Factors influencing tumour uptake and retention of benzonaphthyridine derivatives in mice

Abstract

The DNA-binding benzonaphthyridine derivatives have demonstrated to have high and even curative activity against the Colon-38 murine adenocarcinoma. A homologous series of derivatives with, hydrogen, methyl, ethyl, propyl and butyl groups attached to the 2- position of the chromophore, was used to investigate the role of physicochemical properties on the tissue pharmacokinetics and its contribution to anti-tumour activity. The methyl derivative, SN 28049, N-[2-(dimethylamino)ethyl]-2,6-dimethyl-1-oxo-1,2- dihydrobenzo[b]-1,6-naphthyridine-4-carboxamide, was the most active of the series. A rapid and sensitive LC-MS/MS method with a run time of 7 min for the simultaneous measurement of the five analogues in a pooled sample was developed. The method was validated in mouse plasma and tissues and had a limit of quantitation of 0.001 μM and was linear over the range, 0.001 – 0.3 μM in all matrices with acceptable intra- and interassay precision and accuracy. Plasma protein binding and metabolic stability was assessed by equilibrium dialysis. Plasma and tissue pharmacokinetics were assessed in mice after administration of each analogue (25 μmol/kg) to healthy, as well as mice with Colon-38, Lewis Lung and melanoma tumours. Antiproliferative activity (IC50) was assessed cultures by 3H-thymidine incorporation method. Intracellular concentrations were determined following exposures in vitro. Susceptibility to multi-drug resistance was performed in leukaemia cultures over-expressing P-gp and MRP1. Plasma pharmacokinetics conformed to a model where increasing lipophilicity was associated with a decreasing area under the concentration-time curve (AUC) and an increasing clearance and volume of distribution. In contrast, tumour tissue pharmacokinetics showed a very different relationship where the AUC of the methyl derivative (2334 μM.h) was 89-fold higher than the hydrogen derivative (26.3 μM.h) with other homologues having intermediate values. The tumour AUC correlated with the in vivo anti-tumour activity of the series. The methyl derivative had a 24 min in vitro microsomal half-life, while other analogues ranged from 1.6 - 12.2 min. The plasma free fraction decreased (17 - 5 %) significantly with lipophilicity. SN 28049’s pharmacokinetics were evaluated in mice with LLTC, NZM4, NZM10 and NZM52 tumours. The tumour AUCs were 2.7 to 36-fold greater than the plasma, but higher AUCs were observed in the kidney. Overall, the tumour exposures were significantly lower than the Colon-38 tumour. Cellular uptake and retention varied widely with different cell lines and in general showed a parabolic dependence on lipophilicity, but not clearly related to antiproliferative activity. The analogues were only marginally susceptible to MRP1, 2.3-fold (methyl, ethyl and propyl) and P-gp (2-fold, hydrogen). The plasma pharmacokinetics of this series are related to changes in lipophilicity. However, the tumour pharmacokinetics reveal a strong dependence on the nitrogen substituent on the benzonaphthyridine chromophore, with the methyl group providing by far the best retention in tumour.