N-Docosahexaenoyl Dopamine, an Endocannabinoid-like Conjugate

Nov 21, 2016 - Several studies indicate that the n-3 long-chain polyunsaturated fatty acid docosahexaenoic acid (DHA) contributes to an attenuated ...
0 downloads 0 Views 3MB Size
Research Article pubs.acs.org/chemneuro

N‑Docosahexaenoyl Dopamine, an Endocannabinoid-like Conjugate of Dopamine and the n‑3 Fatty Acid Docosahexaenoic Acid, Attenuates Lipopolysaccharide-Induced Activation of Microglia and Macrophages via COX‑2 Ya Wang,†,‡,⊥ Pierluigi Plastina,†,§,⊥ Jean-Paul Vincken,‡ Renate Jansen,† Michiel Balvers,† Jean Paul ten Klooster,∥ Harry Gruppen,‡ Renger Witkamp,† and Jocelijn Meijerink*,† †

Division of Human Nutrition and ‡The Laboratory of Food Chemistry, Wageningen University, 6700 AA Wageningen, The Netherlands § Department of Chemistry and Chemical Technologies, University of Calabria, 87036 Cosenza, Italy ∥ Research Centre Technology & Innovation, Innovative Testing in Life Sciences and Chemistry, University of Applied Sciences, 3584 CH Utrecht, The Netherlands S Supporting Information *

ABSTRACT: Several studies indicate that the n-3 long-chain polyunsaturated fatty acid docosahexaenoic acid (DHA) contributes to an attenuated inflammatory status in the development of neurodegenerative disorders, such as Alzheimer’s and Parkinson’s disease. To explain these effects, different mechanisms are being proposed, including those involving endocannabinoids and related signaling molecules. Many of these compounds belong to the fatty acid amides, conjugates of fatty acids with biogenic amines. Conjugates of DHA with ethanolamine or serotonin have previously been shown to possess anti-inflammatory and potentially neuroprotective properties. Here, we synthesized another amine conjugate of DHA, N-docosahexaenoyl dopamine (DHDA), and tested its immune-modulatory properties in both RAW 264.7 macrophages and BV-2 microglial cells. N-Docosahexaenoyl dopamine significantly suppressed the production of nitric oxide (NO), the cytokine interleukin-6 (IL-6), and the chemokines macrophage-inflammatory protein-3α (CCL20) and monocyte chemoattractant protein-1 (MCP-1), whereas its parent compounds, dopamine and DHA, were ineffective. Further exploration of potential effects of DHDA on key inflammatory mediators revealed that cyclooxygenase-2 (COX-2) mRNA level and production of prostaglandin E2 (PGE2) were concentration-dependently inhibited in macrophages. In activated BV-2 cells, PGE2 production was also reduced, without changes in COX-2 mRNA levels. In addition, DHDA did not affect NF-kB activity in a reporter cell line. Finally, the immune-modulatory activities of DHDA were compared with those of N-arachidonoyl dopamine (NADA) and similar potencies were found in both cell types. Taken together, our data suggest that DHDA, a potentially endogenous endocannabinoid, may be an additional member of the group of immune-modulating n-3 fatty acid-derived lipid mediators. KEYWORDS: Endocannabinoids, N-docosahexaenoyl dopamine, N-arachidonoyl dopamine, cyclooxygenase-2, prostaglandin E2, interleukin-6

N

inflammatory processes receives considerable attention in brain research as a means to prevent or treat neurological diseases. Several lines of evidence suggest that intake of long chain omega-3 polyunsaturated fatty acids (n-3 LC-PUFAs), especially docosahexaenoic acid (DHA; 22:6 n-3), is associated with a reduced cognitive decline and risk for certain neurological disorders.2−5 Although the picture is not fully

euroinflammation is a common element of several neurological disorders, including Alzheimer’s disease, multiple sclerosis, Parkinson’s disease, stroke, and trauma.1 In brain, resident microglial cells are activated and peripheral inflammatory cells can be mobilized and recruited to the sites of damage in response to an insult or pathological process. The continuous release of molecular mediators (e.g., cytokines, chemokines) from these cells, often together with a reduced integrity of the blood-brain barrier (BBB) and increased bloodborne leukocyte infiltration, leads to a vicious cycle of inflammation in the brain tissue.1−3 Modulation of neuro© 2016 American Chemical Society

Received: September 9, 2016 Accepted: November 21, 2016 Published: November 21, 2016 548

