Breaking self-tolerance by virus infection in the CYP2D6 mouse model
(1) Transgenic CYP2D6 mice expressing the human Cytochrome P450 2D6 isoform (CYP2D6) as self-components in the liver are tolerant or ignorant to the (self)-transgene and thus neither destroy the hepatocytes nor develop hepatitis. (2) Induction of AIH in such mice is initiated by infection with an Adenovirus that expressing human CYP2D6 (Ad-2D6) that induces liver-inflammation and attracts leukocytes to the site of virus infection. This process is mediated by predetermined programs of the innate immune system in response to Adenovirus-infection. (3) Once arrived at the target site Ad-2D6-specific precursor lymphocytes expand and rapidly eliminate the virus. (4) After virus elimination activated CYP2D6-specific lymphocytes attack the only remaining target antigen, namely the human CYP2D6-antigen transgenically expressed in the liver. This results in destruction of hepatocytes, liver fibrosis and cirrhosis. Hence, an initially unspecific inflammation in the target organ proceeds to antigen-specific autoimmunity.

Persistent liver damage after infection of mice by Ad-2D6
Left panel: Paraformaldehyde-fixed livers of CYP2D6 mice infected with either Ad-2D6 or Ad-GFP at week 8 post-infection. Right panel: Hepatic necrosis and massive cellular infiltrations after Ad-2D6-infection. H & E staining of liver sections obtained from CYP2D6 mice infected with either Ad-2D6 or Ad-GFP at week 1, 4 and 8 post-infection. Note that persistent hepatic damage (massive cellular infiltrations and significant necrosis) is only present in Ad-2D6-infected mice.

The etiology of autoimmune hepatitis (AIH) is poorly understood although the major autoantigen, cytochrome P450 2D6 (CYP2D6), has been identified and immunodominant epitopes mapped. Therefore, we generated an animal model for human AIH using the natural autoantigen CYP2D6. We infected transgenic mice expressing human CYP2D6 in the liver (CYP-2D6 mice) with an Adenovirus-CYP2D6 vector (Ad-2D6) to break self-tolerance. Surprisingly, upon infection with Ad-2D6 not only transgenic CYP2D6 mice but also wildtype FVB mice showed several persistent features characteristic for liver damage associated with AIH. These features included massive hepatic fibrosis, ‘fused’ liver lobules, disorganized architecture, cellular infiltrations, elevated serum aminotransferase levels, and focal to confluent necrosis. Further, all Ad-2D6-infected mice (CYP2D6 and FVB) generated high titers of anti-CYP2D6 antibodies. Epitope mapping revealed that such anti-CYP2D6 antibodies predominantly recognized the core peptide sequence AQPPRD (CYP2D6 aa265-270), which is the immunodominant linear epitope recognized by LKM-1 antibodies from AIH patients. In contrast, mice infected with a control Adenovirus expressing green fluorescence protein (Ad-GFP) did neither develop liver damage nor generated anti-CYP2D6 antibodies. Interestingly the kinetics and severity of liver damage as well as antibody formation was enhanced in wildtype FVB mice compared to transgenic CYP2D6 mice. Our data indicate that the autoimmune liver damage was reduced and delayed in transgenic CYP2D6 mice due to a certain degree of tolerance towards human CYP2D6 compared to wildtype FVBs, which do not express the human version of the 2D6 isoenzyme. In wildtype FVB mice, due to the homology of the mouse isoenzymes of the CYP superfamily to human CYP2D6, autoimmune liver damage by Ad-2D6-infection was possibly induced via true ‘molecular mimicry’.