The expression of proliferating cell nuclear antigen (PCNA) was examined in normal human and rat liver fixed in either formaldehyde or methanol, and was compared with the incorporation of bromo‐deoxyuridine (BrdU) in S‐phase cells. Codistribution of PCNA and BrdU was assessed in rat liver by double immunohistochemical staining using PC10 and anti‐BrdU monoclonal antibodies to identify labelled nuclei of parenchymal and sinusoidal cells. In formaldehyde‐fixed human biopsies (n = 13) PCNA‐labelling index (PCNA LI) was 0.43 ± 0.24 (mean ± SEM) for hepatocytes and 0.09 ± 0.03 for sinusoidal cells. A great inter‐specimen variability was observed and a preferential lobular distribution was not evident. In methanol‐fixed human liver (n = 8) the immunostaining was strong. PCNA LI was 0.05 ± 0.01) for hepatocytes and 0.14 ± 0.01 for sinusoidal cells. 75 of labelled hepatocytes and 60 of labelled sinusoidal cells were found in acinar zone 1. In formaldehyde‐fixed rat liver (n = 10) a weak nuclear staining and a great interspecimen variability were evident. LI was 0.13 ± 0.07) for hepatocytes and 0.40 ± 0.21 for sinusoidal cells without preferential acinar distribution. In methanol‐fixed rat liver (n = 10), PCNA LI was 0.14 ± 0.02 for hepatocytes and 0.40 ± 0.04 for sinusoidal cells. 64 of labelled hepatocytes and 50 of labelled sinusoidal cells were found in zone 1. Only on methanol‐fixed material did double immunohistochemistry show an almost complete overlap of BrdU and PCNA labelling. The PCNA LIs and the zonal distribution of labelled nuclei as obtained in methanol‐fixed material are in keeping with previous reports using3H‐thymidine (3H‐Thy) incorporation, suggesting that PCNA immunostaining represents a valid alternative to3H‐Thy. In addition, the present data support the hypothesis that S‐phase associated PCNA is more selectively r
展开▼