DOI: 10.1021/acschemneuro.6b00298 ACS Chem. Neurosci. 2017, 8, 548−557

Research Article

ACS Chemical Neuroscience

However, N-acyl dopamines derived from the polyunsaturated omega-3 fatty acids, such as DHDA and EPDA, have received limited attention so far.12,25 The present study was undertaken to evaluate the antiinflammatory activity of DHDA in both RAW264.7 macrophages and BV-2 microglial cells. We first studied its modulatory effects on the production of several key mediators (NO, IL-6, MCP-1, and CCL-20), known to be involved in neuroinflammation. To investigate the underlying mechanism of DHDA exerted immune-modulatory activity, we further studied its effects on the level of COX-2 mRNA expression, its metabolite PGE2 and the potential involvement of NF-κB. The present study provides new insights in molecular mechanisms by which DHA could be involved in the modulation of neuroinflammatory processes.

clear, part of these beneficial effects of n-3 LC-PUFAs is considered to be related to their effects on inflammatory processes involved. Different potential mechanisms underlying the effects of DHA have been investigated, including DHA as a natural ligand for peroxisome proliferator activated receptor-γ (PPAR-γ)6 and the G-protein coupled receptor GPR120.7 Moreover, it was found that DHA can act as a competitive inhibitor of the conversion of arachidonic acid (AA; 20:4 n-6) into proinflammatory lipid intermediates and as a source for the generation of anti-inflammatory lipid mediators, such as resolvins and protectins.8,9 We have found evidence for an alternative mechanism showing that the anti-inflammatory actions can also be mediated via fatty acid amides (FAAs) derived from n-3 LCPUFAs.10−14 FAAs are a group of lipids formed from fatty acids and ethanolamine, amino acids, or monoamine neurotransmitters, which are widely occurring in nature.15,16 These lipids have been shown to belong to a class of endogenous signaling molecules that are involved in several biological processes, such as pain and inflammation.16−18 N-Docosahexaenoyl ethanolamine (DHEA), the ethanolamide of DHA, was shown to possess considerably stronger anti-inflammatory potency compared to its precursor DHA in activated macrophages through an inhibition of cyclooxygenase-2 (COX-2) derived eicosanoid production.13 A recent study showed that brain levels of N-docosahexaenoyl amines, but not of other N-acyl amines (N-arachidonoyl, N-palmitoyl, N-linoleoyl), were upregulated after acute peripheral injury.19 This provided additional evidence on the molecular level that n-3 fatty acids might be involved in modulation of inflammatory processes via their amide derivatives.19 Dopamine, one of the major neurotransmitters in the CNS, is also involved in regulation of the immune system and regulation of host defense.20 Conjugates of fatty acids with dopamine (N-acyl dopamines) might represent a so far largely unknown class of mediators involved in the regulation of neuroinflammation. It has been shown that some N-acyl dopamines possess cannabimimetic properties, especially if the fatty acid moiety contains four or more double bonds, as in Narachidonoyl dopamine (NADA), N-eicosapentaenoyl dopamine (EPDA), and N-docosahexaenoyl dopamine (DHDA)21 (Figure 1). NADA, an endocannabinoid which acts as an agonist of the CB1 receptor22 and the transient receptor potential V1 (TRPV1),23 was found to be a potent inhibitor of PGE2 synthesis in lipopolysaccharide (LPS) stimulated primary microglial cells, without modifying the expression or enzymatic activity of COX-2 and the production of prostaglandin D2.24



RESULTS

Cytotoxicity of Compounds. The effects of DHDA, NADA, DHA, and dopamine on cell viability and cytotoxicity were examined by performing XTT and LDH assays, in both RAW264.7 and BV2 cells. For all conditions tested, cell proliferation and cytotoxicity did not differ more than 20% compared to the respective vehicle control (VC), indicating that none of these conditions was cytotoxic. Data are shown in Table S1 (Supporting Information). DHDA Reduces Nitric Oxide (NO) Production in LPSStimulated RAW264.7 Macrophages. To investigate the potential anti-inflammatory properties of DHDA, RAW264.7 macrophages stimulated with 0.5 μg/mL LPS were incubated with a concentration series ranging from 10 nM to 2.5 μM of DHDA. The incubation time was 48 h, as NO is a relatively late inflammatory marker of the LPS-induced inflammatory cascade. DHDA concentration-dependently suppressed NO production (Figure 2a) and significantly inhibited NO release at the highest concentration (2.5 μM) up to 54% (F4,15 = 68.28, p < 0.001). The precursors of DHDA, DHA, and dopamine did not affect production of NO (Figure 2b). No or extremely low levels of NO production were measured in the absence of LPSstimulation. The potency of DHDA to inhibit NO production was compared with that of the endocannabinoid compound NADA. Concentrations were tested from 10 nM up to 1 μM (higher concentrations of NADA showed cytotoxicity on RAWs). As shown in Figure 2c, both NADA and DHDA showed similar potency in reducing NO production at 1 μM. DHDA Elicits Concentration-Dependent Suppression of CCL-20, MCP-1 and IL-6 Release in RAW264.7 Macrophages after LPS-Stimulation. We further investigated the effects of DHDA for three proinflammatory markers, namely the inflammatory chemokine monocyte chemotactic protein-1 (MCP-1), the macrophage-inflammatory protein-3α (CCL-20) and the cytokine interleukin-6 (IL-6). Concentration−dependent effects of DHDA on the production of these mediators were observed (Figure 3). After 24 h incubation, 2.5 μM DHDA significantly inhibited release of CCL-20 (Figure 3a), MCP-1 (Figure 3b), and IL-6 (Figure 3c) up to 62% (F4,10 = 57.47, p < 0.001), 58% (F4,10 = 27.67, p < 0.001), and 46% (F4,10 = 38.32, p < 0.001), respectively. Lowest suppressive significant effects were found with 100 nM DHDA for IL-6 release. The precursors of DHDA, DHA and dopamine, had no effect on the cytokine production (data not shown). No or low levels of cytokines production were measured in the absence of LPS-stimulation.

