Investigating the catalytic efficiency of C22-Fatty acids with LOX human isozymes and the platelet response of the C22-oxylipin products

Abstract

Polyunsaturated fatty acids (PUFAs) have been extensively studied for their health benefits because they can be oxidized by lipoxygenases to form bioactive oxylipins. In this study, we investigated the impact of double bond placement on the kinetic properties and product profiles of human platelet 12-lipoxygenase (h12-LOX), human reticulocyte 15-lipoxygenase-1 (h15-LOX-1), and human endothelial 15-lipoxygenase-2 (h15-LOX-2) by using 22-carbon (C22) fatty acid substrates with differing double bond content. With respect to k_cat/K_M values, the loss of Δ⁴ and Δ¹⁹ led to an 18-fold loss of kinetic activity for h12-LOX, no change in kinetic capability for h15-LOX-1, but a 24-fold loss for h15-LOX-2 for both C22–FAs. With respect to the product profiles, h12-LOX produced mainly 14-oxylipins. For h15-LOX-1, the 14-oxylipin production increased with the loss of either Δ⁴ and Δ¹⁹, however, the 17-oxylipin became the major species upon loss of both Δ⁴ and Δ¹⁹. h15-LOX-2 produced mostly the 17-oxylipin products throughout the fatty acid series. This study also investigated the effects of various 17-oxylipins on platelet activation. The results revealed that both 17(S)-hydroxy-4Z,7Z,10Z,13Z,15E,19Z-DHA (17-HDHA) and 17-hydroxy-4Z,7Z,10Z,13Z,15E-DPAn6 (17-HDPAn6) demonstrated anti-aggregation properties with thrombin or collagen stimulation. 17-hydroxy-7Z,10Z,13Z,15E,19Z-DPAn3 (17-HDPAn3) exhibited agonistic properties, and 17-hydroxy-7Z,10Z,13Z,15E-DTA (17-HDTA) showed biphasic effects, inhibiting collagen-induced aggregation at lower concentrationsbut promoting aggregation at higher concentrations. Both 17-hydroxy-13Z,15E,19Z-DTrA (17-HDTrA), and 17-hydroxy-13Z,15E-DDiA (17-HDDiA) induced platelet aggregation. In summary, the number and placement of the double bonds affect platelet activation, with the general trend being that more double bonds generally inhibit aggregation, while less double bonds promote aggregation. These findings provide insights into the potential role of specific fatty acids and their metabolizing LOX isozymes with respect to cardiovascular health.

1. Introduction

Omega-3 fatty acids are a group of essential fatty acids (FA) that are important for preventing hypertriglyceridemia [1], hypertension [2], inflammation [3,4], arrhythmia [5,6], atherosclerosis [7,8], and thrombosis [9,10]. cis-4,7,10,13,16,19-docosahexaenoic acid (DHA) is a long-chain polyunsaturated fatty acid with six double bonds (22:6) that is found in high concentrations in salmon, mackerel, and tuna. DHA plays a role in preventing atherothrombosis, which is the formation of blood clots within arteries [11-13]. Docosapentaenoic acid (DPA), another essential FA, is a 22-carbon long polyunsaturated fatty acid with five double bonds (22:5) [14,15], which can be both an omega-3 (DPAn3) and an omega-6 fatty acid (DPAn6) (15). These two DPA fatty acids have been shown to be important in the prevention of atherosclerosis [16,17]. DPAn3 affects colorectal cancer [18], chronic inflammatory diseases [19,20], and myocardial infarct [21], while DPAn6 affects stroke [22]. In addition, higher levels of cis-7,10,13,16,19-docosapentaenoic acid (DPAn3) have been shown to be associated with improved cognitive function, specifically memory [23-25]. One other essential FA, cis-7,10,13,16-docosatetraenoic acid (DTA), is a 22-carbon long polyunsaturated fatty acid with four double bonds (22:4), it composes 17.5% of all fatty acids found in cerebral grey matter [26,27], has been shown to have an inverse association with fetal growth [28], and is positively associated with primary open-angle glaucoma, a disease that leads to vision loss and blindness [29-33].

The biological benefits of these fatty acids can be attributed to the fact that they can be oxidized into oxylipins by lipoxygenases (LOXs). Lipoxygenases are a family of enzymes that catalyze the oxidation of polyunsaturated fatty acids (PUFAs) to form a variety of biologically active lipid mediators, including leukotrienes and lipoxins [34-39]. There is a wide spectrum of PUFA substrates that can be catalyzed by lipoxygenase isozymes in the human body [40-42]. A key chemical property of LOX substrates is the presence of the 1,4-pentadiene moiety, with its position relative to either the carboxylate or methyl end of the fatty acids dictating which LOX isozyme is a competent catalyst [43-45]. Arachidonic acid (AA) is the canonical substrate of LOX isozymes due to its high levels in the human body and thus the LOX isozymes are named according to their positional specificity with AA [46-50].

