Limitations of the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-2H-tetrazolium bromide (MTT) assay when compared to three commonly used cell enumeration assays.

Background: The tetrazolium-based MTT assay has long been regarded as the gold standard of cytotoxicity assays as it is highly sensitive and has been miniaturised for use as a high-throughput screening assay. However, various reports refer to interference by different test compounds, including the glycolysis inhibitor 3-bromopyruvate, with the conversion of the dye to coloured formazan crystals. This study assessed the linear range and reproducibility of three commonly used cell enumeration assays; the neutral red uptake (NRU), resazurin reduction (RES) and sulforhodamine B (SRB) assays, in comparison to the MTT assay. Interference between the MTT assay and three glycolysis inhibitors, 2-deoxyglucose, 3-bromopyruvate and lonidamine, was investigated.

Results: Data indicate that the NRU, RES and SRB assays showed the smallest variability across the linear range, while the largest variation was observed for the MTT assay. This implies that these assays would more accurately detect small changes in cell number than the MTT assay. The SRB assay provided the most reproducible results as indicated by the coefficient of determination after a limited number of experiments. The SRB assay also produced the lowest variance in the derived 50% inhibitory concentration (IC50), while IC50 concentrations of 3-bromopyruvate could not be detected using either the MTT or RES assays after 24 hours incubation. Interference in the MTT assay was observed for all three tested glycolysis inhibitors in a cell-free environment. No interferences were observed for the NRU, SRB or RES assays.

Conclusions: This study demonstrated that the MTT assay was not the best assay in a number of parameters that must be considered when a cell enumeration assay is selected: the MTT assay was less accurate in detecting changes in cell number as indicated by the variation observed in the linear range, had the highest variation when the IC50 concentrations of the glycolysis inhibitors were determined, and interference between the MTT assay and all the glycolysis inhibitors tested were observed. The SRB assay performed best overall considering all of the parameters, suggesting that it is the most suitable assay for use in preclinical screening of novel therapeutic compounds with oxido-reductive potential.