Archives of Biochemistry and Biophysics 571 (2015) 50–57

Contents lists available at ScienceDirect

Archives of Biochemistry and Biophysics journal homepage: www.elsevier.com/locate/yabbi

Mutagenesis of triad determinants of rat Alox15 alters the specificity of fatty acid and phospholipid oxygenation Mária Pekárová a,b, Hartmut Kuhn a, Ly´dia Bezáková b, Christoph Ufer a, Dagmar Heydeck a,⇑ a b

Institute of Biochemistry, University Medicine Berlin – Charité, Chariteplatz 1, D-10117 Berlin, Germany Department of Cell and Molecular Biology of Drugs, Faculty of Pharmacy, Comenius University, Kalincˇiakova 8, 832 32 Bratislava, Slovakia

a r t i c l e

i n f o

Article history: Received 12 January 2015 and in revised form 11 February 2015 Available online 27 February 2015 Keywords: Eicosanoids ALOX15 Biomembranes Lipoproteins Oxidative stress Reaction mechanism

a b s t r a c t Among lipoxygenases ALOX15 orthologs are somewhat peculiar because of their capability of oxygenating polyenoic fatty acids even if they are incorporated in complex lipid-protein assemblies. ALOX15 orthologs of different species have been characterized before, but little is known about the corresponding rat enzyme. Since rats are frequently employed as models in biomedical research we expressed rat Alox15 as recombinant protein in pro- and eukaryotic expression systems and characterized the enzyme with respect to its enzymatic properties. The enzyme oxygenated free arachidonic acid mainly to 12S-HpETE with 15S-HpETE only contributing 10% to the product mixture. Multiple directed mutagenesis studies indicated applicability of the triad concept with particular importance of Leu353 and Ile593 as specificity determinants. Ala404Gly exchange induced subtle alterations in enantioselectivity suggesting partial applicability of the Coffa/Brash concept. Wildtype rat Alox15 and its 15-lipoxygenating Leu353Phe mutant are capable of oxygenating ester lipids of biomembranes and high-density lipoproteins. For the wildtype enzyme 13S-HODE and 12S-HETE were identified as major oxygenation products but for the Leu353Phe mutant 13S-HODE and 15S-HETE prevailed. These data indicate for the first time that mutagenesis of triad determinants modifies the reaction specificity of ALOX15 orthologs with free fatty acids and complex ester lipids in a similar way. Ó 2015 Elsevier Inc. All rights reserved.

Introduction Lipoxygenases1 are lipid peroxidizing enzymes, which oxygenate polyenoic fatty acids containing at least one bisallylic methylene group to their corresponding hydroperoxy derivatives [1]. They have been implicated in the biosynthesis of bioactive lipids [2] but also play a role in the pathogenesis of hyperproliferative [3], cardio-vascular [4] and neurodegenerative [5] diseases. The LOX family is rather diverse and the human genome involves six functional LOX genes (ALOX15, ALOX15B, ALOX12, ALOX12B, ALOXE3, ALOX5) encoding for six different LOX-isoforms [6]. In mice an additional functional LOX gene (Aloxe12) is present so that this rodent species has seven Alox-isoforms [6]. Although the nomenclature of LOX is based on the reaction specificity of the enzymes with arachidonic acid there is currently no unifying concept explaining the ⇑ Corresponding author. Fax: +49 30 450 528905. E-mail address: [email protected] (D. Heydeck). Abbreviations used: LOX, lipoxygenase; ALOX15, arachidonate 15-lipoxygenase; 12S-HETE, 12S-hydro(pero)xy-5Z,8Z,10E,14Z-eicosatetraenoic acid; 15S-HETE, 15Shydro(pero)xy-5Z,8Z,11Z,13E-eicosatetraenoic acid; 13S-HODE, 13S-hydro(pero)xy9Z,11E-octadecadienoic acid; 9S-HODE, 9S-hydro(pero)xy-10E,12Z-octadecadienoic acid. 1

http://dx.doi.org/10.1016/j.abb.2015.02.029 0003-9861/Ó 2015 Elsevier Inc. All rights reserved.

mechanistic basis of this enzyme property. However, there are two supplementary hypotheses [7,8] but recent data suggested that neither of them is applicable for all LOX-isoforms [9]. According to the ‘‘triade-model’’ Phe353, Ile418/Met419 and Ile593 form the bottom of the substrate-binding pocket of ALOX15 orthologs and introduction of less space-filling residues at these positions favors arachidonic acid 12-lipoxygenation [7]. This model explains the positional specificity of all ALOX15 orthologs tested so far and may partly be applicable for ALOX12 and ALOX5 orthologs [9]. However, it fails to explain the reaction specificity of ALOX15B and ALOX12B [9]. The ‘‘A-vs-G model’’ (Coffa/Brash-concept) explains the enantioselectivity of many ALOX-isoforms [8]. This concept is based on the observation that S-LOXs contain an Ala at a critical position of their primary structure. In contrast, for R-LOXs a smaller Gly is present at this position and site directed mutagenesis studies indicated the principle possibility to interconvert S- and R-LOXs by modifying these residues [8]. However, the zebrafish LOX1 has been characterized as 12S-lipoxygenating enzyme although it contains a Gly at this position [10,11]. Thus, this enzyme does not follow the enantioselectivity concept (Coffa/Brash concept). Most LOX-isoforms strongly prefer free polyenoic fatty acids as substrates. However, the ALOX15 orthologs of rabbits [12], men