Figure 1. Chemical structure of N-arachidonoyl dopamine (NADA) and N-docosahexaenoyl dopamine (DHDA). 549

DOI: 10.1021/acschemneuro.6b00298 ACS Chem. Neurosci. 2017, 8, 548−557

Research Article

ACS Chemical Neuroscience

Figure 2. (a) Effect of N-docosahexaenoyl dopamine (DHDA) on NO production in RAW264.7 macrophages at 48 h. (b) Effect of DHA and dopamine on NO production in RAW264.7 macrophages at 48 h. (c) Effect of DHDA and NADA on NO production in RAW264.7 macrophages at 48 h. RAW 264.7 cells were seeded in 96-well plates (density of 250 000 cells/mL). The test compounds were added to the adherent cells, then stimulated with LPS (0.5 μg/mL) for 48 h. The supernatant of the cells was analyzed for nitric oxide using the Griess assay. Data are expressed as percentage, where LPS stimulation (containing solvent only) was set at 100%. Values are means of three separate experiments (each done in duplicate), with standard errors of the mean represented by vertical bars. Mean values were significantly different from the control: ***p < 0.001, analyzed by one-way ANOVA, Dunnett’s t-test. No significant difference (n.s.) was found between NADA and DHDA, analyzed by two-way ANOVA, Dunnett’s t-test.

DHDA Suppresses LPS-Induced PGE2 Release and COX-2 Gene Expression in RAW264.7 Macrophages. To further examine the possible targets involved in the immunemodulating effects of DHDA in RAW264.7 macrophages, we first determined the level of PGE2, one of COX-2 enzymatic products. Production of PGE2 was concentration-dependently suppressed with a reduction of 25.3% at 100 nM DHDA and a 75% reduction for 1 μM DHDA (Figure 4a). To establish whether this reduction was regulated at the gene-expression level, COX-2 mRNA level was assessed in the corresponding cells by RT-qPCR. After 24 h, the expression level of COX-2 was concentration-dependently decreased, with up to 58% induced suppression by 2.5 μM DHDA (Figure 4b). DHDA Does Not Affect NF-κB Activity. To investigate whether DHDA interacts with targets upstream of the COX-2 enzyme, the possible involvement of NF-κB was assessed. To this end, effects of different DHDA concentrations (0.2−3.2 μM) on LPS-induced NF-κB activity were tested using HEK Blue NF-κB reporter cells. At these different, nontoxic, concentrations no effects of DHDA on NF-κB activity were observed, suggesting that DHDA acts downstream of NF-κB instead of directly affecting the NF-κB pathway (Figure 5). DHDA Elicits Concentration-Dependent Suppression of IL-6 and CCL-20 Production in Microglia Cells after LPS-Stimulation and Exerts Similar Suppression Effects as NADA. Next, to assess the activity of DHDA in LPSactivated microglial cells and compare these effects with antiinflammatory properties of NADA, the release of IL-6, CCL-20

