Proteomic evaluation of cell preparation methods in primary hepatocyte cell culture

In vitro liver preparations are being used increasingly to study various aspects of chemical hepatotoxicity and thus have become powerful alternatives to in vivo toxicologic models. Primary hepatocyte culture systems are especially useful in screening cytotoxic and genotoxic compounds and assessing biochemical lesions associated with chemical exposure. We have begun to use this approach in combination with proteomic analysis to construct a molecular "toxicoproteomic" test system for a broad range of relevant and potentially toxic chemicals. Using a highly parallel two-dimensional electrophoretic (2-DE) protein separation system to analyze cells from culture systems, we previously observed significant variations in protein expression that were unrelated to chemical exposure. We hypothesized these artifactual protein alterations were the result of the variations in the culture conditions or cell manipulations, or both. Therefore, we conducted a study to assess the expression of hepatocyte proteins cultured on 6-well plates and recovered for analysis either by scraping/pelleting or direct in-well solubilization. Following incubation of 1.2 x 10(6) hepatocytes in six-well plate, recovery and solubilization of the cells and 2-DE of the solubilized lysates of 100 000 cells, we detected 1388 proteins in the in-well solubilized samples compared to 899 proteins in the washed/scraped/pelleted cell samples, a loss of 35%. Based on protein identification by peptide mass fingerprinting, the subcellular location of nearly all of the proteins whose abundance decreased were cytosolic and those few that increased were either microsomal, mitochondrial, or cytoskeletal proteins. These results emphasize the variation introduced by cell-handling during recovery of hepatocytes from culture plates and may explain at least some of the artifactual differences observed in earlier in vitro experiments.