Omega-3 fatty acid prevents the development of heart failure by changing fatty acid composition in the heart

Some clinical trials showed that omega-3 fatty acid (FA) reduced cardiovascular events, but it remains unknown whether omega-3 FA supplementation changes the composition of FAs and their metabolites in the heart and how the changes, if any, exert beneficial effects on cardiac structure and function. To clarify these issues, we supplied omega-3 FA to mice exposed to pressure overload, and examined cardiac structure and function by echocardiography and a proportion of FAs and their metabolites by gas chromatography and liquid chromatography-tandem mass spectrometry, respectively. Pressure overload induced cardiac hypertrophy and dysfunction, and reduced concentration of all FAs’ components and increased free form arachidonic acid and its metabolites, precursors of pro-inflammatory mediators in the heart. Omega-3 FA supplementation increased both total and free form of eicosapentaenoic acid, a precursor of pro-resolution mediators and reduced free form arachidonic acid in the heart. Omega-3 FA supplementation suppressed expressions of pro-inflammatory cytokines and the infiltration of inflammatory cells into the heart and ameliorated cardiac dysfunction and fibrosis. These results suggest that omega-3 FA-induced changes of FAs composition in the heart have beneficial effects on cardiac function via regulating inflammation.

In this study, we examined whether omega-3 FA supplementation has any effects on the composition of FA in the heart and whether these changes have beneficial effects on cardiac function and remodeling.

Results
Omega-3 FA ameliorates cardiac dysfunction induced by pressure overload. To assess the change of FA composition during a process of cardiac remodeling under pathological stresses, we firstly made the murine model of cardiac hypertrophy induced by constricting transverse aorta (TAC) 27 . We treated mice with vehicle and omega-3 FAs, major components of FAs in cardiomyocyte 20,28 to clarify whether FA composition had some effects on cardiac function and structure. We administered Omega-3 acid ethyl esters (Omega-3 EE), a compound containing both EPA and DHA, 1.5 mg/g of body weight once a day. Although this dose is ~twice of a clinical dose in human 29 , this dose which had been reported to have a triglyceride-lowering effect in a rodent 30 was adopted in this study. An echocardiogram showed that wall thickness of the left ventricle was hypertrophic with normal systolic function at 1 and 2 weeks after TAC and that left ventricular dimensions were enlarged at systole and cardiac systolic function was impaired at 4 weeks in vehicle-treated group (Fig. 1). On the other hand, Omega-3 EE inhibited the dilatation of cardiac dimension at systole and improved cardiac function at 4 weeks after operation (Fig. 1). Omega-3 EE also attenuated an increase in the heart weight (HW) to body weight (BW) ratio at 4 weeks after TAC (Fig. 2a). Tissue examinations revealed that Omega-3 EE suppressed pressure overload-induced cardiomyocyte hypertrophy and cardiac fibrosis (Fig. 2b,c). These results indicate that Omega-3 EE protects cardiac remodeling under pathological stresses.

