml min21. Purity and identity of all compounds were checkedby analytical HPLC and mass spectrometry. Purity was$50% for lipidatedpeptides and$ 90% for unlipidated peptides.

Cell lines, Abs, and MHC purification

The mouse lymphoma EL-4 was used as the source of Kb and Db class Imolecules. EL-4 cells were maintained as described previously (12, 13).The mAbs used for purification were Y3 (anti-Kb) and 28-14-8S (anti-Db,Ld, and Dq; American Type Culture Collection, Manassas, VA). The mousemastocytoma cell line P815 was used as the source of Dd, Kd, and Ld classI molecules. P815 cells were maintained as described previously (12, 13).The mAbs used for purification were 34-5-8S (anti-H-2Dd), SF1.1.1.1 (anti-H-2Kd), and 28-14-8S (anti-Dd, Ld, and Dq). Cell lysates for MHC puri-fication were prepared as previously described (12, 13). Briefly, cells werelysed at a concentration of 108 cells/ml in 50 mM Tris-HCl (pH 8.5) con-taining 1% Nonidet P-40 (Fluka Biochemika, Buchs, Switzerland), 150mM NaCl, 5 mM EDTA, and 2 mM PMSF. The lysates were then passedthrough 0.45-mm filters and cleared of nuclei and debris by centrifugationat 10,0003 g for 20 min. MHC molecules were then purified by repeatedsequential passage over either 34-5-8S, SF1.1.1.1, or 28-14-8S columns.Protein purity and effectiveness of depletion steps were monitored bySDS-PAGE.

In vitro MHC peptide-binding assays

Quantitative assays for the binding of peptides to detergent-solubilized H-2MHC class I molecules, on the basis of the inhibition of binding of aradiolabeled standard probe peptide, were performed as previously de-scribed for HLA class I molecules (12–16). Briefly, 1–10 nM radiolabeledprobe peptide, iodinated by the chloramine-T method (12), was coincu-bated for 2 days at room temperature with various amounts of MHC in the

presence of 1mM humanb2-microglobulin (Scripps Laboratories, San Di-ego, CA) and a mixture of protease inhibitors. At the end of the incubationperiod, the percentage of MHC-bound radioactivity was determined by sizeexclusion gel filtration chromatography on a TSK 2000 column (TosoHaas,Montgomeryville, PA).

The concentration of peptide yielding 50% inhibition of the binding ofthe radiolabeled probe (IC50) in competitive inhibition assays was calcu-lated. Peptides were usually tested at one or two high doses, and the IC50

values of peptides yielding positive inhibition were determined in subse-quent experiments, in which two to six additional dilutions were tested, asnecessary. MHC concentrations yielding;15% binding of the radiolabeledprobe peptide were used for all competition assays. Each competitor pep-tide was tested in two to four independent experiments. The radiolabeledprobes used, and their average IC50 values in the respective assays, were asfollows: HIV-IIIB Env gpG4 . Y analogue (sequence RGPYRAFVTI),3.7 nM for Dd; 1079.03 (sequence KFNPMKTYI), 1.1 nM for Kd; andB35CON2 (sequence FPFKYAAAF), 30 nM for Ld; adenovirus E1A P7 Yanalogue (sequence SGPSNTYPEI) 4.4 nM for Db; and VSV NP52–59(sequence RGYVFQGL) 3.1 nM for Kb.

Mice

Eight- to 12-wk-old mice were used in all experiments. (C57BL/63B10D2)F1 (B6D2F1) and (C57BL/63 BALB/c)F1 (CB6F1) were pur-chased from The Jackson Laboratory (Bar Harbor, ME). HBV-transgenicmice from lineage 1.3.32 (official designation Tg(HBV 1.3 genome)x32)(inbred C57BL/6) were bred at The Scripps Research Institute. Transgenicmice from lineage 1.3.32 have been previously described (17). The hepa-tocytes from these animals express all of the HBV gene products andreplicate HBV at high levels in the liver without any evidence of cytopa-thology. Lineage 1.3.32 mice C57BL/6 (H-2b) were backcrossed one gen-eration against BALB/c (H-2d) mice to produce H-2bxd F1 (CB6F1) hybridsfor this study.

Peptide formulations and immunizations

Peptides and lipopeptides were resuspended at 20 mg ml21 and 10 mgml21, respectively, in DMSO/0.1% trifluoroacetic acid and stored as stocksolutions at220°C. Two different formulations were used for immuniza-tions. For experiments involving IFA, each peptide (or peptide mixture)stock solution was diluted after vortexing for 30 s with appropriateamounts of DMSO and PBS at room temperature to obtain the desiredpeptide concentrations in a final buffer composition of 90% PBS, 10%DMSO/0.1% trifluoroacetic acid. These samples were emulsified with anequal volume of IFA (Difco, Detroit, MI) for 20 min in a 5100 Spexmixer/mill (Spex Industries, Metuchen, NJ). In experiments involving lipopep-tides, stock solutions were heated for 10 min at 45°C (to ensure solubili-zation of the lipopeptides in DMSO) and then vortexed for 30 s beforedilution with PBS at room temperature. Eight- to 12-wk-old mice wereimmunized s.c. in the base of the tail with the formulations describedabove.

Assay of CTL activity

Eleven to 14 days after immunization, splenocytes were harvested fromindividual mice and 33 107 cells were stimulated with specific HBVpeptides at 10.0mg ml21 in the presence of 107 irradiated (4500 rad)syngeneic LPS blasts. To obtain LPS blast cells, syngeneic splenocyteswere resuspended at a concentration of 1–1.53 106 cells ml21 in completemedia in the presence of 25mg ml21 of LPS (Sigma, St. Louis, MO) and7 mg ml21 of dextran sulfate (Pharmacia Biotech, Uppsala, Sweden) andkept in culture for 72 h at 37°C. After 5–6 days, splenocytes from eachflask were collected and assayed for cytolytic activity using a standard 4-h51Cr release assay.

