International Journal of Molecular Sciences Article Functional Characterization of Mouse and Human Arachidonic Acid Lipoxygenase 15B (ALOX15B) Orthologs and of Their Mutants Exhibiting Humanized and Murinized Reaction Specificities Kumar R. Kakularam 1,† , Miquel Canyelles-Niño 2,3,† , Xin Chen 1 , José M. Lluch 2,4 , Àngels González-Lafont 2,4 and Hartmut Kuhn 1, * 1 Department of Biochemistry, Charité–Universitätsmedizin Berlin, Corporate Member of Freie Universität Berlin and Humboldt Universität zu Berlin, D-10117 Berlin, Germany; kumar.1416@gmail.com (K.R.K.); xin.chen2@charite.de (X.C.) 2 Departament de Química, Universitat Autònoma de Barcelona, 08193 Barcelona, Spain; miquel.canyelles@uab.cat (M.C.-N.); josemaria.lluch@uab.cat (J.M.L.); angels.gonzalez@uab.cat (À.G.-L.) 3 Arquebio S.L., 08005 Barcelona, Spain 4 Institut de Biotecnologia i de Biomedicina (IBB), Universitat Autònoma de Barcelona, 08193 Barcelona, Spain * Correspondence: hartmut.kuehn@charite.de; Tel.: +49-30-450528040 † These authors contributed equally to this work. Abstract: The arachidonic acid lipoxygenase 15B (ALOX15B) orthologs of men and mice form differ- ent reaction products when arachidonic acid is used as the substrate. Tyr603Asp+His604Val double mutation in mouse arachidonic acid lipoxygenase 15b humanized the product pattern and an inverse mutagenesis strategy murinized the specificity of the human enzyme. As the mechanistic basis for these functional differences, an inverse substrate binding at the active site of the enzymes has been suggested, but experimental proof for this hypothesis is still pending. Here we expressed wildtype mouse and human arachidonic acid lipoxygenase 15B orthologs as well as their humanized and Citation: Kakularam, K.R.; murinized double mutants as recombinant proteins and analyzed the product patterns of these Canyelles-Niño, M.; Chen, X.; Lluch, enzymes with different polyenoic fatty acids. In addition, in silico substrate docking studies and J.M.; González-Lafont, À.; Kuhn, H. Functional Characterization of molecular dynamics simulation were performed to explore the mechanistic basis for the distinct Mouse and Human Arachidonic Acid reaction specificities of the different enzyme variants. Wildtype human arachidonic acid lipoxygenase Lipoxygenase 15B (ALOX15B) 15B converted arachidonic acid and eicosapentaenoic acid to their 15-hydroperoxy derivatives but the Orthologs and of Their Mutants Asp602Tyr+Val603His exchange murinized the product pattern. The inverse mutagenesis strategy in Exhibiting Humanized and mouse arachidonic acid lipoxygenase 15b (Tyr603Asp+His604Val exchange) humanized the product Murinized Reaction Specificities. Int. pattern with these substrates, but the situation was different with docosahexaenoic acid. Here, J. Mol. Sci. 2023, 24, 10046. https:// Tyr603Asp+His604Val substitution in mouse arachidonic acid lipoxygenase 15b also humanized the doi.org/10.3390/ijms241210046 specificity but the inverse mutagenesis (Asp602Tyr+Val603His) did not murinize the human enzyme. Academic Editor: Claudiu With linoleic acid Tyr603Asp+His604Val substitution in mouse arachidonic acid lipoxygenase 15b T. Supuran humanized the product pattern but the inverse mutagenesis in human arachidonic acid lipoxygenase 15B induced racemic product formation. Amino acid exchanges at critical positions of human and Received: 16 March 2023 mouse arachidonic acid lipoxygenase 15B orthologs humanized/murinized the product pattern Revised: 6 June 2023 with C20 fatty acids, but this was not the case with fatty acid substrates of different chain lengths. Accepted: 8 June 2023 Published: 12 June 2023 Asp602Tyr+Val603His exchange murinized the product pattern of human arachidonic acid lipoxy- genase 15B with arachidonic acid, eicosapentaenoic acid, and docosahexaenoic acid. An inverse mutagenesis strategy on mouse arachidonic acid lipoxygenase 15b (Tyr603Asp+His604Val exchange) did humanize the reaction products with arachidonic acid and eicosapentaenoic acid, but not with Copyright: © 2023 by the authors. docosahexaenoic acid. Licensee MDPI, Basel, Switzerland. This article is an open access article Keywords: eicosanoids; lipids; metabolism; fatty acids; inflammation; atherosclerosis distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/). Int. J. Mol. Sci. 2023, 24, 10046. https://doi.org/10.3390/ijms241210046 https://www.mdpi.com/journal/ijms Int. J. Mol. Sci. 2023, 24, 10046 2 of 25 1. Introduction In the human genome, six different arachidonate lipoxygenase (ALOX) genes have been detected [1] and these genes encode for six functionally distinct ALOX isoforms). In the reference genome of mice, an orthologous enzyme exists for each human ALOX isoform, but in addition a functional Aloxe12 gene [1,2] encoding for the Aloxe12 has been identified [3]. In humans, the ALOXE12 gene is a corrupted pseudogene [1,2]. The human ALOX15B gene is located in a common ALOX gene cluster present on the short arm of chromosome 17 [2]. The mouse Alox15b gene is located in a similar Alox gene cluster that is localized in a synthetic region on chromosome 11 [1,2]. ALOX-isoforms are lipid peroxidizing enzymes that convert polyunsaturated fatty acids to the corresponding hydroperoxides [4,5]. In cellular systems these peroxides are rapidly reduced to the corresponding alcohols [6]. Alternatively, the peroxides are further converted to more complex secondary ALOX products such as pro-inflammatory leukotrienes [7] or anti-inflammatory resolvins, maresins, and protectins [8,9], which are collectively named specialized pro-resolving mediators (SPMs). Although the biosynthetic pathways of SPMs are rather diverse and although the SPM-induced intracellular signaling mechanisms are far from clear [9], these compounds exhibit a large number of interesting bioactivities [9]. The majority of ALOX isoforms prefer free polyenoic fatty acids as the substrate and under in vivo conditions the formation of ALOX products is limited by the liberation of the polyenoic fatty acids (PUFAs) from cellular ester lipids [10]. However, some isoforms, in particular human ALOX15 and human ALOX15B accept esterified polyenoic fatty acids as the substrate even if they are presented in complex lipid-protein-assemblies, such as nanodiscs [11,12], biomembranes [13–15], and different classes of lipoproteins [16–18]. In 1997, human ALOX15B was first described as AA 15-lipoxygenating enzyme, which is high level expressed in human skin, in the prostate, in the lungs and in the cornea [19]. The enzyme was cloned from hair routes and has been implicated in hair growth and skin development [19]. Later experiments indicated that the enzyme is also expressed in human monocytes when these cells are differentiated in vitro to macrophages. Stimulation of these cells with the Th2 cytokines interleukins-4 and -13 upregulated the expression of the enzyme [20]. When human peripheral monocytes are long-term stimulated with toll-like receptor-2/4 agonists, ALOX15B expression is also upregulated and this treatment induced a switch in the lipid mediator profile produced by these cells [21]. Despite these findings the biological relevance of ALOX15B orthologs is far from clear. A recent review nicely summarizes the potential biological functions of human ALOX15B, but this paper also indicates that there is no unifying concept for the biological role of ALOX15B [22]. Mouse Alox15b was first described in 1991 in phorbol ester-treated mouse skin [23]. This enzyme, which was later on cloned and expressed as recombinant protein [24] con- verted exogenous arachidonic acid to 8-HETE and it was therefore originally called mouse 8-LOX [23]. Today we know that this enzyme constitutes the mouse ortholog of hu- man ALOX15B and it should therefore be called Alox15b. Mouse and human ALOX15B share a high (78%) degree of amino acid identity and the two ALOX15B genes have similar structures [1]. They both span some 10 Mbp and consist of 14 exons and 13 in- trons. Thus, from a structural point of view, the two enzymes are very similar. On the other hand, the two enzymes exhibit different catalytic properties, and these data suggest that their biological functions in mice and men might be different. One of the most striking differences between the two ALOX15B orthologs is their different product pattern when arachidonic acid is sued as the substrate. Human ALOX15B [19,25] catalyzes the stereospe- cific oxygenation of AA to 15S-HpETE (15S-hydroperoxy-5Z,8Z,11Z,13E-eicosatetraenoic acid). In contrast, mouse Alox15b [23] converts the same substrate to 8S-HpETE (8S- hydroperoxy-5Z,9,9E,11Z,14Z-eicosatetraenoid acid). Interestingly, such differences in the product mixture of the two enzymes were not observed when complex substrates such as AA-containing phospholipids and/or AA containing phospholipid loaded nanodiscs were used as the substrate [11]. In fact, with these substrates 15-HETE was identified as the major oxygenation product for mouse and human ALOX15B orthologs [11]. Int. J. Mol. Sci. 2023, 24, 10046 3 of 25 The molecular basis for the differential product patterns of mouse and human ALOX15B has been explored and in vitro mutagenesis studies indicated that the Tyr603Asp+His604Val mutant of mouse Alox15b formed similar reaction products as wildtype human ALOX15B [25]. An inverse mutagenesis strategy (Asp602Tyr+Val603His exchange in human ALOX15B) partially murinized the product pattern of the human enzyme [25]. To test whether the product pattern of mouse Alox15b can also be humanized in vivo, knock-in mice were recently created, which express the humanized AA 15-lipoxygenting Tyr603Asp+His604Val double mutant of mouse Alox15b instead of the AA 8-lipoxygenating mouse wildtype enzyme [26]. These mice are viable and breed normally [26], but the erythrocyte parameters of male knock-in individuals are compromised [26]. As indicated above, the Asp602Tyr+Val603His exchange murinized the product pat- tern of human ALOX15B with arachidonic acid and an inverse mutagenesis strategy on mouse Alox15b (Tyr603Asp+His604Val exchange) humanized the product mixture of this enzyme. However, whether this humanization of mouse Alox15b and murinization of hu- man ALOX15B can also be observed when other polyenoic fatty acid are used as substrate has not been explored in detail. To fill this gap, we expressed the four enzyme variants as recombinant proteins and tested under strictly comparable experimental conditions their reaction specificities with different substrates (oxygenated and non-oxygenated polyenoic fatty acids). We found that human and mouse ALOX15B orthologs exhibit differential reaction specificities with all free polyenoic fatty acid derivatives tested (AA, EPA, DHA, LA, 15-HETE, and 8-HETE) and that the product pattern of the mouse enzyme can be humanized by Tyr603Asp+His604Val exchange. When an inverse mutagenesis strategy was applied for human ALOX15B, the product mixture was murinized for AA and EPA oxy- genation, but the major DHA oxygenation product was identified as 7-HDHA. Thus, with this substrate fatty acid the product pattern was not murinized. To explore the molecular basis for the observed functional differences we carried out substrate docking studies and performed molecular dynamics simulations, and the obtained in silico data are consistent with our experimental observations. 2. Results 2.1. Recombinant Expression of Mouse and Human ALOX15B Orthologs and of Selected Mutants Human ALOX15B oxidizes arachidonic acid to 15-HETE but its D602Y+V603H mu- tant forms a 2:1 mixture of 8-HETE and 12-HETE [25]. Thus, with arachidonic acid the D602Y+V603H double mutant of human ALOX15B forms similar products as wildtype mouse Alox15b. On the other hand, mouse Alox15b forms almost exclusively to 8-HETE from AA but its Y603D+H604V mutant dominantly produces 15-HETE [25] and thus, this enzyme mutant forms humanized reaction products [25]. To test whether murinization of human ALOX15B and humanization of mouse Alox15b ortholog can also be seen when other substrate fatty acids are used, we expressed the four enzyme species (wildtype human ALOX15B, D602Y+V603H mutant of human ALOX15B, wildtype mouse Alox15b, and Y603D+H604V mutant of mouse Alox15b) as his-tag fusion proteins in E. coli and quantified the expression levels of the four enzyme variants by quantitative immunoblotting using an anti-his-tag antibody (Figure 1). The expression levels of the different recombinant proteins are summarized in (Table 1). It can be seen that all ALOX15B variants are well expressed in E. coli and that mouse Alox15b reached an expression level of about 700 mg Alox15b protein per liter bacterial liquid culture. Next, we attempted to purify the recombinant proteins by affinity chromatography on Ni-agarose gel. The two wildtype enzymes could well be purified as catalytically active proteins employing this purification technique. Int.J.J.Mol. Int. Mol.Sci. Sci.2023, 2023,24, 24,10046 x FOR PEER REVIEW 44 of of 25 24 Figure1.1.Quantification Figure Quantificationof ofthe theexpression expressionlevels levelsof ofhuman humanand andmouse mouseALOX15B ALOX15Bvariants. variants.ALOX15B ALOX15B variants were variants were expressed expressed in E. coli coli and and aliquots aliquotsofofthethebacterial bacteriallysis lysissupernatants supernatantswere wereanalyzed analyzedby quantitative by immunoblotting quantitative immunoblotting(see(see Section 4). A4).total Section of 1 µL A total of human 1 µL humanwildtype supernatant; wildtype 1.1 µL supernatant; human 1.1 D602Y+V603H µL human mutantmutant D602Y+V603H supernatant; 0.5 µL0.5 supernatant; wildtype mousemouse µL wildtype supernatant; 1.2 µL 1.2 supernatant; mouse µL Y603D+H604V supernatant. The band intensity scale was calibrated applying mouse Y603D+H604V supernatant. The band intensity scale was calibrated applying known known amounts of the pure recombinant N-terminal his-tag fusion protein of M. fulvus ALOX. (A) Human ALOX15B, amounts of the pure recombinant N-terminal his-tag fusion protein of M. fulvus ALOX. (A) Human (B) Mouse Alox15b. ALOX15B, (B) Mouse Alox15b. Table 1. Expression levels of the different recombinant