5,6-DiHETE attenuates vascular hyperpermeability by inhibiting Ca2+ elevation in endothelial cells[S]

Although more than 100 lipid metabolites have been identified, their bioactivities remain unknown. In a previous study, we discovered that the production of several lipid metabolites in the intestines dramatically changed in colitis. Of these metabolites, 5,6-dihydroxyeicosatetraenoic acid (DiHETE) possesses novel anti-inflammatory activity in the vasculature. In this study, we used mouse and human umbilical vein endothelial cell (HUVEC) models to elucidate the mechanisms underlying the vascular activity of lipid metabolites, particularly those related to the release of histamine, a major proinflammatory mediator that stimulates endothelial cells to produce NO, a mediator of vascular relaxation and hyperpermeability, by activating intracellular Ca2+ concentration-dependent signaling. In a mouse ear, the administration of 5,6-DiHETE did not induce inflammatory reactions, and pretreatment with 5,6-DiHETE inhibited histamine-induced inflammation, specifically vascular dilation and hyperpermeability. In an isolated mouse aorta, 5,6-DiHETE treatment did not influence vascular contraction but attenuated acetylcholine-induced vascular relaxation. In HUVECs, treatment with 5,6-DiHETE inhibited histamine-induced endothelial barrier disruption and inhibited the production of NO. Most notably, 5,6-DiHETE inhibited histamine-induced increases in intracellular Ca2+ concentrations in HUVECs. Our findings suggest that 5,6-DiHETE attenuates vascular hyperpermeability during inflammation by inhibiting endothelial Ca2+ elevation, which might lead to a novel pharmacological strategy against inflammatory diseases.

100 lipids and their metabolites have been identified (1). but their bioactivities remain to be elucidated. In a previous study in mice, we found that the concentrations of several lipid metabolites significantly increased or decreased in colon tissue depending on the progression and healing of dextran sodium sulfate-induced colitis (2). Severely inflamed colon tissue was found to contain higher concentrations of docosahexaenoic acid-derived 7-hydroxydocosahexaenoic acid (HDoHE). The concentrations of -linolenic acid-derived 13-hydroxyoctadecatrienoic acid (HOTrE) and eicosapentaenoic acid-derived 5,6dihydroxyeicosatetraenoic acid (DiHETE) were found to be elevated in the healing colon tissue. Thus, the abovementioned lipid mediators potentially exert bioactivities that participate in the progression and/or resolution of inflammation.
During inflammation, various bioactive substances, including lipid metabolites, are secreted from the vascular tissues, as well as by infiltrating immune cells. Some of these substances stimulate hyperpermeability of the vasculature, which in turn causes plasma leakage and infiltration of immune cells into the interstice of tissue (3,4). This phenomenon is necessary to eliminate foreign substances and to promote tissue repair. However, excessive and/or sustained vascular hyperpermeability can trigger a variety of inflammatory diseases (5,6). Thus, the appropriate modulation of vascular permeability is crucial for managing health and disease.
The vasculature is primarily composed of endothelial cells and smooth muscle cells. Endothelial cells form a monolayer that covers the inner surface of the blood vessels, while smooth muscle cells encapsulate the outer surface of the endothelial layer. These two types of cells undergo coordinated action to modulate vascular reactions. Upon inflammation, some bioactive substances increase vascular permeability by modulating two major factors: blood flow and endothelial barrier function. Histamine is a major proinflammatory mediator that is involved in the acute immune response and strongly increases vascular permeability. It stimulates endothelial cells to produce a vascular relaxant factor NO by activating intracellular Ca 2+ concentration ([Ca 2+ ] i )-dependent signaling (7). The vascular dilations increase the downstream blood flow and intraluminal hydrostatic pressure, which in turn causes vascular hyperpermeability (8). In contrast, the endothelial barrier is formed mainly by intercellular adherens junctions consisting of vascular endothelial cadherin and the cytoskeleton (4). Histamine also disrupts the endothelial adherence junction by increasing [Ca 2+ ] i (9). Certain lipid metabolites, especially cyclooxygenase-derived prostanoids, are known to modulate vascular function (10). Prostaglandins E 2 and I 2 cause vascular dilatation and hyperpermeability (11,12). Likewise, thromboxane A 2 causes vascular hyperpermeability by disrupting endothelial battier function (13). In addition, we recently demonstrated that prostaglandin D 2 causes vascular hypopermeability by enhancing endothelial barrier formation (14). However, there are numerous lipid metabolites whose bioactivities remain unknown.
In this study, we attempted to determine the bioactivities of 7-HDoHE, 13-HOTrE, and 5,6-DiHETE by evaluating their effects on vascular function both in vivo and in vitro.

