Anti-inflammatory action of BT75, a novel RARα agonist, in cultured microglia and in an experimental mouse model of Alzheimer’s disease

Abstract

BT75, a boron-containing retinoid, is a novel retinoic acid receptor (RAR)α agonist synthesized by our group. Previous studies indicated that activation of retinoic acid (RA) signaling may attenuate progression of Alzheimer’s disease (AD). Presently, we aimed to examine the anti-inflammatory effect of BT75 and explore the possible mechanism using cultured cells and an AD mouse model. Pretreatment with BT75 (1 μM–25 μM) suppressed the release of nitric oxide (NO) and IL-1β in the culture medium of mouse microglial SIM-A9 cells activated by LPS. BMS195614, an RARα antagonist, partially blocked the inhibition of NO production by BT75. Moreover, BT75 attenuated phospho-Akt and phospho-NF-κB p65 expression augmented by LPS. In addition, BT75 elevated arginase 1, IL-10, and CD206, and inhibited inducible nitric oxide synthase (iNOS) and IL-6 formation in LPS-treated SIM-A9 cells, suggesting the promotion of M1-to-M2 microglial phenotypic polarization. C57BL/6 mice were injected intracerebroventricularly (icv) with streptozotocin (STZ) (3 mg/kg) to provide an AD-like mouse model. BT75 (5 mg/kg) or the vehicle was intraperitoneally (ip) injected to icv-STZ mice once a day for 3 weeks. Immunohistochemical analyses indicated that GFAP-positive cells and rod or amoeboid-like Iba1-positive cells, which increased in the hippocampal fimbria of icv-STZ mice, were reduced by BT75 treatment. Western blot results showed that BT75 decreased levels of neuronal nitric oxide synthase (nNOS), GFAP, and phosphorylated Tau, and increased levels of synaptophysin in the hippocampus of icv-STZ mice. BT75 may attenuate neuroinflammation by affecting the Akt/NF-κB pathway and microglial M1-to-M2 polarization in LPS-stimulated SIM-A9 cells. BT75 also reduced AD-like pathology including glial activation in the icv-STZ mice. Thus, BT75 may be a promising anti-inflammatory and neuroprotective agent worthy of further AD studies.

1. Introduction

Alzheimer’s disease (AD), the most common form of dementia, is a terminal, progressive disorder leading to memory loss, personality changes and the inability to communicate [1]. Currently, nearly 5.3 million Americans aged 65 and older have AD, 47 million people live with dementia globally, and it is estimated to increase more than threefold (~131 million) by 2050 [2]. Two well-recognized pathological hallmarks are extracellular amyloid-β (Aβ) plaques due to inappropriate digestion of amyloid precursor protein, and intracellular neurofibrillary tangles (NFT) due to hyperphosphorylated tau protein [3]. There is no effective cure for AD, and current treatments are only palliative. Owing to insidious onset, long incubation period, and various equivocal mechanisms, it is difficult to explore an effective AD therapy. Among various hypotheses regarding AD pathology, neuroinflammation has been one of the focuses of recent research [4–6].

Neuroinflammation is considered a pathophysiological process involved in various neurological and neurodegenerative diseases [7, 8]. Microglia activation provides the first line of defense against invading pathogens or injury-related products [9]. Under physiological conditions, microglia have branched morphology and high motility, which are beneficial for monitoring the microenvironment, pruning synapses, and timely removal of apoptotic neurons to maintain CNS homeostasis [10]. In the case of brain injury, microglia are the first to respond to this injury and become active [11]. Activated microglia are currently thought to differentiate into different phenotypes, including deleterious M1 and neuroprotective M2 phenotypes [12]. Specifically, “classically activated” M1 microglia typically release damaging pro-inflammatory cytokines such as IL-1β, TNF-α, and IL- 6, and reactive oxygen species and nitrogen (ROS and RNS) that exacerbate brain damage [13–15]. In contrast, the “alternatively activated” M2 phenotype secretes excess anti-inflammatory cytokines and trophic factors such as IL-4, IL-10, TGF-β, that resolve local inflammation and promote brain recovery [12]. Also, arginase 1 (Arg1), which converts arginine to polyamines, proline, and ornithine, can effectively outcompete iNOS, which uses the same substrate arginine, and downregulates production of pro-inflammatory mediator nitric oxide. Thus, iNOS and Arg1 represent a set of markers for M1 and M2 microglia [16]. CD206, also known as mannose receptor 1 or MR1, involved in the pinocytosis and the phagocytosis of immune cells, has been shown to be positively associated with M2 polarization and to play a role in regulating this phenotype [12,16]. In AD, while phagocytosis by microglia is known to play a major role in Aβ clearance, the presence of proinflammatory cytokines is known to block the phagocytic activity of microglia and may induce proinflammatory astrocytes [17–19]. These elevated inflammatory responses are associated with the cognitive decline and neuronal loss observed in AD patients and in the animal models [20, 21]. Therefore, inhibiting M1 activation and promoting M2 activation is a promising strategy for the treatment of neuroinflammation-related diseases such as AD and ischemic stroke compared with general inhibition of microglial activation.

