Evaluation of the Antimicrobial Activities of Plant Oxylipins Supports Their Involvement in Defense against Pathogens

Screening Oxylipins for Inhibitory Activities against Plant Microbial Pathogens

Fourty-three oxylipins as well as anacardic acid (AA-15:1; Table I; Fig. 1) were prepared and purified in multi-milligram quantities by up scaling previously described procedures (see “Materials and Methods”). They were chosen to represent as much as possible the diversity of structures of this family of metabolites (Fig. 1; Supplemental Fig. 1). Screening of this collection of oxylipins for their effect on microbial plant pathogens was achieved by evaluating the growth of target organisms in liquid media, 24 h after exposure to each oxylipin at 100 μm, in a microspectrophotometric assay. This concentration was selected on the basis of published data on antimicrobial oxylipins, e.g. CLn or 2-hexenal (Croft et al., 1993; Weber et al., 1999), and of preliminary experiments (data not shown). Target organisms were chosen in the major classes of microbial pathogens of crop plants, i.e. bacteria, fungi, and oomycetes (Table II). In vitro growth effects were measured as compared to controls grown in the presence of the ethanol carrier (1%) alone. Results were organized and visualized using a hierarchical clustering software (Seo and Shneiderman, 2002).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1.

Structures of oxylipins and other compounds used in this work. Abbreviated names for oxylipins are as in Table I.

The 44 synthesized compounds as well as 18:2 and 18:3 were assayed at 100 μm on plant bacterial pathogens Pseudomonas syringae, Xanthomonas campestris, and Erwinia carotovora (Fig. 2; Supplemental Table I). Lower concentrations were also tested for the most active compounds. The highly antibacterial oxylipin, 2-hexenal, that prevented the growth of all the bacteria at low concentration (10 μm) was used as a control in these experiments. About half of the tested oxylipins reduced the growth of at least one bacterial strain above 25%. Sensitivity toward oxylipins was contrasted among and within species. P. syringae strains were the most sensitive bacteria, whereas Erwinia and Xanthomonas strains exhibited low sensitivity toward oxylipins. The most active compounds (CL, CLn, ω5(Z)-etherolenic acid [ω5(Z) ELn], 17:3-al, and AA-15:1) were specific to one species. Interestingly, to our knowledge, inhibition of bacterial growth has not been reported before for divinyl ethers. One third of the oxylipins tested did not affect the growth of any bacterial strains.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2.

Antibacterial activities of oxylipins. Each test bacterial strain was exposed to each compound (100 μm) in the appropriate liquid medium, for 24 h. For all compounds reproducibly producing a significant reduction in growth of a given organism compared to the ethanol control (P < 0.05), inhibition of growth over 24 h was expressed as a percentage relative to the ethanol control: 100 × (1 − [growth in the presence of oxylipin/growth in control]). Other compounds were scored as causing 0% growth inhibition for this organism. In the diagram, color intensity is inversely proportional to inhibitory activity of the oxylipin toward the organism. Bacteria names are abbreviated as follows: Ps pv syr, P. syringae pv syringae; Ps pv tom, P. syringae pv tomato; Ps pv mac, P. syringae pv maculicola; Ps pv tab, P. syringae pv tabaci; Xcc, X. campestris pv campestris; and Ecc, E. carotovora subsp. carotovora. Values represent means of at least three independent experiments, with at least three replicates in each experiment.