Animals
Male C57BL/6 mice aged 8-12 weeks were used. All experimental procedures in this study were approved by the Institutional Animal Care and Use Committee at the University of Tokyo.

Cell culture
Human umbilical vein endothelial cells (HUVECs) were purchased from Lonza. HUVECs were cultured in EGM-2. These cells (passages 4-9) were used for experiments after 4 h of starvation in EBM-2 supplemented with 2% FBS.

In vivo microscopy
To visualize the ear vessels of the mice using a confocal microscope (ECLIPSE Ti with C1 confocal system; Nikon, Tokyo, Japan), 70 kDa FITC-dextran (10 mg/kg; Sigma-Aldrich) was injected intravenously. Mice were then positioned on the microscope stage, and their body temperatures were maintained at 37°C. Dextran leakage and vascular diameter were monitored every minute and quantified as described previously (11) using EZ-C1 Free Viewer (Nikon).

Measurement of vascular contraction
The aortic arteries of the mice were excised and placed in a physiological saline solution containing 136.9 mM NaCl, 5.4 mM KCl, 5.5 mM glucose, 23.8 mM NaHCO 3 , 1.5 mM CaCl 2 , 1.0 mM MgCl 2 , and 0.01 mM EDTA. After removing fat and connective tissue, the aortic arteries were cut into rings. The contractile force of the vascular rings was then isometrically recorded with a forcedisplacement transducer (Orientec, Tokyo, Japan) connected to a strain amplifier (Yokogawa, Japan) under a resting tension of 3 mN. Precontraction was induced by norepinephrine (0.3 or 1 M). Acetylcholine (0.03-1 M) was added cumulatively.

Measurement of TER
Endothelial barrier function was evaluated by measuring transendothelial electrical resistance (TER) with the use of the xCEL-Ligence Real-Time Cell Analyzer DP system (Roche, Basel, Switzerland). This system monitors changes in TER over time across an interdigitated microelectrode at the bottom of tissueculture E-plates (Roche). Cells (8,000) were plated on E-plates and incubated until confluent. TER was measured every 60 s. For normalization, cell index values at each time point were shown as a ratio to the initial value. We quantified the maximum normalized cell index after stimulation and represented it as maximal response.

Measurement of NO production
HUVECs were washed three times with HEPES and incubated in HEPES supplemented with 1 mM L-arginine and 10 M tetrahydrobiopterin. After equilibrating for 30 min, the media were exchanged, and the cells were incubated for another 30 min. Cells were subsequently stimulated with the indicated agents for 15 min and stimulated with 10 M histamine for 5 min. Conditioned media before and after stimulation (100 l each) were collected and centrifuged at 300 g for 3 min. The supernatants were collected and used for the measurement of nitrite and nitrate, stable metabolites of NO, by an ENO-20 NOx analyzer (Eicom, Japan). The degrees of increase in nitrite and nitrate levels after stimulation were normalized with cell protein contents.

Measurement of Ca 2+ concentration
HUVECs were incubated with 3 g Fura 2-AM containing 0.11% Cremophor EL for 30 min and washed three times with HEPES-buffered solution. Coverslips were placed into a specialized airtight chamber mounted onto the stage of microscope maintained at 37°C. HUVECs were excited at 340 and 380 nm, and the emitted fluorescence signal was collected every 3 s at 510 nm. The fluorescence ratio was determined using a fluorescence imaging system (AQUACOSMOS; Hamamatsu Photonics, Japan). After the experiments, 1 M ionomycin was added to measure the fluorescence alteration in the presence of 0 or 1.5 mM Ca 2+ . Ca 2+ concentration was shown as the change in the fluorescence ratio. The area under the curve (AUC) was calculated to assess the change in the fluorescence ratio 1 min after each stimulation.

Statistical analysis
The results of the experiments were expressed as means ± SEMs. Data evaluations were conducted using one-way ANOVA, followed by Bonferroni's test for comparison between more than two groups. P < 0.05 was regarded as significant.