With respect to the biological role of LOX isozymes, platelet activation is highly dependent on LOX products, as seen by the varied aggregation response with the addition of oxylipins [22,51-58]. The 12-oxylipin products of AA and eicosapentaenoic acid (EPA, C20:5) do not inhibit platelet aggregation, while the dihomo-γ-linolenic acid (DGLA, C20:3) product does inhibit aggregation, indicating the positional importance of double bonds with respect to 12-oxylipin activity [52,53]. Interestingly, the 15-oxylipins of all three of these C20 fatty acids inhibit platelet aggregation, indicating that both the number of double bonds and the position of the oxidation are critical to the selectivity of receptor binding [51]. This hypothesis is extended to DHA (C22:6), where its 11-, 14-, and 17-oxylipins are all potent aggregation inhibitors, with their potency increasing with the loss of a double bond at either end of the C22-oxylipin (i.e., DPAn3 and DPAn6 oxylipins versus DHA oxylipins) [53,56,59,60]. These data highlight the biological importance of double bond location, length, and oxidation position of oxylipins with respect to platelet activation.

In the current work, the kinetics, product profile, and platelet activity of a family of C22–FAs against h12-LOX, h15-LOX-1, and h15-LOX-2 were performed. This C22 family of fatty acids includes DHA (22:6), DPAn3 (22:5), cis-4,7,10,13,16-docosapentaenoic acid (DPAn6) (22:5), DTA (22:4), cis-13,16,19-docosatrienoic acid (DTrA) (22:3), and cis-13,16-docosadienoic acid (DDiA) (22:2) (Fig. 1). By quantifying the catalytic efficiency of these C22–FAs and examining the distribution of oxylipins, their biological availability was determined. Finally, the platelet activity of these oxylipins was determined, establishing the effect of double bond positioning on platelet aggregation.

Fig. 1.

C22–FAs used in this study.

2. Experimental procedures

2.1. Chemicals

Fatty acids used in this study were purchased from Nu Chek Prep, Inc. (MN, USA) and the oxylipins isolated in house (vide infra). Soybean Lipoxygenase-1 (SLO-1) was purchased from Cayman Chemicals. All other solvents and chemicals were reagent grade or better and were used as purchased without further purification.

2.2. Expression and purification of 12LOX, h15-LOX-1, and h15-LOX-2

Recombinant expression and purification of his-tagged h12-LOX (Uniprot entry P18054), h15-LOX-1 (Uniprot entry P16050) and h15-LOX-2 (Uniprot entry O15296) were performed using nickel-affinity chromatography, as previously described [61,62]. The purity of h12LOX, h15-LOX-1 and h15-LOX-2 were assessed by SDS gel to be greater than 85%. The metal content was assessed on a Finnigan inductively-coupled plasma-mass spectrometer (ICP-MS), via comparison with iron standard solution and applied to the kinetic parameters. Cobalt-EDTA was used as an internal standard.

2.3. Product determination and isolation

h15-LOX-1 (0.125 μM, pH 7.5), h12-LOX (0.300 μM, pH 8.0), h15-LOX-2 (0.5 μM, pH 7.5) were reacted in 2 mL of 25 mM HEPES at room temperature and ambient oxygen. The enzymatic turnover was monitored by absorbance at 234 nm, with 20 μM fatty acid reacted until 20% turnover occurred (DHA, DPAn3, DPAn6, DTA, DTrA and DDiA). The reactions were quenched with 1% glacial acetic acid and extracted three times with dichloromethane (DCM). The products were then reduced with trimethyl phosphite and evaporated under a stream of nitrogen gas. The reaction products were reconstituted in methanol and analyzed via liquid chromatography-tandem mass spectrometry (LC-MS/MS). Chromatographic separation was performed using a C18 column (Phenomenex Kinetex, 4 μm, 150 × 2.0 mm). Mobile phase solvent A consisted of 99.9% water and 0.1% formic acid, and solvent B consisted of 99.9% acetonitrile and 0.1% formic acid. Analysis was carried out over 60 min using isocratic 50:50 A:B for 0–30 min followed by a gradient from 50:50 A:B to 75:25 A:B from 30 to 60 min. The chromatography system was coupled to a Thermo-Electron LTQ LC-MS/MS instrument for mass analysis. All analyses were performed in negative ionization mode at the normal resolution setting. MS/MS was performed in a targeted manner with a mass list containing the following m/z ratios containing the following m/z ratios of 343.2 ± 0.5 (HDHAs), 345.2 ± 0.5 (HDPAs), 347.2 ± 0.5 (HDTAs), 349.2 ± 0.5 (HDTrAs), 351.2 ± 0.5 (HDDiAs) were used (see Supporting Information, Figs. S1-S5). Matching retention times, UV spectra, and fragmentation patterns for four common fragments were used to identify products without standards. To identify the known standards, at least three common fragments were used to identify the products (listed in the supporting information). DHA fragmentations used to identify our oxylipins were from previous literature [54,63]. The commercially obtained standards that were used were: d8-12-HETE, 17-HDHA, 14-HDHA, 11-HDHA, and 12-HETE. It should be noted that the enzyme products coeluted with the known standards, so this fact along with the catalytic nature of LOX isozymes allowed us to assume the products were of the S-configuration.