The following cell lines were used as target cells in the cytotoxicityassays: EL4 (a murine lymphoma), P815 (a mouse mastocytoma), andP815-S (P815 cells that stably express all the HBV Env protein (17). Allof the cell lines were grown in RPMI 1640 (Life Technologies, GrandIsland, NY) containing 10% FCS (Gemini Bioproduct, Calabasas, CA), 4nM L-glutamine (Irvine Scientific, Santa Ana, CA), 10mg ml21 of genta-micin (Irvine Scientific), and 53 1025 nM 2-ME (Sigma). Target cells(3 3 106) were labeled with 300mCi of 51Cr sodium chromate (NENResearch Products, Boston, MA) for 60 min at 37°C, washed three times,and resuspended in RPMI 1640/10% FCS at a concentration of at 105 cellsml21 in the absence or presence of 2mg ml21 of the appropriate peptides.To assay for CTL activity, 100ml of target cells was incubated with 100ml of different numbers of effectors cells in U-bottom 96-well plates. Su-pernatants (100ml) were removed after 4 h at 37°C and the percent lysis

1390 INDUCTION OF HBV-SPECIFIC CTLS IN HBV-TRANSGENIC MICE

by guest on July 6, 2015http://w

ww

.jimm

unol.org/D

ownloaded from

was determined by the following formula: percent release5 100 3 [(ex-perimental release2 spontaneous release)4 (maximum release2 spon-taneous release)].

To allow averaging of different experiments, specific CTL activity wasalso expressed in LU (LU30), in which 1 LU (LU30) corresponds to thenumber of effector cells required to induce 30% lysis of 104 51Cr-labeledtarget cells during the 4-h assay (18). Thus, in the conditions used in theassay 1 LU30 represents 30% lysis at the 100:1 E:T ratio. Ten LU30 rep-resents 30% lysis at the 10:1 ratio, 100 LU30 represents 30% lysis at the 1:1ratio, and so on.

CTL activity was also measured by IFN-g production. A total of 13105 CTLs were cultured with 105 target cells (HBV-Env-transfected or-untransfected P815 cells), with or without relevant peptide, in 96-wellImmulon-2 plates (Dynex Technologies, Chantilly, VA) that were pre-coated with a capture anti-mouse IFN-g Ab (clone R4-6A2; PharMingen,San Diego, CA). After a 20-h incubation at 37°C, wells were washed withPBS/0.5%Tween 20 and incubated sequentially with a biotinylated sec-ondary anti-mouse IFN-g Ab (XMG1.2; PharMingen), followed by strepta-vidin-HRP (Zymed, San Francisco, CA), and finally substrate (3,39,5,59-tetramethylbenzidine1 H2O2; PharMingen). Absorbance of each well wasmeasured spectrophotometrically at 450 nm using an automated ELISAreader and the picograms per milliliter of IFN-g captured in each well wasdetermined by extrapolating from an IFN-g standard curve.

Injection of CTL lines

Short-term CTL lines generated, as described above, were expanded invitro in 6-well plates, each well containing 43 10 6 CTLs and 1.43 107

LPS/dextran sulfate-activated splenocytes previously pulsed with 100mg/ml peptide for 1 h at 37°C and irradiated with 3000 rad. Eighteen hourslater, ConA-activated splenocyte supernatant (10% final concentration,v/v) was added to cultures that were fed and expanded as necessary. Fivedays after the last stimulation, cells were then washed, counted, suspendedin PBS containing 2% FCS, and injected i.v. into recipient 1.3.32 HBV-transgenic mice. One day after injection, mice were sacrificed and theirliver tissue biopsies were harvested for histological and immunohistochem-ical analyses or snap frozen in liquid nitrogen and stored at280°C forsubsequent molecular analyses (see below).

Tissue DNA and RNA analyses

Frozen liver tissue was mechanically pulverized under liquid nitrogen andtotal genomic DNA and RNA were isolated for Southern and Northern blotanalyses exactly as previously described (17). Nylon membranes were an-alyzed for HBV DNA, HBV RNA, GAPDH and 29,59-oligoadenylate syn-thetase as described elsewhere (19). Quantitation of cytokine, T lympho-cyte, and macrophage marker mRNAs was performed by RNase protectionassay exactly as previously described (6, 19).

Biochemical and histological analyses

The extent of hepatocellular injury was monitored by measuring serumalanine aminotransferase (sALT) activity at multiple time points after treat-ment with saline or the CTL lines were measured in a Paramax chemicalanalyzer (Baxter Diagnostics., McGaw Park, IL), exactly as previouslydescribed (6). For histological analysis, liver tissue samples were fixed in10% zinc-buffered Formalin (Anatech, Battle Creek, MI), embedded inparaffin, sectioned (3mm), and stained with hematoxylin and eosin exactlyas described elsewhere (17). To detect 5-bromo-29-deoxyuridine-positive cellsin tissue, Formalin-fixed paraffin-embedded tissue sections were stained withanti-5-bromo-29-deoxyuridine Abs, as previously described (20).

ResultsIdentification of novel HBV-Env CTL epitopes

Previous studies have characterized an HBV-Env-derived, H-2Ld-restricted dominant CTL epitope (17, 21, 22). To identify addi-tional epitopes and thus being able to generate responses of max-imum breadth, the HBV-Env protein sequence was scanned for thepresence of H-2d and H-2b motif peptides (Table I). The corre-sponding peptides were tested for binding to purified MHC mol-ecules in vitro. Consistent with previous studies (23, 24), a peptidewas classified as a good binder when associated with an IC50 of500 nM or less.

Two peptides that carried the Kd motif (Y2(LI)9, 10, (25)) wereidentified, one of which, Env.362, bound with an IC50 of 16 nM.Of three peptides containing the Dd motif (G2P3(ILF)8, 9, 10(25),

two (Env.364 and Env.281) bound with IC50 values of 5 and 115nM, respectively. Of 20 peptides carrying the Kb motif (FY5

(LIVM) 8,9 (25)), three (Env.329, Env.342, and Env.371) boundwith IC50values of 300, 150, and 469 nM, respectively. Finally,neither of the two Db motif (N5 (LIVM) 9,10(23, 24)]-positive pep-tides bound with an IC50 of ,500 nM. For the sake of comparison,the dominant Ld-restricted peptide was also tested for binding ca-pacity and an IC50 of 2.4 nM was recorded.

Next, the immunogenicity of H-2b and H-2d class I binding pep-tides was evaluated in B6D2F1 mice. Analogous to previous stud-ies (26, 27), peptides were emulsified in IFA in the presence of anexcess (140mg) of the I-Ab/d-restricted helper epitope OVA.323–336. Herein, to allow averaging and combination in a single con-sistent format in several different experiments, specific CTL ac-tivity is expressed in LU (LU30), in which 1 LU (LU30)corresponds to the number of effector cells required to induce 30%lysis of 104 51Cr-labeled target cells during the 4-h assay (18).Thus, in the conditions used in the assay 1 LU30 represents 30%lysis at the 100:1 E:T ratio. Ten LU30 represent 30% lysis at the10:1 ratio, 100 LU30 represent 30% lysis at the 1:1 ratio, and so on.Also, as defined previously (23, 24, 26), CTL responses of$2LU30/106 cells were considered positive.