5,6-DiHETE inhibits histamine-induced vascular hyperpermeability
First, we performed the modified Miles assay to assess the effects of the lipid metabolites on vascular permeability in vivo. Figure 1A shows representative images of mouse ears, and Fig. 1B shows the results for the quantification of dye extravasation. The administration of 400 g histamine induced blue dye extravasation (0.14 ± 0.02 g/mg); this amount was twice the amount used in the vehicle treatment (0.06 ± 0.02 g/mg). Histamine stimulates a histamine H1 receptor to increase [Ca 2+ ] i in endothelial cells. This signaling activation promotes endothelial NO production, which in turn causes vascular dilation and hyperpermeability. Pretreatment with the histamine H1 receptor blocker diphenhydramine (2.5 g per ear; 15 min) or a Ca 2+ channel blocker LaCl 3 (250 g per ear; 15 min) consistently and significantly inhibited the dye extravasation.
In subsequent experiments, we focused the investigation on the inhibitory effects of 5,6-DiHETE on histamineinduced vascular activation.

5,6-DiHETE inhibits histamine-induced arterial dilation
We next investigated the effect of 5,6-DiHETE on histamine-activated vasculature in the mouse ears. Results of intravital imaging showed that 400 g histamine caused extravasation of FITC-dextran mainly from the bifurcation area of vasculature and peripheral vascular vessels ( Fig.  2A, B). The mean fluorescence intensity was significantly higher at 30 min following histamine administration (Fig.  2B). Pretreatment with 0.1 g 5,6-DiHETE (15 min) narrowed the area of FITC-dextran extravasation ( Fig. 2A) and significantly decreased the mean intensity relative to that of histamine-treated samples (Fig. 2B).
Smooth muscle contraction reduces downstream blood flow and limits vascular leakage, whereas smooth muscle relaxation increases blood flow, thereby leading to vascular leakage. We next assessed the changes in the vascular diameters of the mouse ears. Figure 2C shows the typical blood vessels of the mouse ear, and Fig. 2D shows the results of quantification of arterial diameters. The administration of 400 g histamine increased the arterial diameter to 178% ± 13% relative to that of nonstimulated arteries, which suggested that histamine induced vascular dilation. Interestingly, pretreatment with 0.1 g 5,6-DiHETE significantly inhibited the histamine-induced increase in arterial diameter by 132% ± 15% relative to that of nonstimulated arteries. On the other hand, Fig. 2E shows the results of the quantification of vein diameters. Histamine administration tended to increase vein diameter, but pretreatment with 5,6-DiHETE did not inhibit the histamine-induced increase in vein diameter.

5,6-DiHETE inhibits histamine-induced endothelial barrier dysfunction
We evaluated endothelial barrier function in vitro by measuring the TER of HUVECs. Barrier disruption resulted in lower TER, whereas barrier enhancement increased TER. As shown in Fig. 4A and B, treatment with 1 U/ml thrombin significantly decreased TER, whereas treatment with an adenylate cyclase activator forskolin (1 M) increased TER. Treatment with 5,6-DiHETE alone (0.1 M and 0.3 M) did not change TER.  We next examined the effects of 5,6-DiHETE under histamine stimulation. As shown in Fig. 4C and D, treatment with 10 M histamine significantly decreased TER (minimum of 0.82 ± 0.05-fold), which was abrogated by pretreatment with 10 M diphenhydramine, a histamine H1 receptor antagonist. In addition, pretreatment with 0.3 M 5,6-DiHETE (15 min) significantly inhibited the histamine-induced barrier disruption (minimum of 0.915 ± 0.012-fold). These results suggest that 5,6-DiHETE attenuated endothelial barrier dysfunction under histamine stimulation.

5,6-DiHETE inhibits histamine-induced eNOS phosphorylation and NO production in endothelial cells
Both histamine and acetylcholine lead to eNOS phosphorylation and NO production (15,16). Thus, we examined whether 5,6-DiHETE inhibited eNOS phosphorylation and NO production in HUVECs. Treatment with 10 M histamine (5 min) resulted in eNOS phosphorylation at serine1177, which is essential for NO synthesis in HUVECs. Figure 5A shows the representative results of Western blot analysis of total eNOS and phosphorylated eNOS levels, while Fig. 5B shows the corresponding phosphorylation levels. Pretreatment with 0.1 M 5,6-DiHETE (15 min) significantly reduced histamine-induced phosphorylation of eNOS. As shown in Fig. 5C, 10 M histamine stimulated NO production (0.31 ± 0.06 pmol/g protein), which was abrogated by pretreatment with the histamine H1 receptor blocker diphenhydramine (10 M; 15 min). Consistent with the results for eNOS phosphorylation, pretreatment with 0.1 M 5,6-DiHETE (15 min) significantly inhibited histamine-induced NO production (0.11 ± 0.06 pmol/g protein).
These findings show that 5,6-DiHETE inhibited eNOS phosphorylation and NO production in endothelial cells.

