- Email: [email protected]
The protective effect of formononetin on cognitive impairment in streptozotocin (STZ)-induced diabetic mice
Biomedicine & Pharmacotherapy 106 (2018) 1250–1257
Contents lists available at ScienceDirect
Biomedicine & Pharmacotherapy journal homepage: www.elsevier.com/locate/biopha
The protective effect of formononetin on cognitive impairment in streptozotocin (STZ)-induced diabetic mice Jinchun Wanga, Lei Wanga, Jie Zhoua, Aiping Qina, Zhujing Chenb, a b
T
⁎
Jiangsu Health Vocational College, No 69. Huangshanling Road, Nanjing, 211800, China The Affiliated Jurong Hospital of Jiangsu University, No 60. West Street, Jurong, 212400, China
A R T I C LE I N FO
A B S T R A C T
Keywords: Formononetin STZ Cognitive impairment HMGB1/TLR4/NF-κB signaling NLRP3 inflammasome
The present study was aimed to elucidate the pharmacological effect of Formononetin (FMN) treatment on STZinduced diabetic cognitive dysfunction. The diabetic model was induced by an intraperitoneally injection of 180 mg/kg STZ. The animals were randomly divided into five groups: control group, streptozocin (STZ, 180 mg/ kg) group, STZ + metformin (Met, 200 mg/kg) group, STZ + FMN (25 mg/kg) group, STZ + FMN (50 mg/kg) group. The mice were intragastrically administrated with metformin (Met, 200 mg/kg) or FMN (25, 50 mg/kg) once daily for 6 weeks. The blood glucose content and body weight were examined. Morris water maze test and Y maze test were used to evaluate the learning and memory abilities. The cognitive decline was reversed by regulating superoxide dismutase (SOD), malondialdehyde (MDA), tumor necrosis factor-a (TNF-α), interleukin (IL)-1β, IL-6 in serum and hippocampus. The protein expressions of high mobility group box-1 protein (HMGB1), toll like receptor 4 (TLR4), myeloid differentiating factor 88 (MyD88), inhibitor of NF-κB (IκBα), p-IκBα, nuclear factor kappa-B(NF-κB), p-NF-κB, NOD-like receptor 3(NLRP3), apoptosis-associated speck-like protein containing CARD(ASC) and caspase-1 were detected. Furthermore, the SH-SY5Y cells were exposed to high glucose stimulation, FMN (2.5, 5 and 10 μM) treatment, and glycyrrhizin, the selective inhibitor of HMGB1. After an incubation for 22 h, the SH-SY5Y cells were harvested for detection. As a result, FMN treatment effectively attenuated the body weight, learning and memory abilities, as well as the levels of blood glucose, SOD, MDA, TNF-α, IL-1β, IL-6. FMN administration also downregulated the protein expressions of HMGB1, TLR4, MyD88, pIκB, p-NF-κB, NLRP3, ASC and caspase-1. The inhibition of HMGB1 by glycyrrhizin also confirmed the involvement of HMGB1/TLR4/NF-κB/NLRP3 pathway in high glucose-induced SH-SY5Y cells. In summary, the results suggested that FMN exhibited the protective effect on STZ-induced cognitive impairment possibly via the mediation of HMGB1/TLR4/NF-κB signaling and NLRP3 inflammasome.
1. Introduction Over the past few decades, Diabetes mellitus (DM) has been considered as a common metabolic disorder, which seriously affects people’s life quality [1]. Some neurological complications are frequently observed in the peripheral and central nervous systems, including brain atrophy and cognitive decline [2]. Chronic hyperglycemia shows the glucose metabolism dysregulation in brain, cognitive function deficiency and excessive inflammatory process [3]. Recent studies have indicated that the worsened metabolic-neurocognitive dysfunction in diabetes mellitus patients may be an increased risk of developing Alzheimer’s disease (AD) [4]. Therefore, early discovery and diagnosis of diabetic cognitive impairment are of great significance for preventing and ameliorating the cognitive dysfunction in DM patients.
⁎
Although the pathogenesis of the STZ-induced cognitive disorder was not fully demonstrated, several studies supported the view that the overexpression of inflammatory cytokines including interleukin (IL)-1β, tumor necrosis factor (TNF)-α and interleukin (IL)-6, were related to diabetes-associated cognitive impairment [5]. Thus, the inflammatory response is activated in the process of cognitive impairment of diabetes. In addition, the increased oxidative stress caused by the chronic hyperglycemia also worsens cognitive functions. The generation of reactive oxygen species (ROS) and the accumulation of free radicals participate in the initiation of the oxidative stress damage [6]. High-mobility group box 1 (HMGB1) is well known as a pivotal mediator in neuroinflammation. HMGB1 binds with the cellular receptors including toll-like receptor (TLR)-4 and phosphorylates the nuclear factor kappa B (NF-κB). NF-κB is reported to activate the NLRP3
Corresponding author at: Jurong Hospital of Jiangsu University, No 60. West Street, Jurong, 212400, China. E-mail address: [email protected] (Z. Chen).
https://doi.org/10.1016/j.biopha.2018.07.063 Received 26 November 2017; Received in revised form 13 July 2018; Accepted 13 July 2018 0753-3322/ © 2018 Published by Elsevier Masson SAS.
Biomedicine & Pharmacotherapy 106 (2018) 1250–1257
J. Wang et al.
animals used. The animals were randomly divided into 2 groups: control group (n = 12) and diabetic group. The mice in diabetic group intraperitoneally received a single dose of 180 mg/kg STZ. Simultaneously, the control group intraperitoneally received the same volume of citrate buffer. Three days later, the fasting blood glucose in a tail-vein sample was determined by a glucose analyzer, a value > 15 mM/l was accepted as diabetes. 3 mice were excluded from the experiment. The diabetic animals were randomly divided into four groups (n = 12) according to the blood glucose content: streptozocin (STZ, 180 mg/kg) group, STZ + metformin (Met, 200 mg/kg) group, STZ + FMN (25 mg/kg) group, STZ + FMN (50 mg/kg) group. Then the levels of blood glucose detected previously were taken as the “before treatment glucose”. The mice were intragastrically administrated with metformin (Met, 200 mg/kg) or FMN (25,50 mg/kg) once daily for 6 weeks. After drug treatment, the levels of blood glucose were detected as the “after treatment glucose”. The animals were applied to Morris water maze (MWM) test. Afterwards, the blood samples were collected and mice were sacrificed. The blood was centrifuged for 10 min at 3000 g and then the serum sample was maintained at −80 °C for pending tests. The brain tissues were harvested for biochemical assays and western bolt analysis.
inflammasome and govern the inflammatory reactions [7]. The NLRP3 inflammasome is the member of Nod-like receptor (NLR) family consisting of nucleotide-binding oligomerization domain-like receptor with a pyrin domain(NLRP3), apoptosis-associated speck-like protein(ASC) and caspase-1 [8]. Former literature displayed that NLRP3 was related to the acceleration of cognitive impairment [9]. Trifolium pratense L (Red clover) isoflavones are longstanding-used treatment for alternative hormone therapy. The bioactive isoflavones, such as calycosin, have been proved to modulate alternative hormone therapy and relieve cognitive impairment in cerebral ischemia/reperfusion rats [10]. Formononetin (FMN), the mainly bioactive isoflavone isolated from red clover, exhibits a variety of properties including anti-apoptotic, anti-oxidant and anti-inflammatory activities [11,12]. It was proved that FMN attenuated alloxan-induced diabetes in previous investigation [13]. Former literatures also demonstrated that FMN exerted neuroprotective effect on cerebral ischemia/reperfusion [14,15]. However, the pharmacological effect of FMN on diabetes-induced cognitive impairment has not been elucidated. The present study was conducted to evaluate the effect of FMN on STZ-induced cognitive dysfunction and investigate its potential mechanism. 2. Methods 2.1. Reagents
2.3. Behavioral tests for cognitive evaluation FMN was purchased from National Institutes for Food and Drug Control (purity more than 99%, Beijing, China, the structural formula of Formononetin was presented in Fig. 1). Metformin hydrochloride (Met) was produced by Jiangsu Suzhong Pharmaceutical Group(Taizhou, China). The drugs were dissolved in the vehicle with the final concentration of DMSO less than 0.1% [v/v]. Streptozocin (STZ) was obtained from Sigma (St. Louis, USA) and dissolved in citrate buffer (Sigma, St. Louis, USA). ACCU-CHEX Blood glucose meters and test strip were produced by Roche (Basel, Switzerland). Glycyrrhizin was obtained from Calbiochem (San Diego, USA). TNF-α, IL-6 and IL-1β enzyme-linked immuno-sorbent assay (ELISA) kits were supplied by Elabscience Biotech. Co. Ltd. (Wuhan, China). Malondialdehyde (MDA) and superoxide dismutase (SOD) commercial kits and BCA protein assay kits were purchased from Jiancheng Bioengineering Institute (Nanjing, China). Anti-HMGB1(#6893), anti-TLR4(#14,358), antiMyD88(#4283), anti-p-IκB(#9246), anti-IκB(#9242), anti-p-NF-κB (#3033), anti-NF-κB(#4764), anti-NLRP3(#1510), anti-ASC(#67,824), anti-GAPDH(#5174) antibodies were obtained from Cell Signaling Technology (Danvers, USA). Anti-caspase-1 (ab138483) antibody was supplied from abcam (Cambridge, UK). DMEM/F12 medium (DMEM) was purchased from Life Technologies (Carlsbad, USA).
The Morris water maze (MWM) test was applied to assess the longterm learning and memory functions. The animals were kept in a circular pool. The pool was divided into four quadrants (N, S, E, and W zones, respectively) according to the markers on the edge for location. The maze was a tank (80 cm in radius and 45 cm high) filled with 25 °C water. A 12 cm diameter circular escape platform was immobilized 1 cm below the water surface in the target quadrant. The mice were tested for 5 consecutive days with four training trials per day with a constant interval of 1 h. Before the first trail, each mouse was put on the platform for 15 s, followed with a 90 s free swim and then was assisted to the platform where it remained for another 15 s rest. The mice were gently placed in water at one of the four quadrants randomly. Each trial lasted for 90 s or ended soon if the mouse reached the submerged platform, thus escaping from the water maze. Whether a mouse found or failed to find the platform within 90 s, it was placed on the plat-form for 30 s. The escape latencies from the water maze (finding the submerged escape platform) were recorded. On the sixth day, a 120 s probe trial with the platform removed was conducted. The numbers of target crossings over the previous location of the target platform, the time spent in the target quadrant, the escape latency before reaching the platform were collected by the video tracking equipment (Viewer 2 Tracking Software, Ji Liang Instruments, China).
