[Adsorption, separation, and purification of cyclosporine A using reversed-phase liquid chromatography].

High performance liquid chromatography (HPLC) is widely used in the separation and analysis of cyclosporine A (CsA). Analyzing the chromatographic behavior of CsA is key to the purification of CsA by preparative HPLC. In this study, the retention behavior of CsA on the C18 column using mobile phases of methanol-water and acetonitrile-water was compared. The retention time of CsA was sensitive to the change in the ratio of the organic solvent. When 84%-88% methanol or 75%-85% acetonitrile was used, the retention factor () was in the range of 3-7. The change in the peak shape of CsA was investigated with loading amounts of 5, 25, 50, 100, and 500 mg. With an increase in sample loading, the peak shape of CsA in both mobile phases changed from symmetric to tailing, and the retention time reduced. Therefore, it is necessary to focus on the removal of impurities that were eluted before CsA during the purification. In addition, the peak shapes of CsA in methanol-water and acetonitrile-water were similar in the tested concentration range. This indicates that it was not possible to tune the peak shape of CsA by changing the organic solvent. Adsorption isotherms were obtained to describe the retention behavior of CsA. When the mass concentration of CsA in the mobile phase was low, the effect of the organic solvent ratio on the adsorption capacity of CsA on the C18 stationary phase was not distinct. With an increase in the solute mass concentration above 0.5 g/L, the reduced proportion of organic solvent helped improve the adsorption capacity of CsA. When the mass concentration of CsA in the mobile phase reached 5 g/L, the adsorption capacities were 24.9 g/L in 88% methanol and 40.8 g/L in 84% methanol. The adsorption capacity of CsA in acetonitrile-water was higher than that in methanol-water. When the mass concentration of CsA was 5 g/L, the adsorption capacity increased to 46.4 g/L in 75% acetonitrile. Scatchard analysis showed that the slope of the adsorption isotherm decreased gradually, which was consistent with trend observed in the Langmuir adsorption isotherm for the shape of the Langmuir peak (i.e. trailing peak). When the mass concentration of CsA in the mobile phase was between 0.01 g/L and 0.03 g/L, the slope of the curve decreased significantly, and the peak shape of CsA rapidly tailed with increasing loading amount. However, when using a mobile phase with a lower proportion of organic solvent (84% methanol or 75% acetonitrile), this trend was weakened. The adsorption data of CsA were fitted to models. The Langmuir model was found to be suitable for the methanol-water mobile phase, and the Moreau model for the acetonitrile-water mobile phase. The model parameters indicated that the monolayer adsorption of CsA occurred on the C18 stationary phase in both mobile phases, the difference being that more intermolecular interactions between CsA occurred in the acetonitrile-water mobile phase, resulting in a higher adsorption capacity. In methanol-water, the intermolecular interactions between CsA were inhibited by methanol due to its role as a proton donor. As an aprotic solvent, acetonitrile could only weakly inhibit these interactions; hence, the interactions could be improved by increasing the acetonitrile proportion. As the proportion of acetonitrile changed from 85% to 75%, the saturated adsorption capacity increased from 123 g/L to 197 g/L, while the interaction constant decreased from 0.618 to 0.588. Finally, CsA was purified using the conditions of 0-60 min 65%-75% acetonitrile, 60-80 min 75% acetonitrile, by which the impurity could be controlled to below 0.2%. The results of this study will aid in the purification of CsA by preparative HPLC.