ALOX15B variants. ALOX15B variants were Table 1. Expression expressed levels in E. coli and of the lysis bacterial different recombinant supernatants wereALOX15B prepared variants. ALOX15B as described variants in Section 4. Ali- were quotsexpressed of the lysis E. coli and bacterial in supernatants lysis supernatants were analyzed were prepared by quantitative as described immunoblotting using in anSection 4. anti-hist- tag antibody Aliquots of thefor specific lysis staining were supernatants of theanalyzed ALOX bands. SDS-PAGEimmunoblotting by quantitative and immunoblotting using were carried an anti-hist- outantibody tag as described in Section for specific 4. Theofband staining intensity the ALOX scaleSDS-PAGE bands. was calibrated by applying known and immunoblotting wereamounts carried (3 µg, out 1.5 µg, 0.75 as described inµg) of the4.pure Section recombinant The band N-terminal intensity scale was his-tag fusion calibrated protein ofknown by applying M. fulvus ALOX amounts (3asµg, the1.5 reference µg, 0.75protein [27].pure recombinant N-terminal his-tag fusion protein of M. fulvus ALOX µg) of the as the reference protein [27]. Expression Level (mg ALOX Protein per L Enzyme Variant Bacterial Expression Liquid Level (mg ALOXCulture) Protein per L Enzyme Wildtype Variant human ALOX15B 200 Bacterial Liquid Culture) Wildtypehuman Wildtype mouseALOX15B Alox15b 700 200 D602Y+V603H mutant Wildtype mouse(human Alox15bALOX15B) 170 700 D602Y+V603H mutant Y603D+H604V (human mutant (mouseALOX15B) Alox15b) 170 292 Y603D+H604V mutant (mouse Alox15b) 292 Unfortunately, the D602Y+V603H mutant of human ALOX15B lost more than 90% of its catalytic activitythe Unfortunately, during the purification D602Y+V603H mutantprocedure. of human To ensure strict ALOX15B comparability lost more than 90% of of the functional data we used crude enzyme preparations (bacterial lysate supernatant) its catalytic activity during the purification procedure. To ensure strict comparability of for functional the characterization. functional data we used crude enzyme preparations (bacterial lysate supernatant) for functional characterization. 2.2. Mouse and Human ALOX15B Variants Prefer C20 and C22 Polyenoic Fatty Acids over 2.2. Mouse Linoleic and(C18) Acid Human ALOX15B Variants Prefer C20 and C22 Polyenoic Fatty Acids over Linoleic Acid (C18) To test the substrate specificity of human and mouse ALOX15B variants in vitro ac- tivityToassays test theusing substrate specificity linoleic of human acid (LA), and mouse ALOX15B variants 5,8,11,14,17-eicosapentaenoic in vitroarachi- acid (EPA), activ- ity assays using linoleic acid (LA), 5,8,11,14,17-eicosapentaenoic acid (EPA), arachidonic donic acid (AA), and 4,7,10,13,16,19-docosahexaenoic acid (DHA) as ALOX substrates acid (AA), and 4,7,10,13,16,19-docosahexaenoic acid (DHA) as ALOX substrates were per- were performed. When we set the catalytic activity of the different enzyme variants with formed. When we set the catalytic activity of the different enzyme variants with AA 100% AA 100% and calculated the relative catalytic activities with the other polyenoic fatty acids and calculated the relative catalytic activities with the other polyenoic fatty acids we found we found (Figure 2A, left columns) that wildtype human ALOX15 prefers DHA as sub- (Figure 2A, left columns) that wildtype human ALOX15 prefers DHA as substrate. EPA strate. EPA and AA were also oxygenated with relatively high reaction rates, but linoleic and AA were also oxygenated with relatively high reaction rates, but linoleic acid was acid was not a good substrate for this enzyme. This data is consistent with previous re- not a good substrate for this enzyme. This data is consistent with previous reports on the ports on the substrate specificity of human ALOX15B [19]. substrate specificity of human ALOX15B [19]. Int. J.J. Mol. Mol. Sci. Sci. 2023, 24, x10046 2023, 24, FOR PEER REVIEW 5 5of of 24 25 Figure Figure 2.2. Substrate Substrate specificity specificityofofhuman humanand and mouse mouse ALOX15 ALOX15 variants variants with with different different polyunsatu- polyunsaturated rated fatty acids. ALOX15B variants were expressed in E. coli and aliquots of the bacterial fatty acids. ALOX15B variants were expressed in E. coli and aliquots of the bacterial lysis supernatantslysis su- pernatants were incubated in vitro for 5 min with different polyunsaturated fatty acids. After the were incubated in vitro for 5 min with different polyunsaturated fatty acids. After the incubation incubation period the produced hydroperoxy fatty acids were reduced to their more stable hydroxy period the produced hydroperoxy fatty acids were reduced to their more stable hydroxy derivatives derivatives and products carrying a conjugated diene chromophore were quantified by RP-HPLC. and products Two incubation carrying a conjugated samples diene were set up forchromophore were each fatty acid quantified substrate andbyeach RP-HPLC. sampleTwo wasincubation analyzed twice. The product formation from AA by each enzyme preparation was set 100% andThe samples were set up for each fatty acid substrate and each sample was analyzed twice. the product relative formationof amounts from AA by each conjugated enzyme dienes preparation formed wasother from the set 100% and the relative substrates amounts of(A) were calculated. conjugated Human dienes formed ALOX15B, from the (B) mouse other substrates were calculated. (A) Human ALOX15B, (B) mouse Alox15b. Alox15b. When we When wetested testedthethe substrate substrate specificity specificity of the of the humanhuman ALOX15B ALOX15B D602Y+V603H D602Y+V603H dou- double mutant DHA, EPA and AA were equally well oxygenated ble mutant DHA, EPA and AA were equally well oxygenated but LA was not a good but LA was not a good substrate (Figure substrate (Figure2A,2A,right rightcolumns). columns).SimilarSimilar experiments experiments werewere subsequently subsequently carried carried out out with mouse Alox15b and its Y603D+H604V double mutant. with mouse Alox15b and its Y603D+H604V double mutant. The wildtype enzyme con- The wildtype enzyme converted verted DHA,DHA, EPA,EPA,andand AAAA with with similar similar reaction reaction ratesbut rates buthere hereagain, again,LA LAwas wasnotnotaa good good substrate (Figure substrate (Figure 2B, 2B, left left columns). columns). For For the the mouse mouse Alox15b Alox15b Y603D+H604V Y603D+H604V double double mutant mutant EPA was the preferred polyenoic fatty acid substrate followed by DHA EPA was the preferred polyenoic fatty acid substrate followed by DHA and AA. For this and AA. For this enzyme variant, LA was also not a good substrate in the absence of detergents enzyme variant, LA was also not a good substrate in the absence of detergents (Figure 2B, (Figure 2B, right columns). The addition of sodium cholate improved the substrate right columns). The addition of sodium cholate improved the substrate behavior of LA behavior of LA [28] and and [28] this this effect might effect be related might to the be related to allosteric properties the allosteric of this properties enzyme of this [29].[29]. enzyme However, How- since cholate is absent in most mammalian cells at these relatively high concentrations, we ever, since cholate is absent in most mammalian cells at these relatively high concentra- did not test the impact of this detergent on the reaction rate and on the product pattern in tions, we did not test the impact of this detergent on the reaction rate and on the product this study. pattern in this study. It should be stressed at this point that the data shown in Figure 2 characterize the product It should be stressed at this point that the data shown in Figure 2 characterize the mixtures formed by the four different recombinant enzymes during a 5 min incubation period product mixtures formed by the four different recombinant enzymes during a 5 min in- (end point measurements). Unfortunately, more detailed kinetic studies could not be carried cubation period (end point measurements). Unfortunately, more detailed kinetic studies out since our crude enzyme preparation did not allow spectrophotometric measurements could not be carried out since our crude enzyme preparation did not allow spectrophoto- following the increase in absorbance at 235 nm (formation of conjugated dienes). metric measurements following the increase in absorbance at 235 nm (formation of conju- gated dienes). 2.3. D602Y+V603H Exchange Partly Murinized the Reaction Products Formed by Human ALOX15B with EPA and AA but Altered the Product Pattern with DHA in a Different Way 2.3. D602Y+V603H Human ALOX15B Exchange PartlyAA converts Murinized the Reaction to 15-H(p)ETE Products [19,25] Formed by Human but D602Y+V603H exchange ALOX15B with EPA partly murinized theand AA but pattern product Altered the Product [25]. Pattern In fact, withdouble for this DHA in a Different mutant Way 8S-H(p)ETE Human ALOX15B was identified converts as major AA AA to oxidation 15-H(p)ETE product. Whether [19,25] suchbut D602Y+V603H murinization of theexchange product partly patternmurinized can also bethe product with observed pattern [25]. other In fact, for substrates this has notdouble mutant 8S-H(p)ETE been explored. To answer was this identified question, as wemajor AA oxidation incubated wildtypeproduct. human Whether ALOX15B such andmurinization of the product its D602Y+V603H double pat- mu- tant with tern different can also polyenoic be observed fatty with acidssubstrates other and analyzed the been has not reaction products explored. Toby RP-HPLC. answer this As shownwe question, in incubated Figure 3A,wildtype 15-H(p)ETE wasALOX15B human the almost exclusive and AA oxygenation its D602Y+V603H doubleproduct mutant formed with by thepolyenoic different wildtype enzyme. fatty acidsHowever, for the and analyzed D602Y+V603H the double reaction products mutant a mix- by RP-HPLC. As ture of 8-H(p)ETE, 15-H(p)ETE, and 5-H(P)ETE was analyzed. This shown in Figure 3A, 15-H(p)ETE was the almost exclusive AA oxygenation product data confirmed the previousbyfinding formed that D602Y+V603H the wildtype exchange enzyme. However, partly for the murinized the D602Y+V603H product double mixture mutant of a mix- human ture ALOX15B [25] of 8-H(p)ETE, but our results 15-H(p)ETE, also indicate and 5-H(P)ETE was that the mutant analyzed. enzyme This data formed confirmed thea previous finding that D602Y+V603H exchange partly murinized the product mixture of Int. J. Mol. Sci. 2023, 24, x FOR PEER REVIEW 6 of 24 Int. J. Mol. Sci. 2023, 24, 10046 6 of 25 human ALOX15B [25] but our results also indicate that the mutant enzyme formed a sig- nificant share significant of of share 5-HETE asasminor 5-HETE minor(15–20% (15–20%ofofthe thetotal totalconjugated conjugated dienes dienes formed) side formed) side product. Interestingly, combined NP/CP-HPLC NP/CP-HPLC indicated that this indicated that this minor side product was completely chiral (5S-HETE), suggesting that the 5S-HETE formation was completely enzyme-controlled. Figure 3. Figure 3. D602Y+V603H D602Y+V603Hexchange exchangepartly murinized partly murinized thethe product mixture product of human mixture ALOX15B of human ALOX15Bwith different polyenoic fatty acids. ALOX15B variants were expressed in E. coli and aliquots with different polyenoic fatty acids. ALOX15B variants were expressed in E. coli and aliquots of the bac- of terial lysis supernatants were used to perform in vitro activity assays were carried out as described the bacterial lysis supernatants were used to perform in vitro activity assays were carried out as in the legend of Figure 2. Representative partial RP-HPLC chromatograms are shown. The chemical described in the legend of Figure 2. Representative partial RP-HPLC chromatograms are shown. identity of the conjugated dienes formed during the 5 min incubation period was concluded by co- The chemical identity chromatography of the conjugated with authentic dienes standards. formed during Oxygenation productsthe of 5AAminandincubation DHA were period was analyzed concluded by co-chromatography with authentic standards. Oxygenation products of isocratically using a solvent system consisting of acetonitrile/water /acetic acid (70/30/0.1, by vol).AA and DHA The EPA were analyzed isocratically oxygenation products were using a solvent analyzed system with isocratically consisting of acetonitrile/water/acetic the solvent system acetonitrile/wa- ter/acetic acid acid (50/50/0.1, (70/30/0.1, by vol). byThe vol). (A)oxygenation EPA Arachidonicproducts acid as substrate, (B) eicosapentaenoic were analyzed acidthe isocratically with as substrate, (C) docosahexaenoic acid as substrate. solvent system acetonitrile/water/acetic acid (50/50/0.1, by vol). (A) Arachidonic acid as substrate, (B) eicosapentaenoic acid as substrate, (C) docosahexaenoic acid as substrate. Similar incubations were performed with EPA (Figure 3B) and DHA (Figure 3C). As Similar expected fromincubations the product were performed profile with EPA (Figure of AA oxygenation, 15-HEPE 3B)was andthe DHA major(Figure 3C). EPA oxy- As expected genation fromformed product the product profile human by wildtype of AA oxygenation, 15-HEPE8-HEPE ALOX15B. In contrast, was thewas major theEPA ma- oxygenation product formed jor EPA oxygenation productbyformed wildtypebyhuman ALOX15B. Indouble the D602Y+V603H contrast, 8-HEPE mutant of was the human major EPA In ALOX15B. oxygenation product addition, smaller formedof amounts by15-H(p)EPE the D602Y+V603H double were and 5-H(p)EPE mutant of human found. Thus, ALOX15B. In addition, the oxygenation productssmaller formedamounts of 15-H(p)EPE from AA and EPA byand 5-H(p)EPE wildtype were human found. Thus, ALOX15B and the oxygenation products formed from AA and EPA by wildtype human its D602Y+V603H double mutant were very similar. In contrast, with DHA the situation ALOX15B and its wasD602Y+V603H doubleWildtype somewhat different. mutant were very human similar. In ALOX15B contrast, with oxygenated DHA the this highly situation unsaturated was somewhat different. Wildtype human ALOX15B oxygenated this highly polyenoic fatty acid almost exclusively to 17-HDHA and this simple product pattern unsaturated polyenoic fatty acid could be predicted onalmost exclusively the basis to 17-HDHA of the analysis of the AAandandthis EPAsimple product oxygenation pattern products. could be predicted Interestingly, on the the major DHA basis of the analysis oxidation productofformed the AAbyand theEPA oxygenation D602Y+V603H products. double mu- Interestingly, the major DHA oxidation product formed by the D602Y+V603H tant was not 10-HDHA (the equivalent DHA oxygenation product of 8-HETE) but 7- double mutant was notThis HDHA instead. 