Proteinase 3 (PR3), also referred to as myeloblastin, is a 26kD serine protease which is overexpressed in precursor cells of chronic myeloid leukaemia (CML). PR3 has previously been identified as the autoantigen of Wegener’s granulomatosis, which is characterized by a self-destructive humoral and T-cell response against PR3. However, high avidity PR3-specific cytotoxic T-cells (CTL) from CML patients specifically lysed CML blasts and their presence in vivo correlated with the achievement of a complete cytogenetic remission under interferon alpha (IFNα) therapy. Thus, overcoming tolerance to PR3 contributed to the elimination of leukaemia without causing auto-immunity (Molldrem et al., Nat. Med. 2000). We have previously shown that the induction of PR3 by IFNα in CD14+ monocytes may contribute to elicit a PR3-specific CTL expansion in CML. Since PR3-upregulation has subsequently been reported in monocytes of HIV- and pneumonia patients as well as in hepatitis C liver specimens, we asked, whether the upregulation of PR3 under inflammatory conditions may be linked to the expansion of PR3-specific T-cells. Therefore, we studied the frequency of PR3 specific CD8 positive CTL in patients with de novo diagnosed hepatitis C, a disease known to be associated with various autoimmune phenomenona. Thirty four HLA-A2.1 positive patients were analyzed by means of staining peripheral blood mononuclear cells using PR3-specific tetramers. 17 patients (50%) had PR3 specific CTL either at diagnosis (n=13) or developed them during follow up (n=4). Of the 13 patients, who had PR3-tetramer positive T-cells at diagnosis, six patients had both high and low aviditiy PR3-CTL, whereas seven patients had only low avidity PR3-specific T-cells. The frequency of these CD8 positive CTL at diagnosis ranged between 0.16% to 0.53% of all T-cells for the high avidity population and between 0.12 to 1.58% for the low avidity population. Upon antiviral therapy with IFNα and Ribaverin, the frequency of PR3-CTL became undetectable within less than 5 month of initiation of therapy in five of six patients of the high avidity PR3-CTL group. Only one patient of the patients with low avidity PR3-CTL at diagnosis was available for follow up testing and became negative. Interestingly, in four patients we saw a transient increase of low avidity PR3 specific CTL after initiation of therapy. Immunohistology of four hepatitis C liver sections and four healthy control specimens revealed a strong infiltration of PR3-overexpressing mononuclear cells in hepatitis liver, but PR3-negativity in normal livers. Hepatocytes did not express PR3. In conclusion: Hepatitis C is significantly associated with the induction of PR3-specific CTL and overexpression of PR3 in mononuclear cells may be important to drive their induction. PR3-specific CTL may be responsible for autoimmune phenomena in hepatitis C such as vasculitis and myelosuppression.