M. Pekárová et al. / Archives of Biochemistry and Biophysics 571 (2015) 50–57

51

[13] and pigs [14] are capable of oxidizing more complex substrates even if they are incorporated in complex lipid-protein assemblies, such as biomembranes and lipoproteins. This catalytic activity has been implicated in the maturational breakdown of mitochondria during red cell maturation [15] and in oxidative lipoprotein modification [16]. The ALOX15 orthologs of men [13], higher primates [17], pigs [18], cattle [19], rabbits [15] and mice [20] have been characterized but little is known on the corresponding rat ortholog. 20 years ago the cDNA of a leukocyte-type 12-LOX was cloned from rat brain and recombinant expression of the corresponding enzyme indicated arachidonic acid 12-lipoxygenation [21]. Although this enzyme shared a high degree of amino acid conservation with rabbit and human ALOX15 site directed mutagenesis studies indicated that the Sloane determinants (Ile418, Met419), which are important for the reaction specificity of the rabbit and human ALOX15, does not play a major role for the reaction specificity of the rat enzyme [21]. Moreover, a recent inhibitor study comparing the susceptibility of different rat Alox-isoforms towards standard LOX-inhibitors indicated considerable differences between rat and human ALOX15 orthologs [22]. Strong inhibitors of human ALOX15 were totally ineffective for the corresponding rat enzyme indicating functional differences between rat and human ALOX15. Since rats are frequently employed as animal models in human biomedical research and since the enzymatic properties of the rat enzyme have not been characterized we expressed rat Alox15 in two (prokaryotic, eukaryotic) different recombinant expression systems and characterized the enzyme with respect to its enzymatic properties. Our data indicate that rat Alox15 is a 12-lipoxygenating enzyme, which follows the triad concept of reaction specificity. It is capable of oxidizing membrane bound ester lipids and mutagenesis of the major triad determinant (Leu353Phe) alters the oxygenation specificity of free fatty acids and membrane bound ester lipids.

To express rat Alox15 in E. coli the corresponding cDNA was inserted into the pET28b (+) prokaryotic expression plasmid. In this plasmid the N-terminus is elongated by additional amino acids including six consecutive His so that the N-terminal protein sequence contains the following amino acid sequence before the Alox15 methionine: MGSSHHHHHHSSGLVPRGSHMASMTGGQQGR DPNSSSVD. The same peptide sequence was used for eukaryotic expression. Site-directed mutagenesis was performed using the QuickChange™ Site-Directed Mutagenesis Kit (Stratagene, Amsterdam, The Netherlands). For each mutant, 5–10 clones were selected, screened for LOX expression and one clone was completely sequenced to confirm mutagenesis.

Material and methods

Enzyme purification

Chemicals

Wild-type rat Alox15 and its mutants expressed in E. coli were purified from the bacterial lysis supernatant by affinity chromatography on a Ni–NTA (Machery-Nagel, Düren, Germany) open bed column. For this purpose a single, well separated bacterial colony grown after plating the transformation sample on a antibiotics containing agar plate was picked and a liquid culture was grown overnight at 37 °C in 1.5 L LB-medium containing 50 lg/ml kanamycin and 35 lg/ml chloramphenicol. Expression of the recombinant enzyme was induced by adding of 1 mM IPTG (final concentration) and the culture was kept under constant shaking at 30 °C for 24 h. Bacteria were collected by centrifugation, the pellet was resuspended in 10 ml of PBS and cells were lysed by sonication (W-3250 D sonifyer; Braun, Melsungen, Germany). Cell debris was removed by centrifugation, and the supernatant was incubated for 1 h at 4 °C with 0.5 ml of Ni–NTA-agarose suspension. The gel was then transferred to an open bed column and the column was washed once with 0.5 ml washing buffer containing 10 mM imidazole to remove nonspecifically bound proteins. Then the column was washed twice with 0.5 ml eluting solution containing 25 mM imidazole to remove more tightly bound proteins and finally the column was rinsed with the cleaning solution containing 200 mM imidazole (6 times 0.5 ml). Applying this chromatographic strategy one washing fraction (1  0.5 ml 10 mM imidazole) and 8 elution fractions (2  0.5 ml 25 mM imidazole, 6  0.5 ml 200 mM imidazole) were obtained and the majority of the recombinant protein was detected in the elution fractions 1 and 2.