10-HDHA product(the equivalent is the equivalentDHA DHA oxygenation oxygenationproduct productofof8-HETE) 5-HETE. but 17- 7-HDHA instead. This product is the equivalent HDHA and 10-HDHA were also detected in minor quantities. DHA oxygenation product of 5-HETE. 17-HDHA and 10-HDHA were also detected in minor quantities. 2.4. Y603D+H604V Mutation Humanized the Product Pattern of Mouse Alox15b with AA, 2.4. Y603D+H604V Mutation Humanized the Product Pattern of Mouse Alox15b with AA, EPA EPADHA and and DHA Mouse Alox15b Mouse Alox15b converts convertsAA AAtoto8-HETE 8-HETE but butthethe Y603D+H604V Y603D+H604V exchange humanized exchange human- the product mixture of this enzyme [25]. For this mutant, 15S-HETE was identified ized the product mixture of this enzyme [25]. For this mutant, 15S-HETE was identified as the major AA oxidation product [25]. When we incubated mouse as the major AA oxidation product [25]. When we incubated mouse Alox15b and itsAlox15b and its Y603D+H604V mutant with AA we confirmed these results. 8-HETE Y603D+H604V mutant with AA we confirmed these results. 8-HETE was the major wild- was the major wildtype type enzyme enzyme product product whereas whereas thethe Y603D+H604Vdouble Y603D+H604V doublemutant mutantproduced produced 15-HETE 15-HETE (Figure 4A). When EPA was used as the substrate, 8-HEPE was analyzed (Figure 4A). When EPA was used as the substrate, 8-HEPE was analyzed as the major as the major oxygenation product oxygenation productofofthe thewildtype wildtypeenzyme, enzyme, but but 15-HEPE 15-HEPE waswas dominantly dominantly formed formed by by the the Y603D+H604V Y603D+H604V double double mutant mutant (Figure (Figure 4B).4B). Int. J. Mol. Sci. 2023, 24, 10046 Int. J. Mol. Sci. 2023, 24, x FOR PEER REVIEW 77 of of 25 24 Figure4.4. The Figure The Y603D+H604V Y603D+H604Vexchange exchangehumanized humanizedthe theproduct pattern product of of pattern mouse mouse Alox15b Alox15bwith dif- with ferent polyenoic fatty acids. Mouse Alox15b variants were expressed in E. coli and aliquots different polyenoic fatty acids. Mouse Alox15b variants were expressed in E. coli and aliquots of the of the bacterial lysis supernatants were used to perform in vitro activity assays as described in the legend bacterial lysis supernatants were used to perform in vitro activity assays as described in the legend of Figure 2. Representative partial RP-HPLC chromatograms are shown. The chemical identity of of Figure 2. Representative partial RP-HPLC chromatograms are shown. The chemical identity of the the conjugated dienes formed during the 5 min incubation period was concluded by co-chromatog- conjugated raphy withdienes formed authentic during Oxygenation standards. the 5 min incubation products period of AAwas andconcluded DHA werebyanalyzed co-chromatography isocratically with authentic standards. Oxygenation products of AA and DHA were analyzed using a solvent system consisting of acetonitrile/water /acetic acid (70/30/0.1, by vol). The isocratically EPA oxy- using a solvent genation system products wereconsisting of acetonitrile/water/acetic analyzed isocratically acid (70/30/0.1, with the solvent system by vol). The EPA acetonitrile/water/acetic acid (50/50/0.1, by oxygenation vol). (A)were products Arachidonic analyzedacid as substrate, isocratically with(B) theeicosapentaenoic acid as substrate, (C) do- solvent system acetonitrile/water/acetic cosahexaenoic acid (50/50/0.1,acid as substrate. by vol). (A) Arachidonic acid as substrate, (B) eicosapentaenoic acid as substrate, (C) docosahexaenoic acid as substrate. Finally, we explored the DHA oxygenation products formed by these two enzyme variants andwe Finally, explored found theexpected that, as DHA oxygenation from the productproducts formed profiles by these of AA and EPAtwo oxygena- enzyme variants and found tion, 10-HDHA wasthat, as expected mainly formedfrom by the product wildtypeprofiles enzyme of whereas AA and EPA oxygenation, 17-HDHA domi- 10-HDHA nated the wasproductmainly formed pattern by the by produced wildtype enzyme whereas the Y603D+H604V double17-HDHA dominated mutant (Figure 4C). the product Taken pattern together theseproduced by the Y603D+H604V data indicated that the Y603D+H604Vdouble mutantexchange(Figure 4C). Taken humanized the together product these pattern data indicated of mouse that the Alox15b notY603D+H604V only with AAexchange humanized but also with EPA and theDHA. productThe pattern outcome of of mouse the DHAAlox15b not only was experiments withofAA but alsointerest particular with EPA andthe since DHA. Themutagenesis inverse outcome of the DHA experiments was of particular interest since strategy of human ALOX15B (D602Y+V603H exchange) murinized the reaction the inverse mutagenesis strategy products of human ALOX15B (D602Y+V603H exchange) murinized the of AA and EPA oxygenation by human ALOX15 (Figure 3A,B) but not the product pattern reaction products of AA and EPA oxygenation of DHA oxygenation by human (FigureALOX15 3C). (Figure 3A,B) but not the product pattern of DHA oxygenation (Figure 3C). 2.5. D602Y+V603H Exchange in Human ALOX15B and Y603D+H604V Mutation of Mouse 2.5. D602Y+V603H Exchange in Human ALOX15B and Y603D+H604V Mutation of Mouse Alox15b also Altered the Reaction Products of Linoleic Acid Oxygenation Alox15b also Altered the Reaction Products of Linoleic Acid Oxygenation Inthe In theabsence absenceofofdetergents, detergents, linoleic linoleic acid acid waswas notnot a good a good substrate substrate forfor either either humanhumanor or mouse ALOX15B (Figure 2). Nevertheless, we analyzed the product mouse ALOX15B (Figure 2). Nevertheless, we analyzed the product mixture formed by the four mixture formed by the four different enzyme variants from these polyenoic fatty different enzyme variants from these polyenoic fatty acids, which abundantly occurs in mostacids, which abundantly occurs in most mammalian cells.mammalian cells. For For this purpose, this activity in vitro purpose, in vitro assays activity were carriedassays were carried out (Figure out 2), and the (Figure 2), and the conjugated dienes formed were prepared conjugated dienes formed were prepared by RP-HPLC and further analyzed by the combined by RP-HPLC and further analyzed normal by the combined phase/chiral phase HPLCnormal phase/chiral phase (NP/CP-HPLC). HPLC (NP/CP-HPLC). With linoleic acid as the substrateWithonlylinoleic four acid as the substrate only four major ALOX products (13S-HODE, major ALOX products (13S-HODE, 13R-HODE, 9S-HODE, 9R-HODE) can be formed and all of 13R-HODE, 9S-HODE, 9R-HODE) them are wellcan be formed separated by our and all of them combined are well separated NP/CP-HPLC analyticalby our combined NP/CP- protocol. HPLC analytical protocol. As expected, the major LA oxygenation product of wildtype human ALOX15B was As expected, 13S-HODE. 13R-HODE the major and the LAtwo oxygenation product of wildtype 9-HODE enantiomers (9S- and human 9R-HODE) ALOX15B were onlywas 13S-HODE. formed in small13R-HODE amountsand the two (Figure 5A).9-HODE enantiomers Interestingly, the major(9S- LA andoxygenation 9R-HODE) were only product formedby formed inthe small mouse amounts wildtype(Figure Alox15b 5A).wasInterestingly, 9S-HODE and the major LA oxygenation here again, the other HODE product iso- formed mers werebyonly the minor mouseside wildtype products Alox15b (Figurewas 5C).9S-HODE The human and here again, ALOX15B the other HODE D602Y+V603H dou- isomers ble mutant were only minor exhibited only side a very products (Figure low catalytic 5C). The activity human with ALOX15B this fatty acid and D602Y+V603H here we ob- double mutant exhibited only a very low catalytic activity served a complex mixture of all four LA oxygenation products. However, 9- and with this fatty acid and here we 13-HODE observed were a complex not complete mixture racemic of all four mixtures sinceLAfor oxygenation products. each positional However, isomer 9- and 13- the S-enantiomer HODE were prevailed not complete (S/R-ratio of 7:3)racemic for the mixtures two positionalsince for each positional isomers. isomer the S-enanti- The Y603D+H604V double omer prevailed mutant of mouse(S/R-ratio of 7:3) for the Alox15b dominantly two positional formed 13S-HODE isomers. The the and thus, Y603D+H604V product pattern doubleof mutant the mouse of Alox15b mouse Alox15b dominantly with linoleic acid was formed 13S-HODE and thus, the product pattern fully humanized. of the mouse Alox15b with linoleic acid was fully humanized. Int.J.J.Mol. Int. Mol.Sci. Sci.2023, 2023,24, 24,10046 x FOR PEER REVIEW 88 of of 25 24 Figure5.5.Product Figure Productpatterns patternsofofmouse mouseandandhuman humanwildtype wildtypeALOX15 ALOX15orthologs orthologsand andofoftheir theirmurinized murinized (D602Y+V603H)and (D602Y+V603H) andhumanized humanized (Y603D+H604V) (Y603D+H604V) double double mutants mutants using using linoleic linoleic acid acid as theassubstrate. the sub- strate. The The different different ALOX15B ALOX15B variantsvariants were expressed were expressed as N-terminal as N-terminal his-tagproteins his-tag fusion fusion proteins in E. coli in andE. coli and aliquots of the bacterial lysis supernatants were used to perform in vitro activity aliquots of the bacterial lysis supernatants were used to perform in vitro activity assays (Figure 2). assays (Figure 2). The conjugated dienes formed during the incubation period were prepared by RP-HPLC The conjugated dienes formed during the incubation period were prepared by RP-HPLC and were and were further analyzed by NP/CP-HPLC. Representative partial RP-HPLC chromatograms are further analyzed by NP/CP-HPLC. Representative partial RP-HPLC chromatograms are shown. The shown. The chemical identity of the conjugated dienes formed during the incubation period was chemical concluded identity of the conjugated dienes by co-chromatography formed during with authentic the incubation standards. period was concluded The chromatographic conditionsby of co-chromatography with authentic standards. The chromatographic conditions of NP/CH-HPLC are described in detail in Section 4. (A) Wildtype human ALOX15B, (B) murinized NP/CH-HPLC are described in detail (D602Y+V603H), human ALOX15B in Section 4. (A) Wildtype human (C) wildtype ALOX15B, mouse Alox15B,(B) murinized (D) humanized human mouseALOX15B Alox15b (Y603D+H604V).(C) wildtype mouse Alox15B, (D) humanized mouse Alox15b (Y603D+H604V). (D602Y+V603H), 2.6. 2.6.MDMDSimulations SimulationsSuggest Suggestan anAlternative AlternativeFatty FattyAcid AcidBinding Bindingatatthe theActive ActiveSite SiteofofHuman Humanand Mouse ALOX15 Orthologs and Mouse ALOX15 Orthologs Human and mouse ALOX15 orthologs share a high degree of amino acid conservation Human and mouse ALOX15 orthologs share a high degree of amino acid conserva- but the catalytic properties of the two enzymes are remarkably different. Human ALOX15B tion but the catalytic properties of the two enzymes are remarkably different. Human oxygenates AA and EPA to the 15-oxygenation products (Figure 3A,B) whereas mouse ALOX15B oxygenates AA and EPA to the 15-oxygenation products (Figure 3A,B) whereas Alox15b dominantly forms the 8-oxygenation products from these two PUFAs. A plau- mouse sible Alox15b dominantly explanation formsproduct for the different the 8-oxygenation patterns was products an inversefromsubstrate these twoalignment PUFAs. A plausible explanation for the different product patterns at the active site of the two enzymes. For 15-HETE formation by human ALOX15B was an inverse substrate align- a ment at the active site of the two enzymes. For 15-HETE formation tail-first substrate alignment at the active site can be suggested [30]. Such substrate bind- by human ALOX15B a tail-first ing allows substrate hydrogenalignment abstraction at the fromactive C13site andcan beradical [+2] suggested [30]. Such substrate rearrangement. bind- In contrast, ing allows hydrogen abstraction from C13 and [+2] radical rearrangement. 