2.2. Animals and experimental design Male C57BL/6 J mice (aged 8–10 weeks) were housed under a 12 h light/dark cycle at 25 ± 2 °C. All the animals consumed standard water and food pellets ad libitum. All animal experiments were carried out in accordance with the National Institutes of Health Guidelines. All efforts were made to reduce animals suffering and the number of
2.4. Y maze The apparatus, a room without noise, was comprised of three arms (34 cm long, 8 cm wide, and 14.5 cm high). Then three arms are at 120° angles from each other and the central platform was 8 cm × 8 cm × 8 cm). The maze wall and floor are constructed of dark grey plastic. 24 h before testing, the animals were placed in the test room to adapt the environment. The test began when a mouse was placed in one arm the apparatus facing the wall and the sequence and number of arm entries were recorded manually over an 8-min period. The automatic tracking system (CSI) was employed to record number of arm entries and trace the position of the mice in real time. The percentage alternation was defined as [(Number of alternations)/(Total arm entries −2)] ×100%. The Y-maze arms were cleaned with 10% v/ v ethanol between trials to remove odors and residues.
Fig. 1. The structural formula of Formononetin. 1251
Biomedicine & Pharmacotherapy 106 (2018) 1250–1257
J. Wang et al.
Fig. 2. The effect of FMN on blood glucose levels before drug treatment(A), blood glucose levels after drug treatment(B) and body weight(C). The mice intraperitoneally received 180 mg/kg STZ. The diabetic mice were intragastrically administrated with metformin (Met, 200 mg/kg) or FMN (25, 50 mg/kg) once daily for 6 weeks. Data were expressed as means ± SDs. ###p < 0.001 compared with control group. *p < 0.05, **p < 0.01, ***p < 0.001 compared with STZ group.
antibodies overnight at 4 °C. After washing for 3 times, the HRP-labeled secondary antibodies were added to incubate with membranes for 2 h. Immunoblots were visualized using the ECL detection system and a gel imaging system.
2.5. Isolation of hippocampus At the end of the experiments, all the mice were sacrificed and the brains were collected. The hippocampus of each mouse was immediately isolated on an ice box and then kept at −80 °C until analysis. The supernatant of 10% hippocampus homogenate in normal saline (0.9% sodium chloride) was prepared by centrifugation at 12 000 g at 4 °C for 10 min.
2.10. Statistical analysis All data were expressed as means ± SDs. Results were analyzed by analysis of variance (ANOVA) with Tukey multiple comparison tests, or two-way repeated measures ANOVA, followed by Bonferroni multiple comparison tests. P < 0.05 was considered as significant.
2.6. Determinations of MDA and SOD in serum and hippocampus The levels of oxidative stress markers MDA and the activities of antioxidants markers SOD in serum and hippocampus were assessed by spectrophotometry using commercially available kits.
3. Results 2.7. Determinations of cytokines in serum and hippocampus by ELISA 3.1. Effects of FMN on blood glucose levels and body weight The levels of IL-6, IL-1β and TNF-α in serum and hippocampus of STZ-induced mice were determined using ELISA kits. All experimental procedures were performed according to the manufacturer’s instructions.
To examine the effects of FMN on diabetes, the levels of blood glucose were detected. As depicted in Fig. 2A, before FMN treatment, the mice in all STZ-treated groups presented hyperglycemic states than those in normal mice(p < 0.01). After the drug administration, it was statistically satisfied that FMN(p < 0.01 or p < 0.05) or metformin (p < 0.001) evidently decreased blood glucose levels in STZ-induced mice. The group treated with FMN (50 mg/kg) showed lower glucose level compared with the group treated with FMN at 25 mg/kg (Fig. 2B). Additionally, FMN(p < 0.01 or p < 0.05) and Met(p < 0.01) also relieved the body weight reduction caused by STZ challenge(Fig. 2C).
2.8. Cell culture and drug treatment The SH-SY5Y cell line, obtained from cell bank of the Chinese Academy of Sciences, (Shanghai, China), was cultured in DMEM/F12 culture medium containing 10% FBS(Hyclone, South America), 100 IU/ ml penicillin and 100 IU/ml streptomycin. The cells were divided into seven groups: control group, high glucose group, high glucose + glycyrrhizin group, high glucose + FMN (2.5 μM) group, high glucose + FMN (5 μM) group, high glucose + FMN (10 μM) group, and high glucose + FMN (10 μM) + glycyrrhizin group. FMN and glycyrrhizin were dissolved in DMSO and further diluted with high-glucose DMEM/F12 (33 mM glucose, 2% FBS) before treatment. Generally, SHSY5Y cells with 80–90% confluency in log-phase were seeded at 4 × 105/ml for 24 h. The cells were exposed to serum-starvation for 4 h, then the medium was replaced with high-glucose DMEM/F12 (33 mM glucose, 2% FBS) containing FMN(2.5, 5 and 10 μM) or normal DMEM/ F12 medium as recorded by previous literature [16]. 2 h later, the cells were treated with glycyrrhizin (20 μM). After an incubation for 22 h, the SH-SY5Y cells were harvested for detection.
3.2. Effect of FMN on Morris water maze Thereafter, we assessed the spatial learning and memory ability of the mice by performing Morris water maze test. Mice in the STZ-stimulated group had significant impairment in spatial learning ability during the five-day place navigation training because of the longer escape latency compared with the control mice [4 trials/mice/day for 5 days. Effect of day, F (4, 99) = 482.8, p < 0.01. Effect of group, F (4, 99) =7.729 p < 0.01; Effect of group-by-day interaction, F (4, 99) = 0.740, p < 0.05]. While the treatment with Met or FMN (25, 50 mg/ kg) significantly reduced the escape latency. (Fig. 3A–E,H) Next day after the navigation training, the mice were carried out with the probe trial. The mice in STZ group crossed the location for fewer times (p < 0.01) and spent less time in the target quadrant (p < 0.01) as compared with control group. The mice treated with Met (p < 0.01) or FMN (p < 0.01 or p < 0.05) presented more preference for the target quadrant compared with those in STZ group. The Met or FMN (50 mg/kg) group also demonstrated more target crossing numbers as compared with STZ group(p < 0.05). These results suggested that the spatial learning-memory ability of STZ-diabetic mice was ameliorated by the FMN administration. (Fig. 3F,G)
2.9. Western blot analysis The hippocampus tissues were chopped into small pieces and homogenized in lysis buffer containing 1 ml of RIPA and 10 μl of PMSF. The protein was quantified by BCA kit(Beyotime, Nanjing, China). Protein samples were loaded on 10% SDS-polyacrylamide gel electrophoresis and transferred to polyvinylidene fluoride membranes (BioRad). The membranes were blocked with 5% milk for 2 h and washed with TBST for three times. The blots were incubated with primary 1252
Biomedicine & Pharmacotherapy 106 (2018) 1250–1257
J. Wang et al.
Fig. 3. Effect of FMN on cognitive impairment in the MWM test. (A–E) Representative images of the swimming paths in the probe trial. (F) The number of target crossing in 2 min (G) The percentage of time in the target quadrant during the probe trial test. (H)Escape latency during the 5-day place navigation training. Data were expressed as means ± SDs. ##p < 0.01, ###p < 0.001 compared with control group. *p < 0.05, **p < 0.01, ***p < 0.001 compared with STZ group.
induced animals. By contrast, the Met or FMN-treated group presented restored activity of SOD(p < 0.01 or p < 0.05) and reduced content of MDA(p < 0.01 or p < 0.05). The study reflected the role of FMN on oxidative response in STZ-induced cognitive impairment. (Fig. 5B,D)
3.3. Effect of FMN on Y-maze test As depicted in Fig. 4, STZ significantly reduced spontaneous alternation compared with that in control group(p < 0.01), indicating impairments of working memory. However, this decreased spontaneous alternation behavior induced by STZ was effectively relieved by Met (p < 0.01) or FMN (25, 50 mg/kg)(p < 0.01 or p < 0.05), suggesting that the treatment of FMN attenuated STZ-induced cognitive impairment in mice.
3.5. Effects of FMN on inflammatory cytokines in serum and hippocampus As presented in Fig. 6, the results demonstrated significant increases in the levels of TNF-α, IL-1β and IL-6 in serum and hippocampus of the STZ-induced mice (p < 0.001 or p < 0.01). Treatment with Met or FMN (50 mg/kg) significantly decreased the elevated levels of IL-1β and IL-6 in serum(p < 0.01) of cognitive impairment mice, which was more potent than those in 25 mg/kg group(p < 0.05). FMN(50 mg/ kg) or Met treatments were effective in TNF-α intervention(p < 0.001), which was more efficient than that in FMN(25 mg/kg) treated group (Fig. 6A,C,E). Moreover, the Met or FMN(50 mg/kg) administrations markedly inhibited the hippocampal expressions of TNF-α and IL-1β(p < 0.01). Our study showed that FMN(25 mg/kg) administration could also evidently suppress the increases of these inflammatory mediators (p < 0.05). Additionally, the Met and FMN(50 mg/kg) interventions significantly reduced IL-6 levels(p < 0.05) in hippocampus. The data suggested that FMN exerted anti-inflammatory property in STZ-stimulated cognitive impairment in mice. (Fig. 6B,D,F)
3.4. Effects of FMN on lipid peroxidation in serum and hippocampus As illustrated in Fig. 5A,C, the STZ group showed a considerable decrease of oxidative enzyme SOD activity as compared with control group in serum(p < 0.001). The serum level of MDA in the STZ group increased robustly compared with that in control group(p < 0.001). However, the STZ+ Met group and the STZ+ FMN(50 mg/kg) groups displayed increases of SOD(p < 0.001) and reductions of MDA(p < 0.01) in serum. The administration of FMN(50 mg/kg) showed more efficient changes than FMN treatment at 25 mg/kg. In addition, the suppressed activity of SOD(p < 0.01) and elevated level of MDA(p < 0.01) in hippocampus were observed in STZ-
3.6. Effects of FMN on the protein expressions of HMGB1/TLR4/NF-κB pathway and NLRP3 inflammasome To investigate the ameliorating effect on cognitive impairment, the expressions of HMGB1, TLR4, MyD88, p-IκB, IκB, p-NF-κB, NF-κB, NLRP3, ASC and caspase-1 in hippocampus were measured. As shown in Figs. 7 and 8, the expression levels of HMGB1, TLR4, MyD88, p-NFκB, NLRP3, ASC, caspase-1 were obviously increased in the STZ group as compared with those in control group(p < 0.001). However, the expressions of these proteins were considerably down-regulated in both Met and FMN(25 mg/kg)-treated groups, which possessed stronger efficiency than those in FMN(25 mg/kg) group. The phosphorylation of IκB was also enhanced by the administration of Met and FMN(50 mg/ kg)(p < 0.05). These findings suggested that the ameliorated effects of FMN on STZ-induced cognitive impairment mice might be due to the HMGB1/TLR4/NF-κB pathway and NLRP3 inflammasome. To further confirm the involvement of HMGB1/TLR4/NF-κB pathway, we detected the correlated protein expressions in high
Fig. 4. Effect of FMN on STZ-induced learning and memory impairment in Ymaze. The diabetic mice were intragastrically administrated with metformin (Met, 200 mg/kg) or FMN (25, 50 mg/kg) once daily for 6 weeks. Data were expressed as means ± SDs. ##p < 0.01 compared with control group. *p < 0.05, **p < 0.01 compared with STZ group. 1253