The chemicals used were obtained from the following sources: arachidonic acid and linoleic acid from Sigma (Taufkirchen, Germany); HPLC standards of 5(±)-HETE, 12S-HETE, 12(±)-HETE, 15S-HETE, 15(±)-HETE, 13S-HODE and 13(±)-HODE from Cayman Chem. (distributed by Biomol, Hamburg, Germany); sodium borohydride from Life Technologies, Inc. (Eggenstein, Germany); HPLC solvents from Baker (Deventer, The Netherlands); antibiotics and isopropyl-ß-thiogalactopyranoside (IPTG) from Carl Roth GmbH (Karlsruhe, Germany); restriction enzymes from Thermo Fisher Scientific-Fermentas (Schwerte, Germany); Escherichia coli strain BL21 (DE3)pLysS from Invitrogen (Carlsbad, USA); E. coli strain XL-1 blue from Stratagene (La Jolla, USA); N2a mouse neuroblastoma cells from LGC Standards GmbH (Wesel, Germany). Oligonucleotide synthesis was performed at BioTez (Berlin, Germany). Nucleic acid sequencing was carried out at Eurofins MWG Operon (Ebersberg, Germany). PCR cloning of rat Alox15 cDNA Total RNA was isolated from rat blood using the blood RNA extraction kit from Qiagen (Hilden, Germany). The RNA was reversely transcribed using Premium Reverse Transcriptase from Thermo Fisher Scientific-Fermentas and the resulting cDNA mixture was employed as template for PCR cloning of rat

Alox15 using the following primer combination: forward 50 – CCGTCGACATGGGTGTCTACCGCATC – 30 and reverse 50 – GCGGCCGCTCATATGGCCACGCTGTT – 30 . These primers flank the rat Alox15 coding region. The coding region was amplified using Advantage PCR Mix (Takara Bio Europe/Clontech, Saint-Germainen-Laye, France) and the resulting 2100 bp PCR fragment was ligated into the TOPO pCR 2.1 cloning vector (Life Technologies GmbH, Darmstadt, Germany). Competent E. coli XL-1 Blue cells (Life Technologies, Darmstadt, Germany) were transformed with the recombinant plasmid, plated on an antibiotics containing agar gel, isolated insert-containing clones were selected and plasmid DNA was prepared. The rat Alox15 coding region was excised and subcloned either into the pET 28b (+) bacterial expression vector (Merck Millipore, Darmstadt, Germany) using the unique Sal I and Not I restriction sites or alternatively into the pcDNA 3.1 eukaryotic expression vector (Life Technologies GmbH, Darmstadt, Germany) using the Xba I and Not I restriction sites. These subcloning strategies lead to a His-tag fusion protein for the bacterial as well as for the eukaryotic expression strategy. Bacterial expression and site-directed mutagenesis

52

M. Pekárová et al. / Archives of Biochemistry and Biophysics 571 (2015) 50–57

Eukaryotic expression To express rat Alox15 in eukaryotic cells murine neuroblastoma cells Neuro-2a (N2a) were employed. For transfection N2a cells were plated into 6-well plates at a density of about 2  105 cells per well in DMEM medium (4.5 g/L glucose, L-glutamine, 1 mM sodium pyruvate, 3.7 g/L NaHCO3, 10% fetal calf serum) 24 h prior to transfection. Cell transfection was performed using the TransIT-LT1 transfection kit (Mirus bio, Madison, USA). For this purpose 2 lg of prepared plasmid DNA were mixed with 6 ll TransITLT1 in 200 ll OptiMEM and transfection complexes were allowed to form according to the manufacturer’s protocols. Then the transfection mixture was added to each well and the cells were incubated with the transfection mixture for 48 h. Then the cells were washed with PBS, harvested by scraping, centrifuged (350 rpm for 3 min) and the pellet was resuspended in 0.1 ml PBS. Next, cells were disrupted by sonication (Dr. Hielscher instrument, 2 times for 10 s at 100% amplitude), debris was pelleted by centrifugation (16,000 rpm, 20 min, 4 °C) and the lysis supernatant was used for activity assays, protein quantification and/or SDS–PAGE. HPLC activity assays For activity assays aliquots of the eukaryotic lysis supernatant or of the Ni–NTA-agarose affinity chromatography elution fractions were incubated for 15 min at 37 °C with 0.1 mM of arachidonic acid in 0.5 ml of PBS. The hydroperoxy compounds formed were reduced with sodium borohydride and after acidification 0.5 ml of ice-cold methanol were added. The protein precipitate was spun down and aliquots of the clear supernatant were injected directly for quantification of the LOX products to RP-HPLC. For purified enzyme preparations LOX activity was assayed spectrophotometrically recording the time-dependent increase in absorbance at 235 nm (Shimadzu UV-2100 photometer).