The results of testing the three newly identified H-2d-bindingpeptides and the known dominant Env.28 epitope for immunoge-nicity are shown in Table II. As expected, vigorous responses wereobserved in the case of Env.28 (17/18 positive mice, 11.2 LUaverage response). Similar responses were observed in the case ofEnv.364, (12/14 mice yielded recall CTL responses, with an av-erage of 13.6 LU). Lower but still appreciable responses werenoted in the case of Env.362 (recall CTL responses in 7/12 wild-type mice, with an average of 5.7 LU). Finally, in the case ofEnv.281, recall CTL responses were observed in only 2 of 10 micetested (6.2 LU average magnitude).

Table I. In vitro binding capacity of HBV-Env-derived peptides

Peptide Sequence MotifBinding Capacity

(IC50, nM)

Env. 362 WYWGPSLYSI Kd 16Env. 368 LYSILSPFL Kd 2500Env. 364 WGPSLYSIL Dd 5.0Env. 281 TGPCRTCMT Dd 115Env. 281 TGPCRTCMTT Dd 4286Env. 28 IPQSLDSWWTSL Ld 2.4Env. 150 LSSIFSRI Kb 3333Env. 155 SRIGDPAL Kb —a

Env. 165 ENITSGFL Kb —Env. 178 LQAGFFLL Kb —Env. 231 TCPGYRWM Kb —Env. 233 PGYRWMCL Kb —Env. 242 RFIIFLFI Kb —Env. 243 FIIFLFIL Kb 5000Env. 244 IIFLFILL Kb 7500Env. 247 LFILLLCL Kb —Env. 252 LCLIFLLV Kb —Env. 253 CLIFLLVL Kb 6000Env. 259 VLLDYQGM Kb 3333Env. 329 ASARFSWL Kb 300Env. 342 FVQWFVGL Kb 150Env. 361 MWYWGPSL Kb 7500Env. 365 GPSLYSIL Kb —Env. 369 YSILSPFL Kb 1111Env. 371 ILSPFLPL Kb 469Env. 378 LLPIFFCL Kb —Env. 159 DPALNMENI Db —Env. 305 PSDGNCTCI Db —

a Dash indicates$10,000.

1391The Journal of Immunology

by guest on July 6, 2015http://w

ww

.jimm

unol.org/D

ownloaded from

Of the three HBV-ENV-derived H-2b binders, only one(Env.371) yielded a relatively weak CTL-positive response (3.0LU), and only in two of the six mice tested. Two control H-2Kb-restricted class I epitopes (lymphocytic choriomeningitis virus(LCMV) Np.396 and OVA.257) yielded vigorous responses (av-erage magnitudes of 6.5 and 16.9 LU, respectively) with 100% ofthe animals responding (data not shown).

Recognition of naturally processed HBV-ENV Ag by CTLsspecific for the Env.362 and Env.364 epitopes

To determine whether the Env.362 and the Env.364 epitopes weregenerated in the course of natural processing of the HBV-Env Ag,short-term CTL lines were derived by peptide immunization ofnontransgenic (B6D2F1) mice and tested for their capacity to rec-ognize P815-S cells that express all HBV-Env proteins (17).

As shown in Fig. 1, CTLs specific for control epitopes werecompletely devoid of cytotoxicity for the HBV-Env-transfectedP815 target cells in multiple independent experiments (Fig. 1,a

andb). As expected, high levels of specific killing were detected inthe case of Env.28 (Fig. 1c), with51Cr release approaching 100%at an E:T of 10:1. Recognition of P815-S target cells was alsodetected in the case of both Env.364- and Env.362-specific CTLlines (Fig. 1,d ande) with 51Cr release approaching 50% at an E:Tof 10:1 for both cultures.

It should be noted that in the case of Env.364 and Env.362,recognition of target cells expressing naturally processed Ag wasrelatively less efficient (as compared with peptide-pulsed targets)for these two epitopes (as compared with the dominant Env.28epitope), perhaps signifying that these epitopes are less abundantlyproduced by natural processing, which might in turn explain theirsubdominant status.

Immunogenicity of HBV-Env CTL epitopes in 1.3.32 HBV-transgenic mice

Next, we compared the immunogenicity of the two newly identi-fied H-2d-restricted epitopes (Env.362 and Env.364) and theknown dominant Env.28 epitope in H-2bxd CB6F1-nontransgenicmice and CB6F1 hybrids of 1.3.32 lineage HBV-transgenic mice.Specifically, we used a pool of HBV-Env-specific lipopeptidesconsisting of three constructs, each composed of a two palmiticacid lipid tail, the optimized PADRE HTL epitope, and an HBV-specific CTL epitope (PAM2-PADRE-Env.28, PAM2-PADRE-Env.362, and PAM2-PADRE-Env.364) to prime responses for thethree HBV-Env epitopes. Eleven to 14 days after injection of atotal of 30mg of the pool of lipopeptides, splenocytes were cul-tured for 6 days with the appropriate CTL epitopes and then testedfor activity in the chromium release assay. As expected, all of thenontransgenic H-2bxd F1 mice tested responded with CTL re-sponses in the 38–88 LU range (Table III). Interestingly, recallCTL responses were observed for a majority of the transgenicanimals tested, with magnitudes in the 10–86 LU range for thethree different HBV-specific CTL epitopes (Table III), comparable

Table II. Immunogenicity of HBV-Env-derived H-2d-binding peptidesa

Peptide

CTL Response

No. positive/No. tested LU30

Env. 28 17/18 11.2 (2.4)b

Env. 364 12/14 13.6 (2.0)Env. 362 7/12 5.7 (1.5)Env. 281 2/10 6.2 (1.5)

a Peptides were dissolved in DMSO/PBS. An IFA emulsion containing 50mg ofeach CTL epitope and 140mg of OVA.323 helper epitope was injected in B6D2F1

mice. Eleven days later, splenocyte cultures were stimulated with 10mg/ml of specificCTL epitopes and 107 syngeneic irradiated LPS blast cells. After 6 days, chromiumrelease CTL assays were performed using51Cr-labeled P815 (Kd) target cells in thepresence or absence of specific CTL epitopes.

b Geometric mean of positive cultures (X/4SD).