For product isolation for platelet activity, the same protocol was used except for the high-performance liquid chromatography (HPLC) purification method and amounts of enzyme and fatty acid used. The products were purified isocratically via HPLC on a Higgins Haisil Semi-preparative (5 μm, 250 mm × 10 mm) C18 column with 45:55 acetonitrile: water, and 0.1% acetic acid. Oxylipin production reactions required at least 2 h to complete. 17-HDHA was synthesized from DHA (50 μM) and h15-LOX-1. 17-DPAn3 was synthesized by SLO-1 and DPAn3 (50 μM), and 17-DPAn6 was synthesized by SLO-1 with DPAn6 (50 μM). 17-HDTA and 14-HDTA were synthesized by reaction of DTA (50 μM) with h15-LOX-1 and h12-LOX respectively. 17-HDTrA was synthesized by reaction of DTrA (50 μM) with SLO-1. 17-HDDiA were synthesized by reaction of DDiA (50 μM) with SLO-1.

2.4. Steady-state kinetics

h12-LOX, h15-LOX-1 and h15-LOX-2 reactions were performed at 22 °C in a 1 cm² quartz cuvette containing 2 mL of 25 mM HEPES (pH 7.5) with substrates (DHA, DPAn3, DPAn6, DTA, DTrA, and DDiA). All substrate concentrations varied from 0.25 to 25 μM. Concentrations of all substrates were determined by measuring the amount of oxylipins produced from complete reaction with SLO-1. Concentrations of oxylipins were determined by measuring the absorbance at 234 nm. Reactions were initiated by the addition of h12-LOX (130 nM), h15-LOX-1 (300 nM) and h15-LOX-2 (325 nM). Product formation was determined by the increase in absorbance at 234 nm for oxylipin products (ε_234nm = 25,000 M⁻¹ cm⁻¹) with a PerkinElmer Lambda 45 UV/Vis spectrophotometer. KaleidaGraph (Synergy) was used to fit initial rates (at less than 10% turnover), to the Michaelis-Menten equation for the calculation of kinetic parameters.

2.5. Oxylipin titration in human platelet aggregometry

All research involving human subjects was carried out in accordance with the Declaration of Helsinki and approved by the University of Michigan Institutional Review Board (approval number: HUM00100677). Written informed consent was obtained prior to blood collection. Washed platelets were isolated from human whole blood via serial centrifugation and resuspended at a concentration of 3.0 × 10⁸ platelets per milliliter in Tyrode’s buffer (10 mM HEPES, 12 mM NaHCO₃, 127 mM NaCl, 5 mM KCl, 0.5 mM NaH₂PO₄, 1 mM MgCl₂, and 5 mM glucose), as previously published [56,64,65]. Platelets (250 μl at 3.0 × 10⁸ platelets per milliliter) were dispensed into glass cuvettes and oxylipin at the indicated concentrations (0–40 μM) was added to the platelet solution. The aggregation response to each oxylipin was measured in a Chrono-log Model 700D lumi-aggregometer for 10 min at 37 °C. Oxylipins that did not induce platelet aggregation were subsequently added to 250 μl of platelets and incubated for 10 min at 37 °C to determine if the oxylipin had an inhibitory effect. Oxylipin-treated platelets were stimulated with an EC₈₀ concentration of either collagen (Chrono-log) or thrombin receptor agonist under stirring conditions (1100 rpm) at 37 °C. Platelet aggregation was recorded for 10 min. Bar graphs are shown in the following figures (vide infra), however, aggregation dose responses are shown as Figs. S6-S9 in the supporting information.

3. Results and discussion

3.1. h12-LOX kinetics and product profile of fatty acids

The k_cat values for h12-LOX with DHA, DPAn3, DPAn6, and DTA at 22 °C compare to that of AA (Table 1) [66]. They were 13 ± 0.4 s⁻¹, 7.1 ± 0.5 s⁻¹, 11 ± 0.9 s⁻¹, and 0.40 ± 0.05 s⁻¹, respectively; however, DTrA and DDiA were not substrates (Table 1). The k_cat/K_M for h12-LOX with DHA, DPAn3, DPAn6, and DTA at 22 °C was 7.4 ± 0.3 μM⁻¹ s⁻¹, 4.0 ± 0.3 μM⁻¹ s⁻¹, 5.3 ± 1 μM⁻¹ s⁻¹, and 0.30 ± 0.03 μM⁻¹ s⁻¹, respectively (Table 1). These data indicate that the loss of either Δ⁴ or Δ¹⁹ double bonds of DHA (i.e., DPAn3 and DPAn6) does not significantly affect the k_cat or k_cat/K_M values. However, losing both Δ⁴ and Δ¹⁹ (i.e., DTA) causes more than a 10-fold decrease in the k_cat and k_cat/K_M values and the loss of additional double bonds (i.e., DTrA and DDiA) leads to complete loss of catalysis. The latter results can be easily explained due to the lack of a 1,4-diene in the appropriate position relative to the active site iron for DTrA and DDiA, however, the 10-fold loss of activity for DTA relative to DPAn3 and DPAn6 is less clear. The loss of both Δ⁴ and Δ¹⁹ would make DTA more flexible and more hydrophobic due to the loss of double bonds, which could affect its positioning in the active site. Another possible explanation is the loss of an aromatic interaction. Previously, it was observed that F414 of h12-LOX pi-pi stacked with Δ¹¹ of AA [66], therefore it is possible that there is an aromatic residue which could pi-pi stack with both Δ⁴ and Δ¹⁹, leading to a loss of catalytic activity for DTA. For example, a loss of activity relative to the number of double bonds was previously observed where the k_cat value for mead acid (20:3) was ten-fold less than that found for AA (20:4), illustrating how the loss of a double bond could have drastic effects to LOX activity [66]. It should be noted that the k_cat and k_cat/K_M values of DHA, DPAn3, DPAn6 with h12-LOX correlated well with our previous work [51,56,63,66,67].