Retinoic acid receptors (RARs), the nuclear receptor superfamily, are involved in the regulation of cell differentiation, proliferation, embryonic development, metabolism, and other life activities [22]. Activation of the retinoic acid (RA) signaling pathway may attenuate progression of AD through up-regulating non-amyloidogenic α-secretase and down-regulating amyloidogenic β-secretase 1 [23]. One study showed RA, the main active vitamin A metabolite, acted primarily through CDK5 to inhibit APP processing and reduce tau phosphorylation [24]. RA is also implicated in exerting anti-inflammatory functions [25]. A recent study indicating the relationship between reduction in BACE1 expression and anti-inflammatory effects of RAR agonists underscores the importance of elucidation of mechanisms of anti-inflammatory effects of RAR agonists [26].

BT75, a boron-containing retinoic acid receptor α (RARα) agonist, was previously synthesized by Dr. Das as a derivative of an RARα agonist Am580 and has been shown to be less toxic compared to all-trans retinoic acid (ATRA) or Am580 [27]. The use of boron atoms in pharmaceutical drug design has a high potential for discovery of new biological activities [28]. In the current studies, we examined possible anti-inflammatory action of BT75 and its mechanisms, using microglial cell culture and an icv-STZ AD mouse model.

2. Materials and Methods

2.1. Synthesis of BT75

Synthesis of BT75 [Fig. 1; (E)-2-(4-(2-(6,8-dichloro-2-phenyl-2H-chromen-3-yl)vinyl)phenyl)-4,4,5,5-tetramethyl-1,3,2-dioxaborolane] was performed as follows: A clean oven dried three-neck round bottom flask (RBF) was charged with aldehyde 1 [27] (1 equiv.) in DMF. Then addition of salt 2 (1.2 equiv.) Sodium tertbutoxide (STB) (3 equiv.) was added into the reaction mixture at 0°C. The reaction mixture was stirred at room temperature for 24 h. Progress of the reaction was monitored by TLC (100% Hexane), In TLC nonpolar spot was observed corresponding to aldehyde. After completion of the reaction, reaction mixture was neutralized by 2N HCl. After neutralization, ethyl acetate was added in the reaction mixture; the combined organic layer was collected and dried with Na₂SO₄ and evaporated under reduced pressure. The crude material was purified by column chromatography (5% ethyl acetate : 95 % hexane). After purification white solid was observed (BT75). The result of BT synthesis was: ( Yield = 72 %), ¹H NMR (600 MHz, DMSO-d₆) δ 7.61 (d, J = 7.8 Hz, 2H), 7.47 (d, J = 7.8 Hz, 2H), 7.42 (d, J = 7.2 Hz, 2H), 7.35 – 7.29 (m, 5H), 7.23 (d, J = 16.4 Hz, 1H), 7.01 (s, 1H), 6.71 (d, J = 16.5 Hz, 1H), 6.64 (s, 1H), 1.26 (s, 12H).

Fig. 1.

BT75 chemical structure and its synthetic formula.

2.2. Other chemicals

BMS195614, RARα antagonist, was purchased from Tocris (Minneapolis, MN). LPS (0111:B4, Cat# LPS25) and streptozotocin (Cat# S0130) were purchased from Sigma (St. Lous, MO). The 3-(4,5-dimethyl-2-thiazolyl) 2,5-diphenyl-2H-tetrazoliumbromide (MTT) was obtained from Sigma [Roche Cell Proliferation Kit I (MTT)].

2.3. Cell culture

Mouse microglial cell line SIM-A9 was obtained from ATCC (CRL3265, Manassas, VA) and cultured in DMEM/F12 containing 10% fetal bovine serum, 5% horse serum and 1% penicillin-streptomycin solution. Neuro-2a cell line was also purchased from ATCC (CCL-131) and cultured in DMEM containing 10% fetal bovine serum and 1% penicillin-streptomycin solution. All cells were cultured under the conditions of 37°C and 5% CO₂ in an incubator.

BT75 (from 25 μM to 1 μM) used for the treatment of SEM-A9 cells was dissolved in DMSO followed by dilution in DMEM. Final concentrations of DMSO were less than 0.1%. Other reagents (LPS and STZ) used were dissolved in DMEM.