Biomedicine & Pharmacotherapy 106 (2018) 1250–1257
J. Wang et al.
Fig. 5. Effect of FMN on MDA in serum(A) and hippocampus(B), as well as SOD in serum(C) and hippocampus(D). The mice intraperitoneally received 180 mg/kg STZ. The diabetic mice were intragastrically administrated with metformin (Met, 200 mg/kg) or FMN (25, 50 mg/kg) once daily for 6 weeks. Data were expressed as means ± SDs. ##p < 0.01, ###p < 0.001 compared with control group. *p < 0.05, **p < 0.01, ***p < 0.001 compared with STZ group.
impairment.
glucose-induced SH-SY5Y cells. The results were depicted in Fig. 9. The high glucose stimulation notably upregulated the protein levels of HMGB1, TLR4, MyD88, p-NF-κB and NLRP3(p < 0.001, or p < 0.01). The high glucose + FMN(5 μM, 10 μM) group effectively inhibited the expressions of HMGB1 and NLRP3(p < 0.05). The FMN(10 μM) treatment also reduced the protein levels of TLR4, p-NF-κB(p < 0.01) and MyD88 (p < 0.05), which were slightly more potent than those in high glucose + FMN(5 μM) group. The incubations with glycyrrhizin, the selective inhibitor of HMGB1, markedly suppressed the expressions of HMGB1, TLR4, MyD88, NLRP3 (p < 0.05) and p-NF-κB (p < 0.01). It was noteworthy that high glucose + FMN(10 μM) + glycyrrhizin group showed more efficiency of HMGB1, p-NF-κB inhibitions than those in high glucose + FMN(10 μM) group (p < 0.05), and more efficiency of HMGB1, TLR4, p-NF-κB and NLRP3 inhibitions than those in high glucose + glycyrrhizin group (p < 0.05). The data confirmed the critical role of HMGB1/TLR4/NF-κB pathway and NLRP3 inflammasome in the protective effect of FMN on diabetic cognitive
4. Discussion In the present study, we investigated the ameliorated effect of FMN on STZ-induced blood glucose and body weight. It was observed that FMN obviously attenuated the cognitive decline in Y maze test and Morris water maze of FMN on STZ-induced diabetic mice. Our study demonstrated that FMN could dramatically prevent oxidative stress and inflammatory response occurring in brain during the process of cognitive dysfunctions. We confirmed that FMN administration resulted in significant improvements on cognitive performance in diabetes through HMGB1/TLR4/NF-κB signaling pathway and NLRP3 inflammasome in vivo and in vitro. Oxidative stress is a key element in the mechanisms of diabetic cognitive decline. SOD is a major antioxidant enzyme converting superoxide radical to hydrogen peroxide [17]. When the antioxidant
Fig. 6. Effects of FMN on IL-1β(A), IL-6(C) and TNF-α(E) in serum, as well as IL-1β(B), IL-6(D) and TNF-α(F) in hippocampus. The mice intraperitoneally received 180 mg/kg STZ. The diabetic mice were intragastrically administrated with metformin (Met, 200 mg/kg) or FMN (25, 50 mg/kg) once daily for 6 weeks. Data were expressed as means ± SDs. ##p < 0.01, ###p < 0.001 compared with control group. *p < 0.05, **p < 0.01, ***p < 0.001 compared with STZ group. 1254
Biomedicine & Pharmacotherapy 106 (2018) 1250–1257
J. Wang et al.
Fig. 7. Effects of FMN on the protein expressions of HMGB1/TLR4/NF-κB pathway in hippocampus. The mice intraperitoneally received 180 mg/kg STZ. The diabetic mice were intragastrically administrated with metformin (Met, 200 mg/kg) or FMN (25, 50 mg/kg) once daily for 6 weeks. Data were expressed as means ± SDs. ## p < 0.01, ###p < 0.001 compared with control group. *p < 0.05, **p < 0.01, ***p < 0.001 compared with STZ group.
stress. ROS overexpression quietly increases the content of MDA, which leads to severe neuronal apoptosis resulting in learning and memory deficiencies [18]. However, our study demonstrated that FMN treatment increased SOD levels and reduced MDA contents in serum and
system is damaged by STZ, ROS is not sufficiently scavenged to mediate oxidative stress. The oxidative stress can induce cell membrane deformation and lipid peroxidation, thus reducing the activity of SOD. MDA is one of the crucial lipid peroxidation production in oxidative
Fig. 8. Effects of FMN on the protein expressions of NLRP3 inflammasome in hippocampus. The mice intraperitoneally received 180 mg/kg STZ. The diabetic mice were intragastrically administrated with metformin (Met, 200 mg/kg) or FMN (25, 50 mg/kg) once daily for 6 weeks. Data were expressed as means ± SDs. ###p < 0.001 compared with control group. *p < 0.05, **p < 0.01 compared with STZ group. 1255
Biomedicine & Pharmacotherapy 106 (2018) 1250–1257
J. Wang et al.
Fig. 9. Effects of FMN on the protein expressions of HMGB1/TLR4/NF-κB/NLRP3 pathway in vitro. The SH-SY5Y cells were exposed to serum-starvation for 4 h, then the medium was replaced with high-glucose DMEM/F12 (33 mM glucose, 2% FBS) containing FMN(2.5, 5 and 10 μM) or normal DMEM/F12 medium. 2 h later, the cells were treated with glycyrrhizin. After an incubation for 22 h, the SH-SY5Y cells were harvested for detection. Data were expressed as means ± SDs. ###p < 0.001, ##p < 0.05 compared with control group. *p < 0.05, **p < 0.01 compared with high glucose group. @p < 0.05 compared with high glucose + FMN(10 μM) group. &p < 0.05 compared with high glucose + glycyrrhizin group.
hippocampus, suggesting its anti-inflammatory effect on a diabetic-associated cognitive impairment model. High mobility group box-1 protein (HMGB1) is considered as an endogenous mediator of inflammatory reactions [24]. HMGB1 amplifies an appropriate inflammatory response through binding to receptor for advanced glycation end products (RAGE) and affecting its upregulation [25]. HMGB1 is a pivotal mediator in neuroinflammation associated with surgery-induced cognitive impairment in aged mice [26]. Extracellular HMGB1 can interact with the toll-like receptor TLR4. After the activation, the expression of TLR4 enhances and induces a potential inflammatory disease [27]. Furthermore, MyD88 is also an essential intracellular adaptor protein in the downstream molecular of TLRs [28]. Once stimulated by TLR4, MyD88 exerts its effects by mediating NF-κB translocation to nuclear and promoting the expressions of pro-inflammatory cytokines, such as TNF-α, IL-1β and IL-6 [29]. The phosphorylation and degradation of IκB are required for the promotion of NF-κB. The activation of NF-κB subsequently promotes the NLRP3 inflammasome [30]. The caspase-1 promotion is required for the inflammasome formation. Upon initiated by stimuli including STZ, NLRP3 proteins polymerize and combine with ASC adaptor, which
hippocampus of STZ-induced mice, suggesting that FMN exerted an ameliorating effect on cognitive dysfunction via suppressing oxidative stress. In addition to the oxidative stress, the inflammatory response is another mechanism involved in the pathogenesis of cognitive function decline. The levels of these inflammatory cytokines play an important role in the development of inflammatory reactions [19]. Of note, TNFα, IL-1β and IL-6 are important pro-inflammatory cytokines in the inflammatory progress. It has been also proved that the inflammatory mediators are closely associated with cognitive impairment [20]. Zhang et al also elicited that intra-hippocampal microinjection of STZ resulted in release of TNF-α and IL-1β [21]. During inflammatory disease procedures, TNF-α may be implicated in the onset of diabetic cognitive dysfunction which induced by Aβ [22]. Besides, Mina et al have found that IL-1β receptor antagonist can relieve sepsis-induced diabetic cognitive impairment [23]. Thus, TNF-α, IL-6 and IL-1β are major inflammatory mediators in regulating inflammatory response procedures. In this study, we observed the up-regulations of these cytokines in STZinduced mice. FMN treatment notably prevented the elevations of proinflammatory cytokines TNF-α, IL-1β and IL-6 levels in serum and 1256
