Oral lichen Planus Pathogenesis: A Role for the HCWpecific Cellular Immune Response

Transcription

1 Oral lichen Planus Pathogenesis: A Role for the HCWpecific Cellular Immune Response Massimo Pilli, Amalia Penna, Alessandro Zerbini,' Paolo Vescovi,2 Maddalena Manfredi,z Francesco Negro,3 Marco Carrozzo,* Cristina Mori,' Tiziana Giuberti,' Carlo Ferrari,' and Gabriele Missale' Hepatitis C virus infection can be associated with different extrahepatic manifestations, including lichen planus; however, no clear role for HCV in their pathogenesis has been established. T cells were isolated from lichen biopsy specimens of 7 HCV positive patients with oral lichen planus. HCV-specific CD4+ T-cell lines were obtained in 4 patients from lichen lesions but only in 2 of them from the peripheral blood. Different clonal populations were found in oral tissue and peripheral blood of individual patients, as shown by TCR-VP analysis of antigen-specific T cells. Frequency of HCV-specific CD8+ cells tested with 4 different HCV tetramers was significantly higher in the lichen tissue than in the circulation; moreover, lichen-derived HCV-specific CD8+ T cells showed the phenotype of recently activated T cells because most of them were CD69+ and produced interferon gamma (IFN-y) but expanded poorly in vitro upon antigen stimulation. The specificity of HCVreactive T-cell recruitment into the lichen tissue was further confirmed by the absence of HBV-specific T cells within lichen lesions in 3 additional patients with lichen planus associated with HBV infection. Our study shows HCV-specific T-cell responses at the site of the lesions of an HCV-associated dermatologic disease, sustained by HCV-specific T cells with phenotypic and functional characteristics of terminally differentiated effector cells. In conclusion, this finding and the detection of HCV RNA strands in the lichen tissue strongly suggest a role for HCV-specific T-cell responses in the pathogenesis of oral lichen planus associated with HCV infection. (HEPATOLOGY 2002;36: ) S everal extrahepatic diseases have been reported to be associated with hepatitis C virus (HCV) infection. However, no clear role for HCV in their pathogenesis has been established, with the exception of mixed cryoglobulinemia for which the interaction between HCV and the CD8 1 molecule has been proposed to play a pathogenetic role.' Among these extrahepatic manifes- Abbreviations: HCV, hepatitis C virus: LP, lichen planus; HBK hepatitis B virus: HBsAg, hepatitis B suface antigen: HBeAg, hepatitis B e antigen: HBcAg, hepatitis R core antigen. From the 'Divisione Malattie Infittiye ed Epatologia, Azienda Ospedaliera Universitaria di Parma, Parma, Italy: 'Dipartimento di Scienze Otorino-Odonto- Ofialmologiche e Cervico-Facciali, Universitu di Parma, Parma, Italy; 'Diuision of Gastroenterology and Hepatology, University Hospital, Geneva. Switzerland: and 4Dipartimento di Fisiopatologia Clinica, Sezione di Medicina Orale, Universitri di Torino, Torino, Italy. Received August 16, 2002; accepted September 24, Supported in part by grants from EC Biomed (BMH , Fondazione Cassa di Risparmio di Parma, Italy, andfrom the Swiss National Science Foundation ( to EN.). Address reprint requests to: G. Missale, M. D., Divisione Malattie Infittive ed Epatologia, Azienda Ospedaliera e Universitaria di Parma, Via Gramsci 14, Parma, Italy. . mis~ale@tin.it:fix: (39) Copyright by the American Association for the Study of Liver Diseases /02/ $35.00/0 doi:lo.i053bhep tations, lichen planus (LP) is well known to be associated with chronic liver diseases,* although its association with HCV infection is still a debated issue, appearing to be proven only in some geographical areas, such as Japan3 and Southern Recently, positive and negative HCV RNA strands were detected in epithelial cells of the