Omega-3 FA supplementation changes the proportion of FAs under pathological condition.
To know the mechanisms of how Omega-3 EE protects the heart from pressure overload, we first examined concentration of FAs before, and 1 and 2 weeks after TAC operation when there was no difference in left ventricular wall thickness and cardiac systolic function between vehicle and Omega-3 EE groups (Fig. 1). We used gas chromatography to evaluate compositional change of each FA in blood and heart 31,32 . Plasma levels of almost all FAs were not affected by pressure overload in vehicle group (Fig. 3a). On the other hand, all FAs were significantly decreased in the heart according to time after pressure overload in vehicle group (Fig. 3b).
In blood, there were less ARA and more EPA in Omega-3 EE group as compared with vehicle group (Fig. 3a). In the heart, concentrations of both ARA and DHA were not different between vehicle and Omega-3 EE groups under stress condition, while concentrations of EPA were higher in Omega-3 EE group than those in vehicle group (Fig. 3b). These results indicate that oral supplementation of Omega-3 EE changes the proportion of FAs and in particular increases EPA concentration in the pressure overloaded heart.  www.nature.com/scientificreports/ Omega-3 FA treatment augments the increase in a free form EPA in the heart exposed with high blood pressure. Metabolites of both omega-6 FA and omega-3 FA have been reported to be involved in regulating inflammation [13][14][15][16][17] . We thus evaluated the concentration of various metabolites during the development of cardiac hypertrophy by using liquid chromatography-tandem mass spectrometry (LC-MS) method 33,34 . The levels of free forms of ARA, EPA and DHA were much lower compared with those in total ARA, EPA and DHA, respectively (Figs. 3b and 4). Although total amounts of ARA and DHA in the heart were decreased after pressure overload (Fig. 3b), free forms of ARA and DHA were increased in the heart at 2 weeks after pressure overload in vehicle group (Fig. 4a,b). Some of ARA-related metabolites such as prostaglandin (PG) E 2 , PGD 2 and PGF 2α were increased at 2 weeks after TAC operation (Fig. 4a). We next examined whether oral treatment of omega-3 FA had an impact on concentration of free form and its metabolites of omega-3 FA and omega-6 FA in the heart. Under the condition without any pathological stresses, Omega-3 EE treatment did not show any effects on the concentration of free form of both omega-6 FA and omega-3 FAs (Fig. 4). Although free form ARA was increased in vehicle-treated group at 2 weeks after TAC operation, it was not increased in Omega-3 EE group (Fig. 4a). Free form DHA was increased under stress similarly both in vehicle and Omega-3 EE groups (Fig. 4b), while free form EPA was increased by pressure overload only in Omega-3 EE group (Fig. 4c).
There was no difference in any metabolites in the heart derived from omega-6 FAs and omega-3 FAs without pressure overload between vehicle and Omega3 EE groups (Fig. 4). Omega-3 EE treatment suppressed the concentrations of PGD 2 , an ARA-metabolite and 5-HEPE, an EPA-metabolite at 2 weeks (Fig. 4a,c), but the treatment did not have any effects on the concentration of DHA-derived metabolites (Fig. 4b).
Since ROS has been reported to be involved in heart failure 6 and regulate the metabolites of omega-3 FA and omega-6 FA 21,35,36 , we examined whether ROS is involved in this process. We first evaluated malondialdehyde (MDA) concentration, a product of lipid peroxidation in the heart using the thiobarbituric acid reactive substance (TBARS) assay 37 . The concentration of MDA in the heart was not changed during the development of cardiac hypertrophy. Furthermore, there was no difference in the MDA concentration between vehicle and Omega-3 EE groups (Fig. 4d). We also evaluated iso-PGF2α, an isoprostane produced by the non-enzymatic www.nature.com/scientificreports/ peroxidation of ARA by LC-MS method 21,36 , but the level was below detection limit (Table 1). So, in this study we did not obtain any evidence suggesting that ROS modulates the metabolites of omega-3 FA and omega-6 FA in the development of cardiac hypertrophy.
Omega-3 FA regulates inflammation in the development of cardiac hypertrophy. We finally examined whether oral administration of Omega-3 EE had some effects on inflammation in the heart. Many www.nature.com/scientificreports/ inflammatory cells such as CD45-positive leukocytes and F4/80-postive macrophages were still observed in hearts treated with vehicle at 2 weeks after operation, while the number of inflammatory cells went back to basal levels in the hearts treated with Omega-3 EE (Fig. 5a,b). Macrophages can be classified into several types; M1 macrophage is a proinflammatory phase macrophage and M2 is a resolution phase macrophage 38,39 . We examined the polarity of macrophages at 1 week after exposure to pressure overload when there was no difference in the number of F4/80-postive macrophages between groups (Fig. 5b). We used anti-iNOS and anti-CD163 antibodies to evaluate M1 and M2 macrophages, respectively 40 . There was no difference in the number of iNOS-positive cells and CD163-positive cells between vehicle and Omega-3 EE groups at 1 week after TAC (Fig. 5c,d), but the ratio of anti-CD163 positive cell numbers to anti-iNOS positive cell numbers was higher in Omega-3 EE treated   www.nature.com/scientificreports/ group than that in vehicle group (Fig. 5e), indicating that omega-3 FA modulates the balance between M1 and M2 macrophages. We also examined expression levels of proinflammatory cytokines in the heart and their concentrations in blood. Omega-3 EE attenuated the induction of inflammatory cytokines such as TNFα and MIP-1α expressions in the heart at 1 week (Fig. 5f). Both TNFα and MIP-1α concentrations in blood were not increased in 1 and 2 weeks after operation (Fig. 5g). These results suggest that the effects on the heart might be direct actions on the heart of omega-3 FA rather than indirect actions via systemic changes of metabolism.

