Characterization of a novel 8R,11S-linoleate diol synthase from Penicillium chrysogenum by identification of its enzymatic products

To identify novel fatty acid diol synthases, putative candidate sequences from Penicillium species were analyzed, and hydroxy fatty acid production by crude Penicillium enzyme extracts was assessed. Penicillium chrysogenum was found to produce an unknown dihydroxy fatty acid, a candidate gene implicated in this production was cloned and expressed, and the expressed enzyme was purified. The product obtained by the reaction of the purified enzyme with linoleic acid was identified as 8R,11S-dihydroxy-9,12(Z,Z)-octadecadienoic acid (8R,11S-DiHODE). The catalytic efficiency of this enzyme toward linoleic acid was the highest among the unsaturated fatty acids tested, indicating that this enzyme was a novel 8R,11S-linoleate diol synthase (8R,11S-LDS). A sexual stage in the life cycle of P. chrysogenum has recently been discovered, and 8R,11S-DiHODE produced by 8R,11S-LDS may constitute a precocious sexual inducer factor, responsible for regulating the sexual and asexual cycles of this fungus.

acid to induce protein expression, with shaking at 150 rpm for 5 days. Mycelia were harvested by fi ltration, washed three times with saline solution, and homogenized on ice for 10 min in 50 mM HEPES buffer (pH 7.5). Mycelial debris and unbroken mycelia were removed by centrifugation at 13,000 g for 30 min at 4°C, and the supernatant was fi ltered through a 0.45 m fi lter. The fi ltrates were then used as crude Penicillium enzyme extracts.
Cloning and site-directed mutagenesis P. chrysogenum and E. coli ER2566 were used as the sources of DNA template for fatty acid diol synthase gene and host cells, respectively. pET-21a(+) plasmid, which contained nucleotides encoding six histidine residues at C-terminal position, was used as an expression vector. Total RNA was isolated from the mycelia of P. chrysogenum using a Hybrid-R total RNA purifi cation kit (GeneAll). Full-length cDNA was synthesized from total RNA by reverse transcription PCR using a reverse transcription kit (TaKaRa). The gene encoding fatty acid diol synthase and the full and partial genes of N-terminal domain or C-terminal domain were cloned using the Gibson assembly method ( 21 ). The sequences of the primers used for gene cloning were based on the DNA sequence of P. chrysogenum fatty acid diol synthase (Gen-Bank accession number CAP97986). Primers were designed to amplify the expression vector and fatty acid diol synthase DNA fragments in supplementary Table 1. The amplifi ed DNA fragments and linearized vector, which were obtained by PCR with Phusion High-Fidelity DNA Polymerase (New England Biolabs), were ligated using Gibson Assembly Master Mix (New England Biolabs). E. coli strain ER2566 was transformed with the ligation mixture and plated on Luria-Bertani (LB) agar containing 20 g/ml ampicillin. An ampicillin-resistant colony was selected, and plasmid DNA from the transformant was isolated using a plasmid purifi cation kit (Intron). Site-directed mutagenesis was performed using a QuikChange kit, according to the manufacturer's protocol (Stratagene).

