Maresin 1 attenuates pro‐inflammatory activation induced by β‐amyloid and stimulates its uptake

Alzheimer's disease (AD) is the most common dementia, characterized by pathological accumulation of β‐amyloid (Aβ) and hyperphosphorylation of tau protein, together with a damaging chronic inflammation. The lack of effective treatments urgently warrants new therapeutic strategies. Resolution of inflammation, associated with beneficial and regenerative activities, is mediated by specialized pro‐resolving lipid mediators (SPMs) including maresin 1 (MaR1). Decreased levels of MaR1 have been observed in AD brains. However, the pro‐resolving role of MaR1 in AD has not been fully investigated. In the present study, human monocyte‐derived microglia (MdM) and a differentiated human monocyte cell line (THP‐1 cells) exposed to Aβ were used as models of AD neuroinflammation. We have studied the potential of MaR1 to inhibit pro‐inflammatory activation of Aβ and assessed its ability to stimulate phagocytosis of Aβ42. MaR1 inhibited the Aβ42‐induced increase in cytokine secretion and stimulated the uptake of Aβ42 in both MdM and differentiated THP‐1 cells. MaR1 was also found to decrease chemokine secretion and reduce the associated increase in the activation marker CD40. Activation of kinases involved in transduction of inflammation was not affected by MaR1, but the activity of nuclear factor (NF)‐κB was decreased. Our data show that MaR1 exerts effects that indicate a pro‐resolving role in the context of AD and thus presents itself as a potential therapeutic target for AD.


| INTRODUC TI ON
Alzheimer's disease (AD) is the leading cause of dementia characterized by neuronal loss, and pathological accumulation of neurotoxic β-amyloid (Aβ) and hyperphosphorylated tau proteins, together with damaging chronic inflammation as indicated by activated microglia, which are the resident immune cells in the central nervous system (CNS) where they are important players in health as well as disease.
In health, microglia execute supporting functions for the nervous tissue by trophic support, synapse maintenance and phagocytic removal of molecular and cellular debris, as well as surveillance of the tissue for pathogenic threats. [1][2][3] In AD, microglia are activated by an overabundance of Aβ, a reaction that is further augmented due to the increase in misfolded and aggregated forms of this peptide, and thus initiate an innate immune response that contributes to the pathogenesis by increasing neurotoxic pro-inflammatory mediators while neuroprotective anti-inflammatory mediators are decreased, and oxidative stress increased. [4][5][6][7] Furthermore, phagocytosis is impaired, [8][9][10] thus limiting the ability of microglia to decrease the amyloid burden.
Considering the harmful consequences of chronic inflammation, 5,[11][12][13][14] it makes sense that the inflammatory response should be ended as soon as the pathogenic threat is neutralized. Under normal physiological conditions, an inflammation is cleared by resolution, 15 which is an active process wherein the beneficial aspects of inflammation are increased while the damaging ones are decreased.
This process is associated with the restoration and regeneration that occurs in healing. Thus, the damaged area where inflammation has acted is returned to homeostasis. Resolution of inflammation is mediated by specialized pro-resolving lipid mediators (SPMs) including lipoxins, resolvins, maresins and protectins are derived from n-3 and n-6 fatty acids. [15][16][17][18] SPMs down-regulate the inflammatory response, normalize chemokine gradients, facilitate the apoptosis of polymorphonuclear leucocytes and initiate the regeneration of local tissue by trophic activity and phagocytosis of molecular and cellular debris. 15,[19][20][21] Most research on resolution has been focused on the periphery, while the CNS has received less attention. There is, however, evidence for a failed and dysfunctional resolution in the AD brain, which can contribute to pathology and pathogenesis, indicated by decreased levels of SPMs in the hippocampus, entorhinal cortex and cerebrospinal fluid (CSF), 20,22,23 together with alterations in the levels of receptors for SPMs. 24,25 An impaired resolution in AD means that the brain is not only continuously exposed to the debilitating effects of chronic neuroinflammation but is also deprived of important trophic support and maintenance. One of the SPMs shown to be decreased in AD brains is maresin 1 (MaR1). 20 MaR1, derived from omega-3 fatty acid docosahexaenoic acid (DHA), was first detected by Serhan et al in mouse peritonitis exudates, using liquid chromatography-tandem mass spectrometry, 15 and subsequently, the pro-resolving roles of MaR1 have been identified in several disease models. [26][27][28][29][30][31] AD is a disabling disease afflicting an estimated 40 million people worldwide. 32 Societal costs amounted to about € 72 500 per person per year for residential care in 2011. 33 However, there is currently no treatment that can reverse the progression of the disease (see 34 ). New therapeutic strategies are therefore urgently needed.
Promoting the progression to resolution, leading to reduced inflammation and at the same time stimulating regeneration may be a successful therapeutic strategy for AD.
Considering the decrease in MaR1 levels in AD, and the promising results from previous in vitro and in vivo studies, 20,23,35 MaR1 is a candidate substance for re-establishing the failed resolution in AD.
However, the molecular mechanisms of action of MaR1 in the brain and in AD are not yet fully known and further studies are necessary before clinical trials in humans can be considered. The aim of the present study is to investigate the effects of MaR1 in a model of Aβ 42 -induced inflammation in two human in vitro models, with the hypothesis that MaR1 will stimulate a phenotype switch, resolve the inflammation, increase phagocytosis of Aβ 42 and improve cell survival and that these effects will be associated with reduced activity of intracellular pro-inflammatory signalling pathways.

