Immunohistochemical determination of hepatic cytochrome P-4502E1 in formalin-fixed, paraffin-embedded sections

Cytochrome P-4502E1 (2E1) is inducible by chronic ethanol consumption that results in enhanced activation of anesthetics and commonly used drugs (such as acetaminophen) to hepatotoxins. Therefore, assessment of hepatic 2E1 is needed in prescribing these drugs for the management of alcoholic patients. Currently, measurement of 2E1 requires either immunohistochemistry on frozen sections or Western blot (WB) analysis of homogenized tissue in excess of that needed for pathology. To obtain a more widely applicable method, we developed a procedure to detect 2E1 by immunohistochemistry in formalin-fixed, paraffin-embedded liver biopsies obtained routinely for diagnosis. Data were collected from rats fed ethanol-containing or control liquid diets for 3 weeks. Immunostaining was performed using anti-human rabbit 2E1 antibody as the primary antibody, and the immunoreaction was detected by the avidin-biotin immunoperoxidase method after treating sections with target unmasking fluid, an antigen retrieval buffer that enhanced the staining of 2E1. In control rats, 2E1 staining was weak and perivenular. After ethanol feeding, it showed a lobular gradient, strongest perivenular and weakest periportal, similar to that seen in frozen sections. The staining intensity was scored as: 0 (no staining) to 3 (strong staining). The zonal staining was scored as follows: 1 = perivenular zonal staining, 2 = midzonal, and 3 = panlobular. With the product of the two scores, a significant difference was found between alcohol-fed and control rats (5.1 +/- 0.3 vs. 0.8 +/- 0.2, p < 0.001). 2E1 assessments by WB were also significantly different for these rat pairs (68.5. +/- 2.1 vs. 7.9 +/- 0.8 arbitrary units/mg protein, p < 0.001), with a parallel increase of immunostaining scores and WB measurement of 2E1 content. This immunohistochemical method was then validated in 14 paraffin-embedded percutaneous human liver biopsy samples. In livers of nonalcoholics, 2E1 staining was seen in the perivenular zone only, whereas in samples of alcoholics, the staining was perivenular to midzonal and sometimes periportal. A significant correlation between the zonal staining scores (r(s) = 0.67, p < 0.005) or intensity x zonal staining scores (r(s) = 0.79, p < 0.001) and WB analysis was found. The immunohistochemical assessments of 2E1 expression in formalin-fixed, paraffin-embedded sections from livers of alcoholics was found to correlate with WB analysis, and lobular distribution was consistent with that seen in frozen sections. The proposed method should therefore be useful for the assessment of 2E1 content in paraffin-embedded liver samples, thereby aiding in the management of heavy drinkers.