8-lipoxygenation by mouse Alox15b may involve head-first substrate alignment, hydrogen In contrast, 8- lipoxygenation abstraction fromby mouse C10, and [Alox15b −2] radicalmayrearrangement. involve head-first substrate alignment, hydrogen abstraction from C10, and [−2] radical rearrangement. To test these hypotheses, we have carried out docking calculations of the substrates AA, EPA, To testDHA,theseandhypotheses, we have LA at the active carried sites of theoutfourdocking enzymecalculations variants. For of each the substrates case, we AA, EPA, DHA, and LA at the active sites of the four enzyme have selected the five more stable structures and classified them according to the variants. For each case, we preference have of selected tail-first (TF) the orfive more stable head-first (HF)structures substrateand classifiedThe alignment. them according results to the in are shown preference Table 2. of tail-first Despite the (TF) or head-first limited accuracy (HF) substrate of such docking alignment. The results calculations, it can be areseen shown thatinthe Table four2. Despite the substrates limited accuracy preferentially tend of to such adoptdocking calculations, a tail-first alignmentitincan be seenhuman wildtype that theALOX15B four sub- strates and preferentially in the Y603D+H604V tend to adoptof mutant a tail-first alignment mouse Alox15b. Inin wildtype contrast, human ALOX15B a head-first and orientation in the Y603D+H604V mutant of mouse Alox15b. In contrast, was preferred in the D602Y+V603H mutant of human ALOX15B and in wildtype mouse a head-first orientation was preferredToinvisualize Alox15b. the D602Y+V603H these results,mutant images of of the human ALOX15B tail-first alignment and in wildtype of AA mouse (the carboxylate Alox15b.toTothe pointing visualize proteinthese results, surface, images Figure 6A) inof wildtype the tail-first alignment human ALOX15B, of AAand(theof carboxylate the head- pointing first to theof alignment protein AA insurface, wildtype Figure mouse 6A) in wildtype Alox15b human ALOX15B, (the carboxylate facing toandtheofbottom the head- of firstenzyme, the alignment of AA Figure 6B)inare wildtype mouse illustrated. In Alox15b both cases,(theAAcarboxylate facing to the adopts a U-shaped bottomin structure of thesubstrate the enzyme, Figure binding 6B) are illustrated. In both cases, AA adopts a U-shaped structure in the pocket. substrate binding pocket. Int.J.J.Mol. Int. Mol.Sci. Sci.2023, 2023,24, 24,10046 x FOR PEER REVIEW 99 of of 25 24 Table2.2.The Table Thefive fivemost moststable stablestructures structures ofof the the substrates substrates AA, AA, EPA,EPA, DHA, DHA, andand LALA at the at the active active sites sites of of the four enzymes, according to docking calculations. Total free energy change (kJ/mol) that the four enzymes, according to docking calculations. Total free energy change (kJ/mol) that occurs occurs on ligand binding and substrate alignment (Tail-first (TF) or Head-first (HF)). The TF solutions are on ligand binding and substrate alignment (Tail-first (TF) or Head-first (HF)). The TF solutions are highlighted in blue, and the HF solutions in pink. highlighted in blue, and the HF solutions in pink. Substrate AA EPA DHA LA Substrate AA EPA DHA LA bind Enzyme ∆G bindOrient. ∆G bindOrient. ∆G bind bind Orient. ∆G Orient. Enzyme ∆Gbind Orient. ∆Gbind Orient. ∆G Orient. ∆Gbind Orient. −57.26 TF −64.38 TF −69.23 TF −58.15 TF −57.26 TF −64.38 TF −69.23 TF −58.15 TF −57.25 TF Human −62.69−57.25 TF TF −62.69 −67.46 TF −67.46 TF TF −55.08 −55.08 TF TF Human ALOX15B −57.21 TF ALOX15B−59.66−57.21 TF TF −59.66 −62.39 TF −62.39 TF TF −54.38 −54.38 TF TF WT WT −51.22 HF −58.93 TF −56.63 TF −53.22 TF TF −51.22 HF −58.93 TF −56.63 TF − 53.22 −51.00 TF −56.36−51.00 TFTF −56.36 −55.85 TF −55.85 HF HF −52.51 −52.51 TF TF −39.43 HF −39.75−39.43 HF TF −39.75 −43.77 TF −43.77 HF HF −35.09 −35.09 HF HF Human −39.13 HF −39.34−39.13 HF HF −39.34 −43.71 HF −43.71 TF TF − 34.82 −34.82 TF TF Human ALOX15B −39.01 HF −38.98−39.01 HF HF −43.56 −38.98 HF −43.56 TF − TF34.78 −34.78 TF TF mut ALOX15B mut −38.97 HF −38.93−38.97 TF HF −43.51 −38.93 TF HF −43.51 −34.70 HF −34.70 TF TF −38.87 TF −38.89−38.87 TFTF −43.44 −38.89 TF TF −43.44 − TF34.68 −34.68 TF TF −41.44 HF −41.47−41.44 HF HF −45.98 −41.47 HF HF −45.98 −36.15 HF −36.15 HF HF Mouse −40.12 TF −41.19 HF −45.23 HF −35.76 TF −40.12 TF −41.19 HF −45.23 HF −35.76 TF Alox15b −39.69 TF Mouse −41.04 HF −44.87 HF −35.57 TF WT −39.69 TF −41.04 HF −44.87 HF −35.57 TF −39.08 Alox15b WT TF −41.02 HF −44.56 TF −35.36 TF −39.08 TF −41.02 HF −44.56 TF −35.36 TF −39.07 TF −40.88 HF −44.16 TF −35.34 TF −42.56 TF −43.64−39.07 TF TF −40.88 −49.03 HF −44.16 TF TF − 49.03 −35.34 TF TF Mouse −42.56 HF −43.28−42.56 HFTF −43.64 −47.95 TF −49.03 TF TF − 47.95 −49.03 TF TF Alox15b −42.52 TF −42.63−42.56 HF TF −43.28 −46.30 HF −47.95 TF TF − 46.30 −47.95 TF TF mut Mouse −42.12 TF −42.51−42.52 TFTF −42.63 −46.08 TF −46.30 TF TF −46.08 −46.30 TF TF Alox15b mut −41.65 TF −42.29−42.12 TFTF −42.51 −46.00 TF −46.08 TF TF − 46.00 −46.08 TF TF −41.65 TF −42.29 TF −46.00 TF −46.00 TF Figure6. Figure 6. Tail-first Tail-first alignment alignmentofofAA AA(A) in in (A) wildtype human wildtype ALOX15B human in green, ALOX15B and head-first in green, align- and head-first ment of AA (B) in wildtype mouse Alox15b in pink. alignment of AA (B) in wildtype mouse Alox15b in pink. Int. J. Mol. Sci. 2023, 24, x FOR PEER REVIEW 10 of 24 Int. J. Mol. Sci. 2023, 24, 10046 10 of 25 After our docking calculations revealed the preferred substrate alignment for each After our docking enzyme-substrate calculations complex, revealed we employed the preferred molecular substrate dynamics (MD)alignment simulations fortoeach ex- enzyme-substrate plore which hydrogen complex, atomsweare employed locatedmolecular dynamics at a suitable (MD) position for simulations to explore productive hydrogen which hydrogen abstraction fromatoms are located a bisallylic at a suitable methylene. position In principle, AAforinvolves productive hydrogen three differentabstraction bisallylic from a bisallylic methylene. In principle, AA involves three different methylene groups and each of these carbon atoms carry two hydrogen atoms, which can bisallylic methylene groups and each be abstracted of these during carbon initial atomsremoval. hydrogen carry twoTohydrogen atoms,ofwhich explore which thesecan be abstracted hydrogen atoms during initial hydrogen removal. To explore which of these are suitable candidates for hydrogen abstraction, we focused on AA and ran hydrogen atoms are suitable eight MD candidates simulationsfor (100hydrogen ns each). abstraction, we focused These simulations mirroron AAHF both and ranTFeight and MD simulations substrate orientation (100 at thens active each). These site ofsimulations mirror both the four different HF and enzyme TF substrate species orientation (wildtype humanatALOX15B, the active site of the four different enzyme species (wildtype human D602Y+V603H mutant of human ALOX15B, wildtype mouse Alox15b, and Y603D+H604V ALOX15B, D602Y+V603H mutant mutant of of mouse humanAlox15b). ALOX15B, Eachwildtype simulationmouse Alox15b, started from theand Y603D+H604V energetically mostmutant favorable of mouse Alox15b). Each simulation started from the energetically most complex obtained in our docking studies. The evolution of the distances between the ox- favorable complex obtained ygen atom inof ourthedocking studies. Fe(III)-OH The evolution − cofactor of the (catalytically distances ALOX competent between the oxygen species) and theatom of clos- the Fe(III)-OH − cofactor (catalytically competent ALOX species) and the closest hydrogen est hydrogen atom attached to C7 (H7), C10 (H10) and C13 (H13) of AA along the MD atom attached simulations to C7 (H7), is pictured C10 (H10) in Figures 7 andandS1C13 and(H13) of AA along the the corresponding MD simulations average distances are is pictured in Figure summarized in Table 3. 7 and Figure S1 and the corresponding average distances are summarized in Table 3. − cofactor to Figure 7. Distances from the Figure 7. the oxygen oxygen atom atom in in the the Fe(III)-OH Fe(III)-OH− cofactor to the closest hydrogen atom attached attached to to C7 C7(H7), (H7),C10 C10(H10), (H10),and andC13 C13(H13) (H13)along thethe along molecular dynamics molecular simulations dynamics for AA simulations for (HF AA (HF or orat TF) TF) theatactive the active site site of of wildtype wildtype human human ALOX15B ALOX15B and wildtype and wildtype mouse mouse Alox15b. Alox15b. Table 3. Table 3. Average Average distances distances (Å) (Å)from fromthe theoxygen oxygenatom atominin the Fe(III)-OH the Fe(III)-OHcofactor − − to the cofactor closest to the hy- closest drogen atom attached to C7 (H7), C10 (H10), and C13 (H13) along the molecular dynamics hydrogen atom attached to C7 (H7), C10 (H10), and C13 (H13) along the molecular dynamics simula- tions for AA (HF or TF) at the active site of the different enzymes. simulations for AA (HF or TF) at the active site of the different enzymes. Human Mouse Substrate Human Mouse Substrate WT Mutant WT Mutant Orientation WT Mutant WT Mutant Orientation H7 H10 H13 H7 H10 H13 H7 H10 H13 H7 H10 H13 H7 H10 H13 H7 H10 H13 H7 H10 H13 H7 H10 H13 AA (HF) 6.6 9.7 11.4 7.2 5.7 3.4 5.8 3.9 3.4 8.5 6.9 5.5 AA (HF) 6.6 9.7 11.4 7.2 5.7 3.4 5.8 3.9 3.4 8.5 6.9 5.5 AA (TF) 6.4 3.4 4.2 6.5 4.6 4.8 8.4 5.8 3.0 7.4 4.8 3.2 AA (TF) 6.4 3.4 4.2 6.5 4.6 4.8 8.4 5.8 3.0 7.4 4.8 3.2 In wildtype human ALOX15B the carboxylate group of AA in the HF alignment is In wildtype destabilized human by D602 andALOX15B the carboxylate V603. A similar group destabilizing of AA effect was in the HF by observed alignment D603 andis destabilized by D602 and V604 in the humanized V603. A similar Y603D+H604V destabilizing double mutant ofeffect mousewas observed Alox15b. by D603 Because and of these V604 in the humanized Y603D+H604V double mutant of mouse Alox15b. Because of these Int. J. Mol. Sci. 2023, 24, 10046 11 of 25 destabilizing effects, AA is unable to establish a strong non-covalent interaction with any residue in the depth of the enzyme cavity and it tends to move away towards the protein surface. The large average distances given in Table 3 confirm this trend. Conversely, the strong interaction between the AA carboxylate and Arg428 stabilizes the position of TF AA alignment inside the substrate binding pocket. Thus, the tail-first orientation turns out to be the preferred way of substrate alignment and this data is consistent with the docking calculations and with the observed 15-HETE formation by wildtype human ALOX15B and the humanized mouse mutant. In our MD simulations many structures with H13 and/or H10 at distances suitable for hydrogen abstraction appear for both wildtype human ALOX15B, and the humanized Y603D+H604V double mutant of mouse Alox15b. Consequently, the average distances varying between 3.2 and 4.8 Å are rather small. For those enzyme variants carrying a Tyr and a His in the core of the enzymes (wildtype mouse Alox15b and D602Y+V603H mutant of human ALOX15B) the AA carboxylate can form stabilizing hydrogen bonds independently of whether AA is aligned with HF or TF. In both cases, the enzyme-substrate complex remains stable, with H13 and H10 being located close to the oxygen of the cofactor (Figure 7 and Figure S1, Table 3). In any alignment for any of the four studied enzymes H7 was always rather dis- tant from Fe(III)-OH− cofactor and thus hydrogen abstraction from this pro-chiral center is unlikely. However, our MD simulations cannot decide if the HF or the TF substrate alignment is more favorable, but our docking results already suggested that HF substrate orientation is the preferred choice for mouse Alox15b and the murinized human ALOX15B double mutant. The potential energy barriers for hydrogen abstractions from the pro-chiral bisallylic methylenes of AA (H7, H10, H13) tend to increase with the distance between the hydrogen to be abstracted and the oxygen atom of the Fe(III)-OH− cofactor. However, there is no precise correlation between the energy barriers and the distances since many other factors are involved. AA 15-lipoxygenation by wildtype human ALOX15B and by the mouse Alox15b Tyr603Asp+His604Val double mutant involves hydrogen abstraction from C13 (H13). The nascent pentadienyl radical formed during hydrogen abstraction is close to planar (sp2 hybridization of the carbon atoms) and the π-electron density is completely de- localized over the entire pentadienoic system (C11 to C15). In contrast, AA 8-lipoxygenation by wildtype mouse Alox15b and by the human ALOX15B Asp602Tyr+Val603His double mutant involves hydrogen abstraction from C10 (H10). Here the π-electron density of the pentadienyl radical is delocalized over C8 to C12 We have previously suggested that the extent of conversion of the nonplanar structure of the fatty acid substrate to the planar structure of the evolving pentadienyl radical (geometric motion) severely impacts the energy barrier of the hydrogen abstraction. In fact, a high degree of geometric motion appears to increase the energy barrier [31]. The differential steric hindrance of this geometric motion by the C-terminal amino acids (Ile675 for human ALOX15B and Ile677 for mouse Alox15) as well as by Leu609 (for human ALOX15B) and Leu611 (for mouse Alox15b) appears to favor C13 hydrogen abstraction by human ALOX15B but C10 hydrogen abstraction by its mouse ortholog. For TF binding of AA in wildtype human ALOX15B (Figure 8) the dihedral angles C11-C12-C13-C14 and C12-C13-C14-C15 are closer to planarity (0◦ or 180◦ ) than the dihedral angles C8-C9-C10-C11 and C9-C10-C11-C12. Thus, abstraction of H10 requires a high degree of geometric motion of AA in the region of the C8-C9 double bond but the side chains of Ile675 and Leu609 significantly hinder this motion. Thus, H13 abstraction and subsequent [+2] radical rearrangement is sterically preferred. For tail-first AA alignment at the active site of the Y603D+H604V double mutant of mouse Alox15b we observed a similar steric situation (Figure S2). Int. Int. J.J. Mol. Mol. Sci. Sci.2023, 2023,24, 24,x10046 FOR PEER REVIEW 1212ofof 24 25 Figure 8. Figure Shortestdistances 8. Shortest distances(Å) (Å)from fromthe theC8-C9 C8-C9and andC11-C12 C11-C12double double bonds bonds toto terminal terminal Ile675 Ile675 and and Leu609, and Leu609, anddihedral dihedralangles C8-C9-C10-C11, angles C8-C9-C10-C11, C9-C10-C11-C12, C9-C10-C11-C12, C11-C12-C13-C14, and C12-C13-C14-C15 C11-C12-C13-C14, and C12-C13- C14-C15 along thealong the molecular molecular dynamicsdynamics simulationsimulation foratAA for AA (TF) the (TF) at site active the of active site ofhuman wildtype wildtype human ALOX15B. ALOX15B. On the other hand, the scenario is the opposite for head-first AA alignment. For wildtype On themouse otherAlox15b (Figure hand, the S3) and scenario for the is the D602Y+V603H opposite double for head-first AA mutant of human alignment. For ALOX15Bmouse wildtype (Figure S4) the (Figure Alox15b steric effects S3) and of the above-mentioned for the D602Y+V603Hamino doubleacids mutantfavor of the H10 human abstraction(Figure ALOX15B and subsequent [−2] S4) the steric radical effects of rearrangement. the above-mentioned amino acids favor the H10 abstraction and subsequent [−2] radical rearrangement. 