20 min under argon atmosphere. Then the sample was cooled down, acidified with 0.15 ml of acetic acid and aliquots of this mixture were directly injected into RP-HPLC for quantification of conjugated diene formation. For reliable analysis of the reaction products we reduced the hydroperoxy fatty acids (HpETE, HpODE isomers) formed by the enzyme to the corresponding hydroxy compounds (HETE, HODE) by adding sodium borohydride to the reaction mixture after the reaction period. Thus, we analyzed the HETE-isomers although the enzyme forms hydroperoxides. Analytics HPLC analysis of the LOX products was performed on a Shimadzu instrument equipped with a Hewlett–Packard diode array detector 1040 A by recording the absorbance at 235 nm. Reverse phase-HPLC was carried out on a Nucleodur C18 Gravity column (Machery-Nagel, Düren, Germany; 250  4 mm, 5 lm particle size) coupled with a guard column (8  4 mm, 5-lm particle size). A solvent system of methanol/water/acetic acid (85/15/0.1, by volume) was used at a flow rate of 1 ml/min. Straight phaseHPLC (SP-HPLC) was performed on a Nucleosil 100-5 column (250  4 mm, 5 lm particle size) with the solvent system n-hexane/2-propanol/acetic acid (100/2/0.1, by volume) and a flow rate of 1 ml/min. Hydroxy fatty acid enantiomers were separated by chiral phase HPLC (CP-HPLC). 15-HETE enantiomers were separated as free fatty acids on a Chiralcel OD column (Daicel Chem. Ind., Ltd.) using a solvent system consisting of hexane/2propanol/acetic acid (100/5/0.1, by volume) and a flow rate of 1 ml/min. 12-HETE enantiomers were separated as methyl esters on a Chiralcel OB column (Daicel Chem. Ind., Ltd.) with a solvent system consisting of n-hexane/2-propanol/acetic acid (100/2/0.1, by volume) and a flow rate of 1 ml/min. Results

Ester lipid preparations and alkaline hydrolysis

Recombinant expression of rat Alox15

To test the oxygenase activity of wildtype and mutant rat Alox15 with complex lipid protein assemblies the enzyme preparations were incubated with vesicles of inner mitochondrial membranes [23], human low density lipoprotein and human high density lipoprotein [24]. The volume of the assay system was 1 ml. The amount of the lipid-protein-assemblies added to the assay mixture was calculated so that the concentration of the two major polyenoic fatty acids (arachidonic acid, linoleic acid) present in the assay mixture was adjusted to about 0.165 mM. To quantify the content of endogenous polyenoic fatty acids in biomembranes and lipoproteins we extracted the total lipids from the untreated preparations and hydrolyzed them under alkaline conditions (20 min at 60 °C under argon atmosphere). Hydrolysis was stopped by acidification to pH 4 (acetic acid) and aliquots of the acidified hydrolysis mixture were directly injected to RP-HPLC for quantification of the polyenoic fatty acids using the solvent system methanol/water/acetic acid (85:15:0.1, by volume). To detect the polyenoic fatty acids the absorbance at 205 nm was recorded. Under these conditions arachidonic acid and linoleic acid were eluted at 25.3 min and 28.1 min, respectively. The chromatographic scale was calibrated by injecting known amounts of the two fatty acids (5 point calibration curves for each fatty acid). For activity assays the substrates were incubated with the enzyme preparations for 5 min at 25 °C and the reaction was stopped by the addition of sodium borohydride. After acidification the total lipids were extracted from the reaction mixtures according to [25]. After evaporation of the solvents the remaining ester lipids were reconstituted in 0.85 ml of methanol and 0.15 ml of 40% KOH was added. The lipids were hydrolyzed at 60 °C for

The bacterial lysate supernatants (20,000g) were run over a Ni-AT agarose column and aliquots of elution fraction 1 were used for in vitro activity assays. As indicated in Fig. 1A (left inset) recombinant rat Alox15 is expressed in E. coli only at low levels (no prominent protein band in the lysis supernatant) but can strongly be enriched by affinity chromatography on a Ni-matrix. As indicated, the majority of the recombinant protein was eluted from the affinity columns in fractions 1 and 2. An aliquot of elution fraction 1 was subsequently incubated with exogenous arachidonic acid (0.1 mM) for 15 min and the reaction products were analyzed by RP-HPLC. The majority of the conjugated dienes co-chromatographed with an authentic standard of 12-HETE (Fig. 1A). Much smaller amounts (11%) of 15-HETE were analyzed. Chiral phase HPLC of the separated HETE-isomers indicated that both 12-HETE (right inset Fig. 1A) and 15-HETE (right inset to Fig. 1B) were mainly the S-enantiomers. Taken together these data confirm the previous findings that rat Alox15 as the mouse [20], pig [18], cattle [19] and macaca [17] orthologs is an arachidonic acid 12S-lipoxygenating ALOX15-isoform. To explore the possible impact of the expression system on the catalytic properties of recombinant rat Alox15 we subcloned the cDNA into a mammalian expression vector (pcDNA3.1) and expressed the enzyme in cultured N2a mouse neuroblastoma cells. After harvesting the cell lysates were incubated in PBS with exogenous arachidonic acid (0.1 mM) and the reaction products were analyzed by RP-HPLC. As indicated in Fig. 1B the enzyme also converts arachidonic acid mainly to 12-HETE with 15-HETE being a minor reaction product. These products were virtually absent when lysate supernatants of mock-transfected (empty transfection

