Glutamine synthetase heterogeneous expression as a marker for the cellular lineage of preneoplastic and neoplastic liver populations.

The distribution of glutamine synthetase (GS) in rat liver and in putative preneoplastic and neoplastic lesions was studied at stages of hepatocarcinogenesis induced by diethylnitrosamine (DEN), 2-acetylaminofluorene (AAF) and aflatoxin B1 (AFB1) using immunohistochemistry. In control rats, GS was localized entirely to rings 1-3 cells deep surrounding the central veins. Exposure to DEN or AAF resulted in a reduction of these GS-positive (GS+) zones by 96 and 61% respectively, due to necrosis of the cells in this compartment, while AFB1 had little effect (less than 14%). At later times GS+ foci, adenomas and carcinomas developed in animals exposed to DEN or AAF, but not AFB1, corresponding to the acute effects. Small GS+ foci also identified by other phenotypic abnormalities (e.g. gamma-glutamyltransferase) were exclusively associated with the regenerating GS+ zone as were a few collections of GS+ hepatocytes, the nature of which was uncertain. Although accounting for a very small fraction (less than 3%) of total foci, these GS+ foci or GS+ hepatocytes apparently gave rise to a substantial fraction of adenomas that were GS+ (DEN: 43% and AAF: 33%) and an even higher percentage of GS+ carcinomas (DEN: 59% and AAF: 39%). No GS+ neoplasms were induced by AFB1. Statistical evaluation of these data strongly suggests that GS+ neoplasms originate through initiation of hepatocytes which possess this particular phenotype. Thus, GS might serve as a specific marker for tracing cell lineage relationships during hepatocarcinogensis. The GS+ phenotype which confers glutamine independence may also provide a growth advantage.