The same set of compounds was assayed on five fungi, Alternaria brassicicola, B. cinerea, Cladosporium herbarum, Fusarium oxysporum, and Rhizopus sp., and on two oomycetes, Phytophthora infestans and P. parasitica. Spores of some of the target organisms, namely, P. infestans, P. parasitica, B. cinerea, and C. herbarum, which showed a prolonged lag phase in liquid medium (data not shown), were pregerminated for 16 h prior to oxylipin treatment to actually test the activity of oxylipins on mycelial growth. The results obtained after 24 h of treatment with each oxylipin at 100 μm are gathered in Figure 3 and Supplemental Table II. Eukaryotic microbes appeared to be more sensitive to oxylipins than bacteria. About two-thirds of the molecules tested reduced the growth of at least one eukaryotic plant pathogen above 25% and about half of the compounds were strongly active (around 50% growth inhibition). As observed with bacteria, the eukaryotic microorganisms displayed contrasting levels of overall sensitivity to oxylipins, ranging from high (P. parasitica and C. herbarum) to intermediate (B. cinerea, P. infestans, and F. oxysporum) or rather low (A. brassicicola and Rhizopus sp.). One of the most active compounds was 12-Oxo-PDA. This compound is a well established plant signal molecule in wound and stress responses (Stintzi et al., 2001), but its antifungal and antioomycete activity was not previously reported. Four PUFA hydroperoxides (13-HPOT, 9-HPOT, 9-HPOD, and 13-HPOD) as well as their reduced forms (13-HOT, 9-HOT, 13-HOD, and 9-HOD) and dehydration products (notably the ketotrienes, 13-KOT and 9-KOT) were also among the most active oxylipins. In this category, 18:3-derived products seemed slightly more efficient on the strains of the intermediate class of sensitivity to oxylipins than the corresponding 18:2-derived products. Comparison of the effects of 9-LOX-derived compounds to their 13-LOX-derived counterparts, in contrast, did not reveal any obvious relationship between positional specificity and activity.

• Download figure
• Open in new tab
• Download powerpoint

Figure 3.

Antifungal and antioomycete activities of oxylipins. Each test fungal or oomycete strain was exposed to each compound (100 μm) in the appropriate liquid medium for 24 h, starting from the time of spore addition to the wells (F. oxysporum, Rhizopus sp., and A. brassicicola) or 16 h later (P. parasitica, C. herbarum, B. cinerea, and P. infestans). For all compounds reproducibly producing a significant reduction in growth of a given organism compared to the ethanol control (P < 0.05), inhibition of growth over 24 h was expressed as a percentage relative to the ethanol control: 100 × (1 − [growth in the presence of oxylipin/growth in control]). Other compounds were scored as causing 0% growth inhibition for this organism. In the diagram, color intensity is inversely proportional to inhibitory activity of the oxylipin toward the organism. The names of tested organisms are abbreviated as follows: P. parasitica, Phytophthora parasitica nicotianae; C. herbarum, Cladosporium herbarum; B. cinerea, Botrytis cinerea; P. infestans, Phytophthora infestans; F. oxysporum, Fusarium oxysporum; R. spp., Rhizopus species; and A. brassicicola, Alternaria brassicicola. Values represent means of at least three independent experiments, with at least three replicates in each experiment.

Apart from only two oxylipins, 9(S),12(S),13(S)-Trihydroxy-10(E),15(Z)-octadecadienoic acid and 11(R),12(S),13(S)-Trihydroxy-9(Z)-octadecenoic acid, it is clear that most oxylipins exhibited antimicrobial activities. Inhibitory activity was very often restricted to a small number of microbes, but 26 oxylipins out of 43 significantly reduced the in vitro growth of at least three different microbes in this study, and nine oxylipins, 9-HOT, 13-HOD, 9-KOT, 13-KOT, 12(R),13(S)-Epoxy-9(Z)-octadecenoic acid, 12-Oxo-PDA, 13-Hydroxy-12-keto-9(Z),15(Z)-octadecadienoic acid, CLn, and ω5(Z)-ELn, inhibited the growth of six or more different microbes. We next focused on a small number of oxylipins showing inhibitory activities against eukaryotic microbes.

Concentration Required for 50% Growth Inhibition by Selected Oxylipins

The concentration required to inhibit the in vitro mycelial growth of B. cinerea, C. herbarum, P. infestans, and P. parasitica by 50% (IC₅₀) was determined for 12-Oxo-PDA, 13-KOT, 13-HPOT, 13-HOT, 9-HPOT, and 9-HOT. A widely used antifungal molecule (benomyl) and an antioomycete compound (metalaxyl) were assayed in the same conditions against the two fungi and the two oomycetes, respectively. For oxylipins, IC₅₀ values ranged from 25 to 70 μm with C. herbarum and P. parasitica as target organisms, from 50 to 130 μm with B. cinerea, and were estimated to be higher than 150 μm for P. infestans. The IC₅₀ for metalaxyl or benomyl were about 1 to 30 μm for the same target organisms, except for metalaxyl and P. infestans (100 μm). This suggested that natural oxylipins were slightly less efficient than synthetic molecules at inhibiting fungal (or oomycete) growth. We also observed that the growth inhibition effect of these oxylipins was generally transient, with most target organisms overcoming inhibition and resuming growth after 48 to 72 h (data not shown). These observations prompted us to address the stability of the oxylipins in the assay.