M. Pekárová et al. / Archives of Biochemistry and Biophysics 571 (2015) 50–57

53

determinant (Val593Ile) led to complete inversion of the reaction specificity of rat Alox15. For rabbit ALOX15 this determinant does not play such prominent role [26]. In summary, one may conclude that Leu353 and Val593 are the most important triad determinants for rat Alox15. To explore the functional interplay of the two relevant triad determinants (Leu353, Val593) of rat Alox15, which are far distant in the primary structure, we applied a sequential mutagenesis strategy. As shown above, wildtype rat Alox15 exhibits a 12lipoxygenating activity (Fig. 3A) and when we introduced a more space filling Phe at Leu353 arachidonic 15-lipoxygenation was forced (Fig. 3B). Next, we introduced a less space-filling Gly at Val593 of the Leu353Phe mutant and expected partial reversion of the effects induced by the Leu353Phe exchange. Indeed, as indicated in Fig. 3C the Leu353Phe + Val593Gly mutant exhibited a dual positional specificity. Although 15-HETE was still the major oxygenation product (64%) the share of 12-HETE was strongly increased when compared with the Leu353Phe mutant. These data indicates that the alterations in the positional specificity that were introduced by the Leu353Phe mutation were partly reversed by Val593Gly exchange, suggesting a functional interplay of Leu353 and Val593. Fig. 1. Expression of rat Alox15 in bacterial and eukaryotic expression systems and reaction specificity of arachidonic acid oxygenation. Rat Alox15 was expressed as His-tag fusion protein in E. coli and N2a cells. The reaction specificity of the recombinant enzymes was determined analyzing the reaction products formed during a 5 min incubation period with 100 lM arachidonic acid in PBS. (A) RP-HPLC of reaction products formed by enriched (elution fraction 1 of affinity chromatography) rat Alox15 (bacterial expression) during 15 min incubation with 100 lM arachidonic acid. Left inset: SDS–PAGE of aliquots of different stages of enzyme preparation. 10 ll of protein preparations were applied. Right inset: chiral phase HPLC of the 12-HETE fraction collected during RP-HPLC of the reaction products formed by wildtype rat Alox15 expressed in E. coli. (B) RP-HPLC of reaction products formed by rat Alox15 (lysis supernatant of transfected N2a cells) during 15 min incubation with 100 lM arachidonic acid. Inset: chiral phase HPLC of the 15-HETE fraction collected during RP-HPLC of the reaction products formed by wildtype rat Alox15 expressed in E. coli.

plasmid) cells were employed as enzyme source (data not shown). Here again, chiral HPLC indicated the 12S-enantiomer as major reaction product.

Expression of rat Alox15 and its triad mutants in eukaryotic cells To test the impact of the expression system on the relative importance of the triad determinants we expressed wildtype rat Alox15, its Leu353Ile, Ala418Ile and Val593Ile mutants in N2a cells. From Table 2 it can be seen that the wildtype enzyme and its Ala418Ile mutant exhibit major 12-lipoxygenating activity. In contrast, Leu353Phe exchange induced 15-lipoxygenation, which is consistent with the data obtained in the bacterial expression system. For the Val593Ile mutant we also observed an increase in the share of 15-HETE (from 8% in wildtype to 30% in the mutant) but this alteration was incomplete since the major reaction product was still 12-HETE. Thus, when expressed in eukaryotic cells Val593 of rat Alox15 may not play a major role for the positional specificity of the enzyme. Mutagenesis of the enantioselectivity determinant

Mutagenesis of triad determinants Since mammalian ALOX15 orthologs follow the triad concept [7] we explored whether mutagenesis of the triad determinants alters the reaction specificity of rat Alox15. For this purpose we first identified the triad determinants by multiple sequence alignments (Fig. 2). We found that the Borngräber 1 determinant was occupied by a Leu, which is less bulky than the Phe found at this position in 15-lipoxygenating ALOX15 orthologs (humans, orangutans). At the position of the major Sloane determinant a small Ala was found, whereas a more space-filling Ile was present in the 15lipoxygenating enzymes. The Borngräber 2 determinant was occupied by a small Ala. Taken together these data suggest that the volume of the substrate-binding pocket of rat Alox15 should be bigger than that of the human ortholog, which is consistent with the 12lipoxygenating capacity of the rat enzyme. To test the applicability of the triad concept, we first mutated the triad determinants to the corresponding residues present at these positions in human ALOX15. As indicated in Table 1 Leu353Phe exchange completely inverted the reaction specificity since the major (96%) arachidonic acid oxygenation product was 15-HETE. Mutagenesis of the Sloane determinants (Ala418Ile) did hardly impact the product pattern (Table 1). These data are consistent with previous mutagenesis results obtained for rat Alox15 [21] suggesting that the Sloane determinants may not be of major importance for the specificity of this enzyme. Surprisingly, mutation of the Borngräber 2