Biomedicine & Pharmacotherapy 106 (2018) 1250–1257
J. Wang et al.
deficient mice, J. Neurosci. 31 (2011) 3871–3879. [8] X. Zhang, Q. Du, Y. Yang, J. Wang, S. Dou, C. Liu, et al., The protective effect of luteolin on myocardial ischemia/reperfusion (I/R) injury through TLR4/NFkappaB/NLRP3 inflammasome pathway, Biomed. Pharmacother. 91 (2017) 1042–1052. [9] M. Li, H. Li, F. Fang, X. Deng, S. Ma, Astragaloside IV attenuates cognitive impairments induced by transient cerebral ischemia and reperfusion in mice via antiinflammatory mechanisms, Neurosci. Lett. 639 (2017) 114–119. [10] Y. Wang, X. Dong, Z. Li, W. Wang, J. Tian, J. Chen, Downregulated RASD1 and upregulated miR-375 are involved in protective effects of calycosin on cerebral ischemia/reperfusion rats, J. Neurol. Sci. 339 (2014) 144–148. [11] F. Jin, C. Wan, W. Li, L. Yao, H. Zhao, Z. Yuan, et al., Formononetin protects against acetaminophen-induced hepatotoxicity through enhanced NRF2 activity, Plos One 12 (2017) e0170900. [12] Lizhi Chen, Ou Shanshan, Lingqi Zhou, Hai Tang, Xu Jie, Kaihua Guo, Formononetin attenuates Aβ25-35-induced cytotoxicity in HT22 cells via PI3K/Akt signaling and non-amyloidogenic cleavage of APP, Neurosci. Lett. 639 (2017) 36–42. [13] G. Qiu, W. Tian, M. Huan, J. Chen, H. Fu, Formononetin exhibits anti-hyperglycemic activity in alloxan-induced type 1 diabetic mice, Exp. Biol. Med. (Maywood) 242 (2017) 223–230. [14] H. Zhu, L. Zou, J. Tian, F. Lin, J. He, J. Hou, Protective effects of sulphonated formononetin in a rat model of cerebral ischemia and reperfusion injury, Planta Med. 80 (2014) 262. [15] K. Liang, Y. Yu, W. Yong, J. Zhang, C. Li, Formononetin mediates neuroprotection against cerebral ischemia/reperfusion in rats via downregulation of the Bax/Bcl-2 ratio and upregulation PI3K/Akt signaling pathway, J. Neurol. Sci. 344 (2014) 100–104. [16] Z. Yin, H. Yu, S. Chen, C. Ma, X. Ma, L. Xu, et al., Asiaticoside attenuates diabetesinduced cognition deficits by regulating PI3K/Akt/NF-κB pathway, Behav. Brain Res. 292 (2015) 288–299. [17] S. Nurdiana, Y.M. Goh, A. Hafandi, S.M. Dom, A.N. Syimal’Ain, N.M.N. Syaffinaz, et al., Improvement of spatial learning and memory, cortical gyrification patterns and brain oxidative stress markers in diabetic rats treated withFicus deltoidealeaf extract and vitexin, J. Trad. Complement. Med. 8 (2018) 190–202. [18] C. Jiao, F. Gao, L. Ou, J. Yu, M. Li, P. Wei, et al., Tetrahydroxy stilbene glycoside (TSG) antagonizes abeta-induced hippocampal neuron injury by suppressing mitochondrial dysfunction via Nrf2-dependent HO-1 pathway, Biomed. Pharmacother. 96 (2017) 222–228. [19] L.A. Johnson, K.L. Zuloaga, T.L. Kugelman, K.S. Mader, J.T. Morre, D.G. Zuloaga, et al., Amelioration of metabolic syndrome-associated cognitive impairments in mice via a reduction in dietary fat content or infusion of non-diabetic plasma, EBioMedicine 3 (2016) 26–42. [20] J. Zhang, C. Yu, X. Zhang, H. Chen, J. Dong, W. Lu, et al., Porphyromonas gingivalis lipopolysaccharide induces cognitive dysfunction, mediated by neuronal inflammation via activation of the TLR4 signaling pathway in C57BL/6 mice, J. Neuroinflamm. 15 (2018) 37. [21] C.T. Zhang, J.R. Lin, F. Wu, A. Ghosh, S.S. Tang, M. Hu, et al., Montelukast ameliorates streptozotocin-induced cognitive impairment and neurotoxicity in mice, Neurotoxicology 57 (2016) 214–222. [22] C. Yang, B. Zhu, J. Ding, Z.G. Wang, Isoflurane anesthesia aggravates cognitive impairment in streptozotocin-induced diabetic rats, Int. J. Clin. Exp. Med. 7 (2014) 903–910. [23] F. Mina, C.M. Comim, D. Dominguini, O.J. Cassol Jr, D.M. Dall Igna, G.K. Ferreira, et al., Il1-beta involvement in cognitive impairment after sepsis, Mol. Neurobiol. 49 (2014) 1069–1076. [24] J.S. Thinschmidt, L.M. Colon-Perez, M. Febo, S. Caballero, M.A. King, F.A. White, et al., Depressed basal hypothalamic neuronal activity in type-1 diabetic mice is correlated with proinflammatory secretion of HMBG1, Neurosci. Lett. 615 (2016) 21–27. [25] E. Song, J.W. Jahng, L.P. Chong, H.K. Sung, M. Han, C. Luo, et al., Lipocalin-2 induces NLRP3 inflammasome activation via HMGB1 induced TLR4 signaling in heart tissue of mice under pressure overload challenge, Am. J. Transl. Res. 9 (2017) 2723–2735. [26] Z.H. Kong, X. Chen, H.P. Hua, L. Liang, L.J. Liu, The Oral pretreatment of glycyrrhizin prevents surgery-induced cognitive impairment in aged mice by reducing neuroinflammation and Alzheimer’s-related pathology via HMGB1 inhibition, J. Mol. Neurosci. 63 (2017) 385–395. [27] S.H. Zhu, B.Q. Liu, M.J. Hao, Y.X. Fan, C. Qian, P. Teng, et al., Paeoniflorin suppressed high glucose-induced retinal microglia MMP-9 expression and inflammatory response via inhibition of TLR4/NF-kappaB pathway through upregulation of SOCS3 in diabetic retinopathy, Inflammation 40 (2017) 1475–1486. [28] K. Ohno, A. Kuno, H. Murase, S. Muratsubaki, T. Miki, M. Tanno, et al., Diabetes increases the susceptibility to acute kidney injury after myocardial infarction through augmented activation of renal Toll-like receptors in rats, Am. J. Physiol. Heart Circ. Physiol. 313 (2017) H1130–H1142. [29] D. Xueyang, M. Zhanqiang, M. Chunhua, H. Kun, Fasudil, an inhibitor of Rho-associated coiled-coil kinase, improves cognitive impairments induced by smoke exposure, Oncotarget 7 (2016) 78764–78772. [30] G. Chen, W. Sun, Y. Liang, T. Chen, W. Guo, W. Tian, Maternal diabetes modulates offspring cell proliferation and apoptosis during odontogenesis via the TLR4/NFkappaB signalling pathway, Cell Prolif. 50 (2017). [31] P. Xiang, T. Chen, Y. Mou, H. Wu, P. Xie, G. Lu, et al., NZ suppresses TLR4/NF-κB signalings and NLRP3 inflammasome activation in LPS-induced RAW264.7 macrophages, Inflamm. Res. 64 (2015) 799–808. [32] J. Zhuang, X. Wen, Y.Q. Zhang, Q. Shan, Z.F. Zhang, G.H. Zheng, et al., TDP-43 upregulation mediated by the NLRP3 inflammasome induces cognitive impairment in 2 2’,4,4’-tetrabromodiphenyl ether (BDE-47)-treated mice, Brain Behav. Immun. 65 (2017) 99–110.
Fig. 10. The pathway illustration.
consequently induce the translocation and activation of caspase-1. Caspase-1 is reported to trigger pro-inflammatory cytokines allowing for the mature form secretions. NLRP3 inflammasome is the critical inflammatory target for diabetes [31]. Evidence has emerged indicating that knocking down NLRP3 in the hippocampus attenuates the cognitive impairment [32]. Our data displayed that HMGB1, TLR4, MyD88, NF-κB, NLRP3, ASC and caspase-1 were implicated in the pathological process of FMN-mediated diabetic cognitive dysfunction in STZ-induced mice. The inhibition of HMGB1 using glycyrrhizin in high glucose-stimulated SH-SY5Y cells further confirmed the role of HMGB1/TLR4/NFκB pathway and NLRP3 inflammasome in vitro. The pathway illustration was presented in Fig. 10. In conclusion, our research provided evidences that FMN exhibited protective effects on STZ-induced diabetic cognitive impairment by attenuating oxidative stress and inflammatory response through HMGB1/TLR4/NF-κB signaling and NLRP3 inflammasome. It is suggested to further elucidate the mechanism by conducting the experiment in HMGB1−/− transgenic animal. Further researches are warranted before clinical application. References [1] Y.J. Wu, C.C. Lin, C.M. Yeh, M.E. Chien, M.C. Tsao, P. Tseng, et al., Repeated transcranial direct current stimulation improves cognitive dysfunction and synaptic plasticity deficit in the prefrontal cortex of streptozotocin-induced diabetic rats, Brain Stimul. 10 (2017) 1079–1087. [2] P. Sanguanmoo, P. Tanajak, S. Kerdphoo, T. Jaiwongkam, W. Pratchayasakul, N. Chattipakorn, et al., SGLT2-inhibitor and DPP-4 inhibitor improve brain function via attenuating mitochondrial dysfunction, insulin resistance, inflammation, and apoptosis in HFD-induced obese rats, Toxicol. Appl. Pharmacol. 333 (2017) 43–50. [3] Y.B. Li, W.H. Zhang, H.D. Liu, Z. Liu, S.P. Ma, Protective effects of Huanglian Wendan decoction aganist cognitive deficits and neuronal damages in rats with diabetic encephalopathy by inhibiting the release of inflammatory cytokines and repairing insulin signaling pathway in hippocampus, Chin. J. Nat. Med. 14 (2016) 813–822. [4] Q. Wang, J. Yuan, Z. Yu, L. Lin, Y. Jiang, Z. Cao, et al., FGF21 attenuates High-fat diet-induced cognitive impairment via metabolic regulation and anti-inflammation of obese mice, Mol. Neurobiol. (2017) 1–16. [5] C. Ma, H. Long, Protective effect of betulin on cognitive decline in streptozotocin (STZ)-induced diabetic rats, Neurotoxicology 57 (2016) 104–111. [6] T. Baluchnejadmojarad, Z. Kiasalari, S. Afshin-Majd, Z. Ghasemi, M. Roghani, Sallyl cysteine ameliorates cognitive deficits in streptozotocin-diabetic rats via suppression of oxidative stress, inflammation, and acetylcholinesterase, Eur. J. Pharmacol. 794 (2017) 69–76. [7] D.A. Costello, M.B. Watson, T.R. Cowley, N. Murphy, C. Murphy Royal, C. Garlanda, et al., Interleukin-1alpha and HMGB1 mediate hippocampal dysfunction in SIGIRR-
1257
Contents lists available at ScienceDirect
Biomedicine & Pharmacotherapy journal homepage: www.elsevier.com/locate/biopha
The protective effect of formononetin on cognitive impairment in streptozotocin (STZ)-induced diabetic mice Jinchun Wanga, Lei Wanga, Jie Zhoua, Aiping Qina, Zhujing Chenb, a b
T
⁎
Jiangsu Health Vocational College, No 69. Huangshanling Road, Nanjing, 211800, China The Affiliated Jurong Hospital of Jiangsu University, No 60. West Street, Jurong, 212400, China
A R T I C LE I N FO
A B S T R A C T
Keywords: Formononetin STZ Cognitive impairment HMGB1/TLR4/NF-κB signaling NLRP3 inflammasome