normal oral mucosa and in lesional tissue of oral LP from anti-hcv positive patients by both strand-specific reverse-transcription polymerase chain reaction (RT- PCR)s and in situ hybridization.9 LP is a mucocutaneous disease characterized by a cellular inflammatory infiltrate enriched in CD4+ cells, by the presence of acidophilic bodies that may represent apoptotic epithelial cells, and by vacuolating degeneration of the basal epithelial layer. Because replicative forms of HCV have been detected within LP lesions and a cellmediated damage of basal layer cells is believed to be the crucial pathogenetic mechanism responsible for LP lesions, we asked whether the cell-mediated immune response against HCV can play a pathogenetic role in LP either by a direct effect or by triggering an autoimmune reaction. For this purpose, the aims of our study were to test for the presence and to analyze the phenotypic and

2 HEPATOLOGY, Vol. 36, No. 6, 2002 PILL1 ET AL functional characteristics of HCV-specific CD4 and CD8 T cells at the site of the oral LP lesions. Patients and Methods Patients. Seven patients with anti-hcv positive chronic hepatitis and 3 patients with chronic HBV infection and oral LP were studied. Sera were tested for qualitative HCV RNA by the commercially available COBAS Amplicor test (Roche-Diagnostics, Basel, Switzerland), for quantitative HCV RNA by commercially available HCV RNA Quantiplex 2.0 Assay (Chiron Corporation, Emeryville, CA), and for HBV DNA by PCR (Cobas Amplicor Test, Roche Diagnostic). Genotyping was performed by line probe assay (Immunogenetics, Zwijnaarde, Belgium). Patients with chronic HBV infection were hepatitis B surface antigen (HBsAg) positive and hepatitis B e antigen (HBeAg) negative; HBV DNA, tested by quantitative PCR (Roche Diagnostics), was 2,770, <200 and, 656 copies/ml in patients A, B, and C, respectively, on the day of the lichen biopsy. All patients underwent oral mucosa biopsy as part of their diagnostic evaluation for lichen planus. Patients gave written informed consent before entering the study, and the study protocol conformed to the ethical guidelines of the 1975 Declaration of Helsinki. HBV and HCV Antigens and Peptides. Recombinant hepatitis B core antigen (HBcAg) and HBsAg were purchased from DiaSorin (Saluggia, Italy). HBeAg was purchased from Biogen (Geneva, Switzerland). The HCV antigens core, El, E2, NS3, NS4, and NS5'O were kindly provided by Chiron Corporation. Nineteen synthetic 20mer peptides overlapping by 10 amino acid residues and covering the entire HCV core antigen of genotype la and 8 peptides, derived from regions of NS3 conserved across the different HCV genotypes,l' were purchased from Chiron Mimotopes (Clayton, Australia). Synthetic peptides with the HLA-A2 and HLA-A3 binding motif HCV NS3 1,073 to 1,081, NS3 1,406 to 1,415, NS3 1,391 to 1,399, and NS5 2,588 to 2,596 and HBV core 18 to 27, envelope were also purchased from Chiron Mimotopes. Isolation of Peripheral Blood Mononuchar CeZh and Lichen-Infiltrating Lymphocytes, Antigen-SpecijFc Polyclonal T-cell Lines Production and In Vitro Expansion of HBV and HC%Specz$c Cytotoxic T Cefi. Peripheral blood mononuclear cells (PBMC) were isolated from fresh heparinized blood and resuspended in WMI containing 10% human serum (complete medium). Excess biopsy tissue not needed for diagnostic purposes was washed in RPMI to remove contaminant blood and digested with collagenase (I mg/ml) and DNAase (25 pg/ml) (Sigma, St.Louis, MO) for 1 hour at 37 C. This method yielded from 50,000 to 300,000 lichen-infiltrating lymphocytes (LIL) depending on the different patients. Antigen-specific polyclonal T-cell lines were generated by PBMC (2 X 105/well) or LIL (at least 2,000 LILlwell cocultured with autologous irradiated PBMC 1 X 105/well) stimulation with individual recombinant