S-LOX contain a highly conserved Ala residue, which is a Gly in R-LOXs [8,27]. To test the applicability of this concept (Coffa/Brash) for rat Alox15 we identified the corresponding Ala by multiple amino acid alignment (Fig. 2) and mutated this residue to Gly. We found that Ala406Gly exchange impacted the product pattern (Table 3). The share of 8-HETE went up from 2% to 23% on the expense of 12-HETE formation, which declined from 83% to 62%. However, 12HETE remained the major oxygenation product. Next, we carried out chiral phase HPLC on the three major reaction products [15-HETE, 12-HETE, 8-HETE]. For 15-HETE the S-enantiomer prevailed and for 12-HETE no R-enantiomer was present for both wildtype rat Alox15 and its Ala406Gly mutant. In contrast, the 8-HETE formed by the Ala406Gly was exclusively the R-isomer. This switch in enantiomer specificity could be predicted assuming the applicability of the Coffa/Brash concept [8,27]. Unfortunately, the observed alterations were incomplete. Similar partial changes have previously been reported for 15-lipoxygenating ALOX15 orthologs from other species [17] (see Table 4). Substrate specificity of rat Alox15 and its 15-lipoxygenating Leu353Phe mutant In most mammalian cells linoleic acid and arachidonic acid are the dominating polyenoic fatty acids. In theory arachidonic acid 12-lipoxygenating LOXs should not be capable of effectively

54

M. Pekárová et al. / Archives of Biochemistry and Biophysics 571 (2015) 50–57

Fig. 2. Multiple amino acid sequence alignment of mammalian ALOX15 orthologs. The regions of the primary structure containing the three sequence determinants for the positional specificity, which have previously been identified for the rabbit, pongo and mouse ALOX15 isoforms are aligned. 15-lipoxygenating isoforms (pongo, human) carry more space filling residues at these positions, whereas the 12-lipoxygenating isoforms (mouse, rat) contain smaller residues.

Table 1 Positional specificity of recombinant wildtype and mutant ratAlox15 expressed in E. coli. Rat Alox15 and the corresponding mutants were expressed as recombinant His-tag fusion proteins in E. coli, purified by affinity chromatography on Ni–NTAagarose and the positional specificity of arachidonic acid oxygenation was analyzed by RP-HPLC. Three independent activity assays were performed for each enzyme variant and the means are listed in the table. The relative specific activity was estimated relating the intensity of the immunoreactive band at 75 kDa (immunoblot) to the amount of conjugated dienes (HPLC) formed during a 15 min incubation period. The specific activity of the wild-type enzyme was set 100%. Means ± standard deviations (SD) are given. For quantification we only considered the two major reaction products (15-HETE, 12-HETE), which were separated by RP-HPLC. Additional SP-HPLC analysis of the material in the 12-HETE/8-HETE peak of wildtype rat Alox15 and its A418I mutant indicated that the share of 8-HETE did not exceed 3% and thus we did not quantify this product separately. LOX variant

Specific activity (%)

Share of 15-HETE (%)

Share of 12-HETE (%)

Wild-type L353F A418I V593I

100.0 ± 8.1 15.1 ± 2.0 40.8 ± 3.9 24.3 ± 5.9

11.9 ± 1.1 96.5 ± 1.4 3.7 ± 0.1 91.0 ± 1.1

88.1 ± 1.1 3.5 ± 1.4 96.3 ± 0.1 9.0 ± 1.1

oxygenating linoleic acid because of the lacking n  11 bisallylic methylene in this fatty acid. When we incubated wildtype rat Alox15 in an activity assay containing both, arachidonic acid and linoleic acid at identical concentrations, arachidonic acid is preferentially oxygenated as indicated by HPLC quantification of the reaction products (Table 5). In contrast, using the 15-lipoxygenating Leu353Phe mutant both fatty acids are oxygenated with comparable rates. For the Leu353Phe + Val593Gly double mutant linoleic acid turned out to be the preferred substrate. Taken together, these data indicate that despite of subtle quantitative differences among the different enzyme variants 12-lipoxygenating wildtype and 15-lipoxygenating mutant variants of rat Alox15

oxygenate both, linoleic acid and arachidonic acid, when offered simultaneously in a single assay mixture. Rat Alox15 oxygenates biomembranes and HDL Rabbit [12], human [13] and pig [14] ALOX15 orthologs are capable of oxidizing complex lipid-protein assemblies such as biomembranes and lipoproteins and 13-HODE as well as 15- and/ or 12-HETE have been identified as major oxygenation products [13,14,28]. Since the rat Alox15 oxygenates both, arachidonic acid and linoleic acid, we next tested the reactivity of this enzyme with biomembranes and lipoproteins. For this purpose we first incubated wildtype rat Alox15 (affinity chromatography purified bacterial lysis supernatant) with vesicles of beef heart inner mitochondrial membranes and analyzed by combined RP/SPHPLC the hydrolyzed lipid extracts for the occurrence of conjugated dienes. As indicated in Fig. 4A the formation of conjugated dienes, co-migrating in RP-HPLC with authentic standards of 13HODE and 12-HETE, were observed. These products were not detected in the membrane vesicles incubated in the absence of the LOX preparation (Fig. 4B). Consecutive SP- (inset to Fig. 4A) and CP-HPLC (data not shown) of the RP-HPLC purified conjugated dienes indicated 13S-HODE and 12-HETE as major oxygenation products, but we also observed the formation of 9-HODE and of the all-E HODE isomers and the UV-spectra of the E,E isomers recorded during the run confirm the structure of these products (inset to Fig. 4B). A similar product pattern has previously been described when the rabbit ALOX15 oxygenated mitochondrial membranes [28]. Next we compared the product patterns formed by wildtype rat Alox15 and its 15-lipoxygenating Leu353Phe mutant during the interaction with biomembranes. As expected, 12S-HETE was a major oxygenation product of the wildtype enzyme (Fig. 4A inset and Table 6) but 15-HETE was only formed in trace amounts. In contrast, 15S-HETE was among the dominant