Table 1.

Kinetics values of h12-LOX with various C22–FAs. NA = no activity.^a previously published kinetic values (Aleem et al., 2019).

	kcat (s−1)	kcat/KM (μM−1 s−1)	KM (μM)
AAa	28 (1)	14 (1)	2 (0.2)
DHA	13 (0.4)	7.4 (0.3)	1.7 (0.3)
DPAn3	7.1 (0.5)	4.0 (0.3)	1.8 (0.4)
DPAn6	11 (0.9)	5.3 (1)	2.2 (0.2)
DTA	0.40 (0.05)	0.30 (0.03)	1.1 (0.09)
DTrA	NA	NA	NA
DDiA	NA	NA	NA

With respect to product formation, h12-LOX with DHA and DPAn6 produced mostly the 14-oxylipin products, with the 14-oxylipin: 11-oxylipin ratios being 90:10 and 100:0, respectively (Table 2), consistent with previous work [55,58,64]. There is an increase in promiscuity in product formation with DPAn3 and DTA, with 17-oxylipins, 14-oxylipins, and 11-oxylipins being made in the ratios of 4:75:21 and 11:76:12, respectively. Nevertheless, h12-LOX still produces mainly the 14-oxylipin product for all the C22–FAs which had activity. Considering that an increase of the 17-oxylipin relative to the 14-oxylipin occurs when the substrate enters the active site less deeply, and the 11-oxylipin occurs when the fatty acid enters more deeply, it appears that loss of either Δ⁴ or both Δ⁴ and Δ¹⁹ leads to less precise oxidation. This is consistent with the substrate not being held as tightly in the active site, possibly due to an increase in flexibility/hydrophobicity and/or a loss of pi-pi stacking (vide supra) [51,56,63,66-70]. It is interesting to note that this hypothesis for product formation parallels the lowered rates observed for DTA but not DPAn3. This could indicate that the product ratios are more sensitive to enzyme/substrate interactions, indicating that the double bond at Δ⁴ is critical for product profiling, while the loss of both Δ⁴ and Δ¹⁹ is more critical for kinetic efficiency. We are currently investigating specific active site amino acid interactions with Δ⁴ and Δ¹⁹ through mutagenesis in order to tease out the specific interactions.

Table 2.

Product distribution of h12-LOX with various fatty acids. n/d = not detected, NA = no activity.

	17-product	14-product	11-product
DHA	n/d	90 ± 1	10 ± 1
DPAn6	n/d	100 ± 1	n/d
DPAn3	4 ± 1	75 ± 1	21 ± 0.5
DTA	11 ± 0.2	76 ± 3	12 ± 1
DTrA	NA	NA	NA
DDiA	NA	NA	NA

3.2. h15-LOX-1 kinetics and product profile of fatty acids

The k_cat values for h15-LOX-1 with DHA, DPAn3, DPAn6, DTA, DTrA, and DDiA at 22 °C compare to that of AA (Table 3) [71]. They were 1.6 ± 0.1 s⁻¹, 1.0 ± 0.1 s⁻¹, 1.7 ± 0.2 s⁻¹, and 2.5 ± 0.04 s⁻¹, 1.6 ± 0.06 s⁻¹, and 3.3 ± 0.04 s⁻¹ respectively. The k_cat/K_M with DHA, DPAn3, DPAn6, DTA, DTrA, and DDiA at 22 °C was 0.49 ± 0.03 μM⁻¹ s⁻¹, 0.72 ± 0.1 μM⁻¹ s⁻¹, 0.47 ± 0.1 μM⁻¹ s⁻¹, and 0.35 ± 0.01 μM⁻¹ s⁻¹, 0.37 ± 0.02 μM⁻¹ s⁻¹, and 0.81 ± 0.06 μM⁻¹ s⁻¹, respectively (Table 3). Surprisingly, there were no significant changes in the k_cat or k_cat/K_M values with the loss of multiple double bonds in the C22-FA scaffold, indicating that the degree and/or placement of the double bonds does not affect h15-LOX-1 kinetics appreciably. This is consistent with past literature results which reported that the loss of the Δ⁵ double bond in AA did not affect the kinetic values for h5-LOX-1 [51]. It should be noted that the k_cat and k_cat/K_M values of DHA with h15-LOX-1 correlated well with our previous work [51,56,63,66,67].

Table 3.