2.4. AD-like icv-STZ model

Three-month-old male C57BL/6 mice were purchased from Jackson Laboratory (Bar Harbor, ME). All procedures described below were approved by the Nathan Kline Institute IACUC and were in accordance with NIH guidelines for the proper treatment of animals. The mice were injected (icv) with STZ at 3 mg/kg body weight to induce AD-like model using a slightly modified version of previously reported methods [29–31]. The scalp of the mouse was incised to locate bregma. Small burr holes (1 mm in diameter) were drilled, and STZ in citrate buffer (50 mM, pH 4.5) (volume was calculated according to the mouse weight) was injected into both sides of the lateral ventricle with a microsyringe fitted with a 26-gauge blunt cannula at a rate of 0.5 μL/min with the following coordinates: 0.2 mm posterior to the bregma, ±1.0 mm lateral to the sagittal line, and 2.5 mm ventral to the surface of the skull. The cannula was withdrawn after remaining for 5 min in the injection position. The incision was closed with wound clips, and the animal was allowed to recover under a heat lamp. Sham mice had received the same surgical procedures except with injection of the same volume of the vehicle (50 mM citrate buffer).

Three days after the icv-STZ injection, BT75 (in 5% DMSO in saline, 5mg/kg body weight, STZ+BT75 group) or the vehicle (5% DMSO in saline) (STZ group) was given to the mice with intraperitoneal (ip) injection once a day for 3 weeks. Sham mice were also given vehicle (Control group) or BT75 (5 mg/kg body weight, BT75 group). After drug administration, the hippocampus was collected for western blot or the brain was perfusion-fixed for immunofluorescence staining.

2.5. Cell viability detection

MTT method was used for cell viability detection. SIM-A9 cells (2 × 10⁵ cells/well) were propagated in 96-well plates. SIM-A9 cells were treated with different doses of BT75 for 24 h or pretreated with different doses of BT75 (25, 10, 5 and 1μM) for 30 min and then subjected to 24 h incubation with 1 μg/ml LPS. MTT was added to cell culture with the final concentration of 0.5 mg/ml. After 4 h incubation, the cell media was removed and 100 μl DMSO was added to dissolve the crystal. The absorbance was measured at 570 nm with reference wave at 630 nm. Neuro-2a cells were treated with BT75 (1 mM, 100 μM, 10 μM and 1 μM) or STZ (1 mM, 500 μM, 100 μM, 50 μM and 10 μM) for 24 h, and the cell viability was detected by MTT. The cell viability of Neuro-2a cells pretreated with BT75 (25, 10, 5 and 1 μM) for 30 min followed by treatment with STZ (1 mM) for another 24 h. Cell viability was assessed using the MTT assay.

2.6. Nitric Oxide (NO) detection

Griess reaction method was used for the detection of NO production. Due to its extreme instability, NO is quickly and proportionally metabolized to nitrite. Therefore, nitrite has been extensively used as an indicator of NO. SIM-A9 cells (2 × 10⁵ per well) were cultured in 96-well plates overnight. Then the cells were subjected to 30 min incubation of BT75 (25, 10, 5 and 1 μM) followed by 24 h stimulation with 100 ng/mL LPS. Same volume of the cell culture medium and Griess agent solution (Sigma-Aldrich, Cat# G4410) were mixed together and the absorbance at 540 nm was detected after 30 min.

2.7. Inflammatory cytokine detection

Enzyme-Linked Immunosorbent Assay (ELISA) was used to detect inflammatory factors in the cell culture medium. SIM-A9 cells (2 × 10⁵ per well) were cultured in 96-well plates overnight. Then the cells were pretreated with BT75 (25, 10, 5 and 1 μM) for 30 min followed by treatment with LPS (100ng/mL) for another 24 h. TNF-α, IL-1β l, and IL-10 levels in the cell culture medium were detected by commercial ELISA kits, TNFα Quantikine ELISA kit (R&D Systems, Minneapolis, MN, Cat# MTA00B ), mouse IL-10 Quantikine ELISA kit (R&D Systems), and IL-1β mouse ELISA kit (Thermo Fisher Scientific, Watham, MA, Cat# BMS6002), respectively, according to the manufacturers’ instructions.

2.8. Western blot

Western blot assay was done for the detection of target proteins involved in the cell signal pathway both in SIM-A9 cells and brain hippocampus tissue. Cells were pretreated with BT75 (10μM) for 30 min before 100 ng LPS treatment for 24 h, then the cells were collected and lysed with RIPA lysis buffer plus protease and phosphatase inhibitors to detect Akt, phospho (p)-Akt, NF-κB p65, p-NF-κB p65, iNOS, Arginase 1 (Arg1), IL-6, CD206, and β-actin expression by western blot. The hippocampus was also lysed and nNOS, GFAP, synaptophysin, p-Tau and β-actin expression were detected by western blot. Briefly, total protein amounts were measured by the Pierce BCA protein assay, and 20 μg amount of total protein from each sample was separated by SDS-PAGE gel and transferred to nylon membranes. After blocking with 5% non-fat milk powder in TBST for 2 h at room temperature, these membranes were incubated with primary antibodies against Akt (Cat#4685), p-Akt (Cat#4060), NF-κB p65 (Cat#8242), p-NF-κB p65 (Cat#8242), iNOS (Cat#13120), Arg1 (Cat#93668), β-actin (Cat#3700), nNOS (Cat#4231), GFAP (Cat#3670), p-Tau (p-Thr231, Cat#71429), synaptophysin (Cat#36406), CD206/MRC1 (Ca#24595), and IL-6 (#12912) [all from Cell Signaling Technology (Danvers, MA) with dilution of 1:1000] and corresponding secondary antibodies [Millipore Sigma (Burlington, MA) with dilution of 1:2500]. The bands were shown with an enhanced chemiluminescence (ECL) method and analyzed by Image J software.