The present study was aimed to elucidate the pharmacological effect of Formononetin (FMN) treatment on STZinduced diabetic cognitive dysfunction. The diabetic model was induced by an intraperitoneally injection of 180 mg/kg STZ. The animals were randomly divided into five groups: control group, streptozocin (STZ, 180 mg/ kg) group, STZ + metformin (Met, 200 mg/kg) group, STZ + FMN (25 mg/kg) group, STZ + FMN (50 mg/kg) group. The mice were intragastrically administrated with metformin (Met, 200 mg/kg) or FMN (25, 50 mg/kg) once daily for 6 weeks. The blood glucose content and body weight were examined. Morris water maze test and Y maze test were used to evaluate the learning and memory abilities. The cognitive decline was reversed by regulating superoxide dismutase (SOD), malondialdehyde (MDA), tumor necrosis factor-a (TNF-α), interleukin (IL)-1β, IL-6 in serum and hippocampus. The protein expressions of high mobility group box-1 protein (HMGB1), toll like receptor 4 (TLR4), myeloid differentiating factor 88 (MyD88), inhibitor of NF-κB (IκBα), p-IκBα, nuclear factor kappa-B(NF-κB), p-NF-κB, NOD-like receptor 3(NLRP3), apoptosis-associated speck-like protein containing CARD(ASC) and caspase-1 were detected. Furthermore, the SH-SY5Y cells were exposed to high glucose stimulation, FMN (2.5, 5 and 10 μM) treatment, and glycyrrhizin, the selective inhibitor of HMGB1. After an incubation for 22 h, the SH-SY5Y cells were harvested for detection. As a result, FMN treatment effectively attenuated the body weight, learning and memory abilities, as well as the levels of blood glucose, SOD, MDA, TNF-α, IL-1β, IL-6. FMN administration also downregulated the protein expressions of HMGB1, TLR4, MyD88, pIκB, p-NF-κB, NLRP3, ASC and caspase-1. The inhibition of HMGB1 by glycyrrhizin also confirmed the involvement of HMGB1/TLR4/NF-κB/NLRP3 pathway in high glucose-induced SH-SY5Y cells. In summary, the results suggested that FMN exhibited the protective effect on STZ-induced cognitive impairment possibly via the mediation of HMGB1/TLR4/NF-κB signaling and NLRP3 inflammasome.
1. Introduction Over the past few decades, Diabetes mellitus (DM) has been considered as a common metabolic disorder, which seriously affects people’s life quality [1]. Some neurological complications are frequently observed in the peripheral and central nervous systems, including brain atrophy and cognitive decline [2]. Chronic hyperglycemia shows the glucose metabolism dysregulation in brain, cognitive function deficiency and excessive inflammatory process [3]. Recent studies have indicated that the worsened metabolic-neurocognitive dysfunction in diabetes mellitus patients may be an increased risk of developing Alzheimer’s disease (AD) [4]. Therefore, early discovery and diagnosis of diabetic cognitive impairment are of great significance for preventing and ameliorating the cognitive dysfunction in DM patients.
⁎
Although the pathogenesis of the STZ-induced cognitive disorder was not fully demonstrated, several studies supported the view that the overexpression of inflammatory cytokines including interleukin (IL)-1β, tumor necrosis factor (TNF)-α and interleukin (IL)-6, were related to diabetes-associated cognitive impairment [5]. Thus, the inflammatory response is activated in the process of cognitive impairment of diabetes. In addition, the increased oxidative stress caused by the chronic hyperglycemia also worsens cognitive functions. The generation of reactive oxygen species (ROS) and the accumulation of free radicals participate in the initiation of the oxidative stress damage [6]. High-mobility group box 1 (HMGB1) is well known as a pivotal mediator in neuroinflammation. HMGB1 binds with the cellular receptors including toll-like receptor (TLR)-4 and phosphorylates the nuclear factor kappa B (NF-κB). NF-κB is reported to activate the NLRP3
Corresponding author at: Jurong Hospital of Jiangsu University, No 60. West Street, Jurong, 212400, China. E-mail address: [email protected] (Z. Chen).
https://doi.org/10.1016/j.biopha.2018.07.063 Received 26 November 2017; Received in revised form 13 July 2018; Accepted 13 July 2018 0753-3322/ © 2018 Published by Elsevier Masson SAS.
Biomedicine & Pharmacotherapy 106 (2018) 1250–1257
J. Wang et al.
animals used. The animals were randomly divided into 2 groups: control group (n = 12) and diabetic group. The mice in diabetic group intraperitoneally received a single dose of 180 mg/kg STZ. Simultaneously, the control group intraperitoneally received the same volume of citrate buffer. Three days later, the fasting blood glucose in a tail-vein sample was determined by a glucose analyzer, a value > 15 mM/l was accepted as diabetes. 3 mice were excluded from the experiment. The diabetic animals were randomly divided into four groups (n = 12) according to the blood glucose content: streptozocin (STZ, 180 mg/kg) group, STZ + metformin (Met, 200 mg/kg) group, STZ + FMN (25 mg/kg) group, STZ + FMN (50 mg/kg) group. Then the levels of blood glucose detected previously were taken as the “before treatment glucose”. The mice were intragastrically administrated with metformin (Met, 200 mg/kg) or FMN (25,50 mg/kg) once daily for 6 weeks. After drug treatment, the levels of blood glucose were detected as the “after treatment glucose”. The animals were applied to Morris water maze (MWM) test. Afterwards, the blood samples were collected and mice were sacrificed. The blood was centrifuged for 10 min at 3000 g and then the serum sample was maintained at −80 °C for pending tests. The brain tissues were harvested for biochemical assays and western bolt analysis.
inflammasome and govern the inflammatory reactions [7]. The NLRP3 inflammasome is the member of Nod-like receptor (NLR) family consisting of nucleotide-binding oligomerization domain-like receptor with a pyrin domain(NLRP3), apoptosis-associated speck-like protein(ASC) and caspase-1 [8]. Former literature displayed that NLRP3 was related to the acceleration of cognitive impairment [9]. Trifolium pratense L (Red clover) isoflavones are longstanding-used treatment for alternative hormone therapy. The bioactive isoflavones, such as calycosin, have been proved to modulate alternative hormone therapy and relieve cognitive impairment in cerebral ischemia/reperfusion rats [10]. Formononetin (FMN), the mainly bioactive isoflavone isolated from red clover, exhibits a variety of properties including anti-apoptotic, anti-oxidant and anti-inflammatory activities [11,12]. It was proved that FMN attenuated alloxan-induced diabetes in previous investigation [13]. Former literatures also demonstrated that FMN exerted neuroprotective effect on cerebral ischemia/reperfusion [14,15]. However, the pharmacological effect of FMN on diabetes-induced cognitive impairment has not been elucidated. The present study was conducted to evaluate the effect of FMN on STZ-induced cognitive dysfunction and investigate its potential mechanism. 2. Methods 2.1. Reagents
2.3. Behavioral tests for cognitive evaluation FMN was purchased from National Institutes for Food and Drug Control (purity more than 99%, Beijing, China, the structural formula of Formononetin was presented in Fig. 1). Metformin hydrochloride (Met) was produced by Jiangsu Suzhong Pharmaceutical Group(Taizhou, China). The drugs were dissolved in the vehicle with the final concentration of DMSO less than 0.1% [v/v]. Streptozocin (STZ) was obtained from Sigma (St. Louis, USA) and dissolved in citrate buffer (Sigma, St. Louis, USA). ACCU-CHEX Blood glucose meters and test strip were produced by Roche (Basel, Switzerland). Glycyrrhizin was obtained from Calbiochem (San Diego, USA). TNF-α, IL-6 and IL-1β enzyme-linked immuno-sorbent assay (ELISA) kits were supplied by Elabscience Biotech. Co. Ltd. (Wuhan, China). Malondialdehyde (MDA) and superoxide dismutase (SOD) commercial kits and BCA protein assay kits were purchased from Jiancheng Bioengineering Institute (Nanjing, China). Anti-HMGB1(#6893), anti-TLR4(#14,358), antiMyD88(#4283), anti-p-IκB(#9246), anti-IκB(#9242), anti-p-NF-κB (#3033), anti-NF-κB(#4764), anti-NLRP3(#1510), anti-ASC(#67,824), anti-GAPDH(#5174) antibodies were obtained from Cell Signaling Technology (Danvers, USA). Anti-caspase-1 (ab138483) antibody was supplied from abcam (Cambridge, UK). DMEM/F12 medium (DMEM) was purchased from Life Technologies (Carlsbad, USA).
The Morris water maze (MWM) test was applied to assess the longterm learning and memory functions. The animals were kept in a circular pool. The pool was divided into four quadrants (N, S, E, and W zones, respectively) according to the markers on the edge for location. The maze was a tank (80 cm in radius and 45 cm high) filled with 25 °C water. A 12 cm diameter circular escape platform was immobilized 1 cm below the water surface in the target quadrant. The mice were tested for 5 consecutive days with four training trials per day with a constant interval of 1 h. Before the first trail, each mouse was put on the platform for 15 s, followed with a 90 s free swim and then was assisted to the platform where it remained for another 15 s rest. The mice were gently placed in water at one of the four quadrants randomly. Each trial lasted for 90 s or ended soon if the mouse reached the submerged platform, thus escaping from the water maze. Whether a mouse found or failed to find the platform within 90 s, it was placed on the plat-form for 30 s. The escape latencies from the water maze (finding the submerged escape platform) were recorded. On the sixth day, a 120 s probe trial with the platform removed was conducted. The numbers of target crossings over the previous location of the target platform, the time spent in the target quadrant, the escape latency before reaching the platform were collected by the video tracking equipment (Viewer 2 Tracking Software, Ji Liang Instruments, China).