Kinetic values of h15-LOX-1 with various C22–FAs.^a previously published kinetic values (Wecksler et al., 2008).

	kcat (s−1)	kcat/KM (μM−1 s−1)	KM (μM)
AAa	5.3 (0.5)	2.0 (0.2)	2.7 (0.3)
DHA	1.6 (0.1)	0.49 (0.03)	3.3 (0.7)
DPAn3	1.0 (0.1)	0.72 (0.1)	1.4 (0.5)
DPAn6	1.7 (0.2)	0.47 (0.1)	3.7 (0.3)
DTA	2.5 (0.04)	0.35 (0.01)	7.2 (0.1)
DTrA	1.6 (0.06)	0.37 (0.02)	4.3 (0.2)
DDiA	3.3 (0.04)	0.81 (0.06)	4.1 (0.2)

With respect to product profiling, h15-LOX-1 reacts with DHA, DPAn6, and DPAn3 producing both the 14-oxylipin and 17-oxylipin products, with a preference for the 14-oxylipin products with DPAn3 and DPAn6 (Table 4). With DHA as the substrate, the ratio of 17HDHA:14HDHA:11HDHA was 48:45:7, which is consistent with previous work [51,56,63,66,67]. With DPAn6 as the substrate, the ratio of 17-HDPAn6:14-HDPAn6 was 37:63, with no 11-product being made. With DPAn3 as the substrate, the ratio of 17-HDPAn3:14-HDPAn3 was 23:77. However, with DTA as the substrate, the product preference shifts to the 17-oxylipin product, with the ratio of 17-HDTA:14-HD-TA:11-HDTA being 68:23:5. For DTrA and DDiA, the ratio of 17-HDTrA:13-HDTrA is 82:18 and 86:14, respectively. The 17-oxylipin was the majority product because the 14-oxylipin is not catalytically competent due the position of the 1,4-diene relative to the methyl end of the substrate.

Table 4.

Product distribution of h15-LOX-1 with various C22–FAs. n/d = not detected.

	17 -product	14 -product	13-product	11-product
DHA	48 ± 3	45 ± 2	n/d	7 ± 0.5
DPAn6	37 ± 3	63 ± 3	n/d	n/d
DPAn3	23 ± 3	77 ± 3	n/d	n/d
DTA	68 ± 2	23 ± 1	n/d	5 ± 0.4
DTrA	82 ± 1	n/d	18 ± 1	n/d
DDiA	86 ± 2	n/d	14 ± 2	n/d

Analyzing the kinetic and product data for h15-LOX-1, it is observed that the number and placement of double bonds do not affect enzymatic rates significantly, but the product profile is affected. When both the Δ⁴ and Δ¹⁹ double bonds are removed, there is a novel shift in the product specificity that has not been reported previously. For DHA, DPAn6, and DPAn3, the 14-oxylipin is the majority product, but for DTA, the 17-oxylipin is the majority product, indicating that DTA, slides less deep into the cavity site of h15-LOX-1 than DHA, DPAn6, and DPAn3, thus making more 17-oxylipins. Since DTA is both more flexible and more hydrophobic than either DPAn3 or DPAn6, this effect could be due to a general binding mode change. Another explanation could be that there is a specific residue interacting with the double bonds of DTA, pulling it further into the active site. However, considering that the effect is only observed when both Δ⁴ and Δ¹⁹ are removed, it appears there is not one interaction that is important, but rather that there is a synergistic effect between the two double bonds and the active site residues. Previously in the literature, a pi-pi interaction between the Δ¹¹ or Δ¹⁴ double bond of AA and F414 was observed, but the Δ¹¹ and Δ¹⁴ double bonds are in close proximity so in theory, F414 could interact with both double bonds [69,72]. The distance between Δ⁴ and Δ¹⁹ double bonds of DHA is approximately 4.5 Å, greater than that between Δ¹¹ and Δ¹⁴ and thus one critical amino acid interaction seems unlikely. Nevertheless, H425 is within 5 Å of F414, suggesting the possibility that H425 could be at least partially responsible for the observed kinetic effects. We are currently beginning a mutagenesis campaign to probe the role of H425 and other active site residues that affect C22 substrate catalysis. With respect to DTrA and DDiA, 17-oxylipin is the major product because 14-oxylipin is not catalytically feasible. Due to improper double bond positioning relative to the active site iron, both fatty acids appear to flip and enter the active site in the opposite orientation. This reverse orientation could produce the 13-oxylipin from DTrA and DDiA, similar to that seen for h5-LOX catalysis [54,55,71,72], It is also possible that this change in product profile could also be due to improper substrate positioning in the active site, but this is less likely based on previous results (vide supra).

