This is the accepted version of the journal article: Golovanov, Alexey; Zhuravlev, Alexander; Cruz, Alejandro; [et al.]. «N- Substituted 5-(1H-Indol-2-yl)-2-methoxyanilines Are Allosteric Inhibitors of the Linoleate Oxygenase Activity of Selected Mammalian ALOX15 Or- thologs : Mechanism of Action». Journal of Medicinal Chemistry, Vol. 65, Issue 3 (January 2022), p. 1979-1995. American Chemical Society. DOI 10.1021/acs.jmedchem.1c01563 This version is available at https://ddd.uab.cat/record/263124 under the terms of the license Page 1 of 28 Journal of Medicinal Chemistry 1 2 3 4 5 6 7 N-Substituted 5-(1H-Indol-2-yl)-2-methoxyanilines are Allosteric 8 9 Inhibitors of the Linoleate Oxygenase Activity of Selected 10 Mammalian ALOX15 Orthologs: Mechanism of Action 11 12 Alexey Golovanov†#, Alexander Zhuravlev†#, Alejandro Cruz‡, Vladislav Aksenov†, Rania 13 Shafiullina†, Kumar R. Kakularam§, José M. Lluch‡,ǁ, Hartmut Kuhn§, Àngels González-Lafont‡,ǁ 14 15 and Igor Ivanov†* 16 †Lomonosov Institute of Fine Chemical Technologies, MIREA - Russian Technological University, Vernadskogo Pr. 17 86, 119571 Moscow, Russia 18 ‡Departament de Química, Universitat Autònoma de Barcelona, 08193 Bellaterra, Barcelona, Spain 19 §Departmentof Biochemistry, Charite - University Medicine Berlin, Corporate member of Free University Berlin, 20 21 Humboldt University Berlin and Berlin Institute of Health Charitéplatz 1, D-10117 Berlin, Germany 22 ǁInstitut de Biotecnologia i de Biomedicina (IBB), Universitat Autònoma de Barcelona, 08193 Bellaterra, Barcelona, 23 Spain 24 25 26 ABSTRACT: Here we describe the first systematic study on the mechanism of substrate selective inhibition of mammalian 27 ALOX15 orthologs. For this purpose we prepared a series of N-substituted 5-(1H-indol-2-yl)anilines and found that (N-(5- 28 (1H-indol-2-yl)-2-methoxyphenyl)sulfamoyl)carbamates and their monofluorinated analogues are potent and selective 29 inhibitors of the linoleate oxygenase activity of rabbit and human ALOX15. Introduction of a 2-methoxyaniline moiety into 30 the core pharmacophore plays a crucial role for substrate selective inhibition of ALOX15-catalyzed oxygenation of linoleic 31 acid at submicromolar concentrations without affecting arachidonic acid oxygenation. Steady state kinetics, mutagenesis 32 studies and molecular dynamics simulations suggested an allosteric mechanism of action. Using a dimer model of ALOX15 33 our MD-simulations suggest that binding of the inhibitor at the active site of one monomer induces conformational 34 alterations in the other monomer so that the formation of a productive enzyme-linoleic acid complex is energetically 35 compromised. 36 37 38 KEYWORDS: 5-(1H-indol-2-yl)-2-methoxyanilines, lipoxygenase inhibitors, allosterism, protein–protein interactions, 39 molecular dynamics 40 41 42 the other hand, reports suggesting anti-inflammatory 43 INTRODUCTION activities of ALOX15 orthologs challenged the strategy of 44 Lipoxygenases form a family of lipid peroxidizing enzymes developing ALOX15 inhibitors for the treatment of human 45 that have been implicated in cell differentiation,1-3 in the diseases.19-21 46 pathogenesis of inflammatory and anaphylactic diseases,4- 47 8 in atherogenesis,9-10 in ferroptosis,11-12 in insulin resistance The putative pathophysiological roles of ALOX15 made this 48 and in adipose tissue inflammation, which is a key element enzyme a target for pharmacological research. A number 49 of obesity.13-15 In humans there are six functional LOX genes ALOX15 inhibitors that have been developed in the past 50 (ALOX15, ALOX15B, ALOX12, ALOX12B, ALOX5, ALOXE3) involve nitrogen containing heterocycles. Among them are 51 and their expression leads to the formation of six highly potent oxadiazole or oxazole derivatives,22 pyrazole- 52 functionally distinct ALOX-isoforms. One of these based sulfonamides and sulfamides,23 imidazol-based enzymes, ALOX15, is somewhat peculiar because of its sulfonamide analogs,24 indole-based inhibitors25-28 as well 53 capability of oxygenating complex lipid-protein assemblies as tryptamine sulfonamides29. Most of these compounds 54 such as biomembranes and lipoproteins16 and the inhibit ALOX15-catalyzed oxygenation of linoleic acid (LA) 55 emerging role of this isoenzyme in human pathologies has with a similar efficiency as arachidonic acid conversion.24, 56 29 However, for some compounds their inhibitory potency recently been reviewed17. In inflammation ALOX15 exhibits 57 depends on the fatty acid substrate29 and this substrate dual functionality18. In various inflammation models 58 specific inhibition may be of biological relevance. In fact, ALOX15 products induced pro-inflammatory effects. On 59 60 ACS Paragon Plus Environment Journal of Medicinal Chemistry Page 2 of 28 substrate specific ALOX inhibitors may impact the (Figure 1A). The imidazole based N-(2-(5-(4- 1 formation of deleterious linoleic acid derived mediators methoxyphenyl)-2-(pyridin-2-yl)-1H-imidazol-4-yl)ethyl)- 2 without impacting the biosynthesis of beneficial signalling 4-pentylbenzenesulfonamide (Figure 1B)24 exhibited a 3 molecules formed from other polyenoic fatty acids. When similar degree of substrate specific inhibition for rabbit 4 we tested isobuthyl (N-(5-(5,6-difluoro-1H-indol-2-yl)-2- ALOX15. Interestingly, 3-substitution at the indole ring 5 methoxyphenyl)sulfamoyl)carbamate26 as inhibitor for (Figure 1C) reduced the degree of substrate selectivity of 6 purified human ALOX15 we observed selective inhibition the inhibitor against rabbit ALOX15 and the absence of a 7 of the linoleate oxygenase activity of the enzyme with an methoxyaryl moiety completely abolished substrate 8 IC50(LA)/IC50(AA) ratio of 0.03 (SI Figure S1). The specific inhibition.29 Based on this data one might 9 observed difference could not be explained by the conclude that a 2-methoxyaniline moiety attached to 10 conventional one ligand-binding model,30 in which position C2 of the core pharmacophore may play a crucial 11 inhibitor and substrates compete for binding at the role for the degree of substrate specific inhibition of the 12 catalytic centre of the monomeric enzyme. Since rabbit linoleate oxygenase activity of ALOX15. When we carried 13 ALOX15 may occur as non-covalent dimer in aqueous out docking studies of the sulfamoylcarbamate depicted in 14 solution31-32, such weak intermolecular association may Figure 1A into the putative substrate binding pocket of 15 contribute to the allosteric character.33-34 monomer A of the rabbit ALOX15 dimer (PDB entry 2P0M) 16 we found that the inhibitor in its best docking pose is To explore the molecular basis for substrate-specific 17 located in the hydrophobic substrate binding pocket of inhibition of the linoleate-oxygenase activity of ALOX15 by monomer A with its 5-(1H-indol-2-yl)-2-methoxyaniline 18 isobuthyl (N-(5-(5,6-difluoro-1H-indol-2-yl)-2- motif ahead (SI Figure S2). The carbamic acid ester group 19 methoxyphenyl)sulfamoyl)carbamate26 we first performed was located close to the entrance of the substrate binding 20 in silico docking studies to fit the compound into the pocket of monomer A and partly interconnected the two 21 substrate binding pocket of rabbit ALOX15. Since no monomers within the enzyme dimer. The NH2 groups of 22 crystal data is currently available for human ALOX15 most Arg599 and Arg403 interact electrostatically with the 23 of our mechanistic experiments were carried out with oxygen atom of the methoxy group and the NH2 group of 24 rabbit ALOX15. After the initial docking studies we Gln596 forms a hydrogen bond with that same oxygen 25 selected (N-(3-(1H-indol-2-yl)phenyl)sulfamoyl)carbamic atom. Finally, it is worth mentioning that one of the two 26 acid as core pharmacophore, prepared a set of modified fluorine atoms presents unfavourable electrostatic 27 analogues and tested their activities against rabbit, human interactions with Glu357 (SI Figure S2).This binding mode 28 and mouse recombinant mammalian ALOX15 orthologs. suggests that the chemical nature of residues I, II and III 29 Considering our previous finding that docking of (Figure 1A) might be of critical relevance for the allosteric 30 arachidonic acid (AA) into the active site of conformer A properties of the inhibitor. On this basis three lines of 31 of rabbit ALOX15 resulted in an unstable enzyme-substrate modifications of the inhibitor have been suggested: i) Since 32 complex (unpublished data), we used a new dimer model OCH3-group of the methoxyaniline moiety may limit the for rabbit ALOX15, in which the substrate binding pocket 33 mobility of this ligand inside the catalytic cavity we of conformer A is occupied by the inhibitor, whereas 34 replaced this residue by a small hydrogen atom. ii) The conformer B involved the substrate fatty acid at its catalytic 35 binding mode of the inhibitor we identified during our centre. Finally, we performed site-directed mutagenesis 36 docking studies offers sufficient space around the tail of studies and MD simulations to explore the consequences 37 carbamic acid ester moiety for chemical modification of of inhibitor binding on substrate alignment at the active 38 substituent II. Thus, we reasoned that incorporation of a site of the other monomer. 39 long chain aliphatic ester group at this position might 40 RESULTS increase the lipohylicity of this residue and, hence, might 41 improve the binding affinity of Targeted drug design and structure activity relation 42 43 (SAR). In aqueous solutions rabbit ALOX15 exhibits a conformational heterogenity35 and is present in a dynamic 44 equilibrium between monomeric and dimeric fractions at 45 physiological salt concentrations31. Although the allosteric 46 character of ALOX15 was first reported more than a decade 47 ago33, the precise mechanism of allosteric regulation of the 48 enzyme remains elusive. Thus, it was impossible to create 49 powerful allosteric inhibitors employing the conventional 50 methods of rational drug design. On the other hand, a 51 literature search for ALOX15 inhibitors revealed that only 52 for a limited number of compounds the IC50(LA)/IC50(AA) 53 ratios has been determined in the past23, 29. For the present 54 study we quantified the IC50 values for isobuthyl (N-(5- 55 (5,6-difluoro-1H-indol-2-yl)-2- 56 methoxyphenyl)sulfamoyl)carbamate for linoleic acid and Figure 1. Selected structures of the substrate specific and non- 57 arachidonic acid oxygenation by the pure recombinant specific ALOX15 inhibitors. 58 human ALOX15 (SI Figure S1) and obtained a ratio of 0.03 59 60 ACS Paragon Plus Environment Page 3 of 28 Journal of Medicinal Chemistry Scheme 1. Synthetic Route of compounds 6-8a 1 2 O R2 NO2 NH2 3 i iii path B 4 1 3a,b 2a (R2=H) 5 O O 2b (R2=OCH3) R2 6 R1 R1 7 path A NH2 8 N NH2 H ii N N H 9 3a (R1=H) 3b (R1=F) iv 4a R1= H, R2=H 4a-c 10 4b R1= H, R2=OMe 4c R2= F, R2=OMe 11 R1 12 N R2 v 13 R1 H 5a-c NH2 v 14 R O N O 15 H HN S N O v R1 H 16 6a-c O N R2 O O 17 R1 H HN S N O 7a-e H 18 5a-8a R1= H, R2= H N R2 O O O 5b-8b R1= H, R2= OMe 19 5c-8c R1= F, R2= OMe H HN S N O H 20 8a-e O 21 aReagents and conditions: (i) 10% Pd/C, HCOONH4, EtOH, reflux; (ii) PPA, 130 oC, 0.5h; (iii) AcOH, EtOH, reflux; (iv) ZnCl2,180 22 oC, 2h; (v) ClSO2NCO, HN(C2H5)2, isobuthyl-, allyl - or octylalcohol, rt. 23 24 25 the inhibitor at the hydrophobic environment of the unfortunately, the overall yields of the target 2-arylindoles 26 catalytic centre. Moreover, introduction of an allylic ester 5a-c did not exceed 10%. Similar results were obtained 27 moiety may force additional π- interactions between the when the synthesis was performed according to a two-step 28 inhibitor and aromatic active site amino acids and thus, procedure via isolation of the intermediate hydrazones 4a- 29 may strengthen the protein-inhibitor interactionn. iii) c (Scheme 1, Path B, for analytical data see supporting 30 Since fluorination of the indole ring may not only enhance information). The low yields (5-10%) at the cyclization step the metabolic stability of the pharmacophore towards of the corresponding hydrazone can be explained by the 31 oxidation, but also might be important for the alignment formation of a large number of side-products due to a 32 of the inhibitor at the enzyme’s active site, we also relatively low reaction rate (2-5 h) and the high 33 prepared non fluorinated and mono C5-fluorinated temperature (180°C). In contrast, in polyphosphoric acid 34 inhibitor derivatives. (PPA)37 all targeted 2-arylindoles were obtained at 35 relatively high yields of 50-60% within 15-30 minutes at 36 Chemical synthesis of substituted N-(3-(1H-indol-2- rather moderate temperatures (120°C-130°C) (Scheme 1, 37 yl)phenyl)sulfamoylcarbamates. To test the validity of Path A). The physical data of compounds 5a and 5b were 38 our in silico predictions we synthesized a set of nine 2- identical to those reported in the literature.27, 38 The 39 arylindolderivatives involving the following precursors: 3- resulting indoles 5a-c were subsequently transformed to 40 (1H-indol-2-yl)aniline (5a), 5-(1H-indol-2-yl)-2- the derivatives 6-8 by the reaction with 41 methoxyaniline (5b) and its mono halogenated derivative chlorosulfoisocyanate and the corresponding alcohol in 42 5c. Previously reported methods for preparation of the presence of diethylamine.26 As expected, in 13C-NMR 43 substituted 3-(1H-indol-2-yl)anilines employ the canonic spectra of the fluorinated compounds 6c-8c the signal of 44 Bischler-Möllau indol cyclization reaction26-27. Here were C5 (158 ppm) was split into a doublet with J = 232 Hz. In 45 report that 5a-c may be prepared via the Fischer indole contrast, the signals of C4 and C6 at 105 and 110 ppm were 46 synthesis using commercially available starting split with lower coupling constants (J = 23.5 and 26 Hz, 47 compounds (Scheme 1) in a one-pot procedure under respectively; SI Figure S3). 48 solvent-free conditions. For the synthesis we used Inhibitory potency of substituted (N-(3-(1H-indol-2- 49 commercial phenylhydrazines 3, 1-(3-aminophenyl)ethan- 50 1-one (2a) or 1-(3-amino-4-methoxyphenyl)ethan-1-one yl)-phenyl)sulfamoyl)carbamates 6-8 for ALOX15. 