FIGURE 1. Recognition of naturally processed HBV-Env epitopes by short-term CTL lines. Groups of three individual B6D2F1 mice were injected with50mg of indicated CTL epitopes (a,LIS.91;b, LCMV Np.118;c, Env. 28;d, Env.364; ande,Env.362) and 140mg of the OVA.323 HTL epitope emulsifiedin IFA. Eleven days later, spleens were removed and pooled. Splenocytes were stimulated with 10mg/ml of each specific CTL epitope. After 10 days,cultures were restimulated using syngeneic irradiated LPS blast cells pulsed with 10mg/ml of each specific CTL epitope. Cultures were assayed 5 dayslater for cytolytic activity using a standard 4-h51Cr release assay using P815 target cells in the presence (E) or absence (‚) of the specific peptide orPreS1-transfected P815 target cells (M).

1392 INDUCTION OF HBV-SPECIFIC CTLS IN HBV-TRANSGENIC MICE

by guest on July 6, 2015http://w

ww

.jimm

unol.org/D

ownloaded from

to those induced in nontransgenic mice. These results demonstratethat T cell tolerance can be overcome by lipopeptide immunizationat the level of induction of recall CTL activity. However, unlikethe CB6F1-derived CTLs, the CTLs derived from the HBV-trans-genic mice did not recognize endogenously synthesized Ag sincethey failed to kill HBV-Env-transfected target cells (Table IV). In theHBV-transgenic mouse system, in addition to monitoring for the in-duction of recall memory CTL responses after immunization, the ca-pacity of the induced CTLs to lyse HBV-transgenic hepatocytes in theimmunized mice in vivo can be monitored biochemically by mea-suring sALT activity, a marker of hepatocellular lysis. Despitedetectable recall memory CTLs, indicative of in vivo CTL induc-tion, sALT activity remained at background levels in all of themice (50–60 U/L; data not shown). Furthermore, no significantinflammatory foci were detected by immunohistochemistry, and nodown-regulation of viral DNA was detected in the liver of theseanimals (data not shown). This was in keeping with their failure tokill endogenously synthesized Ags in vitro (see above).

HBV-specific CTLs generated in 1.3.32-transgenic mice are oflower avidity relative to those generated in nontransgenic mice

In the next series of experiments, we compared the avidity of theCTLs generated in nontransgenic CB6F1 mice and in 1.3.32 lin-eage CB6F1 mice using a dose titration of the various epitopes.One of two independent experiments yielding similar results isshown in Fig. 2. In the case of the Env.28 epitope, the concentra-tion of peptide necessary to achieve 30% lysis by CTLs from

1.3.32 CB6F1-transgenic mice and nontransgenic mice was 5.9 and0.13 pg/ml, respectively. In the case of the Env.364 epitope, theconcentrations necessary for 30% lysis were 276 and 0.256 ng/mlfor 1.3.32 and CB6F1-derived CTL lines, respectively. Thus, inboth cases, CTL lines derived from transgenic mice displayed 100-to 1000-fold lower avidity for their targets. In contrast, little or nodifference in avidity was detected in the case of the Env.362epitope.

Transfer of CTL derived from 1.3.32-transgenic mice isassociated with HBV Ag down-regulation in the absence of liverpathology

Adoptive CTL transfer experiments were performed next to furthercharacterize the quality of the recall CTL activity elicited by li-popeptide immunization in the transgenic and nontransgenic mice.Groups of three CB6F1 mice were immunized with a pool of thethree lipidated PADRE-CTL constructs encompassing the Env.28,362, and 364 CTL epitopes. After 1 mo, all groups were boostedwith the same immunogens. Eleven to 14 days later, splenocyteswere harvested and CTL-specific lines were expanded in vitro withthe appropriate CTL epitopes.

Five days after the second in vitro stimulation, CTL culturesfrom each group were tested for activity in a51Cr release assay.Data from one representative experiment is shown in Table V.Cultures derived from normal CB6F1 mice yielded strong CTLactivity in the 60–220 LU range, and cultures derived from 1.3.32-transgenic mice yielded slightly less CTL activity (in the 40–160LU range) against P815 target cells pulsed with the various HBV-Env-derived epitopes.

Equal numbers of the nontransgenic or transgenic CTL linesspecific for the three different HBV-Env CTL epitopes werepooled, and a total of 13 107 nontransgenic or transgenic cellswere injected in groups (six mice per group) of 8- to 10-wk-oldfemales and serum hepatitis B e Ag-matched animals from lineage1.3.32 (CB6F1 hybrids) HBV-transgenic mice. An equal numberof cells from CTLs specific for the LCMV Np.118 CTL epitopewas also injected into the HBV-transgenic mice to control for non-specific effects.

Transfer of nontransgenic HBV-specific CTL lines into trans-genic mice was associated with increased levels of sALT activity(Table VI), and the presence of small, scattered necroinflammatoryfoci in the liver parenchyma (data not shown), both at day 1 andday 3 after transfer (data not shown). As shown in Fig. 3, 1 dayafter injection, HBV DNA replication was strongly inhibited in theliver of these animals when compared with saline-injected con-trols. This was associated with the intrahepatic induction of IFN-g,TNF-a, CD8, CD4, CD3, and F480 mRNA (Fig. 3). These resultsare not surprising since using the same mouse model we havepreviously shown that the antiviral potential of the CTLs is pri-marily mediated by noncytolytic mechanisms that involve the in-trahepatic production of IFN-g (6, 7).