Stability of Selected Oxylipins

Many oxylipins are chemically unstable and/or prone to degradation in the presence of biomolecules. Hydroperoxides, for example, are very easily decomposed by metal ions or converted into hydroxides by thiol groups or other reducing agents. Many of the keto-containing oxylipins possess an electrophilic α,β-unsaturated carbonyl partial structure that is prone to Michael addition of glutathione and other thiol-containing compounds (Vollenweider et al., 2000). Additionally, oxylipins may be metabolized by several enzymes (Schaller, 2001; Chechetkin et al., 2004) or enter the mitochondrial or peroxisomal β-oxidation helix, resulting in successive shortening of the carbon chain. To get some insight into their chemical and metabolic stability, the proportion of oxylipins remaining in the assay together with their corresponding metabolites was identified after a 24-h incubation period. Experiments carried out with B. cinerea, C. herbarum, or P. parasitica indicated that the oxylipins introduced were generally hardly detectable in the assay mixture after 24 h (Table III). The structures of the recovered metabolites showed that the hydroperoxides were reduced by the three microorganisms, and 9-HOT, 13-HOT, and 12-Oxo-PDA were degraded through conversion into 15,16-dihydroxylated derivatives or β-oxidation processes, in B. cinerea and C. herbarum. In contrast, the two hydroxides were recovered nearly unmodified after incubation with P. parasitica. CLn and 13-KOT showed low chemical stability in the absence of the fungi, and metabolites could not be detected. CLn may have suffered hydrolysis and chain cleavage into shorter-chain compounds (Galliard et al., 1974), which would have escaped detection by the method used, whereas 13-KOT may have formed Michael adducts with cellular constituents to form water-soluble derivatives (Vollenweider et al., 2000; Almeras et al., 2003). In the case of P. infestans, stability of 9,12,13-THOE, ω5(Z)-ELn, and CL as well as above-mentioned compounds was assayed. Most oxylipins, including 9-HOT and 13-HOT, showed a very low chemical stability in the noninoculated growth medium of P. infestans (data not shown), which might partly explain their low inhibitory activity toward this oomycete as compared to P. parasitica. Only 12-Oxo-PDA and 9,12,13-THOE, a chemically quite stable oxylipin (Hamberg, 1991), were less degraded (44% and 83% remaining after 24 h versus 67% and 90% in medium alone, respectively). Altogether, the data suggested that the half-life of oxylipins in the presence of the target organism might be quite short. Early degradation might therefore explain why many oxylipins were found to reproducibly but only partially affect the growth of the target microorganisms. In spite of this, the selected oxylipins still displayed significant growth-inhibiting activity toward actively growing mycelium of several target organisms.

Effects on Spore Germination

A microscopic visualization approach was used to investigate the effects of oxylipins on spore germination, an important developmental stage in the life cycle of fungi and oomycetes. Interestingly, this method allowed evaluation of oxylipin activity after shorter time exposure. Oxylipins were added (100 μm final concentration) to freshly prepared spores of P. parasitica, P. infestans, or B. cinerea. The percentage of germinated spores was calculated at optimized time points following treatment, for each organism, and compared to the percentage of germination in controls mock treated with the ethanol carrier (1%) alone or untreated (Fig. 4). Except for CL, the compounds tested highly impaired spore germination of oomycetes such as P. parasitica and P. infestans, with a maximum of 5% residual spore germination. B. cinerea spore germination was also efficiently inhibited by five oxylipins. 12-Oxo-PDA was again the most active oxylipin. It is also noteworthy that some oxylipins that produced the highest levels of inhibition at these early time points, durably inhibited spore germination with no further germination being observed during the next 3 to 4 d (data not shown).

• Download figure
• Open in new tab
• Download powerpoint

Figure 4.

Effect of selected oxylipins on spore germination of oomycetes and fungi. Freshly prepared spores of P. parasitica, P. infestans, or B. cinerea (5.10⁴ spores/mL) were exposed to each selected oxylipin (100 μm) and assessed for germination status after 5 h (P. infestans), 6 h (P. parasitica), or 8 h (B. cinerea). Each oxylipin was tested in at least three replicates, in each experiment. Each bar represents mean of three independent experiments ± sd.