51 (2b). The latter was prepared from the corresponding nitro During the first round of testing, we determined the IC50 derivative by reduction of NO2-group using ammonium values for nine compounds using the pure recombinant 52 formiate in the presence of 10%-Pd/C catalyst. Next, we rabbit ALOX15 as target enzyme. Inhibition was measured 53 optimized the reaction conditions for compounds 2a-b and employing the standard spectrophotometric assay 54 3a-b using different Lewis acids. In our hands, under the monitoring the rate of conjugated diene formation when 55 previously described reaction conditions (ZnCl2/180°C)36 arachidonic or linoleic acids were used as substrates. The 56 IC50 values varied between 0.04 for linoleic acid and >50 57 we observed the formation of a complex product mixture as concluded from HPLC and 1H NMR analyses and µM for 58 59 60 ACS Paragon Plus Environment Journal of Medicinal Chemistry Page 4 of 28 Table 1. Inhibitory activity of substituted 2-arylindolsulfocarbamates 6-8 against rabbit ALOX15. 1 2 R1 3 R2 O N O a R1= H, R2= H 4 H R3 b R1= H, R2= OMe HN S N O 5 H c R1= F, R2= OMe O 6 7 6, R3= -CH2-CH(CH3)2 7, R3= -CH-CH=CH2 8, R3= -(CH2)7-CH3 8 Indole LA IC50, AA IC50, IC50LA/I LA IC50, AA IC50, IC50LA/ LA IC50, AA IC50, LA/AA 9 Type µMa µMa C50AA µMa µMa IC50AA µMa µMa ratio 10 ratio ratio 11 12 a 29.97±12.20 > 50 > 0.60 15.38±10.20 > 50 > 0.30 1.55±0.34 2.79±1.53 0.55 13 b 1.45±0.27 > 50 > 0.030 0.63±0.32 43.34±13.09 0.014 0.040±0.001 2.06±0.14 0.020 14 c 0.42±0.03 20.82±2.12 0.020 0.54±0.06 23.26±3.46 0.023 0.047±0.001 1.08±0.06 0.043 15 aThe data have been obtained on the basis of dose response curves for spectrophotometric ALOX15 activity assay for purified 16 17 rabbit ALOX15 (see Experimental Section). All data represent means ± SEM of three independent (n=3) measurements. 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 Figure 2. Dose response curves for inhibition of LA- and AA-oxygenase activities of mammalian ALOX15 orthologs. The LA- and 42 AA oxygenase activities of pure recombinant rabbit (A) and pure recombinant human (B) ALOX15 orthologs were assayed in the presence or absence of compound 8b. Spectrophotometric assay were performed a described in Experimental Section and half- 43 maximal inhibitory concentrations (IC50) were determined. (C) Effect of detergent (Tween 20) on inhibitory potency of 8b on 44 rabbit ALOX15. Spectrophotometric assays were performed a described in Experimental Section. (D) Inhibition of different 45 mammalian ALOX15 orthologs by 8b using the HPLC-based activity assay. The assay was performed a described in Experimental 46 Section .The ALOX activity of the incubation sample lacking the inhibitor was set 100%. Finally, we calculated the ratios of residual 47 activities of LA oxygenation vs. AA oxygenation (formation of 13-HODE vs. formation of 15-HETE) as suitable measure for the 48 substrate selectivity of the inhibitory effect. (E) Inhibition of the intracellular human ALOX15 by 8b using L1236 Hodgkin 49 lymphoma cells. The assay was performed a described in Experimental Section. The ALOX activity of the incubation sample lacking 50 the inhibitor was set 100%. As suitable measure for the substrate selectivity of the inhibitory effect of 8b we calculated the ratios 51 of residual activities of LA oxygenation vs. AA oxygenation (formation of 13-HODE vs. formation of 15-HETE). (F) Inhibition of the 52 intracellular human ALOX15 by 5µM ETYA using L1236 Hodgkin lymphoma cells as enzyme source. The assay was performed 53 under strictly comparable conditions as described for compound 8b. All data represent means ± SEM of n=3 measurements. 54 55 arachidonic acid (Table 1, SI Figure S4). Thus, a presence of the methoxy group at C2 of the aniline moiety 56 pronounced substrate specific enzyme inhibition was (indole types b and c) is important not only for the 57 observed. Comparison of the data indicates that the inhibitory potency but also for the substrate selectivity of 58 59 60 ACS Paragon Plus Environment Page 5 of 28 Journal of Medicinal Chemistry the inhibitory effect (IC50(LA)/IC50(AA) ratio < 0.05). mixture of 15-HpETE and 12-HpETE in the ratio 85:15 1 Replacement of the butyl moiety (6b-c) with an allyl group (Figure 2E), which is consistent with the product 2 (7b-c) resulted in a moderate change in inhibitory potency specificity of purified recombinant human ALOX15.42 3 of the compounds. In contrast, introduction of a more Although somewhat higher inhibitor concentrations were 4 lipophilic octyl residue (compounds 8b-c) resulted in app. required to inhibit the cellular ALOX15 activity, 5 10-fold increase the inhibitory potency. Modification of the preferential inhibition of the LA-oxygenase activity by 8b 6 indol ring at C5 (introduction of a fluorine atom, 6c-8c) was observed (LA/AA ratios 0.14-0.18) (Figure 2E). On the 7 moderately improved the inhibitor potency against both other hand, the standard ALOX/COX inhibitor ETYA43 did 8 fatty acids, but had no effect on the substrate selectivity of not show a major degree of substrate selectivity (Figure 9 inhibition. In summary, with IC50 values of 0.04 µM 2F). Taken together, these data indicate that the substrate- 10 compounds 8b (Figure 2A) and 8c appeared to be the specific inhibitory effect of human ALOX15 by 8b can also 11 same powerful as the best ALOX15 inhibitors described so be observed for the intracellular enzyme. 12 far.29 ML351, which has previously been identified as paralog-specific ALOX15 inhibitor was with an IC50 of 200 Mechanism of inhibition. In order to explore the 13 mechanism of the substrate-specific inhibition of the 14 nM less potent.39 It should be stressed at this point that the assay conditions used in the different literature studies and linoleate oxygenase activity of ALOX15 we employed 15 compounds 6c and 8b, which exhibit low 16 in our experiments were not strictly comparable (different enzyme preparations, different substrate concentrations, IC50(LA)/LC50(AA) ratios and carried out extensive kinetic 17 measurements using pure recombinant rabbit ALOX15 use of detergents) and thus, direct comparison of the 18 (Figure 3A-D). The following kinetic constants were determined IC50-values may be misleading. However, most 19 evaluated for this enzyme without inhibitor: kcatLA=32.5±1.0 interestingly, compounds 6c, 7b-c and 8b exhibited a high 20 s-1, KMLA=24.5±5.0 µM, kcatAA=21.9.5±2.0 s-1, KMAA=13.2±2.0 degree of substrate selectivity with IC50(LA)/IC50(AA) 21 µM for LA and AA, respectively, and these data are ratios varying between 0.014 and < 0.030 (Table 1, SI 22 consistent with our previous reports42, 44. Using LA as Figure S4). A similar trend for substrate specific inhibition 23 was observed when 8b was tested with recombinant substrate both compounds (6c and 8b) induced a marked 24 human ALOX15 (Figure 2B). decrease in kcat but the KM-values remained largely 25 unaffected (SI Figures S5 and S6). Finally, we fitted these 26 Owing to their limited water solubility the activity of inhibitor data (Figure 3A,C) to different inhibition models 27 different inhibitors may depend on their aggregate state in and found that the best fits were obtained for a model of a 28 aqueous solutions.40 To exclude such physiologically non-competitive inhibition of LA oxygenation with a Ki of 29 irrelevant promiscuous effects we performed additional 21.9±0.5 nM (R2= 0.977) for 8b and Ki of 0.29±0.08 µM (R2 30 spectrophotometric inhibitor studies in PBS containing = 0.985) for 6c. For AA oxygenation 6c and 8b induced a 31 0.02% Tween 20 solubilizer (0.1 and 1µM for LA and AA, marked decrease in both, kcat and KM values (SI Figures S5 32 respectively, Figure 2C). Here we found that substrate and S6) and here the best fit of our experimental data was specific inhibition of ALOX15 was also seen in the presence obtained for an uncompetitive mode of inhibition with a Ki 33 of detergents (LA/AA ratio < 0.1). Since surface active of 0.95±0.06 µM (R2 = 0.989) for 8b and Ki of 6.53±0.49 µM 34 compounds may also impact the enzyme structure we (R2 = 0.979) for 6c (Figure 3B,D). On the other hand, 35 refrained from including detergents into the assay systems inhibition of both substrates by the compound 8a (SI 36 for the other experiments. Figure S7) could be equally well described by a mixed 37 model of inhibition (R2= 0.977) with a Ki of 0.79±0.02 µM 38 Next, we explored whether a similar substrate selective (α=6.6) and 3.13±0.40 µM (α=7.9) for LA and AA, 39 inhibition can be seen for mouse Alox15 ortholog. Since respectively. In this case the dissociation constant ratios 40 this enzyme could not be purified to homogenieity we KdEI to KdESI of o.075 and 0.153 for LA and AA, respectively, 41 employed crude enzyme preparations (bacterial lysate suggested a higher binding affinity of 8a to the enzyme (E) 42 supernatants) for comparative experiments (Figure 2D) rather than enzyme-substrate complex (ES). 43 and a assayed the catalytic activities by a HPLC based 44 system. Although enzyme inhibition was less pronounced Next, we tested the reversibility of 8b induced inhibition 45 when compared with our spectrophotometric assays a of ALOX15 catalyzed LA oxygenation. For this purpose, a 46 clear-cut substrate specific inhibition of rabbit and human solution of pure rabbit ALOX15 was pre-incubated with a 47 ALOX15 was confirmed (Figure 2D). However, for mouse 100-fold molar excess of 8b and activity assays suggested Alox15 no substrate selectivity could be detected. In fact, at complete inhibition of the enzyme. Next, we removed the 48 2 µM 8b we hardly observed any inhibition of mouse Alox15 inhibitor by gel-filtration. The procedure was repeated 49 neither with LA nor with AA as substrate. These data several times until the final ratio of maximal UV- 50 suggest a pronounced ortholog specific inhibition of absorbance at 312 nm (inhibitor specific chromophore) vs. 51 ALOX15 orthologs of different species by 8b. the absorbance of protein (280nm) remained constant 52 (Figure 3G). The amount of inhibitor left in the pre- 53 Finally, we tested whether the substrate specific inhibition incubation mixture suggested that on average one protein 54 of human ALOX15 can also be demonstrated in a cellular molecule in the pre-incubation mixture binds one 55 system. For this purpose we employed the Hodgkin molecule of 8b. However, activity assays using the 56 lymphoma cells L1236, which constitutively express human spectrophotometric system 57 ALOX15 at high levels.41 These cells oxygenate AA to a 58 59 60 ACS Paragon Plus Environment Journal of Medicinal Chemistry Page 6 of 28 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 Figure 3. Mechanism of ALOX15 inhibition. (A-F) Eadie-Hofstee plots demonstrating the effects of 6c, 8b and 8a on rabbit ALOX15 33 when incubated with 25 µM linoleic (A, C, E) or arachidonic (B, D, F) acids. Incubations were repeated at different inhibitor 34 concentrations. All data points represent means ± SEM with n=3. (G) On average 1 molecule of 8b binds to 1 molecule of rabbit 35 ALOX15. To estimate the rALOX15:8b ratio molecular extinction coefficient of 1.78 mg-1cm-1 (280 nm) for the enzyme was used. 36 37 38 indicated almost complete recovery of the catalytic activity contrast, 8a tends to compete with both substrates for the 39 and thus, a reversible mode of action was concluded. substrate binding site (Figure 3E,F). 40 Similar data were obtained with the compound 6c (data not shown). Effect of inhibitors on ALOX15 product pattern. 41 Human and rabbit ALOX15 exhibit dual reaction specificity 42 In general, sulfamoylcarbamates may undergo degradation and arachidonic acid is converted by these two enzymes to 43 in acidic aqueous solutions in the presence of Fe3+ (SI a mixture of 15- and 12-HETE (90:10 for human and 95:5 for 44 Figure S8A-D). This reaction is associated with a rabbit enzyme). Next, we explored whether 8a and 8b alter 45 reduction of Fe3+ ions to the divalent ferrous form. Since no the reaction specificity of the rabbit enzyme. We found 46 inhibitor degradation occurs during enzymatic catalysis that both inhibitors do not modify the product pattern of 47 with either substrate (AA or LA) (SI Figure S8E), one may AA oxygenation even at high inhibitor concentrations. 48 exclude substrate specific decomposition reactions of the Independent on the inhibitor concentration in the assay 49 inhibitor. Moreover, preincubation of the ALOX15-8b mixture we always quantified a 15-HETE / 12-HETE ratio of 50 complex with large amounts of organic peroxydes (13- 95.5±0.2% / 4.5±0.2% (SI Figure S9A, left columns). A 51 HpODE, 20 µM) did neither impact the inhibitory activity similar product pattern has previously reported for this 52 (data not shown). These data exclude that substrate- enzyme in the absence of any inhibitor45.However, 8b 53 specific enzyme inhibition may be related to the redox induced subtle alterations in the product specificity of LA 54 properties of the inhibitors. oxygenation (SI Figure S9B). When we applied the AA 12- 55 lipoxygenating Ile418Ala mutant of rabbit ALOX1546 for Taken together, our data indicate that compounds 8b and 56 comparative experiments (SI Figure S9A, middle columns) 6c inhibit LA and AA oxygenation by rabbit ALOX15 in a 57 we confirmed that this enzyme variant oxygenated AA reversible manner via different allosteric mechanisms. In 58 predominantly to 12-HETE. Only a minor share (25 %) of 59 60 ACS Paragon Plus Environment Page 7 of 28 Journal of Medicinal Chemistry 15-HETE formation was detected and the presence of 8a 8b(LA) and 8a(LA) were almost identical. These data 1 did not alter this product pattern. However, 8b reduced suggest that the structural integrity of the inter-monomer 2 the relative share of 12-HETE formation when compared interface within the wild-type ALOX15 dimer may play a 3 with the no inhibitor control (SI Figure S9A, middle critical role for substrate specific inhibition of ALOX15 4 columns). Here again, the inhibitor-induced alterations in orthologs by 8b (Figure 4). 5 the product specificity were rather subtle but statistically 6 significant. With human ALOX15 as catalyst we observed 7 that both 8a and 8b reduced the relative share of the major 8 oxygenation product (15S-HETE) when compared with no- 9 inhibitor controls (SI Figure S9A, right columns). The 10 differences observed between rabbit and human ALOX15 11 may result from the structural differences of both enzymes, 12 which share 71% of amino acid sequence identity. Taken 13 together, these data show that inhibitor induces rather 14 subtle but statistically significant alterations in the product 15 specificity of ALOX15 that may be regarded as indication 16 for enzyme-inhibitor interaction. 