Chromatographic retention assisted in-silico prediction of anticancer potential of glucocorticoids on cancer cell lines.

Abstract

Glucocorticoids (GCs) are hallmarks of anti-inflammatory activity. They are used as adjuvant therapy in oncology medications to alleviate some of the associated side effects. Although recent research has indicated that GCs have favorable anticancer potential, some scientific evidence suggests a pro-proliferation impact of GCs on cancer cells. This may create a scientific confusion on the utility of GCs and the choice of which GC enhances the anticancer potential and mitigates the negative effect. Accurate in-silico prediction and correct ranking of biological activity may be limited by the impact of the physicochemical interaction between small molecules and biological membranes. Chromatographic retention is inherently dependent on the physicochemical properties of the test molecule. It can scale the relative hydrophobicity and tentatively indicate the membrane permeability. In this study, the relationship between the in-silico binding affinity, the chromatographic retention and the in-vitro anticancer activity was investigated. Fifteen GCs were chromatographically separated on an Inertsil® C18 (4.6 * 250 mm; 5 μm) HPLC column. The binding affinity of the test GCs was determined on three receptors involved in cancer cell proliferation (topoisomerase II (TOPII), glucocorticoid receptor (GR) and ATP-binding cassette (ABCG2)). The antiproliferative potential of the test steroids was determined as per their IC₅₀. The correlation between chromatographic retention and the binding affinity to the observed IC₅₀ was investigated by multiple linear regression (MLR). Results revealed that some GCs exhibited a remarkably favorable inhibitory potential against cancer cell lines over normal cell lines. Our data indicated a significant correlation between the retention times of different GCs and the determined binding affinity, especially to the GR (r² = 0.677; p = 0.011) and the estrone sulphate (ESS) binding site of the ABCG2 (r² = 0.643; p = 0.018). Concurrently, the retention times were well correlated to the observed IC₅₀ on the colorectal cancer cell lines (r² = 0.580; p = 0.038) and the hypopharyngeal cell lines (r² = 0.638; p = 0.019). Significant MLR models (n = 4) correlating the retention times of the tested steroids and the binding affinity to the observed IC₅₀ were created. The MLR model correlating the retention times and the ABCG2_ESS binding affinity to the IC₅₀ on lung cancer was the most significant (p = 0.004). The accuracy of the model was 107.12 ± 29.18 % indicating good IC₅₀ prediction abilities. In conclusion, chromatographic retention can be employed as a low-cost and simple auxiliary tool for improving the in-silico prediction of the in-vitro activities of small molecules.