Biological Material and Growth Conditions

The plant pathogens used in this work are listed in Table II. The bacterial strains were Pseudomonas syringae pv syringae (NCPPB2686), P. syringae pv tabaci (NCPPB1427), P. syringae pv tomato DC3000 (Whalen et al., 1991), P. syringae pv maculicola (NCPPB 1820), Xanthomonas campestris pv campestris 8004, X. campestris pv campestris 147 (Lummerzheim et al., 1993), and Erwinia carotovora subsp. carotovora SCC3193 (Vidal et al., 1997). The fungal strains were Cladosporium herbarum (Pers.:Fr.) Link, Botrytis cinerea (strain MUCL30158), Fusarium oxysporum f. sp. radicis-lycopersici, Rhizopus strain unspecified, and Alternaria brassicicola (strain MUCL20297; Thomma et al., 1999), kindly provided by B. Mauch-Mani, University of Neuchatel, Switzerland. The oomycete strains were Phytophthora parasitica Dastur var. nicotianae (Breda de Haan) Tucker race1 (Apple's isolate 1452; Hendrix and Apple, 1967) and Phytophthora infestans strain CRA 208 transformed with a green fluorescent protein gene under the control of the Ham promoter (Si-Ammour et al., 2003), kindly provided by Dr Felix Mauch, University of Fribourg, Switzerland. Cultures of A. brassicicola, F. oxysporum, Rhizopus sp., and B. cinerea were maintained on potato (Solanum tuberosum) dextrose agar (39 g/L, Oxoid) and subcultured on a weekly basis. P. parasitica and C. herbarum were grown at 24°C on V8 agar medium (50 mL of V8 juice and 20 g agar per liter, pH 5.0) and subcultured every 14 d. P. infestans was maintained on oat bean agar and subcultured every 11 d. A. brassicicola, F. oxysporum, and Rhizopus sp. spores were harvested in potato dextrose broth (4.8 g/L, DIFCO) and diluted to a concentration of 1 to 2 × 10⁵ spores/mL and used immediately. For C. herbarum conidia production, actively growing mycelium was suspended in sterile distilled water and spread onto a fresh V8 agar plate. After 5 d at 24°C in the dark, a spore suspension was collected by gently scrapping the spores into sterile distilled water. Spore suspension was filtered through a 60 μm scrynell membrane and concentration was adjusted to 4 × 10⁴ spores/mL with water. B. cinerea spore suspension was obtained in a similar way, after growing mycelium for 3 to 4 weeks on potato dextrose agar medium (24 g/L, Sigma) with a photoperiod of 12 h to induce sporulation. Concentration was adjusted to 10⁵ spores/mL. P. parasitica zoospore production was initiated by starvation of 7-d-old mycelium on 1.5% agar plates and zoospores were released in sterile water by cold shock, at 4°C for 30 min, according to Gooding and Lucas (Gooding and Lucas, 1959). The zoospore suspension was collected and density adjusted to 4 × 10⁴ or 10⁵ zoospores/mL with sterile distilled water. Similarly, P. infestans zoospores were collected in sterile distilled water, after 3 h at 4°C, and spore density was adjusted to 5 × 10⁵ zoospores/mL.