HBV or HCV antigens at 1 p,g/ml, as previously described." For CTLexpansion, PBMC (2 X 106/mL) or LIL (in the presence of autologous irradiated PBMC) were resuspended in complete medium and stimulated with peptides used at 1pmollL final concentration. Recombinant interleukin 2 (IL-2; 50 U/mL) was added on day 4 of culture. Proliferation Assay. T-cell lines were extensively washed, and 2 to 5 X lo4 cells/well were incubated for 3 days with 1 X lo5 irradiated (4,000 rads) autologous APC in complete medium, in the presence or absence of HBV or HCV antigens at 1 pg/ml. 3H-TdR (0.5 pci/well; Amersham, United Kingdom) was added 8 hours before harvesting. The results are expressed as the mean counts per minute of triplicate determinations. The stimulation index was calculated as the ratio between mean counts per minute obtained in the presence and absence of antigen. Peptide-HLA Class I Tetramers. Tetrameric peptide-hla class I complexes were purchased from Proimmune LTD (Oxford, United Kingdom). Tetramers contained the HLA-M or HLA-A3 restricted HCV peptides: HLA A2-NS3 1,073 to 1,081 (CINGVCWTV), HLA A2-NS3 1,406 to 1,415 (KLVALGINAV), HLA A3-NS3 1,391 to 1,399 (LIFCHSKKK), and HLA A3- NS5 2,588 to 2,596 (RVCEKMALY) and the HLA-A2 restricted HBV peptides: core 18 to 27(FLPSDFFPSV), envelope (WLSLLVPFV). T-cell &?$ace Markers Determination, Tetramer Staining, and FACSAnalysis. PBMC, LIL, or antigenspecific T-cell lines were stained with the following monoclonal antibodies (MoAb) conjugated with different Auorochromes (FITC, PE, PerCP, or APC): anti-cd3, anti-cd8, anti-cd4, anti-cd69 (Becton Dickinson, San Jose, CA), anti-cd45ra (Sigma Chemical Co), anti- TCR: Vp 3a, 3.1, 5.1, 5.3, 6.7, 8, 12.1, 13, Va 2, 12.1 (Endogen, Woburn, MA); VP I, 2,5.2,7,9, 11, 13.6, 14, 16, 17, 18, 20, 21, 22, 23, Va 24 (Immunotech, Marseille, France). For tetramer staining, 5 X lo5 PBMC or 5 X lo4 LIL were incubated for 30 minutes at 37 C with 1 p,l of PE-labeled tetrameric complex. Cells were washed and incubated at 4 C for 30 minutes with saturating concentrations of anti-cd8 and anti-cd69. Analysis was performed on a 4-color multiparametric flow cytometer (FACS-Calibur, Becton Dickinson) using the CELLQuest (Becton Dickinson) software.

3 ~~ ~ 1448 PILL1 ET AL. HEPATOLOGY, December 2002 IFN-y and IL-4 Producing T-cell Detection by FLOW Cytometry. Intracellular cytokine assessment was performed on antigen-specific CD4' or CD8+ cells using flow cytometry as previously described, with slight modifications.12 Briefly, antigen-specific polyclonal T-cell lines were resuspended at a concentration of 1 X 106/mL in complete medium and cocultured with autologous adherent cells as APC (previously incubated overnight with HCV peptides or antigens at 1 pg/ml) for 5 to 8 hours in a humidified 5% C02 incubator at 37 C. Brefeldin-A (10 pg/ml; Sigma) was added 1 hour after beginning of coculture. As a control, the same numbers of effector cells were incubated with autologous APC either without addition of HCV antigens or peptides or in the presence of an irrelevant antigen or peptide. After incubation, cell suspensions were stained with APC-labeled anti-cd4 or anti-cd8 MoAb, fixed, permeabilized (Fix&Perm, Caltag Laboratories), and stained with PE-labeled anti-human IFN-.)I or FITC-labeled anti-human IL-4 (Sigma). Intracellular cytokine expression was analyzed on the flow cytometer. Intracelluhr HCV RNA in Oral Mucosa. The genomic and negative strand HCV RNA titers were determined by semiquantitative RT-PCR as described. l3 RT was performed using the Tth reverse transcriptase at 70 C in the presence of the forward primer (negative strand detection) or the backward primer (genomic strand detection) on different dilutions of RNA, starting from 100 ng of total lichen-derived RNA. The second primer was added for the first round of PCR. Semiquantitation was achieved by performing a nested RT-PCR to the endpoint on 2- to 4-fold dilutions (in 10 pglml of Escherichia coli trna; Sigma) of an initial amount of 100 ng of total RNA. Titers were expressed as the last dilution giving a visible band of the appropriate size on a 1.6% agarose gel stained by ethidium bromide. Intracellular genomic and negative strand HCV RNA titers were normalized to an arbitrary p-actin messenger RNA (mrna) titer of 1,024, as measured on the same specimen.'