Docosahexaenoic acid, but not eicosapentaenoic acid, improves septic shock-induced arterial dysfunction in rats

Introduction Long chain n-3 fatty acid supplementation may modulate septic shock-induced host response to pathogen-induced sepsis. The composition of lipid emulsions for parenteral nutrition however remains a real challenge in intensive care, depending on their fatty acid content. Because they have not been assessed yet, we aimed at determining the respective effects of eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA) during septic shock-induced vascular dysfunction. Methods In a peritonitis-induced septic shock model, rats were infused with EPA, DHA, an EPA/DHA mixture or 5% dextrose (D5) during 22 hours. From H18, rats were resuscitated and monitored during 4 hours. At H22, plasma, aorta and mesenteric resistance arteries were collected to perform ex vivo experiments. Results We have shown that septic rats needed an active resuscitation with fluid challenge and norepinephrine treatment, while SHAM rats did not. In septic rats, norepinephrine requirements were significantly decreased in DHA and EPA/DHA groups (10.6±12.0 and 3.7±8.0 μg/kg/min respectively versus 17.4±19.3 μg/kg/min in D5 group, p<0.05) and DHA infusion significantly improved contractile response to phenylephrine through nitric oxide pathway inhibition. DHA moreover significantly reduced vascular oxidative stress and nitric oxide production, phosphorylated IκB expression and vasodilative prostaglandin production. DHA also significantly decreased polyunsaturated fatty acid pro-inflammatory mediators and significantly increased several anti-inflammatory metabolites. Conclusions DHA infusion in septic rats improved hemodynamic dysfunction through decreased vascular oxidative stress and inflammation, while EPA infusion did not have beneficial effects.


Introduction
Long chain n-3 fatty acid supplementation may modulate septic shock-induced host response to pathogen-induced sepsis. The composition of lipid emulsions for parenteral nutrition however remains a real challenge in intensive care, depending on their fatty acid content. Because they have not been assessed yet, we aimed at determining the respective effects of eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA) during septic shock-induced vascular dysfunction.

Methods
In a peritonitis-induced septic shock model, rats were infused with EPA, DHA, an EPA/DHA mixture or 5% dextrose (D5) during 22 hours. From H18, rats were resuscitated and monitored during 4 hours. At H22, plasma, aorta and mesenteric resistance arteries were collected to perform ex vivo experiments.

Results
We have shown that septic rats needed an active resuscitation with fluid challenge and norepinephrine treatment, while SHAM rats did not. In septic rats, norepinephrine requirements were significantly decreased in DHA and EPA/DHA groups (10.6±12.0 and 3.7±8.0 μg/kg/ min respectively versus 17.4±19.3 μg/kg/min in D5 group, p<0.05) and DHA infusion significantly improved contractile response to phenylephrine through nitric oxide pathway inhibition. DHA moreover significantly reduced vascular oxidative stress and nitric oxide production, phosphorylated IκB expression and vasodilative prostaglandin production. DHA also significantly decreased polyunsaturated fatty acid pro-inflammatory mediators and significantly increased several anti-inflammatory metabolites. PLOS

Introduction
Over the past decades, the composition of lipid emulsions for parenteral nutrition has emerged as a real challenge in intensive care. Indeed, lipid emulsions may differently interfere with inflammation and immune pathways, depending on their fatty acid content [1][2][3]. N-3 polyunsaturated fatty acid (PUFA) supplementation, including eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA), was thus proposed as a potential beneficial adjunctive therapy [4,5]. Yet, in critically ill patients, lipid emulsions for parenteral nutrition containing n-3 PUFA are used daily in intensive care units (ICUs) if enteral nutrition is contraindicated [1]. Septic shock is one of the first causes of ICU mortality, and is characterized by severe hypotension with a profound vasodilation and vascular hyporesponsiveness to vasoconstrictors [6,7]. Although n-3 PUFA supplementation is considered as a promising therapeutic in several inflammatory diseases [4,8,9], including septic shock patients [1], the respective effects of EPA and DHA in vascular dysfunction of septic shock have never been established yet. Whether only their combination or both of their individual effects are beneficial has not been determined either. Marine n-3 PUFAs may be beneficial in septic shock patients by decreasing the deregulated hyper-inflammatory response. Indeed, n-3 PUFAs have anti-inflammatory effects by modulating cell membrane and lipid raft compositions, thus notably interfering with innate and adaptive immune cell differentiation and functions [10][11][12]. They are also responsible for the inhibition of inflammatory gene expression, the activation of anti-inflammatory transcription factor, and the modulation of transcription by binding to peroxisome proliferator-activated receptors (PPARs) [1]. Their effective potential benefits however remain contradictory and largely unexplained. Thus, while Heller et al. [13] showed a significant mortality reduction in critically ill patients treated with parenteral fish oil emulsion, there was no benefit in septic patients [14]. Both Barbosa et al. [15] and Gultekin et al. [16] supported these results in smaller cohorts of sepsis or septic shock patients.
EPA and DHA could have synergistic effects, as assessed by recent data. Zhang et al. [17] thus showed that EPA and/or DHA supplementation reduced the inflammatory response in a rat ischemia-reperfusion model and this effect is more important when EPA and DHA are combined. Interestingly, Allaire et al. [18] showed that DHA is more effective than EPA in modulating inflammation markers and blood lipids in humans at risk of cardiovascular disease.
The purpose of our experimental study was therefore to determine in an original way the dissociated and combined effects of the EPA and DHA parenteral administration in a rat peritonitis-induced septic shock model and assess the mechanisms involved in their hemodynamic and vascular effects.