Antimicrobial Tests

In vitro antimicrobial activities of oxylipins were evaluated in a microspectrophotometric assay, according to Broekaert et al. (1990). Microorganisms were grown in sterile, flat-bottom, 96-well microplates, sealed with Parafilm, in a final volume of 100 μL. Bacteria were grown at 28°C starting with approximately 2,000 bacteria per well. Pseudomonas strains were grown in King's medium (2% proteose peptone, 2% glycerol, 6.5 mm K₂HPO₄, and 6 mm MgSO₄ 7H₂O), Erwinia in LPG medium (0.3% yeast [Saccharomyces cerevisiae] extract, 0.5% Bacto-peptone, and 0.5% Glc), and Xanthomonas in Kado medium (1% yeast extract, 0.8% casamino acids, 1% saccharose, 6.5 mm K₂HPO₄, and 6 mm MgSO₄ 7H₂O). Fungi and oomycetes were grown in the dark at 25°C, except for P. infestans, which was grown at 18°C. A. brassicicola, F. oxysporum, and Rhizopus sp. were grown in potato dextrose broth (4.8 g/L, DIFCO) in an orbital shaker. B. cinerea, C. herbarum, and P. parasitica were grown in clarified V8 juice broth (5% V8 juice in water, pH 5.0, centrifuged three times at 3,000 rpm, for 20 min) without shaking. P. infestans was grown in water, onto solid oat bean medium (100 μL each by well), without shaking. Cultures were started with 2,000 spores (C. herbarum and P. parasitica) or 5,000 spores (B. cinerea, A. brassicicola, F. oxysporum, Rhizopus sp., and P. infestans). Fatty acids, oxylipins, AA-15:1, metalaxyl (methyl n-methoxyacetyl-n-2,6-xylyl-DL-alaninate, Riedel-deHaën) and benomyl [methyl 1-(butylcarbamoyl)benzimidazol-2-ylcarbamate, Riedel-deHaën] were dissolved in ethanol. Each treatment was performed in three to eight replicates in the same plate, and experiments were repeated three or four times independently. Ethanol concentration was kept to a maximum of 1% final concentration in each well; controls were treated with 1% ethanol.

Except for P. infestans, growth was monitored by measuring the absorbance of the microcultures at 595 nm with a microplate reader (Bio-Rad or DYNEX Technologies), at 0 h and after 24 to 72 h of incubation in the presence or absence of compounds to be tested. Growth of P. infestans transformed with green fluorescent protein was determined by measuring fluorescence emission at 530 nm after excitation at 485 nm, with a CytoFluor II fluorescence reader (BioSearch, Millipore). For C. herbarum, B. cinerea, P. parasitica, and P. infestans, oxylipins were added after 16 h of growth and A₅₉₅ (or fluorescence) measured at 16 h and after a further 24 to 72 h period. At least three wells were used in each experiment and the experiments repeated at least three times independently. Analysis of variance of antimicrobial test results was performed on final absorbance (or fluorescence) values within experiments. The test compounds were tested against different organisms in groups of up to 12 compounds, with all compounds being tested on each of a number of replicate plates alongside control treatments. As a result, individual plates were treated as replicates and the effect of different compounds on the growth of individual organisms was analyzed for small groups of compounds and compared to the values gained for control treatments in that group of plates only. Therefore, the different analyses of variance do not provide comparisons of the effectiveness of one compound against another, but do provide a measure of the impact of individual compounds against the control treatment (1% ethanol) in that experiment.

Growth inhibition for each compound and each target organism was calculated using the following expression:

In the experiments involving P. infestans, fluorescence was used instead of absorbance. Results were organized and visualized using the Hierarchical Clustering Explorer software (version 2.0 beta; Human-Computer Interaction Lab, University of Maryland [http://www.cs.umd.edu/hcil/hce/]; Seo and Shneiderman, 2002).

Spore Germination Assay

Spore germination in the presence of oxylipins was assessed in 96-well microplates, in clarified V8 juice broth (P. parasitica and B. cinerea) or water (P. infestans). Oxylipins in ethanol were added in each well to a final concentration of 100 μm (1% ethanol final concentration) together with 5,000 spores or zoospores, in triplicate, in a total volume of 100 μL, and the plates incubated for a few hours. For each target organism, time points for spore status assessment were optimized for high number of germinated spores in the controls together with short germ tubes (average length of germ tubes not higher than 8- to 10-fold the spore diameter) for better visualization of each germling. The plates were observed in an inverted microscope (DMIRBB, Leitz or Leitz DM IL, Leica), and for each well two to four nonoverlapping pictures (magnification: 100×) were acquired with a charge-coupled device camera (Color Coolview, Photonic Sciences or Nikon D1x) and 100 to 200 spores or zoospores subsequently assessed for germination status. Spores with germ tubes as long or longer than the spore were counted as being germinated.