* The intracellular p-actin titer was expressed as the last 4-fold dilution giving a visible band by ethidium bromide staining of RT-PCR products obtained from 100 ng of total biopsy RNA and run on a 1.6% agarose gel. Results HCV Antigen Recognition by Lichen-InJLtrating and Peripheral Blood HCV-Specifc CD4' T CeLls in Lichen Patients. To assess the presence of HCV-specific CD4 cells within lichen lesions and to compare their functional features with those of circulating HCV-specific CD4' T cells, PBMC and oral LIL were stimulated Table 1. Clinical and Virological Characteristics of Anti-HCV-Positive Patients Patient Quantitative Age Genomic No. HCV-RNA GenotVpe (yr) Sex HLA-A2/A3 StrandTiter c 60 F Neg ND c 59 F A lb 62 F Neg ND lb 69 F A lb 76 F A F A c 73 F A2 ND Abbreviation: ND, not determined. with HCV proteins to generate antigen-specific T-cell lines. Lichen specimens from 7 patients with chronic HCV infection (Table 1) were processed, and isolated lymphomononuclear cells were stained with anti-cd3, CD4, CD8, CD69, and CD45RA monoclonal antibodies for phenotypic characterization of the infiltrating T- cell population. An increased percentage of CD69' T cells and a reduced frequency of CD45RA+ T cells compared with the peripheral blood were detected among lichen infiltrates (61.2% i 6.5% vs. 3.7% f 2.5% and 8.7%? 7% vs. 49.7%? 13%, respectively), confirming the isolation of activated cells actually deriving from lichen lesions (data not shown). PBMCs from patients No. 1, 2, 3, and 4 were stimulated with recombinant structural and nonstructural HCV antigens to generate antigen-specific T-cell lines. HCV-specific polyclonal T-cell lines were produced from the oral mucosa in all these patients; the response was focused on a single antigen in each patient (Table 2). Even starting from a higher number of T cells, antigen-specific T-cell lines could be generated from the peripheral blood of only 2 of the 4 patients (Table 2), suggesting a higher frequency of virusspecific CD4' T cells within the inflamed oral mucosa. These results further demonstrate that HCV-specific T cells isolated from oral tissue are not derived from contaminated blood. Fine Speczjicity, Cytokine Pattern, and TCR- Vj3 Expression of Lichen- and Peripheral Blood-Derived PoLyclonal T-CeLl Lines. To analyze the fine specificity of the T-cell response to core, and NS3, antigens, we used 2 panels of synthetic peptides. The first panel was composed of 2Omer peptides overlapping by 10 residues and covering the whole core sequence; the second set comprised 8 peptides corresponding to highly conserved regions of NS3. Fine-specificity analysis of core and NS3- specific T-cell lines was performed by assessing the ability of synthetic peptides to restimulate HCV-specific T cells. Peripheral blood and lichen-derived core-specific T-cell lines recognized the same amino acidic sequences core 3 1 to 60 in patient 1 and core 101 to 120 in patient 4. The

