Fluoxetine Decreases Phagocytic Function via REV-ERBα in Microglia

Although fluoxetine (FLX) is a commonly used drug in psychiatric disorders, such as major depressive disorder, anxiety disorder, panic disorder, and obsessive–compulsive disorder, the mechanism by which FLX exerts its therapeutic effect is not completely understood. In this study, we aimed to determine the possible mechanism by which FLX focuses on microglial phagocytosis. FLX reduced phagocytic function in BV2 cells and increased REV-ERBα without affecting other microglia-related genes, such as inflammation and phagocytosis. Although FLX did not change BMAL1 protein levels, it restricted the nucleocytoplasmic transport (NCT) of BMAL1, leading to its cytosolic accumulation. REV-ERBα antagonist SR8278 rescued the decreased phagocytic activity and restricted NCT of BMAL1. We also found that REV-ERBα mediates the effect of FLX via the inhibition of phospho-ERK (pERK). The ERK inhibitor FR180204 was sufficient to reduce phagocytic function in BV2 cells and restrict the NCT of BMAL1. These results were recapitulated in the primary microglia. In conclusion, we propose that FLX decreases phagocytic function and restricts BMAL1 NCT via REV-ERBα. In addition, ERK inhibition mimics the effects of FLX on microglia.


Introduction
Fluoxetine (FLX) is an anti-depressant that belongs to the selective serotonin reuptake inhibitor (SSRI) class, which is a commonly prescribed drug for various neuropsychiatric diseases including major depressive disorder (MDD), anxiety disorder, panic disorder, and obsessive-compulsive disorder (OCD) [1]. Despite the well-known mechanism that elevates serotonin concentration by repressing serotonin reuptake in presynaptic neurons [2], the exact mechanism by which FLX exerts its therapeutic effect is not completely understood because FLX requires a longer medication time (at least 4 weeks) to achieve its therapeutic effect in patients regardless of serotonin concentration [3]. As additional mechanisms, FLX have neuroprotective, antioxidant, and anti-inflammatory effects [4] and it accelerates neurogenesis [5]. Considering that neuroinflammation is a multifaceted pathophysiology in psychiatric disorders [6], FLX mechanisms related to microglia, which are central nervous system (CNS) innate immune cells that play a crucial role in neuroinflammation, could be a promising therapeutic strategy in psychiatric disorders [7]. Notably, FLX inhibited IL-1β, TNF-α, and IL-6 elevation by nuclear factor kappa-light chain-enhancer of activated B cells (NF-κB) in microglia [8,9]. Indeed, treatment with FLX has been shown to decrease inflammatory cytokine elevations, such as IL-1β, IL-6, and TNF-α, in patients with MDD [10]. However, other clinical studies obtained opposite results, demonstrating no or even pro-inflammatory effects of antidepressant treatment [11]. In addition, experimental studies have shown that FLX may have a pro-inflammatory effect [12]. This discrepancy suggests that FLX may not have a significant effect on inflammatory processes [13]. Thus, an additional mechanism for FLX should be explored in microglia.
In addition to its anti-inflammatory effects, microglial phagocytosis is a crucial function in psychiatric disorders [14]. Microglial phagocytosis is required for proper Da-Yoon Jang and Bohyun Yang contributes to this study equally. neural circuit maintenance by the removal of cellular debris. Microglia phagocytize apoptotic neural progenitor cells and oligodendrocytes throughout development [15]. Even if transient, phagocytic dysfunction during the critical developmental period can contribute to aberrant CNS architecture and vulnerability to stress later in life [16]. Microglia isolated from mice with chronic unpredictable stress showed increased phagocytosis of neuronal components in the prefrontal cortex [17]. Chronic social defeat also increases the number of CD68 hi microglia, a marker of phagocytic activity [18]. Microglia can also remove synaptic terminals and axons during pruning, and impairment of microglial pruning leads to disrupted neuronal circuit formation and neuroplasticity via a complement system composed of C1q and C3 [19]. Microglia are unique CNS cells that express the C3 receptor, targeting immature synapses for phagocytosis [20] and are the dominant source of C1q [21]. Nevertheless, there are few reports on the effects of FLX on microglial phagocytosis.
Despite the tremendous number of studies on FLX, little is known about the role of FLX in the regulation of the circadian genes of microglia. In this study, we found that FLX decreased microglial phagocytic function via the elevation of REV-ERBα. In addition, the increase in REV-ERBα inhibited ERK phosphorylation, leading to reduced nucleocytoplasmic transport (NCT) of Bmal1.

