Hydrogen Sulde Ameliorates High Glucose-induced Inammation Through SIRT1-mTOR/NF-κB Signaling Pathway in HT-22 cells

Background: Hyperglycemia-induced neuroinammation promotes the progression of diabetic encephalopathy (DE). Hydrogen sulde (H 2 S) exerts anti-inammatory and neuroprotective activities against neurodegenerative diseases. However, its role in hyperglycemia-induced neuronal inammation has not been investigated. Herein, we examined the effects and its related signaling pathway of H 2 S on inammatory response in high glucose-treated HT-22 cells. Methods: A hippocampal neuronal cell line, HT-22, was used as an in vitro model to explore the function of H 2 S on inammatory response triggered by high glucose. A dicyanoisophorone-based near-infrared uorescent probe (NIR-NP) was synthesized to detect H 2 S levels in HT-22 cells. Western blotting, immunouorescence and real time-qPCR were carried out to study the mechanism of action for H 2 S. Results: We found that high glucose (85 mM) decreased the level of endogenous H 2 S and the expression of cystathionine-β-synthase (CBS) which is the main enzyme for H 2 S production in the brain. Sodium hydrosulde (NaHS, a H 2 S donor) or S-adenosylmethionine (SAMe, an allosteric activator of CBS) administration restored high glucose-induced downregulation of CBS and H 2 S levels. Importantly, high glucose upregulated the level of pro-inammatory factors (IL-1β, IL-6, TNF-α) in HT-22 cells. Treatment with NaHS or SAMe alleviated this enhanced transcription of these pro-inammatory factors, suggesting that H 2 S might ameliorate high glucose-induced inammation in HT-22 cells. We also found that high glucose reduced SIRT1 protein levels. SIRT1 reduction elevated the level of p-mTOR, p-NF-κB and pro-inammatory factors, which were restored by resveratrol (a SIRT1 agonist). These results suggested that SIRT1 might be an upstream mediator of mTOR/NF-κB signaling pathway. Furthermore, NaHS or SAMe treatment reversed the expression of SIRT1, mTOR and NF-κB under high glucose conditions. Conclusions: Our study revealed that high glucose decreased CBS to reduce the production of H 2 S, which in turn decreased the expression of SIRT1. The reduction of SIRT1 activated mTOR/NF-κB signaling to promote inammation. Given that promoting H 2 S production using NaHS or SAMe can reverse high glucose-induced inammatory response, our study might shed light on the prophylactic treatment of DE. RT-qPCR: real-time quantitative polymerase chain reaction; SAMe: S-adenosylmethionine; SIRT1: silent information regulator-1; TBST: tris-buffered saline Tween-20; TNF-α: tumor necrosis factor-α.