Material and methods
This study was performed with the approval of the Strasbourg Regional Committee of Ethics in Animal Experimentation (CREMEAS, AL/69/76/02/13).

Septic shock model and infusion with n-3 PUFAs
Rats underwent cecal ligation and puncture (CLP) as previously described [19]. During surgical procedures, rats were anesthetized with isoflurane 1-2% (Baxter S.A.S, Maurepas, France) and analgesia was performed with subcutaneous buprenorphine (Sogeval, Laval, Fance; 0.1 mg/kg of body weight;). A subcutaneous injection of 0.1 mL lidocaine 1% (AstraZeneca, Rueil-Malmaison, France) was performed before skin incision. Under aseptic conditions, a 3-cm midline laparotomy was performed to allow exposure, ligation and puncture of the caecum with a 21-gauge needle. A small amount of feces was extruded, the caecum returned into the peritoneal cavity and the laparotomy closed. SHAM rats underwent a midline laparotomy and cecal exposure without further manipulation. All rats received a subcutaneous injection of 0.9% NaCl after the surgical procedure (30 ml/kg of body weight).
The right femoral vein was then catheterized, and the catheter was tunneled between the ears. Purified EPA and DHA (> 95% purity) from Pivotal Therapeutics, Inc. (Woodbridge, ON, Canada) were emulsified in 10% DMSO (Sigma Aldrich, Saint-Quentin Fallavier, France) (EPA:DHA 1:1). The randomly allocated 2% emulsion (EPA or DHA or EPA:DHA 1:1) was blindly infused at identical rates during 22 hours (0.15 g/kg/day as recommended in humans). The control groups (SHAM-D5 and CLP-D5) were infused with 5% dextrose (D5) at identical rates (S1 Fig). It has to be noted that all the experiments have also been performed using only DMSO as a control at the same concentration as the lipid mediator. DMSO did not affect any parameter considered.

Hemodynamic monitoring and resuscitation
The CLP-rats developed septic shock within the 16-20 hours after CLP. Septic shock was considered established on the basis of clinical criteria (lethargy, piloerection and glassy eyes) and confirmed if MAP was below 90 mmHg and plasma lactate level elevated (> 2 mmol/L). After 18 hours, rats were anesthetized with an intraperitoneal injection of pentobarbital sodium (60 mg/kg), tracheotomised and mechanically ventilated (Series Small animal Ventilator, SAR-830/P; CWE, Ardmore, PA). Analgesia was ensured as described above.
The left femoral artery was cannulated and used to measure mean arterial pressure (MAP) and heart rate (HR), and collect blood samples. The homolateral femoral vein was cannulated for fluid maintenance and drug infusion. In septic rats, fluid resuscitation was performed by bolus of 0.9% NaCl (500 μL/10 min. when needed) and norepinephrine infused to target a MAP above 90 mmHg. DHA, EPA, EPA/DHA or dextrose infusion was continued at the same rate during the 4-hour resuscitation.
After a 4-hour resuscitation (S1 Fig), blood samples were collected and rats were sacrificed by intravenous pentothal injection (100 mg/Kg). The presence of caecum necrosis and intraabdominal pus in CLP-rats and their absence in SHAM rats were systematically checked. Aorta was collected for electron paramagnetic resonance measurements. Mesenteric resistance arteries (MRAs) were collected for vascular reactivity study and Western blotting analysis.