Transfer of transgenic HBV-Env-specific CTLs into transgenicmice was not associated with increased sALT activity and veryfew necroinflammatory foci were detected in the liver (data notshown). The signal for CD8, CD4, CD3, and F480 RNAs (whichindicates the number of liver-infiltrating inflammatory cells, in-cluding the adoptively transferred CTLs) in these animals washigher than that of saline-injected controls and lower than that ofmice that received nontransgenic CTLs (Fig. 3). This is consistentwith their lower avidity (except Env 362) and poorer recognitionof endogenously synthesized Ag than the nontransgenic CB6F1

CTLs. Nonetheless, HBV replication was strongly inhibited in theanimals that received transgenic CTLs, almost to the same degreeobserved in the mice that received nontransgenic CTLs. This was

Table III. Immunogenicity of a lipopeptide pool in nontransgenic and1.3.32 HBV-transgenic mice: recognition of peptide-pulsed target cellsa

Peptide

CTL Response

Nontransgenic 1.3.32 HBV Transgenic

No. positive/No. tested LU30

No. Positive/No. Tested LU30

Env. 28 3/3 79.6 (1.5)b 3/3 86.1 (1.2)Env. 364 3/3 38.9 (1.3) 3/3 10.8 (1.7)Env. 362 3/3 87.7 (1.3) 3/3 58.9 (1.0)

a Lipopeptides were formulated as described inMaterials and Methodsand in-jected as a pool into nontransgenic CB6F1 or 1.3.32 HBV CB6F1-transgenic mice.Eleven days later, splenocyte cultures were stimulated with 10mg/ml of specific CTLepitopes and 107 syngeneic irradiated LPS blast cells. After 6 days, chromium releaseCTL assays were performed using51Cr-labeled P815 (Kd) target cells in the presenceor absence of specific CTL epitopes or P815-S target cells.

b Geometric mean of positive cultures (X/4SD).

Table IV. Immunogenicity of a lipopeptide pool in nontransgenic and1.3.32 HBV-transgenic mice: recognition of P815-S-transfected targetcellsa

Peptide

CTL Response

Nontransgenic 1.3.32 HBV Transgenic

No. positive/No. tested LU30

No. Positive/No. Tested LU30

Env. 28 3/3 33.9 (1.0)b 0/3 —b

Env. 364 3/3 14.2 (1.2) 0/3 —Env. 362 3/3 8.2 (1.2) 0/3 —

a Lipopeptides were formulated as described inMaterials and Methodsand in-jected as a pool into nontransgenic CB6F1 or 1.3.32 HBV CB6F1-transgenic mice.Eleven days later, splenocyte cultures were stimulated with 10mg/ml of specific CTLepitopes and 107 syngeneic irradiated LPS blast cells. After 6 days, chromium releaseCTL assays were performed using51Cr-labeled P815 (Kd) target cells in the presenceor absence of specific CTL epitopes or P815-S target cells.

b Geometric mean of positive cultures (X/4SD). Dash indicates#2.0 LU30.

1393The Journal of Immunology

by guest on July 6, 2015http://w

ww

.jimm

unol.org/D

ownloaded from

associated with a moderate induction of IFN-g, indicating that Agrecognition has occurred in vivo (Fig. 3). Finally, no liver diseasewas detected in the control group that received LCMV-specificCTLs. As expected, HBV replication was not affected and IFN-gproduction was not induced in the liver of these mice (Fig. 3).

These in vivo observations were paralleled by in vitro experi-ments which demonstrated that CTL derived from HBV-transgenicmice did not kill HBV-Env-transfected target cells, but were ca-pable of secreting IFN-g, albeit in small amounts, when confrontedwith HBV-Env-transfected targets. Control CTL lines derivedfrom normal mice were capable, by contrast, of both killing andIFN-g secretion in response to HBV-Env-transfected p815 targetcells (Fig. 4).

DiscussionImmunogenicity screening in nontransgenic mice of H-2d-binding,HBV-Env-derived CTL peptides, emulsified in IFA, revealed twonew subdominant CTL epitopes (Env.362 and Env.364) whichelicited CTL specificities that recognized naturally processed Agin the form of HBV-Env-transfected target cells. In contrast, onlythree H-2b-binding peptides were identified, and they bound withrelatively low affinity and were found to be essentially nonimmu-nogenic. These results are consistent with the observation thatH-2b mice are CTL nonresponders for the Env protein (28, 29).Overall, these results, along with the observation that the previ-ously known dominant Ld-restricted Env.28 epitope was the high-est affinity binder identified, confirm the important role of classI-binding capacity in determining immunodominance.

Relative to Env.28, recognition of endogenously synthesized Agwas less efficient in the case of the subdominant epitopes (Env.364and Env.362), suggesting that these epitopes are produced lessefficiently by natural HBV-Env processing. This, along with their

lower binding affinity relative to Env.28, might explain their sub-dominant status (30). Analysis of lineage 1.3.32 HBV-transgenicmice, in terms of CTL responsiveness directed against the newH-2d-restricted epitopes and the previously defined dominantEnv.28 epitope, demonstrated T cell hyporesponsiveness in 1.3.32mice for the Env.28 and Env.364, but less so in the case of theEnv.362 epitope. It is possible that this subdominant epitope mightbe a less effective tolerogen because it is naturally expressed inlower amounts in hepatocytes of 1.3.32 HBV-transgenic mice.These results suggest that targeting subdominant epitopes may bean effective way to overcome CTL tolerance.

Studies in nontransgenic H-2d mice demonstrated that a mixtureof HTL-CTL epitope lipoconstructs was capable of simultaneouslyinducing responses for each of these CTL epitopes. In 1.3.32HBV-transgenic mice, this immunization strategy was also effec-tive in inducing recall CTL activity. These results are significantbecause previous attempts aimed at overcoming CTL tolerance inthis lineage of transgenic mice had not met with success (8). Acrucial element of this strategy appears to be the use of prepro-cessed epitopes, as in this system whole HBV-Env DNA immu-nization does not break CTL tolerance (8), but epitope-based mini-genes do (A. D. Sette, C. Oseroff, and J. Alexander, unpublishedobservations). However, the induction of HBV-specific CTLs wasnot associated with either down-regulation of expression of HBVAgs in the liver or with increased sALT activity, indicating lack ofhepatocyte damage. Thus, the induced CTLs either did not reachthe liver in sufficient numbers to exert detectable effector func-tions, or the CTLs were functionally defective in the transgenicmice, or the target Ags were not efficiently processed. The lastexplanation is unlikely since all three epitopes are presented byEnv-transfected target cells (Fig. 1) and because Env.28 CTLs are

Table V. Activity of CTL lines used for adoptive transfer experimentsa

ImmunizationIn Vitro

Restimulation

CTL Activityb (LU30)

Nontransgenic1.3.32 HBVtransgenic

Pool of HBV-Env CTL lipopeptides Env. 28 216.5 (1.05) 155.4 (1.32)Env. 364 71.8 (1.1) 43.8 (1.5)Env. 362 64.7 (1.2) 51.3 (1.2)

LCMV Np.118 lipopeptide LCMV Np.118 ND 53.2 (1.2)

a Nontransgenic CB6F1 or 1.3.32 HBV-transgenic CB6F1 mice were injected and boosted with a pool of lipopeptidesencoding the three HBV-Env epitopes as described inMaterials and Methods.Splenocytes were restimulated twice in vitro witheach epitope separately and then tested for activity in a standard chromium release assay.

b Geometric mean (X/4SD) of the lytic activity for peptide-pulsed P815 target cells of three different cultures.