17 Leu183Glu+Leu192Glu mutation converts the effect 18 of 8b on LA oxygenation to that of 8a. In a previous 19 report we have shown that Leu183 and Leu192, which are 20 oriented outwards the monomeric protein molecules, are 21 important for the integrity of the inter-monomer interface 22 of the ALOX15 dimer observed in the crystal structure 23 (PDB ID 2P0M). Mutation of these residues to negatively 24 charged glutamates (Leu183Glu+Leu192Glu mutant) 25 affected the mode of self-association of the enzyme and 26 this alteration was mirrored by modified catalytic 27 properties (substrate inhibition above 10 µM)44. In this 28 study we created a structural in silico model for the 29 Leu183Glu+Leu192Glu double mutant of rabbit ALOX15 30 and performed a 200 ns MD simulation. The results show 31 that His585(A), Trp181(B), and His585(B) form a network 32 of π-π interactions even though this is rather weak. The 33 hydrophobic amino acid cluster that stabilizes the inter- 34 monomer interaction of the wild-type enzyme has been 35 dismounted. In the double mutant Leu188 (B) was moved Figure 4. Leu183Glu+Leu192Glu mutant interface (A) vs. that 36 away and pointed outwards. Leu179 (A) did also moved of the wild-type enzyme (B). (C) Inhibition of linoleate and 37 away but pointed inwards the interface. Glu183 (A) and arachidonate oxygenase activity of rabbit ALOX15 38 Glu192 (B) formed a salt bridge with Lys189 (B). Trp181 (A) L183E+L192E mutant (10 µg purified enzyme, 15 µM LA) in the moved outwards and did not contribute any more to the presence of 8a and 8b. Spectrophotometric assays were 39 intermonomer interaction. In the wild-type ALOX15 dimer performed as described in Experimental section. PDB ID 40 the α2(A) and α2(B) helices are located parallel, while in 2P0M of crystal structures of rabbit ALOX15 was used as 41 the Leu183Glu+ Leu192Glu double mutant they are not starting points for molecular modeling. 42 parallel any more (Figure 4A,B). 43 MD simulations of enzyme-inhibitor-substrate 44 To obtain experimental evidence whether the effect of 8b complexes. To explain the observation that compound 8b 45 binding to one ALOX15 monomer may be translated to induces substrate selective enzyme inhibition but 8a does 46 another LA-bound monomer via joint interface observed not (Figures 2 and3), we hypothesized that inhibitor 47 in the crystal structure we compared the inhibitor binding at one monomer of the ALOX15 dimer may induce 48 efficiency of compounds 8a and 8b on the LA- and AA- conformational alterations in the other monomer, which 49 oxygenase activity of wild-type rabbit ALOX15 and its modifies the alignment of the fatty acid substrate at the 50 Leu183Glu+Leu192Glu double mutant. If allosteric active site. To test this hypothesis, we carried out MD 51 mechanisms are involved in the mode of action of 8b the simulations and employed a structural model, in which 52 Leu183Glu+Leu192Glu exchange should induce a right- compounds 8a and 8b were placed inside the putative 53 shift of the dose-response curve of 8b approaching that of substrate binding pocket of conformer A and LA in the compound 8a. When we performed this experiment, we substrate binding pocket of conformer B [complexes 54 found that the dose response curves of LA and AA ALOX15-8a(A)-LA(B), ALOX15-8b(A)-LA(B)] by means of 55 inhibition by 8b were close to each other (Figure 4C) with docking calculations. Finally, we selected the energetically 56 IC50 values of 4.95±0.38 µM and 7.02±0.91 µM for AA and most favourable binding poses and initiated an MD 57 LA, respectively. In fact, the dose-response curves for trajectory. When LA was 58 59 60 ACS Paragon Plus Environment Journal of Medicinal Chemistry Page 8 of 28 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 Figure 5. Effect of inhibitor on the substrate binding. (A) Most representative binding mode of LA in the ALOX15-8a(A)-LA(B) (in 22 green) and ALOX15-8b(A)-LA(B) (in purple) complexes. (B) Most representative binding mode of AA in the ALOX15-8a(A)-AA(B) 23 (in green) and ALOX15-8b(A)-AA(B) (in purple) complexes. The centroids of the most representative cluster for LA of the ALOX15- 8a(A)-LA(B) and AA of the ALOX15-8a(A)-AA(B) complex simulation is depicted. The clustering has been carried out using an 24 RMSD of 0.5 Å for the heavy atoms of the LA (AA) substrate. (C) Cavity surfaces of monomer B in ALOX15-8a(A)-LA(B) and (D) 25 ALOX15-8b(A)-LA(B) complexes. Plots were carried out with the CavityPlus web server. PDB ID 2P0M of crystal structures of 26 rabbit ALOX15 was used as starting points for molecular modeling. 27 28 29 bound at the conformer B in the ALOX15-8a(A)-LA(B) ALOX15-8b(A)-LA(B) complexes the cavity surfaces of 30 complex the most stable interactions along the MD conformer B for both complexes are shown in Figure 5C- 31 between 8a and the protein were a hydrogen bond of the D. 32 side chain of Arg403 with the carbonyl moiety of 8a. 33 Finally, we performed docking studies to define the most Another hydrogen bond was visible between the backbone 34 favourable binding mode for AA in the pocket of oxygen atom of Arg403 and the NH-group of substituted 35 conformer B in the ALOX15-8a(A)-AA(B) and ALOX15- aniline moiety of 8a. A third stable hydrogen bond was 36 8b(A)-AA(B) complexes. Here again, the best AA poses observed between the NH-group (carbamate moiety) of 8a 37 were taken as the initial structures for MD simulations of and the backbone oxygen atom of Gly407 (SI Figure S12A). 38 both systems. Along the MD trajectory of the ALOX15- In the complex ALOX15-8a(A)-LA(B) the LA molecule is 39 8a(A)-AA(B) complex we found that 8a adopts a different stretched out in the putative substrate binding pocket of 40 conformation when compared with its shape in ALOX15- conformer B (Figure 5A). Its carboxylate group interacted 41 8a(A)-LA(B) MD simulation. Thus, inhibitor 8a may be with the enzyme protein via two hydrogen bonds to two 42 present in two different poses in the conformer A Arg side chains: i) Arg403, ii) Arg599. Both interactions 43 depending on the chemistry of the substrate (AA vs. LA) were maintained during the entire MD simulation period. that is present at the active site of conformer B. The 44 The methyl tail of LA is located close to the bottom of the structure of the enzyme-inhibitor-complex appears to be 45 substrate binding cavity that is formed by Ile418, Met419 more rigid when 8b is bound in the putative substrate 46 and Ile593 as suggested by the Triad Concept46. The binding pocket of conformer A. In fact, this inhibitor 47 presence of the inhibitor 8b in substrate binding pocket of presents the same binding mode in the three complexes 48 conformer A (ALOX15-8b(A)-LA(B) complex) (SI Figure studied here [ALOX15-8b(A)-empty(B) (see SI), ALOX15- 49 S12B) modifies the shape of the substrate binding cavity of 8b(A)-LA(B), ALOX15-8b(A)-AA(B)]. The AA’s carboxylate 50 conformer B (SI Figure S12D). Now, the two Arg residues group interacts with Arg599 via two hydrogen bonds 51 (Arg403, Arg599) that anchor the carboxylic group of the (Figure 5B) and both interactions are maintained during 52 substrate fatty acid, are blocked in this complex. the whole MD simulation period. The tail of AA moves 53 Consequently, the mode of LA binding is altered and the along the trajectory but remains at the bottom of the cavity 54 substrates carboxylate forms new hydrogen bonds with surrounded by residues Ile418, Met419 and Ile593 as 55 Arg405 and Asn152. The tail of linoleic acid remains in a suggested by the Triad 56 similar position regardless whether 8a or 8b was present in 57 the cavity of conformer A. For better visualization of the global differences between ALOX15-8a(A)-LA(B) and 58 59 60 ACS Paragon Plus Environment Page 9 of 28 Journal of Medicinal Chemistry Table 2. Average distances of C11 (and its hydrogen atoms) of LA to the OH-group for the system without inhibitor, 1 with inhibitor 8a and inhibitor 8b along with the percentage of well-oriented structures 2 3 System d(C11-OH) (Å) d(H11proS-OH) (Å) d(H11proR-OH) (Å) Well-oriented structures (%) 4 No inhibitor* 4.6 4.7 4.6 90.09 5 6 8a** 5.6 5.8 5.6 78.89 7 8b 5.5 6.2 5.6 34.00 8 *These results correspond to the MD simulation of the ALOX15-LA(B) complex. **In the ALOX15-8a(A)-LA(B) system the 9 averages were calculated over the second 100ns-period of the MD trajectory. 10 11 12 Table 3. Average distances of C13 (and its hydrogen atoms) of AA to the OH-group for the system without inhibitor, 13 with inhibitor 8a and inhibitor 8b along with the percentage of well-oriented structures 14 15 System d(C13-OH) (Å) d(H13proS-OH) (Å) d(H13proR-OH) (Å) Well-oriented structures 16 (%) 17 No inhibitor* 5.2 4.8 5.5 97.78 18 88.47 8a 5.4 4.9 6.0 19 8b 6.4 6.0 7.0 87.90 20 21 *These results correspond to the MD simulation of the ALOX15-AA(B) complex. 22 23 24 Concept46. In contrast, in the ALOX15-8b(A)-AA(B) and 15-H(p)ETE, formed from LA and AA, respectively33-34. 25 complex the presence of inhibitor 8b in the conformer A Although the existence of allosteric sites that bind 26 modifies the shape of the cavity of conformer B. Arg403 endogenous ALOX products47 or other molecules48-50 have 27 and Arg599 are blocked in this complex. Although the previously been suggested there is no direct structural data 28 binding mode of AA changes somewhat with respect to its supporting this hypothesis. As COX2 that forms stable 29 head (new hydrogen bonds are formed between AA catalytically active dimers in aqueous solutions, in which 30 carboxylate group and Arg405, Asn152, and Lys146), the tail one monomer may function as allosteric regulator whereas 31 of the substrate fatty acid remains at the same region as the other monomer is catalytically active51, rabbit ALOX15 32 observed for ALOX15-8a(A)-AA(B) complex. Thus, in the might also form dimers. However, in contrast to COX2 and 33 ALOX15-8a(A)-LA(B) and the ALOX15-8b(A)-LA(B) humanALOX552 or ALOX1231 these dimers (weak self- 34 complexes the substrate molecules adopt different associates) may be formed in solution via transient 35 conformations. These differences are related to the geometrical interfaces of structurally different enzyme 36 number of well-oriented structures found in both cases conformations.31, 35 Moreover, there is a dynamic 37 (Table 2). When 8b is present in the system it causes a equilibrium between monomeric and dimeric fractions31. 38 clear reduction of the well-oriented LA structures (those Under hypotonic conditions (20 mM Tris) the enzyme is 39 structures with the C11-H bond well-oriented for the H- mainly present as monomer, whereas, at physiological salt abstraction) in comparison to the number of well-oriented concentrations dimerization is favoured31. 40 structures obtained for 8a system. In the ALOX15-8a(A)- 41 In the crystal structure of rabbit ALOX15 (PDB entry 2P0M) LA(B) simulation we have found 79% of well-oriented 42 the enzyme is present as mixed dimer, in which two structures for the H11-abstraction reaction from the total 43 conformers A and B interact with each other via their α2 number of selected snapshots, whereas there is only 34% 44 and α18 helices.32 Mutagenesis studies combined with in the case of the ALOX15-8b(A)-LA(B) system (Table 2, SI 45 SAXS measurements in aqueous solution suggested the Figure S10). This data indicates that LA is bound to 46 importance of integrity of the hydrophobic inter-monomer ALOX15 in a less reactive conformation with respect to the 47 interface for ALOX15 stability and its catalytic properties. orientation of the hydrogen atoms attached to C11. For AA 48 Numerous contacts between the two helices α2 and α18 the situation is somewhat different. With this substrate the 49 restrain the mobility of the two monomers within the number of well-oriented structures at C13 (the structures 50 ALOX15 dimer. Mutagenesis studies suggested that with the C13-H bond well-oriented for the H-abstraction) 51 rearrangements of inter- and intra-molecular contacts of in the system appears to be similar for both, the ALOX15- 52 helix 18 may correlate with the volume of the putative 8a(A)-AA(B) (88.5%) and the ALOX15-8b(A)-AA(B) (88%) 53 substrate binding cavity of the catalytic monomer in the complexes (Table 3, SI Figure S11). ligand bound ALOX 15 dimer and, thus, may fine-tune the 54 DISCUSSION reaction specificity of the enzyme.42 This hypothesis is 55 consistent with a recent report that showed that 56 Mammalian ALOX15 are allosteric enzymes and the dimerization of human ALOX12 affect the enzymatic 57 substrate specificity and kinetic parameters of oxygenation activity of this enzyme via allosteric mechanism.53 58 are affected by their own products, 13-H(p)ODE and 12- 59 60 ACS Paragon Plus Environment Journal of Medicinal Chemistry Page 10 of 28 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 Figure 6. Schematic representation of the mechanism of proposed for action of 8b. 19 20 21 Here we report that N-substituted 5-(1H-indol-2-yl)-2- information, SI Figure S12). When EI is formed, no AA 22 methoxyanilines 6b, 6c, 8b and 8c may selectively inhibit binding may be possible any more owing to 23 linoleate oxygenase activity of rabbit and human ALOX15 conformational restrictions (Figure 6B). Which of these 24 with IC50-values in the lower two-digit nanomolar range mechanistic scenarios may correctly reflect the 25 without affecting the rate of AA oxygenation. Since experimental data will be studied in a follow-up paper. 