4 ~ ~~~~~~~~~~~ HEPATOLOGY, Vol. 36, No. 6, 2002 PILL1 ET AL Table 2. T-cell Response to HCV Antigens of LIL and PBMC-Derived T-cell Lines HCV Antigen Patlent 1 Patient 2 Patient 3 Patlent 4 Lichen El E2 CORE NS3 NS4 NS5 Blood El E2 CORE NS3 NS4 NS5 *Results are expressed as stimulation index. fine specificity of the T-cell lines was confirmed by intracellular staining with anti-ifn-y monoclonal antibodies after peptide stimulation and flow cytometry analysis. In the same experiments, TCR V-/3 chain usage was also assessed. Interestingly, even in presence of the same fine specificity, TCR-VP usage was different in the 2 compartments. The NS3-specific T-cell line derived from the lichen lesion of patient 2 was specific for the previously described immunodominant epitope NS3 1,253 to 1,272.'5 Core- and NS3-specific T-cell lines from lichen and peripheral blood did not produce IL-4 upon antigen stimulation, indicating a Thl cytokine pattern (Table 3). Frequency and Phenotypic Characterization of HCVSpeczf?c CD8' T Celt%. Freshly isolated licheninfiltrating lymphocytes and PBMC from patients No. 2, 4, 5, 6, and 7 were stained with HLA-A2 and HLA-A3 tetrameric complexes containing the HLA-A2-restricted HCVepitopesNS3 1,073 to 1,081, NS3 1,406 to 1,415, and HLA-A3 (patient 2) restricted epitopes NS3 1,391 to 1,399, NS5 2,588 to 2,596. Lymphomononuclear cells were also stained with anti-cd8, CD3, CD69, monoclonal antibodies. HCV-specific CD8+ T cells were detected in the lichen lesions of 3 of the 5 patients analyzed (Fig. I). Ex vivo frequency of tetrf/cd8+ T cells was higher in the lichen lesions compared with the peripheral blood (not shown). Furthermore, a substantially higher percentage of tetr+/cd8+ T cells from the lichen lesions was positive for the activation marker CD69' compared with peripheral blood HCV-specific CD8 cells (Fig. 2). However, in vitro T-cell stimulation with the specific peptides (performed in patients 4 and 7) showed different efficiencies of expansion of PBMC and LIL, as detected by tetramer staining. Indeed, tetr+/cd8+ T cells from lichen lesions expanded less efficienrly upon peptide stimulation in vitro than the homologous circulating counterpart (data not shown), confirming a different differentiation state of HCV-specific CD8' T cells in the 2 compartments. HCV RNA Genomic Strand Titer in the Oral Lichen Planus. To assess whether HCV replication was occurring in the oral mucosa of the patients studied, the HCV genomic strand titer was tested by a semiquantitative PCR assay'3 in 4 patients from whom a large amount of oral tissue was available from biopsy. By this approach, the titers of positive and negative HCV RNAstrands were assessed on the total RNA extracted from the lichen tissues. The positive genomic strand was detected at low titers in 3 of 4 patients; the negative strand that is a marker of ongoing HCV replication was not detectable in any of the tested samples. HBV Antigen Recognition by Lichen-Infiltrating and Circulating HBV-Spec$% CD4' and CD8' T Ceh. Three patients with chronic HBV infection and oral lichen planus were studied to define whether, also in these patients, virus-specific T cells are recruited into the oral mucosa. Lichen-derived lymphornononuclear cells and PBMC from 2 of the 3 patients (patient A and B) were stimulated with HBV core (HBc), e (HBe), and surface (HBs) antigens to induce HLA class I1 restricted, HBV-specific polyclonal T-cell lines. LIL and PBMC from patient C were tested ex vivo and after in vitro peptide stimulation with HLA-A2 tetrameric complexes containing the HLA-A2 restricted HBV epitopes core 18 to 27 and envelope Table 3. Fine Specificity, Cytokine Pattern and TCR-VP Usage of HLA Class Il-Restricted HCV-Specific T-cell lines Patient 1 Patient 2 Patlent 3 Patient 4 LiL PBMC LI 1 PBMC LIL PBMC LIL PBMC Protein Core Core NS3 - NS4 - Core Core Peptide(s) ND Cytokine pattern Th 1 Th 1 Thl - ND - Th 1 Th 1 Vp usage ; 13a 22 - ND - Vp8 neg* 8 Abbreviation: ND, not determined. *The LIL-derived, core-specific T-cell line did not stain with any of the available anti-vp monoclonals, including Vp8, that stained the core-specific T cells derived from the peripheral blood.