2.7. D602Y+V603H Exchange in Human ALOX15B and Y603D+H604V Mutation in the Mouse Ortholog also Altered the Substrate Behavior of HETE Isomer 2.7. D602Y+V603H Exchange in Human ALOX15B and Y603D+H604V Mutation in the MouseAll canonic Ortholog HETE-isomers also involve Altered the Substrate bisallylic Behavior methylenes of HETE Isomer and thus they constitute suitable ALOX substrates. In theory, 15-HETE should be a good substrate for mouse All canonic HETE-isomers involve bisallylic methylenes and thus they constitute Alox15b, and 8,15-diHETE is predicted as major reaction product. On the other hand, suitable ALOX substrates. In theory, 15-HETE should be a good substrate for mouse 8-HETE should not be accepted as substrate by wildtype mouse Alox15b. In contrast, Alox15b, and 8,15-diHETE is predicted as major reaction product. On the other hand, 8- human ALOX15B should well oxygenate 8-HETE and here again, 8,15-diHETE should be HETE should not be accepted as substrate by wildtype mouse Alox15b. In contrast, hu- the major reaction product. However, 15-HETE should not be oxygenated by wildtype man humanALOX15B ALOX15B. should wellthese To test oxygenate 8-HETE hypotheses, inand vitrohere again, activity 8,15-diHETE assays should were carried outbewith the major reaction product. However, 15-HETE should not be oxygenated wildtype human and wildtype mouse ALOX15 orthologs and the reaction products were by wildtype human ALOX15B. analyzed byTo test these hypotheses, in vitro activity assays were carried out with RP-HPLC. wildtype As indicatedand human wildtype in Figure mouse human 9, wildtype ALOX15 orthologs ALOX15B and the8-HETE accepted reactionbut products were not 15-HETE analyzed by RP-HPLC. as the substrate. As expected, mouse Alox15b exhibited an inverse substrate specificity. As indicated 15-HETE was a good in Figure 9, wildtype substrate, but 8-HETE human wasALOX15B not (Figureaccepted 8-HETE butexchange 9). D602Y+V603H not 15- HETE as the substrate. As expected, mouse Alox15b exhibited an inverse murinized the substrate specificity of the human ALOX15B since 15-HETE was a good substrate specific- ity. 15-HETE substrate for was this aenzyme good substrate, variant. but 8-HETE Y603D+H604V In contrast, was not (Figure mutation 9). D602Y+V603H in mouseexchange Alox15b murinized humanizedthe thesubstrate substratespecificity specificityof ofthe human mouse ALOX15B Alox15b since since 15-HETE for this enzymewas a good mutant sub- 8-HETE strate was a for goodthis enzyme variant. oxygenation In contrast, substrate whereas Y603D+H604V 15-HETE was not mutation in mouse oxygenated Alox15b (Figure 9). hu- manized the substrate specificity of mouse Alox15b since for this enzyme mutant 8-HETE was a good oxygenation substrate whereas 15-HETE was not oxygenated (Figure 9). Int. Int.J.J.Mol. Mol.Sci. Sci.2023, 2023, 24, 10046 x FOR PEER REVIEW 13 of 13 25 of 24 Figure 9. Substrate specificity of mouse and human wildtype ALOX15 orth murinized (D602Y+V603H) and humanized (Y603D+H604V) double mutants usi HETE as the substrate. The different ALOX15B variants were expressed as N-term proteins in E. coli and aliquots of the bacterial lysis supernatants were used to p tivity assays using 15S-HETE and 8S-HETE (20 µM final concentration) as the su pearance Figure of the specificity Figure9.9. Substrate Substrate substrate during specificityofof mouse and mouse the incubation human and wildtype human period ALOX15 wildtype wasorthologs measured orthologs ALOX15 and of bytheir their and of RP-H assay was murinized murinized carried outand (D602Y+V603H) (D602Y+V603H) in duplicate andhumanized humanized and each sample (Y603D+H604V) (Y603D+H604V) double was double analyzed mutants mutants using using twice 15-HETE andby 15-HETE andRP-H 8- HETE as 8-HETE asthe the substrate. Thedifferent substrate. The differentALOX15B ALOX15B variants variants were were expressed expressed as N-terminal as N-terminal his-tag his-tag fusionfusion proteinsinin proteins E. E. colicoli andand aliquots aliquots ofbacterial of the the bacterial lysis supernatants lysis supernatants were were used used to in to perform perform in vitro ac- vitro activity 2.8. tivityWildtype assays using Mouse 15S-HETE Alox15b and Forms 8S-HETE (20 Significant µM final Amounts concentration) as of the 8,15-diHETE substrate. assays using 15S-HETE and 8S-HETE (20 µM final concentration) as the substrate. The disappearance The disap-fro pearance Long of the substrate Term Incubations of the substrate during the incubation period was measured by RP-HPLC. during the incubation period was measured by RP-HPLC. Each activity assay was Each activity assay was carried out in duplicate and each sample was analyzed carried out in duplicate and each sample was analyzed twice by RP-HPLC. twice by RP-HPLC. When we analyzed the mixture of oxygenation products formed 2.8. 2.8.Wildtype WildtypeMouse MouseAlox15b Alox15bForms Significant Forms Amounts Significant of 8,15-diHETE Amounts from from of 8,15-diHETE AA during Long AA during short-term Term Incubationsincubations by mouse Alox15b, we only detected the formatio Long Term Incubations 15-HETE. When weHowever, analyzed theafter longer mixture incubation of oxygenation periods products formed we observed from AA duringthe add Whenincubations short-term we analyzed the mixture by mouse Alox15b,ofweoxygenation products only detected formed the formation of from 8-HETE AAandduring of conjugated short-term trienes incubations that co-migrated with an authentic standard of 8S,15 15-HETE. However, afterby mouse longer Alox15b, periods incubation we onlywedetected observed the the formation of 8-HETE additional forma- and 10). tion ofThis 15-HETE. compound, However, conjugated afterthat trienes which longer was with incubation co-migrated absent periods in wethe product observed an authentic standard mixtureformation theofadditional 8S,15S-diHETE of short- of conjugated (Figure 10). This involved trienes thecompound,that canonic whichco-migrated with was absent conjugated an authentic standard in the product triene of 8S,15S-diHETE mixture of (inset chromophore short-term (Figure 10 to incu- Figure 10). Thisinvolved bations, compound, which conjugated the canonic was absenttriene in thechromophore product mixture (inset of to short-term incubations, Figure 10) and this consistent involved data the with canonic is consistent with itsitschemical conjugated structure. triene chromophore (inset to Figure 10) and this data is chemical structure. consistent with its chemical structure. Figure 10. Formation of 8,15-diHETE during long-term incubations of AA with mouse Alox15b. Wildtype mouse Alox15b was expressed as N-terminal his-tag fusion protein in E. coli and aliquots of the bacterial lysis supernatant were used to perform in vitro AA oxygenation activity assays. The Figure reaction Formation 10.was stopped ofafter 8,15-diHETE 1of (red) during and long-term 10 min (blue) incubations by of AA thelong-term addition with mouse of solid sodiumAlox15b. borohydride Figure 10. Formation 8,15-diHETE during incubations of AA wi Wildtype mouse Alox15b was expressed as N-terminal his-tag fusion protein and the reaction products were analyzed by RP-HPLC (legends to Figures 3 and in E. coli and4) aliquots of the recording Wildtype the bacterial mouse lysis Alox15b supernatant were was used expressed to perform in as N-terminal vitro AA his-tag oxygenation activity fusion assays. protein The in absorbance at 270 nm. Inset: UV-spectrum of the peak eluting at about 4 min indicating the presence of of the reaction bacterial was stopped the canonic lysis after supernatant conjugated 1triene (red) and 10 min were chromophore. (blue)used by theto perform addition in sodium of solid vitro AA oxygenation borohydride reaction wasproducts and the reaction stopped wereafter 1 (red) analyzed and 10(legends by RP-HPLC min (blue) by 3the to Figures and addition 4) recordingof thesolid so In absorbanceprinciple, at 270 there nm. Inset: are two distinct UV-spectrum of biosynthetic the peak eluting pathways at about 4 and the reaction products were analyzed by RP-HPLC (legends to Figures 3 an for min 8,15-diHETE indicating the formation presence of bythemouse canonicAlox15b. conjugatedThe triene chromophore. absorbance at 270 nm. first mechanism Inset: UV-spectrum involvesof primary the peakformation eluting at of about 8-HETE, which 4 min indi can subsequently be used as substrate for secondary 15-lipoxygenation. The second path- of the canonic conjugated triene chromophore. way involves primary AA 15-lipoxygenation followed by 8-lipoxygenation. In principle, there are two distinct biosynthetic pathways for 8,15-d Int. J. Mol. Sci. 2023, 24, 10046 14 of 25 Int. J. Mol. Sci. 2023, 24, x FOR PEER REVIEW 14 of 24 In principle, there are two distinct biosynthetic pathways for 8,15-diHETE formation by mouse Alox15b. The first mechanism involves primary formation of 8-HETE, which can subsequently be used as substrate for secondary 15-lipoxygenation. The second pathway involves primary AA 15-lipoxygenation followed by 8-lipoxygenation. 2.9. The Product Pattern of AA Oxygenation by Mouse Alox15b Depends on the Duration of the In Vitro 2.9. Incubation The Product Period Pattern andOxygenation of AA on the Amounts of Enzyme by Mouse Alox15bAdded Depends on the Duration of the In Vitro Incubation Period and on the Amounts of Enzyme After a 3 min incubation period of mouse Alox15b Added with AA, a 93:7 mixture of 8-HETE After a 3 min and 15-HETE wasincubation analyzedperiod (Figureof4A). mouse Alox15b Next, with AA,whether we explored a 93:7 mixture of 8-HETE the product pattern and 15-HETE was analyzed (Figure 4A). Next, we explored whether formed by mouse Alox15b depends on the enzyme concentration. For this purpose, the product pattern we formed by mouse first incubated Alox15b low depends amounts on the enzyme of enzymes concentration. for different For this time periods andpurpose, analyzedwethefirst prod- incubated low amounts of enzymes for different time periods and analyzed uct pattern (Figure 11A). Here we found that at all time points 8-HETE was the dominant the product pattern (Figure 11A). Here we found that at all time points 8-HETE was the dominant reaction product (black symbols) whereas 15-HETE as well as 8,15-diHETE were only reaction product (black symbols) whereas 15-HETE as well as 8,15-diHETE were only formed in small amounts. However, at higher enzyme concentrations this product pattern formed in small amounts. However, at higher enzyme concentrations this product pattern was remarkably different (Figure 11B). When activity assays were carried out at a 10-fold was remarkably different (Figure 11B). When activity assays were carried out at a 10-fold higher enzyme higher enzymeconcentration concentrationthetherelative relative share share of of 8-HETE 8-HETE formation formation waswas down, down, whereas whereas 8,15-diHETEformation 8,15-diHETE formation waswas augmented. augmented. In at In fact, fact, thisatenzyme this enzyme concentrations concentrations 8,15-diHETE 8,15-di- HETE was the dominant AA oxygenation product was the dominant AA oxygenation product (Figure 11B). (Figure 11B). Figure 11. Formation of different Figure 11. different oxygenation oxygenationproducts productsbybywildtype wildtypemouse mouseAlox15b Alox15bduring duringthethe ox- ygenation ofofAA. oxygenation AA.Wildtype Wildtype mouse Alox15bwas mouse Alox15b wasexpressed expressedas as N-terminal N-terminal his-tag his-tag fusion fusion protein protein in in coli and E. coli andaliquots aliquotsofofthe thebacterial bacterial lysis lysis supernatant supernatant were were usedused to perform to perform in vitro in vitro AA oxygenation AA oxygenation activity assays. activity assays. The The formation formation of 8-HETE, 8-HETE, 15-HETE, and 8,15-diHETE (nmoles) was 8,15-diHETE (nmoles) was quantified. quantified. (A) Time (A) dependence Time dependenceof product formation; of product single formation; values single are plotted. values (B) Product are plotted. (B) Productformation at different formation at enzyme concentrations. different Each point enzyme concentrations. Eachrepresents the mean point represents the four meandifferent assays; four different means assays; ± SD ± means areSDgiven. are given. 2.10. MD Simulations of 15-HETE and 8-HETE Binding 2.10. MD Simulations of 15-HETE and 8-HETE Binding To explore the structural basis for differential oxygenation of 15- and 8-HETE by hu- man To andexplore mouse theALOX15B structural basis for differential orthologs, we ran oxygenation of 15- and four independent MD 8-HETE by human simulations (100 ns and mouse ALOX15B orthologs, we ran four independent MD simulations (100 ns each). each). For wildtype human ALOX15B and for the humanized mouse Alox15b For wildtype human ALOX15B and for the humanized mouse Alox15b Y603D+H604V Y603D+H604V double mutant we analyzed TF alignment of 8-HETE. Since 15-HETE was double mutant we analyzed TF alignment of 8-HETE. Since 15-HETE was not a good not a good substrate forsubstrate these twofor these (Figure enzymes two enzymes 9) we did(Figure 9) we the not analyze didbinding not analyze of thisthe binding of substrate. this substrate. For wildtype mouse Alox15b and the For wildtype mouse Alox15b and the murinized human ALOX15B D602Y+V603H double murinized human ALOX15B D602Y+V603H double mutant, 15-HETE (HF) was the preferred substrate mutant, 15-HETE (HF) was the preferred substrate (Figure 9) and thus, we did not initially (Figure 9) and thus, we8-HETE explore did notbinding. initially explore 8-HETE binding. The evolution ofof The evolution thethe distances distances between between the oxygen the oxygen atom atom in the in − cofactor the Fe(III)-OH Fe(III)-OH − cofactor to to the the closest closest hydrogen hydrogen atomatom attached attached to C7C10 to C7 (H7), (H7), C10or(H10) (H10) or C13 C13 (H13) (H13) along the along the MD MD simu- lations is pictured simulations in Figure is pictured in 12, and the Figure 12, average and thedistances average are summarized distances in Table 4. Itin are summarized canTable be − 4. Itseen canthat H13 in be seen 8S-HETE that H13 in (TF) is localized 8S-HETE (TF) isinlocalized close proximity in closetoproximity the Fe(III)-OH to thecofactor Fe(III)-OH− and thus,and cofactor it is thus, readyittoisbe abstracted ready by the twoby to be abstracted enzymes. the twoThese stericThese enzymes. configurations are steric configura- consistent with the formation of 