26 ALOX15 exhibits comparable binding affinities for LA and When no OCH3-group was present at the aniline moiety of 27 AA54-55 a conventional mode of inhibition predicts that the core pharmacophore (compound 8a) no substrate 28 oxygenation of the two substrates should be inhibited to a selectivity in enzyme inhibition was observed. The 29 similar extent at a given inhibitor concentration. However, experimental data for 8a were best fitted to a mixed model 30 we observed remarkable substrate-specific ALOX15 inhibition by 8b or 6c and this substrate specificity can of inhibition and analysis of the experimental data (SI 31 best be explained by an allosteric enzyme-inhibitor Figure S7) suggest that 8a may compete with both AA and 32 interaction. Extensive kinetic measurement suggested that LA for binding at the active site of the enzyme (Figure 33 3E,F). Our MD simulations indicated that rigid location of 34 inhibition of LA oxygenation by 8b or 6c may be best described by a non-competitive inhibition model. Here the 8b in the putative substrate binding pocket of one 35 monomer may fix the structure of the binding cavity of 36 substrate affinity (KM) remains constant, but kcat goes down (Figure 3A,C). This data suggests that binding of LA as another monomer in dimer. In turn, 8a adopts different 37 positions depending on the chemistry of the substrate (AA oxygenation substrate and binding of the inhibitor are 38 vs. LA). Thus, 8a that is not bounded as tightly when independent processes, but that the ternary complex 39 compared with 8b, does not cause this structural (ESILA) may not be productive. In contrast, for AA the best 40 modification. In the presence of 8b LA is bound at the fit of the experimental data suggested an uncompetitive 41 substrate binding pocket of the catalytic monomer in a less mode of inhibition. In this case KM and kcat were reduced 42 to similar extents (SI Figures S5 and S6) at increasing productive conformation. In contrast, for AA the number 43 inhibitor concentrations. A reduction of the KM-value of well-oriented substrate structures (those structures in 44 suggests a better enzyme-substrate binding owing to which the C13-H bond are well-oriented for the H- 45 constant depletion of the ESAA complex upon inhibitor abstraction) is not affected by the inhibitor (Tables 2 and 46 binding (ESIAA formation). These data suggest that the 3). 47 inhibitor can preferentially bind to the enzyme-substrate Human and rabbit ALOX15 orthologs exhibit dual 48 complex (ESAA), forming a non-productive ternary complex positional specificity as they form both 15- and 12- 49 (ESIAA). Here, one might suggest that when LA binds to hydroperoxides during AA oxygenation.56-57 In fact, a 15- to 50 ALOX15 or when its product (13-H(p)ODE) is formed 12-H(p)ETE ratio of 95.3±0.2 vs. 4.7±0.2 and 90.2±0.6 vs. 51 during lipoxygenase reaction the ESLA complex is formed 9.8±0.8 were previously reported for recombinant rabbit 52 (Figure 6A). The later binds 8b to form the non- and human ALOX15, respectively.42 Here we found that 53 productive ESILA complex. On the other hand, it might be that neither 8a nor 8b impacted the product specificity of 54 possible that 8b binds first to form the dimeric ALOX15- the rabbit enzyme with AA, but the product specificity of 55 8b(A)-empty(B) complex EI (Figure 6A) and this scenario human enzyme was slightly altered (SI Figure S9). In other 56 will resemble the situation observed in the crystal words, in the presence of 8a or 8b the human enzyme 57 structure (2P0M). In fact, our MD simulations supported allows AA to adopt different conformation in the substrate 58 the rigidity and stability of such a complex (see supporting binding pocket, which slightly favours AA 12- 59 60 ACS Paragon Plus Environment Page 11 of 28 Journal of Medicinal Chemistry lipoxygenation. Human and rabbit ALOX15 orthologs inhibitors and suggest an allosteric model for their 1 share an amino acid identity of 71 % and, thus, there are inhibitory activity. This model is based on our previous 2 structural differences between the two enzymes. These observations that ALOX15 may form transient dimers in 3 differences might be responsible for these observations. aqueous solutions so that inhibitor binding at the substrate 4 Although the structural basis for this subtle difference has binding pocket of one monomer may induce 5 not been explored in detail it might be possible that conformational changes in the catalytic cavity of the other 6 inhibitor 8b binding at one monomer of the ALOX15 dimer monomer. The 2-methoxyaniline moiety attached to the 7 may induce minor structural alterations in the geometry of core pharmacophore appears to be crucial for the substrate 8 the catalytic cavity of the other monomer and, hence, selectivity of ALOX15 inhibition. 9 might impact the alignment of the substrate fatty acid at 10 the active site. 11 Additional evidence for the proposed allosteric mechanism EXPERIMENTAL SECTION 12 of 8b induced substrate-specific ALOX15 inhibition was Chemistry. The solvent and reagents were purchased 13 provided by mutagenesis studies. When we explored the from Acros or Sigma-Aldrich (Schnelldorf, Germany) and 14 inhibition behaviour of the Leu183Glu+L192Glu double were used without further purification unless otherwise 15 mutant of rabbit ALOX15, for which disruption of the inter- noted. 1H and 13C NMR spectra were recorded with a 300 16 monomer interface within the ALOX15 dimer (PDB ID MHz Brucker MSL spectrometer in CDCl3 or acetone-d6 17 2P0M) has previously been reported,44 we found that these using tetramethylsilane as the internal standard for 1H- 18 mutations prevented the substrate specific inhibition of NMR. Chemical shifts in 13C-NMR were referenced to the 19 the LA oxygenase activity by 8b (Figure 4). For the mutant residual carbon signal of CDCl3 and acetone-d6 at δ 13C = 20 enzyme, the IC50 values obtained for LA and AA 77.19 and 206.26 ppm, respectively. Chemical shifts are 21 oxygenation were very similar. Although the IC50-values of given in ppm, spin-spin interaction constants in Hz. IR 22 8a and 8b obtained for the enzyme mutant were higher spectra were recorded using a Brucker Equinox 55 23 than those for the wild-type enzyme, there was no spectrometer (Germany) at 4 cm-1 resolution. Elemental 24 pronounced substrate specific inhibition. analysis was performed with a Thermo Finnigan FLASH EA 25 In this study we observed that 8b selectively inhibited 1112 analyzer (Italy). Flash column chromatography was 26 linoleate oxygenase activity of human and rabbit ALOX15. carried out using silica gel (Acros, Germany, particle size 27 Interestingly, no difference in the degree of inhibition of 60-200 μm) as stationary phase. Silica gel 60 F254 plates 28 LA and AA was detected when mouse ALOX15 was used as (Merck, Germany) were used for thin-layer 29 catalyst. Although human ALOX15 shares about 74% of chromatography (TLC). Spots were detected under UV 30 amino acid identity with its mouse ortholog, the murine light or after staining with an ethanolic (3%) solution of 31 enzyme lacks Gln596 that is conserved in the ALOX15 vanillin. Analytical HPLC of all compounds was performed 32 orthologs of rabbits, men and other higher primates. on a Shimadzu LC-10Avp liquid chromatograph equipped 33 Recently, this residue has been reported to contribute to with a SPD-10Advp UV detector (Japan) using a Nucleodur 34 the allosteric properties of rabbit ALOX1542. In contrast, in C18 Gravity column (Macherey-Nagel, Düren, Germany; 35 rodents (rat, mouse) a Leu is present at position 596 but it 250 x 4 mm, 5 μm particle size) coupled with a guard 36 remains unclear whether this amino acid exchange may column (8 x 4 mm, 5 μm particle size) at a flow rate of 1 37 affect the allosteric property of the enzyme. mL/min. For compounds 5a-5c the solvent system 38 consisted of a mixture of acetonitrile : 10 M ammonium There are reports in the literature suggesting that ALOX15 formate (50:50, v/v) and its pH was adjusted to pH 3.0 with 39 monomers involve allosteric binding sites48-49 but for the formic acid. For all other compounds we used acetonitrile 40 time being such binding sites have not convincingly been : water : acetic acid (70:30:0.1, v/v). Preparative HPLC was 41 characterized. On the other hand, the results obtained in performed using a Knauer HPLC pump 64 system coupled 42 this study are consistent with the formation of non- with a differential refractometer and UV-VIS detector 43 covalently linked dimeric enzyme associates in aqueous (Knauer, Germany). A Luna C18(2) column (75 × 30.0 mm, 44 solutions. Although sedimentation velocity analytical particle size 5 μm, Phenomenex, USA) and isocratic 45 ultracentrifugation established the gold standard in systems of different compositions of solvents A and B at a 46 aggregation analytics58-59, it cannot be applied to study flow rate of 25 mL / min were applied to achieve the best 47 reversible ALOX15 dimerization in diluted protein chromatographic performance. Solvent A was HPLC-grade 48 solutions. This method requires protein amounts that are acetonitrile. A polar component of the system (solvent B) 49 at least 100-fold higher than those used in our kinetic was a 10 M solution of ammonium formate adjusted with 50 assays (2-4 µg/ml). Unfortunately, for the time being, there formic acid to pH 3.0. All compounds were >95% as 51 is no direct experimental proof for either of these two indicated by analytical HPLC. Mass spectra (EI) were 52 possible explanations. recorded on an Agilent a 6890N gas chromatograph 53 coupled with 5973N mass spectral detector (Agilent, USA) 54 using a DB-5ms column (30 m, coating thickness 0.5 μm, 55 CONCLUSIONS Agilent J&W, USA). An injector temperature of 250 °C, an 56 ion source temperature of 230 °C and the electron energy Here we describe the chemical synthesis and functional 57 characterization of novel substrate-specific ALOX15 of 70 eV were set. Helium was used as carrier gas at a flow 58 rate of 1.0 mL / min. Samples were eluted using the 59 60 ACS Paragon Plus Environment Journal of Medicinal Chemistry Page 12 of 28 following temperature program: isothermally at 70 °C for 5 aqueous solution of ammonia until pH 13 was reached. The 1 min, then from 70 to 290 °C at a rate of 30 °C/min, followed precipitate was filtered off and recrystallized from a 2 by isothermal step at 290°C for 10 min. Finally, the column mixture ethanol/water (1:1, v/v). The raw product that 3 was conditioned at 310 °C for 30 min. HRMS were obtained remained in the mother liquor was chromatographed on 4 on the Bruker micrOTOF II mass spectrometer with an silica gel using gradient elution with THF/Pet (from 20 to 5 electrospray ionization (ESI). The measurements were 66% THF). 6 performed using either positive or negative ionization 7 mode. Interface capillary voltage was set up at 4500 V and 3-(1H-indol-2-yl)aniline (5a): Yield by method B: 60%. the mass range from m/z 50 to 3000 Da. External or m.p.: 168 °C (lit. 166 °C)38. Rf = 0.53 (EtOAc / Pet 1: 1, v/v). 8 internal calibration was performed with Electrospray Analytical RP-HPLC: Rt 11.27 min (235 and 310 nm). IR 9 Calibration Solution (Fluka). N2 was applied as a drying (KBr): ν 3440 cm-1 (s), 3399 cm-1 (s), 1611 cm-1 (s). UV/Vis 10 gas. λmax 228 nm, 310 nm. 1H NMR (300 MHz, acetone-d6) δ 11 10.48 (s, 1H), 7.53 (dm, J = 7.76 Hz, 1H), 7.39 (dq, J = 8.06, 12 1-(4-methoxy-3-aminophenyl)ethanone (2). To an 0.88 Hz, 1H), 7.16 - 6.97 (m, 5H), 6.76 (dd, J = 2.26, 0.92 Hz, 13 ethanolic solution (100 mL) of 1-(4-methoxy-3- 1H), 6.64 (ddd, J = 7.49, 2.28, 1.50 Hz, 1H), 4.66 (s, 2H). 13C 14 nitrophenyl) ethanone 1 (11.1 g, 56.92 mmol) 0.55 g (5 % NMR (75 MHz, acetone-d6) δ 149.75, 139.74, 138.24, 134.33, 15 m/m) of 10% Pd/C-catalyst and ammonium formate (10.8 130.33, 122.28, 120.90, 120.28, 114.77, 114.70, 111.95, 99.38. MS 16 g, 171.26 mmol) were added. The resulting mixture was kept (70 eV) m/z (%): 208.1 (100), 180 (18). Elemental analysis 17 for 3 h under reflux. After the reduction was complete the calcd: C 80.74, H 5.81, N 13.45; found: C 80.68, H 5.73, N 18 catalyst was filtered off, the solvent was evaporated under 13.36. 19 reduced pressure and the dry residue was dissolved in 50 20 mL of saturated K2CO3 solution. The organic products were 5-(1H-indol-2-yl)-2-methoxyaniline (5b): Yield by method 21 extracted with CH2Cl2 (4 x 20 mL), the combined organic B: 52%. m.p.: 208.2 °C (lit. 208-209 °C).27 Rf = 0.41 (EtOAc / 22 extracts were dried over Na2SO4 and concentrated. The raw Pet 1: 1, v/v). Analytical RP-HPLC: Rt 11.42 min (235 and 310 23 product was chromatographed using gradient elution with nm). IR (KBr): ν 3440 cm-1 (s), 3399 cm-1 (s), 1611 cm-1 (s), 24 EtOAc/Pet (from 25 to 50% EtOAc) to yield 8.5 g (90.5%) 1449 cm-1 (m). UV/Vis λmax 230 nm, 312 nm. 1H NMR (300 of compound 2. Rf = 0.66 (EtOAc/Pet 1: 1). IR (KBr): ν 3446 MHz, acetone-d6) δ 10.49 (s, 1H), 7.50 (dm, J = 7.73 Hz, 1H), 25 cm-1 (s), 3354 cm-1 (s). 1H NMR (300 MHz, CDCl3) δ 7.35 (d, 7.36 (dm, J = 7.96 Hz, 1H), 7.20 (d, J = 2.23 Hz, 1H), 7.12 - 26 J = 8.32, 2.12 Hz, 1H), 7.31 (d, J = 2.13 Hz, 1H), 3.88 (s, 3H), 6.97 (m, 3H), 6.88 (d, J = 8.35 Hz, 1H), 6.66 - 6.65 (m, 1H), 27 2.49 (s, 3H). 13C NMR (75 MHz, CDCl3) δ = 197.34, 151.29, 4.50 (s, 2H), 3.85 (s, 3H). 13C NMR (75 MHz, acetone-d6) δ 28 136.31, 130.75, 120.44, 113.99, 109.31, 55.67, 26.28. Elemental 147.80, 139.91, 138.77, 138.10, 130.43, 126.64, 121.83, 120.58, 29 analysis calcd: C 65.44, H 6.71, N 8.48, found: C 65.34, H 120.15, 114.92, 111.99, 111.73, 111.44, 98.25, 55.89. MS (70 eV) 30 m/z (%): 223 (100), 238 (75). Elemental analysis calcd: C 31 6.77, N 8.54. MS (70 eV): m/z (%): 150 (100), 165 (84), 122 (29). 74.61, H 5.92, N 11.76; found: C 74.57, H 5.91, N 11.69. 32 33 5-(5-fluoro-1H-indol-2-yl) -2-methoxyaniline (5c): Yield by General method for preparation of hydrazones 4a-с. A method B: 58%. m.p.: 221-223 °C. Rf = 0.45 (CH2Cl2 with 5% 34 solution of phenylhydrazine 3a-b (1.20 mmol) and 35 NH3). Analytical RP-HPLC: Rt = 11.82 min (235 and 310 nm). ethanone 2a-b (0.6 mmol) in 6 mL of a mixture of ethanol- IR (KBr):  3439 cm-1 (m), 3351 cm-1 (m), 3291 cm-1 (m) 36 acetic acid (5: 1, v/v) was refluxed for 18 h. After the reaction 1611cm-1 (s), 1449 cm-1 (vs), 1186 cm-1 (m). UV/Vis λmax 228 37 mixture was cooled down it was stirred at room nm, 322 nm. 1H NMR (300 MHz, acetone-d6) δ 10.58 (s, 1H), 38 temperature for one additional hour. The precipitated 7.34 (ddt, J = 8.76, 4.58, 0.73 Hz, 1H), 7.22 - 7.17 (m, 2H), 39 hydrazones 4a-b were filtered off, washed with hexane (10 7.09 (dd, J = 8.29, 2.20 Hz, 1H), 6.88 ( d, J = 8.33 Hz, 1H), 40 mL), dried and used for further synthesis without 6.86 – 6.80 (m, 1H), 6.66 - 6.65 (m, 1H), 4.52 (s, 2H), 3.86 41 additional purification. (s, 3H). 