5 1450 PILL1 ET AL. HEPATOLOGY. December 2002 Patient 4 Upon peptide stimulation in vitro Tetramer NS pe.. 1m-y PIlC Tetramer NS5 I I TetramerNS pE 2588% = Tetramer NS IFN-y wc I Patient 7 I Ex-vivo : Upon peptide stimulation in vitro 1 8 Tetramer NS i Tetramer NS A IFN-y mc./ TetramerNS R j TetramerNS R IFN-y mc Fig. 1. x vivo and in vitro analysis of lichen HCV-specific CD8 cells. The frequency of intralesional HCV-specific CD8 cells ex vivo and after 10 days of peptide stimulation was analyzed by tetramer staining. IFN--y secretion was studied by intracellular cytokine staining after peptide stimulation of in vitro expanded T-cell lines. Percentage of tetramer and IFN--y positive CD8 cells is reported in the upper right quadrant of each panel. HBcAg and HBsAg-specific T-cell lines were derived from the peripheral blood of patient B, whereas no HBVspecific T cell lines were generated from lichen-derived T cells (Table 4). Analysis of HBV-specific CD8 cells showed low frequency of core 18 to 27 specific CD8' T cells in the peripheral blood of patient C that could be efficiently expanded in vitro, whereas no HBV-specific CD8+ T cells were detectable among lichen infiltrating T cells ex vivo and after peptide stimulation (Table 4). Discussion Association of HCV infection with different extrahepatic manifestations has been described, including mixed cryoglobulinaemia, membranoproliferative glomerulonephritis, lymphoma, autoimmune thyroiditis, aplastic anemia, porphyria cutanea tarda, sicca syndrome, and lichen planus. The pathogenetic role of HCV in the development of these extrahepatic manifestations is undefined. Mixed cryoglobulinemia has been proposed to be the result of a polyclonal activation of B cells triggered by the engagement of CD8 1, a tetraspanin molecule expressed on various cell types including B lymphocytes, with the HCV envelope protein E2I; however, this hypothesis still awaits definitive demonstration. Recently, ongoing HCV replication was demonstrated in mucosal tissue from anti-hcv positive patients with oral lichen planus by semiquantitative strand-specific RT-PCR8 and in situ hybridization,' suggesting the possibility of a direct or immune-mediated role of HCV in lichen planus pathogenesis. In this study, we tested the hypothesis that HCV-specific T cells are present in the oral mucosa, at the site of the lichen lesions, and that the HCV-specific T-cell response plays a role in lichen planus pathogenesis. Recruitment of HCV-specific CD4+ andlor CD8' T cells was demonstrated in the lichen tissue of 5 out of 7 patients with chronic HCV infection. CD4+ polyclonal T-cell lines were generated more efficiently from lichen-infiltrating lymphomononuclear cells than from PBMC of the same patients, suggesting a higher frequency of HCV-specific T

6 ~~~~ ~ ~~ HEPATOLOGY, Vol. 36, No. 6, 2002 PILL1 ET AL Lichen Peripheral blood PI 2 Pt 4 n Pt I + z 0,2 E - t I a, 0,I * 0 I n?: 60 E E 8 40 Fig. 2. Comparison of the ex vivo frequency of HCV-specific CD8/tetr cells from LIL and PSMC and frequency of CD8/tetr+ T cells expressing the activation marker CD69. *Below the limit of detection. 