Background
Diabetic encephalopathy (DE) is recognized as a complex and elusive complication of diabetes nowadays [1]. Patients with DE exhibit progressive alterations in brain structures and cognitive decline.
These clinical manifestations are primarily caused by hyperglycemia-induced neurotoxicity [2]. Although multiple pathophysiological processes are involved in the progression of DE, accumulating evidence suggests that neuroin ammation plays a critical role in neurodegeneration induced by hyperglycemia [3,4]. In diabetic animals, the expression of neuroin ammatory cytokines is signi cantly increased in the hippocampus, an important brain region for cognition [3]. High glucose increases pro-in ammatory response in the immune cells such as microglia and astrocyte in the central nervous system [5,6].
However, in neurons, changes triggered by high glucose in the expression of in ammatory cytokines and associated signaling molecules are less reported. The underlying molecular mechanisms about how hyperglycemia activates in ammation in the brain remain further elucidate.
Hydrogen sul de (H 2 S) is a noteworthy endogenous gasotransmitter which participates in regulating various pathophysiological processes [7]. During the progression of diabetes, the H 2 S levels declined gradually in the patients' plasma [8,9]. In line with these clinical ndings, preclinical research revealed that hyperglycemia reduced H 2 S biosynthesis and increased its degradation in rodent models of diabetes [10,11]. While, supplementation of H 2 S donor sodium hydrosul de (NaHS) protected hyperglycemiainduced myocardial tissue damage in vivo and in vitro [12]. In the central nervous system, the production of endogenous H 2 S is mainly catalyzed by cystathionine-β-synthase (CBS) [13]. Exogenous H 2 S inhibited the expression of pro-in ammatory factors and decreased the generation of oxidative indicator reactive oxygen species in the mouse model of Alzheimer's disease [14,15]. However, it has not been fully investigated the effects and mechanisms of H 2 S on the in ammation in neurons under hyperglycemia.
Recent research has shown that H 2 S regulated in ammation-associated signaling molecules, such as silent information regulator-1 (SIRT1), mammalian target of rapamycin (mTOR) and nuclear factor-κB (NF-κB). H 2 S upregulated the expression of SIRT1 to prevent homocysteine-induced neurotoxicity in PC12 cells [16]. H 2 S increased SIRT1 levels in the rat hippocampus, which reduced chronic mild stress-induced depressive-like behavior [17]. However, it is unknown if SIRT1 and its downstream signaling in the neuron are interrupted by high glucose. mTOR is a crucial protein kinase involved in regulating multiple diseases including cancer, metabolic and neurological diseases. mTOR signaling was dysregulated in type 2 diabetes as well as in ammation [18]. Upregulation of SIRT1 inhibited the expression of mTOR and subsequently attenuated in ammation in bleomycin-induced scleroderma mice and high-fat-diet mice [19,20], suggesting SIRT1 might be an upstream signaling molecule of mTOR. However, SIRT1-mTOR singling pathway has not been investigated in neurons with high glucose. Activation of mTOR modulated its downstream ubiquitous transcriptional factor NF-κB to facilitate the production of pro-in ammatory cytokines [21,22]. H 2 S plays an inhibitory role in lipopolysaccharide (LPS)-triggered in ammatory response via blocking the transactivation of NF-κB in endothelial cells [23]. Collectively, we predicted that H 2 S might modulate in ammation in neuronal cells by inhibiting the activated SIRT1/mTOR/NF-κB signaling pathway.
In the current study, we applied a neuronal cell line HT-22, which was widely used as an in vitro model of neurodegenerative disease [24,25], to explore the function of high glucose on H 2 S and its endogenous synthase CBS, as well as pro-in ammatory cytokines expression and SIRT1/mTOR/NF-κB signaling pathway. Furthermore, we examined whether NaHS or S-adenosylmethionine (SAMe, an allosteric activator of CBS) could attenuate high glucose-induced in ammatory effect.

Cell line and reagents
The mouse hippocampal neuronal cell line HT-22 cells were kindly gifted from the Research Center for Neurobiology of Xuzhou Medical University (Xuzhou, China). NaHS and SAMe (97% purity) was purchased from Aladdin Reagent Co., Ltd (Shanghai, China) and MedChemExpress (Monmouth Junction, NJ, USA), respectively. Resveratrol and glucose were purchased from Sigma-Aldrich (St. Louis, MO, USA).

Synthesis of H 2 S probe NIR-NP
A dicyanoisophorone-based near-infrared uorescent probe (NIR-NP) for detection of H 2 S was synthesized as described in the previous study [26] and its structure was veri ed by 1 H NMR and HRMS. The spectroscopic properties of NIR-NP probe such as selectivity, sensitivity and stability were validated using a Hitachi F-4600 uorescence spectrophotometer.

Cell culture and treatments
The culture of HT-22 cells was conducted in DMEM medium containing 10% (v/v) FBS and 1% (v/v) penicillin/streptomycin in an incubator (37 ℃, 5% CO 2 ). HT-22 cells were grown in 96-well plate (5×10 3 cells/well) for 24 h and incubated with glucose, SAMe, NaHS, resveratrol and NIR-NP at a range of concentrations for another 24 h, respectively. Then the cells were incubated with MTT (0.1 mg/well) at 37 ℃ for 4 h. The Microplate reader (ELX808IU, Bio-tek Instruments Inc., USA) was used to detect the optical density at 550 nm.