2.9. Immunofluorescence staining

Immunofluorescence staining was done as previously described [32]. Mice were perfused with a solution containing 4% paraformaldehyde and 4% sucrose in cacodylate buffer (pH 7.2), and the heads were removed and further fixed in the perfusion solution overnight. Then brains were removed, transferred to phosphate buffered saline (PBS) solution, and kept at 4°C for 2–5 days until cut with a vibratome into 50 μm thick coronal sections. The free-floating sections were rinsed in PBS, permeabilized in methanol for 10 min, and incubated for 30 min in blocking solution (PBS containing 5% BSA and 0.1% Triton X-100), followed by incubation overnight with antibodies against GFAP (GA5) (mouse mAb Cat#3670, Cell Signaling with dilution of 1:500) and Iba1 (rabbit polyclonal Cat#019–19741, FujifilmWako Chemicals, Richmond, VA with dilution of 1:500), rinsed in PBS three times, followed by incubation with Alexa Fluoro594 goat anti-rabbit IgG and Alexa Fluoro488 goat anti-mouse IgG (Life Technologies, Grand Island, NY) in 0.1% Triton X-100 in PBS containing 1% BSA for 1 h at r.t.. Sections were finally rinsed in PBS three times, mounted, and coverslipped using ProLong Gold Antifade Reagent (Life Technologies). Glial activation was evaluated in the dorsal hippocampus region, especially in the fimbria, where activation was prominent. For quantitative measurement, GFAP-positive cell number or rod (or amoeboid)-like Iba1-positive cell number was counted in the hippocampal fimbria using 3 sections (around bregma −1.34 to −1.82) from each brain, and 3–4 brains per each treatment group were analyzed. Cell counting and the measurement of area of interest (AOI) were done by Image J software.

2.10. Statistical analysis

Data were presented in the form of mean ± SD. One-way analysis of variance (ANOVA) with a Tukey test was performed to analyze the data using GraphPad Prism software, and p < 0.05 was defined as statistically significant.

3. Results

3.1. Effects of BT75 on cell viability of SIM-A9 cells treated by a high dose of LPS

The cytotoxicity test of BT75 (500, 250, 100 and 50 μM) was done using SIM-A9 cells. BT75 was dissolved in DMSO followed by dilution in DMEM. Final concentrations of DMSO were less than 0.1%. the results showed that 50 μM BT75 was not toxic to cells (Fig. 2A). Therefore, we chose lower than 50 μM BT75 for further experiments. The cell viability of SIM-A9 cells pretreated with BT75 (25, 10, 5 and 1 μM) for 30 min followed by treatment with LPS (1 μg/mL) for another 24 h indicated that BT75 partially inhibited the decrease in cell viability induced by high dose of LPS (Fig. 2B).

Fig. 2.

Effects of BT75 on the cytotoxicity and cell viability of SIM-A9 cells stimulated by LPS. (A) Cell viability of SIM-A9 cells treated with BT75 (500, 250, 100 and 50 μM) for 24 h. The results are expressed as the mean ± SD (n = 8). **p < 0.01 as compared with control group. (B) The cell viability of SIM-A9 cells pretreated with BT75 (25, 10, 5 and 1 μM) for 30 min followed by treatment with LPS (1 μg/mL) for another 24 h. Cell viability was assessed using the MTT assay. The results are expressed as the mean ± SD (n = 8). ##p < 0.01 as compared with control group, **p < 0.01 as compared with LPS-treated group.

3.2. Effects of BT75 on NO production and inflammatory factors induced by LPS in SIM-A9 cells

SIM-A9 cells were pretreated with BT75 (25, 10, 5 and 1 μM) for 30 min followed by treatment with LPS (100 ng/mL) for another 24 h. LPS treatment increased NO production in the cell culture medium, and BT75 inhibited NO production at a dose-dependent manner (Fig. 3A). TNF-α and IL-1β levels detected by ELISA in the cell culture medium also increased in the LPS group, and BT75 at the concentrations of 25 and 10 μM was effective in decreasing IL-1β levels induced by LPS (Fig. 3C), but not TNF-α levels at these concentrations (Fig. 3B). In contrast, IL-10 levels in the cell culture medium did not change by LPS, but significantly increased by BT75 treatment (Fig. 3D).

Fig. 3.