and MCP-1 was investigated after DHDA stimulation. A concentration series from 10 nM up to 2 μM DHDA was tested. After 24 h incubation, 2 μM DHDA significantly inhibited IL-6 and CCL-20 release up to 49% (F4,20 = 43.12, p < 0.001) (Figure 6a) and 37% (F4,20 = 7.52, p < 0.01) (Figure 6b), respectively. There were no effects of DHDA on MCP-1 production in the microglial cells (Figure 6c). The lowest concentration at which significant effects of DHDA on IL-6 production was observed was at 100 nM (F4,20 = 43.12, p < 0.05); while in case of CCL-20, effects of DHDA became significant at 2 μM (F4,20 = 7.52, p < 0.01). The precursors of DHDA, DHA, and dopamine did not affect cytokine production (Figure 6d and e). No or extremely low levels of cytokine production were measured in the absence of LPSstimulation. Effects of DHDA on production of IL-6 (Figure 6f) and CCL-20 (Figure 6g) were compared with those of NADA in a concentration range of 10 nM to 1 μM. As shown in Figure 6, both DHDA and NADA reduced IL-6 with similar potencies, whereas they both did not affect CCL-20 production at the concentrations tested. DHDA Reduces PGE2 Levels, But Not COX-2 Gene Expression in LPS-Activated Microglial Cells. As DHDA significantly inhibited PGE2 level and COX-2 gene expression in RAW264.7 macrophages, it was studied whether the same potential targets are involved in the immune modulatory process of DHDA in activated microglial cells. A concentrationdependent effect was found for PGE2 with a significant reduction of 25.3% (F6,14 = 145.9, p < 0.001) for a 550

DOI: 10.1021/acschemneuro.6b00298 ACS Chem. Neurosci. 2017, 8, 548−557

Research Article

ACS Chemical Neuroscience

Figure 3. Effect of N-docosahexaenoyl dopamine (DHDA) on CCL-20 (a), MCP-1 (b), and IL-6 (c) production in RAW264.7 macrophages at 24 h. RAW 264.7 cells were seeded in 48-well plates (density of 250 000 cells/mL). The test compounds were added to the adherent cells, then stimulated with LPS (0.5 μg/mL) for 24 h. The supernatant of the cells was analyzed for cytokines by ELISA. Data are expressed as percentage, where LPS stimulation (containing solvent only) was set at 100%. Values are means of three separate experiments (each done in duplicate), with standard errors of the mean represented by vertical bars. Mean values were significantly different from the control: *p < 0.05, **p < 0.01, ***p < 0.001, analyzed by one-way ANOVA, Dunnett’s t-test.

Figure 4. Effect of N-docosahexaenoyl dopamine (DHDA) on PGE2 (a) production level in the culture medium and COX-2 (b) mRNA expression level in activated RAW264.7 macrophages at 24 h. RAW 264.7 cells were seeded in 48-well plates (density of 250 000 cells/mL). The test compounds were added to the adherent cells, then stimulated with LPS (0.5 μg/mL) for 24 h. The cells were harvested for analyzed COX-2 mRNA level by RT-qPCR and the culture medium was analyzed for PGE2 by ELISA. Data are expressed as percentage, where LPS stimulation (containing solvent only) was set at 100%. Values are means of three separate experiments (each done in duplicate), with standard errors of the mean represented by vertical bars. Mean values were significantly different from the control: *p < 0.05, ***p < 0.001, analyzed by one-way ANOVA, Dunnett’s t-test.

studies, randomized control trails and animal studies generated a large body of evidence showing that an increased intake of n− 3 PUFA is associated with a lower risk of cognitive decline and Alzheimer’s disease.28 Moreover, it is also shown that DHA provides a protective mechanism in human, primates and rodents models of Parkinson’s disease.29−32 These effects might be partly due to the anti-inflammatory effects of n-3 PUFAs, which are involved in both the reduction in and resolution of inflammation. To date, several mechanisms underlying these

concentration as low as 100 nM DHDA and a 62% (F6,14 = 145.9, p < 0.001) reduction for 2 μM DHDA (Figure 7a). However, gene-expression of COX-2 was not altered by DHDA in activated microglial cells after 24 h incubation (Figure 7b).



DISCUSSION Dietary consumption of DHA has been linked to a decreased grade of neuroinflammation and improvement of certain neurological disorders.2,5,26,27 Results from observational 551