The detection of H 2 S levels in HT-22 cells using NIR-NP
The levels of H 2 S in HT-22 cells were examined using NIR-NP according to the reported method [31]. Cells were treated with high glucose and/or other agents and then collected in PBS buffer. Cells were homogenized and centrifuged at 12,000 g at 4 ℃ for 15 min. The Pierce BCA Protein Assay Kit (Beyotime Institute of Biotechnology, Shanghai, China) was used to determine total protein concentrations of the collected supernatant. The H 2 S content in the treated cells was determined using supernatant spiked with Na 2 S solution at a series of concentrations as the internal criterion. To obtain zero point, ZnCl 2 was added to scavenge the endogenous H 2 S within the sample. In Eppendorf tubes, 20 μL homogenates supernatant and 1 µL 1.0 mM NIR-NP probe were mixed with 69 μL PBS, which was spiked with 0, 0.4, 0.8,1.2, 1.6 µL of 100 µM Na 2 S solution and double distilled water (10, 9.6, 9.2, 8.8, 8.4 µL) correspondingly. After incubation at 37 ℃ for 20 min, the uorescence intensity was detected using the Hitachi F4600 Fluorescence Spectrophotometer. Each sample was measured at least three times in parallel. According to the calibration curve of Na 2 S, the concentration of each sample was calculated and expressed as μmol/g protein.

Real-time quantitative polymerase chain reaction (RT-qPCR)
The TRIzol reagent (Invitrogen Co., Carlsbad, CA, USA) was used for the extraction of total RNA. The determination of total RNA concentration was performed on NanoDrop 1000 spectrophotometer (Thermo Scienti c). ReverTra Ace qPCR RT kit (TOYOBO CO., LTD.) was used to synthesize cDNA. The forward and reverse primers sequences of IL-1β, TNF-α, IL-6 and β-actin (Sangon Biotech Co. Ltd., Shanghai, China) were presented in Table 1. 10 μL reaction solution including SYBR Green MIX, cDNA, forward primer, reverse primer and RNase-free water was centrifuged and subject to LightCycle 480 (Roche Applied Science). LightCycler 480 software was used to determine the relative mRNA levels.
Western BlottingA Pierce BCA Protein Assay Kit was used to determine the supernatant protein concentrations of HT-22 cell lysis buffer. After being electrophoresed in SDS-PAGE gels, the separated proteins were transferred to nitrocellulose membranes, which were blocked at 37 ℃ for 2 h in Trisbuffered saline with 5% BSA and 0.05% Tween 20 and incubated with the primary antibodies against CBS, SIRT1, mTOR, p-mTOR, NF-κB p65, p-NF-κB p65 and β-actin (all antibodies were diluted at 1:1000) at 4 ℃ overnight, respectively. Subsequently, the membranes were incubated with the secondary horseradish peroxidase-linked anti-rabbit (1:10000) or anti-mouse (1:10000) antibodies (ZSGB-BIO, Beijing, China) at 37 ℃ for 1 h. After the membranes were washed with TBST, chemiluminescent reagent was applied and the images were captured using Odyssey infrared uorescence imaging system.Immuno uorescenceHT-22 cells were cultured on glass coverslips in a 24-well plate (1.5×10 4 cells/well). After being rinsed with PBS, HT-22 cells were xed in 4% paraformaldehyde for 15 min, and then penetrated with 0.5% Triton X-100 (PBS) for 5 min and blocked at room temperature in 1 mL of PBS solution containing 5% FBS for 1 h. After blocking, HT-22 cells were incubated at 4 ℃ with the primary antibodies against CBS, SIRT1, p-mTOR and p-NF-κB p65 (All antibodies were diluted at 1:200) overnight and then incubated at room temperature in the dark with 300 μL Alexa Fluor-labeled uorescent antirabbit (1:200) or anti-mouse (1:200) antibodies for 1 h. 300 μL DAPI solution (1 mg/mL) were then applied for the following 5 min. The coverslips were placed on a slide with antifade mounting medium.
Images were acquired using the OLYMPUS upright uorescence microscope and analyzed with Image J.Statistical AnalysisAll results are analyzed by SPSS 16.0 (SPSS, Inc., Chicago, IL, USA). The intergroup differences were assessed by one-way ANOVA accompanied by LSD test or Dunnett's T3 post hoc analysis. Data from all experiments were presented as mean ± SD (n = 3). P < 0.05 represented a statistical signi cance. Table 1 Primer sequences for the RT-qPCR analysis.