Effects of BT75 on NO production and inflammatory factors induced by LPS in SIM-A9 cells. SIM-A9 cells were pretreated with BT75 (25, 10, 5 and 1 μM) for 30 min followed by treatment with LPS (100 ng/mL) for another 24 h. (A) NO production in the cell culture medium was assessed by the Griess method. TNF-α (B), IL-1β (C), and IL-10 (D) levels in the cell culture medium were detected by ELISA. The results are expressed as the mean ± SD (n = 4–5). ##p < 0.01 as compared with control group, *p < 0.05 and **p<0.01 as compared with LPS-treated group.

3.3. BT75 may ameliorate inflammation induced by LPS in SIM-A9 cells via Akt/NF-κB pathway

SIM-A9 cells pretreated with BT75 (10 μM) or the vehicle for 30 min were cultured in the presence or absence of 100 ng/ml LPS for 24 h. Then, the expression of p-Akt, Akt, p-NF-κB p65, NF-κB p65 and β-actin in the cells were examined by Western blot. Fig. 4 showed the ratio of p-Akt/Akt and p-NF-κB p65/NF-κB p65 increased in the LPS group as compared with control group, and BT75 prevented the increase.

Fig. 4.

BT75 may ameliorate inflammation induced by LPS in SIM-A9 cells via regulation of the AKT/NF-κB pathway. Cells were pretreated with BT75 (10μM) for 30 min before 100 ng/ml LPS treatment for 24 h. Akt/p-AKT and NF-κb/p- NF-κb expression in the cells were detected by Western blot. (A) Represented immunoblot bands for p-Akt, Akt, p- NF-κB, NF-κB, and β-actin. (B and C) Quantitative analysis of immunoblot bands. Protein expression levels were normalized to β-actin. The data were expressed as mean ± SD (n = 3–4). ##p < 0.01 as compared with control group, *p < 0.05 as compared with LPS-treated group.

3.4. BT75 may ameliorate inflammation induced by LPS in SIM-A9 cells via promoting M1 to M2 phenotypic polarization

SIM-A9 cells pretreated with BT75 (10 μM) or the vehicle for 30 min were further incubated with 100 ng LPS for 24 h, and iNOS, IL-6, Arg1, and CD206 expression in the cells were analyzed by Western blot. The results showed that LPS treatment increased the expression of iNOS and IL-6, while BT75 inhibited iNOS and IL-6 expression (Fig. 5A, B and C). Interestingly, BT75 alone totally inhibited iNOS expression (Fig. 5A and B). At the same time, we found that Arg1 and CD206 expression decreased in the LPS group, and BT75 inhibited this reduction (Fig. 5A, D and E). iNOS and IL-6 are considered M1 markers, and Arg1 and CD206 are considered M2 markers. These results, combined with the results of ELISA for IL-1β and IL-10 expression (Fig. 3), suggest that BT75 promoted microglia from M1 to M2 phenotypic polarization.

Fig. 5.

BT75 ameliorated inflammation induced by LPS in SIM-A9 cells via promoting M1-M2 phenotypic polarization. Cells were pretreated with BT75 (10 μM) for 30 min before 100 ng/ml LPS treatment for 24 h. iNOS, IL-6, Arg1, and CD206 expression in the cells were detected by Western blot. (A) Representative immunoblots for iNOS, IL-6, Arg-1, CD206, and β-actin. (B to E) Quantitative analysis of immunoblot bands. Protein expression levels were normalized to β-actin. The data were expressed as mean ± SD (n = 3–4). ##p < 0.01 as compared with control group, **p < 0.01 as compared with LPS-treated group.

3.5. BMS195614, a selective RARα antagonist, partially blocked the inhibition of NO production by BT75 in SIM-A9 cells activated by LPS

BMS195614 (10 μM and 20 μM) was administrated for 24 h and MTT assay showed that 20 μM BMS195614 had a little toxic effect on cells (Fig. 6A). The cells were pretreated with BMS195614 (10 μM and 20 μM) for 30 min, then treated with 100 ng LPS for another 24 h, and NO production in the cell medium was measured. The results showed that BMS195614 had no effect on NO production in LPS-stimulated SIM-A9 cells (Fig. 6B). In Fig. 6C, after pretreatment with BMS195614 (10 μM) for 30 min, cells were incubated in the presence of BT75 (10μM) or the vehicle for 30 min, and further treated with 100 ng LPS for another 24 h. The result indicates that BMS195614 incompletely but significantly blocked the inhibition of NO production by BT75 in LPS-stimulated SIM-A9 cells, suggesting that BT75 inhibited LPS-induced NO production at least partially through the RARα activation.

Fig. 6.

BMS195614, an RARα antagonist, partially blocked the inhibition of NO production by BT75 in LPS-activated SIM-A9 cells. Cells were pretreated with BMS195614 (10 μM) for 30 min and then incubated with BT75 (10 μM) for another 30 min. Then, LPS (100 ng/ml) was added and incubated for 24 h. (A) The cytotoxicity test of BMS195614. (B) Effect of BMS195614 on NO production in LPS-induced SIM-A9 cells. (C) BMS195614 partially blocked the inhibition of NO production by BT75 in LPS-activated SIM-A9 cells. The data were expressed as mean ± SD (n = 4). ##p < 0.01 as compared with control group, **p < 0.01 as compared with LPS-treated group. &&p<0.01 as compared to LPS+BT75 group.