| Materials
Human THP-1 and primary human monocytes were obtained from LGC Standards (Teddington, UK) and Lonza (Basel, Switzerland),

| Cell culture and stimulations
Human primary monocytes from adult healthy donors were purchased or isolated from fresh blood samples obtained from adult healthy volunteers who had given informed consent to participate in the study. All blood collection and experimental procedures were performed in compliance with the protocols approved by the Regional ethical review board in Stockholm (2019-04340). In brief, monocytes were positively selected from whole blood samples using CD14 + microbeads and were isolated using a whole blood column kit, and subsequently plated at a density of 10 5 cells/cm 2 in 24-well plates. To induce differentiation the monocytes were incubated at standard humidified culture condition (37°C in 5% CO 2 ) for 10 days  The production and purification of WT Aβ 42 monomers was performed as previously described. 37 Table S1). Purified Aβ 42 monomers were aliquoted in low-binding tubes and stored at −20℃. Before use, Aβ 42 was slowly thawed on ice to avoid aggregation.

| Immunocytochemistry
Immunocytochemistry for the macrophage and microglial marker Iba-1 was performed on d-THP-1 cells. In brief, the cells were fixed with 4% para-formaldehyde (PF) in PBS. After blocking by serum, the cells were incubated with Iba-1 antibodies (dilution 1:100) at +4°C overnight, rinsed in PBS for 15 minutes and then incubated at room temperature for 1 hour with Alexa Fluor 546 secondary antibodies (dilution 1:1000). The cell nuclei were stained with 4',6-diamidino-

2-phenylindole (DAPI). Microscopy was performed using a Zeiss
Fluorescence microscope, and images captured with a Nikon Ds-Fi1 camera controlled by NIS-Elements D software (both from Bergman-Labora, Sweden).

| Viability assessment
Cell death was analysed by LDH assay using a cytotoxicity detection kit according to the manufacturer's instructions. The absorbance was measured at 340 nm. The absorbance of the cell culture supernatant was obtained by subtracting the absorbance of the negative control. The appearance of the d-THP-1 cells was evaluated in 10× magnification using a light microscope (EVOS Cell Imaging System, Thermo Fisher Scientific, Stockholm, Sweden).

| Cytokine and IL-6Rα assays
Concentrations of TNF-α, IL-1β, IL-6 and IL-6Rα in d-THP-1 culture supernatants were determined using ELISA kits according to the instructions supplied with the kits, with the exception that QuantaRed TM fluorescent detection agent was used. The intensity of the emitted fluorescence at 575 nm (10 nm bandwidth) after excitation at 620 nm (10 nm bandwidth), with a gain of 60 V, was measured using a plate reader (Tecan Safire 2 , Tecan Nordic, Stockholm, Sweden). The cytokine concentrations were extrapolated from the standard curve within the recommended range for TNF-α (15.6-1000 pg/mL), IL-1β

| Western blot analysis
Analysis of phosphorylation of p38, p44/42, Akt and JNK was per-

| Surface biomarkers
The levels of the membrane-associated cellular biomarkers CD40,  Table S2. The cells were washed once and resuspended in 200 μL PBS and analysed with a BD Accuri C6 Plus flow cytometer. The percentage of labelled cells for each marker was determined as the proportion of cells with a signal that was stronger than that for the isotype control. Flow cytometric data were analysed by the BD Accuri C6 Plus software (v1.0).

| Statistical analysis
Statistical analyses were conducted using Statistica 12 (Dell Software, Aliso Viejo, USA). All data were normalized to the mean of each individual experiment. Kruskal-Wallis ANOVA was used to test for group differences, with the built-in post hoc test, or manually with Mann-Whitney with Bonferroni correction for multiple comparisons, used to test for differences between treatments. A P value of <0.05 was considered statistically significant.

| RE SULTS
In the present study, we have analysed the effects of MaR1 on   Figure 2D).

| MaR1 reduced Aβ 42 -induced cell death
The effect of MaR1 on cell survival in d-THP-1 cells incubated with

| MaR1 decreased pro-inflammatory surface biomarkers
In order to investigate whether MaR1 could alter the phenotype of d-THP-1 cells in the context of AD, cells were incubated for 2 hours with 5 μM Aβ 42 alone or together with 5 μM MaR1, followed by analysis of the pro-inflammatory (CD40 and CD86) and anti-inflammatory (CD163 and CD200R) biomarkers using flow cytometry (Figure 7). A total of six experiments were performed. Aβ 42 significantly increased the expression of CD40 and CD80 about 3-and 2-fold, respectively ( Figure 7A and B), whereas there was no significant effect on CD163 nor CD200R ( Figure 7C and D). MaR1 significantly reduced the Aβ 42 -induced increase in CD40, nearly to baseline level (P < 0.05) ( Figure 7A), whereas no effect of MaR1 was observed on the Aβ 42 -induced levels of CD86 ( Figure 7B).