FIGURE 2. Avidity of HBV-Env-specificCTL lines. Splenocytes were derived fromnontransgenic CB6F1 mice (E) or 1.3.32 HBVCB6F1-transgenic mice (F) immunized withthe pool of HBV-Env CTL epitope lipopep-tides. Lines were derived by in vitro stimula-tion with Env.28 (a), Env.362 (b), and Env.364(c) as described inMaterials and Methods.Their activity was assessed using P815 targetcells and a dose titration of the relevantpeptide.

1394 INDUCTION OF HBV-SPECIFIC CTLS IN HBV-TRANSGENIC MICE

by guest on July 6, 2015http://w

ww

.jimm

unol.org/D

ownloaded from

known to recognize their target epitopes very efficiently in vivo (6,20, 21).

Direct visualization and/or detection of CTL activity in freshlyharvested splenocytes by means of tetrameric staining was not at-tempted in the current set of experiments. Future studies mightinvestigate the relationship between these CTLs and the function-ally impaired CTLs observed in the course of chronic infectionwith various viruses including LCMV, hepatitis C virus, and HBV(2, 31, 32). It is important to point out, however, that in thesecases, functionally impaired CTLs were associated with chronicexposure to viral Ags, whereas in our cases, these CTLs are in-

duced de novo after immunization of animals already chronicallyexpressing large amounts of HBV Ags.

Further adoptive transfer studies characterized the CTLs in-duced in 1.3.32-transgenic mice in more detail. It was foundthat these CTLs were effective in inhibiting viral replication.However, unlike the control CTLs induced in nontransgenicmice, they did not cause hepatitis. This antiviral effect is likelymediated by noncytolytic mechanisms that involve intrahepaticproduction of IFN-g (6, 33). Peptide titration experiments dem-onstrated that CTLs derived from 1.3.32 HBV-transgenic micewere characterized by low avidity, thus suggesting a possiblemechanism to account for these observations if the lower avid-ity recognition can trigger the release of antiviral cytokineswithout significantly activating the cytolytic function of theCTLs. The CTLs induced in nontransgenic mice were moreeffective than those induced in nontransgenic mice, possibly areflection of both cytolytic and noncytolytic mechanisms aswell as potentially higher levels of production of noncytolyticmediators. Nevertheless, it is obvious that selective induction ofantiviral effects in chronically infected individuals, in the ab-sence of tissue damage, would be of significant therapeuticinterest.

These results are also consistent with previous observationsby Sherman and coworkers (34) and von Herrath et al. (35) anddemonstrated that low-avidity CTLs can escape tolerance inac-tivation and might be stimulated by immunization with vaccine

Table VI. Transfer of 1.3.32-derived CTL lines is not associated withincreased sALT levels in 1.3.32 HBV-transgenic recipientsa

Specificity and Origin of CTL LineTransferred Mouse

Day After Transfer

1(U/ml)

2(U/ml)

HBV-specific, 1.3.32 transgenic 1 51 632 51 54

HBV-specific, H-2-matched control 1 174 3482 258 330

LCMV-specific, 1.3.32 transgenic 1 63 452 48 69

a sALT levels in two control 1.3.32 HBV-transgenic mice that did not receive anycell transfer were 54 and 66 U/L.

FIGURE 3. Transfer of CTL derived from 1.3.32-transgenic mice in-hibits HBV replication. Age-matched, sex-matched, and serum hepatitis Be Ag-matched lineage 1.3.32 HBV-transgenic mice were injected i.v. with1 3 107 nontransgenic or transgenic CTLs and sacrificed 24 h later. Con-trol mice were injected with either saline (NaCl) or CTL derived fromnontransgenic mice immunized with an LCMV-specific peptide. Total he-patic DNA was analyzed for HBV DNA by Southern blot analysis. AllDNA samples were RNase treated before quantitation and gel electro-phoresis. Bands corresponding to the integrated transgene, relaxed circular(RC) double-stranded HBV DNA, and single-stranded (SS) linear HBVDNA replicative forms are indicated. The integrated transgene can be usedto normalize the amount of DNA bound to the membrane. The filter washybridized with a32P-labeled HBV-specific DNA probe. Total hepaticRNA was analyzed for cytokine, T cell, and macrophage marker transcriptsby RNase protection assay, as indicated. The mRNA encoding the ribo-somal protein L32 was used to normalize the amount of RNA loaded ineach lane of the RNase protection assay.

FIGURE 4. IFN-g secretion and killing activity of CTL lines derivedfrom 1.3.32-transgenic mice. The capacity of short-term CTL lines gener-ated from 1.3.32-transgenic and -nontransgenic mice against three Envepitopes to display epitope-specific cytotoxic activity or IFN-g productionwas determined. CTLs were tested for effector function against P815 tumorcells in the presence or absence of peptide, or against P815 cells transfectedwith the HBV Env gene. Cytotoxic activity was measured using the stan-dard51Cr release assay at a 10:1 E:T ratio while cytokine production wasdetermined using an in situ capture ELISA. Net CTL responses are shownafter subtracting background. CTL activity against P815 cells in the ab-sence of exogenous peptide (background) ranged from 4 to 16% (cytotox-icity) and 17 to 96 pg/well (IFN-g).

1395The Journal of Immunology

by guest on July 6, 2015http://w

ww

.jimm

unol.org/D

ownloaded from

constructs using preprocessed optimal epitopes. A crucial point,which remains to be addressed, is why, in the current study,induction of this CTL specificity did not affect viral replicationin the immunized animals, but only in the adoptive transferexperiments. Our results are compatible with two nonmutuallyexclusive explanations. First, it is possible that in vitro expan-sion might in some way reverse a functional inactivation state.Alternatively, and in our view more likely, it is possible that themagnitude, in terms of the number and expansion of CTL spec-ificities induced in the transgenic mice, is suboptimal and invitro expansion is necessary to achieve sufficient numbers ofCTL effectors.

