To sharpen the search for new lipoxygenase inhibitors, we designed a screen to probe for both potency and selectivity. The assay utilized 12-human (12-HLO), 15-human (15-HLO), and 15-soybean (15-SLO) lipoxygenases. The IC(50) value data obtained provided new insights about structure-activity relationships (SAR) for redox and nonredox inhibitors. All of the compounds tested were isolated from sponges and consisted of a novel terpenoid, hyrtenone A (1), and 12 known terpenoids. Potent compounds were defined as those having IC(50) values < 1 microM, and selectivity was assessed from the three possible IC(50) value ratios. One of the four terpenoid redox inhibitors studied, puupehenone (2), was equivalent to or better in potency than the well-known redox inhibitor nordihydroguarierate acid (NDGA, 14). However, none of the terpene redox inhibitors exhibited a selectivity ratio on a par with that of 14. Several potent nonredox inhibitors were identified, and one, dimethoxypuupehenol (5), exhibited notable selectivity. The structural elucidation of 1 and the SAR results for 13 natural products are reported. This study suggests that sponge-derived terpenes are a promising source for new lipoxygenase inhibitors.

Yeast damage-associated response protein (Dap1p) and mouse progesterone receptor membrane component-1 protein (mPGRMC1p) belong to a highly conserved class of putative membrane-associated progesterone binding proteins (MAPR), with Dap1p and inner zone antigen (IZA), the rat homologue of mPGRMC1p, recently being reported to bind heme. While primary structure analysis reveals similarities to the cytochrome b(5) motif, neither of the two axial histidines responsible for ligation to the heme is present in any of the MAPR proteins. In this paper, EPR, MCD, CD, UV-vis, and general biochemical methods have been used to characterize the nature of heme binding in both Dap1p and a His-tagged, membrane anchor-truncated mPGRMC1p. As isolated, Dap1p is a tetramer which can be converted to a dimer upon addition of 150 mM salt. The heme is noncovalently attached, with a maximal, in vitro, heme loading of approximately 30%, for both proteins. CD and fluorescence spectroscopies indicate a well-ordered structure, suggesting the low level of heme loading is probably not due to improperly folded protein. EPR confirmed a five-coordinate, high-spin, ferric resting state for both proteins, indicating one axial amino acid ligand, in contrast to the six-coordinate, low-spin, ferric state of cytochrome b(5). The MCD spectrum confirmed this conclusion for Dap1p and indicated the axial ligand is most likely a tyrosine and not a histidine, or a cysteine; however, an aspartic acid residue could not be conclusively ruled out. Potential axial ligands, which are conserved in all MAPRs, were mutated (Y78F, D118A, and Y138F) and purified to homogeneity. The Y78F and D118A mutants were found to bind heme; however, Y138F did not. This result is consistent with the MCD data and indicates that Tyr138 is most likely the axial ligand to the heme in Dap1p.

Lipoxygenase (LO) is a biological target for many diseases such as asthma, atherosclerosis, and cancer. Our labs have synthesized and investigated nordihydroguaiaretic acid (NDGA) derivatives and have established that the reductive inhibition of soybean and 15-human LO can be affected by the strength of the electron-withdrawing substituents on the phenyl rings of NDGA. In addition, we have determined that hydrophobic NDGA derivatives activate 15-HLO, suggesting a hydrophobic allosteric site.