Purifi ed enzyme preparation
Recombinant E. coli cells were cultivated in a 2 liter fl ask containing 500 ml of LB medium and 20 g/ml ampicillin at 37°C with shaking at 200 rpm. When the optical density at 600 nm of the bacterial culture reached 0.6, isopropyl-␤ -D -thiogalactopyranoside was added to a fi nal concentration of 0.1 mM to induce enzyme expression. The culture was then incubated at 16°C with shaking at 150 rpm for 16 h to express the enzyme. Cells from the culture broth were harvested by centrifugation at 10,000 g for 30 min at 4°C and washed three times with 0.85% saline solution. The washed recombinant cells were subsequently resuspended in 50 mM phosphate buffer (pH 7.0) containing 300 mM NaCl and 1 mg/ml lysozyme. The resuspended cells were disrupted by sonication on ice for 10 min. The unbroken cells and cell debris were removed by centrifugation at 13,000 g for 20 min at 4°C, and the supernatant was passed through a 0.45 m fi lter. The fi ltrate was applied to a HisTrap HP affi nity chromatography column (GE Healthcare) equilibrated with 50 mM phosphate buffer (pH 7.0) on a fast-protein liquid chromatography (Bio-Rad) in a cold room at 4°C. The bound protein was eluted at 4°C using the same buffer containing 250 mM imidazole at a fl ow rate of 1 ml/min. The active fractions were collected and dialyzed against 50 mM 3-[4-(2-hydroxyethyl)-1-piperazinyl]-propane sulfonic acid (EPPS) buffer (pH 8.0) at 4°C for 16 h. After dialysis, the resulting solution was used as the purifi ed enzyme.
In the present study, we isolated and characterized 8 R ,11 S -linoleate diol synthase (8 R ,11 S -LDS) from P. chrysogenum for the fi rst time by assessing the hydroxy fatty acids produced by several Penicillium species, cloning the gene involved in the production of an unknown dihydroxy fatty acid according to sequence analysis of putative fatty acid diol synthase genes, identifying hydroxy fatty acids produced from unsaturated fatty acids by the expressed enzyme, and determining kinetic parameters for unsaturated fatty acids.
LC/MS/MS analysis of oxylipins was performed using a Thermo-Finnigan LCQ Deca XP Plus ion trap mass spectrometer (Thermo Scientifi c). The instrument consisted of an LC pump, an autosampler, and a photodiode array detector. Ionization of the samples was carried out using ESI. The operation parameters were as follows: capillary temperature, 275°C; ion source voltage, 5 kV; nebulizer gas, 206.84 kPa; capillary voltage, 46 V in positive mode and 15 V in negative ionization mode; average scan time, 0.6 s; average time to change polarity, 1.2 s; and collision energy, ‫ف‬ 35% abundance of the precursor ion.

Assessment of the hydroxy fatty acids produced from linoleic acid by crude Penicillium enzyme extracts
Crude enzyme extracts from P. chrysogenum , P. digitatum , P. marneffei , and P. oxalicum containing putative diol synthases were used for the conversion of linoleic acid to hydroxy fatty acids. The reaction products were analyzed by HPLC using a reversed phase Nucleosil C 18 column and reference standards, including 5 S ,8 R -DiHODE, 7 S ,8 S -DiHODE, 8 R ,11 S -DiHODE, 8 R -HODE, and 10 R -HODE. Crude enzyme extracts from P. digitatum, P. oxalicum , and P. marneffei converted linoleic acid to products, with the same retention times as those of 8 R -HODE, 7 S ,8 S -DiHODE, and 10 R -HODE, respectively ( Fig. 1 ). However, the retention time (4.3 min) of the compound produced by crude P. chrysogenum enzyme extract was the same as that of 8 R ,11 S -DiHODE, suggesting that the compound was 8,11-DiHODE.