DOI: 10.1021/acschemneuro.6b00298 ACS Chem. Neurosci. 2017, 8, 548−557

Research Article

ACS Chemical Neuroscience

proinflammatory activity.42−47 Cyclooxygenase-2 plays a central role in the inflammation cascade and is considered an important therapeutic target in many neuroinflammatory disorders.48 Observational studies in humans point to a protective role of NSAIDs (nonsteroidal anti-inflammatory drugs) in AD and cognitive impairment.49 Studies in animals support this observation. For example, recently it was reported that COX-2 and PGE2 levels are increased in hippocampus in an colchicine induced rat model of AD (cAD) and that COX-2 mediated neuroinflammation was linked to neurodegeneration.50 Preadministration of a COX-2 inhibitor in these cAD rats resulted in protection of memory impairments, neurodegeneration and neuroinflammation in hippocampus.50 In our study, DHDA was found to affect COX-2 activity (RAW264.7 and microglia) and expression (RAW264.7 only) significantly. Unlike its analogue DHEA, DHDA reduced COX-2 expression in RAW264.7 at the transcriptional level up to 60%; DHEA has been shown to suppress COX-2 derived prostaglandin levels in RAW264.7, without altering expression of COX-2. DHDA appears to exert its effects downstream of NF-κB, as no effects on NF-κB activity were found in the reporter assay up to 3.2 μM DHDA, suggesting that NF-κB is not involved in mediating the effects of DHDA. Additionally, we demonstrated that the precursors of DHDA, DHA and dopamine, did not induce any effects at 2.5 μM, confirming that DHDA-elicited effects were not due to its possible hydrolysis products. Interestingly, DHA has also been previously reported to decrease the expression of proinflammatory factors in BV-2 microglial cells, such as inducible NO synthase (iNOS), IL-1β and IL-6,51,52 and to be associated with attenuated neuroinflammation. However, the effective concentrations for DHA were approximately 10−100 times higher compared to the effective concentrations we found here for its dopamine conjugate. It might be the case that DHDA is functioning as a potential metabolite of DHA thereby contributing to the beneficial effects of DHA involved in neuroinflammatory related disorders, such as Alzheimer’s disease and stroke.36. Interestingly, DHDA displayed a somewhat different antiinflammatory pattern in the macrophage cell line compared with the effects found in the microglia cell line. This may be explained by the primary role of microglia as they mainly serve as surveillance cells in the brain to support normal tissue function and provide housekeeping functions.53 Once they are activated by invading agents or internal enemies, microglia and many other immune cells including peripheral macrophages can be attracted to cross the BBB and they become involved in immune defense of the brain. To attract other immune cells, chemokines, like MCP-1, might be crucial for microglia cells’ functionality in the natural physiological condition of CNS.54 DHDA has Immune-Modulating Properties Comparable to Its Analogue NADA. We have shown that DHDA exhibited similar immune-modulatory activity as NADA in both RAW264.7 and microglial cells. Interestingly, concentrations of DHDA higher than 1 μM, namely, 2 and 2.5 μM, displayed no signs of cytotoxicity in our assays for both RAW264.7 cells and microglial cells, which is in contrast to NADA. NADA is an endogenous arachidonoylamine that has been shown to induce anti-inflammatory and immunosuppressive effects in several cells types, e.g., human endothelial cells, primary glial cells, and human T cells.23,24,40,55 We further assessed that DHDA acts as potent inhibitor of PGE2 production, however, without modifying the expression of COX-2 in LPS-stimulated microglial cells. Similarly, its arachidonoyl analogue NADA

Figure 5. Effect of N-docosahexaenoyl dopamine (DHDA) on NF-κB activity measured in a HEK blue cells reporter assay. Cells were seeded overnight and different concentrations of DHDA (0, 0.2, 0.4, 0.8, 1.6, and 3.2 μM) were added for 18 h with in the presence of LPS (1 μg/ mL). Supernatant was used for measuring NF-κB activity. Experimental details are described in the Methods. Data are expressed as percentage, where LPS stimulation (containing solvent only) was set at 100%. Values are means ≥3 separate experiments (each done in duplicate), with standard errors of the mean represented by vertical bars.

anti-inflammatory effects of DHA have been proposed.33−36 Data from our and other research groups have provided evidence for the existence of an as yet largely unexplored mechanism, involving DHA-derived amides as a class of endogenous signaling molecules that act as modulators of inflammation.10,37 In the current study, we show that DHDA shares this capacity by demonstrating that it is a potent antiinflammatory compound in both LPS-activated RAW264.7 macrophages and microglial cells. The results suggest that DHDA has anti-inflammatory effects in immune cells, thereby acting downstream of NF-κB by modulating expression or activity of COX-2. DHDA has Potent Immune-Modulating Properties and Its Mechanism of Action Involves the Enzyme COX2. Compared to other more studied FAAs, in particular those that belong to the endocannabinoid class of signaling molecules, N-acyl dopamines have received considerably less attention thus far. Although dopamine conjugates with arachidonic acid (NADA), oleic acid (OLDA), stearic acid (STEARDA), and palmitic acid (PALDA) have been described in the literature before,23,38−40 hardly anything is known on the n-3 PUFA-derived conjugates with dopamine. Interestingly, our data provide evidence that DHDA has strong immunemodulatory potency, when compared to some other structurally related FAA compounds. So far DHEA (the ethanolamine analogue) has been found to display the highest efficacy in attenuating LPS-induced NO-production among a series of related N-acylethanolamines, including the endocannabinoid anandamide (AEA).10 The data reported here suggest that DHDA possesses an even stronger capacity to suppress NO, IL6, and MCP-1 production in LPS-stimulated RAW264.7 macrophages. These results are in line with observations showing that N-acyl dopamines with polyunsaturated fatty acids, including DHDA, exhibited more potent inhibitory activity on the production of NO in LPS stimulated RAW264.7 macrophages than N-acylethanolamines and their analogues.41 Depending to a certain extent on the cell type, our data showed that DHDA exerted potent inhibitory effects on the release of key enzymes and mediators involved in the pathogenesis of neuroinflammation, including COX-2, PGE2, NO, IL-6, MCP-1, and CCL-20, which are all reflecting a 552