13C NMR (75 MHz, acetone-d6) δ 160.24, 157.17, 42 General procedure for the preparation of aminoarylindoles 147.99, 141.95, 138.83, 134.65, 130.79, 130.66, 126.19, 115.00, 43 5a-c. Method A: A mixture of the hydrazone 4a-c (1.89 112.54, 112.41, 111.95, 111.39, 109.79, 109.44, 105.18, 104.87, 44 mmol) and ZnCl2 (15.13 mmol) was heated to 180 °C for 2 98.44, 98.38, 55.87. MS (70 eV) m/z (%): 241 (100), 256 (75). 45 h. The reaction mixture was then cooled down and Elemental analysis calcd: C 70.30, H 5.11, N 10.93; found: C 46 quenched with 10% HCl (50 mL). The precipitate was 69.51, H 5.29, N 10.66. 47 filtered off, the filtrate was alkalized with 25% aqueous 48 General procedure for the preparation of substituted 2- ammonia until a pH 13 was reached and extracted with 49 CH2Cl2 (3 x 25 mL). The combined organic extracts were arylindolesulfocarbomats 6a-c−8a-c. To a solution of 50 evaporated under reduced pressure. The raw product was chlorosulfonyl isocyanate (0.27 mmol) and the 51 chromatographed on silica gel using gradient elution with corresponding alcohol (0.25 mol) in 3 mL of dry CH2Cl2 52 THF/Pet (from 20 to 66% THF). The yields of the Et3N (0.38 mmol) was added and the mixture was stirred 53 aminoarylindoles 5a-c were only 5-10%. Method B: A for 15 min at rt. After that a solution of arylindole 5a-e (1.0 54 mixture of 80% polyphosphoric acid (2.07 mmol), mol) in 3 mL of dry CH2Cl2 was added to the mixture and 55 phenylhydrazine 3a-b (1.38 mmol) and ethanone 2a-b (1.38 the reaction was left overnight at room temperature. After 56 mol) was heated to 130 °C and kept at this temperature for the reaction was complete, it was quenched with water (10 57 30 minutes. After the reaction mixture was cooled down, it mL). The organic layer was separated and dried with was quenched with water (100 mL) and alkalized with 25% Na2SO4, concentrated under reduced pressure and the raw 58 59 60 ACS Paragon Plus Environment Page 13 of 28 Journal of Medicinal Chemistry product was purified on silica gel using a gradient elution Allyl (N-(3-(1H-indol-2-yl)phenyl)sulfamoyl)carbamate 1 with THF/Pet (from 20 to 66% THF) followed by (7a): Yield: 36.1% Rf = 0.57 (THF / Pet 1: 1). Analytical RP- 2 preparative RP-HPLC (solvent A/ solvent B 60:40 or 65:35, HPLC: Rt 3.78 min (235 and 320 nm), purity 98.5%. UV/Vis 3 v/v, 235 or 310 nm) for all compounds. λmax 233 nm, 307 nm. 1H NMR (300 MHz, acetone-d6) δ = 4 7.83 (t, J = 1.98 Hz, 1H), 7.64 (dm J = 7.80 Hz, 1H), 7.58 (dm, 5 Isobutyl (N-(3-(1H-indol-2- J = 7.78 Hz, 1H), 7.47 – 7.38 (m, 2H), 7.26 (ddd, J = 8.05, 6 yl)phenyl)sulfamoyl)carbamate (6a): Yield: 29.9%. Rf = 0.4 2.20, 1.05 Hz, 1H), 7.12 (ddd, J = 8.21, 7.02, 1.29 Hz, 1H), 7.03 7 (THF / Pet 1: 1, v/v). Analytical RP-HPLC: Rt 4.64 min (235 (ddd, J = 8.08, 7.04, 1.12 Hz, 1H), 6.88 – 6.84 (m, 1H), 5.84 8 and 320 nm), purity 99.0%. UV/Vis λmax 233 nm, 309 nm. 1H NMR (300 MHz, acetone-d ) δ = 10.66 (s, 1H), 7.82 (t, J = (ddt, J = 17.28, 10.64, 5.44 Hz, 1H), 5.21 (dq, J = 17.29, 1.61 Hz, 9 6 1H), 5.09 (dq, J = 10.58, 1.48 Hz, 1H), 4.57 (dt, J = 5.50, 1.51 1.73 Hz, 1H), 7.65 (ddd, J = 7.79, 1.74, 1.02 Hz, 1H), 7.58 (dm, 10 Hz, 2H). 13C NMR (75 MHz, acetone-d6) δ 152.58, 139.15, J = 7.88 Hz, 1H), 7.45 - 7.40 (m, 2H), 7.25 (ddd, J = 8.04, 2.20, 11 138.29, 138.10, 134.63, 132.97, 130.58, 129.95, 122.81, 122.06, 1.01 Hz, 1H), 7.12 (ddd, J = 8.17, 7.05, 1.26 Hz, 1H), 7.03 (ddd, 12 121.13, 120.51, 120.36, 118.27, 117.89, 112.10, 100.35, 66.90. J = 8.05, 7.09, 1.12 Hz, 1H), 6.87 (dd, J = 2.25, 0.92 Hz, 1H), 13 HRMS (ESI): calculated for [M+H]+ = 372.1013, found 3.88 (d, J = 6.64 Hz, 2H), 1.85 (hept, J = 13.37, 6.74 Hz, 1H), 14 [M+H]+ = 372.1004 (100 %), error ppm = 2.418696199; 0.81 (d, J = 6.73 Hz, 6H). 13C NMR (75 MHz, acetone-d6) δ 15 calculated for [M+Na]+ = 394.0832, found [M+Na]+ = 152.31, 139.10, 138.58, 138.33, 134.87, 130.66, 130.15, 122.94, 16 394.0824 (53.5 %), error ppm = 2.030028177; calculated for 122.33, 121.23, 120.67, 120.64, 118.28, 112.19, 100.58, 72.67, 17 [M+K]+ = 410.0571, found [M+K]+ = 410.0562 (15.1 %), error 28.58, 18.97. HRMS (ESI): calculated for [M+H]+ = 388.1326, 18 ppm = 2.194816283. found [M+H]+ = 388.1330 (50.6 %), error ppm = 1.030575633; 19 calculated for [M+Na]+ = 410.1145, found [M+Na]+ = 410.1154 Allyl (N-(3-(1H-indol-2- 20 (100 %), error ppm = 2.194509094. yl)methoxyphenyl)sulfamoyl)carbamate (7b): Yield: 21 49.9% Rf = 0.58 (THF / Pet 1: 1). Analytical RP-HPLC: Rt 4.29 22 Isobutyl (N-(3-(1H-indol-2-yl)methoxyphenyl)sulfamoyl)- min (235 and 320 nm), purity 98.0%. UV/Vis λmax 225 nm, 23 carbamate (6b): Yield: 55.7%. Rf = 0.31 (THF / Pet 1: 1, v/v). 311 nm. 1H NMR (300 MHz, acetone-d6) δ 7.92 (d, J = 2.24 24 Analytical RP-HPLC: Rt 5.43 min (235 and 320 nm), purity Hz, 1H), 7.64 (dd, J = 8.56, 2.24 Hz, 1H), 7.55 (ddd, J = 7.71, 25 99.5%. UV/Vis λmax 225 nm, 311 nm. 1H NMR (300 MHz, 1.33, 0.79 Hz, 1H), 7.40 (dq, J = 8.07, 0.93 Hz, 1H), 7.15 (d, J 26 acetone-d6) δ 7.92 (d, J = 1.93 Hz, 1H), 7.63 (dd, J = 8.53, 2.23 = 8.59 Hz, 1H), 7.12 – 6.98 (m, 2H), 6.77 (d, J = 0.90 Hz, 1H), 27 Hz, 1H), 7.54 (ddd, J = 7.71, 1.33, 0.79 Hz, 1H), 7.40 (dm, J = 5.89 (ddt, J = 17.26, 10.64, 5.43 Hz, 1H), 5.26 (dq, J = 17.26, 28 10.05 Hz, 1H), 7.15 (d, J = 8.57 Hz, 1H), 7.12 – 6.98 (m, 2H), 1.63 Hz, 1H’), 5.17 (dq, J = 10.52, 1.41 Hz, 1H’’), 4.62 (dt, J = 29 6.77 (d, J = 0.91 Hz, 1H), 3.90 (d, J = 6.61 Hz, 2H), 1.87 (hept, 5.41, 1.51 Hz, 2H). 13C NMR (75 MHz, acetone-d6) δ 151.91, 30 J = 6.71 Hz, 1H), 0.85 (d, J = 6.74 Hz, 6H). 13C NMR (75 MHz, 150.66, 138.35, 138.19, 132.94, 130.23, 127.40, 126.73, 123.17, acetone-d6) δ 152.07, 150.51, 138.36, 138.19, 130.23, 127.42, 31 122.41, 120.87, 120.45, 118.92, 118.30, 112.72, 111.90, 99.32, 126.76, 123.10, 122.42, 120.87, 120.46, 118.64, 112.68, 111.90, 32 67.09, 56.69. HRMS (ESI): calculated for [M+H]+ = 402.1118, 99.33, 72.69, 56.69, 28.53, 18.97. HRMS (ESI): calculated for 33 found [M+H]+ = 402.1113 (100 %), error ppm = 1.2434356283; [M+H]+ = 418.1430, found [M+H]+ = 418.1426 (81.2 %), error 34 calculated for [M+Na]+ = 424.0938, found [M+Na]+ = ppm = 0.956610537; calculated for [M+Na]+ = 440.1250, 35 424.0934 (48.7 %), error ppm = 0.943187568; calculated for found [M+Na]+ = 440.1242 (100 %), error ppm = 36 [M+K]+ = 440.0677, found [M+K]+ = 440.0673 (4.2 %), error 1.817665436; calculated for [M+K]+ = 456.0990, found 37 ppm = 0.908951055. [M+K]+ = 456.0980 (40.4 %), error ppm = 2.192506451. 38 Allyl (N-(5-(5-fluoro-1H-indol-2-yl)-2-methoxyphenyl)- 39 Isobutyl (N-(5-(5-fluoro-1H-indol-2-yl)-2- sulfamoyl)carbamate (7c): Yield: 52.3%. Rf = 0.32 (THF / 40 methoxyphenyl)-sulfamoyl)carbamate (6c): Yield: 63.1%. Pet 2: 1, v/v). Analytical RP-HPLC: Rt 4.35 min (235 and 320 41 Rf = 0.21 (THF / Pet 1: 1, v/v). Analytical RP-HPLC: Rt 5.55 nm), purity 98.2%. UV/Vis λmax 224 nm, 314 nm. 1H NMR 42 min (235 and 320 nm), purity 98.5%. UV/Vis λmax 224 nm, (300 MHz, acetone-d6) δ 10.81 (s, 1H), 7.91 (d, J = 2.18 Hz, 43 315 nm. 1H NMR (300 MHz, acetone-d6) δ 10.72 (s, 1H), 7.92 1H), 7.64 (dd, J = 8.56, 2.20 Hz, 1H), 7.39 (dd, J = 8.76, 4.54 44 (d, J = 2.24 Hz, 1H), 7.62 (dd, J = 8.57, 2.28 Hz, 1H), 7.39 (dd, Hz, 1H), 7.25 (dd, J = 9.85, 2.57 Hz, 1H), 7.15 (d, J = 8.63 Hz, 45 J = 8.77, 4.50 Hz, 1H), 7.24 (dd, J = 9.86, 2.55 Hz, 1H), 7.16 1H), 6.89 (ddd, J = 9.59, 8.76, 2.54 Hz, 1H), 6.78 (dd, J = 2.22, 46 (d, J = 8.56 Hz, 1H), 6.87 (td, J = 9.16, 2.56 Hz, 1H), 6.76 (s, 0.90 Hz, 1H), 5.88 (ddt, J = 17.19, 10.69, 5.41 Hz, 1H), 5.25 47 1H), 3.94 (s, 3H), 3.89 (d, J = 6.58 Hz, 2H), 1.87 (hept, J = (dq, J = 17.24, 1.58 Hz, 1Ha), 5.17 (dq, J = 10.51, 1.40 Hz, 1Hb), 48 13.43, 6.70 Hz, 1H), 0.85 (d, J = 6.73 Hz, 6H). 13C NMR (75 4.62 (dt, J = 5.41, 1.48 Hz, 2H), 3.92 (s, 3H). 13C NMR (75 49 MHz, acetone-d6) δ 160.37, 157.30, 152.25, 150.73, 140.41, MHz, acetone-d6) δ 160.41, 157.33, 151.78, 151.02, 140.50, 50 134.78, 130.63, 130.49, 127.57, 126.36, 123.20, 118.68, 112.81, 134.95, 132.88, 130.68, 130.55, 127.45, 126.42, 123.47, 119.15, 51 112.70, 110.50, 110.15, 105.46, 105.14, 99.48, 99.42, 72.68, 56.71, 118.39, 112.87, 110.53, 110.18, 105.48, 105.16, 99.59, 99.52, 67.17, 28.54, 18.98. HRMS (ESI): calculated for [M+H]+ = 436.1337, 52 56.74. HRMS (ESI): calculated for [M+H]+ = 420.1024, found found [M+H]+ = 436.1332 (100 %), error ppm = 1.146437434; 53 [M+H]+ = 420.1037 (100 %), error ppm = 3.094483631; calculated for [M+Na]+ = 458.1156, found [M+Na]+ = 458.1151 54 calculated for [M+Na]+ = 442.0843, found [M+Na]+ = (46.9 %), error ppm = 1.091427579; calculated for [M+K]+ = 55 442.0856 (100 %), error ppm = 2.940615625; calculated for 474.0896, found [M+K]+ = 474.0888 (2.0 %), error ppm = 56 [M+K]+ = 458.0583, found [M+K]+ = 458.0591 (50 %), error 1.687444736. 57 ppm = 1.746502574. 58 59 60 ACS Paragon Plus Environment Journal of Medicinal Chemistry Page 14 of 28 Octyl (N-(3-(1H-indol-2-yl)phenyl)sulfamoyl)carbamate Germany), HPLC grade methanol, acetonitril and acetic 1 (8a): Yield: 38.6%. Rf = 0.53 (THF / Pet, 2:1, v/v). Analytical acid from Fisher Scientific (Nidderau, Germany), 2 RP-HPLC: Rt 16.68 min (235 and 320 nm), purity 98.9%. isopropyl-β-d-thiogalactopyranoside (IPTG) from Carl 3 UV/Vis λmax 233 nm, 308 nm. 1H NMR (300 MHz, acetone- Roth GmbH (Karlsruhe, Germany). The E. coli strain XL-1 4 d6) δ 10.69 (s, 1H), 7.82 (t, J = 1.73 Hz, 1H), 7.65 (dm, J = 8.25 blue was purchased from Stratagene (La Jolla, CA), the 5 Hz, 1H), 7.57 (dm, J = 8.13 Hz, 1H), 7.45 – 7.40 (m, 2H), 7.25 Rosetta 2 strain BL21(DE3)pLysS were purchased from 6 (ddd, J = 8.03, 2.19, 1.02 Hz, 1H), 7.12 (ddd, J = 8.16, 7.05, 1.26 Invitrogen (Carlsbad, California, USA). 7 Hz, 1H), 7.03 (ddd, J = 8.18, 2.57, 1.25 Hz, 1H), 6.88 – 6.87 Preparation of rabbit ALOX15. Wild-type rabbit 8 (m, 1H), 4.08 (t, J = 6.61 Hz, 2H), 1.56 – 1.52 (m, 2H), 1.31 – ALOX15 and several enzyme mutants were expressed as N- 9 1.15 (m, 10H), 0.84 (t, J = 6.84 Hz, 3H).13C NMR (75 MHz, terminal His-tag fusion proteins in E. coli using the pQE-9 10 acetone-d6) δ = 152.35, 139.16, 138.56, 138.17, 134.80, 130.64, prokaryotic expression plasmid and XL-1 blue competent 11 130.11, 122.92, 122.20, 121.22, 120.61, 120.48, 118.04, 112.19, cells according to the protocol described previously.60-61 12 100.52, 66.94, 32.59, 30.19, 29.95, 29.90, 26.34, 23.31, 14.34. 13 HRMS (ESI): calculated for [M-H]+ = 442.1806, found [M-H]+ = Preparation of human ALOX15. Human ALOX15 used in 14 442.1819 (100 %), error ppm = 2.939975205. this study was expressed as N-terminal his-tag fusion 15 Octyl (N-(5-(1H-indol-2-yl)-2- proteins employing the E. coli strain Rosetta2(DE3)pLysS 16 methoxyphenyl)sulfamoyl)carbamate (8b): Yield: 65.7%. as overexpression system. Competent bacteria were 17 Rf = 0,43 (THF / Pet 2: 1, v/v). Analytical RP-HPLC: Rt 15.16 transformed with 100 ng of the recombinant pET28b 18 min (235 and 320 nm), purity 98.5%. UV/Vis λmax 226 nm, expression plasmid and the cells were grown overnight on 19 312 nm. 1H NMR (300 MHz, acetone-d6) δ 11.05 (s, 1H), 8.10 kanamycin/chloramphenicol containing agar plates. An 20 (d, J = 2.21 Hz, 1H), 7.49 (dm, J = 7.68 Hz, 1H), 7.41 (dm, J = isolated clone was selected and two 1 mL bacterial liquid 21 7.97 Hz, 1H), 7.36 (dd, J = 8.42, 2.22 Hz, 1H), 7.05 – 6.93 (m, cultures (LB medium with 50 μg/mL kanamycin/35 µg/mL 22 2H), 6.86 (d, J = 8.50 Hz, 1H), 6.70 – 6.69 (m, 1H), 3.88 (t, J chloramphenicol) were grown at 37 °C and 180 rpm 23 = 6.82 Hz, 2H), 3.88 (s, 3H), 1.52 – 1.43 (m, 2H), 1.28 – 1.15 agitation for 6 h. Aliquots of these pre-cultures were added 24 (m, 10H), 0.83 (t, J = 6.76 Hz, 3H).13C NMR (75 MHz, to a 50 to 500 mL main culture to reach an optical density acetone-d6) δ 161.23, 148.64, 139.80, 138.35, 131.78, 130.37, (A600) of 0.10-0.15. Bacterial expression was carried out in 25 126.47, 121.71, 120.50, 119.99, 119.25, 115.63, 112.08, 111.70, 50 or 500 mL Ultra Yield culture flasks (Thomson 26 98.37, 65.33, 56.31, 32.56, 30.15, 29.97, 29.94, 26.79, 23.27, Instrument Company, Oceanside, USA) using the 27 14.33. HRMS (ESI): calculated for [M+H]+ = 474.2057, found optimized Enpresso-Expression system (Enpresso GmbH, 28 [M+H]+ = 474.2064 (100 %), error ppm = 1.476152649; Berlin, Germany) following the instructions of the vendor. 29 calculated for [M+Na]+ = 496.1877, found [M+Na]+ = After overnight incubation at 30°C and 250 rpm agitation, 30 the OD600 had reached values above 5 and expression of 31 496.1881 (100 %), error ppm = 0.806146545; calculated for [M+K]+ = 512.1616, found [M+K]+ = 512.1625 (50 %), error the recombinant his-tag fusion proteins was induced by 32 the addition of 1 mM (final concentration) IPTG. Then the ppm = 1.757257865. 33 cultures were maintained at 22 °C for 24 h at 230-250 rpm 34 Octyl (N-(5-(5-fluoro-1H-indol-2-yl)-2-methoxyphenyl)- agitation. Bacteria were harvested, the resulting pellet was 35 sulfamoyl)carbamate (8c): Yield: 62.5%. Rf = 0.15 (THF / reconstituted in a total volume of 5 to 30 mL PBS and 36 Pet 2: 1, v/v). Analytical RP-HPLC: Rt 15.28 min (235 and 320 bacteria were disrupted at 15,000 psi with an Emulsiflex-C5 37 nm), purity 99%. UV/Vis λmax 222 nm, 314 nm. 1H NMR (300 high-pressure cell homogenizer (Avestin, Ottawa, 38 MHz, acetone-d6) δ 10.83 (s, 1H), 7.91 (d, J = 2.23 Hz, 1H), Canada). After the debris was removed by centrifugation 39 7.63 (dd, J = 8.55, 2.24z Hz, 1H), 7.38 (dd, J = 8.87, 4.63 Hz, (30 min, 14,000 g), the lysis supernatant was passed 40 1H), 7.24 (dd, J = 9.87, 2.53 Hz, 1H), 7.16 (d, J = 8.56 Hz, 1H), through a Ni-NTA column (Qiagen, Hilden, Germany) for 41 6.87 (ddd, J = 9.59, 8.75, 2.52 Hz, 1H), 6.78 (dd, J = 2.27, 0.90 purification (gel bed volume of 1.8 mL). The column was 42 Hz, 1H), 4.09 (t, J = 6.59 Hz, 2H), 3.93 (s, 3H), 1.58 – 1.53 (m, washed twice with 2 mL of washing buffer W1 (100mM Tris, 43 2H), 1.28 – 1.18 (m, 10H), 0.85 (t, J = 6.74 Hz, 3H).13C NMR pH 8.0, 200 mM NaCl, 10 mM imidazol), twice with 2 mL 44 (75 MHz, acetone-d6) δ 160.31, 157.24, 151.78, 150.64, 140.49, with washing buffer W2 (100 mM Tris, pH 8.0, 200 mM 45 134.87, 130.60, 130.46, 127.62, 126.30, 123.03, 118.48, 112.83, NaCl, 25 mM imidazol) and the adherent proteins were 46 112.70, 112.63, 110.44, 110.09, 105.41, 105.10, 99.44, 99.38, eluted 5 times with 0.6 ml of elution buffer (100mM Tris, 47 66.94, 56.65, 32.48, 29.88, 29.84, 29.26, 26.32, 23.25, 14.32. pH 8.0, 200 mM NaCl, 200 mM imidazol). Enzyme 48 HRMS (ESI): calculated for [M+H]+ = 492.1963, found containing fractions (E3-E5) were combined, 49 [M+H]+ = 492.1960 (35.5 %), error ppm = 0.609512912; supplemented with glycerol (10% final concentration), 50 calculated for [M+Na]+ = 514.1782, found [M+Na]+ = 514.1773 shock frozen in liquid nitrogen and stored at −80°C. After 51 (100 %), error ppm = 1.750365924; calculated for [M+K]+ = thawing the pooled Ni-NTA fractions were desalted on an 530.1522, found [M+K]+ = 530.1523 (22.4 %), error ppm = Econo-Pac 10DG column (Bio Rad, Munich, Germany) and 52 0.188625078. the recombinant proteins were further purified by FPLC on 53 a Resource Q column (GE Healthcare Bio-Sciences AB, 54 Biochemistry general. The chemicals used were obtained Uppsala, Sweden) as described previously for rabbit 55 from the following sources: arachidonic acid ALOX1560-61. The fractions that contained active protein 56 (5Z,8Z,11Z,14Z-eicosatetraenoic acid), 13(S)-HpODE,13(S)- were pooled together and used as final enzyme 57 HODE and HPLC standards of 15(S)-HETE and 12(S)-HETE preparations. 