Gene
Forward

Results
The e cacy of NIR-NP to detect endogenous H 2 S in HT-22 Cells The compound NIR-NP was successfully synthesized and con rmed by 1 H NMR ( Figure S2) and HRMS ( Figure S3). The probe exhibited excellent properties including an off-on response to H 2 S ( Figure S4), linearity ( Figure S5), fast response time ( Figure S6), high selectivity for H 2 S over other analytes ( Figure   S7) and pH stability ( Figure S8). We then examined the effect of NIP-NP on the cell viability using MTT assay. Different concentrations of NIP-NP (5, 10, 20, 40, and 80 μM) exhibited no signi cant cytotoxicity ( Figure 1A). These results suggest that the NIP-NP probe is less toxic and biocompatible for intracellular bioimaging. As shown in Figure 1B, when HT-22 cells were incubated with NIR-NP and Na 2 S (H 2 S donor), obvious red uorescence in the cytoplasm was observed. When Na 2 S concentration was increased, the uorescent intensity was enhanced, suggesting that NIR-NP was able to detect the H 2 S levels in HT-22 cells.
NaHS and SAMe prevented high glucose-induced reduction of H 2 S and CBS levels in HT-22 cells The probe NIR-NP was used to detect H 2 S contents in HT-22 cells. The H 2 S level in the HG group was evidently decreased than that of the control group (P < 0.001, Figure 2A). SAMe or NaHS application at 30 min prior to high glucose treatment signi cantly increased H 2 S levels compared with the HG group (both P < 0.001). The results indicated that the SAMe and NaHS improved the H 2 S level, which was reduced by the high glucose.
CBS is an important enzyme to catalyze H 2 S [13]. Next, we found that high glucose reduced CBS protein levels in HT-22 cells ( Figure 2B, P < 0.001).The CBS level was obviously elevated in HG + SAMe and HG + NaHS groups compared with the HG group ( Figure 2B, both P < 0.001). The results of the quanti cation of immuno uorescent intensity were consistent with this nding and showed the alteration of CBS in the HT-22 cells ( Figure 2C, P < 0.01, P < 0.001).
High glucose triggered in ammatory response via SIRT1/mTOR/NF-κB p65 pathway Studies have shown that SIRT1 negatively modulates mTOR [19] and NF-κB [32]. Therefore, SIRT1 may serve as an important regulator of high glucose-mediated in ammatory signaling and pro-in ammatory cytokines. RT-qPCR results showed that high glucose upregulated the mRNA expression of IL-6, IL-1β and TNF-α in HT-22 cells (Figure 3, all P < 0.001), while resveratrol (an agonist of SIRT1) inhibited the increase of these pro-in ammatory cytokines ( Figure 3, all P < 0.001).

H 2 S prevented high glucose-induced increase of proin ammatory cytokines
It is reported that the H 2 S level is negatively associated with in ammation [33]. Following the nding that high glucose reduced H 2 S level and increased pro-in ammatory response in neuronal HT-22 cells, we next examined if SAMe or NaHS can prevent these processes. RT-qPCR results showed that IL-6, IL-1β and TNF-α levels were signi cantly increased in high glucose-treated HT-22 cells ( Figure 5, all P < 0.001). Importantly, the upregulation of pro-in ammatory cytokines was attenuated by either SAMe or NaHS treatment ( Figure 5, all P < 0.001), suggesting that increasing H 2 S can prevent high glucose-induced neuroin ammation in HT-22 cells.