2.2. Animals and experimental design Male C57BL/6 J mice (aged 8–10 weeks) were housed under a 12 h light/dark cycle at 25 ± 2 °C. All the animals consumed standard water and food pellets ad libitum. All animal experiments were carried out in accordance with the National Institutes of Health Guidelines. All efforts were made to reduce animals suffering and the number of
2.4. Y maze The apparatus, a room without noise, was comprised of three arms (34 cm long, 8 cm wide, and 14.5 cm high). Then three arms are at 120° angles from each other and the central platform was 8 cm × 8 cm × 8 cm). The maze wall and floor are constructed of dark grey plastic. 24 h before testing, the animals were placed in the test room to adapt the environment. The test began when a mouse was placed in one arm the apparatus facing the wall and the sequence and number of arm entries were recorded manually over an 8-min period. The automatic tracking system (CSI) was employed to record number of arm entries and trace the position of the mice in real time. The percentage alternation was defined as [(Number of alternations)/(Total arm entries −2)] ×100%. The Y-maze arms were cleaned with 10% v/ v ethanol between trials to remove odors and residues.
Fig. 1. The structural formula of Formononetin. 1251
Biomedicine & Pharmacotherapy 106 (2018) 1250–1257
J. Wang et al.
Fig. 2. The effect of FMN on blood glucose levels before drug treatment(A), blood glucose levels after drug treatment(B) and body weight(C). The mice intraperitoneally received 180 mg/kg STZ. The diabetic mice were intragastrically administrated with metformin (Met, 200 mg/kg) or FMN (25, 50 mg/kg) once daily for 6 weeks. Data were expressed as means ± SDs. ###p < 0.001 compared with control group. *p < 0.05, **p < 0.01, ***p < 0.001 compared with STZ group.
antibodies overnight at 4 °C. After washing for 3 times, the HRP-labeled secondary antibodies were added to incubate with membranes for 2 h. Immunoblots were visualized using the ECL detection system and a gel imaging system.
2.5. Isolation of hippocampus At the end of the experiments, all the mice were sacrificed and the brains were collected. The hippocampus of each mouse was immediately isolated on an ice box and then kept at −80 °C until analysis. The supernatant of 10% hippocampus homogenate in normal saline (0.9% sodium chloride) was prepared by centrifugation at 12 000 g at 4 °C for 10 min.
2.10. Statistical analysis All data were expressed as means ± SDs. Results were analyzed by analysis of variance (ANOVA) with Tukey multiple comparison tests, or two-way repeated measures ANOVA, followed by Bonferroni multiple comparison tests. P < 0.05 was considered as significant.
2.6. Determinations of MDA and SOD in serum and hippocampus The levels of oxidative stress markers MDA and the activities of antioxidants markers SOD in serum and hippocampus were assessed by spectrophotometry using commercially available kits.
3. Results 2.7. Determinations of cytokines in serum and hippocampus by ELISA 3.1. Effects of FMN on blood glucose levels and body weight The levels of IL-6, IL-1β and TNF-α in serum and hippocampus of STZ-induced mice were determined using ELISA kits. All experimental procedures were performed according to the manufacturer’s instructions.
To examine the effects of FMN on diabetes, the levels of blood glucose were detected. As depicted in Fig. 2A, before FMN treatment, the mice in all STZ-treated groups presented hyperglycemic states than those in normal mice(p < 0.01). After the drug administration, it was statistically satisfied that FMN(p < 0.01 or p < 0.05) or metformin (p < 0.001) evidently decreased blood glucose levels in STZ-induced mice. The group treated with FMN (50 mg/kg) showed lower glucose level compared with the group treated with FMN at 25 mg/kg (Fig. 2B). Additionally, FMN(p < 0.01 or p < 0.05) and Met(p < 0.01) also relieved the body weight reduction caused by STZ challenge(Fig. 2C).
2.8. Cell culture and drug treatment The SH-SY5Y cell line, obtained from cell bank of the Chinese Academy of Sciences, (Shanghai, China), was cultured in DMEM/F12 culture medium containing 10% FBS(Hyclone, South America), 100 IU/ ml penicillin and 100 IU/ml streptomycin. The cells were divided into seven groups: control group, high glucose group, high glucose + glycyrrhizin group, high glucose + FMN (2.5 μM) group, high glucose + FMN (5 μM) group, high glucose + FMN (10 μM) group, and high glucose + FMN (10 μM) + glycyrrhizin group. FMN and glycyrrhizin were dissolved in DMSO and further diluted with high-glucose DMEM/F12 (33 mM glucose, 2% FBS) before treatment. Generally, SHSY5Y cells with 80–90% confluency in log-phase were seeded at 4 × 105/ml for 24 h. The cells were exposed to serum-starvation for 4 h, then the medium was replaced with high-glucose DMEM/F12 (33 mM glucose, 2% FBS) containing FMN(2.5, 5 and 10 μM) or normal DMEM/ F12 medium as recorded by previous literature [16]. 2 h later, the cells were treated with glycyrrhizin (20 μM). After an incubation for 22 h, the SH-SY5Y cells were harvested for detection.
3.2. Effect of FMN on Morris water maze Thereafter, we assessed the spatial learning and memory ability of the mice by performing Morris water maze test. Mice in the STZ-stimulated group had significant impairment in spatial learning ability during the five-day place navigation training because of the longer escape latency compared with the control mice [4 trials/mice/day for 5 days. Effect of day, F (4, 99) = 482.8, p < 0.01. Effect of group, F (4, 99) =7.729 p < 0.01; Effect of group-by-day interaction, F (4, 99) = 0.740, p < 0.05]. While the treatment with Met or FMN (25, 50 mg/ kg) significantly reduced the escape latency. (Fig. 3A–E,H) Next day after the navigation training, the mice were carried out with the probe trial. The mice in STZ group crossed the location for fewer times (p < 0.01) and spent less time in the target quadrant (p < 0.01) as compared with control group. The mice treated with Met (p < 0.01) or FMN (p < 0.01 or p < 0.05) presented more preference for the target quadrant compared with those in STZ group. The Met or FMN (50 mg/kg) group also demonstrated more target crossing numbers as compared with STZ group(p < 0.05). These results suggested that the spatial learning-memory ability of STZ-diabetic mice was ameliorated by the FMN administration. (Fig. 3F,G)
2.9. Western blot analysis The hippocampus tissues were chopped into small pieces and homogenized in lysis buffer containing 1 ml of RIPA and 10 μl of PMSF. The protein was quantified by BCA kit(Beyotime, Nanjing, China). Protein samples were loaded on 10% SDS-polyacrylamide gel electrophoresis and transferred to polyvinylidene fluoride membranes (BioRad). The membranes were blocked with 5% milk for 2 h and washed with TBST for three times. The blots were incubated with primary 1252
Biomedicine & Pharmacotherapy 106 (2018) 1250–1257
J. Wang et al.
Fig. 3. Effect of FMN on cognitive impairment in the MWM test. (A–E) Representative images of the swimming paths in the probe trial. (F) The number of target crossing in 2 min (G) The percentage of time in the target quadrant during the probe trial test. (H)Escape latency during the 5-day place navigation training. Data were expressed as means ± SDs. ##p < 0.01, ###p < 0.001 compared with control group. *p < 0.05, **p < 0.01, ***p < 0.001 compared with STZ group.
induced animals. By contrast, the Met or FMN-treated group presented restored activity of SOD(p < 0.01 or p < 0.05) and reduced content of MDA(p < 0.01 or p < 0.05). The study reflected the role of FMN on oxidative response in STZ-induced cognitive impairment. (Fig. 5B,D)
3.3. Effect of FMN on Y-maze test As depicted in Fig. 4, STZ significantly reduced spontaneous alternation compared with that in control group(p < 0.01), indicating impairments of working memory. However, this decreased spontaneous alternation behavior induced by STZ was effectively relieved by Met (p < 0.01) or FMN (25, 50 mg/kg)(p < 0.01 or p < 0.05), suggesting that the treatment of FMN attenuated STZ-induced cognitive impairment in mice.
3.5. Effects of FMN on inflammatory cytokines in serum and hippocampus As presented in Fig. 6, the results demonstrated significant increases in the levels of TNF-α, IL-1β and IL-6 in serum and hippocampus of the STZ-induced mice (p < 0.001 or p < 0.01). Treatment with Met or FMN (50 mg/kg) significantly decreased the elevated levels of IL-1β and IL-6 in serum(p < 0.01) of cognitive impairment mice, which was more potent than those in 25 mg/kg group(p < 0.05). FMN(50 mg/ kg) or Met treatments were effective in TNF-α intervention(p < 0.001), which was more efficient than that in FMN(25 mg/kg) treated group (Fig. 6A,C,E). Moreover, the Met or FMN(50 mg/kg) administrations markedly inhibited the hippocampal expressions of TNF-α and IL-1β(p < 0.01). Our study showed that FMN(25 mg/kg) administration could also evidently suppress the increases of these inflammatory mediators (p < 0.05). Additionally, the Met and FMN(50 mg/kg) interventions significantly reduced IL-6 levels(p < 0.05) in hippocampus. The data suggested that FMN exerted anti-inflammatory property in STZ-stimulated cognitive impairment in mice. (Fig. 6B,D,F)
3.4. Effects of FMN on lipid peroxidation in serum and hippocampus As illustrated in Fig. 5A,C, the STZ group showed a considerable decrease of oxidative enzyme SOD activity as compared with control group in serum(p < 0.001). The serum level of MDA in the STZ group increased robustly compared with that in control group(p < 0.001). However, the STZ+ Met group and the STZ+ FMN(50 mg/kg) groups displayed increases of SOD(p < 0.001) and reductions of MDA(p < 0.01) in serum. The administration of FMN(50 mg/kg) showed more efficient changes than FMN treatment at 25 mg/kg. In addition, the suppressed activity of SOD(p < 0.01) and elevated level of MDA(p < 0.01) in hippocampus were observed in STZ-
3.6. Effects of FMN on the protein expressions of HMGB1/TLR4/NF-κB pathway and NLRP3 inflammasome To investigate the ameliorating effect on cognitive impairment, the expressions of HMGB1, TLR4, MyD88, p-IκB, IκB, p-NF-κB, NF-κB, NLRP3, ASC and caspase-1 in hippocampus were measured. As shown in Figs. 7 and 8, the expression levels of HMGB1, TLR4, MyD88, p-NFκB, NLRP3, ASC, caspase-1 were obviously increased in the STZ group as compared with those in control group(p < 0.001). However, the expressions of these proteins were considerably down-regulated in both Met and FMN(25 mg/kg)-treated groups, which possessed stronger efficiency than those in FMN(25 mg/kg) group. The phosphorylation of IκB was also enhanced by the administration of Met and FMN(50 mg/ kg)(p < 0.05). These findings suggested that the ameliorated effects of FMN on STZ-induced cognitive impairment mice might be due to the HMGB1/TLR4/NF-κB pathway and NLRP3 inflammasome. To further confirm the involvement of HMGB1/TLR4/NF-κB pathway, we detected the correlated protein expressions in high
Fig. 4. Effect of FMN on STZ-induced learning and memory impairment in Ymaze. The diabetic mice were intragastrically administrated with metformin (Met, 200 mg/kg) or FMN (25, 50 mg/kg) once daily for 6 weeks. Data were expressed as means ± SDs. ##p < 0.01 compared with control group. *p < 0.05, **p < 0.01 compared with STZ group. 1253