DOI: 10.1021/acschemneuro.6b00298 ACS Chem. Neurosci. 2017, 8, 548−557

Research Article

ACS Chemical Neuroscience

Figure 6. Effect of N-docosahexaenoyl dopamine (DHDA) on IL-6 (a), CCL-20 (b), and MCP-1 (c) production in LPS stimulated microglial cells at 24 h; effect of DHA and dopamine on IL-6 (d) and CCL-20 (e) production in LPS stimulated microglial cells at 24 h; comparison of the effects of N-docosahexaenoyl dopamine (DHDA) and N-arachidonoyl dopamine (NADA) on IL-6 (f) and CCL-20 (g) production in LPS stimulated microglial cells at 24 h. Microglial cells were seeded in 24-well plates (density of 250 000 cells/mL). The test compounds were added to the semiadherent cells, then stimulated with LPS (1 μg/mL) for 24 h. The supernatant was analyzed for cytokines by ELISA. Data are expressed as percentage, where LPS stimulation (containing solvent only) was set at 100%. Values are means ≥3 separate experiments (each done in duplicate), with standard errors of the mean represented by vertical bars. Mean values were significantly different from the control: *p < 0.05, **p < 0.01, ***p < 0.001, analyzed by one-way ANOVA, Dunnett’s t-test.

Figure 7. Effect of N-docosahexaenoyl dopamine (DHDA) on PGE2 production level (a) in the culture medium and COX-2 mRNA level (b) in activated microglial cells at 24 h. Microglial cells were seeded in 24-well plates (density of 250 000 cells/mL). DHDA was added to the semiadherent cells and then stimulated with LPS (1 μg/mL) for 24 h. The cells were harvested for analyzed COX-2 mRNA level by RT-qPCR and the culture medium was analyzed for PGE2 by ELISA. Data are expressed as percentage, where LPS stimulation (containing solvent only) was set at 100%. Values are means of three separate experiments (each done in duplicate), with standard errors of the mean represented by vertical bars. Mean values were significantly different from the control: *p < 0.05, **p < 0.01, ***p < 0.001, analyzed by one-way ANOVA, Dunnett’s t-test. 553

DOI: 10.1021/acschemneuro.6b00298 ACS Chem. Neurosci. 2017, 8, 548−557

ACS Chemical Neuroscience



was also shown to reduce PGE2, but not prostaglandin D2 production in LPS-stimulated primary microglial cells, without modifying the expression or enzymatic activity of COX-2.24 Altogether, this indicates that DHDA functions as a lipid mediator that could modulate CNS inflammation. Can DHDA Act As a Bioactive Compound in Vivo? As DHA is the most abundant fatty acid in the brain and dopamine is a pivotal neurotransmitter in the brain, it can be hypothesized that DHDA can be endogenously formed and exert its bioactivities underlying the neuroprotective effects associated with DHA.56,57 It has been reported that other N-acyldopamines, like OLDA, PALDA, and STEARDA, are present in brains from different species, including mouse, rat, pig, and bovine.23,38,39 The NADA analogue, which is most similar to DHDA because of its poly unsaturated fatty acid moiety, has proven to be more difficult to detect than the analogues carrying a saturated fatty moiety.40 Unfortunately, despite several attempts, we were so far not able to detect DHDA in dopamine-rich areas (striatum and hypothalamus) of the brains from both rat and mice after applying different variations of our previously published method,58 which we deemed suitable to analyze the compound. This might suggest that the poly unsaturated fatty acid moiety affects the stability of the conjugates. Nevertheless, the ethanolamines of arachidonic acid (AEA and DHEA) are normally detected in brain,59 suggesting that the putative instability of DHDA cannot be caused by the poly unsaturated character of the fatty acid alone but rather by the dopamine group. Several reports underline DHDA’s promising biological activities. Besides its pronounced cannabimimetic properties in mice,21 it may also serve as a dopamine carrier molecule to the brain, contributing significantly to brain dopamine content.57 Furthermore, DHDA reduced the development of Parkinson’s disease-like symptoms in a mouse model and produced a concentration-dependent protective effect on cultured cerebellar granule cells from rat cerebellum under conditions of oxidative stress.60 Moreover, it has been reported that DHDA exhibits other effects like antiautophagic and antiapoptotic effects in breast cancer cells.61 The immunemodulating effects of DHDA might have contributed to some of these in vivo bioactivities described above, or alternatively, DHDA-exhitited bioactivities on different pathways or processes might yield combined or synergistic effects resulting in beneficial effects for some neuroinflammatory or cancer related disorders. In conclusion, in the present study, we showed that DHDA, derived from the n-3 polyunsaturated fatty acid DHA displays marked immune-modulatory effects by attenuating NO, IL-6, MCP-1, and CCL-20 levels in LPS-stimulated macrophages and in microglial cells. We further revealed that the underlying mechanism of action involves the enzyme COX-2 as its geneexpression and/or production of its metabolite PGE2 were down-regulated by DHDA. However, activity of the upstream inflammatory mediator NF-κB was not affected by DHDA. Our results provide new insights in potentially alternative molecular mechanisms by which DHA can modulate inflammatory processes in brain. Moreover, elucidating the immunemodulatory properties of the dopamine conjugate of DHA, a compound with reported pharmacological activity in vivo, might lead to novel routes to design therapeutic strategies to manage neuroinflammatory diseases.