Amino acid sequence analysis of putative diol synthases from Penicillium species
The production of monohydroxy and dihydroxy linoleic acids from linoleic acid by Penicillium species resulted from the reactions of the corresponding putative fatty acid diol synthases in these species. To analyze such enzymes, amino acid sequences of putative fatty acid diol synthases from Penicillium species, including PC_1 and PC_2 from P. chrysogenum , PD_1 and PD_2 from P. digitatum , PM from P. marneffei , and PO_1 and PO_2 from P. oxalicum , were aligned with that of PpoA from A. nidulans ( Fig. 2 ). The YRWH motif of the N-terminal domain containing the catalytic residue Tyr 374 of A. nidulans PpoA ( 2 ) and the Glu and Arg residues in EXXR motif of the C-terminal domain were found to be conserved across all of the Penicillium candidate proteins. However, the two crucial residues for the hydroperoxide isomerase activity in the C-terminal heme signature motif, His 1004 and Cys 1006 of A. nidulans unsaturated fatty acid and 50 g/ml purifi ed enzyme or 5 mg/ml crude enzyme extract. The effect of pH and temperature on the activity of P. chrysogenum 8 R ,11 S -LDS toward linoleic acid was investigated by varying the pH from 5.5 to 9.0, using 50 mM MES buffer (pH 5.5 Ϫ 6.0), 50 mM HEPES buffer (pH 6.0 Ϫ 8.0), 50 mM EPPS buffer (pH 8.0 Ϫ 8.5), and 50 mM 2-(cyclohexylamino)ethanesulfonic acid buffer (pH 8.6 Ϫ 9.0) at 25°C and by varying the temperature from 5°C to 65°C at pH 6.5. The effect of temperature on enzyme stability was investigated by varying the temperature from 15°C to 55°C in 50 mM HEPES buffer (pH 6.5). Enzyme activity was determined after incubating the solution at each temperature for 2 h. To determine the kinetic parameters of the DOX activity in 8 R ,11 S -LDS, O 2 uptake was measured by a Clark-type 5300A biological oxygen monitor (YSI) with 1 g/ml purifi ed enzyme and unsaturated fatty acids such as palmitoleic acid, oleic acid, linoleic acid, and ␣ -linolenic acid (12.5 M to 1 mM). The concentrations of linoleic acid and ␣ -linolenic acid as substrates in the range of 50 to 500 M were used to determine the kinetic parameters of whole enzyme using an HPLC system (Agilent 1100). The kinetic parameters, K m and k cat , were determined using a Hanes-Woolf plot based on the Michaelis-Menten equation.

Structural analysis
Homology modeling of C-terminal domains of 8 R ,11 S -LDS from P. chrysogenum and 5 S ,8 R -LDS from A. nidulans was performed using Build Homology Models module in the MODELER application of Discovery Studio (DS) 4.0 (Accerlys) based on the crystal structure of allene oxide synthase (AOS) from guayule [Protein Data Bank (PDB) entry 3DBM] as a template. Comparative modeling was used to generate the most probable structure of the query protein by aligning it with the template sequence, simultaneously considering spatial restraints, and local molecular geometry. The generated structure was improved by subsequent refi nement of the loop conformations by assessing the compatibility of amino acid sequences with the known PDB structures using Protein Health module in DS 4.0. The geometry of the loop region was corrected using the Refi ne Loop/MOD-ELER, and the best model was selected. The quality of the model was analyzed by PROCHECK ( 22 ). Hydrogen atoms were added to the model and minimized to have a stable energy conformation and to relax the conformation from close contacts. 8 R -HPODE as a substrate was docked into the active-site pocket in the model of C-terminal domains of 8 R ,11 S -LDS from P. chrysogenum and 5 S ,8 R -LDS from A. nidulans using C-DOCKER module, and a sphere with a radius of 4.5 Å around the ligand-binding pocket of the enzyme was defi ned as the active site. Candidate poses were created using random rigid-body rotations, followed by simulated annealing. The structures of the protein, substrate, and their complexes were subjected to energy minimization using the CHARMM force fi eld in DS 4.0 ( 23 ). Full-potential fi nal minimization was used to refi ne the substrate poses. The energy-docked conformation of the substrate was retrieved for postdocking analysis using C-DOCKER module. The substrate orientation giving the lowest interaction energy was chosen for subsequent rounds of docking.