Discussion
Hearts use a lot of FAs and produce adenosine triphosphate (ATP) to sustain beating and ion exchange. Decreases of ATP have been reported to result in the development of heart failure 41 , but it remains unclear whether FAs were changed in the heart during the process of cardiac remodeling. This study showed that almost all FAs in the heart were decreased from 1 week after exposure to high blood pressure when cardiac systolic function was still preserved (Figs. 1 and 3b). There are three possibilities for the decrease of FAs in the heart; the decrease in influx of FAs into the heart, the increase in consumption, and the dilution of total FAs. Since it has been reported that FA uptake into the hypertrophied heart was inhibited by various factors such as the increase of ketone bodies and the decrease of FA transporters' expressions 42,43 , the impairment of FA uptake might be a reason for the decrease of FAs in the heart. On the other hand, the second reason may be not the case, because there are reports indicating that the consumption of FAs is decreased in the pathological heart due to the impairment of mitochondrial function 7,44,45 . During the development of cardiac hypertrophy, protein synthesis is robustly enhanced and the cellular edema is also induced, the concentration of FAs might be relatively decreased.
There are some studies to examine the molecular mechanisms of how omega-3 FA preserves cardiac function under various stresses. Especially, Stanley WC and his colleagues demonstrated the role of omega-3 FA 7,45-47 . They showed that fish oil including much omega-3 FA increased the proportion of omega-3 FAs such as DHA and EPA and decreased that of ARA in mitochondrial phospholipids in the heart 7,45 and suppressed the mitochondrial permeability transition pore opening 45 , which led to inhibition of cardiac dilatation under the pressure overload. We examined composition of total cardiac FAs at shorter time points after exposure to stresses. The change of total cardiac FA composition in our study (Fig. 3b) was almost similar to that in their studies, suggesting that Omega-3 EE also ameliorates cardiac dysfunction through the preservation of mitochondrial function.
Inflammation is a key factor in the development of cardiac hypertrophy and dysfunction induced by high blood pressure [7][8][9]48 . Inflammation has two processes, initiation and resolution, both of which are active reactions 39,49 . Omega-6 FA-and omega-3 FA-derived metabolites have been reported to be critically involved in both processes [13][14][15][16][17]50 , and the imbalance of these metabolites causes amplification and long-term continuation of inflammation, leading to tissue damage 49 . Accelerating resolution of inflammation contributes to maintaining organ homeostasis. Omega-3 FA's metabolites have been reported to promote resolution [14][15][16][17] and protect the heart from myocardial infarction and ischemia-reperfusion injury 51 . Fat1 transgenic mice, which have an enzyme of C. elegans to convert omega-6 FAs to omega-3 FAs, preserved cardiac function under pressure overload via production of 18-HEPE, an EPA derivative 20 . These results suggest that promoting resolution by omega-3 FA metabolites has a therapeutic potential for cardiac diseases. In this study, we clearly showed that the oral administration of omega-3 FA prevented the development of heart failure by suppression of inflammation ( Figs. 1 and 5).
In addition to the increase in free form EPA, it is an important point that Omega-3 EE had no effect on concentration of free form ARA and its metabolites at 1 week after imposing pressure overload (Fig. 4a). Until now, many drugs have been designed to inhibit the initial phase of inflammation when ARA-derived metabolites such as PGs and leukotrienes play crucial roles, and clinical trials and animal studies were performed to use these inhibitors to treat inflammation-related diseases. Recent clinical studies, however, revealed that a selective inhibitor of COX2, which is a key enzyme to convert a free form ARA to various types of lipid mediators, increased the risk of cardiovascular events 52,53 . Furthermore, deletion of PLA2, which is an enzyme to release omega-3 and omega-6 FAs from cellular membrane phospholipids, deteriorated ischemia-reperfusion injury in mice 54 . These results suggest that acute inflammatory responses are protective for hearts from various stresses.
There are some reports showing the role of PGD 2 in both cardiomyocytes and macrophages in the ischemic heart. In cardiomyocytes, because of a low expression of PGD 2 receptors (DP1 and DP2), PGD 2 bound to PGF 2α receptor, and then showed a protective effect against ischemia-reperfusion injury via anti-oxidative effects 55 . In macrophages, the PGD 2 / DP1 axis changed the polarity of macrophage to M2 macrophage, resulted in accelerated resolution of inflammation 56 . In this study, PGD 2 concentration in the heart was higher in vehicle group than in Omega-3 EE group at 2 weeks after operation, while PGF 2α concentration was almost same between two groups (Fig. 4a), suggesting that PGD 2 and PGF 2α might be not involved in cardiac remodeling in this study.
Regarding the omega-3 FA metabolites, Omega-3 EE decreased 5-HEPE concentration in the heart at 2 weeks after surgery (Fig. 4c). A report showed that 5-HEPE attenuated the expression of some proinflammatory cytokines induced by palmitic acid in macrophage 57 , but another showed that 5-HEPE did not reduce IL-6 production induced by macrophage-conditioned media in cardiac fibroblast 20 . We need further studies to examine the meaning of the reduction of 5-HEPE by Omega-3 EE during the process of cardiac hypertrophy.
Taken together, this study demonstrates that omega-3 resolves the inflammation at later stage but does not inhibit the initiation of inflammation and suggests that omega-3 might be an ideal drug to treat heart failure where inflammation is critically involved.