DISCUSSION
Our findings revealed that 5,6-DiHETE attenuates inflammation in the vasculature by inhibiting vascular hyperpermeability. Our study is the first to demonstrate the bioactivity of 5,6-DiHETE, a cytochrome P450-catalyzed metabolite produced from eicosapentaenoic acid following epoxidation of the -5 double bond. Isomers of 5,6-DiHETE have been reported to promote aryl hydrocarbon receptor-mediated transcriptional activity, which potentially stimulates inflammatory signaling (17). However, no studies have investigated the bioactivity of 5,6-DiHETE. We originally showed that 5,6-DiHETE production is increased during the healing stage of murine colitis. 5,6-DiHETE may have a crucial role in the promotion of healing by inhibiting excessive or sustained inflammation.
Although 5,6-DiHETE concentrations produced in inflamed tissue have not been documented, Yamada et al. (1) reported that 5,6-DiHETE concentrations in normal, noninflamed mouse heart, liver, and kidney tissues were 30-70 pg/mg. In this study, we utilized a considerably higher concentration of 5,6-DiHETE (10 ng/mg; 0.1 g per ear) to evaluate its bioactivity. 5,6-DiHETE is a lipid metabolite that can be produced in and affect the local vasculature. Further investigations are required to identify the pathophysiological role of endogenous 5,6-DiHETE in inflammation.
As described above, vascular permeability is primarily determined by two factors: blood flow and endothelial barrier function. Vascular dilation increases blood flow in peripheral tissue. In turn, the increased blood flow exerts intraluminal hydrostatic pressure, which leads to plasma leakage from the blood vessels (18). In particular, nonactivated eNOS binds to caveolin found on the endothelial cell membrane and is subsequently translocated into the cytoplasm in response to an increase in [Ca 2+ ] i . Afterward, a serine/threonine kinase Akt phosphorylates/activates eNOS to induce NO production (19,20). Endothelial cellderived NO causes the dilation of smooth muscle cells through cyclic guanosine monophosphate-protein kinase G signaling. The increase in [Ca 2+ ] i also disrupts the endothelial barrier directly by forming the actin stress fiber and destabilizing adherence junctions (19,21). In this study, we found that 5,6-DiHETE inhibited the increase in [Ca 2+ ] i , which in turn attenuated vascular dilation and hyperpermeability. Interestingly, 5,6-DiHETE decreased endothelial cell activation but did not influence smooth muscle contractility. 5,6-DiHETE appeared to modulate endothelial cell-specific Ca 2+ channels, which attenuates both the dilation of smooth muscle cells and endothelial barrier dysfunction, resulting in the inhibition of vascular hyperpermeability.
Appropriate modulation of vascular barrier is indispensable for maintaining tissue homeostasis. Vascular dysfunction is a critical step in the pathogenesis of inflammatory disease. Given that endothelial barrier enhancement is likely to inhibit an excessive inflammatory response, identifying a novel barrier enhancer and elucidating its mechanism of action will potentially provide new therapeutic strategies for the treatment of various diseases.
This study is the first to verify the anti-inflammatory effects of 5,6-DiHETE and the mechanisms underlying its activity. However, further studies are required to identify the source cells, receptor involved, endogenous physiological activity, and precise mechanisms underlying the function of 5,6-DiHETE. In addition, we were unable to identify the bioactivities of 0.1 g 7-HDoHE or 13-HOTrE by assessing vascular function. However, we cannot exclude the possibility that these lipid metabolites modulate the progression and/or healing of inflammation through mechanisms other than modulating vascular function. Further investigations are needed to clarify the abovementioned points. Fig. 6. 5,6-DiHETE inhibits His-induced increase in Ca 2+ concentrations in endothelial cells. A: Effect of pretreatment with 10 M Diphe or 0.03, 0.1, and 0.3 M 5,6-DiHETE on 10 M His-induced increase in Ca 2+ concentrations in HUVECs. B: Quantification by AUC at 1 min after stimulation (n = 4-5). *P < 0.05 and ***P < 0.001 relative to His. Diphe, diphenhydramine; His, histamine.