58 from Cayman Chem (distributed by Biomol, Hamburg, 59 60 ACS Paragon Plus Environment Page 15 of 28 Journal of Medicinal Chemistry Preparation of mouse ALOX. Recombinant mouse After 3 min the reaction was stopped by addition of 1 Alox15 (mAlox15), human Alox12 (hALOX12), mouse Alox12 NaBH4. Samples were acidified, proteins were precipitated 2 (mAlox12), human ALOX15B and mouse Alox15b were by the addition of 0.5 ml of acetonitrile and the precipitate 3 expressed as N-terminal his-tag fusion proteins in was spun down. The extent of the ALOX activity was 4 employing the E. coli strain Rosetta2(DE3)pLysS strain as measured by HPLC quantification of the ALOX products 5 overexpression system. Competent bacteria were (13-HODE for LA oxygenation, sum of 12- + 15-HETE for AA 6 transformed with 50 ng of the recombinant pET28b oxygenation). For each inhibitor concentration 7 expression plasmid and the cells were grown overnight on measurements were carried out in triplicate. 8 agar plates containing antibiotics. A well-isolated bacterial Cellular ALOX Activity Assay. To test the impact of the 9 clone was picked using a wooden sterile toothpick and two novel ALOX inhibitors on the cellular activity of human 10 1 mL liquid cultures (LB medium with 50 μg/mL ALOX15 we employed L1236 human Hodgkin lymphoma 11 kanamycin/35 µg/mL chloramphenicol) were grown for 7 cell lines. These cells constitutively express native human 12 hours at 37° C. Aliquots of these pre-culture were added to ALOX15 at high levels41. In these experiments the cultured 13 a 50 mL main culture to reach an optical density (A600) of L1236 cells were pre-incubated in the absence or presence 14 0.10-0.15. Bacterial expression was carried out in 50 mL of different inhibitor concentrations (8b or ETYA) at 25 °C 15 Ultra Yield culture flasks (Thomson Instrument Company, for 5 min (0.25 mL of PBS with 0.1 mill. cells in total) and 16 Oceanside, USA) using the optimized Enpresso-Expression then the ALOX15 reaction was started by the addition of 25 17 system (Enpresso GmbH, Berlin, Germany). After µM of either AA or LA. After 10 min the reaction was 18 overnight incubation at 30°C the OD600 had reached stopped and the amounts of eicosanoids formed from the 19 values above 5 and expression of the recombinant his-tag two fatty acid substrates were quantified by HPLC. For LA 20 fusion proteins was induced by the addition of IPTG oxygenation the formation of 13-HODE was quantified. For reraching a final concentration of 1 mM. Then the cultures 21 AA oxygenation we separately evaluated the formation of were kept at 22 °C for 24 h at 230-250 rpm agitation. 22 12-HETE and 15-HETE as independent measures. The Bacteria were harvested, the resulting pellet was 23 ALOX activity of the incubation sample lacking the reconstituted in a total volume of 5 mL PBS and bacteria 24 inhibitor was set 100%. As suitable measure for the were lyzed by sonication using a digital sonifier (W-250D 25 substrate selectivity of the inhibitory effect we calculated Microtip Model 102, 50% maximal sonication amplitude; 26 the ratios of residual activities of LA oxygenation vs. AA Branson Ultraschall, Fürth, Germany). Cell debris was 27 oxygenation. spun down (15 min, 15,000 x g, 4°C) and the lysate 28 supernatants were employed as enzymes source for the Reversibility of inhibition. To test the reversibility of 29 inhibitor studies without further purification. inhibition a solution of rabbit ALOX15 (3.3 µM) was pre- 30 incubated in 1 mL phosphate buffer (pH 7.4) containing 31 ALOX Activity Assay (purified enzymes). The 300 µM (app. 100 molar excess) of the inhibitor for 2 min 32 oxygenation kinetics of linoleic acid [LA] or arachidonic at 5 oC. The excess of inhibitor was removed by gel- 33 acid [AA] (25 µM final concentrations) were assayed filtration using a 1 mL PD MidiTrap 25-G column (GE 34 spectrophotometrically measuring the increase in Healthcare, Marlborough, USA). For this purpose, 1 mL of 35 absorbance at 235 nm. The assay mixture was a 0.1 M protein-inhibitor mixture was applied and the column was 36 phosphate buffer, pH 7.4 containing various eluted with 2 mL of phosphate buffer (pH 7.4) according to 37 concentrations of inhibitors. For this purpose, compounds the recommendation of the manufacturer. The column was 38 tested were reconstituted in DMSO and serial dilutions washed and re-equilibrated with 50 mL of the elution were carried out so that from each dilution 1 µL was applied 39 buffer until no detectable traces of the inhibitor remained for the measurement. Purified rabbitALOX15 [5 µg, specific 40 in the eluate. The procedure was repeated 3-4 times until activity with LA - 25 s-1, electrophoretical purity (>98%)] or 41 the final ratio of a maximum of UV-absorbance at 318 nm human ALOX15 [5 µg, specific activity with LA - 10 s-1, 42 (inhibitor specific chromophore) to the absorbance of electrophoretical purity (>98%)] were pre incubated with 43 protein (280 nm) remained constant. The amount of LOX a testing compound for 1 min and the reaction was started 44 protein left and that of the inhibitor bound was estimated by the addition of the substrate. The linear part of the 45 spectrophotometricaly using molecular extinction kinetic progress curve was evaluated and the activity of the 46 coefficient ε=1,78 (mg-1 cm-1) for LOX. The residual solvent controls (DMSO) was set as 100 %. All 47 enzymatic activity was assayed spectrophotometrically measurements were carried out at the room temperature 48 comparing to the activity of the control (ALOX preparation as triplicates. 49 without inhibitor). 50 ALOX Activity Assay (bacterial lysate Inhibitor recovery. Purified rabbit ALOX15 (10 µg) was 51 supernatants).Since mouse Alox15 is difficult to purify pre-incubated with an aliquot of inhibitor 6-8 (10 µM final 52 these comparative experiments were carried out with concentration) in 0.5 mL PBS for 1 min and the reaction 53 crude enzyme preparations. For this purpose the was started by addition of 25 µM substrate. After 3 min the 54 recombinant enzymes were expressed in E. coli and reaction was quenched with 0.5 mL of ice-cold methanol, 55 bacterial lysate supernatants were used as enzyme source. the protein precipitate was spun down and aliquots of the 56 2 to 10 µL of bacterial cell lysate supernatants were pre- clear supernatant (300µL) were injected for reverse-phase 57 incubated with 2 µM of 8b at 25° C for 1 min and the ALOX HPLC for quantification of the remaining inhibitor. For 58 reaction was started by the addition of 25 µM AA or LA. 59 60 ACS Paragon Plus Environment Journal of Medicinal Chemistry Page 16 of 28 each inhibitor concentration measurements were carried program. The resulting systems contain nearly 200000 1 out in triplicate. atoms being about 21000 of them of the protein. The 2 remaining atoms represent water molecules and salt ions. 3 Molecular docking studies. Docking studies were carried All molecular dynamics (MD) simulations were run with 4 out with the program GOLD5.8.0.62 Concerning the 8a and AMBER 18 GPU (CUDA) version of the PMEMD package73- 5 8b inhibitors, those calculations were restricted to the 74. Initially, the systems were submitted to 22000 energy 6 binding cavity of monomer A of rabbit ALOX15 considering minimization steps using the steepest-descent method to the crystallographic x-ray structure of the dimeric enzyme avoid close contacts. In the first 6000 steps, harmonic 7 (PDB entry 2P0M)32 as the receptor. Before initiating the restraints were applied to the protein and substrate atoms 8 docking protocol, the ligand bound at the active site of the with a force constant of 5.0 kcal mol-1 Å-2 so only the solvent 9 crystal structure was removed. Hydrogen coordinates were and ions were relaxed. In the following 6000 steps, 10 generated with the H++ web-server63-64 using a pH 6.0 for harmonic restraints were applied to the protein backbone 11 titrable residues. By contrast, with AA and LA, the docking and the substrate heavy atoms with a force constant again 12 calculations were restricted to the binding cavity of 13 of 5.0 kcal mol-1 Å-2. In the last 10000 steps the whole monomer B of rabbit ALOX15 considering the relaxed system was kept free of restraints. After that, MD 14 ALOX15-8a(A)-empty(B) and ALOX15-8b(A)-empty(B) 15 simulations using periodic boundary conditions were complexes as the receptors. The binding site cavity used in carried out. The system was gradually heated from 0 K to 16 the docking runs is a 20 Å radius sphere centred around 17 300 K for a period of 200 ps. Next, an MD run of 1 ns, at the iron atom of monomer A, when inhibitor 8a or 8b act constant temperature and pressure (300 K, 1 bar), has been 18 as the ligand, or the iron atom of monomer B when AA or calculated to adjust the volume of the orthorhombic box 19 LA act as the ligand. The receptor was kept fixed but so that a density of around 1 g cm-3 is reached. Throughout 20 complete flexibility was given to the ligand in the the heating and the compressing, harmonic restraints were 21 conformational search. We activated the option in GOLD applied to the protein backbone and substrate heavy atoms 22 that considers the interactions of organic ligands with with a force constant of 5.0 kcal mol-1 Å-2 whereas no 23 metal ions in metalloenzymes but limiting the docking restraints were applied to the rest of the system. The 24 exploration to hexacoordinated geometries of iron. The temperature was controlled by Langevin dynamics 75, while 25 most efficient genetic algorithm was used to ensure an the pressure was adjusted by the Berendsen barostat76. 26 extensive search of the conformational space of all Then, an equilibration stage of 10 ns, at constant 27 substrates. To estimate the binding free energies of temperature (300 K) and volume, was carried out. Finally, 28 substrates the ChemScore fitness function has been a production period was run within the same isothermal- 29 selected. isochoric ensemble. A time step of 2 fs was used along the 30 Molecular dynamics (MD) simulations. The system has whole MD trajectory. All bonds and bends containing 31 been assembled using the recommended procedure by the hydrogen atoms were constrained by the SHAKE 32 AMBER program package.65 The ff14SB66 force field was algorithm.77 Non-bonding interactions have been 33 used for the protein atoms. The force field parameters for calculated with a cut-off of 9 Å. The best docking pose of 34 AA,67 LA68 and iron and its first coordination sphere42 8a and 8b into monomer A were used as the starting 35 (His361, His366, His541, His545, Ile 663 and OH-) were structure for the MD simulations of the ALOX15-8a(A)- 36 taken from previous works while the specific parameters of empty(B) and ALOX15-8b(A)-empty(B) complexes, 37 8a and 8b inhibitors were developed here. The calculations respectively. The length of the production period for those 38 to generate those specific parameters were carried out two MD trajectories was 200 ns for letting the system relax 39 following the standard protocol in AMBER with and adapt to the presence of the inhibitor. The last 40 Antechamber and Parmchk2 modules even though, as structures of those two MD simulations were then taken as 41 both inhibitors are far from common substrates used with receptors for docking AA and LA into monomer B of the 42 AMBER, a procedure developed by MacKerell et al.69 had ALOX15-8a(A)-empty(B) and ALOX15-8b(A)-emty(B) 43 to be employed in order to overcome some large dihedral complexes. Next, the best docking pose of AA into 44 penalties. As the source for those parameters, the GAFF265, monomer B was selected as the starting structure for the 45 70 library was used. The B3LYP/6-31G(d) level of theory was MD simulations of ALOX15-8a(A)-AA(B) and ALOX15- 46 employed to optimize the 8a and 8b inhibitors’ structures 8b(A)-AA(B) as well as the best docking pose of LA that and their atomics charges were assigned using Merz- was taken to initiate the simulation of the ALOX15-8a(A)- 47 Kollman RESP procedure.71 Besides, the protonation states LA(B) and ALOX15-8b(A)-LA(B) complexes. In these cases, 48 for all ligands were established by hand to ensure that they a production period of 100 ns was run except for the 49 match with the protonation state in physiological ALOX15-8a(A)-LA(B) system that needed additional 100 ns 50 conditions. to equilibrate. For the sake of comparison, we have also 51 included the results of the MD simulation of the ALOX15- 52 All MD simulations followed the same protocol, the only AA(B) complex (that is, without inhibitor) from our 53 differences are found in the length of the production previous study.42 Also, a 100 ns MD trajectory for the 54 period and the starting structures. After combing enzyme ALOX15-LA(B) system has been calculated here to 55 and substrate files, the different protein-substrate compare with the simulations including an inhibitor and 56 complexes were solvated with an orthorhombic box of pre- LA. The Leu183Glu+Leu192Glu mutant has been built using 57 equilibrated TIP3P72 waters and their total charge was the CHIMERA77 program from the wild-type ALOX15 dimer 58 neutralized by adding sodium cations using the tLeap 59 60 ACS Paragon Plus Environment Page 17 of 28 Journal of Medicinal Chemistry model. With the aim of relaxing the system in the presence REFERENCES 1 of the double mutation, the length of the MD production 1. Cyrus, T.; Witztum, J. L.; Rader, D. J.; Tangirala, R.; Fazio, 2 period was 200 ns. Analysis of the MD simulations was S.; Linton, M. F.; Funk, C. D., Disruption of the 12/15-lipoxygenase 3 carried out with AmberTools18, whereas visualization of gene diminishes atherosclerosis in apo E-deficient mice. J Clin Invest 4 those trajectories was performed with VMD78 and USCF 1999, 103 (11), 1597-1604. 