The presence of a specific acetyl-11-keto-β-boswellic acid (AKBBA)-binding site on 5-lipoxygenase (5-LO) that is distinct from the arachidonate substrate-binding site was determined using photoaffinity labeling. A novel radio-iodinated photosensitive analog of AKBBA, azido-125I-KBBA (4-azido-5-125iodo-salicyloyl-β-alanyl-11-keto-β-boswellic acid), was used to specifically label human 5-LO. azido-125I-KBBA inhibits 5-LO as efficiently as AKBBA but contains a 4-azido-5-125iodo-salicyloyl-β-alanyl moiety at the C3 position instead of an acetoxy group (Sailer et al., 1998).
Photolabeling experiments were performed using protein fractions from supernatants of human leukocytes that contained approximately 5% of 5-LO. Two to three proteins, one of them 5-LO, has been labeled by azido-125I-KBBA. When certain variables of the system were changed, the following was observed:
- Heat-denatured protein pools did not support photo-incorporation into 5-LO, however, labeling by azido-125I-KBBA increased in other proteins present
- Incubation of protein pools with increasing concentrations of AKBBA or arachidonic acid prior to photolabeling selectively prevented the incorporation of azido-125I-KBBA into 5-LO
- The presence of calcium influenced labeling. When calcium was omitted, almost no azido-125I-KBBA was incorporated into 5-LO, whereas other proteins remained unchanged
- Addition of other pentacyclic triterpenic compounds such as amyrin, β-boswellic acid, and 18α-glycyrrhetinic acid completely prevented azido-125I-KBBA from labeling 5-LO, which suggests competition for the same binding site (Sailer et al., 1998)
Blockage of the substrate-binding site of 5-LO with a competitive-type synthetic inhibitor, prior to the addition of arachidonic acid and azido-125I-KBBA, did not modulate the inhibitory effect of arachidonic acid on labeling by azido-125I-KBBA. These results suggest that the calcium-mediated pentacyclic-triterpene binding site characterized using the AKBBA analog is distinct from the substrate-binding site (Sailer et al., 1998). In fact, it may be identical with a second arachidonic binding site postulated previously (Sailer et al., 1998; Aharony and Stein,1986).
Among the several compounds classified as leukotriene synthesis inhibitors, nonredox inhibitors, such as boswellic acids, are preferred. Unlike redox type inhibitors, they do not interact with other biological redox systems, lessening the likelihood of side effects like methemoglobin formation (Wrez et al., 1998; Lau et al., 1992). AKBBA has been identified as the only leukotriene synthesis inhibitor so far that inhibits 5-LO activity by noncompetitive, nonredox mechanisms (Sailer et al., 1998). Because of the success with boswellic acids in treating chronic inflammatory conditions, several synthetic nonredox type inhibitors, such as L-739,010 and ZM-230,487, have been prepared. Unlike AKBBA, they are competitive type inhibitors (Werz et al., 1998).
The 5-LO inhibitors were tested in intact cells and homogenates (broken cell preparations) of human granulocytes or HL-60 cells. Iron ligand inhibitor (BWA4C) furnished an inhibitory concentration value 3-fold lower in intact cells than in broken cell preparations of HL-60 cells [IC50= 0.05 µM (intact cells) and 0.15 µM (homogenates)].
Similar results were obtained for AKBBA [IC50 = 15 µM (intact cells) and 50 µM (homogenates)]. Synthetic compounds ZM-230,487 and L-739,010 inhibited 5-LO activity in intact cells in the nanomolar range IC50 = 50 nM and 20 nM, respectively. However, in cell homogenates or preparations of the purified enzyme, the concentrations necessary to inhibit 5-LO activity were up to 150-fold higher (Werz et al., 1998).
The inhibitory potency of the synthetic nonredox inhibitors in cell homogenates was increased by the addition of thiols, glutathione (GSH) and dithiothreitol (DTT), such that 5-LO inhibition was comparable to that of intact cells. Millimolar concentrations of thiols act as co-substrates for peroxidases that are involved in the regulation of cellular 5-LO activity. The thiols were successful in reversing the loss of 5-LO inhibition in cell homogenates of the synthetic nonredox inhibitors. However, the IC50 values for AKBBA and BWA4C did not change with the addition of thiols. The 5-LO activity of ZM-230,487 and AKBBA in the presence and absence of thiols are shown (Werz et al., 1998).
Dose response curves of AKBA in granulocyte homogenates in the presence and absence of thiols
Dose response curves of ZM-230,487 in granulocyte homogenates and intact cells in the presence and absence of thiols
Further experiments revealed that low hydroperoxide concentrations were necessary for efficient 5-LO inhibition by ZM-230,487. Physiological conditions such as inflammation that are characterized by oxidative stress and increased peroxide levels reduce the 5-LO inhibitory activity of nonredox inhibitors ZM-230,487 and L-739,010 and may explain their reported lack of activity in chronic inflammatory processes (Werz et al.,1998).