Analytical methods
The reaction products were extracted using an equal volume of ethyl acetate, the solvent was then removed with a rotary evaporator, and methanol was added to the dried extracts. Fatty acids and oxylipins were quantitatively analyzed using the HPLC system mentioned previously, with a UV detector at a detection wavelength of 202 nm and a reversed-phase Nucleosil C 18 column (3.2 × 150 mm, 5 m particle size; Phenomenex). The column was eluted at 35°C with a gradient of solvent A (acetonitrile-wateracetic acid, 50:50:0.1, v/v/v) and solvent B (acetonitrile-acetic GenBank (accession number CAP97986) (supplementary Fig. 1), and the full and partial genes encoding the N-terminal or C-terminal domain (supplementary Fig. 2) were cloned and expressed in E. coli , respectively. The N-terminal domain of the putative fatty acid diol synthase from P . chrysogenum (supplementary Fig. 3A) was identifi ed as the amino acid residues of 1 Ϫ 669 by sequence alignment (supplementary Fig. 3B) with that of 5 S ,8 R -LDS from A. fumigatus ( 4 ), which was the amino acid residues of 1 Ϫ 674 (supplementary Fig. 3C). The expressed protein of the residues of 1 Ϫ 669 was active, whereas the expressed protein of the residues of 93 Ϫ 642, which was predicted as the N-terminal peroxidase-like superfamily by NCBI, exhibited no activity for unsaturated fatty acids. Thus, the Cterminal domain was the amino acid residues of 670 Ϫ 1,074 PpoA, were conserved in P. chrysogenum PC_1, P. digitatum PD_1, P. marneffei PM, and P. oxalicum PO_1, but not in P. chrysogenum PC_2, P. digitatum PD_2, and P. oxalicum PO_2 ( Fig. 2 ). These results suggest that PC_1, PD_1, PM, and PO_1 convert linoleic acid to dihydroxy linoleic acids, whereas PC_2, PD_2, and PO_2 convert linoleic acid to monohydroxy linoleic acids. As P. chrysogenum converted linoleic acid to putative 8,11-DiHODE, the gene encoding PC_1 was cloned, with the aim of identifying a novel 8,11diol synthase.
Gene cloning and purifi cation of the putative fatty acid diol synthase from P . chrysogenum A 3,225 bp gene encoding PC_1 from P. chrysogenum (1,074 amino acids), with the same sequence as that in 311 for peak 1 (4.3 min retention time in HPLC), m / z 311 for peak 2 (5.5 min), m / z 296 for peak 3 (14.9 min), and m / z 311 for peak 4 (15.3 min). Data regarding the fragments formed by cleavage of hydroxylated fatty acids have been reported previously ( 24 ). The m / z 157 and 213 for peak 1, m / z 115 and 173 for peak 2, m / z 157 for peak 3, and m / z 173 for peak 4 resulted from ␣ -cleavage of the hydroxyl groups at the C8 and C11 positions, the C5 and C8 positions, and the C8 position, and the hydroperoxyl group at the C8 position, respectively. The m / z 293 for peak 1, m / z 293 for peak 2, m / z 278 for peak 3, and m / z 293 for peak 4 were derived from the loss of water from molecular anion of each product. Based on this analysis, the compounds represented by peaks 1, 2, 3, and 4 were identifi ed as 8,11-DiHODE, 5,8-DiHODE, 8-HODE, and 8-HPODE, respectively. The LC/MS/MS spectrum of peak 1 consisted of the same pattern as that of the previously reported 8 R ,11 S -DiHODE ( 7 ). 8,11-DiHODE in the HPLC profi le was a major dihydroxy fatty acid product, whereas 5,8-DiHODE was a minor reaction product ( Fig.  3A ), suggesting that the enzyme was 8,11-diol synthase.
Enantiomeric The expressed protein exhibited the catalytic residue Tyr 369 in its N-terminal domain and the crucial residues for hydroperoxide isomerase activity such as His 999 and Cys 1001 in its C-terminal domain. These residues were critical for the conversion of unsaturated fatty acid to dihydroxy fatty acid. The recombinant enzyme was purifi ed as a soluble protein from crude E. coli extract by HisTrap affi nity chromatography. The putative fatty acid diol synthase from P. chrysogenum was purifi ed with a fi nal purifi cation of 6.4-fold, a yield of 21.2% compared with crude enzyme, and the specifi c activity of the purifi ed enzyme for the conversion of linoleic acid to dihydroxy fatty acid was 1.43 mol/min/mg.