Materials and methods
Mouse experiments. All animal experiments were approved by the Ethics Committee for Animal Experiments of the University of Tokyo, and the procedures were adhered strictly to the guidelines for animal experiments of the University of Tokyo. Ten-week old male C58BL/6J mice were purchased from CLEA Japan Inc. (Tokyo, Japan) and housed under controlled temperature with a 12 h light/ dark cycle and provided with standard food and water ad libitum. Mice were exposed to isoflurane and then were subjected to pressure overload by ligating aortic arch to induce cardiac hypertrophy 27 . Mice were treated with Omega-3 EE 1.5 mg/g of body weight 30 via feeding needle every day from 5 days before operation to the end of study. Just before administration, a proper dose of Omega-3 EE was diluted to 100 µL with 0.5% methylcellulose to keep their stability. 0.5% methylcellulose was used as a vehicle. Omega-3 EE is a clinical drug containing both EPA and DHA, obtained from BASF AS (Norway) and Takeda Pharmaceutical Company (Japan) 29 . To check cardiac function and morphology, echocardiogram was performed without anesthesia using the Vevo2100 ultrasound system (FUJIFILM VisualSonics, Japan) 27 . We obtained hearts at 0, 1, 2 and 4 weeks after operation.
Morphology. Hearts were fixed with 20% formalin (Sakura Finetek Japan Inc.) and embedded in paraffin.
After cutting 4 µm each, paraffin sections were stained with anti-CD45 (BD Biosciences, USA), anti-F4/80 antibodies (Bio-Rad, USA), anti-iNOS antibody (abcam, UK), and anti-CD163 antibody (abcam) to evaluate cell infiltration. Elastica van Gieson staining and wheat germ agglutinin staining were performed to examine fibrosis proportion and calculate cell area, respectively. Cross-sectional area (CSA) of myocytes with a centrally located nucleus and overall circular shape was measured in the left ventricular free walls using Image J (NIH, USA) 27 .
Fatty acid composition in plasma and heart. Samples for FA levels in the plasma and heart were prepared and analyzed by a modification of the one-step reaction analysis using gas chromatography 31,32 . For analysis of the FA composition in plasma and heart, 0.005% butylated hydroxytoluene /methanol and tricosanoic acid as internal standard, were added to samples and then kept at − 30 °C. The samples were heated at 98 °C for 1 h after addition of acetyl chloride. After the sequential addition of 0.5 M sodium hydroxide /10% sodium chloride and octane, samples were shaken for 3 min, centrifuged at 950×g for 10 min at 20 °C and the top layer was collected. The FA composition was measured using the Agilent 6850 A gas chromatograph (Agilent Technologies, USA) equipped with a flame ionization detector and an automatic sampler. Each FA methly esters were determined compared with retention time of standard materials. In the present study, we have determined FA levels in total FAs but not in the phospholipid 58 . Fatty Acid Metabolites in heart 33,34 . Sample preparation. Heart tissues were homogenized with 70% methanol and centrifuged at 5000×g for 10 min at 4 °C. The supernatants were diluted with distilled water. PGE 2 -d4, PGD 2 -d4, PGF 2α -d4, 5-HETE-d8, and ARA-d8 were added to each sample as internal standards. After adjusting the PH of samples to 4.0 using 0.1 M HCl, samples were applied to preconditioned solid-phase extraction cartridges (Sep-Pak C18, Waters, USA). Sep-Pak cartridges were washed with water and n-hexane. Finally, methyl formate was applied to elute FA metabolites. The precise method was described as previously 33,34 .

LC-ESI-MS-MS-Based Analysis.
High-performance liquid chromatography (HPLC) was combined with electrospray ionization-mass spectrometry (ESI-MS). HPLC was performed using a Luna 3u C18(2) 100 Å LC column (100 × 2.0 mm, Phenomenex, USA). Samples were eluted in a mobile phase comprising acetonitrilemethanol and water-acetic acid in a 27:73 ratio for 5 min, ramped up to a 100:0 ratio after 25 min, and held for 10 min. Tandem mass spectrometry (MS-MS) analyses were conducted in negative ion mode. FA metabolites were detected and quantified by selected reaction monitoring (SRM as listed in Table 2). Each area was calculated using the Xcalibur 2.1 software (Thermo Fisher Scientific, USA). The precise method was described as previously 33,34 .