8S,15S-diHETE from the 8S-HETE substrate. tions are consistent with the formation of 8S,15S-diHETE from the 8S-HETE substrate. On On the other hand, the otherH10 hand, in 15S-HETE H10 in(TF) is also localized 15S-HETE in close (TF) is also proximity localized Fe(III)-OH−tocofactor to the proximity in close the Fe(III)- OH− cofactor and thus, it may also be ready to be abstracted by both wildtype mouse Alox15b and the murinized D602Y+V603H double mutant of human ALOX15B,. These conformations of the enzyme-substrate complexes explain the formation of 8S,15S-di- HETE from 15-HETE by these two enzymes, which was experimentally observed. Int. J. Mol. Sci. 2023, 24, 10046 15 of 25 and thus, it may also be ready to be abstracted by both wildtype mouse Alox15b and the murinized D602Y+V603H double mutant of human ALOX15B. These conformations of the Int. J. Mol. Sci. 2023, 24, x FOR PEER REVIEW enzyme-substrate complexes explain the formation of 8S,15S-diHETE from 15-HETE by these two enzymes, which was experimentally observed. Figure 12. Distances from the oxygen atom in the Fe(III)-OH− cofactor to the closest hydrogen atom Figure attached12. Distances to C7 (H7), C10 from (H10), the oxygen and C13 (H13)atom in the along the Fe(III)-OH molecular − cofactor to the closest hydro dynamics simulations of tail-first attached 8S-HETE in wildtype human ALOX15B and Y603D+H604V mutant of mouse Alox15b,dynamics to C7 (H7), C10 (H10), and C13 (H13) along the molecular and head-firstsimulatio first 8S-HETE in wildtype human ALOX15B and Y603D+H604V 15S-HETE in wildtype mouse Alox15b and D602Y+V603H mutant of human ALOX15B. mutant of mouse Alo head-first 15S-HETE in wildtype mouse Alox15b and D602Y+V603H mutant of human A Table 4. Average distances (Å) between the oxygen atom in the Fe(III)-OH− cofactor and the closest hydrogen Table atom attached 4. Average to C7 (Å) distances (H7),between C10 (H10),theand C13 (H13) oxygen atomalong the Fe(III)-OH in the molecular dynamics − cofactor and simulations of TF 8S-HETE alignment in wildtype human ALOX15B, wildtype mouse Alox15b, and hydrogen atom attached to C7 (H7), C10 (H10), and C13 (H13) along the molecular dynami Y603D+H604V mutant of mouse Alox15b, and head-first 15S-HETE in wildtype mouse Alox15b and tions of TF 8S-HETE alignment in wildtype human ALOX15B, wildtype mouse Alo D602Y+V603H mutant of human ALOX15B. n.p., not possible since neither C13 nor C10 are bisallylic Y603D+H604V mutant of mouse Alox15b, and head-first 15S-HETE in wildtype mouse Al methylenes in the corresponding substrates, n.c., such MD simulations have not been performed. D602Y+V603H mutant of human ALOX15B. n.p., not possible since neither C13 nor C10 are methylenes in theHuman corresponding substrates, n.c., such MD simulations Mouse have not been perfo Substrate and Orientation WT mut WT mut Human Mouse Substrate and H7 H10 H13 H7 H10 H13 H7 H10 H13 H7 H10 H13 15S-HETE (HF) WT n.c. n.c. n.c. 6.7 mut n.p. 4.0 5.6 4.5 WT n.p. n.c. n.c. n.c. mut Orientation 8S-HETE (TF) H7 5.8H10 n.p.H132.8 H7 n.c.H10 n.c. H136.9 H7 n.c. n.p. H10 6.3 H13n.p. H7 4.5 5.2 H10 15S-HETE (HF) n.c. n.c. n.c. 6.7 4.0 n.p. 5.6 4.5 n.p. n.c. n.c. 8S-HETE (TF) 5.8 n.p. 2.8 needs Another aspect n.c.to be considered n.c. n.c. when the6.9 n.p. formation 4.5 from 8S,15S-diHETE 6.3AA n.p. by wildtype mouse Alox15b is explored. We have shown that the primary formation of Another aspect needs to be considered when the formation 8S,15S-diHETE by wildtype mouse Alox15b is explored. We have shown that the primary format HETE is a favorable and fast process. In contrast, the formation of 8S,15S-diHETE Int. J. Mol. Sci. 2023, 24, 10046 16 of 25 8S-HETE is a favorable and fast process. In contrast, the formation of 8S,15S-diHETE from 8S-HETE is an unfavorable and slow process, but it is in fact possible. In effect, according to Table 2, the tail-first alignment of AA in wildtype mouse Alox15b is not the most favorable one, but it is possible. So, the tail-first alignment of 8S-HETE is also possible, and the corre- sponding MD simulation (Table 4 and Figure S5) shows that if this is the case, H13 is ready to be abstracted to yield the secondary 15-lipoxygenation. Moreover, Figure 9 shows that 8S- HETE is not a good substrate of wildtype mouse Alox15b, but it is somewhat metabolized. A second pathway would involve an initial slow 15-lipoxygenaton (it would imply an unfavorable tail-first orientation of AA), but a fast secondary 8-lipoxygenation of 15-HETE. Both pathways seem to be possible, as both are globally slow, because 15-lipoxygenation is always unfavorable. 2.11. Impact of Reaction Conditions on the Product Pattern of Mouse and Human ALOX15B For the soybean LOX1 it has previously been reported [32] that the pattern of the reaction products formed from linoleic acid depends on the pH of the reaction mixture. On the other hand, the product patterns of a number of vertebrate ALOX isoforms [33] were remarkably stable when the pH of the reaction mixture was altered. To explore the possible impact of the reaction conditions on the rate of EPA oxy- genation, we incubated human and mouse ALOX15B orthologs with EPA at different pH (6.4, 7.4, and 8.4), different temperatures (15 ◦ C, 25 ◦ C, and 35 ◦ C) and different substrate concentrations (50 µM, 100 µM, and 200 µM) for 5 min, quantified the reaction products formed during the incubation period and determined the chemical structure of the major EPA oxygenation products. As indicated in Table 5 (upper panel), human ALOX15B converted EPA most rapidly at pH 7.4. At pH 6.4 and 8.4 the oxygenation rates were gradually reduced but at all pH-values 15-HEPE was the exclusive oxygenation product. As expected, for mouse Alox15b 8-HEPE was the dominant EPA oxygenation product and 15-HEPE was only pro- duced in small quantities. Interestingly, in contrast to human ALOX15B, which exhibited a higher EPA oxygenation rate at pH 7.4, mouse Alox15b was more active at pH 6.4. (Table 5, upper panel). When we explored the temperature dependence of EPA oxygenation (Table 5, middle panel) we found that both human and mouse ALOX15B oxygenated EPA with highest rates at 37 ◦ C. At all temperatures 15-HEPE was the exclusive EPA oxygenation product of the human enzyme, whereas 8-HEPE was dominant for mouse Alox15b. Finally, we also tested the impact of substrate concentrations (Table 5, lower panel). Human and mouse ALOX15B orthologs exhibited the highest EPA oxygenase activities at 200 µM substrate concentration. Here again, the exclusive EPA oxygenation product was identified as 15-HEPE for human ALOX15B at all substrate concentrations tested. For mouse Alox15b the situation was somewhat different. At low substrate concentrations (50 µM) a 1:2 mixture of 15- and 8-HEPE was analyzed but the relative share of 8-HEPE did increase with the increasing substrate concentrations. In fact, at 200 µM EPA 8-HEPE was rather dominant (Table 5, lower panel). Int. J. Mol. Sci. 2023, 24, 10046 17 of 25 Table 5. Impact of reaction conditions on the rate of EPA oxygenation and the product pattern of human and mouse wildtype ALOX15B orthologs. Aliquots of bacterial lysis supernatants (5 µL) of recombinant ALOX15B preparations (wildtype human and mouse ALOX15B) were incubated for 5 min at different conditions (pH 6.4, 7.4, 8.4, 100 µM EPA, room temperature; temperatures 15 ◦ C, 25 ◦ C, 35 ◦ C, 100 µM EPA, room temperature; EPA concentrations 50 µM, 100 µM, 200 µM, pH 7.4, room temperature). Sample workup and RP-HPLC analysis were carried out as described in the legend to Figure 4B. Three independent incubations were carried out under each condition (n = 3) and means ± SD are given. pH Dependence Product Pattern rel. Activity Enzyme pH 15-HEPE 8-HEPE (%) 6.4 77.1 ± 3.6 100.0 ± 0.0 0.0 ± 0.0 Human 7.4 100.0 ± 6.3 100.0 ± 0.0 0.0 ± 0.0 ALOX15B 8.4 98.4 ± 6.6 100.0 ± 0.0 0.0 ± 0.0 6.4 100 ± 2.6 2.6 ± 0.5 97.4 ± 0.5 Mouse Alox15b 7.4 62.1 ±1.0 1.5 ± 0.1 98.5 ± 0.1 8.4 65.9 ± 10.4 3.8 ± 0.1 96.2 ± 0.1 Temperature Dependence Product Pattern rel. Activity Enzyme Temperature (◦ C) 15-HEPE 8-HEPE (%) 15 47.6 ±2.4 100.0 ± 0.0 0.0 ± 0.0 Human 25 73.7 ± 6.1 100.0 ± 0.0 0.0 ± 0.0 ALOX15B 35 100.0 ± 8.5 100.0 ± 0.0 0.0 ± 0.0 15 9.8 ± 0.9 5.8 ± 0.5 94.2 ± 0.5 Mouse Alox15b 25 54.2 ± 7.0 6.7 ± 0.6 93.3 ± 0.6 35 100.0 ±13.4 7.2 ± 1.3 92.8 ± 1.3 Substrate Concentration Dependence Product Pattern EPA Concentration rel. Activity Enzyme 15-HEPE 8-HEPE (µM) (%) 50 27.0 ± 3.6 100.0 ± 0.0 0.0 ± 0.0 Human 100 70.6 ± 0.8 100.0 ± 0.0 0.0 ± 0.0 ALOX15B 200 100.0 ± 10.9 100.0 ± 0.0 0.0 ± 0.0 50 11.6 ± 1.7 32.5 ± 1.3 67.5 ± 1.3 Mouse Alox15b 100 88.7 ± 2.0 15.1 ± 1.2 84.9 ± 1.2 200 100.0 ± 14.6 5.1 ± 0.5 94.9 ± 0.5 3. Discussion 3.1. Degree of Novelty, Advancement of Knowledge, and Limitations The product pattern of wildtype mouse and human ALOX15B orthologs with AA as the substrate has previously been explored and product analysis indicated that human ALOX15B converts AA mainly to 15S-HETE [19,25] whereas 8S-HETE was the dominant AA oxygenation product of mouse Alox15b [25]. Y603D+H604V exchange in mouse Alox15b humanized the product pattern with AA as the substrate and the inverse mutagenesis strategy on the mouse enzyme partly humanized this enzyme property [25]. For wildtype human ALOX15B the product pattern with EPA and DHA has also been explored [34], but a number of important questions remained unanswered: (i) Which products are formed from other PUFAs by mouse Alox15b. (ii) Does D602Y+V603H exchange in human ALOX15B murinize the product pattern of this enzyme when EPA, DHA, and LA are used as the Int. J. Mol. Sci. 2023, 24, 10046 18 of 25 substrate. (iii) Does Y603D+H604V exchange in mouse Alox15b humanize the product mixture of this enzyme when EPA, DHA, and LA are used as substrate. (iv) Do wildtype mouse and human ALOX15B orthologs and their humanized/murinized double mutants accept hydroxylated PUFAs (8-HETE and 15-HETE) as the substrate and what are the major products of these reactions. (v) Can the different reaction specificities of mouse and human ALOX15B orthologs and their humanized/murinized double mutants be explained by the different ways of substrate binding at the active site of the enzyme variants (docking studies and MD simulations). To address these questions, we designed appropriate experiments and obtained the following results: (i) 8-HEPE and 10-HDHA were the dominant oxygenation products of mouse Alox15b with EPA and DHA, respectively (Figure 3). (ii) Mouse Alox15b (Figure 4) con- verted AA, EPA, and DHA to their n-12 oxygenation products (8-HETE, 8-HEPE, 10-HDHA) and the Y603D+H604V exchange humanized the product patterns with all three PUFAs. A similar humanization of the product mixture of mouse Alox15b was observed when linoleic acid was used as the substrate (Figure 5C,D). (iii) When an inverse mutagenesis strategy was employed for human ALOX15B, the D602Y+V603H exchange murinized the AA and EPA oxygenation products (Figure 4A,B). In contrast, when DHA was used as the substrate for the D602Y+V603H double mutant of human ALOX15B (Figure 4C) we did not see such murinization. Instead of 10-HDHA, which was expected to be formed by a murinized en- zyme, we identified 7-HDHA as the major DHA oxygenation product. In other words, the D602Y+V603H double mutant of human ALOX15B converted DHA to the same reaction product that is formed from this substrate by ALOX5. This data suggests that the way of DHA binding at the active site of the D602Y+V603H double mutant of human ALOX15B should be similar in the way of substrate binding at ALOX5. When LA was used as the substrate, we neither observed murinization of the reaction products. In fact, the D602Y+V603H double mutant of human ALOX15B produced an unspecific product mixture of 13S-, 13R-, 9S-, and 9R-HODE and this data suggests that the oxygenation reaction with this fatty acid was not strictly controlled by the enzyme (Figure 5A,B). (iv) 15-HETE is a good substrate for wild- type mouse Alox15b and for the murinized human ALOX15B double mutant Y603D+H604V. In contrast, it is not well oxygenated by wildtype human ALOX15B and the humanized D602Y+V603H double mutant of mouse Alox15b. In contrast, 8-HETE is well oxygenated by wildtype human ALOX15B and the humanized mouse Alox15b double mutant Y603D+H604V. 8,15-diHETE was the major oxygenation product formed from both substrates. (v) Docking calculations indicate that the four polyenoic fatty acid substrates (AA, EPA, DHA, and LA) tested in this study preferentially tend to adopt a tail-first (TF) substrate alignment in wildtype human ALOX15B and in the humanized Y603D+H604V double mutant of mouse Alox15b. In contrast, a head-first (HF) substrate alignment was calculated for murinized D602Y+V603H mutant of human ALOX15B and wildtype mouse Alox15b. Molecular dynamics simulations using AA as the substrate revealed that binding of carboxylate group of the head-first oriented substrate in the substrate binding pocket would be destabilized by D602 and V603 in wildtype human ALOX15B, and by D603 and V604 in the humanized Y603D+H604V double mutant of mouse Alox15b. Thus, the substrate fatty acid is unable to establish a strong non-covalent interaction with any active site amino acid residue present at the bottom of the substrate binding pocket. Conversely, the strong interaction between the negatively charged carboxylic group of the substrate and the negatively charged side-chain of Arg428 stabilizes the TF substrate alignment. When Tyr and a His residues are present in the active site (wildtype mouse Alox15b and the murinized D602Y+V603H double mutant of human ALOX15B), the fatty acid carboxylate can form stabilizing hydrogen bonds with amino acid residues present either in the bottom region of the substrate binding cavity (HF substrate alignment) or with amino acids located close to the entrance of the substrate binding pocket (TF alignment). However, the docking results suggest that HF substrate orientation appears to be preferred. (vi) For each preferred substrate alignment both H13 (attached to C13) and H10 (attached to C10) of AA are located sufficiently close to the oxygen of the Fe(III)-OH− cofactor for hydrogen abstraction, and steric hindrance by the N-terminal Ile675 and by the side-chain of Leu609 is Int. J. Mol. Sci. 2023, 24, 10046 19 of 25 important. For TF, AA alignment at the active site of wildtype human ALOX15B and of the humanized Y603D+H604V double mutant of mouse Alox15b the H13 abstraction is preferred. In contrast, for HF, AA alignment at the active site of wild type mouse Alox15B and of the murinized D602Y+V603H double mutant of human ALOX15B H10 abstraction is preferred. The most serious limitation of this study is that functional characterization of the four enzyme variants was carried out with crude enzyme preparations. Although we were able to purify wildtype human and mouse ALOX15B as well as theY603D+H604V double mutant of mouse Alox15b by affinity chromatography on Ni-agarose as catalytically active proteins, we experienced a >90% loss in catalytic of the D602Y+V603H double mutant of human ALOX15B during the purification procedure. The mechanistic basis for this loss in catalytic activity has not been clarified but to ensure strict comparability of the functional data we decided to use crude enzyme preparations (bacterial lysate supernatant) for functional characterization. For human and mouse wildtype ALOX15B orthologs we carried out similar experiments with the purified enzyme preparations and obtained similar data as for the bacterial lysates. 