Identifi cation of the products obtained from the conversion of unsaturated fatty acids by the putative fatty acid diol synthase from P . chrysogenum
The products obtained from the conversion of linoleic acid by the putative fatty acid diol synthase from P. chrysogenum are represented in the HPLC profi le given in Fig. 3A . Unknown peaks 1, 2, 3, and 4, which were assumed to be oxylipins, were analyzed by LC/MS/MS ( Fig. 3B ). The molecular anions of the products were represented by m / z  truncated C-terminal domain for 8 R -HPODE were evaluated by whole cells containing each domain protein or crude enzyme extracts. Truncated N-terminal domain exhibited activity, whereas truncated C-terminal domain showed no activity. However, the activity of truncated N-terminal domain reduced to 60% of the DOX activity of whole enzyme. Therefore, the DOX activity of whole enzyme was measured to determine the activity of N-terminal domain.
Heme occupancy of the wild-type and H999A, C1001S, and H999A-C1001S variant enzymes of 8 R ,11 S -diol synthase from P . chrysogenum was determined by measuring the ratio of heme absorption to protein absorption ( A 410 / A 280 ) with a UV-visible spectroscopy (supplementary Fig. 4). The H999A, C1001S, and H999A-C1001S variant enzymes showed 0.8-, 0.6-, 0.4-fold lower heme occupancy than the wild-type enzyme, respectively. The hydroperoxide isomerase activity of the double-site H999A-C1001S variant was completely abolished, whereas those of the H999A and C1001S variants remained. The ratios of heme absorption to protein absorption of the wild-type and variant enzymes were closely related with the hydroperoxide isomerase activities.
The effect of pH on the enzymatic production of 8 R -HODE or 8 R ,11 S -DiHODE from linoleic acid was investigated at 25°C. The maximum rates of 8 R -HODE production by DOX activity and 8 R ,11 S -DiHODE production by whole enzyme were observed at pH 8.0 and 6.5, respectively (supplementary Fig. 5). The effect of temperature on enzyme activity was assessed at pH 6.5, and the maximum activities of both DOX and whole enzyme were observed at 25°C (supplementary Fig. 6). The thermal stability of the enzyme was examined by measuring activity after incubation for 2 h at temperatures ranging from 15°C to 55°C. As temperature increased, 8 R ,11 S -DiHODE production by whole enzyme decreased and stopped at 55°C (supplementary Fig. 7), indicating that the activity of C-terminal domain was completely abolished. However, 8 R -HODE production by DOX activity was exhibited below 65°C. Thus, the N-terminal domain demonstrated greater thermal stability than the C-terminal domain.