Experimental Animals
All experimental procedures were approved by the Institutional Animal Care and Use Committee (IACUC) of CHA University (IACUC210089). All procedures were approved by the Institutional Ethics Committee for the Care and Use of Animals.

Primary Microglia Culture
Primary microglial cultures were conducted as described in our previous study [35]. Briefly, Uteri from pregnant mice (female C57BL/6, Orient Bio Inc. Seoul, Korea) were dissected and soaked in Hanks' Balanced Salt Solution (HBSS, Invitrogen, USA). The umbilical cord and yolk sac in the mesometrial surface of the uterus were removed using microsurgical instruments under a microscope and the embryo was taken out of the uterus. Then, the neuroepithelial layer (NEL) of C57BL/6 mice (Embyronic 13.5 days) was carefully dissected using a pair of microsurgical scissors. The shredded tissue was incubated with 1 mL of 1X trypsin-EDTA (Gibco, #15090-046) for 3 min in a 37 °C water bath. After centrifugation at 300×g for 5 min, the cells were plated in 25-T flasks coated with poly-d-lysine (Sigma, #P7280). The neuroepithelial layer was cultured in DMEM containing 10% FBS, 1X penicillin/streptomycin, and 0.1% GlutaMAX (Gibco, #35,050-061) at 37 °C in a 5% CO 2 incubator. The culture medium was replaced with 5 ml of fresh growth medium after 24 h. Subsequently, half of the culture medium was replaced with an equal volume of fresh growth medium twice per week. On day 21, when 1 3 stratification was reached, the NEL culture flasks were incubated with 1X trypsin-EDTA for 3 min, resulting in the detachment of an intact layer of cells in one piece. The pellet was resuspended in cold MACS PB buffer (containing PBS, pH 7.2, and 0.5% BSA), and cells were processed for microglia magnetic sorting by incubating with CD11b microbeads (Miltenyi Biotec, 130-093-634) for 15 min at 4 °C and then the cells were washed and resuspended in 500 μL MACS PB buffer. After washing, the cells were suspended in MACS PB buffer and applied to the LS column on a magnetic separator. The column was washed three times with PBS, and the separated cells (primary microglia) were obtained via positive selection. Primary microglia were resuspended in DMEM with 1% penicillin/streptomycin and 10 ng/ml recombinant mouse IL-34 (Tocris, #5195-ML-CF) and transferred to PDL-coated 24-well plates at a density of 2 × 10 5 cells/well.

Phagocytosis Assay
To assess phagocytic activity, cells were seeded on a 12 mm coverslip in 24-well cell culture dishes. BV2 cells were seeded at a density of 1 × 10 5 cells/ml and primary microglia were seeded at a density of 2 × 10 5 cells/ml. The cells were treated with 2 μL of red fluorescent latex beads (Sigma-Aldrich, #L3030) for 2 h at 37 °C. Phagocytic activity was inhibited by the addition of ice-cold PBS. Cells were washed twice with ice-cold PBS, fixed, stained with a microglial marker (Iba-1), and counterstained with DAPI. The cells were analyzed using a confocal microscope (TCS SP5, Leica).

Quantitative Reverse Transcription-Polymerase Chain Reaction (qRT-PCR)
Total RNA was extracted using TRIzol reagent (Invitrogen). Chloroform was added to separate the phase-containing RNA, and isopropyl alcohol was added to precipitate RNA. Each precipitated RNA pellet was air-dried and dissolved in DEPC-treated water (Invitrogen, Carlsbad, CA, USA). The RNA concentration was quantified using a spectrophotometer (DeNovix, DS-11 FX). Messenger RNA (mRNA) was reverse-transcribed with 20 μL of reaction mix using a RevertAid First Strand cDNA Synthesis kit (Thermo Fischer Scientific, Waltham, MA, USA). qPCR was performed using Power SYBR® Green PCR Master Mix (Life Technologies). The primer sequences are listed in Table 1. The cycling conditions consisted of initial enzyme activation at 95 °C for 5 min, followed by 50 cycles of denaturation at 95 °C for 20 s, annealing, and extension, including detection of SYBR Green bound to PCR product at 56 °C for 40 s. Glyceraldehyde 3-phosphate dehydrogenase (GAPDH) was used as an internal control for normalization. Relative quantities of PCR fragments were calculated using the comparative CT method.