55

M. Pekárová et al. / Archives of Biochemistry and Biophysics 571 (2015) 50–57

substrates. In fact, although rabbit ALOX15 oxygenates LDL and HDL these lipoproteins were no effective substrates for rat Alox15. Discussion

Fig. 3. Alterations of rat Alox15 specificity by sequential mutagenesis of triad determinants. Wildtype and mutant rat Alox15 variants were incubated with 100 lM arachidonic acid in PBS and reaction specificity of the enzyme variant was quantified by RP-HPLC analysis of the reaction products. (A) Wildtype enzyme, (B) Leu353Phe mutant, (C) Leu353Phe + Val593Gly double mutant.

products of the 15-lipoxygenating Leu353Phe mutant (Table 5). These data indicate for the first time that mutagenesis of the triad determinants, which alters the reaction specificity of free fatty acid oxygenation, does also modify the reaction specificity with membrane-bound ester lipids. To compare the extent of membrane oxygenase activity of rat Alox15 with that of the well-characterized rabbit enzyme, we performed identical experiments with the rabbit ortholog. As shown in Table 6 rat and rabbit ALOX15 exhibit similar membrane oxygenase activities. This was, however, not the case for other complex Table 2 Positional specificity of recombinant wildtype and mutant Alox15 expressed in eukaryotic cells. Rat Alox15 and the corresponding mutants were expressed in N2a mouse neuroblastoma cells and the positional specificity of arachidonic acid oxygenation was analyzed by RP-HPLC. Five independent activity assays were performed for each enzyme variant and the means are listed in the table. The relative specific activity was estimated relating the intensity of the immunoreactive band at 75 kDa (immunoblot) to the amount of conjugated dienes (HPLC) formed during a 15 min incubation period. The specific activity of the wild-type enzyme was set 100%. Means ± standard deviations (SD) are given. As in Fig. 1 we only considered the two major reaction products (15-HETE, 12-HETE) for quantification, which were separated by RP-HPLC. Additional SP-HPLC analysis of the material in the 12-HETE/8-HETE peak of wildtype rat ALOX15 and its A418I mutant indicated that the share of 8-HETE did not exceed 3% and thus we did not quantify this product separately. Enzyme

Specific activity (%)

Share of 15-HETE (%)

Share of 12-HETE (%)

Wildtype L353F A418I V593I

100 ± 6.8 85 ± 6.7 114 ± 8.2 194 ± 29.2

8 ± 2.2 77 ± 11 5 ± 2.3 30 ± 6.5

91 ± 5.8 23 ± 1.8 95 ± 6.7 70 ± 21.8

ALOX15 orthologs occur in most mammals and they can be classified according to their positional specificity of arachidonic acid oxygenation in 12- and 15-lipoxygenating enzymes [2]. In many mammals (mice, pigs, cattle, macaca) 12-lipoxygenating enzymes have been identified but in higher primates such as orangutan [17], chimpanzee [9], bonobo [9] and men [13] the corresponding orthologs are 15-lipoxygenating. The only exceptions from this rule are rabbits, which express both 12- and 15lipoxygenating enzyme variants in a tissue specific manner [9]. In immature red cells the 15-lipoxygenating enzyme variant prevails, whereas in monocytes a 12-lipoxygenating enzyme is expressed [29]. Interestingly, there is only a single ALOX15 gene in the rabbit genome. The database sequences suggest that this gene encodes for a 15-lipoxygenating enzyme since bulky residues (Phe353, Ile418, Ile593) occupy the positions of the triad determinants. However, from peripheral rabbit monocytes a second ALOX15 cDNA was isolated and the predicted amino acid sequence of this enzyme shared a 99% identity with the 15-lipoxygenating enzyme species. Among the six amino acid residues different in both enzymes a Phe-Leu exchange was detected at position 353 and when the corresponding protein was expressed in E. coli arachidonic acid 12-lipoxygenation was observed. For the time being the mechanistic details of the recoding process, which interconverts mRNAs of 12- and 15-lipoxygenating LOXs have not been explored and the biological relevance of this interconversion remains elusive. It has previously been reported [21] that rat Alox15 belongs to the 12-lipoxygenating ALOX15 orthologs and our data confirm this conclusion. However, so far it remained unclear whether or not this enzyme follows the triad concept since mutagenesis studies of the Sloane determinants did not alter the positional specificity of the enzyme [21]. For this paper we systematically explored this question by site-directed mutagenesis and found that the triad concept is applicable for rat Alox15. However, in contrast to the human enzyme, for which Ile418 and Met419 play a major role, Leu353 is the most important determinant for the reaction specificity. In this respect the rat enzyme resembles the mouse ortholog [30]. Rabbit, human and porcine ALOX15 orthologs are capable of oxygenating ester lipid substrates, even if they are incorporated in complex lipid protein assemblies, such as biomembranes and lipoproteins [12–14]. Here we show that rat Alox15 is also capable of oxidizing biomembranes. When normalized to identical arachidonic acid oxygenase activity, the rat enzyme exhibits about 60% of