3.3. h15-LOX-2 kinetics and product profile of fatty acids

The k_cat values for h15-LOX-2 with DHA, DPAn3, DPAn6, DTA, DTrA, and DDiA at 22 °C compare to that of AA (Table 5) [73]. They were 3.0 ± 0.1 s⁻¹, 1.5 ± 0.07 s⁻¹, 2.0 ± 0.1 s⁻¹, and 0.13 ± 0.3 s⁻¹, 0.029 ± 0.08 s⁻¹, and 0.014 ± 0.02 s⁻¹, respectively. The k_cat/K_M with DHA, DPAn3, DPAn6, DTA, DTrA, and DDiA at 22 ◦C was 0.78 ± 0.07 μM⁻¹ s⁻¹, 0.38 ± 0.08, 0.76 ± 0.2 μM⁻¹ s⁻¹, and 0.032 ± 0.004 μM⁻¹ s⁻¹, 0.022 ± 0.003 μM⁻¹ s⁻¹, and 0.045 ± 0.005 μM⁻¹ s⁻¹, respectively (Table 5). There is no significant difference in k_cat or k_cat/K_M values with the loss of either Δ⁴ or Δ¹⁹ from DHA (i.e., DPAn3 and DPAn6). However, the loss of both Δ⁴ and Δ¹⁹, such as for DTA, DTrA and DDiA, leads to more than a greater than 10-fold difference in the k_cat and k_cat/K_M for h15-LOX-2. As previously proposed for h12-LOX (vide supra), this loss of reactivity could be due to an increase in flexibility/hydrophobicity of the fatty acid, or a loss of a specific interaction with an active site residue, such as pi-pi stacking [70,72,74].

Table 5.

Kinetic values of h15-LOX-2 with various C22–FAs.^a previously published kinetic values (Joshi et al., 2013).

	kcat (s−1)	kcat/KM (μM−1 s−1)	KM (μM)
AAa	1.5 (0.1)	0.40 (0.05)	3.8 (0.4)
DHA	3.0 (0.1)	0.78 (0.07)	3.9 (0.3)
DPAn3	1.5 (0.07)	0.38 (0.08)	3.8 (0.4)
DPAn6	2.0 (0.1)	0.76 (0.2)	2.6 (0.5)
DTA	0.13 (0.3)	0.032 (0.004)	4.0 (0.4)
DTrA	0.029 (0.08)	0.022 (0.003)	1.3 (0.2)
DDiA	0.014 (0.02)	0.045 (0.005)	0.31 (0.03)

With all C22–FAs, h15-LOX-2 produced mostly the 17-oxylipins, consistent with its high level of specificity (Table 6). With DHA, DPAn6, DPAn3, and DTA, the ratios between the 17-product: 14-product were 94:6, 96:4, 92:8, and 100:0, respectively. An exception to this trend is the production of the 13-product with DTrA and DDiA, with the ratio of 17-HDDiA:13-HDDiA being 88:12 and 37:63, respectively. As discussed previously for h15-LOX-1 (vide supra), the production of the 14-oxylipin is not possible due to the positions of the double bonds relative to the terminal methyl, and thus 13-oxylipin from h15-LOX-2 and DTrA and DDiA is most likely due to the fatty acid entering the active site carboxylic acid end first presumably making the R-product, as seen for h5-LOX [75].

Table 6.

Product distribution of h15-LOX-2 with various C22–FAs. n/d = not detected.

	17-product	14 -product	13-product
DHA	94 ± 1	6 ± 1	n/d
DPAn6	96 ± 1	4 ± 1	n/d
DPAn3	92 ± 4	8 ± 4	n/d
DTA	100 ± 1	n/d	n/d
DTrA	88 ± 1	n/d	12 ± 1
DDiA	37 ± 5	n/d	63 ± 5

Excluding the production of the 13-oxylipin, the number of double bonds does not affect the product profile for h15-LOX-2 substantially, which could be due to a lack of pi-pi stacking. Previous literature has shown that F414 is involved in pi-pi stacking with the substrate for h12-LOX and h15-LOX-1, but this residue does not exist in the active site of h15-LOX-2 [70,75], which could explain the lack of dependence on double bond content for product production.

3.4. Platelet aggregation effect of C22-oxylipins with variation of oxidation position and double bond number/placement

The C22-oxylipins were assessed for their ability to regulate platelet aggregation, either through a stimulatory role, an inhibitory role, or a biphasic capacity. Previously, it was demonstrated that 17-HDHA was a potent anti-aggregation effector molecule [53,59,76]. In the current investigation, this family of 17-oxylipins with 22 carbons was investigated further, with 17-HDPAn3, 17-HDPAn6, 17-HDTA, 17-HDTrA, and 17-HDDiA being assessed. In the present investigation, 17-HDHA demonstrated inhibition of collagen-stimulated platelets at relatively low concentrations (24±10% aggregation at 5 μM), while higher concentrations of 17-HDHA were required to inhibit thrombin-stimulated platelets (63±10% aggregation at 25 μM) (Fig. 2). 17-DPAn6 had comparable potency with 14±4% aggregation at 5 μM when stimulated with collagen and 29±9% at 25 μM when stimulated with thrombin (Fig. 2). Interestingly, 17-DPAn3 showed no inhibitory affects but did have agonistic properties with 26±10% aggregation at 5 μM with no platelet stimulation. 17-HDDiA and 17-HDTrA were also agonists with 55±10% and 84±2% at 5 μM, respectively (Fig. 3). 17-HDTA is the unusual oxylipin of this family of 17-oxylipins. At 0.5 μM of 17-HDTA, 20±10% platelet aggregation was observed with collagen added, making it a more potent inhibitor than either 17-HDHA or 17-HDPAn6. However, at higher concentrations, 17-HDTA was an agonist. At 5 μM of 17-HDTA, 41±10% aggregation was observed with no platelet stimulation (Fig. 4).