Protein Extraction and Western Blot Analysis
The cells were washed twice with cold phosphate-buffered saline (PBS). Immediately after washing, cells were lysed with RIPA buffer (#R0278; Sigma). To collect the nuclear and cytoplasmic proteins, NE-PER Nuclear and Cytoplasmic Extraction Reagents (Thermo, #78833) were used according to the manufacturer's instructions. Briefly, BV2 cells were harvested using trypsin-EDTA and centrifuged at 500×g for 5 min. Ice-cold CER I was added to the cell pellet. The tube was vortexed vigorously for 15 s to completely suspend the cell pellet. After incubation on ice for 10 min, ice-cold CER II was added to the tubes. The tube was vortexed twice for 5 s and centrifuged for 5 min at 15,000×g. The supernatant was used as the cytoplasmic extract. After collecting the supernatant, the insoluble fraction was resuspended in icecold NER, and vortexed for 15 s every 10 min, for a total of 40 min. The tube was then centrifuged at 15,000×g for 10 min. The supernatant was used as the nuclear extract.
Protein concentration was determined using a detergentcompatible protein assay reagent (Bio-Rad Laboratories) with bovine serum albumin as the standard. After adding dithiothreitol (5 mM) and bromophenol blue (0.1% w/v), the proteins were boiled, separated by electrophoresis on 8-10% polyacrylamide gels (Bio-Rad), and transferred onto a polyvinylidene difluoride (PVDF) membrane (Bio-Rad Laboratories). Membranes were blocked in a shaker for 1 h at room temperature. The blocking buffer consisted of TBST (Tris-buffered saline/0.1% Tween-20) and 5% skim milk. (1:10,000, Abcam, #ab16048) for overnight at 4 °C. The membranes were incubated with anti-rabbit HRP (1:10,000, Bethyl, A120-101P) dissolved in blocking buffer at room temperature for 1 h. The membranes were visualized using an ECL-plus solution (BD Bioscience). The membranes were then exposed to chemiluminescence (LAS-4000, Fujifilm) to detect the light emission.

Migration Assay
For the migration assay, BV2 cells generated a single scratch wound with a sterile 200 μL pipette tip. The cells were replenished with fresh growth medium, and the gap area of wound closure was captured by photographing the same region at 0, 0.5, 1, 2, and 4 h. ImageJ was used to quantify the gap area of wound closure.

Statistical Analysis
Data are presented as mean ± standard error of the mean (SEM). The statistical significance of differences between groups was assessed with an unpaired t-test, one-way analysis of variance (ANOVA), and Kruskal-Wallis test using GraphPad Prism version 5 for Mac (GraphPad, La Jolla, CA,  CCT TCA CCA TTT TGT CTACA  GCT TAA GAG ACA GCC GCA TCT  IL-1β  GGC TGG ACT GTT TCT AAT GC  ATG GTT TCT TGT GAC CCT GA  iNOS  CAT TGG AAG TGA AGC GTT TCG  CAG CTG GGC TGT ACA AAC CTT  TNF-α  GAG TCC GGG CAG GTC TAC TTT  CAG GTC ACT GTC CCA GCA TCT  IL-6  AGT CCG GAG AGG AGA CTT CA  ATT TCC ACG ATT TCC CAG AG  IL-10  CCC TGG GTG AGA AGC TGA AG  CAC TGC CTT GCT CTT ATT TTC ACA  TGF-β  TGA CGT CAC TGG AGT TGT ACGG  GGT TCA TGT CAT GGA TGG TGC  Arginase 1  GGA ATC TGC ATG GGC AAC CTG TGT  AGG GTC TAC GTC TCG CAA GCCA  CX3CR1  TGG CCC AGC AAG CATAG  CAT GTC TGC TAC CCT CAC AAA  Trem2  TGG GAC CTC TCC ACC AGT T  GTG GTG TTG AGG GCT TGG  Tweak  TGC  The CT values were normalized as a ratio of control to 1, and the RQ value refers to the ratio of the respective genes as a percentage of the control (n = 6-9). E The migration of BV2 cells after FLX treatment was investigated at 0, 0.5, 1, 2, and 4 h time points after scratch. The percentage of wound area at indicated time points was calculated by ImageJ. Data are shown as the mean ± SEM. *p < 0.05, **p < 0.01 compared with the control USA). Tukey's post hoc test was performed using one-way ANOVA when the p values were < 0.05 and p < 0.05 was considered statistically significant.