2. Pidgeon, G. P.; Lysaght, J.; Krishnamoorthy, S.; Reynolds, J. 5 CHIMERA79 programs. The cavity surfaces were calculated V.; O'Byrne, K.; Nie, D.; Honn, K. V., Lipoxygenase metabolism: roles 6 with the CavityPlus web server.80 in tumor progression and survival. Cancer Metastasis Rev 2007, 26 (3- 7 4), 503-524. 8 SUPPORTING INFORMATION 3. Cimen, I.; Tuncay, S.; Banerjee, S., 15-Lipoxygenase-1 9  Dose response curves for all inhibitors tested, Michaelis– expression suppresses the invasive properties of colorectal carcinoma cell lines HCT-116 and HT-29. Cancer Sci 2009, 100 (12), 2283-2291. 10 Menten plots at different inhibitor concentrations, 4. Serhan, C. N.; Petasis, N. A., Resolvins and protectins in 11 inhibitor stability assays, impact of inhibitor on product inflammation resolution. Chem Rev 2011, 111 (10), 5922-5943. 12 specificity of ALOX15, docking studies and MD 5. Yoo, S.; Lim, J. Y.; Hwang, S. W., Resolvins: Endogenously- 13 simulations of enzyme-inhibitor complexes, spectral and generated potent painkilling substances and their therapeutic analytical data (1H- and 13C-NMR spectra, IR-spectra, UV- perspectives. Curr Neuropharmacol 2013, 11 (6), 664-676. 14 6. Ivanov, I.; Kuhn, H.; Heydeck, D., Structural and functional spectra, HRMS spectra, GC-MS- and HPLC-data) (PDF) 15 biology of arachidonic acid 15-lipoxygenase-1 (ALOX15). Gene 2015,  Molecular string formulas (CSV) 16 573 (1), 1-32. 17 AUTHOR INFORMATION 7. Pace, S.; Pergola, C.; Dehm, F.; Rossi, A.; Gerstmeier, J.; 18 Troisi, F.; Pein, H.; Schaible, A. M.; Weinigel, C.; Rummler, S.; Corresponding Author Northoff, H.; Laufer, S.; Maier, T. J.; Radmark, O.; Samuelsson, B.; 19 Koeberle, A.; Sautebin, L.; Werz, O., Androgen-mediated sex bias 20 *Igor Ivanov − Lomonosov Institute of Fine Chemical impairs efficiency of leukotriene biosynthesis inhibitors in males. J 21 Technologies, MIREA - Russian Technological University, Clin Invest 2017, 127 (8), 3167-3176. 22 Vernadskogo pr. 86, 119571 Moscow, Russia, orcid.org/0000- 8. Larsen, J. S.; Acosta, E. P., Leukotriene-receptor antagonists 0003-0543-2067; E-mail: igor_ivanov@gmx.de; Tel: +7 (495) and 5-lipoxygenase inhibitors in asthma. Ann Pharmacother 1993, 27 23 (7-8), 898-903. 246-05-55 (supl. 884) 24 9. Mochizuki, N.; Kwon, Y. G., 15-lipoxygenase-1 in the 25 Authors vasculature: expanding roles in angiogenesis. Circ Res 2008, 102 (2), 26 #These 143-145. authors equally contributed to the article 10. Zhao, L.; Grosser, T.; Fries, S.; Kadakia, L.; Wang, H.; Zhao, 27 Author Contributions J.; Falotico, R., Lipoxygenase and prostaglandin G/H synthase 28 cascades in cardiovascular disease. Expert Rev Clin Immunol 2006, 2 29 I.I., H.K., J.M.L. and A.G.L designed the study. A.G, A.Z, R.S (4), 649-658. 30 and V.A. performed synthesis and characterization of all 11. Kapralov, A. A.; Yang, Q.; Dar, H. H.; Tyurina, Y. Y.; 31 compounds. I.I., A.Z, and K.R.K. carried out expression of the Anthonymuthu, T. S.; Kim, R.; St. Croix, C. M.; Mikulska-Ruminska, K.; Liu, B.; Shrivastava, I. H.; Tyurin, V. A.; Ting, H.-C.; Wu, Y. L.; Gao, 32 recombinant enzymes. I.I, A.Z and K.R.K. performed inhibitor Y.; Shurin, G. V.; Artyukhova, M. A.; Ponomareva, L. A.; Timashev, P. 33 testing and determined reaction specificity of the enzyme S.; Domingues, R. M.; Stoyanovsky, D. A.; Greenberger, J. S.; 34 variants with polyenoic fatty acids. A.C. performed modelling Mallampalli, R. K.; Bahar, I.; Gabrilovich, D. I.; Bayır, H.; Kagan, V. E., and MD-simulations. I.I., A.G.L., J.M.L and H.K. drafted the Redox lipid reprogramming commands susceptibility of macrophages 35 manuscript and all authors contributed to prepare its final and microglia to ferroptotic death. Nat Chem Biol 2020, 16 (3), 278- 36 version. 290. 37 12. Kagan, V. E.; Mao, G.; Qu, F.; Angeli, J. P. F.; Doll, S.; Croix, 38 Notes C. S.; Dar, H. H.; Liu, B.; Tyurin, V. A.; Ritov, V. B.; Kapralov, A. A.; Authors declare no competing financial interest. Amoscato, A. A.; Jiang, J.; Anthonymuthu, T.; Mohammadyani, D.; 39 Yang, Q.; Proneth, B.; Klein-Seetharaman, J.; Watkins, S.; Bahar, I.; 40 ACKNOWLEDGMENT Greenberger, J.; Mallampalli, R. K.; Stockwell, B. R.; Tyurina, Y. Y.; 41 Conrad, M.; Bayır, H., Oxidized arachidonic and adrenic PEs navigate Deutsche Forschungsgemeinschaft (KU 961/14-1), Russian 42 cells to ferroptosis. Nat Chem Biol 2017, 13 (1), 81-90. Foundation for Basic Research (19-54-12002) and Russian 43 Ministry of Science and High Education (0706-2020-0019) 13. de Luca, C.; Olefsky, J. M., Inflammation and insulin 44 resistance. FEBS Lett 2008, 582 (1), 97-105. supported the experimental part of this work. We thank the 14. Sears, D. D.; Miles, P. D.; Chapman, J.; Ofrecio, J. M.; 45 Spanish “Ministerio de Ciencia, Innovación y Universidades” Almazan, F.; Thapar, D.; Miller, Y. I., 12/15-lipoxygenase is required for 46 (Grant CTQ2017-83745-P). We also acknowledge CSUC for the early onset of high fat diet-induced adipose tissue inflammation 47 computational facilities and Shared Science and training and insulin resistance in mice. PLoS One 2009, 4 (9), e7250. 48 Center for Collective Use of RTU MIREA. Authors also wish to 15. Lieb, D. C.; Brotman, J. J.; Hatcher, M. A.; Aye, M. S.; Cole, thank Prof. Alexander V. Aksenov from North Caucasus B. K.; Haynes, B. A.; Wohlgemuth, S. D.; Fontana, M. A.; Beydoun, H.; 49 Nadler, J. L.; Dobrian, A. D., Adipose tissue 12/15 lipoxygenase pathway 50 Federal University (Russian Federation) for his valuable in human obesity and diabetes. J Clin Endocrinol Metab 2014, 99 (9), 51 advises and comments regarding indole synthesis. Chemical E1713-E1420. synthesis of the inhibitors was carried out in the Ivanov lab in 52 16. Brinckmann, R.; Schnurr, K.; Heydeck, D.; Rosenbach, T.; Moscow. Expression and purification of the recombinant Kolde, G.; Kühn, H., Membrane translocation of 15-lipoxygenase in 53 enzymes, mutagenesis studies and the characterization hematopoietic cells is calcium-dependent and activates the oxygenase 54 experiments were carried out by I.I and his Russian colleagues activity of the enzyme. Blood 1998, 91 (1), 64-74. 55 during several sabbaticals in the Kühn lab in Berlin. In silico 17. Singh, N. K.; Rao, G. N., Emerging role of 12/15- 56 docking studies and MD simulations were performed in the Lipoxygenase (ALOX15) in human pathologies. Prog Lipid Res 2019, 57 73, 28-45. González-Lafont lab in Barcelona. 58 59 60 ACS Paragon Plus Environment Journal of Medicinal Chemistry Page 18 of 28 18. Ackermann, J. A.; Hofheinz, K.; Zaiss, M. M.; Kronke, G., 35. Ivanov, I.; Cruz, A.; Zhuravlev, A.; Di Venere, A.; Nicolai, E.; 1 The double-edged role of 12/15-lipoxygenase during inflammation and Stehling, S.; Lluch, J. M.; González-Lafont, À.; Kuhn, H., 2 immunity. Biochim Biophys Acta 2017, 1862 (4), 371-381. Conformational heterogeneity and cooperative effects of mammalian 19. Zheng, Y.; Liu, Y.; Karatas, H.; Yigitkanli, K.; Holman, T. R.; ALOX15. Int J Mol Sci 2021, 22 (6), 3285. 3 van Leyen, K., Contributions of 12/15-Lipoxygenase to Bleeding in the 36. Gaikwad, R.; Bobde, Y.; Ganesh, R.; Patel, T.; Rathore, A.; 4 Brain Following Ischemic Stroke. Adv Exp Med Biol 2019, 1161, 125-131. Ghosh, B.; Das, K.; Gayen, S., 2-Phenylindole derivatives as anticancer 5 20. Serhan, C. N.; Jain, A.; Marleau, S.; Clish, C.; Kantarci, A.; agents: synthesis and screening against murine melanoma, human 6 Behbehani, B.; Colgan, S. P.; Stahl, G. L.; Merched, A.; Petasis, N. A.; lung and breast cancer cell lines. Synth Commun 2019, 49 (17), 2258- Chan, L.; Van Dyke, T. E., Reduced inflammation and tissue damage 2269. 7 in transgenic rabbits overexpressing 15-lipoxygenase and endogenous 37. Aksenov, A. V.; Smirnov, A. N.; Aksenov, N. A.; Bijieva, A. 8 anti-inflammatory lipid mediators. J Immunol 2003, 171 (12), 6856- S.; Aksenova, I. V.; Rubin, M., Benzimidazoles and benzoxazoles via 9 6865. the nucleophilic addition of anilines to nitroalkanes. Org Biomol 10 21. Il Lee, S.; Zuo, X.; Shureiqi, I., 15-Lipoxygenase-1 as a tumor Chem 2015, 13 (14), 4289-4295. suppressor gene in colon cancer: is the verdict in? Cancer Metastasis 38. Pigerol, C.; de Cointet de Fillain, P.; Nanthavong, S.; Le 11 Rev 2011, 30 (3-4), 481-491. Blay, J. 2-Phenyl-indole derivatives and process for preparing the 12 22. Rai, G.; Joshi, N.; Jung, J. E.; Liu, Y.; Schultz, L.; Yasgar, A.; same. U.S. Patent 4,057,530, November 8, 1977. 13 Perry, S.; Diaz, G.; Zhang, Q.; Kenyon, V.; Jadhav, A.; Simeonov, A.; 39. Rai, G.; Joshi, N.; Perry, S.; Yasgar, A.; Schultz, L.; Jung, J. E.; 14 Lo, E. H.; van Leyen, K.; Maloney, D. J.; Holman, T. R., Potent and Liu, Y.; Terasaki, Y.; Diaz, G.; Kenyon, V.; Jadhav, A.; Simeonov, A.; 15 selective inhibitors of human reticulocyte 12/15-lipoxygenase as anti- van Leyen, K.; Holman, T. R.; Maloney, D. J., Discovery of ML351, a stroke therapies. J Med Chem 2014, 57 (10), 4035-4048. potent and selective inhibitor of human 15-lipoxygenase-1. In Probe 16 23. Ngu, K.; Weinstein, D. S.; Liu, W.; Langevine, C.; Combs, D. Reports from the NIH Molecular Libraries Program, National Center 17 W.; Zhuang, S.; Chen, X.; Madsen, C. S.; Harper, T. W.; Ahmad, S.; for Biotechnology Information (US): Bethesda (MD), 2010. 18 Robl, J. A., Pyrazole-based sulfonamide and sulfamides as potent 40. McGovern, S. L.; Caselli, E.; Grigorieff, N.; Shoichet, B. K., A 19 inhibitors of mammalian 15-lipoxygenase. Bioorg Med Chem Lett 2011, Common mechanism underlying promiscuous inhibitors from virtual 21 (14), 4141-4145. and high-throughput screening. J Med Chem 2002, 45 (8), 1712-1722. 20 24. Weinstein, D. S.; Liu, W.; Ngu, K.; Langevine, C.; Combs, D. 41. Claesson, H. E.; Griffiths, W. J.; Brunnström, A.; Schain, F.; 21 W.; Zhuang, S.; Chen, C.; Madsen, C. S.; Harper, T. W.; Robl, J. A., Andersson, E.; Feltenmark, S.; Johnson, H. A.; Porwit, A.; Sjöberg, J.; 22 Discovery of selective imidazole-based inhibitors of mammalian 15- Björkholm, M., Hodgkin Reed-Sternberg cells express 15- 23 lipoxygenase: highly potent against human enzyme within a cellular lipoxygenase-1 and are putative producers of eoxins in vivo: novel 24 environment. Bioorg Med Chem Lett 2007, 17 (18), 5115-5120. insight into the inflammatory features of classical Hodgkin 25. Sendobry, S. M.; Cornicelli, J. A.; Welch, K.; Bocan, T.; Tait, lymphoma. FEBS J 2008, 275 (16), 4222-4234. 25 B.; Trivedi, B. K.; Colbry, N.; Dyer, R. D.; Feinmark, S. J.; Daugherty, 42. Cruz, A.; Di Venere, A.; Mei, G.; Zhuravlev, A.; Golovanov, 26 A., Attenuation of diet-induced atherosclerosis in rabbits with a A.; Stehling, S.; Heydeck, D.; Lluch, J. M.; González-Lafont, À.; Kuhn, 27 highly selective 15-lipoxygenase inhibitor lacking significant H.; Ivanov, I., A role of Gln596 in fine-tuning mammalian ALOX15 28 antioxidant properties. Br J Pharmacol 1997, 120 (7), 1199-1206. specificity, protein stability and allosteric properties. Biochim Biophys 26. Barvian, N. C.; O'brian, P. M.; Patt, W. C.; Picard, J. A.; Acta Mol Cell Biol Lipids 2020, 1865 (7), 158680. 29 Sliskovic, D. R. 1,2,4-trisubstituted benzenes as inhibitors of 15- 43. Kühn, H.; Holzhütter, H. G.; Schewe, T.; Hiebsch, C.; 30 lipoxygenase. Patent WO 096298 A2, 20 December, 2001. Rapoport, S. M., The mechanism of inactivation of lipoxygenases by 31 27. Connor, D.; Roark, W.; Sorenson, R. Indole and acetylenic fatty acids. Eur J Biochem 1984, 139 (3), 577-583. 32 benzimidazole 15-lipoxygenase inhibitors. U.S. Patent 0038943 A1, 44. Ivanov, I.; Shang, W.; Toledo, L.; Masgrau, L.; Svergun, D. February 26, 2004. I.; Stehling, S.; Gomez, H.; Di Venere, A.; Mei, G.; Lluch, J. M.; 33 28. Eleftheriadis, N.; Neochoritis, C. G.; Leus, N. G.; van der Skrzypczak-Jankun, E.; González-Lafont, À.; Kuhn, H., Ligand- 34 Wouden, P. E.; Dömling, A.; Dekker, F. J., Rational development of a induced formation of transient dimers of mammalian 12/15- 35 potent 15-lipoxygenase-1 inhibitor with in vitro and ex vivo anti- lipoxygenase: a key to allosteric behavior of this class of enzymes? 36 inflammatory properties. J Med Chem 2015, 58 (19), 7850-7862. Proteins 2012, 80 (3), 703-712. 37 29. Weinstein, D. S.; Liu, W.; Gu, Z.; Langevine, C.; Ngu, K.; 45. Kuhn, H.; Heydeck, D.; Sprecher, H., On the mechanistic Fadnis, L.; Combs, D. W.; Sitkoff, D.; Ahmad, S.; Zhuang, S.; Chen, X.; reasons for the dual positional specificity of the reticulocyte 38 Wang, F. L.; Loughney, D. A.; Atwal, K. S.; Zahler, R.; Macor, J. E.; lipoxygenase. Biochim Biophys Acta 1991, 1081 (2), 129-134. 39 Madsen, C. S.; Murugesan, N., Tryptamine and homotryptamine- 46. Vogel, R.; Jansen, C.; Roffeis, J.; Reddanna, P.; Forsell, P.; 40 based sulfonamides as potent and selective inhibitors of 15- Claesson, H. E.; Kuhn, H.; Walther, M., Applicability of the triad 41 lipoxygenase. Bioorg Med Chem Lett 2005, 15 (5), 1435-1440. concept for the positional specificity of mammalian lipoxygenases. J 30. Ludwig, P.; Holzhutter, H. G.; Colosimo, A.; Silvestrini, M. Biol Chem 2010, 285 (8), 5369-5376. 42 C.; Schewe, T.; Rapoport, S. M., A kinetic model for lipoxygenases 47. Freedman, C.; Tran, A.; Tourdot, B. E.; Kalyanaraman, C.; 43 based on experimental data with the lipoxygenase of reticulocytes. Eur Perry, S.; Holinstat, M.; Jacobson, M. P.; Holman, T. R., Biosynthesis 44 J Biochem 1987, 168 (2), 325-337. of the maresin intermediate, 13S,14S-Epoxy-DHA, by human 15- 45 31. Shang, W.; Ivanov, I.; Svergun, D. I.; Borbulevych, O. Y.; lipoxygenase and 12-lipoxygenase and its regulation through negative Aleem, A. M.; Stehling, S.; Jankun, J.; Kuhn, H.; Skrzypczak-Jankun, allosteric modulators. Biochemistry 2020, 59 (19), 1832-1844. 