3.2. Complex Ester Lipids as Substrate for Mouse and Human ALOX15B Despite the differential product patterns with free AA human and mouse ALOX15, orthologs convert AA-containing phospholipids incorporated in nanodiscs [35] mainly to the 15-HETE derivatives [11]. Thus, the product mixture of the two enzymes with this substrate appears to be similar. We carried out similar studies using AA containing phospholipid liposomes and mitochondrial membranes as substrates. These preliminary experiments indicated that 15S-HETE containing phospholipids were the major oxygenation products formed by mouse and human ALOX15B and this data is consistent with the results obtained in the nanodisc experiment [11]. However, when we employed mitochondrial membranes or liposomes involving linoleic acid as substrate for the two enzymes, we observed different product patterns. For human ALOX15B, esterified 15S-HETE and 13S-HODE were the major oxygenation products, and the corresponding R-enantiomers were only formed in small amounts. When incubations were carried out with mouse Alox15b, esterified AA oxygenation products were only detected in small amounts. As the dominant PUFA oxygenation product we identified 9S-HODE. Thus, when mitochondrial membranes and LA-containing liposomes were used as substrate the product specificities of human and mouse ALOX15B are different. Whether similar differences in the reaction specificities of the two enzymes can be observed when other lipid-protein such as plasma membranes of different cells or lipoproteins are used as substrate must be tested experimentally in the future. 3.3. Impact of Reaction Conditions on the Product Pattern of ALOX15B Variants The composition of the reaction products of most ALOX-isoforms with a given sub- strate is well defined and alterations in the reaction conditions in the near physiological range do not dramatically alter this enzyme property. There may be minor modifications but in principle the reaction specificities of ALOX-isoforms are very robust. This conclusion is supported by the experimental date shown in Table 5. On the other hand, we found that the pattern of the arachidonic acid oxygenation products strongly depends on the amount of enzyme added to the incubation mixture (Figure 11B). However, these data do not indicate that the specificity of the enzyme has changed. In fact, the differences in the product pattern are the consequence of enzyme kinetics. At high enzyme concentra- tions the fatty acid substrate (AA) becomes rate limiting and the primary reaction product accumulates. Under these conditions the primary oxygenation product is used as the substrate for secondary oxygenation (formation of 8,15-diHETE). Thus, also at high enzyme concentration the specificity of the enzyme largely remains unchanged. There are, however, rare exceptions from this rule. For instance, the specificity of soybean LOX1 is more variable since the pattern of linoleic acid oxygenation products formed by this enzyme is pH-sensitive [32]. This result might be related to the alkaline pHopt of this ALOX isoform. On the other hand, the specificity of AA oxygenation by Int. J. Mol. Sci. 2023, 24, 10046 20 of 25 most vertebrate ALOX-isoforms (human ALOX5, human ALOX12, human ALOX15, rabbit ALOX15, human ALOX15B, mouseAlox15, mouse Alox15b, zebrafishALOX12) was remark- able stable when the pH of the incubation mixture was varied in the near physiological range (pH 6.4–8.0) [33]. Here we observed (Table 5) that the product pattern of human ALOX15B was remarkably stable at different reaction conditions. At different pH-values (pH 6.4, 7.4, 8.4), at different temperatures (15 ◦ C, 25 ◦ C, 35◦ C), and at different substrate concentrations (50 µM, 100 µM, 200 µM) 15-HEPE was always identified as the exclusive EPA oxygenation product. In other words, in the near physiological range alterations in the reaction conditions did not impact the product pattern of human ALOX15B. For mouse Alox15b the situation was somewhat different. For this enzyme we also observed that alterations in pH and temperature did hardly modify the product pattern of EPA oxygenation. Under all experimental conditions tested in our experiments (Table 5) 8-HEPE was identified as the dominant EPA oxygenation product (>90%) while 15-HEPE was only formed in smaller quantities (<10%). Interestingly, modification of the substrate concentration induced alterations in the reaction products of mouse Alox15b. At high substrate concentrations (200 µM) the relative share of the major EPA oxygenation product (8-HEPE) was about 95%. In contrast, at lower substrate concentrations we observed a gradual decline in the relative share of 8-HEPE formation. In fact, at 100 µM EPA 8-HEPE only contributed 85% to the product mixture and at 50 µM the 8-HEPE/15-HEPE ratio was only about 2:1. Thus, although at lower substrate concentrations 8-HEPE was still the major EPA oxygenation product, the relative share of 15-HEPE did significantly increase at lower substrate concentrations. The molecular basis for this observation was not explored but our findings need to be considered when the biosynthesis of oxygenated lipids in mouse cells, tissues, and body fluids (blood palsma, urine, liquor) are interpreted. 4. Materials and Methods 4.1. Chemicals The chemicals used in this study were purchased from the following sources: Phosphate buffered saline without calcium and magnesium (PBS) from PAN Biotech (Aidenbach, Germany); nitrocellulose blotting membrane from Serva Electrophoresis GmbH (Heidelberg, Germany); EDTA (Merck KG, Darmstadt, Germany). Fatty acid substrates [5,8,11,14-eicosatetraenoic acid (AA), 5,8,11,14,17-eicosapentaenoic acid (EPA), 4,7,10,13,16,19-docosahexaenoic acid (DHA), 9,12-octadecadienoic acid (LA), 9,12,15- octadecatrienoic acid (ALA), 6,9,12-octadecatrienoic acid (GLA) 15S-HETE, 8S-HETE], and authentic HPLC standards of HETE-isomers used for HPLC analysis [15S-HETE, 15S/R-HETE, 12S/R-HETE, 12S-HETE, 8S/R-HETE, 5S-HETE, 8S,15S-diHETE] were ob- tained from Cayman Chem (distributed by Biomol GmbH, Hamburg, Germany). Acetic acid from Carl Roth GmbH (Karlsruhe, Germany); sodium borohydride from Life Tech- nologies, Inc (Eggenstein, Germany); isopropyl-β-thiogalactopyranoside (IPTG) from Carl Roth GmbH (Karlsruhe, Germany); restriction enzymes from ThermoFisher (Schw- erte, Germany); the E. coli strain Rosetta2 DE3 pLysS from Novagen (Merck-Millipore, Darmstadt, Germany). Oligonucleotide synthesis was performed at BioTez Berlin Buch GmbH (Berlin, Germany). Nucleic acid sequencing was carried out at Eurofins MWG Operon (Ebersberg, Germany). HPLC grade methanol, acetonitrile, n-hexane, 2-propanol, and water were from ThermoFisher (Schwerte, Germany). 4.2. Cloning and Expression of Mouse and Human ALOXB Orthologs The cDNA sequences of mouse and human ALOX15B orthologs were extracted from the NCBI cDNA database and were chemically synthesized (BioCat GmbH, Heidelberg, Germany). For subcloning from the initial pUC57 synthesis vector into pET28b(+) (No- vagen/Merck, Darmstadt, Germany) expression vector, a SalI restriction site was introduced immediately upstream of the start codon and a HindIII restriction site was generated im- mediately downstream of the stop codon. The sequences were optimized for bacterial expression by silent mutations. Int. J. Mol. Sci. 2023, 24, 10046 21 of 25 The enzyme variants (wildtype and double mutants) were N-terminal hexa-his-tag fusion proteins, as described in Ref. [36]. Briefly, E. coli Rosetta2 (DE3)-pLysS cells were transformed with the recombinant plasmid pET28b and grown on an agar plate containing kanamycin and chloramphenicol. An isolated clone was selected and grown under shaking in a glucose free MSM with added trace elements and glucoamylase. Protein expression was induced by addition of IPTG. After 18 h, cells were harvested and homogenized by sonication, and the cell free supernatant was used as enzyme source. 4.3. Mutagenesis Studies Site directed mutagenesis was performed as described in Ref. [36]. Plasmid DNA containing the coding sequence of the bony fish ALOX15 isoforms was incubated with specific primer pairs containing the required changes in their nucleotide sequence to achieve the intended amino acid exchanges and Pfu Ultra II Hot Start 2 × PCR Master Mix (Agilent Technologies, Waldbronn, Germany). After the PCR protocol (18 cycles), parent DNA was digested using DpnI. E. coli XL-1 Blue competent cells (Agilent Technologies Inc., Santa Clara, CA, USA) were transformed with the mutated plasmid and after replication the plasmid DNA of a selected bacterial colony was sequenced (Eurofins Genomics Germany GmbH, Ebersberg, Germany). 4.4. In Vitro Activity Assays and HPLC Analyses of the Reaction Products Enzyme activity assays and analysis of the reaction products by HPLC were performed as described in Ref. [36]. Different amounts of the cell free supernatants were incubated with 100 µM substrate (AA, EPA, and DHA) and the reduced reaction products were analyzed by RP-HPLC on a Shimadzu instrument connected with a Hewlett Packard diode array detector 1040 A. Metabolites were separated on a Nucleodur C18 Gravity column (Macherey-Nagel, Düren, Germany; 250 × 4 mm, 5 µm particle size) coupled with a corresponding guard column (8 × 4 mm, 5 µm particle size). A solvent system consisting of acetonitrile:water:acetic acid (70:30:0.1, by volume for AA and DHA derivatives and 50:50:0.1, by volume for EPA derivatives) was used at a flow rate of 1 mL·min−1 . For more detailed analysis of the reaction products of linoleic acid oxygenation the conjugated dienes formed during the incubation period were prepared by RP-HPLC and further analyzed by normal-phase HPLC (NP-HPLC) and/or chiral-phase HPLC (CP-HPLC). Normal-phase HPLC was performed using the solvent system n-hexane/2-propanol/acetic acid (100/2/0.1, by volume) on a Nucleosil 100-5 column (250 × 4.6 mm, 5 µm particle size). 12-HETE enantiomers were resolved on a Chiralpak AD-H column (Daicel Corp., Osaka, Japan) with a solvent system consisting of n-hexane/methanol/ethanol/acetic acid (96:3:1:0.1, by vol; 1 mL/min). 4.5. Miscellaneous Methods including Statistics For SDS/PAGE, approximately 100 µg denatured protein of the bacterial lysate super- natants were analyzed on a 7.5% polyacrylamide gel. The separated protein bands were then transferred onto a Protran BA 85 Membrane (Carl Roth GmbH, Karlsruhe, Germany) and the blots were probed with an anti-his-tag-HRP conjugated antibody (Miltenyi Biotec GmbH, Bergisch Gladbach, Germany). Immunoreactive bands were visualized using the SERVALight Polaris CL HRP WB Substrate Kit (Serva Electrophoresis GmbH, Heidelberg, Germany). Chemiluminescence was detected on a FUJIFILM Luminescent Image Analyzer LAS-1000plus & Intelligent Dark Box II. For testing the pH-dependence of the product mixture of wildtype human and mouse ALOX15B orthologs, a 1:1 mixtures of 0.05 M sodium phosphate buffer and 0.05 M sodium borate buffer were used and the different pH values were adjusted at room temperature by the addition of 5 M NaOH or HCl, respectively. The protein concentrations in the bacterial lysates were quantified using Bradford Reagent for quantitative protein de- termination (AppliChem, VWR International GmbH, Darmstadt, Germany) according to the instructions of the vendor. For statistical calculations and figure design we used the GraphPad prism program (version 8.00, GraphPad Software, La Jolla, CA, USA). Int. J. Mol. Sci. 2023, 24, 10046 22 of 25 4.6. Structural Modeling Four ALOX15B 3D models were built. Human ALOX15B coordinates were obtained from the X-ray structure with PDB ID 4NRE [37] and the mouse Alox15b model was obtained from AlphaFold DataBase (UniProt ID O35936) [38,39]. The double mutant versions of both human ALOX15B and mouse Alox15b were manually constructed using the Rotamers module by UCSF Chimera. Protonation was carried out at pH 7.4 using the ProteinPrepare tool by PlayMolecule.org [40]. 4.7. Docking Calculations The GOLD5.8 program [41] was used for the docking calculations. ChemScore [42] was employed for scoring purposes. 