Electron paramagnetic resonance (EPR) analysis
Values were expressed in signal amplitude/mg weight of dried tissue (Amplitude/Wd).

Quantification of pro-resolving EPA and DHA metabolites
To 100 μL of plasma 300 μL of cold methanol and 5 μL of internal standard (Deuterium labeled compounds) were added. After centrifugation at 900 g for 15 min at 4˚C, supernatants were transferred into 2 mL 96-well deep plates and diluted in H 2 O to 2 mL. Samples were then submitted to solid phase extraction (SPE) using OASIS HLB 96-well plate (30 mg/well, Waters) pretreated with MeOH (1mL) and equilibrated with 10% MeOH (1 mL). After sample application, extraction plate was washed with 10% MeOH (1 mL). After drying under aspiration, lipids mediators were eluted with 1 mL of MeOH. Prior to LC-MS/MS analysis, samples were evaporated under nitrogen gas and reconstituted in 10 μL on MeOH. LC-MS/MS analyses of eicosanoids were performed as described [20]. Briefly, lipid mediators were separated on a ZorBAX SB-C18 column (2.1 mm, 50 mm, 1.8 μm) (Agilent Technologies) using Agilent 1290 Infinity HPLC system (Technologies) coupled to an ESI-triple quadruple G6460 mass spectrometer (Agilent Technologies). Data were acquired in Multiple Reaction Monitoring (MRM) mode with optimized conditions (ion optics and collision energy). Peak detection, integration and quantitative analysis were done using Mass Hunter Quantitative analysis software (Agilent Technologies) based on calibration lines built with commercially available eicosanoids standards (Cayman Chemicals).

Quantification of thromboxane A 2 and prostacyclin I 2
Thromboxane B 2 (TXB 2 ), derived from rapid non-enzymatic hydrolysis of TXA 2 , and prostacyclin I 2 , which is rapidly non-enzymatically hydrated to a stable product, 6-keto PGF1α, were measured in plasma using the respective immunoassay kit (ELISA) according to manufacturer's instructions (Cayman Chemical Company, Ann Arbor, MI, USA).

Statistical analysis
Data obtained during measurements of arterial pressure and heart rate were compared using a nonparametric Wilcoxon signed rank test. Two-way analysis of variance for repeated measurements was used for comparison of vascular reactivity (data were tested for homogeneity of variance by Levene's statistics), and nonparametric Kruskal-Wallis test or Mann-Whitney test were used for comparison Western blotting, NO, superoxide, immunostaining signal measurements and metabolite quantification between the four groups. When a significant difference was found between groups, subsequent post hoc tests were performed. All these tests were performed with the Statview version 5.0 software (SAS Institute, Cary, NC). P < 0.05 was considered statistically significant. The effects of Ach were quantified as the percentage of relaxation in preparations previously contracted with Phe. All values are presented as mean ± SD for n experiments. n represents the number of animals.