The production of monohydroxy fatty acids from substrates by the enzyme followed the order ␣ -linolenic acid > linoleic acid > palmitoleic acid > oleic acid ( Table 1 ). However, no activity was observed for conjugated linoleic acid, ␥ -linolenic acid, eicosadienoic acid, dihomo-␥linolenic acid, arachidonic acid, docosapentaenoic acid, and docosahexaenoic acid. Dihydroxy fatty acids were only produced by P. chrysogenum 8 R ,11 S -diol synthase when ␣linolenic acid and linoleic acid were used as substrates, and the specifi c production using linoleic acid was higher than that using ␣ -linolenic acid. The Michaelis-Menten constants ( K m ), turnover numbers ( k cat ), and catalytic effi ciencies ( k cat / K m ) of P . chrysogenum 8 R ,11 S -diol synthase for unsaturated fatty acids are presented in Table 2 . The k cat order of DOX activity for unsaturated fatty acids ( ␣ -linolenic acid > linoleic acid > palmitoleic acid > oleic acid) was the same as that observed for its specifi c activity for the production of monohydroxy fatty acids. However, fatty acid diol synthase from P. chrysogenum was confi rmed by CP-HPLC with 8 R -HODE produced by the H1004A-C1006S variant of 5 S ,8 R -LDS from A. nidulans ( 19 ) and enantiomeric mixture [8-2 H]8-HODE. The retention time of the earlier eluent of enantiomeric mixture  H]8-HODE ( Fig. 4A ) was the same as those of 8 R -HODE obtained from the H1004A-C1006S variant ( Fig. 4B ) and 8-HODE produced by the putative fatty acid diol synthase ( Fig. 4C ). Therefore, the later eluent of enantiomeric mixture [8-2 H]8-HODE was 8 S -HODE, and the product of the enzyme was 8 R -HODE. The diastereoisomers of enantiomeric mixture [8,11-2 H 2 ]8,11-DiHODE were separated as two eluents by NP-HPLC. The retention time of the earlier eluent of enantiomeric mixture [8,11-2 H 2 ]8,11-DiHODE ( Fig. 4D ) was the same as those of 8 R ,11 S -DiHODE ( Fig.  4E ), which was prepared by crude enzyme extract from A. clavatus , and 8,11-DiHODE, which was produced from linoleic acid by the putative fatty acid diol synthase from P. chrysogenum ( Fig. 4F ). The results indicated that the later eluent of enantiomeric mixture [8,11-2 H 2 ]8,11-DiHODE was 8 R ,11 R -or 8 S ,11 S -DiHODE and that 8,11-DiHODE was 8 R ,11 S -or 8 S ,11 R -DiHODE. Because 8 R -HPODE produced by the putative fatty acid diol synthase from P. chrysogenum was a precursor of 8 R ,11 S -DiHODE, the produced 8,11-DiHODE was 8 R ,11 S -DiHODE.

Biochemical properties of 8 R ,11 S -diol synthase from P . chrysogenum
The expressed protein was visualized by SDS-PAGE as a single band of ‫ف‬ 120 kDa (supplementary Fig. 1), which was consistent with the calculated value of 120,998 Da based on the 1,074 amino acid residues plus 6 histidine residues. Truncated N-terminal and C-terminal domains were expressed in cell debris as the bands of ‫ف‬ 74 and 45 kDa, respectively, in SDS-PAGE. However, the C-terminal domain protein was not observed as soluble enzyme in crude extract, and thus it was not purifi ed by HisTrap affi nity chromatography (supplementary Fig. 2). The activities of truncated N-terminal domain for linoleic acid and

DISCUSSION
The C-terminal crucial residues for the hydroperoxide isomerase activity (His 1004 and Cys 1006 ) of A. nidulans 5,8-LDS (PpoA) ( 2 ) were conserved across P. chrysogenum PC_1, P. digitatum PD_1, P. marneffei PM, and P. oxalicum k cat / K m values of DOX activity were in the order linoleic acid > palmitoleic acid > oleic acid > ␣ -linolenic acid, while the order of K m of DOX activity was exactly the reverse order of k cat / K m . The catalytic effi ciency of whole enzyme for linoleic acid was higher than that for ␣ -linolenic acid.  ( 19 ). The 8 R -HPODE produced was converted to 8 R -HODE by the addition of cysteine as a reducing agent (supplementary Fig. 9). These results demonstrated that 8 R ,11 S -diol synthase from P . chrysogenum converted linoleic acid to 8 R ,11 S -DiHODE via 8 R -HPODE and shared a common mechanism with other fatty acid diol synthases ( 5 ). Among the substrates tested, both DOX activity and whole enzyme activity exhibited the highest catalytic effi ciency for linoleic acid ( Table 2 ). Therefore, this enzyme is characterized as an 8 R ,11 S -LDS. P. chrysogenum 8 R ,11 S -LDS converted ␣ -linolenic acid, linoleic acid, palmitoleic acid, and oleic acid to the monohydroxy fatty acids 8-HOTrE, 8-HODE, 8-HHME, and 8-HOME, respectively. Moreover, it produced the dihydroxy fatty acids 8 R ,11 S -DiHODE and 8,11-DiHOTrE from linoleic acid and ␣ -linolenic acid, respectively ( Table 1 ). Based on this activity, we observed that unsaturated fatty PO_1 but differed from those of P. chrysogenum PC_2, P. digitatum PD_2, and P. oxalicum PO_2 ( Fig. 2 ). These results suggest that PC_1, PD_1, PM, and PO_1 can convert linoleic acid to dihydroxy linoleic acids, while PC_2, PD_2, and PO_2 are only able to convert it to monohydroxy linoleic acids. The crude enzyme extracts from P. digitatum and P. oxalicum converted linoleic acid to a monohydroxy linoleic acid (8 R -HODE) and a dihydroxy linoleic acid (7 S ,8 S -DiHODE), respectively. Thus, P. digitatum PD_2 and P. oxalicum PO_1 may be involved in this activity. Although P. marneffei PM containing the crucial residues His and Cys for hydroperoxide isomerase activity was expected to produce a dihydroxy linoleic acid, the enzyme produced monohydroxy linoleic acid (10 R -HODE), indicating that the His and Cys residues in the C-terminal domain may not be critical for 10 R -DOX activity. We proceeded to clone the gene P. chrysogenum PC_1 containing one catalytic residue for DOX activity (Tyr 369 ) and two crucial residues for hydroperoxide isomerase activity (His 999 and Cys 1001 ) because PC_1 might be responsible for the production of the putative 8,11-DiHODE, which was produced without 10 R -HODE accumulation by crude P. chrysogenum enzyme extract ( Fig. 1 ).
The homology models for the active-site in the C-terminal domains of P. chrysogenum 8 R ,11 S -LDS and A. nidulans 5 S ,8 R -LDS were generated based on the known structure of AOS from guayule (PDB entry 3DBM) as a template. A ligand docking study with 8 R -HPODE was conducted using these homology models to identify the specific residues involved in substrate binding. The proposed active-site residues within a sphere of 4.5 Å radius near to acids in which the fi rst cis double bond was located at the C9 and C10 positions were converted to hydroxy fatty acids. Unsaturated fatty acids containing one double bond and two or more double bonds were converted to monohydroxy fatty acids and dihydroxy fatty acids via monohydroxy fatty acids, respectively. Whether monohydroxy or dihydroxy fatty acids are formed by P. chrysogenum 8 R ,11 S -LDS may therefore depend on the position and number of carbon-carbon double bonds in the unsaturated fatty acid substrate.
Some fungi such as A. fumigatus , A. niger , and A. clavatus convert linoleic acid to the by-product 8 R ,11 S -DiHODE as well as 5 S ,8 R -DiHODE ( 5,8,12 ); however, the enzymes involved in 8 R ,11 S -DiHODE production have not been identifi ed. It was previously assumed that PpoB was responsible for 8 R ,11 S -diol synthase activity in A. fumigatus . However, the biosynthesis of 8 R ,11 S -DiHODE by A. fumigatus has only been observed in cytosolic fractions of A. fumigatus , and the enzyme involved in 8 R ,11 S -DiHODE biosynthesis (PpoB) is not found ( 12 ). Thus, the 8,11-LDS isolated from P. chrysogenum in the present study constitutes a novel fatty acid diol synthase.
Amino acid sequences of fatty acid diol synthase-related fungal proteins, including that of 8 R ,11 S -LDS from P. chrysogenum , were used to construct a phylogenetic tree ( Fig. 6 ). Five groups were recovered, with each protein being categorized as a 5 S ,8 R -LDS ( 2,4,9 ), 7 S ,8 S -LDS