FLX Reduced Phagocytosis with the mRNA Elevation of REV-ERBα in BV2 Cells
To determine the non-toxic and effective concentrations, BV2 cells were incubated with 5 μM or 10 μM FLX for 24 h. Calcein/ethidium staining showed that 10 μM FLX induced BV2 cell death (Fig. 1A). Consistent with this result, 10 μM FLX increased LDH activity (Fig. 1B). However, 5 μM FLX did not induce cell death. Therefore, we conducted subsequent experiments using 5 μM FLX. Next, we conducted a latex bead phagocytosis assay using BV2 cells. FLX treatment for 6 h did not affect phagocytic function, but FLX treatment for 24 h reduced phagocytic function (Fig. 1C). Next, we investigated the effect of FLX on gene expression in BV2 cells. In qRT-PCR, inflammation-related genes including il1β, iNOS, TNF-α, il6, il10, TGF-β, and arg-1 did not change after either 6 h or 24 h treatment with FLX. Although phagocytic function was changed by FLX, as shown in Fig. 1C, phagocytosis-related genes such as cx3cr1, trem2, tweak, c3ar1, and c1qa did not change. Interestingly, among the circadian genes, REV-ERBα mRNA expression was increased after 24 h treatment with FLX, but not other core circadian genes such as period1 (PER1), period2 (PER2), Cry1, Cry2, and BMAL1 expression (Fig. 1D). Although period3 (PER3) was increased at 6 h and reduced at 24 h, we focused on REV-ERBα because its mechanism was less reported compared to other period circadian protein homologues (PER1 and 2). Also, PER1 and PER1/2 were not changed in 24 h FLX-treated BV2 cells. In migration assay, FLX did not alter the migrating activity of BV2 cells (Fig. 1E).

FLX Restricted NCT of BMAL1 with REV-ERBα Elevation in BV2 Cells
To confirm the mRNA expression results, we performed western blotting. As shown in Fig. 2A, REV-ERBα expression was increased in 24 h FLX, but BMAL1 and PER2 were not. Additionally, we wondered whether FLX affected the NCT of BMAL1 and PER2, although FLX did not alter the protein levels of BMAL1 and PER2. To determine the circadian protein distribution between the cytosol and nucleus, cytoplasmic and nuclear proteins were fractionated. To confirm whether fractionation was successful, a nuclear membrane marker, lamin B1, was used. As shown in Fig. 2B, FLX increased the cytosol/nuclear (C/N) ratio of BMAL1, but not PER2. This result was confirmed in the immunofluorescence study. Immunostaining of circadian proteins showed an increased intensity of BMAL1 ( Fig. 2C) in the cytosol of microglia, but not in the case of PER2 (Fig. 2D). We also confirmed that nuclear architecture was not affected by FLX by investigating nuclear circularity (Fig. 2E).

REV-ERBα Antagonist Abrogated the Effect of FLX in BV2 Cells
Based on these results, we hypothesized that REV-ERBα might be a major factor related to reduced phagocytic function and increase of BMAL1 C/N ratio by FLX. To address this issue, we treated the antagonist of REV-ERBα (SR8278; SR) to BV2 cells with 5 μM FLX for 24 h and then assessed phagocytic function. As shown in Fig. 3A, SR rescued FLXinduced phagocytosis reduction in a concentration-dependent manner. In addition, SR restored the NCT restriction of BMAL1 by FLX, leading to a control level of the C/N ratio (Fig. 3B). In contrast, FLX and SR did not alter the migration of BV2 cells (Fig. 3C).