| MaR1 decreased Aβ 42 -induced chemokine secretion
The effect of MaR1 on the levels of Aβ 42 -induced secretion of chemokines was investigated. The d-THP-1 cells were incubated for 24 hours with 5 μM Aβ 42 alone or together with 5 μM MaR1 (Figure 8).   Figure 8D and E) (all P values < 0.01), while no effect of MaR1 was seen on the other markers.

| MaR1 decreased Aβ 42 -induced NF-κB activation
To analyse the effect of MaR1 on NF-κB activation, an NF-κB reporter cell line based on THP-1 cells was differentiated in the same manner as the non-reporter cells and incubated with 5 µM Aβ 42 alone or together with 5 µM MaR1 for 24 hours, after which the luminescence was analysed in the conditioned medium from a total of 7 experiments. Aβ 42 increased the NF-κB activity 3-fold (P value < 0.01) and co-incubation with MaR1 reduced this elevation by approximately 50% (P value < 0.05) (Figure 9).

| MaR1 did not affect Aβ 42 -induced kinase activation
The was performed with the non-parametric Kruskal-Wallis (K-W) test, using the built-in post hoc test for multiple comparisons to find significant differences between treatments. ***P < 0.005 vs. vehicle. # P < 0.05, ## P < 0.01, ### P < 0.005 vs. 5 μM Aβ 42 . Aβ = β-amyloid; MaR1 = maresin 1 (See Figure S1). The analysis showed that the phosphorylation of p38 MAPK was increased by 2 or 5 μM Aβ 42 , whereas no effect of MaR1 was observed. Chemokines guide microglial migration to inflammatory areas and enhance the neuroinflammation in AD. 48 The release of chemokines from microglia is up-regulated upon Aβ stimulation. 49 We found Our studies show that MaR1 reduced pro-inflammatory surface biomarkers. Thus, the Aβ 42 -induced increase in CD40 was reduced almost to baseline levels by MaR1, supporting that the phenotype of microglia may have changed from pro-inflammatory to pro-resolving, in line with our earlier findings that MaR1 reduced Aβ 42 -induced CD11b in human CHME-3 microglia. 20 Gone et al reported that MaR1 attenuated elevation of the pro-inflammatory surface marker CD24 in the acute lung injury mice model induced by LPS. 31 In order to investigate the down-stream mediators of MaR1, we analysed the effects on NF-κB activation, and found that it reduced Aβ-induced NF-κB activation, similarly to effects seen in other disease models. 25,[50][51][52] This may be one of the mechanisms for the beneficial effects of MaR1 since NF-κB is a transcription factor for many inflammatory genes and for amyloid precursor protein (APP). 53 In order to further analyse the signal transduction mechanisms for the activities of MaR1, we analysed certain kinases related to pro-inflammatory activation. However, p38 phosphory-  45 We show that MaR1 increased Aβ 42 -uptake, which is a proposed therapeutic strategy for AD that is associated with the past and present clinical trials based on treatment with anti-Aβ antibodies.

| D ISCUSS I ON
Using SPMs to stimulate Aβ-uptake by microglia to achieve Aβ removal from the extracellular space and subsequent degradation may be a safer alternative than passive or active immunization, which in some cases has been linked to severe side effects. 54 Supporting evidence for SPMs to reduce Aβ 42 burden in vivo comes from studies on an AD mouse model in which co-administration of two SPMs significantly reduced brain amyloid levels. 55 Interestingly, MaR1 reduced Aβ 42

ACK N OWLED G EM ENTS
The study is supported by grants from the China Scholarship Council

CO N FLI C T O F I NTE R E S T
The authors declare that there are no conflicts of interest with regard to this study and its publication.

R E FE R E N C E S F I G U R E 9
MaR1 reduced Aβ 42 -induced NF-κB activation. Differentiated THP-1 (d-THP-1) cells were incubated with 5 µM Aβ 42 alone or together with 5 µM MaR1 for 24 h, after which the luminescence in the conditioned medium was measured. Incubation with vehicle served as control. Co-incubation with MaR1 showed a significant reduction of the Aβ 42 -induced increase in NF-κB activation. Analysis of variance (ANOVA) of the data was performed using the non-parametric Kruskal-Wallis (K-W) test. The built-in post hoc test for multiple comparisons was used to find significant differences between treatments. *P < 0.05 vs. vehicle. #P < 0.05 vs. 5 μmol/L Aβ 42 . Aβ = β-amyloid; MaR1 = maresin 1