0 NS3 NS5 NS3 NS NS3 NS cells in the oral compartment. Fine-specificity analysis of the core-specific T-cell lines derived from the 2 compartments led to the identification of the same epitopes; however, T-cell clones present in the oral mucosa showed a different TCR-VP chain usage than those circulating in the peripheral blood, suggesting a specific compartmentalization at the site of the lichen lesions. Also, HCVspecific cytotoxic T cells were present with higher frequency in the lichen tissue compared with the circulating compartment, as shown by HCV tetramer analysis, and produced IFN-7 upon peptide stimulation. In view of the prior demonstration of HCV genomic strands in the lichen lesions, including replicative inter- Table 4. (A) Proliferative T-cell Response to HBV Antigens of Lichen- and Peripheral Blood-Derived T-cell Lines and (B) Frequency of CDB/Tetramer+ Core and Envelope Within 111 and PBMC Ex Vivo and After Peptlde Stimulation A. HBV Antigen Patient A Patient B LIL HBsAg - - HBcAg - - HBeAg - - PBMC HBsAg 3.6* - HBcAg HBeAg - - B. After Peptlde Patient C HBV Tetramer Ex Vivo Stimulation Ll L Core <0.005 (0.005 Env <0.005 <0.005 PBMC Core Env <0.005 (0.005 *Results are expressed as stimulation index mediates, our results suggest that HCV-specific T cells are attracted into the lesional tissue by local antigen expression. This interpretation is confirmed by our molecular analysis of lichen tissue showing the presence of HCV RNA positive strands at the site of cellular damage and lymphocyte infiltration. Negative replicative intermediates were previously detected at low titers,'6 consistent with low levels of virus replication. The fact that negative strands were not detected in our study does not exclude virus replication but is in line with a low local level of replication at the limit of the sensitivity of our methods of detection. An alternative interpretation of the presence of HCVspecific T cells among lichen-infiltrating lymphocytes is that they are recruited into the lesions as a result of inflammation simply because they are in an activated state, irrespective of HCV replication and viral antigen expression. This possibility is unlikely because it does not explain the enrichment in HCV-specific T cells within the lichen infiltrate in comparison with the circulating blood. To assess more directly this possibility, we asked whether HBV-specific T cells were present and enriched among lichen-infiltrating lymphocytes of patients with chronic HBV infection and lichen planus. If HBV-specific T cells are present in the circulation of these patients and inflammation can induce their recruitment into the lichen tissue by a chemokine-mediated unspecific mechanism, HBVspecific T cells should be detectable within the lichen tissue as in patients with HCV-associated lichen planus. Even when detectable in the peripheral blood, CD4' and CD8+ HBV-specific T cells could not be derived from

7 1452 PILL1 ET AL. HEPATOLOGY, December 2002 the oral mucosa of 3 patients with oral lichen planus and chronic HBV infection. In conclusion, we have provided evidence that HCVspecific CD4+ and CD8+ cells are present within intralesional infiltrates in oral lichen planus associated with chronic HCV infection. These cells are specifically recruited into the oral lesions because their frequency in this compartment is higher than in the peripheral blood and because HBV-specific T cells from patients with lichen planus associated with chronic HBV infection are undetectable within lichen infiltrates, although present and detectable in the peripheral blood. In contrast to circulating HCV-specific T cells that are in great majority CD69-, intralesional HCV-specific T cells (CD69') show the phenotype of recently activated T cells, are able to produce IFN-y but display poor expansion capability, that are all distinctive features of terminally differentiated effector cells. Although final and direct proof cannot be provided by studies in human natural diseases, our results strongly suggest that CD4' and CD8' HCV-specific T cells play a role in the pathogenesis of epithelial cell damage in oral lichen