REV-ERBα Mediated the Effect of FLX via Inhibition of Phospho-ERK in BV2 Cells
Next, we explored the downstream signaling pathway of REV-ERBα to gain deeper mechanistic insight. Previous reports have shown that circadian genes are regulated by ERK and CREB [36][37][38]. In addition, IκB and NF-κB signaling are related to phagocytic functions [39]. Therefore, we identified the signaling molecules by western blotting. As shown in Fig. 4A, phospho-ERK (pERK) levels were reduced by FLX and recovered by SR treatment. However, phospho-CREB (pCREB), phospho-IκB (p IκB), and NF-κB levels were not significantly altered by FLX and Fig. 2 Change of BMAL1 localization by fluoxetine. Fluoxetine (FLX; 5 μM) was treated in BV2 cells for 24 h. Control (CON) cells were treated with vehicle. Western blotting and immunofluorescence studies were then performed. A Circadian protein REV-ERBα, BMAL1, and PER2 levels were assessed by western blot analysis and their expression was quantified using ImageJ (n = 3). B The expression of the circadian proteins BMAL1 and PER2 in the cytoplasm and nucleus was assessed by western blot analysis, and their expression was quantified using ImageJ (n = 6). C, D BV2 cells were stained for microglial markers (Iba-1, green) and circadian proteins (red) BMAL1 (C) or PER2 (D). Immunofluorescence study shows nuclear circularity using labmin B (green) in BV2 cells (E). Cytosol and nuclear protein intensity and roundness in single cells was analyzed using ImageJ (n = 100). Data are shown as the mean ± SEM. *p < 0.05, ***p < 0.001 compared with the CON ◂ SR. Thus, we tested whether ERK inhibition mimicked the effect of FLX on BV2 cells. The ERK inhibitor, FR180204 FR (10 μM), was sufficient to reduce phagocytic function in BV2 cells (Fig. 4B). In addition, ERK inhibition led to BMAL1 accumulation in the cytoplasm, similar to that observed with FLX (Fig. 4C). However, the ERK inhibitor did not change the PER2 NCT in BV 2 cells.

The Possible Mechanism of FLX was Recapitulated in Primary Microglia
Next, we investigated the possible mechanism underlying FLX recapitulation in primary microglia. Consistent with BV2 cells, FLX reduced latex bead phagocytosis, and this effect was abrogated by SR (Fig. 5A). In addition, the C/N ratios of the circadian proteins BMAL1 and PER2 were consistent in the BV2 cells (Fig. 5B). Furthermore, we found that the ERK inhibitor was sufficient to recapitulate the effects of FLX on phagocytosis (Fig. 5C). Taken together, the possible mechanism based on our results is summarized in Fig. 5D.

