Inversely-correlated inhibition of human 5-lipoxygenase activity by bay X1005 and other quinoline derivatives in intact cells and a cell-free system—implications for the function of 5-lipoxygenase activating protein

Abstract

A series of quinoline derivatives were analysed for the influence on leukotriene synthesis as a parameter for 5-LOX (EC 1.13.11.34) activity in a cell-free system of the 10,000 g supernatant of human PMNL (polymorphonuclear leukocytes). The ratios of the ic 50 values for leukotriene synthesis inhibition in this cell-free system and in A23187-stimulated intact PMNL ranged from 1–1100. Consequently, plotting of the two values resulted in a random distribution ( r = −0.281, N = 18) suggesting that no relationship between the inhibition of leukotriene synthesis in the cell-free system and in intact cells exists. At first sight this finding was not surprising since we have shown earlier that in intact cells this class of quinoline derivatives shares the same mode of action as MK-886, i.e. an indirect inhibition of 5-LOX activity by binding to FLAP. However, we found that the potency of these compounds in intact cells is strongly influenced by the K value (partition coefficient) which is a parameter for the ability of a substance to accumulate in a lipid (membrane) phase compared to the water phase. Therefore, the ic 50 values for leukotriene synthesis inhibition in intact PMNL were corrected for the corresponding K value of the compounds and the resulting values again plotted against the ic 50 values for inhibition of leukotriene synthesis in the cell-free system. As a result, a significant correlation ( r = −0.878, N = 18) was obtained. In order to simplify this relationship the influence of the partition coefficient was eliminated by comparing compounds with about the same K value ( K = 7243 ± 1646, N = 7). As a result, the ic 50 values for inhibition of leukotriene synthesis in the 10,000 g supernatant fraction (indicative for the affinity of the compounds to 5-LOX) and in intact cells (indicative for the affinity of the compounds to FLAP) were highly, but inversely correlated ( r = −0.992). That means that a compound with a high affinity to 5-LOX will have a low affinity to FLAP and vice versa. We hypothesized that this pharmacologically obtained relationship could be indicative of a physiologically occurring equivalent. We therefore propose a model in which FLAP binds arachidonic acid as its physiological substrate with low affinity and allows 5-LOX to get access to its substrate (assuming a higher affinity of 5-LOX to arachidonic acid) after 5-LOX translocation from the cytosol to the membrane. In support of this model we provide evidence that arachidonic acid and other cis-unsaturated fatty acids, but neither a trans-unsaturated nor a saturated fatty acid, inhibit BAY X1005 binding to FLAP in intact human PMNL.