planus and provide the first strong evidence for the involvement of HCV in the pathogenesis of an extrahepatic manifestation during HCV infection. Oral cell damage may be the result ofa direct immune aggression of epithelial cells expressing HCV antigens or may be sustained by a cytokine environment favorable to trigger and maintain autoimmune reactions. Aknowledpeizt: The authors thank Patrizia Latorre at the Division of Gastroenterology and Hepatology, University Hospital Geneva, for her help in the experiments of HCV-RNA oral tissue expression analysis. References 1. I'ileri P, Uemawu I., Campagnoli S, Galli G, Falugi F, Petracca R, WeinerAJ, et al. Binding of Hepatitis C Virus to CD81. Science 1998;282: Gruppo Italian0 Studi Epidemiologici in Dermatologia (GISED) Lichen planus and liver disease: a multicentre case-control study. Br Med J 300: Nagao Y, Sata M, Tanikawa K, Itoh K, Kameyama T. Lichen planus and hepatitis C virus in the northern Kyushu region ofjapan. Eur J Clin Invest 1995;25: Carrozzo M, Gandolfo S, Carbone M, Colombatto P, Broccoletti R, Garzino-Demo P, Ghisetti V. Hepatitis C virus infection in Italian patients with oral lichen planus: a prospective case-control study. J Oral Pathol Med 1996;25: Mignogna MD, Lo Muzio L, Favia G, Mignogna RE, Carbone R, Bucci E. Oral lichen planus and HCV infection: a clinical evaluation of 263 cases. Int J Dermatol 1998;37: Del Olmo JA, Pascual I, Bagan JV, Serra MA, Escudero A, Rodriguez F, Rodrigo JM. Prevalence of hepatitis C virus in patients with lichen planus of the oral cavity and chronic liver disease. Eur J Oral Sci 2000;108: Bagan JV, Ramon C, Gonzales L, Diago M, Milian MA, Cors R, Lloria E, et al. Preliminary investigation of the association of oral lichen planus and hepatitis C. Oral Surg Oral Med Oral Pathol Oral Radio1 Endod 1998;85: Nagao Y, Kanieyama T, Sata M. Hepatitis C virus RNA detection in oral lichen planus tissue. Am J Gastroenterol 1998;93: Arrieta JJ, Rodriguez-Inigo E, Casqueiro M, Bartolome J, Manzarbeitia F, Herrero M, Pardo M, et al. Detection of hepatitis C virus replication by in situ hybridization in epithelial cells of anti-hepatitis C virus-positive patients with and without oral lichen planus. H~~~~0~0~~2000;32: Ferrari C, Valli A, Galati L, Penna A, Scaccaglia P, Giuberti T, Schianchi C, et al. T-cell response to structural and non-structural hepatitis C virus antigens in persistent and self-limited hepatitis C virus infection. HEPA- TOLOGY 1994;1!): Lamonaca V, Missale G, Urbani S, Pilli M, Boni C, Mori C, Sette A, et al. Conserved hepatitis C virus sequences are highly immunogenic for CD4( +) T cells: implications for vaccine development. HEPATCXOCY 1999;30: Jung T, Schauer U, Heusser C, Neumann C, Rieger C. Detection of intracellular cytokines by flow cytometry. J Immunol Methods 1993; 159: Negro F, Krawczynski K, Quadri R, Rubbia-Brandt L, Mondelli M, Zarski JI'. Hagengue A. Detection of genomic- and minus-strand of hepatitis C virus RNA in the liver of chronic hepatitis C patients by strand-specific semi-quantitative RT-PCR. HEPA~OLOGY 1999;29: Nakajima-lijinxi S, Hamada H, Reddy P, Kakunaga T. Molecular structure of the human cytoplasmic beta-actin gene: interspecies homology of sequences in the introns. Proc Natl Acad Sci U S A 1985;82: Diepolder HM, Gerlach JT, Zachoval R, Hoffmann RM, Jung MC, Wierenga EA, Scholz S, et a]. Immunodominant CD4' T-cell epitope within nonstructural protein 3 in acute hepatitis C virus infection. J Virol 1997;71 :GO Carrozzo M, Quadri R, Pentenero M, Gandolfo S, Negro F. Molecular evidence that the hepatitis C virus replicates in the oral mucosa. J Hepatol 2002;37;