3.6. BT75 inhibited icv-STZ-induced glial activation in mouse brain hippocampus

Double Immunofluorescence staining on mouse brain sections was performed for the detection of GFAP-positive astrocytes and Iba1-positive microglia in the dorsal hippocampus area. In Fig. 7A, brighter staining by GFAP was shown in the hippocampus in STZ group compared to control group, while BT75 inhibited this increase. BT75 alone was not different from the control group. Fig. 7A also showed that STZ increased Iba-1 positive microglia especially in the fimbria of hippocampus, while BT75 treatment decreased STZ-stimulated Iba-1 positive microglia. As compared to control group, BT75 alone had no effect on Iba-1 positive microglia. In the fimbria, STZ increased rod or amoeboid-like Iba1-positive cells with large cell body, and those seemingly activated microglia were reduced by BT75 treatment (Fig. 7B). The quantitative measurement of the densities of the activated cells in the fimbria (Fig. 7C) indicates that BT75 significantly reduced the activated cell densities. We also measured densities of GFAP+ astrocytes in the fimbria (Fig. 7D), which indicate that BT75 reduced GFAP+ cell densities augmented by STZ treatment.

Fig. 7.

BT75 reduced icv-STZ-induced glial activation in mouse brain. (A) GFAP-positive astrocytes and Iba1-positive microglia in the dorsal hippocampus are shown here. The representative image from each group, Ctr (n=3), STZ (n=3), STZ+BT75 (n=3), and BT75 (n=4) group, is shown here. The bar indicates 500 μm. (B) Iba1-positive microglia in the hippocampus fimbria are shown. The bar indicates 50 μm. (C) Larger round or rod-like cell densities in the fimbria were compared among the control, STZ, STZ+BT75, and BT75 groups. * indicates that the STZ group is different from all other groups significantly (p<0.05). (D) The densities of GFAP-positive cells were compared among the control, STZ, STZ+BT75, and BT75 groups. * indicates that the STZ group is different from all other groups significantly (p<0.05).

3.7. Effect of BT75 on nNOS, GFAP, Synaptophysin and p-Tau expression in the hippocampus of icv-STZ-treated mice

Western blot was done for the analyses of GFAP, synaptophysin, nNOS and p-Tau. The results indicated that BT75 inhibited glial cell activation and p-Tau expression in the hippocampus of icv-STZ mice. BT75 also increased synaptophysin and decreased nNOS expression in the hippocampus of icv-STZ mice (Fig. 8A–E).

Fig. 8.

Effect of BT75 on nNOS, GFAP, synaptophysin and p-Tau expression in the hippocampus of icv-STZ-treated mice. (A) Representative immunoblot bands for synaptophysin and p-Tau and β-actin. (B and C) Quantitative analysis of immunoblot bands. Protein expression levels were normalized to β-actin. The data were expressed as mean ± SD (n = 3). #p < 0.05 as compared with control group, *p < 0.01 as compared with STZ-treated group.

3.8. BT75 had no protective effect on STZ-induced Neuro-2a cytotoxicity

Neuro-2a cells were treated with BT75 (1mM, 100μM, 10μM and 1 μM) for 24 h and cell viability was detected. The results showed that the concentration of BT75 lower than 100 μM was not toxic to Neuro-2a cells (Fig. 9A). STZ (1mM, 500μM, 100μM, 50μM and 10 μM) was administrated to Neuro-2a cells for 24 h to see the cell damage. In Fig. 9B, the cell viability decreased at a dose dependent manner, and 1mM STZ inhibited cell viability to 74% of the control. Neuro-2a cells were pretreated with BT75 (25, 10, 5 and 1 μM) for 30 min followed by treatment with STZ (1 mM) for another 24 h. The cell viability was determined by MTT. The results showed BT75 had no effect on the decreased cell viability induced by STZ, indicating that BT75 may exert the protective effect not directly through neurons injured by STZ (Fig. 9C).

Fig. 9.

BT75 had no protective effect on STZ-induced Neuro-2a cytotoxicity. (A) Cell viability of Neuro-2a cells treated with BT75 (1mM, 100μm, 10μm and 1 μM) for 24 h. (B) Cell viability of Neuro-2a cells treated with STZ (1mM, 500μm, 100μm, 50μm and 10 μM) for 24 h. (C) The cell viability of Neuro-2a cells pretreated with BT75 (25, 10, 5 and 1 μM) for half an hour followed by treatment with STZ (1 mM) for another 24 h. Cell viability was assessed using the MTT assay. The results are expressed as the mean ± SD (n = 4). **p < 0.01 as compared with control group.