Oxylipins

The oxylipins used in this study are listed in Table I and Figure 1. Many of them were prepared from the hydroperoxides 9(S)-HPOD, 13(S)-HPOD, 9(S)-HPOT, or 13(S)-HPOT, which were in turn prepared by incubation of 18:2 or 18:3 (Nu-Chek-Prep) with LOX from tomato (Lycopersicon esculentum) or soybean (Glycine max; Gardner, 1996; Lie and Pasha, 1998). Secondary enzymes used for transformations of hydroperoxides into other oxylipins included allene oxide synthase (Hamberg, 1987; Baertschi et al., 1988), allene oxide cyclase (Hamberg and Fahlstadius, 1990), and divinyl ether synthase (Galliard and Phillips, 1972; Hamberg, 1998; Supplemental Fig. 1). Peroxygenase (Blée and Schuber, 1990; Hamberg and Hamberg, 1990) and epoxide hydrolase (Hamberg and Hamberg, 1996) can be used for the preparation of various epoxy alcohols and trihydroxy oxylipins, however, chemical epoxidation and hydrolysis methods (Hamberg, 1991) are generally more convenient to use in these cases. α-DOX derivatives were produced as previously described (Hamberg et al., 1999). A number of oxylipins, e.g. monoepoxy fatty acids, were isolated as natural products from various seeds (Gunstone et al., 1994), and another group including short-chain aldehydes was prepared from commercially available starting materials using chemical methods. 2-Hexenal and 1-penten-3-ol were purchased from Sigma-Aldrich. 6-[8(Z)-Pentadecenyl]salicylic acid (AA-15:1), a nonoxylipin belonging to the anacardic acid family of antibacterial compounds (Kubo et al., 2003), was isolated from cashew nut shell liquid (Paramashivappa et al., 2001). Final purification of each compound was in most cases carried out by straight-phase HPLC using 2-propanol/hexane-based solvent systems. The structures of the compounds obtained, as well as the stereochemical fidelity and purity in case of optically active compounds, were determined by chemical and physical methods as described either in the references given in Table I or in other studies cited by these references. The purity of each compound was in excess of 97% as determined by straight phase-HPLC and/or gas chromatography-mass spectrometry (GC-MS). Most of the compounds used have recently become commercially available through Larodan Fine Chemicals and/or Cayman Chemical or Biomol.

Estimation of the Stability of Oxylipins

Oxylipins (13-HPOT, 9-HPOT, 13-HOT, 9-HOT, 12-Oxo-PDA, 13-KOT, and CLn, 100 μm each) were added to cultures of P. parasitica, C. herbarum, or B. cinerea in liquid growth medium, or to medium alone, and kept at 25°C. Aliquots were removed at 0 and 24 h and added to ethanol. 14(S)-Hydroxy-10(Z),12(E),16(Z)-nonadecatrienoic acid (3 μg) was added as an internal standard and the mixtures extracted with diethyl ether. The material obtained was derivatized by treatment with diazomethane followed by trimethylchlorosilane/hexamethyldisilazane/pyridine and analyzed by GC-MS using a Hewlett-Packard model 5970B mass-selective detector connected to a Hewlett-Packard model 5890 gas chromatograph. A capillary column of 5% phenylmethylsiloxane (12 m, 0.33 μm film thickness) was used with helium as the carrier gas. The column temperature was raised at 10°C/min from 120°C to 260°C. The derivatized samples were analyzed first in the scan mode to characterize metabolites or degradation products formed from the various oxylipins. A second analysis performed in the selected ion monitoring mode was used to determine the levels of oxylipins present in the 24 h samples relative to those in the 0 h samples. The following mass spectral ions were used in these analyses: mass-to-charge ratio (m/z) 325 (internal standard; M⁺, CH₂-CH=CH-C₂H₅), 311 (13-HOT; M⁺, CH₂-CH=CH-C₂H₅), 225 [9-HOT; Me₃SiO⁺=CH-(CH=CH)₂-C₅H11], 238 (12-Oxo-PDA; M⁺, [CH₂-CH=CH-C₂H₅ minus H], 306 (CLn; M⁺), and 237 (13-KOT; M⁺, CH₂-CH=CH-C₂H5). The percentage level of oxylipin remaining at 24 h relative to the level at 0 h was calculated using the following expression:

In a second series, selected oxylipins were added to cultures of P. infestans grown in water on solid medium, or to water on solid medium alone, and samples obtained at 0 and 24 h were treated as described above. The oxylipins used in this case included the above-mentioned ones as well as 9,12,13-THOE (m/z 173), ω5(Z)-ELn (m/z 306), and CL (m/z 308).

Estimation of the stability of the hydroperoxides 13-HPOT and 9-HPOT could not be performed using the method described due to their thermal instability, which precluded analysis by GC-MS; however, a number of stable metabolites formed from them could be identified.