Table 3 Mutagenesis of the enantioselectivity determinants of rat Alox15. Rat Alox15 and the Ala406G mutant were expressed as recombinant His-tag fusion proteins in E. coli and purified by affinity chromatography on Ni–NTA-agarose. Aliquots of these enzyme preparations were used and 3–5 independent activity assays were carried out. The arachidonic acid oxygenation products were separated by RP-HPLC (see Fig. 3), the hydroxy fatty acid fractions of all activity assays were pooled and 15-HETE, 12-HETE and 8-HETE were prepared by SP-HPLC. Enantiomer composition of 15-HETE was directly analyzed CP-HPLC (see Materials and methods). 12-HETE and 8-HETE were first methylated, repurified by RP-HPLC and subsequently analyzed by CP-HPLC for enantiomer separation. The relative shares of S- and R-enantiomers for each positional isomer are given in %. Enzyme species

Wildtype A404G

15-HETE

12-HETE

8-HETE

S (%)

R (%)

S (%)

R (%)

S (%)

R (%)

88.5 82.1

11.5 17.9

96.3 99.4

3.7 0.6

86.7 13.2

13.3 86.8

56

M. Pekárová et al. / Archives of Biochemistry and Biophysics 571 (2015) 50–57

Table 4 Quantification of arachidonic acid and linoleic acid oxygenase activity of wildtype and mutant rat Alox15. Aliquots of the enriched (Ni agarose elution fractions) rat Alox15 variants were incubated for 5 min in PBS containing 100 lM linoleic acid and 100 lM arachidonic acid (three different activity assays for each enzyme variant). The hydroxy fatty acids formed during the incubation period in each activity assay were separately prepared by RP-HPLC but then pooled together. Since linoleic and arachidonic acid oxygenation products are not well separated by RP-HPLC the combined hydroxy fatty acids were subsequently analyzed by SP-HPLC to separate the oxygenation products formed from the two substrate fatty acids. The sums of the conjugated dienes observed in SP-HPLC (12-HETE + 15-HETE as arachidonic acid oxygenation products; 9-HODE + 13-HODE as linoleic acid oxygenation products) were set 100%. Enzyme variant

Arachidonic acid oxygenation (%)

Linoleic acid oxygenation (%)

12-HETE/15HETE ratio (%)

Wildtype Leu353Phe Leu353Phe + V593Gly

66 48 35

34 53 65

86:14 10:90 39:61

Table 5 Composition of the reaction products formed by wildtype rat Alox15 and its 15lipoxygenating Leu353Phe mutant during oxygenation of biomembranes. Enriched wildtype rat Alox15 (13.3 lkat arachidonic acid oxygenase activity/ml) and its 15lipoxygenating Leu353Phe mutant were incubated for 5 min in PBS with vesicles of inner beef heart mitochondrial membranes (5.3 mg/ml protein in the assay sample). Three separate activity assays were run. After the incubation period the reaction products were reduced with borohydride, the lipids were extracted and the extracts were hydrolyzed under alkaline conditions. The hydroxy fatty acids in each activity assay were separately prepared by RP-HPLC and then pooled together for each enzyme variant. The pooled products were further analyzed by SP-HPLC to quantify the hydroxy fatty acids given. Enzyme variant

Wildtype enzyme Leu353Phe mutant

Fig. 4. Oxygenation of inner mitochondrial membranes by wildtype rat Alox15. Enriched wildtype rat Alox15 (13.3 nkat arachidonic acid oxygenase activity/ml) was incubated for 5 min in PBS with vesicles of inner beef heart mitochondrial membranes (5.3 mg/ml protein in the assay sample). After the incubation period the reaction products were reduced with borohydride, the lipids were extracted and the extracts were hydrolyzed under alkaline conditions. Aliquots of the acidified hydrolysis mixture were injected to RP-HPLC. The conjugated dienes eluted in RPHPLC between 6 and 9 min were collected and further analyzed by SP-HPLC. (A) Hydrolyzed lipid extracts of membrane preparations incubated for 5 min with wildtype rat Alox15. Inset: SP-HPLC (solvent system: hexane/isopropanol/acetic acid; 100:2:0.1, by volume) of the conjugated dienes purified by RP-HPLC (fractions eluting in RP-HPLC between 6 and 9 min). The asterix (⁄) indicates retention time of 15-HETE. (B) Lipid extracts of membrane preparations incubated for 5 min without LOX (controls). Inset: UV-spectra of the conjugated dienes recorded during SP-HPLC of the LOX-treated membranes.

Relative share of major oxygenation products (%) 9HODE (E,Z)

9HODE (E,E)

12-HETE (Z,Z,E,Z)

15-HETE (Z,Z,Z,E)

13HODE (Z,E)

13HODE (E,E)

36