AIM We found in experiment that the relation of cell number-OD in MTS cytotoxicity assay is non-linearity. In this paper we probe into the relations of cell number-OD and drug dose-inhibition rate. METHODS Cells HL-60 and Raji were used in this experiment. Assay on the relation of cell number-OD: Cells of different concentrations were seeded into 96 well plates. OD was read at 490 nm after incubating with MTS for 1～5 hour. The regression curves were estimated from the data with the software SPSS 11.0. Cytotoxicity assay: Cells were seeded into 96 well plate and incubated with drug of different doses for 72 hours. OD was read at 490 nm after incubating with MTS for 3 hour. The regression curves were estimated from the data with the software SPSS 11.0. RESULTS The cubic curves fit well the relations of cell number-OD of the 2 strains of cells at 5 time points and the coefficient R2 of these cubic curves is 0.997～1.000, greater than the R2 0.938～0.993 of linear model. For the relations of dose-inhibition rate, the cubic curves also have best fitness and their R2 is 0.998～1.000 greater than the R2 0.948～0.987 of linear model of logarithm dose-probit. CONCLUSION For the relations of cell number-OD in MTS cytotoxicity assay, the fitness of cubic curves are better than the linear model. A more accurate IC 50 would be obtained by the cubic curve equation than by probit regression that is in common use.

0.05). The generation of ROS increased obviously after the cells were incubated with them for 12 h, and the increase of ROS reached the peak treated for 24 h. The levels of ?? m were time-dependently decreased after treating with four compounds for 12, 24, and 48 h. Conclusion The grouth of both MDR KBv200 cells and parental drug-sensitive KB cells were inhibited to the treatment of four anthraquinone derivative in vitro. The mechanism of their effects is associated with the increase of the cellular ROS level and the decrease of ?? m.