Research Article

METHODS

Chemicals and Materials. Arachidonoyl dopamine, docosahexaenoic acid, PGE2 ELISA kits, Griess reagents, and nitrite standard were purchased from Cayman Chemical (Ann Arbor, MI). Immobilized Candida antarctica Lipase B (Novozym 435) was supplied by Novozymes A/S (Bagsværd, Denmark). Lipopolysaccharide (O111:B4; LPS), phenylmethylsulfonyl fluoride (PMSF), dopamine hydrochloride, 2-methyl-2-butanol, formic acid, TMB (3,3′,5,5′’-tetramethylbenzidine), triethylamine, trifluoroacetic acid (TFA), and HPLC grade acetonitrile (ACN) were from SigmaAldrich (Schnelldorf, Germany). Water for HPLC was purified with a Milli-Q system (Millipore, Billerica, MA). Dulbecco’s modified Eagle’s medium (DMEM), Roswell Park Memorial Institute 1640 (RPMI1640) medium, penicillin−streptomycin (pen−strep), and fetal calf serum (FCS) were obtained from Lonza (Verviers SPRL, Belgium). ELISA kits for IL-6, CCL-20, and MCP-1 were purchased from R&D Systems (Abingdon, U.K.). One-Step Enzymatic Synthesis of N-Docosahexaenoyl Dopamine. N-docosahexaenoyl dopamine (DHDA) was obtained by using a lipase-catalyzed N-acylation method,62 in which dopamine was used instead of ethanolamine. In particular, docosahexaenoic acid (DHA, 131 mg, 0.4 mmol), dopamine hydrochloride (76 mg, 0.4 mmol), and triethylamine (60 mg, 0.6 mmol) were incubated in an orbital shaker at 50 °C in 2-methyl-2-butanol (2 mL) for 48 h, using Novozym 435 as the catalyst (100 mg). After cooling, the solvent was evaporated under reduced pressure, and DHDA was purified by column chromatography on silica gel (using hexane-acetone as the eluent) with a yield of 102 mg (55%). Authenticity of the product was assessed by ESI-MS, 1H NMR, 13C NMR, and IR. Characterization data are reported in Figure S1. Cell Culture. RAW264.7 mouse macrophage cells (American Type Culture Collection, Teddington, U.K.) were cultured in DMEM supplemented with 10% (v/v) fetal bovine serum. The BV-2 mouse microglial cell line (Banca Biologica e Cell Factory, Genova, Italy) was cultured in RPMI supplemented with 10% (v/v) FCS. Under a humidified 5% CO2/95% air atmosphere, cells were plated in 75 cm2 cell culture flask (Corning, Acton, MA, USA) and passaged twice a week. Depending on the different experimental purposes, cells were seeded into different cell culture plates, either 96-well plates (100 μL per well), 48-well plates (300 μL per well), or 24-well plates (600 μL per well), with a density of 250 000 cells/mL. Viability and Cytotoxicity Assays. Effects of the compounds on the viability and cytotoxicity of the cells were evaluated as described before.63 The viability assessment was carried out by an XTT Cell Proliferation Kit II (Roche Applied Science, Almere, The Netherlands) according to the manufacturer’s instructions. In short, cells were incubated for either 24 or 48 h with the compounds and 0.5 μg/mL of LPS in 48-well plates. Thereafter, the XTT assay was performed, where the viable cells’ ability to metabolize XTT to formazan was a measure for cell viability. Conditions were considered toxic if metabolic activity to form formazan was decreased by >20%. As a control, Triton X-100 was added to the cells, yielding total cell lysis. Cytotoxicity of the samples was evaluated through an LDH Cytotoxicity Detection Kit (Roche Applied Science, Almere, The Netherlands) according to the manufacturer’s protocol. Cells were treated similarly as described for the viability assays. After the supernatant had been carefully removed, a mixture of the catalyst (diaphorase/NAD mixture) and a dye solution (iodotetrazolium chloride and sodium lactate) was added to the supernatant. The tetrazolium salt was reduced to formazan by leaked LDH in cell supernatant and this formazan dye was assayed by using an ELISA plate reader at 490 nm. Data of LDH and XTT are given in the Supporting Information. Effects of DHDA on Stimulated Release of Nitric Oxide, Cytokines (IL-6, MCP-1, and CCL-20), and PGE2. The production of nitrite (as an indicator for nitric oxide (NO) production), cytokines/chemokines (IL-6, MCP-1 and CCL-20), and PGE2 was measured as described previously.13 Briefly, DHDA was dissolved in absolute ethanol and then 1000× diluted with culture medium. The final concentration of solvent did not exceed 0.1% (v/v) in the culture 554