4.8. MD Simulations The best-scored solutions for each substrate-enzyme complex (enzyme complexed with AA in head-first (HF) and in tail-first (TF) conformations, 15S-HETE in head-first conformation and 8S-HETE in tail-first conformation) from docking calculations were used as the initial structures for MD simulations. Ff19SB has been used as a force field for standard residues, while Fe and its coordinates have been parametrized following the Seminario Method from MCPB [43]. Force constants and RESP charges have been obtained using the Gaussian16 program at the B3LYP/6-31G(d) level of theory. As recommended in Ref. [44], OPC waters [45] were chosen as the water model. An orthorhombic solvation box of pre-equilibrated water molecules was built with a buffer of 10 Å around the molecule and Na+ ions have been added for neutralization. The AMBER20’s CUDA (GPU) pmemd package was used for running the simulations. The initial structure was minimized, heated at a rate of 50 K from 0 to 300 K, and the system was then equilibrated for 20 ps after each step. A total of 1 ns of NPT ensemble at 300 K and 1 atm was run for stabilization of the water box density. Prior to production, an equilibration of the system using the NVT ensemble at 300 K during 10 ns was performed. After equilibration, 100 ns of production was calculated using the same NVT ensemble. 5. Conclusions Functional experiments with different polyenoic fatty acids, molecular docking studies and molecular dynamics simulations of substrate binding at the active site of human and mouse ALOX15B orthologs and their murinized/humanized double mutants suggest that AA 15-lipoxygenation by human ALOX15 involves tail-first (TF) substrate alignment. In contrast, for AA 8-lipoxygenation by mouse Alox15b head-first (HF) substrate orienta- tion is preferred. The Asp602Tyr+Val603His exchange murinized the product mixture of human ALOX15B with arachidonic acid and eicosapentaenoic acid but not with docosahex- aenoic acid. An inverse mutagenesis strategy on mouse Alox15b (Tyr603Asp+His604Val exchange) did humanize the product pattern of arachidonic acid, eicosapentaenoic acid, and docosahexaenoic acid oxygenation. Supplementary Materials: The supporting information can be downloaded at: https://www.mdpi. com/article/10.3390/ijms241210046/s1. Author Contributions: Conceptualization, H.K. and À.G.-L.; methodology, À.G.-L., K.R.K., H.K. and M.C.-N.; validation, H.K. and À.G.-L.; investigation, K.R.K., M.C.-N., X.C., J.M.L., À.G.-L. and H.K.; data curation, K.R.K., H.K., M.C.-N., J.M.L. and À.G.-L.; writing—original draft preparation, H.K. and À.G.-L.; writing—review and editing, all co-authors; visualization, K.R.K., H.K., M.C.-N., J.M.L. and À.G.-L.; supervision, H.K. and À.G.-L.; funding acquisition, À.G.-L. and J.M.L. All authors have read and agreed to the published version of the manuscript. Funding: The experimental part of the study was supported by research grants of Deutsche Forschungs- gemeinschaft to H.K. and D.H. (KU961/14-1, HE8295/1-1). The in silico part of this study was supported by a grant (PID2020-113764GB-I00) from the Spanish Ministerio de Ciencia e Innovación. Int. J. Mol. Sci. 2023, 24, 10046 23 of 25 Institutional Review Board Statement: Not applicable. Informed Consent Statement: Not applicable. Data Availability Statement: All data generated or analyzed during this study are included in this published article. Original experimental raw data can be obtained upon request from K.R.K. (in vitro experiments) and A.G.L. (in silico experiments). Acknowledgments: The authors acknowledge the methodological support of S. Stehling (Department of Biochemistry, Charité). Conflicts of Interest: The authors declare that they do not have any conflict of interest with the content of this article. Abbreviations ALOX, arachidonic acid lipoxygenases; 5-HETE, 5S-hydroxy-6E,8Z,11Z,14Z-eicosatetraenoic acid; 8-HETE, 8-hydroxy-5Z,9E,11Z,14Z-eicosatetraenoic acid; 15-HETE, 15S-hydroxy-5Z,8Z,11Z, 13E-eicosatetraenoic acid; 12-HETE, 12-hydroxy-5Z,8Z,10E,14Z-eicosatetraenoic acid; AA, arachi- donic acid; EPA, 5,8,11,14,17-eicosapentaenoic acid; DHA, 4,7,10,13,16,19-docosahexaenoic acid; LA, linoleic acid; 8-HEPE, 8-hydroxy-5Z,9E,11Z,14Z,17Z-eicosapentaenoic acid; 15-HEPE, 15S- hydroxy-5Z,8Z,11Z,13E,17Z-eicosapentaenoic acid; 5-HEPE, 5-hydroxy-6E,8Z,11Z,14Z,17Z- eicosapentaenoic acid; 17-HDHA, 17-hydroxy DHA, 17-hydroxy-4Z,7Z,10Z,13Z,15E,19Z- docosahexaenoic acid; 10-HDHA, 10-hydroxy-4Z,7Z,11E,13Z,16Z,19Z-docosahexaenoic acid; 7-HDHA, 7-hydroxy-4Z,8E,10Z,13Z,16Z,19Z-docosahexaenoic acid; 8,15-diHETE, 8S,15S- dihydroxy-5Z,9E,11Z,13E-eicosatetraenoic acid; TF, tail-first substrate alignment; HF, head-first substrate alignment. References 1. Funk, C.D.; Chen, X.-S.; Johnson, E.N.; Zhao, L. Lipoxygenase genes and their targeted disruption. Prostaglandins Other Lipid Mediat. 2002, 68–69, 303–312. [CrossRef] 2. Kuhn, H.; Humeniuk, L.; Kozlov, N.; Roigas, S.; Adel, S.; Heydeck, D. The evolutionary hypothesis of reaction specificity of mammalian ALOX15 orthologs. Prog. Lipid Res. 2018, 72, 55–74. [CrossRef] 3. Funk, C.D.; Keeney, D.S.; Oliw, E.H.; Boeglin, W.E.; Brash, A.R. Functional expression and cellular localization of a mouse epidermal lipoxygenase. J. Biol. Chem. 1996, 271, 23338–23344. [CrossRef] 4. Haeggstrom, J.Z.; Funk, C.D. Lipoxygenase and leukotriene pathways: Biochemistry, biology, and roles in disease. Chem. Rev. 2011, 111, 5866–5898. [CrossRef] [PubMed] 5. Kuhn, H.; Banthiya, S.; van Leyen, K. Mammalian lipoxygenases and their biological relevance. Biochim. Biophys. Acta 2015, 1851, 308–330. [CrossRef] 6. Flohe, L.; Toppo, S.; Orian, L. The glutathione peroxidase family: Discoveries and mechanism. Free Radic. Biol. Med. 2022, 187, 113–122. [CrossRef] [PubMed] 7. Kanaoka, Y.; Austen, K.F. Roles of cysteinyl leukotrienes and their receptors in immune cell-related functions. Adv. Immunol. 2019, 142, 65–84. [CrossRef] [PubMed] 8. Jordan, P.M.; Werz, O. Specialized pro-resolving mediators: Biosynthesis and biological role in bacterial infections. FEBS J. 2022, 289, 4212–4227. [CrossRef] [PubMed] 9. Schebb, N.H.; Kuhn, H.; Kahnt, A.S.; Rund, K.M.; O’Donnell, V.B.; Flamand, N.; Peters-Golden, M.; Jakobsson, P.J.; Weylandt, K.H.; Rohwer, N.; et al. Formation, Signaling and Occurrence of Specialized Pro-Resolving Lipid Mediators-What is the Evidence so far? Front. Pharmacol. 2022, 13, 838782. [CrossRef] 10. Aloulou, A.; Rahier, R.; Arhab, Y.; Noiriel, A.; Abousalham, A. Phospholipases: An Overview. Methods Mol. Biol. 2018, 1835, 69–105. [CrossRef] 11. Bender, G.; Schexnaydre, E.E.; Murphy, R.C.; Uhlson, C.; Newcomer, M.E. Membrane-dependent Activities of Human 15-LOX-2 and Its Murine Counterpart: Implications for Murine Models of Atherosclerosis. J. Biol. Chem. 2016, 291, 19413–19424. [CrossRef] [PubMed] 12. Ulusan, S.; Sheraj, I.; Stehling, S.; Ivanov, I.; Das, A.; Kuehn, H.; Banerjee, S. Structural and functional evaluation mammalian and plant lipoxygenases upon association with nanodics as membrane mimetics. Biophys. Chem. 2022, 288, 106855. [CrossRef] [PubMed] 13. Schewe, T.; Halangk, W.; Hiebsch, C.; Rapoport, S.M. A lipoxygenase in rabbit reticulocytes which attacks phospholipids and intact mitochondria. FEBS Lett. 1975, 60, 149–152. [CrossRef] [PubMed] Int. J. Mol. Sci. 2023, 24, 10046 24 of 25 14. Kuhn, H.; Belkner, J.; Wiesner, R.; Brash, A.R. Oxygenation of biological membranes by the pure reticulocyte lipoxygenase. J. Biol. Chem. 1990, 265, 18351–18361. [CrossRef] 15. Aldrovandi, M.; Banthiya, S.; Meckelmann, S.; Zhou, Y.; Heydeck, D.; O’Donnell, V.B.; Kuhn, H. Specific oxygenation of plasma membrane phospholipids by Pseudomonas aeruginosa lipoxygenase induces structural and functional alterations in mammalian cells. Biochim. Biophys. Acta 2017, 1863, 152–164. [CrossRef] 16. Belkner, J.; Wiesner, R.; Rathman, J.; Barnett, J.; Sigal, E.; Kuhn, H. Oxygenation of lipoproteins by mammalian lipoxygenases. Eur. J. Biochem. 1993, 213, 251–261. [CrossRef] 17. Pirillo, A.; Uboldi, P.; Kuhn, H.; Catapano, A.L. 15-Lipoxygenase-mediated modification of high-density lipoproteins impairs SR-BI- and ABCA1-dependent cholesterol efflux from macrophages. Biochim. Biophys. Acta 2006, 1761, 292–300. [CrossRef] 18. Kuhn, H.; Belkner, J.; Suzuki, H.; Yamamoto, S. Oxidative modification of human lipoproteins by lipoxygenases of different positional specificities. J. Lipid Res. 1994, 35, 1749–1759. [CrossRef] 19. Brash, A.R.; Boeglin, W.E.; Chang, M.S. Discovery of a second 15S-lipoxygenase in humans. Proc. Natl. Acad. Sci. USA 1997, 94, 6148–6152. [CrossRef] 20. Wuest, S.J.; Crucet, M.; Gemperle, C.; Loretz, C.; Hersberger, M. Expression and regulation of 12/15-lipoxygenases in human primary macrophages. Atherosclerosis 2012, 225, 121–127. [CrossRef] 21. Ebert, R.; Cumbana, R.; Lehmann, C.; Kutzner, L.; Toewe, A.; Ferreiros, N.; Parnham, M.J.; Schebb, N.H.; Steinhilber, D.; Kahnt, A.S. Long-term stimulation of toll-like receptor-2 and -4 upregulates 5-LO and 15-LO-2 expression thereby inducing a lipid mediator shift in human monocyte-derived macrophages. Biochim. Biophys. Acta Mol. Cell Biol. Lipids 2020, 1865, 158702. [CrossRef] [PubMed] 22. Benatzy, Y.; Palmer, M.A.; Brune, B. Arachidonate 15-lipoxygenase type B: Regulation, function, and its role in pathophysiology. Front. Pharmacol. 2022, 13, 1042420. [CrossRef] [PubMed] 23. Furstenberger, G.; Hagedorn, H.; Jacobi, T.; Besemfelder, E.; Stephan, M.; Lehmann, W.D.; Marks, F. Characterization of an 8-lipoxygenase activity induced by the phorbol ester tumor promoter 12-O-tetradecanoylphorbol-13-acetate in mouse skin in vivo. J. Biol. Chem. 1991, 266, 15738–15745. [CrossRef] [PubMed] 24. Jisaka, M.; Kim, R.B.; Boeglin, W.E.; Nanney, L.B.; Brash, A.R. Molecular cloning and functional expression of a phorbol ester-inducible 8S-lipoxygenase from mouse skin. J. Biol. Chem. 1997, 272, 24410–24416. [CrossRef] 25. Jisaka, M.; Kim, R.B.; Boeglin, W.E.; Brash, A.R. Identification of amino acid determinants of the positional specificity of mouse 8S-lipoxygenase and human 15S-lipoxygenase-2. J. Biol. Chem. 2000, 275, 1287–1293. [CrossRef] 26. Schafer, M.; Kakularam, K.R.; Reisch, F.; Rothe, M.; Stehling, S.; Heydeck, D.; Puschel, G.P.; Kuhn, H. Male Knock-in Mice Expressing an Arachidonic Acid Lipoxygenase 15B (Alox15B) with Humanized Reaction Specificity Are Prematurely Growth Arrested When Aging. Biomedicines 2022, 10, 1379. [CrossRef] 27. Goloshchapova, K.; Stehling, S.; Heydeck, D.; Blum, M.; Kuhn, H. Functional characterization of a novel arachidonic acid 12S- lipoxygenase in the halotolerant bacterium Myxococcus fulvus exhibiting complex social living patterns. Microbiologyopen 2019, 8, e00775. [CrossRef] 28. Wecksler, A.T.; Kenyon, V.; Deschamps, J.D.; Holman, T.R. Substrate specificity changes for human reticulocyte and epithelial 15-lipoxygenases reveal allosteric product regulation. Biochemistry 2008, 47, 7364–7375. [CrossRef] 29. Wecksler, A.T.; Kenyon, V.; Garcia, N.K.; Deschamps, J.D.; van der Donk, W.A.; Holman, T.R. Kinetic and structural investigations of the allosteric site in human epithelial 15-lipoxygenase-2. Biochemistry 2009, 48, 8721–8730. [CrossRef] 30. Suardiaz, R.; Jambrina, P.G.; Masgrau, L.; Gonzalez-Lafont, A.; Rosta, E.; Lluch, J.M. Understanding the Mechanism of the Hydro- gen Abstraction from Arachidonic Acid Catalyzed by the Human Enzyme 15-Lipoxygenase-2. A Quantum Mechanics/Molecular Mechanics Free Energy Simulation. J. Chem. Theory Comput. 2016, 12, 2079–2090. [CrossRef] 31. Saura, P.; Suardíaz, R.; Masgrau, L.; Lluch, J.M.; González-Lafont, À. Unraveling How Enzymes Can Use Bulky Residues To Drive Site-Selective C–H Activation: The Case of Mammalian Lipoxygenases Catalyzing Arachidonic Acid Oxidation. ACS Catal. 2014, 4, 4351–4363. [CrossRef] 32. Gardner, H.W. Soybean lipoxygenase-1 enzymically forms both (9S)- and (13S)-hydroperoxides from linoleic acid by a pH- dependent mechanism. Biochim. Biophys. Acta 1989, 1001, 274–281. [CrossRef] [PubMed] 33. Walther, M.; Roffeis, J.; Jansen, C.; Anton, M.; Ivanov, I.; Kuhn, H. Structural basis for pH-dependent alterations of reaction specificity of vertebrate lipoxygenase isoforms. Biochim. Biophys. Acta 2009, 1791, 827–835. [CrossRef] 34. Kutzner, L.; Goloshchapova, K.; Heydeck, D.; Stehling, S.; Kuhn, H.; Schebb, N.H. Mammalian ALOX15 orthologs exhibit pronounced dual positional specificity with docosahexaenoic acid. Biochim. Biophys. Acta 2017, 1862, 666–675. [CrossRef] [PubMed] 35. Ravula, T.; Hardin, N.Z.; Ramamoorthy, A. Polymer nanodiscs: Advantages and limitations. Chem. Phys. Lipids 2019, 219, 45–49. [CrossRef] 36. Reisch, F.; Kakularam, K.R.; Stehling, S.; Heydeck, D.; Kuhn, H. Eicosanoid biosynthesis in marine mammals. FEBS J. 2021, 288, 1387–1406. [CrossRef] 37. Kobe, M.J.; Neau, D.B.; Mitchell, C.E.; Bartlett, S.G.; Newcomer, M.E. The structure of human 15-lipoxygenase-2 with a substrate mimic. J. Biol. Chem. 2014, 289, 8562–8569. [CrossRef] 38. Jumper, J.; Evans, R.; Pritzel, A.; Green, T.; Figurnov, M.; Ronneberger, O.; Tunyasuvunakool, K.; Bates, R.; Zidek, A.; Potapenko, A.; et al. Highly accurate protein structure prediction with AlphaFold. Nature 2021, 596, 583–589. [CrossRef] Int. J. Mol. Sci. 2023, 24, 10046 25 of 25 39. Varadi, M.; Anyango, S.; Deshpande, M.; Nair, S.; Natassia, C.; Yordanova, G.; Yuan, D.; Stroe, O.; Wood, G.; Laydon, A.; et al. AlphaFold Protein Structure Database: Massively expanding the structural coverage of protein-sequence space with high-accuracy models. Nucleic Acids Res. 2022, 50, D439–D444. [CrossRef] 40. Martinez-Rosell, G.; Giorgino, T.; De Fabritiis, G. PlayMolecule ProteinPrepare: A Web Application for Protein Preparation for Molecular Dynamics Simulations. J. Chem. Inf. Model. 2017, 57, 1511–1516. [CrossRef] 41. Jones, G.; Willett, P.; Glen, R.C.; Leach, A.R.; Taylor, R. Development and validation of a genetic algorithm for flexible docking. J. Mol. Biol. 1997, 267, 727–748. [CrossRef] [PubMed] 42. Verdonk, M.L.; Cole, J.C.; Hartshorn, M.J.; Murray, C.W.; Taylor, R.D. Improved protein-ligand docking using GOLD. Proteins 2003, 52, 609–623. [CrossRef] [PubMed] 43. Li, P.F.; Merz, K.M. MCPB.py: A Python Based Metal Center Parameter Builder. J. Chem. Inf. Model. 2016, 56, 599–604. [CrossRef] [PubMed] 44. Tian, C.; Kasavajhala, K.; Belfon, K.A.A.; Raguette, L.; Huang, H.; Migues, A.N.; Bickel, J.; Wang, Y.Z.; Pincay, J.; Wu, Q.; et al. ff19SB: Amino-Acid-Specific Protein Backbone Parameters Trained against Quantum Mechanics Energy Surfaces in Solution. J. Chem. Theory Comput. 2020, 16, 528–552. [CrossRef] 45. Izadi, S.; Anandakrishnan, R.; Onufriev, A.V. Building Water Models: A Different Approach. J. Phys. Chem. Lett. 2014, 5, 3863–3871. [CrossRef] Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.