4. Discussion

Several retinoid-based drugs such as tamibarotene, bexarotene and acitretin, which activate RA signaling, have been repurposed for treatments of CNS diseases like AD [33–35]. These drugs have shown promising effects on improvement of AD pathology in mainly animal models by reducing Aβ production and neuroinflammation. To improve the efficacy and safety of those retinoid-based compounds, we synthesized BT75, a boron-containing RARα agonist. BT75, a derivative of the RARα-specific agonist Am580, binds to RARα with a higher affinity and exhibits less toxicity than Am580 and all-trans RA (ATRA) [28]. The use of boron atoms in pharmaceutical drug design has a high potential for discovery of new biological activities. Boron-containing compounds have been shown to have antioxidant and neuroprotective properties in nature by scavenging Reactive Oxygen Species (ROS) to produce carbon-oxygen bond [36]. Coban et. al showed that boron decreases malathion-induced oxidative stress, enhances the antioxidant defense mechanism and regenerates tissues in rats [37]. In this study, we aimed to explore the anti-inflammation effect and its mechanisms of BT75, and its treatment possibility for AD, since recent studies suggest that anti-inflammatory effects of RARα agonists may attenuate AD pathology [38].

We examined anti-inflammatory effects of BT75 using mouse microglial cell line, SIM-A9, because this spontaneously immortalized cell line stably exhibits microglial phenotypes, including expression of Iba1 and CD68, secretion of inflammatory mediators, and phagocytic activity, which are similar to cultured primary microglia [39, 40]. We found that BT75 attenuated NO production, elevation in the iNOS level, and proinflammatory cytokine IL-1β secretion induced by LPS in SIM-A9 cells (Figs. 3 and 5). This agrees with previous studies indicating that retinoic acid (RA) exerts potent anti-inflammatory effects under conditions of neuroinflammation. RA levels decrease in LPS-activated primary cultured microglia by increased protein expression of RA degradative enzymes, and the pharmacological inhibition of intracellular RA degradation or addition of ATRA exerts potent anti-inflammatory effects in these microglia [41, 42]. Also, inflammatory chemokine release of cultured astrocytes treated with LPS is reduced by ATRA [43]. RARα activation significantly decreased TNFα release from Aβ-treated cultured microglia and induces Aβ clearance in these cultures [44]. In AD mouse models, ATRA has been shown to reduce microglial and astrocytic activation, although these effects may be due to reduced Aβ levels by ATRA [44].

Our study also showed that BT75 decreased p-AKT and p-NF-κB p65 expression enhanced by LPS in these cells (Fig. 4). NF-κB is a family of nuclear transcription factors that play a dominant role in the pathogenesis of many chronic inflammatory processes and regulation of immune responses and is involved in the pro-inflammatory activation of microglia and astrocytes [45]. Normally, NF-κB is in the cytoplasm, and once activated, phosphorylated NF-κB dimers are released and translocated to the nucleus, where they induce cytokines, iNOS, and NO production [46]. It has been reported that the levels of NF-κB p65 are significantly elevated in the brains of AD patients [47], and NF-κB activity also increases in AD mouse models [48, 49]. Therefore, inhibition of NF-κB activation in microglia can modulate the progression of AD caused by neuroinflammation. Our results indicate that BT75 inhibited the activation of the NF-κB pathway, thereby inhibiting LPS-induced neuroinflammation in SIM-A9 cells. We also observed that BT75 induced a small but significant decrease in p-Akt augmented by LPS treatment (Fig. 4). Since Akt is shown to be the upstream signal molecule of NF-κB/IκBα [50], BT75 may ameliorate inflammation induced by LPS in SIM-A9 cells via the Akt/NF-κB pathway.

Our results also suggest that BT75 promoted M1-to-M2 phenotypic polarization in LPS-stimulated SIM-A9 cells. Microglia are classified as M1/M2 after activation, which helps to understand the functional status of microglia during the progression of injury and helps us explore new therapeutic strategies, although there are limitations of the M1/M2 classification system because of much higher heterogeneity in functions and morphologies found in microglia [51]. Previous literature supports the dual role of polarized microglial populations in a variety of neurological diseases, such as focal stroke and AD [52, 53]. Since activation of microglia with an anti-inflammatory M2 phenotype leads to brain repair and regeneration, inhibition of M1 activation and promotion of M2 activation is a promising strategy for the treatment of neuroinflammation-related diseases compared to inhibition of microglia activation in general. The present study showed that BT75 increased an M2 marker, Arg1, and decreased M1 marker, iNOS, in SIM-A9 cells activated with LPS (Fig. 5). Since Arg1 and iNOS utilize the same substrate arginine, and Arg1 can effectively outcompete iNOS to reduce NO production, iNOS and Arg1 are considered a typical set of M1 and M2 markers. In addition, our studies indicated that proinflammatory factors, IL-1 and IL-6, elevated by LPS were reduced by BT75 treatment, and levels of IL-10 and CD206, which are considered M2 markers, increased by BT75. These results suggest that BT75 may exert anti-inflammatory activity by promoting M1-to-M2 phenotypic polarization.