DOI: 10.1021/acschemneuro.6b00298 ACS Chem. Neurosci. 2017, 8, 548−557

ACS Chemical Neuroscience



medium. Before stimulation with LPS and compounds, RAW 264.7 cells were seeded at a density of 250 000 cells/mL in 96-well plates, and incubated overnight. The test compounds were added to the adherent cells in combination with LPS (0.5 μg/mL) for 48 h. To measure nitrite production, 100 μL of the supernatant was mixed with 100 μL of Griess reagent. This mixture was incubated for 10 min at room temperature. The absorbance at 540 nm was measured using an ELISA plate reader and the results were compared with a calibration curve using sodium nitrite as the standard. The inhibitory effects of DHDA on IL-6, MCP-1, and CCL-20 production were determined as described before.13 The incubation procedure was similar to the nitrite assay experiment described above. RAW 264.7 cells or BV-2 cells were seeded in 24-well plates and incubated overnight. The test compounds were added to the adherent cells in combination with LPS for 24 h. Thereafter, the supernatant was used for assessment of IL-6, MCP-1, and CCL-20 levels by mouse ELISA kits according to the manufacturer’s instructions. PGE2 production in the cell culture was also measured using a PGE2 EIA kit following an incubation of 24 h. RNA Purification and Quantitative Reverse Transcription Real-Time PCR. Total RNA was extracted using TRIzol reagent (Invitrogen, Breda, The Netherlands). RNA (1 mg/sample) was reverse transcribed to give complementary DNA using the reverse transcription system from Promega (Leiden, The Netherlands). Complementary DNA was amplified by PCR using the master-mix Sensimix SYBR (Bioline Reagents Ltd., London, UK) on a CFX Real Time system apparatus (Bio-Rad, Veenendaal, The Netherlands). The following primer pairs were used for amplification of COX-2:5′-TGAGCA-ACT-ATT-CCA-AAC-CAG-C-3′ (forward) and 5′-GCA-CGTAGT-CTT-CGA-TCA-CTA-TC-3′ (reverse). Samples were analyzed in duplicate, and mRNA levels of the different genes were normalized to RPS27A2. Primer pairs for RPS27A2 were 5′-GGT-TGA-ACCCTC-GGA-CAC-TA-3′ (forward) and 5′-GCC-ATC-TTC-CAGCTG-CTT-AC-3′ (reverse). NF-κB Reporter Assay. HEK-Blue Null1 cells (Invivogen, CA) were seeded in 96 wells plate with 100 000 cells/well in DMEM containing 10% (v/v) FCS. After overnight incubation, medium was replaced with serum free medium and DHDA was added at the indicated concentrations (0.2−3.2 μM). Subsequently, LPS (1 μg/ mL) was added and cells were incubated at 37 °C at 5% CO2/95% air atmosphere for 18 h. The next day, 20 μL supernatant per well was collected and mixed with 180 μL of QUANTI-Blue (Invivogen). NFkB activity was quantified using a spectrophotometer at 620−655 nm. Statistical Analysis. All cell culture experiments were performed in duplicate and repeated at least three times in independent experiments performed on different days. Data from all experiments are expressed as percentage of the LPS-treated controls (set at 100%) and presented as means ± standard errors of the mean (SEM) (see legends of the figures). Statistical differences between treatments and controls were evaluated by one-way ANOVA followed by Dunnett’s ttest. To test statistical differences between effects of NADA and DHDA, data were evaluated with two-way ANOVA followed by Dunnett’s t-test. A p value of