Fig. 2.

Inhibition of platelet aggregation with 17-HDHA and 17-HDPAn6.

Fig. 3.

Agonistic effect of platelet aggregation with 17-HDPAn3, 17-HDTrA and 17-HDDiA.

Fig. 4.

Biphasic effect of platelet aggregation with 17-HDTA.

In general, these data indicate that as the 17-oxylipin loses double bond content, it converts from a platelet activation inhibitor to a platelet activation agonist (Table 7). The transition point appears to be between 17-HDPAn3 and 17-HDPAn6, with loss of Δ¹⁹ maintaining the inhibitory properties of the 17-HDPAn6, but loss of Δ⁴ leads to agnostic properties (i.e., 17-HDPAn3). Interestingly, 17-HDTA, which has lost both Δ⁴ and Δ¹⁹, is both an inhibitor and an agonist. It is tempting to speculate that Δ⁴ and Δ¹⁹ on the 17-oxylipin regulate two distinct receptor pathways. The mechanism of action of these 17-oxylipins is currently being investigated to better understand these intriguing results.

Table 7.

Effects of oxylipins on platelet aggregation.

Oxylipin	Inhibition Effect (Thrombin Stimulation)	Inhibition Effect Collagen Stimulation	Agonistic Effect No Stimulation
17-HDHA	63 ± 10% @ 25 μM	24 ± 10% @ 5 μM	None
17-DPAn6	29 ± 9% @ 25 μM	14 ± 4% @ 5 μM	None
17-DPAn3	None	None	26 ± 10% @ 5 μM
17-HDTA	None	20 ± 10% @ 0.5 μM	41 ± 10% @ 5 μM
17-HDTrA	None	None	84 ± 2% @ 5 μM
17-HDDiA	None	None	55 ± 10% @ 5 μM
14-HDTA	49 ± 10% @ 10 μM	4.8 ± 1% @ 5 μM	None

With respect to the 14-oxylipins, it has previously been observed that 14-HDHA is a potent aggregation inhibitor with 20% aggregation at 5 μM with collagen stimulation [56]. 14-HDPAn6 was also shown to be a potent inhibitor, with 5% at 1 μM when stimulated with collagen and 10% at 1 μM, when stimulated with thrombin [22]. Considering the similar potency between the 14-oxylipins and the 17-oxylipins, the platelet aggregation effect was assessed for 14-HDTA to determine if the same transition to agonist as seen for 17-HDTA would also be observed for 14-HDTA. Contrary to the data for 17-HDTA, 14-HDTA was only an inhibitor, with 4.8±1% aggregation at 5 μM when stimulated with collagen and 49±10% at 10 μM when stimulated with thrombin (Fig. 5). These data indicate that the 14-oxylipins do not follow a similar platelet effect mechanism as do the 17-oxylipins and thus, the change in location of oxidation affects their mechanism of actions. This result is like that seen for the 12- and 15-oxylipins derived from AA, where 12-HETE induces pro-thrombotic platelet activity, while 15-HETE induces anti-thrombotic activity [51,53]. It should be noted that due to the lack of activity of h12-LOX against DTrA and DDiA, the isolation of 14-oxylipin products from these fatty acids was not achieved.

Fig. 5.

Inhibition of platelet aggregation with 14-HDTA.

4. Conclusion

In summary, the kinetic rates and product preference of each of the three LOX isozymes, h12-LOX, h15-LOX-1, and h15-LOX-2, is generally not affected if either Δ⁴ or Δ¹⁹ is removed from DHA (i.e., DPAn3 or DPAn6). However, if both Δ⁴ and Δ¹⁹ are removed, the rate for both h12-LOX and h15-LOX-2 is reduced more than 10-fold, suggesting that the change in carbon hybridization increases the flexibility and hydrophobicity of DTA, which leads to a less-productive ES complex. For h15-LOX-2, the rate and product distribution are affected even further with additional loss of double bonds (i.e., DTrA and DDiA), indicating the ES complex becomes even more un-productive with additional flexibility and hydrophobicity. In contrast, the kinetic values for h15-LOX-1 do not change appreciably with a decrease in the number of double bonds, indicating an ES complex that is less dependent on flexibility and hydrophobicity than that for h12-LOX and h15-LOX-2. Interestingly, while the kinetics of h15-LOX-1 are not affected by the double bond content of the substrate, the product profile is affected. As the C22–FAs have fewer double bonds, more of the 17-oxylipin is produced indicating a reduced penetration of the substrate into the active site. It is unclear why these C22–FAs have reduced active site penetration as their flexibility and hydrophobicity are increased. It is possible that the loss of a specific pi-pi interaction with an active site residue is lost, lowering the energetics of active site insertion. We are currently investigating if a specific active site residue interacts with the fatty acids and accounts for these data through mutagenesis.