Discussion
A tremendous number of epidemiological and experimental studies have shown that circadian disruption is strongly associated with psychiatric disorders [40]. In experimental studies, suprachiasmatic nucleus (SCN)-specific BMAL1 knockdown induced depression-like behaviors in mice [41]. Per2 Brdm1 mutations also induce depressive and anxiety-like behaviors [42]. Genetic deletion or pharmacological inhibition of REV-ERBα triggers mania-like behaviors and maintains a high dopamine concentration in the caudate putamen and nucleus accumbens [43]. These studies suggest that circadian genes are closely related to the clinical manifestations of various neuropsychiatric diseases, including mood and anxiety disorders. In the present study, we identified, for the first time, that FLX reduces phagocytosis via REV-ERBα in microglia. REV-ERBα inhibited ERK phosphorylation, and the ERK inhibitor was sufficient to mimic the effects of FLX in microglia. Together, we speculate that microglial phagocytosis decreased by FLX via REV-ERBα may be associated with a therapeutic target in various psychiatric disorders.
The over-activation of microglia, which includes increased phagocytosis and soma size, is linked to anxietyand depression-like behaviors [44]. In the post-traumatic stress disorder (PTSD) model, fear conditioning induced synapse loss, microglial activation, and synaptic phagocytosis of activated microglia in the hippocampal dentate Western blotting was performed. A Phospho-ERK, phospho-CREB, phospho-IκB, and NF-κB levels were assessed by western blot analysis and their expression was quantified using ImageJ (n = 3). B The ERK inhibitor FR180204 (FR, 10 μM) was treated for 24 h, and then a phagocytosis assay was performed. The phagocytic activity of latex beads of BV2 cells was presented as the number of phagocytosed beads per cell. Two blinded observers manually counted the number of beads to minimize bias. Iba-1 (green), latex beads (yellow), and DAPI (blue) (n = 200). C The expression of BMAL1 and PER2 in the cytosol and nucleus was assessed by western blot analysis and quantified using ImageJ software (n = 3). Data are shown as the mean ± SEM. **p < 0.01 compared with the control. ##p < 0.01 compared with FLX gyrus [45]. In addition, autistic-like behavior induced by propionic acid is accompanied by an increase in microglial phagocytic marker CD68 [46,47]. In a human study, positron emission topography labeling of the translocator protein (TSPO) was enhanced in patients with MDD [48]. In addition, the enhanced density of phagocytic microglia is associated with high trait anxiety [49]. Thus, proper regulation of phagocytic function may be involved in the main effect of FLX in psychiatric disorders. In contrast, we found that another important function of microglia, migration, was not changed by either FLX treatment or REV-ERBα antagonist (SR8278) treatment. Microglial migration is regulated by fractalkine signaling. Fractalkine (Cx3cl1) is known to promote chemotaxis of microglia for neuroprotection and phagocytosis by binding microglial Cx3cr1 [50]. However, FLX did not change the expression of Cx3cr1 in BV2 cells. In addition to Cx3cr1, the characteristics of BV2 cells that originally showed maximum migratory activity under basal conditions, regardless of Cx3cr1 expression, might contribute to the effect of FLX on migration [51]. Therefore, further studies are needed to confirm the effects of FLX on cell migration.
To the best of our knowledge, few studies have examined the exact mechanism of FLX in microglia phagocytosis, focusing on circadian gene expression. Several studies have reported that circadian genes can modulate microglial phagocytosis. BMAL1 deficiency increased phagocytosis of BV2 microglia and decreased inflammatory gene expression after LPS stimulation [26]. In addition, the REV-ERBα agonist SR9009 suppressed the inflammatory response and phagocytosis by disrupting CLOCK gene rhythmicity in microglia [25]. Our results also showed that FLX reduced phagocytosis via REV-ERBα elevation, in accordance with the above study. Additionally, FLX significantly reduced ERK phosphorylation (p-ERK) via REV-ERBα elevation, suggesting that the ERK pathway is regulated by REV-ERBα. Furthermore, the ERK inhibitor was sufficient to mimic the effects of FLX on microglia. The ERK pathway can enhance or inhibit phagocytic function depending on disease model, in vitro microenvironment, and the type of microglia stimulus. Rock inhibitor increased phagocytosis via pERK elevation in BV2 cells [52]. Also, inhibition of ERK significantly reduced microglial phagocytosis of N2A cells with or without IFNγ treatment [53]. Furthermore, CB2 receptor activation inhibits the phagocytic function of microglia through activating ERK/AKT-Nurr1 signal pathways [54]. On the other hands, under persistent acidic environment, pERK was increased, leading to the reduced phagocytosis in microglia [51]. Thus, we