In this light, we are currently investigating different immuniza-tion protocols such as prime/boost strategies (36) and the com-bined use of multiple epitopes from various HBV Ags as a way tooptimize the magnitude and breadth of CTL responses. Other stud-ies have also emphasized the importance of optimized helper ac-tivity in overcoming CTL unresponsiveness (37). Taken together,we believe that these results are very encouraging in terms of po-tential immunotherapy of HBV infection in humans in which, as aresult of the recent availability of antiviral compounds such aslamivudine, problems related to viral expression and its suppres-sive effects might be less severe.

In conclusion, we have shown, for the first time, that immuni-zation of HBV 1.3.32-transgenic mice can result in induction ofCTL activity capable, in passive transfer experiments, of down-regulating expression of viral Ags in the absence of hepatocytedamage. These observations have obvious implications for devel-opment of therapeutic strategies to treat chronic HBV infection.

Finally, we would also like to point out that these results pre-sented in this manuscript may also have implications for autoim-mune disease and cancer treatments. In several instances, it hasbeen reported that T cells involved in autoimmune phenomena (38,39) or directed against tumor-associated Ags (34) are associatedwith altered avidity or patterns of cytokine release. Thus, under-standing the structural basis of the noncytolytic mechanisms de-scribed herein may reveal interesting avenues for treatment of sev-eral complex diseases.

AcknowledgmentsWe acknowledge the expert and patient secretarial assistance of DorothyWharry and Robin Delp. We also thank Monte Hobbs for providing thecytokine gene and T cell marker probe sets used in the RNase protectionassays and Mingsheng Qin, Rick Koch, Christina Whitten, Heike Heiser,and Margie Chadwell for excellent technical assistance.

References1. Tiollais, P., C. Pourcel, and A. Dejean. 1985. The hepatitis B virus.Nature

6:317..

2. Chisari, F. V., and C. Ferrari. 1995. Hepatitis B virus immunopathogenesis.Annu. Rev. Immunol. 13:29.

3. Nayersina, R., P. Fowler, S. Guilhot, G. Missale, A. Cerny, H. J. Schlicht,A. Vitiello, R. Chesnut, J. L. Person, A. G. Redeker, and F. V. Chisari. 1993.HLA A2 restricted cytotoxic T lymphocyte responses to multiple hepatitis Bsurface antigen epitopes during hepatitis B virus infection.J. Immunol. 150:4659.

4. Rehermann, B., P. Fowler, J. Sidney, J. Person, A. Redeker, M. Brown, B. Moss,A. Sette, and F. V. Chisari. 1995. The cytotoxic T lymphocyte response to mul-tiple hepatitis B virus polymerase epitopes during and after acute viral hepatitis.J. Exp. Med. 181:1047.

5. Guidotti, L. G., K. Ando, M. V. Hobbs, T. Ishikawa, L. Runkel,R. D. Schreiber,and F. V. Chisari. 1994. Cytotoxic T lymphocytes inhibit hepa-titis B virus gene expression by a non-cytolytic mechanism in transgenic mice.Proc. Natl. Acad. Sci. USA 91:3764.

6. Guidotti, L. G., T. Ishikawa, M. V. Hobbs, B. Matzke, R. Schreiber, andF. V. Chisari. 1996. Intracellular inactivation of the hepatitis B virus by cytotoxicT lymphocytes.Immunity 4:25.

7. Guidotti, L. G., R. Rochford, J. Chung, M. Shapiro, R. Purcell, and F. V. Chisari.1999. Viral clearance without destruction of infected cells during acute HBVinfection.Science 284:825.

8. Shimizu, Y., L. G. Guidotti, P. Fowler, and F. V. Chisari. 1998. Dendritic cellimmunization breaks cytotoxic T lymphocyte tolerance in hepatitis B virus trans-genic mice.J. Immunol. 161:4520.

9. Mancini, M., M. Hadchouel, H. L. Davis, R. G. Whalen, P. Tiollais, andM. L. Michel. 1996. DNA-mediated immunization in a transgenic mouse modelfor the hepatitis B surface antigen chronic carrier state.Proc. Natl. Acad. Sci.USA 93:12496.

10. Pourcel, C., P. Tiollais, H. Farza. 1990. Transcription of the S gene in transgenicmice is associated with hypomethylation at specific sites and with DNase I sen-sitivity. J. Virol. 64:931.

11. Atherton, E., and R. C. Sheppard. 1989. Fluorenylmethoxycarbonylpolyamidesolid phase peptide synthesis: general principles and development.In Solid PhasePeptide Synthesis: A Practical Approach.IRL Press, Eynsham, Oxford, pp.25–37.

12. Sidney, J., M. F. del Guercio, S. Southwood, V. H. Engelhard, E. Appella,H. G. Rammensee, K. Falk, O. Rotzschke, M. Takiguchi, R. T. Kubo, et al. 1995.Several HLA alleles share overlapping peptide specificities.J. Immunol. 154:247.

13. Sette, A., J. Sidney, M. F. del Guercio, S. Southwood, J. Ruppert, C. Dahlberg,H. M. Grey, and R. T. Kubo. 1994. Peptide binding to the most frequent HLA-Aclass I alleles measured by quantitative molecular binding assays. Mol. Immunol.31:813.

14. Ruppert, J., J. Sidney, E. Celis, R. T. Kubo, H. M. Grey, and A. Sette. 1993.Prominent role of secondary anchor residues in peptide binding to HLA-A2.1molecules.Cell 74:929.

15. Kubo, R. T., A. Sette, H. M. Grey, E. Appella, K. Sakaguchi, N. Z. Zhu,D. Arnott, N. Sherman, J. Shabanowitz, H. Michel, et al. 1994. Definition ofspecific peptide motifs for four major HLA-A alleles.J. Immunol. 152:3913.

16. Kast, W. M., R. M. Brandt, J. Sidney, J. W. Drijfhout, R. T. Kubo, H. M. Grey,C. J. Melief, and A. Sette. 1994. Role of HLA-A motifs in identification ofpotential CTL epitopes in human papillomavirus type 16 E6 and E7 proteins.J. Immunol. 152:3904.

17. Guidotti, L. G., B. Matzke, H. Schaller, and F. V. Chisari. 1995. High-levelhepatitis B virus replication in transgenic mice.J. Virol. 69:6158.

18. Oseroff, C., A. Sette, P. Wentworth, E. Celis, A. Maewal, C. Dahlberg, J. Fikes,R. T. Kubo, R. W. Chesnut, H. M. Grey, and J. Alexander. 1998. Pools oflipidated HTL-CTL constructs prime for multiple HBV and HCV CTL epitoperesponses.Vaccine 16:823.