Discussion
In this study, the NIR-NP probe was successfully synthesized and showed excellent selectivity and sensitivity for the detection of endogenous H 2 S levels in HT-22 cells. Using this probe, we showed that high glucose reduced the H 2 S levels in neurons. Furthermore, high glucose exposure decreased the level of CBS, SIRT1 and elevated the level of pro-in ammatory cytokines IL-6, IL-1β and TNF-α along with mTOR/NF-κB signaling activation. Importantly, NaHS or SAMe reversed the level of H 2 S, CBS, proin ammatory cytokines and signaling molecules under high glucose conditions. As neuroin ammation occupies a critical position in the development of DE [34], our results might provide novel insight that H 2 S decline and neuroin ammation in neurons are involved in the pathophysiology of DE.
DE is a complication that occurred in long-term diabetes [35]. The hyperglycemia-induced in ammation is an emerging factor that contributes to the progression of DE. In the brain, hyperglycemia-induced in ammatory response included microglia activation and increased pro-in ammatory cytokines production [36,37]. In the present study, high glucose treatment is demonstrated to elevate the mRNA levels of pro-in ammatory cytokines in HT-22 cells directly. We further revealed that this increased in ammation was mediated through the downregulation of H 2 S, as restoring H 2 S levels using NaHS or SAMe attenuated the upregulation of pro-in ammatory cytokines induced by high glucose. These results are consistent with previous ndings that H 2 S attenuates LPS-induced in ammatory response in endothelial cells and microglia [23,38]. Our previous preclinical research has also proved that H 2 S attenuates hippocampal in ammation in the AlCl 3 -induced mouse model of Alzheimer's disease [14].
Collectively, these ndings suggest that the reduction of H 2 S in neurons contributes to high-glucose induced neuroin ammation.
We also found that high glucose reduced H 2 S synthesizing enzyme, CBS protein level in HT-22 neurons, while NaHS or SAMe restored CBS and endogenous H 2 S levels. The endogenous H 2 S is mainly regulated by three enzymes including CBS, CSE and 3-MST [39]. In the brain, CBS is the major enzyme to modulate the endogenous H 2 S synthesis. NaHS treatment has been shown to restore H 2 S production and CBS expression in a mouse model of intracerebral hemorrhage [40]. SAMe, a CBS activator, increased the H 2 S synthesis in the central nervous system [28]. In the current study, our results showed that NaHS or SAMe upregulated CBS expression and restored high glucose-induced reduction of H 2 S in neuronal cells. The reduced expression of CBS has been found in the brain of Alzheimer's disease patients [41]. Interestingly, impaired glycemia is associated with the progression from mild cognitive impairment to Alzheimer's disease [42]. Therefore, these ndings suggest that enhancement of CBS is a potential target to increase H 2 S to prevent hyperglycemia associated neurodegenerative disease.
Increasing evidence revealed that SIRT1, a NAD + -dependent deacetylase, modulated in ammatory response. For example, in ammation stimulator, e.g. LPS, reduced the SIRT1 expression and activated pro-in ammatory cytokines levels in vascular endothelial cells [43]. The activation of SIRT1 facilitated the formation of facultative heterochromatin to silence pro-in ammatory genes, for instance, TNF-α and IL-1β in THP-1 promonocytes [44][45][46]. Here, we found that SIRT1 agonist, resveratrol prevented high glucoseinduced pro-in ammatory cytokines expression in HT-22 neurons. Therefore, SIRT1 is an important mediator of glucose-induced neuroin ammation. Research has shown H 2 S upregulated SIRT1 to protect neurotoxicity in PC12 cells induced by homocysteine [16]. SIRT1 mediated the roles of H 2 S in attenuating cognitive impairment induced by chronic restraint stress in rats [47]. In the present work, we found that H 2 S enhancer, SAMe and NaHS upregulated SIRT1 expression and downregulated pro-in ammatory cytokines expression in high glucose-exposed HT-22 cells. These results indicated that SIRT1 mediates the protection of H 2 S against high glucose-triggered neuronal in ammation in HT-22 cells.
Previously, studies have shown that SIRT1 negatively regulated mTOR signaling in a variety of cells, such as MEFs cells, HeLa cells and cardiomyocytes [48,49]. In the current work, we showed that SIRT1 regulate SIRT1-mTOR to provide anti-neuroin ammation effects. NF-κB, a well-known nuclear transcription factor, regulates the mRNA levels of in ammation-related factors [52]. NF-κB is a downstream mediator of mTOR in immune cells [22]. Furthermore, enhancing SIRT1 expression using resveratrol suppressed NF-κB activation and inhibited the secretion of pro-in ammatory mediators in macrophage RAW264.7 cells [32]. Therefore, NF-κB is a downstream molecule of SIRT1-mTOR for in ammatory response. Our results revealed that resveratrol or SAMe/NaHS restored high glucoseinduced NF-κB activation in HT-22 cells. Overall, these ndings suggest that the H 2 S supplement might modulate SIRT1-mTOR-NF-κB signaling pathway and thus inhibit high glucose-induced pro-in ammatory cytokine expression.

Conclusions
In summary, our results have shown that high glucose activated in ammation through reducing H 2 S and its synthesizing enzyme CBS in neurons. Conversely, restore H 2 S levels inhibited in ammation activation.       immuno uorescence results showed that HG reduced the expression of SIRT1 protein, which was reversely by NaHS or SAMe. (B and C) Western blotting and immuno uorescence results showed that HG promoted the phosphorylation of mTOR (B) and NF-κB p65 (C), which was restored by NaHS or SAMe.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download. SupplementaryMaterial.docx