DHA infusion improved septic shock-induced arterial dysfunction
Ex vivo MRAs of CLP-D5 rats showed a significant hyporeactivity to phenylephrine (Table 1). Both DHA and EPA/DHA, but not EPA alone, improved ex vivo contractile response to phenylephrine compared to CLP-D5 group (p<0.05) (Fig 1A). Nitric oxide synthase and COXs pharmacological inhibition significantly restored the vascular reactivity to phenylephrine in CLP-D5 rats; without inhibitors, DHA and EPA/DHA, but not EPA alone, exerted the same effect in term of contractility levels (Table 1).
Compared to basal contraction, NO synthase inhibition did not significantly enhance the contractile response in treated septic rats, whatever the group considered (Table 1). COX inhibition alone, as well as the association of NOS and COX inhibition, significantly enhanced contractile response in CLP-EPA rats, but not in in CLP-EPA/DHA rats ( Table 1).
The sensitivity to acetylcholine was significantly decreased in CLP group compared to SHAM rats (half maximal effective concentration EC 50 : 115 nM vs. 20 nM respectively, p<0.05). The treatment by EPA-DHA (EC 50 : 91nM) but not EPA or DHA alone (EC 50 : 91 nM and 97 nM respectively) significantly affected the MRAs from CLP group (p<0.05) (Fig 1B).
DHA infusion reduced septic shock-induced vascular inflammation, oxidative stress and nitric oxide production NF-κB activation was assessed by IκB phosphorylation, resulting in free NF-κB nuclear translocation and pro-inflammatory gene transcription. In the MRAs, NF-κB p65 expression was significantly decreased in CLP-EPA/DHA group compared to CLP-D5 group (Fig 2A). In MRAs, pIκB expression was significantly increased in CLP-D5 and CLP-EPA rats compared to SHAM-D5 rats, and decreased in CLP-DHA and CLP-EPA/DHA rats compared to CLP-D5 rats (Fig 2B).
Inducible NOS expression was significantly increased in MRAs from septic rats compared to SHAM rats and decreased in MRAs from CLP-DHA and CLP-EPA/DHA groups, compared to CLP-D5 and CLP-EPA groups (Fig 2C). Cyclo-oxygenase-1 expression did not significantly differ between the groups (not shown). Compared to SHAM rats, COX-2 expression was significantly increased in MRAs from septic rats and decreased in MRAs from CLP-DHA and CLP-EPA/DHA groups, but not in CLP-EPA group (Fig 2D). Compared to SHAM group, NO and O 2 productions were significantly increased in the aorta from CLP-D5 rats. DHA and EPA/DHA infusions significantly decreased this production. Nitric oxide and O 2 productions were not significantly different between CLP-D5 and CLP-EPA (Fig 3A and 3B). In MRAs, HO-1 expression was significantly increased in MRAs from CLP-DHA and CLP-EPA/  DHA rats, compared to MRAs from CLP-D5 and SHAM-D5 rats, and not significantly different between SHAM-D5, CLP-D5 and CLP-EPA groups (Fig 3C).

DHA infusion reduced septic shock-induced vasodilative prostacyclin I 2 production
Thromboxane B 2 level was significantly increased in CLP groups compared to SHAM-D5 group, and decreased in CLP-EPA group compared to CLP-D5 group (Fig 4A).