targeted REV-ERBα rather than p-ERK on FLX mechanism. In addition, ERK phosphorylation directly regulates the NCT of BMAL1 via post-transcriptional modification of BMAL1, indicating that ERK inhibition restricts the transport of BMAL1 from the cytosol to the nucleus, leading to an increase in the BMAL C/N ratio [55,56]. We also found that the BMAL1 C/N ratio was increased by FLX or an ERK inhibitor, indicating that FLX caused BMAL1 accumulation in the cytosol by ERK inhibition. Based on the fact that BMAL1 plays a role in its function in the nucleus, the impaired NCT of BMAL1, leading to cytosolic accumulation, might be similar to BMAL1 knockdown in microglia. However, in contrast to our results, a previous study reported that microglia-specific knockdown of BMAL1 increased microglial phagocytic capacity [57]. In addition, BMAL1 knockdown BV2 cells showed decreased gene expression of inflammation and nutrient utilization, along with an increase in phagocytosis [26]. However, contrastingly to our results, the BMAL1 knockdown BV2 showed alteration of inflammatory-related genes (IL-1β, TNF-α, and IL-10) and expression of other circadian genes (Cry1, Cry2, Per2). Thus, BMAL1 C/T ratio elevation by FLX seems to induce a different mechanism from BMAL1 knockdown in microglia.
Although we report a new mechanism of FLX in microglial phagocytosis, there are some limitations to its translational interpretation. First, because this was an in vitro study, we could not confirm whether 24 h could reflect the long-term effect of FLX on microglia, similar to long-term therapy in the clinic. Long-term treatment is required to examine the positive effects of FLX in psychiatric disorders [58]. However, at least, 5 μM seems to be a measurable concentration in the brain when patient intake clinically reliable FLX dose [59,60]. In addition, there is a report inconsistent Fig. 5 Recapitulation of the effect of fluoxetine in the primary microglia. A To assess the effect of REV-ERBα antagonist on reduced phagocytosis and altered cytosol/nucleus (C/N) ratio of BMAL1 by fluoxetine (FLX) in BV2 cells, latex beads were added for 2 h after vehicle, FLX, or FLX + SR8278 (5 μM) treatment for 24 h in BV2 cells. Iba-1 (green), latex bead (yellow), DAPI (blue). The latex beads phagocytic activity of BV2 cells were presented as the number of phagocytosed beads per cell. The number of beads was counted manually by two blinded observers to minimize the bias (n = 200). B In addition, immunofluorescence staining was conducted in BV2 cells. The expression intensity of circadian protein BMAL1 and PER2 in cytoplasmic and nucleic proteins were assessed by immunofluorescence study, and their expression was quantified using ImageJ (n = 6). BV2 cells were stained to microglia marker (Iba-1, green) and circadian protein (red: BMAL1 and PER2). Cytosol and nucleus protein intensity in single cell was analyzed by ImageJ (n = 100). C ERK inhibitor, FR180204 (FR, 10 μM), was treated for 24 h, and then the phagocytosis assay was performed. The number of beads was counted manually by two blinded observers to minimize the bias (n = 200). D FLX increases the expression of the circadian protein REV-ERBα. The increased expression of REV-ERBα inhibits ERK phosphorylation. ERK inhibition reduces microglial phagocytosis and restricts BMAL1 transport from the cytosol to the nucleus. Data are shown as the mean ± SEM. *p < 0.05 **p < 0.01 **p < 0.001 compared with the control. #p < 0.05, ##p < 0.01 compared with FLX ◂ with our result, showing that FLX increased phagocytosis in BV2 cells [61]. This discrepancy might be associated with the difference of experimental design (concentration of FLX, incubation time of FLX, and ligand used for phagocytosis function). Of note, the receptor and downstream mechanism for phagocytosis function depend on the ligand [62]. Thus, a variety of ligands should be tested to determine the role of FLX on phagocytosis in further study. Second, we examined the effect of FLX in naïve microglia, which does not reflect the pathomechanism of psychiatric disorders. FLX is usually used in patients who are likely to have already distorted circadian rhythms, neuroinflammatory states, and neurotransmitter dysregulation. Thus, based on the finding that FLX influences circadian genes and microglial phagocytic function, further studies are needed in microglia models reflecting the pathomechanism of neuropsychiatric disorders.
In conclusion, we identified a new mechanism by which FLX decreased microglial phagocytosis via REV-ERBα elevation. REV-ERBα elevation inhibited the ERK pathway and increased the C/N ratio of BMAL1, leading to the cytosolic accumulation of BMAL1. Finally, the ERK inhibitor mimicked the mechanism of FLX in microglial phagocytosis. Thus, we propose that targeting REV-ERBα in microglia might be an alternative therapeutic strategy in psychiatric disorders, especially similar to ketamine, a rapid-acting antidepressant [40].