BMS195614, a specific RARα antagonist, partially blocked the inhibition of NO production by BT75 in LPS-induced SIM-A9 cells (Fig. 6), indicating that BT75 inhibited LPS-induced NO production at least partially through RARα activation. The cause of the incomplete action of BMS195614 is presently unknown. Experimental conditions such as concentrations of antagonists and length of the preincubation time may influence the result. Furthermore, since RA is indicated to elicit non-canonical (RAR/RXR-independent) activities beyond regulating gene expression [54], the possibility of such RAR/RXR-independent activity of BT75 cannot be eliminated.

In our in vivo experiment, we used 3 month-old male mice injected via icv with STZ (3 mg/kg) once. Although STZ has been originally used peripherally to induce animal models of diabetes, STZ injection via icv at a low dosage to rodents does not alter peripheral glucose levels and is used as an experimental model for sporadic AD. These mice show AD-like features including disturbances in brain insulin signaling and glucose utilization, accumulation of Aβ plaques, tau phosphorylation, inflammation, and progressive cognitive impairments [55–57]. Compared to transgenic mouse models, the icv-STZ model shows more enhanced neuroinflammation, and the effects of anti-inflammatory agents have been tested using the icv-STZ model [58,59]. Therefore, we examined the effects of BT75 in this icv-STZ mouse model.

Immunofluorescence staining showed that BT75 administration inhibited icv-STZ-induced activation of microglia (Iba-1) and astrocytes (GFAP) mainly in the hippocampus (Fig. 7). Microglia are resident macrophages widely distributed in the brain, and Iba-1 is a marker of microglial activation. Reactive astrocytes are characterized by increased expression of glial fibrillary acidic protein (GFAP). Activated microglia and astrocytes are known to be involved in Aβ deposition and neurofibrillary tangle accumulation during AD progression [60, 61]. Western blot experiments also showed that BT75 decreased levels of GFAP enhanced in hippocampus of icv-STZ mice (Fig. 8).

Tau is the major microtubule-associated protein (MAP) of normal mature neurons and is essential for the assembly and stabilization of the structural integrity of microtubules [62]. In the brains of AD patients, tau is more hyperphosphorylated than normal neurons, loses its ability to bind to microtubules, and subsequently aggregates into pairs of helical filaments that mix with straight filaments to form neurofibrillary tangles [63]. Our results showed that BT75 attenuated p-Tau expression in the hippocampus of icv-STZ mice. BT75 also increased synaptophysin suggesting the better synaptic integrity by BT75 and decreased nNOS expression in the hippocampus of icv-STZ mice. Thus, BT75 showed anti-inflammatory activity not only in vitro but also in vivo and reduced some of the AD-like neuropathology observed in icv-STZ mice. However, we may need to increase sample size of the animal experiments to evaluate possible M1-to-M2 phenotypic polarization of microglia by BT75 in vivo. Also, further studies which include female samples, would be necessary because sex differences in immune responses, including sex differences in microglial functionality, have been reported [64].

It has been shown that in vitro STZ decreases cell viability and choline levels, and increases acetylcholinesterase activity, tau phosphorylation and amyloid aggregation in N2a cells, and is used as a model for analyzing diabetes or AD-related cellular processes [65, 66]. We found that BT75 had no significant protective effect against STZ-induced Neuro-2a cytotoxicity. These results indicate that BT75 may not exert its direct neuroprotective effect against STZ-induced toxicity in neurons.

These results, together with the literature indicating that RARα agonists attenuate AD pathology suggest that BT75 would be a promising small molecule for treatment of diseases caused or exacerbated by neuroinflammation [26, 33, 44, 67].

5. Conclusion

BT75, a novel RARα agonist, attenuated the cytotoxicity induced by LPS in SIM-A9 cells and suppressed the releases of NO and IL-1β induced by LPS in SIM-A9 cells. The anti-inflammatory effects of BT75 may be mediated by the AKT/NF-κB pathway. Also, BT75 may promote microglial M1 to M2 polarization, because BT75 elevated Arg1, IL-10, and CD206, and inhibited iNOS and IL-6 expression in LPS-treated SIM-A9 cells. Also, BT75 shows anti-inflammatory effects in the icv-STZ-induced AD mouse model. BT75 reduced densities of GFAP-positive astrocytes and rod or amoeboid-like Iba1-positive cells, which were elevated in the hippocampal fimbria of icv-STZ mice. Furthermore, BT75 ameliorated AD-like pathology by decreasing tau phosphorylation and increasing synaptophysin expression in the icv-STZ mouse hippocampus. Thus, BT75 is a promising anti-inflammatory agent worthy of further studies towards treatment of neuroinflammation-related diseases, such as AD.

Data availability:

All data and materials are available in the article or upon request to corresponding authors.