With respect to the effect the 17-oxylipins had on platelet activity, the number of double bonds on the C22-FA-derived oxylipins had substantial effects, on both inhibition and aggregation. 17-HDHA and 17-HDPAn6 exhibited inhibitory effects on platelet aggregation, with the effect being larger with collagen activation. However, as the number of double bonds decreases, the effect of the 17-oxylipin switches to agonistic. The transition occurs with 17-DPAn3, which is purely agonistic, and 17-HDTA, which is mixed inhibitor/agnostic, with the key chemical feature being the loss of Δ⁴. Surprisingly, the 14-oxylipin of DTA, 14-HDTA, has a different effect on platelet aggregation. While 14-HDTA inhibits platelet aggregation with both thrombin and collagen, 17-HDTA is inhibitory at low concentrations but agonistic at high concentrations. These findings highlight the complexity of platelet regulation by oxylipins and may have implications for their biological significance. For example, since h15-LOX-1 reacts equally well with each of these four C22–FAs, but their 17-oxylipin effects on platelet aggregation are orthogonal to each other, the micro-distribution of these C22–FAs in one’s diet could have distinct consequences with regards to platelet activity and hence cardiovascular indications. In addition, although platelets do not contain 15-LOX isozymes, the oxylipins produced by 15-LOXs are highly potent biomolecules towards platelet activation, suggesting that other cells, such as leukocytes may influence platelet function significantly. We are currently investigating the inhibitory and agonistic mechanism of action of these oxylipins to determine the biochemical requirements for these distinct platelet effects.

Abbreviations

LOX

lipoxygenase

h12-LOX

human platelet 12S-lipoxygenase

h15-LOX-1

human reticulocyte 15-lipoxygenase-1

h15-LOX-2

human epithelial 15-lipoxygenase-2

r15-LOX-1 or r12/15-LOX or rALOX15

rabbit 15S-LOX-1

COX

cyclooxygenase

WT h12-LOX

wild-type human platelet 12S-lipoxygenase

SLO-1

soybean lipoxygense-1

ML355

h12-LOX specific inhibitor

NSAIDs

nonsteroidal anti-inflammatory drugs

coxibs

COX-2 selective inhibitors

ICP-MS

inductively coupled plasma mass spectroscopy

SEC

size exclusion chromatography

BSA

bovine serum albumin

DCM

dichloromethane

EDTA

Ethylenediaminetetraacetic acid

HEPES

N-2-hydroxyethylpiperazine-N′-2-ethanesulfonic acid

SDS-PAGE

sodium dodecyl sulfate-polyacrylamide gel electrophoresis

PDB

Protein Data Bank

AA

arachidonic acid

12(S)-HpETE

12(S)-hydroperoxyeicosatetraenoic acid

12(S)-HETE

12(S)-hydroxyeicosatetraenoic acid

12(S)-HETrE

12(S)-hydroxyeicosatrienoic acid

d8-12HETE

12(S)-Hydroxyeicosatetraenoic Acid-d8

C22–FAs

22 carbon long fatty acids

DHA

cis-4,7,10,13,16,19-docosahexaenoic acid

11-HDHA

11(S)-hydroxy-4Z,7Z,9E,13Z,16Z,19Z-DHA

14-HDHA

14(S)-hydroxy-4Z,7Z,10Z,12E,16Z,19Z-DHA

17S-HDHA

17(S)-hydroxy-4Z,7Z,10Z,13Z,15E,19Z-DHA

20-HDHA

20-hydroxy-4Z,7Z,10Z,13Z,16Z,18E-DHA

DPAn3

cis-7,10,13,16,19-docosapentaenoic acid

11-HDPAn3

11-hydroxy-7Z,9E,13Z,16Z,19Z-DPAn3

14-HDPAn3

14-hydroxy-7Z,10Z,12E,16Z,19Z-DPAn3

17-HDPAn3

17-hydroxy-7Z,10Z,13Z,15E, 19Z -DPAn3

DPAn6

cis-4,7,10,13,16-docosapentaenoic acid

11-HDPAn6

11-hydroxy-4Z,7Z,9E,13Z,16Z-DPAn6

14-HDPAn6

14-hydroxy-4Z,7Z,10Z,12E, 16Z – DPAn6

17-HDPAn6

17-hydroxy-4Z,7Z,10Z,13Z,15E-DPAn6

DTA

cis-7,10,13,16-docosatetraenoic acid

11-HDTA

11-hydroxy-7Z,9E,13Z,16Z-DTA

14-HDTA

14-hydroxy-7Z,10Z,12E,16Z-DTA

17-HDTA

17-hydroxy-7Z,10Z,13Z,15E-DTA

DTrA

cis-13,16,19-docosatrienoic acid

13-HDTrA

13-hydroxy-14E,16Z,19Z-DTrA

17-HDTrA

17-hydroxy-13Z,15E,19Z-DTrA

DDiA

cis-13,16-docosadienoic acid

13-HDDiA

13-hydroxy-14E,16Z-HDDiA

17-HDDiA

17-hydroxy-13Z,15E-DDiA

Data availability

No data was used for the research described in the article.