Sulforhodamine B (SRB), an aminoxanthene dye, has been used as an assay for total cell protein, initially developed as an endpoint assay for in vitro screening of antitumour agents. In this paper it was investigated as a possible endpoint for a cytotoxicity assay using CHO cells. It is a robust assay with a stable colorimetric endpoint, capable of semi-automation using microtitre equipment. At optimum concentrations (0.05-0.1% SRB) the assay is linear with respect to cell number over a range of 5 x 10(3) to 10(5) cells. In comparative studies with the neutral red assay the SRB assay was more sensitive, and in cytotoxicity assays with test compounds gave comparable dose-response curves. The cytotoxicity of five divalent metal chlorides was assessed using the SRB assay. The order of toxicity was Cd > Hg > Zn > Mn > Mg, that is similar to the expected in vivo ranking. 16 compounds with reported oral LD(50) (rat) ranging from 25,800 mg/kg (glucose) to 1 mg/kg (mercuric chloride) were tested in the assay. The relative toxicities of the compounds in the in vitro SRB assay were similar to the relative in vivo toxicities. The exceptions could be explained by the chemistry of the compounds and could be attributed to pharmacokinetic properties or mechanism of action. This assay can therefore be used to rank chemically similar compounds but is unsuitable as a precise predictor of in vivo toxicity.