Biomedicine & Pharmacotherapy 106 (2018) 1250–1257
J. Wang et al.
Fig. 5. Effect of FMN on MDA in serum(A) and hippocampus(B), as well as SOD in serum(C) and hippocampus(D). The mice intraperitoneally received 180 mg/kg STZ. The diabetic mice were intragastrically administrated with metformin (Met, 200 mg/kg) or FMN (25, 50 mg/kg) once daily for 6 weeks. Data were expressed as means ± SDs. ##p < 0.01, ###p < 0.001 compared with control group. *p < 0.05, **p < 0.01, ***p < 0.001 compared with STZ group.
impairment.
glucose-induced SH-SY5Y cells. The results were depicted in Fig. 9. The high glucose stimulation notably upregulated the protein levels of HMGB1, TLR4, MyD88, p-NF-κB and NLRP3(p < 0.001, or p < 0.01). The high glucose + FMN(5 μM, 10 μM) group effectively inhibited the expressions of HMGB1 and NLRP3(p < 0.05). The FMN(10 μM) treatment also reduced the protein levels of TLR4, p-NF-κB(p < 0.01) and MyD88 (p < 0.05), which were slightly more potent than those in high glucose + FMN(5 μM) group. The incubations with glycyrrhizin, the selective inhibitor of HMGB1, markedly suppressed the expressions of HMGB1, TLR4, MyD88, NLRP3 (p < 0.05) and p-NF-κB (p < 0.01). It was noteworthy that high glucose + FMN(10 μM) + glycyrrhizin group showed more efficiency of HMGB1, p-NF-κB inhibitions than those in high glucose + FMN(10 μM) group (p < 0.05), and more efficiency of HMGB1, TLR4, p-NF-κB and NLRP3 inhibitions than those in high glucose + glycyrrhizin group (p < 0.05). The data confirmed the critical role of HMGB1/TLR4/NF-κB pathway and NLRP3 inflammasome in the protective effect of FMN on diabetic cognitive
4. Discussion In the present study, we investigated the ameliorated effect of FMN on STZ-induced blood glucose and body weight. It was observed that FMN obviously attenuated the cognitive decline in Y maze test and Morris water maze of FMN on STZ-induced diabetic mice. Our study demonstrated that FMN could dramatically prevent oxidative stress and inflammatory response occurring in brain during the process of cognitive dysfunctions. We confirmed that FMN administration resulted in significant improvements on cognitive performance in diabetes through HMGB1/TLR4/NF-κB signaling pathway and NLRP3 inflammasome in vivo and in vitro. Oxidative stress is a key element in the mechanisms of diabetic cognitive decline. SOD is a major antioxidant enzyme converting superoxide radical to hydrogen peroxide [17]. When the antioxidant
Fig. 6. Effects of FMN on IL-1β(A), IL-6(C) and TNF-α(E) in serum, as well as IL-1β(B), IL-6(D) and TNF-α(F) in hippocampus. The mice intraperitoneally received 180 mg/kg STZ. The diabetic mice were intragastrically administrated with metformin (Met, 200 mg/kg) or FMN (25, 50 mg/kg) once daily for 6 weeks. Data were expressed as means ± SDs. ##p < 0.01, ###p < 0.001 compared with control group. *p < 0.05, **p < 0.01, ***p < 0.001 compared with STZ group. 1254
Biomedicine & Pharmacotherapy 106 (2018) 1250–1257
J. Wang et al.
Fig. 7. Effects of FMN on the protein expressions of HMGB1/TLR4/NF-κB pathway in hippocampus. The mice intraperitoneally received 180 mg/kg STZ. The diabetic mice were intragastrically administrated with metformin (Met, 200 mg/kg) or FMN (25, 50 mg/kg) once daily for 6 weeks. Data were expressed as means ± SDs. ## p < 0.01, ###p < 0.001 compared with control group. *p < 0.05, **p < 0.01, ***p < 0.001 compared with STZ group.
stress. ROS overexpression quietly increases the content of MDA, which leads to severe neuronal apoptosis resulting in learning and memory deficiencies [18]. However, our study demonstrated that FMN treatment increased SOD levels and reduced MDA contents in serum and
system is damaged by STZ, ROS is not sufficiently scavenged to mediate oxidative stress. The oxidative stress can induce cell membrane deformation and lipid peroxidation, thus reducing the activity of SOD. MDA is one of the crucial lipid peroxidation production in oxidative
Fig. 8. Effects of FMN on the protein expressions of NLRP3 inflammasome in hippocampus. The mice intraperitoneally received 180 mg/kg STZ. The diabetic mice were intragastrically administrated with metformin (Met, 200 mg/kg) or FMN (25, 50 mg/kg) once daily for 6 weeks. Data were expressed as means ± SDs. ###p < 0.001 compared with control group. *p < 0.05, **p < 0.01 compared with STZ group. 1255
Biomedicine & Pharmacotherapy 106 (2018) 1250–1257
J. Wang et al.
Fig. 9. Effects of FMN on the protein expressions of HMGB1/TLR4/NF-κB/NLRP3 pathway in vitro. The SH-SY5Y cells were exposed to serum-starvation for 4 h, then the medium was replaced with high-glucose DMEM/F12 (33 mM glucose, 2% FBS) containing FMN(2.5, 5 and 10 μM) or normal DMEM/F12 medium. 2 h later, the cells were treated with glycyrrhizin. After an incubation for 22 h, the SH-SY5Y cells were harvested for detection. Data were expressed as means ± SDs. ###p < 0.001, ##p < 0.05 compared with control group. *p < 0.05, **p < 0.01 compared with high glucose group. @p < 0.05 compared with high glucose + FMN(10 μM) group. &p < 0.05 compared with high glucose + glycyrrhizin group.
hippocampus, suggesting its anti-inflammatory effect on a diabetic-associated cognitive impairment model. High mobility group box-1 protein (HMGB1) is considered as an endogenous mediator of inflammatory reactions [24]. HMGB1 amplifies an appropriate inflammatory response through binding to receptor for advanced glycation end products (RAGE) and affecting its upregulation [25]. HMGB1 is a pivotal mediator in neuroinflammation associated with surgery-induced cognitive impairment in aged mice [26]. Extracellular HMGB1 can interact with the toll-like receptor TLR4. After the activation, the expression of TLR4 enhances and induces a potential inflammatory disease [27]. Furthermore, MyD88 is also an essential intracellular adaptor protein in the downstream molecular of TLRs [28]. Once stimulated by TLR4, MyD88 exerts its effects by mediating NF-κB translocation to nuclear and promoting the expressions of pro-inflammatory cytokines, such as TNF-α, IL-1β and IL-6 [29]. The phosphorylation and degradation of IκB are required for the promotion of NF-κB. The activation of NF-κB subsequently promotes the NLRP3 inflammasome [30]. The caspase-1 promotion is required for the inflammasome formation. Upon initiated by stimuli including STZ, NLRP3 proteins polymerize and combine with ASC adaptor, which
hippocampus of STZ-induced mice, suggesting that FMN exerted an ameliorating effect on cognitive dysfunction via suppressing oxidative stress. In addition to the oxidative stress, the inflammatory response is another mechanism involved in the pathogenesis of cognitive function decline. The levels of these inflammatory cytokines play an important role in the development of inflammatory reactions [19]. Of note, TNFα, IL-1β and IL-6 are important pro-inflammatory cytokines in the inflammatory progress. It has been also proved that the inflammatory mediators are closely associated with cognitive impairment [20]. Zhang et al also elicited that intra-hippocampal microinjection of STZ resulted in release of TNF-α and IL-1β [21]. During inflammatory disease procedures, TNF-α may be implicated in the onset of diabetic cognitive dysfunction which induced by Aβ [22]. Besides, Mina et al have found that IL-1β receptor antagonist can relieve sepsis-induced diabetic cognitive impairment [23]. Thus, TNF-α, IL-6 and IL-1β are major inflammatory mediators in regulating inflammatory response procedures. In this study, we observed the up-regulations of these cytokines in STZinduced mice. FMN treatment notably prevented the elevations of proinflammatory cytokines TNF-α, IL-1β and IL-6 levels in serum and 1256