46 E., Probing dimerization and structural flexibility of mammalian 48. Meng, H.; Dai, Z.; Zhang, W.; Liu, Y.; Lai, L., Molecular 47 lipoxygenases by small-angle X-ray scattering. J Mol Biol 2011, 409 (4), mechanism of 15-lipoxygenase allosteric activation and inhibition. 48 654-668. Phys Chem Chem Phys 2018, 20 (21), 14785-14795. 49 32. Choi, J.; Chon, J. K.; Kim, S.; Shin, W., Conformational 49. Meng, H.; McClendon, C. L.; Dai, Z.; Li, K.; Zhang, X.; He, 50 flexibility in mammalian 15S-lipoxygenase: Reinterpretation of the S.; Shang, E.; Liu, Y.; Lai, L., Discovery of novel 15-lipoxygenase crystallographic data. Proteins 2008, 70 (3), 1023-1032. activators to shift the human arachidonic acid metabolic network 51 33. Wecksler, A. T.; Kenyon, V.; Deschamps, J. D.; Holman, T. toward inflammation resolution. J Med Chem 2016, 59 (9), 4202-4209. 52 R., Substrate specificity changes for human reticulocyte and epithelial 50. Zhou, J.; Li, Q.; Wu, M.; Chen, C.; Cen, S., Progress in the 53 15-lipoxygenases reveal allosteric product regulation. Biochemistry rational design for polypharmacology drug. Curr Pharm Des 2016, 22 54 2008, 47 (28), 736473-75. (21), 3182-3189. 34. Wecksler, A. T.; Jacquot, C.; van der Donk, W. A.; Holman, 51. Yuan, C.; Rieke, C. J.; Rimon, G.; Wingerd, B. A.; Smith, W. 55 T. R., Mechanistic investigations of human reticulocyte 15- and L., Partnering between monomers of cyclooxygenase-2 homodimers. 56 platelet 12-lipoxygenases with arachidonic acid. Biochemistry 2009, 48 Proc Natl Acad Sci U S A 2006, 103 (16), 6142-6147. 57 (26), 6259-6267. 52. Hafner, A. K.; Cernescu, M.; Hofmann, B.; Ermisch, M.; 58 Hornig, M.; Metzner, J.; Schneider, G.; Brutschy, B.; Steinhilber, D., 59 60 ACS Paragon Plus Environment Page 19 of 28 Journal of Medicinal Chemistry Dimerization of human 5-lipoxygenase. Biol Chem 2011, 392 (12), 1097- Salomon-Ferrer, R.; Swails, J.; Walker, R. C.; Wang, J.; Wei, H.; Wolf, 1 1111. R. M.; Wu , X.; Xiao, L.; York, D. M.; Kollman, P. A. AMBER 2018, 2 53. Tsai, W.-C.; Aleem, A. M.; Whittington, C.; Cortopassi, W. University of California: San Francisco, 2018. A.; Kalyanaraman, C.; Baroz, A.; Iavarone, A. T.; Skrzypczak-Jankun, 66. Maier, J. A.; Martinez, C.; Kasavajhala, K.; Wickstrom, L.; 3 E.; Jacobson, M. P.; Offenbacher, A. R.; Holman, T., Mutagenesis, Hauser, K. E.; Simmerling, C., ff14SB: Improving the accuracy of 4 hydrogen–deuterium exchange, and molecular docking investigations protein side chain and backbone parameters from ff99SB. J Chem 5 establish the dimeric interface of human platelet-type 12- Theory Comput 2015, 11 (8), 3696-3713. 6 lipoxygenase. Biochemistry 2021, 60 (10), 802-812. 67. Tosco, P., A mechanistic hypothesis for the aspirin-induced 54. Burrall, B. A.; Cheung, M.; Chiu, A.; Goetzl, E. J., Enzymatic switch in lipid mediator production by cyclooxygenase-2. J Am Chem 7 properties of the 15-lipoxygenase of human cultured keratinocytes. J Soc 2013, 135 (28), 10404-10410. 8 Invest Dermatol 1988, 91 (4), 294-297. 68. Ivanov, I.; Golovanov, A. B.; Ferretti, C.; Canyelles-Niño, M.; 9 55. Di Venere, A.; Horn, T.; Stehling, S.; Mei, G.; Masgrau, L.; Heydeck, D.; Stehling, S.; Lluch, J. M.; González-Lafont, À.; Kühn, H., 10 González-Lafont, À.; Kuhn, H.; Ivanov, I., Role of Arg403 for Mutations of triad determinants changes the substrate alignment at thermostability and catalytic activity of rabbit 12/15-lipoxygenase. the catalytic center of human ALOX5. ACS Chem Biol 2019, 14 (12), 11 Biochim Biophys Acta 2013, 1831 (6), 1079-1088. 2768-2782. 12 56. Bryant, R. W.; Bailey, J. M.; Schewe, T.; Rapoport, S. M., 69. Guvench, O.; MacKerell, A. D., Jr., Automated 13 Positional specificity of a reticulocyte lipoxygenase. Conversion of conformational energy fitting for force-field development. J Mol 14 arachidonic acid to 15-S-hydroperoxy-eicosatetraenoic acid. J Biol Model 2008, 14 (8), 667-679. 15 Chem 1982, 257 (11), 6050-6055. 70. Wang, J.; Wolf, R. M.; Caldwell, J. W.; Kollman, P. A.; Case, 57. Kuhn, H.; Sprecher, H.; Brash, A. R., On singular or dual D. A., Development and testing of a general amber force field. J 16 positional specificity of lipoxygenases. The number of chiral products Comput Chem 2004, 25 (9), 1157-1174. 17 varies with alignment of methylene groups at the active site of the 71. Bayly, C. I.; Cieplak, P.; Cornell, W. D.; Kollman, P. A., A 18 enzyme. J Biol Chem 1990, 265 (27), 16300-16305. Well-behaved electrostatic potential based method using charge 19 58. Chaturvedi, S. K.; Ma, J.; Brown, P. H.; Zhao, H.; Schuck, P., restraints for deriving atomic charges - the Resp model. J Phys Chem- Measuring macromolecular size distributions and interactions at high Us 1993, 97 (40), 10269-10280. 20 concentrations by sedimentation velocity. Nat Commun 2018, 9 (1), 72. Jorgensen, W. L.; Chandrasekhar, J.; Madura, J. D.; Impey, 21 4415. R. W.; Klein, M. L., Comparison of simple potential functions for 22 59. Chaturvedi, S. K.; Sagar, V.; Zhao, H.; Wistow, G.; Schuck, simulating liquid water. J Chem Phys 1983, 79 (2), 926-935. 23 P., Measuring ultra-weak protein self-association by non-ideal 73. Salomon-Ferrer, R.; Gotz, A. W.; Poole, D.; Le Grand, S.; 24 sedimentation velocity. J Am Chem Soc 2019, 141 (7), 2990-2996. Walker, R. C., Routine microsecond molecular dynamics simulations 60. Kozlov, N.; Humeniuk, L.; Ufer, C.; Ivanov, I.; Golovanov, with AMBER on GPUs. 2. Explicit solvent particle Mesh Ewald. J Chem 25 A.; Stehling, S.; Heydeck, D.; Kuhn, H., Functional characterization of Theory Comput 2013, 9 (9), 3878-3888. 26 novel ALOX15 orthologs representing key steps in mammalian 74. Le Grand, S.; Gotz, A. W.; Walker, R. C., SPFP: Speed 27 evolution supports the Evolutionary Hypothesis of reaction without compromise-A mixed precision model for GPU accelerated 28 specificity. Biochim Biophys Acta Mol Cell Biol Lipids 2019, 1864 (3), molecular dynamics simulations. Comput Phys Commun 2013, 184 (2), 372-385. 374-380. 29 61. Ivanov, I.; Di Venere, A.; Horn, T.; Scheerer, P.; Nicolai, E.; 75. Leach, A. R., Molecular Modeling: Principles and 30 Stehling, S.; Richter, C.; Skrzypczak-Jankun, E.; Mei, G.; Maccarrone, Applications Eddison Wesley Longman Limited: Essex, England, 1996; 31 M.; Kuhn, H., Tight association of N-terminal and catalytic subunits p 595. 32 of rabbit 12/15-lipoxygenase is important for protein stability and 76. Berendsen, H. J. C.; Postma, J. P. M.; Vangunsteren, W. F.; catalytic activity. Biochim Biophys Acta 2011, 1811 (12), 1001-1010. Dinola, A.; Haak, J. R., Molecular-dynamics with coupling to an 33 62. Jones, G.; Willett, P.; Glen, R. C.; Leach, A. R.; Taylor, R., external bath. J Chem Phys 1984, 81 (8), 3684-3690. 34 Development and validation of a genetic algorithm for flexible 77. Ryckaert, J.; Ciccotti, G.; Berendsen, H., Numerical- 35 docking. J Mol Biol 1997, 267 (3), 727-748. integration of cartesian equations of motion of a system with 36 63. Anandakrishnan, R.; Aguilar, B.; Onufriev, A. V., H++3.0: contraints - molecular dynamics of n-alkanes. J Comput Phys 1977, 23 37 automating pK prediction and the preparation of biomolecular (3), 327-341. structures for atomistic molecular modeling and simulations. Nucleic 78. Humphrey, W.; Dalke, A.; Schulten, K., VMD: visual 38 Acids Res 2012, 40 (W1), W537-W541. molecular dynamics. J Mol Graph 1996, 14 (1), 33-8. 39 64. Gordon, J. C.; Myers, J. B.; Folta, T.; Shoja, V.; Heath, L. S.; 79. Pettersen, E. F.; Goddard, T. D.; Huang, C. C.; Couch, G. S.; 40 Onufriev, A., H++: a server for estimating pK(a)s and adding missing Greenblatt, D. M.; Meng, E. C.; Ferrin, T. E., UCSF chimera - A 41 hydrogens to macromolecules. Nucleic Acids Res 2005, 33, W368- visualization system for exploratory research and analysis. J Comput W371. Chem 2004, 25 (13), 1605-1612. 42 65. Case, D. A.; Ben-Shalom, I. Y.; Brozell, S. R.; Cerutti, D. S.; 80. Xu, Y.; Wang, S.; Hu, Q.; Gao, S.; Ma, X.; Zhang, W.; Shen, 43 Cheatham, I., T.E.; Cruzeiro, V. W. D.; Darden, T. A.; Duke, R. E.; Y.; Chen, F.; Lai, L.; Pei, J., CavityPlus: a web server for protein cavity 44 Ghoreishi, D.; Gilson, M. K.; Gohlke, H.; Goetz, A. W.; Greene, D.; detection with pharmacophore modelling, allosteric site 45 Harris, R.; Homeyer, N.; Izadi, S.; Kovalenko, A.; Kurtzman, T.; Lee, T. identification and covalent ligand binding ability prediction. Nucleic S.; LeGrand, S.; Li, P.; Lin, C.; Liu, J.; Luchko, T.; Luo, R.; Mermelstein, Acids Res 2018, 46 (W1), W374-W379. 46 D. J.; Merz, K. M.; Miao, Y.; Monard, G.; Nguyen, C.; Nguyen, H.; 47 Omelyan, I.; Onufriev, A.; Pan, F.; Qi, R.; Roe, D. R.; Roitberg, A.; 48 Sagui, C.; Schott-Verdugo, S.; Shen, J.; Simmerling, C. L.; Smith, J.; 49 50 51 52 53 54 55 56 57 58 59 60 ACS Paragon Plus Environment Journal of Medicinal Chemistry Page 20 of 28 Table of Contents graphic 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 ACS Paragon Plus Environment 20 Page 21 of 28 Journal of Medicinal Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 Figure 1. Selected structures of the substrate specific and non-specific ALOX15 inhibitors 35 133x111mm (300 x 300 DPI) 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 ACS Paragon Plus Environment Journal of Medicinal Chemistry Page 22 of 28 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 Scheme 1. Synthetic Route of compounds 6-8. Reagents and conditions: (i) 10% Pd/C, HCOONH4, EtOH, 30 reflux; (ii) PPA, 130 oC, 0.5h; (iii) AcOH, EtOH, reflux; (iv) ZnCl2,180 oC, 2h; (v) ClSO2NCO, HN(C2H5)2, 31 isobuthyl-, allyl - or octylalcohol, rt. 32 33 226x160mm (300 x 300 DPI) 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 ACS Paragon Plus Environment Page 23 of 28 Journal of Medicinal Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 Figure 2. Dose response curves for inhibition of LA- and AA-oxygenase activities of mammalian ALOX15 24 orthologs. The LA- and AA oxygenase activities of pure recombinant rabbit (A) and pure recombinant human 25 (B) ALOX15 orthologs were assayed in the presence or absence of compound 8b. Spectrophotometric assay were performed a described in Experimental Section and half-maximal inhibitory concentrations (IC50) were 26 determined. (C) Effect of the detergent on inhibitory potency of 8b in rabbit ALOX15 assay. 27 Spectrophotometric assay was performed a described in Experimental Section. (D) Inhibition of different 28 mammalian ALOX15 orthologs by 8b using HPLC-based assay. The assay was performed a described in 29 Experimental Section .The ALOX activity of the incubation sample lacking the inhibitor was set 100%. 30 Finally, we calculated the ratios of residual activities of LA oxygenation vs. AA oxygenation (formation of 13- 31 HODE vs. 15-HETE) as suitable measure for the substrate selectivity of the inhibitory effect. (E) Inhibition of the intracellular human ALOX15 by 8b using L1236 Hodgkin lymphoma cells. The assay was performed a 32 described in Experimental Section. The ALOX activity of the incubation sample lacking the inhibitor was set 33 100%. As suitable measure for the substrate selectivity of the inhibitory effect of 8b we calculated the ratios 34 of residual activities of LA oxygenation vs. AA oxygenation (formation of 13-HODE vs. 15-HETE). (F) 35 Inhibition of the intracellular human ALOX15 by 5µM ETYA using L1236 Hodgkin lymphoma cells. The assay 36 was performed under strictly comparable con-ditions as described for compound 8b. All data represent 37 means ± SEM of n=3 measurements. 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 ACS Paragon Plus Environment Journal of Medicinal Chemistry Page 24 of 28 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 Figure 3. Mechanism of ALOX15 inhibition. (A-F) Eadie-Hofstee plots demonstrating the effects of 6c, 8b and 32 8a on rabbit ALOX15 when incubated with 25 µM linoleic (A, C, E) or arachidonic (B, D, F) acids. Incubations 33 were repeated at different inhibitor concen-trations. All data points represent means ± SEM with n=3. (G) On average 1 molecule of 8b binds to 1 molecule of rabbit ALOX15. To estimate the rALOX15:8b ratio 34 molecular extinction coefficient of 1.78 mg-1cm-1 (280 nm) for the enzyme was used. 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 ACS Paragon Plus Environment Page 25 of 28 Journal of Medicinal Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 Figure 4. Leu183Glu+Leu192Glu mutant interface (A) vs. that of the wild-type enzyme (B). (C) Inhibition of 46 linoleate and arachidonate oxygenase activity of rabbit ALOX15 L183E+L192E mutant (10 µg purified 47 enzyme, 15 µM LA) in the presence of 8a and 8b. Spectrophotometric assays were performed as described in Experimental section. PDB ID 2P0M of crystal structures of rabbit ALOX15 was used as starting points for 48 molecular modeling. 49 50 51 52 53 54 55 56 57 58 59 60 ACS Paragon Plus Environment Journal of Medicinal Chemistry Page 26 of 28 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 Figure 5. Effect of inhibitor on the substrate binding. (A) Most representative binding mode of LA in the 28 ALOX15-8a(A)-LA(B) (in green) and ALOX15-8b(A)-LA(B) (in purple) complexes. (B) Most representative binding mode of AA in the ALOX15-8a(A)-AA(B) (in green) and ALOX15-8b(A)-AA(B) (in purple) complexes. 29 The centroids of the most representative cluster for LA of the ALOX15-8a(A)-LA(B) and ALOX15-8b(A)-LA(B) 30 complex simulations and for AA of the ALOX15-8a(A)-AA(B) and ALOX15-8b(A)-AA(B) complex simulations 31 are depicted. The clustering has been carried out using an RMSD of 0.5 Å for the heavy atoms of the LA 32 (AA) substrate. (C) Cavity surfaces of monomer B in ALOX15-8a(A)-LA(B) and (D) ALOX15-8b(A)-LA(B) 33 complexes. Plots were carried out with the CavityPlus web server. 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 ACS Paragon Plus Environment Page 27 of 28 Journal of Medicinal Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 Figure 6. Schematic representation of the mechanism of proposed for action of 8b. 24 150x73mm (300 x 300 DPI) 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 ACS Paragon Plus Environment Journal of Medicinal Chemistry Page 28 of 28 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 Table of Contents graphic 39 40 50x49mm (400 x 400 DPI) 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 ACS Paragon Plus Environment