19. Guidotti, L. G., P. Borrow, M. V. Hobbs, B. Matzke, I. Gresser, M. B. Oldstone,and F. W. Chisari. 1996. Viral cross talk: intracellular inactivation of the hepatitisB virus during an unrelated viral infection of the liver.Proc. Natl. Acad. Sci. USA93:4589.

20. Ando, K., L. G. Guidotti, A. Cerny, T. Ishikawa, and F. V. Chisari. 1994. Accessto tissue antigen is restricted in vivo.J. Immunol. 153:482.

21. Ando, K., T. Moriyama, L. G. Guidotti, S. Wirth, R. D. Schreiber, H. J. Schlicht,S. N. Huang, and F. V. Chisari. 1993. Mechanisms of class I restricted immu-nopathology: a transgenic mouse model of fulminant hepatitis.J. Exp. Med. 178:1541.

22. Chisari, F. V., P. Filippi, J. Buras, A. McLachlan, H. Popper, C. A. Pinkert,R. D. Palmiter, and R. L. Brinster. 1987. Structural and pathological effects ofsynthesis of hepatitis B virus large envelope polypeptide in transgenic mice.Proc. Natl. Acad. Sci. USA 84:6909.

23. Sette, A., A. Vitiello, B. Reherman, P. Fowler, R. Nayersina, W. M. Kast,C. J. Melief, C. Oseroff, L. Yuan, J. Ruppert, et al. 1994. The relationship be-tween class I binding affinity and immunogenicity of potential cytotoxic T cellepitopes.J. Immunol. 53:5586.

24. Wentworth, P. A., A. Vitiello, J. Sidney, E. Keogh, R. W. Chesnut, H. M. Grey,and A. Sette. 1996. Differences and similarities in the A2.1-restricted cytotoxic Tcell repertoire in humans and human leukocyte antigen-transgenic mice.Eur.J. Immunol. 26:97.

25. Falk, K., O. Rotzschke, S. Stevanovic, G. Jung, H. G. Rammensee. 1991. Allele-specific motifs revealed by sequencing of self-peptides eluted from MHC mol-ecules.Nature 351:290.

26. Wentworth, P. A., A. Sette, E. Celis, J. Sidney, S. Southwood, C. Crimi,S. Stitely, E. Keogh, N. C. Wong, B. Livingston, et al. 1996. Identification ofA2-restricted hepatitis C virus-specific cytotoxic T lymphocyte epitopes fromconserved regions of the viral genome.Int. Immunol. 8:651.

27. Alexander, J., C. Oseroff, J. Sidney, P. Wentworth, E. Keogh, G. Hermanson,F. V. Chisari, R. T. Kubo, H. M. Grey, and A. Sette. 1997. Derivation of HLA-A11/Kb transgenic mice: functional CTL repertoire and recognition of humanA11-restricted CTL epitopes. [Published erratum appears in1999 J. Immunol.162:3104.]J. Immunol. 159:4753.

28. Fuller, J. T., D. H. Fuller, D. McCabe, J. R. Haynes, and G. Widera. 1995.Immune responses to hepatitis B virus surface and core antigens in mice, mon-keys, and pigs after Accell particle-mediated DNA immunization.Ann. NY Acad.Sci. 772:282.

1396 INDUCTION OF HBV-SPECIFIC CTLS IN HBV-TRANSGENIC MICE

by guest on July 6, 2015http://w

ww

.jimm

unol.org/D

ownloaded from

29. Whitton, J. L., A. Tishon, H. Lewicki, J. Gebhard, T. Cook, M. Salvato, E. Joly,and M. B. Oldstone. 1989. Molecular analyses of a five-amino-acid cytotoxicT-lymphocyte (CTL) epitope: an immunodominant region which induces nonre-ciprocal CTL cross-reactivity. Virology 63:4303.

30. Chen, W., L. C. Anton, J. R. Bennink, and J. W. Yewdell. 2000. Dissecting themultifactorial causes of immunodominance in class I-restricted T cell responsesto viruses.Immunity 12:83.

31. Borrow, P., and M. B. A. Oldstone. 1997. Lymphocytic choriomeningitis virus.In Viral Pathogenesis.N. Nathanson, ed. Lippincott-Raven, Philadelphia, p. 593.

32. Rehermann, B., and F. V. Chisari. 2000. Cell mediated immune response to thehepatitis C virus.Curr. Top. Microbiol. Immunol. 242:299.

33. McClary, H., R. Koch, F. V. Chisari, and L. G. Guidotti. 2000. Relative sensi-tivity of hepatitis B virus and other hepatotropic viruses to the anti-viral effectsof cytokines.J. Virol. 74:2255.

34. Morgan, D. J., H. T. Kreuwel, S. Fleck, H. I. Levitsky, D. M. Pardoll, andL. A. Sherman. 1998. Activation of low avidity CTL specific for a self epitoperesults in tumor rejection but not autoimmunity.J. Immunol. 160:643.

35. von Herrath, M. G., D. P. Berger, D. Homann, T. Tishon, A. Sette, andM. B. Oldstone. 2000. Vaccination to treat persistent viral infection.Virology268:411.

36. Schneider, J., S. C. Gilbert, C. M. Hannan, P. Degano, E. Prieur, E. G. Sheu,M. Plebanski and A. V. Hill. 1999. Induction of CD81 T cells using heterologousprime-boost immunisation strategies.Immunol. Rev. 170:29.

37. Livingston, B. D., J. Alexander, C. Crimi, C. Oseroff, E. Celis, K. Daly,L. G. Guidotti, F. V. Chisari, J. Fikes, R. W. Chesnut, and A. Sette. 1999. Alteredhelper T lymphocyte function associated with chronic hepatitis B virus infection andits role in response to therapeutic vaccination in humans.J. Immunol. 162:3088.

38. von Herrath, M. G., J. Dockter, M. Nerenberg, J. E. Gairin, andM. B. A. Oldstone. 1994. Thymic selection and adaptability of cytotoxic T lym-phocyte responses in transgenic mice expressing a viral protein in the thymus.J. Exp. Med. 180:1901.

39. von Herrath, M. G., J. Dockter, and M. B. A. Oldstone. 1994. How virus inducesa rapid or slow onset insulin-dependent diabetes mellitus in a transgenic model.Immunity 1:231.

1397The Journal of Immunology

by guest on July 6, 2015http://w

ww

.jimm

unol.org/D

ownloaded from