Biomedicine & Pharmacotherapy 106 (2018) 1250–1257
J. Wang et al.
deficient mice, J. Neurosci. 31 (2011) 3871–3879. [8] X. Zhang, Q. Du, Y. Yang, J. Wang, S. Dou, C. Liu, et al., The protective effect of luteolin on myocardial ischemia/reperfusion (I/R) injury through TLR4/NFkappaB/NLRP3 inflammasome pathway, Biomed. Pharmacother. 91 (2017) 1042–1052. [9] M. Li, H. Li, F. Fang, X. Deng, S. Ma, Astragaloside IV attenuates cognitive impairments induced by transient cerebral ischemia and reperfusion in mice via antiinflammatory mechanisms, Neurosci. Lett. 639 (2017) 114–119. [10] Y. Wang, X. Dong, Z. Li, W. Wang, J. Tian, J. Chen, Downregulated RASD1 and upregulated miR-375 are involved in protective effects of calycosin on cerebral ischemia/reperfusion rats, J. Neurol. Sci. 339 (2014) 144–148. [11] F. Jin, C. Wan, W. Li, L. Yao, H. Zhao, Z. Yuan, et al., Formononetin protects against acetaminophen-induced hepatotoxicity through enhanced NRF2 activity, Plos One 12 (2017) e0170900. [12] Lizhi Chen, Ou Shanshan, Lingqi Zhou, Hai Tang, Xu Jie, Kaihua Guo, Formononetin attenuates Aβ25-35-induced cytotoxicity in HT22 cells via PI3K/Akt signaling and non-amyloidogenic cleavage of APP, Neurosci. Lett. 639 (2017) 36–42. [13] G. Qiu, W. Tian, M. Huan, J. Chen, H. Fu, Formononetin exhibits anti-hyperglycemic activity in alloxan-induced type 1 diabetic mice, Exp. Biol. Med. (Maywood) 242 (2017) 223–230. [14] H. Zhu, L. Zou, J. Tian, F. Lin, J. He, J. Hou, Protective effects of sulphonated formononetin in a rat model of cerebral ischemia and reperfusion injury, Planta Med. 80 (2014) 262. [15] K. Liang, Y. Yu, W. Yong, J. Zhang, C. Li, Formononetin mediates neuroprotection against cerebral ischemia/reperfusion in rats via downregulation of the Bax/Bcl-2 ratio and upregulation PI3K/Akt signaling pathway, J. Neurol. Sci. 344 (2014) 100–104. [16] Z. Yin, H. Yu, S. Chen, C. Ma, X. Ma, L. Xu, et al., Asiaticoside attenuates diabetesinduced cognition deficits by regulating PI3K/Akt/NF-κB pathway, Behav. Brain Res. 292 (2015) 288–299. [17] S. Nurdiana, Y.M. Goh, A. Hafandi, S.M. Dom, A.N. Syimal’Ain, N.M.N. Syaffinaz, et al., Improvement of spatial learning and memory, cortical gyrification patterns and brain oxidative stress markers in diabetic rats treated withFicus deltoidealeaf extract and vitexin, J. Trad. Complement. Med. 8 (2018) 190–202. [18] C. Jiao, F. Gao, L. Ou, J. Yu, M. Li, P. Wei, et al., Tetrahydroxy stilbene glycoside (TSG) antagonizes abeta-induced hippocampal neuron injury by suppressing mitochondrial dysfunction via Nrf2-dependent HO-1 pathway, Biomed. Pharmacother. 96 (2017) 222–228. [19] L.A. Johnson, K.L. Zuloaga, T.L. Kugelman, K.S. Mader, J.T. Morre, D.G. Zuloaga, et al., Amelioration of metabolic syndrome-associated cognitive impairments in mice via a reduction in dietary fat content or infusion of non-diabetic plasma, EBioMedicine 3 (2016) 26–42. [20] J. Zhang, C. Yu, X. Zhang, H. Chen, J. Dong, W. Lu, et al., Porphyromonas gingivalis lipopolysaccharide induces cognitive dysfunction, mediated by neuronal inflammation via activation of the TLR4 signaling pathway in C57BL/6 mice, J. Neuroinflamm. 15 (2018) 37. [21] C.T. Zhang, J.R. Lin, F. Wu, A. Ghosh, S.S. Tang, M. Hu, et al., Montelukast ameliorates streptozotocin-induced cognitive impairment and neurotoxicity in mice, Neurotoxicology 57 (2016) 214–222. [22] C. Yang, B. Zhu, J. Ding, Z.G. Wang, Isoflurane anesthesia aggravates cognitive impairment in streptozotocin-induced diabetic rats, Int. J. Clin. Exp. Med. 7 (2014) 903–910. [23] F. Mina, C.M. Comim, D. Dominguini, O.J. Cassol Jr, D.M. Dall Igna, G.K. Ferreira, et al., Il1-beta involvement in cognitive impairment after sepsis, Mol. Neurobiol. 49 (2014) 1069–1076. [24] J.S. Thinschmidt, L.M. Colon-Perez, M. Febo, S. Caballero, M.A. King, F.A. White, et al., Depressed basal hypothalamic neuronal activity in type-1 diabetic mice is correlated with proinflammatory secretion of HMBG1, Neurosci. Lett. 615 (2016) 21–27. [25] E. Song, J.W. Jahng, L.P. Chong, H.K. Sung, M. Han, C. Luo, et al., Lipocalin-2 induces NLRP3 inflammasome activation via HMGB1 induced TLR4 signaling in heart tissue of mice under pressure overload challenge, Am. J. Transl. Res. 9 (2017) 2723–2735. [26] Z.H. Kong, X. Chen, H.P. Hua, L. Liang, L.J. Liu, The Oral pretreatment of glycyrrhizin prevents surgery-induced cognitive impairment in aged mice by reducing neuroinflammation and Alzheimer’s-related pathology via HMGB1 inhibition, J. Mol. Neurosci. 63 (2017) 385–395. [27] S.H. Zhu, B.Q. Liu, M.J. Hao, Y.X. Fan, C. Qian, P. Teng, et al., Paeoniflorin suppressed high glucose-induced retinal microglia MMP-9 expression and inflammatory response via inhibition of TLR4/NF-kappaB pathway through upregulation of SOCS3 in diabetic retinopathy, Inflammation 40 (2017) 1475–1486. [28] K. Ohno, A. Kuno, H. Murase, S. Muratsubaki, T. Miki, M. Tanno, et al., Diabetes increases the susceptibility to acute kidney injury after myocardial infarction through augmented activation of renal Toll-like receptors in rats, Am. J. Physiol. Heart Circ. Physiol. 313 (2017) H1130–H1142. [29] D. Xueyang, M. Zhanqiang, M. Chunhua, H. Kun, Fasudil, an inhibitor of Rho-associated coiled-coil kinase, improves cognitive impairments induced by smoke exposure, Oncotarget 7 (2016) 78764–78772. [30] G. Chen, W. Sun, Y. Liang, T. Chen, W. Guo, W. Tian, Maternal diabetes modulates offspring cell proliferation and apoptosis during odontogenesis via the TLR4/NFkappaB signalling pathway, Cell Prolif. 50 (2017). [31] P. Xiang, T. Chen, Y. Mou, H. Wu, P. Xie, G. Lu, et al., NZ suppresses TLR4/NF-κB signalings and NLRP3 inflammasome activation in LPS-induced RAW264.7 macrophages, Inflamm. Res. 64 (2015) 799–808. [32] J. Zhuang, X. Wen, Y.Q. Zhang, Q. Shan, Z.F. Zhang, G.H. Zheng, et al., TDP-43 upregulation mediated by the NLRP3 inflammasome induces cognitive impairment in 2 2’,4,4’-tetrabromodiphenyl ether (BDE-47)-treated mice, Brain Behav. Immun. 65 (2017) 99–110.
Fig. 10. The pathway illustration.
consequently induce the translocation and activation of caspase-1. Caspase-1 is reported to trigger pro-inflammatory cytokines allowing for the mature form secretions. NLRP3 inflammasome is the critical inflammatory target for diabetes [31]. Evidence has emerged indicating that knocking down NLRP3 in the hippocampus attenuates the cognitive impairment [32]. Our data displayed that HMGB1, TLR4, MyD88, NF-κB, NLRP3, ASC and caspase-1 were implicated in the pathological process of FMN-mediated diabetic cognitive dysfunction in STZ-induced mice. The inhibition of HMGB1 using glycyrrhizin in high glucose-stimulated SH-SY5Y cells further confirmed the role of HMGB1/TLR4/NFκB pathway and NLRP3 inflammasome in vitro. The pathway illustration was presented in Fig. 10. In conclusion, our research provided evidences that FMN exhibited protective effects on STZ-induced diabetic cognitive impairment by attenuating oxidative stress and inflammatory response through HMGB1/TLR4/NF-κB signaling and NLRP3 inflammasome. It is suggested to further elucidate the mechanism by conducting the experiment in HMGB1−/− transgenic animal. Further researches are warranted before clinical application. References [1] Y.J. Wu, C.C. Lin, C.M. Yeh, M.E. Chien, M.C. Tsao, P. Tseng, et al., Repeated transcranial direct current stimulation improves cognitive dysfunction and synaptic plasticity deficit in the prefrontal cortex of streptozotocin-induced diabetic rats, Brain Stimul. 10 (2017) 1079–1087. [2] P. Sanguanmoo, P. Tanajak, S. Kerdphoo, T. Jaiwongkam, W. Pratchayasakul, N. Chattipakorn, et al., SGLT2-inhibitor and DPP-4 inhibitor improve brain function via attenuating mitochondrial dysfunction, insulin resistance, inflammation, and apoptosis in HFD-induced obese rats, Toxicol. Appl. Pharmacol. 333 (2017) 43–50. [3] Y.B. Li, W.H. Zhang, H.D. Liu, Z. Liu, S.P. Ma, Protective effects of Huanglian Wendan decoction aganist cognitive deficits and neuronal damages in rats with diabetic encephalopathy by inhibiting the release of inflammatory cytokines and repairing insulin signaling pathway in hippocampus, Chin. J. Nat. Med. 14 (2016) 813–822. [4] Q. Wang, J. Yuan, Z. Yu, L. Lin, Y. Jiang, Z. Cao, et al., FGF21 attenuates High-fat diet-induced cognitive impairment via metabolic regulation and anti-inflammation of obese mice, Mol. Neurobiol. (2017) 1–16. [5] C. Ma, H. Long, Protective effect of betulin on cognitive decline in streptozotocin (STZ)-induced diabetic rats, Neurotoxicology 57 (2016) 104–111. [6] T. Baluchnejadmojarad, Z. Kiasalari, S. Afshin-Majd, Z. Ghasemi, M. Roghani, Sallyl cysteine ameliorates cognitive deficits in streptozotocin-diabetic rats via suppression of oxidative stress, inflammation, and acetylcholinesterase, Eur. J. Pharmacol. 794 (2017) 69–76. [7] D.A. Costello, M.B. Watson, T.R. Cowley, N. Murphy, C. Murphy Royal, C. Garlanda, et al., Interleukin-1alpha and HMGB1 mediate hippocampal dysfunction in SIGIRR-
1257