Discussion
Although growing evidence suggests that marine n−3 PUFAs may be beneficial in septic shock-induced hyper-inflammatory state, their effective potential benefits remain nonetheless controversial [1,21]. Part of the contradictory results from recent literature may arise from the  heterogeneity of the n-3 PUFAs compared. We have thus shown that n-3 PUFAs have differential effects, depending on their nature. For the first time to our knowledge, we describe the respective effects of EPA and DHA supplementations on vascular arterial dysfunction in a septic shock model in rats and the mechanisms involved in these vascular effects. We indeed bring evidence that DHA supplementation decreased vascular inflammation and oxidative stress, improving hemodynamic and arterial dysfunction, while EPA did not. N-3 PUFA supplementation was shown to decrease organ dysfunction in septic shock patients, ICU and hospital lengths of stay [15,22]. In the present study, we have shown that DHA significantly reduced norepinephrine needs to maintain the MAP target in septic rats, while EPA did not. Barton et al. had already shown that dietary n-3 PUFAs decreased mortality in a rat model of sepsis, but they neither investigated the hemodynamic function, nor the respective effects of EPA and DHA [23]. Consistent with our results, some experimental data suggest that n-3 PUFA supplementation might improve septic shock-induced organ dysfunctions. Indeed, the septic cardiac dysfunction was reduced in rats supplemented with parenteral n-3 PUFAs [24], like in animal models of ischemia-reperfusion induced shock [25,26]. Interestingly, Tunctan et al. [27] have demonstrated that a stable synthetic analog of 20-hydroxyeicosatetraenoic acid (20-HETE), derived from arachidonic acid, prevents vascular hyporeactivity, hypotension and inflammation in endotoxemic rats. Moreover, EPA and DHA may have several effects, notably through the production of different anti-inflammatory mediators, including the E-and D-series resolvins or protectins, depending on the metabolic pathway involved [28]. Interestingly, EPA and DHA could have synergistic effects on inflammation in a model of intestinal ischemia-reperfusion, compared to EPA or DHA alone [17]. In this line, we have shown that only DHA but not EPA could have the same beneficial effect in septic shock-induced vascular dysfunction and these beneficial hemodynamic effects are likely to be in part due to differences in derived vasoactive metabolites after DHA or EPA or EPA+DHA infusion. Indeed, herein DHA infusion significantly decreased the production of vasodilative 6kPGF1 in septic rats and was also associated to the production of some anti-inflammatory D-series resolvins, resulting in the hemodynamic and vascular beneficial effects, through decreased sepsis-induced vasodilation. Consistent with these data, we have shown that supplementing septic rats with DHA or EPA+DHA, but not EPA alone, significantly improved the contractile response and endothelial function of resistance arteries, through nitric oxide pathway inhibition. The result of COX inhibition on isolated arteries permitting to restore the contractile response highlighted the role of derived vasoactive metabolites when septic rats were treated by EPA. Thus, several reports showed that DHA and EPA might have different vascular effects in hypertensive rats and hyperlipidemic patients. Indeed, DHA enhanced vasodilator mechanisms and attenuated constrictor responses, while EPA did not [29]. Thus, it was suggested that DHA alters endothelial function, involving different mechanisms depending on the pathology considered [29]. Moreover, DHA was already reported to exert antioxidative activity, decreasing reactive oxygen species production and NO synthase activity [30][31][32]. Consistent with these data, we have shown that arterial NO production and oxidative stress were both decreased in arteries from septic rats supplemented with DHA. HO-1 expression is induced by both inflammatory cytokines and oxidative stress [33] during sepsis and other inflammatory diseases. It was shown to improve resistance to oxidative stress and modulate inflammatory response [34][35][36], through anti-inflammatory and antioxidant properties [37]. In this context, we have shown that HO-1 expression is increased in mesenteric arteries from septic rats treated with DHA and EPA/DHA and may contribute to their anti-inflammatory and antioxidant properties. Furthermore, DHA supplementation decreased inflammation via NF-κB down regulation septic rat arteries and these results are consistent with the literature, which has shown that n-3 PUFAs down regulate NF-κB pathway, inflammatory cytokine and adhesion molecule production, and COX-2 gene transcription [2,38].
Finally, we have shown that EPA supplementation increased the production of vasodilative arachidonic acid-derived prostacyclin I2 in the plasma, which mainly depends on COX-2 activity in the vascular wall [39], consistent with vascular COX-2 increase in this group. Interestingly, the plasma concentration of its biological opposite, thromboxane A 2 , mainly depending on COX-1 and thromboxane synthase, remained unaltered by n-3 PUFAs. The balance between PGI 2 and TXA 2 is notably responsible for vascular tone regulation, with PGI 2 being able to limit the response to TXA 2 in vivo. PGI 2 /TXA 2 ratio might be more informative than the absolute level of these prostanoids [39]. Interestingly, the 6-keto PGF1α to TXB 2 ratio shows a major imbalance with an excessive vasodilative product formation in rats treated with EPA. Mayer et al. [40] demonstrated this beneficial cytokine-shifting effect in septic patients, but once again the EPA and DHA effects were not analyzed separately.
The fact that NO and superoxide was assessed in aorta instead of mesenteric arteries constitutes a limitation of the study. Although both vascular beds depend on different regulatory mechanisms, we have previously shown that sepsis induced the same kind of hyporeactivity to Phe in aorta and small mesenteric arteries [6,41]. Finally, septic shock is a systemic syndrome, in which vascular inflammation, oxidative and nitrosative stresses are ubiquitous, as we showed in our previous studies [42][43][44][45][46].

Conclusion
The beneficial effects of n-3 PUFAs may be mainly attributable to docosahexaenoic acid, improving vascular dysfunction by decreasing inflammatory mediator production and increasing anti-inflammatory metabolites, and by decreasing oxidative stress and NO production in a rat septic shock model. Thus, purified DHA supplementation might be a promising adjunctive immunonutrition therapy during septic shock, which would deserve further investigations in humans.
Supporting information S1  Table. Plasma levels of thromboxane B2 and 6-keto PGF1α. (PDF) S1 Fig. Protocol design. At H0, rats were randomly allocated to a group and underwent a surgical procedure (either cecal ligation and puncture, CLP, or SHAM) and were infused until H22 by 5% dextrose (D5), purified EPA, DHA or a mixture of EPA and DHA at identical rates. From H18 to H22, rats were anesthetized, ventilated and resuscitated with fluid challenge and norepinephrine to reach the mean arterial pressure (MAP) target over 90 mmHg. At H22, rats were sacrificed. n = 10 rats/group. (TIF)