PI3K Inhibitor Library

Mebhydrolin ameliorates glucose homeostasis in type 2 diabetic mice by functioning as a selective FXR antagonist
Tong Zhao a,1, Jie Wang b,1, Anxu He a, Shan Wang a, Yidi Chen a, Jian Lu a, Jianlu Lv a, Shiliang Li b, Jiaying Wang a,
Minyi Qian a,⁎, Honglin Li b,⁎, Xu Shen a,⁎
a School of Medicine& Holistic Integrative Medicine, Nanjing University of Chinese Medicine, Nanjing 210023, China
b Shanghai Key Laboratory of New Drug Design, School of Pharmacy, East China University of Science and Technology, 130 Meilong Road, Shanghai 200237, China

a r t i c l e i n f o

Article history:
Received 6 November 2020
Accepted 28 March 2021

Keywords:
Farnesoid X receptor Mebhydrolin
Type 2 diabetes miR-22-3p Gluconeogenesis Glycogen synthesis

a b s t r a c t

Introduction: Type 2 diabetes mellitus (T2DM) is a chronic disease with hallmarks of hyperglycemia and hyper- lipidemia. Long-term hyperglycemia damages the functions of multiple tissues and organs leading to a series of complications and disability or even death. Nuclear receptor farnesoid X receptor (FXR) antagonism has been re- cently discovered to exhibit beneficial effect on glucose metabolism in T2DM mice, although the underlying mechanisms remain unclear. Here, we performed the study on the discovery of new FXR antagonist and investi- gated the mechanism underlying the amelioration of FXR antagonism on glucose homeostasis in T2DM mice by using the determined FXR antagonist as a probe.
Methods: FXR antagonist Mebhydrolin was discovered by screening against the lab in-house FDA approved drug library through surface plasmon resonance (SPR), microscale thermophoresis (MST), AlphaScreen, mammalian one-hybrid and transactivation assays. Activity of Mebhydrolin in improving glucose homeostasis was evaluated in db/db and HFD/STZ-induced T2DM mice, and the mechanisms governing the regulation of Mebhydrolin were investigated by assays of immunostaining, Western blot, ELISA, RT-PCR against liver tissues of both T2DM mice and the T2DM mice with liver-specific FXR knockdown injected via adeno-associated-virus AAV-FXR-RNAi and mouse primary hepatocytes. Finally, molecular docking and molecular dynamics (MD) technology-based study was performed to investigate the structural basis for the antagonistic regulation of Mebhydrolin against FXR at an atomic level.
Findings: Mebhydrolin ameliorated blood glucose homeostasis in T2DM mice by both suppressing hepatic gluco- neogenesis via FXR/miR-22-3p/PI3K/AKT/FoxO1 pathway and promoting glycogen synthesis through FXR/miR- 22-3p/PI3K/AKT/GSK3β pathway. Structurally, residues L291, M332 and Y373 of FXR were required for Mebhydrolin binding to FXR-LBD, and Mebhydrolin induced H2 and H6 shifting of FXR potently affecting the reg- ulation of the downstream target genes.
Conclusions: Our work has revealed a novel mode for the regulation of FXR against glucose metabolism in T2DM mice and highlighted the potential of Mebhydrolin in the treatment of T2DM.
© 2021 Elsevier Inc. All rights reserved.

1. Introduction

Type 2 diabetes mellitus (T2DM) is a metabolic syndrome with hall- marks of hyperglycemia and hyperlipidemia. Long-term hyperglycemia damages the functions of multiple tissues and organs leading to a series of complications and disability or even death [1]. Although drugs against T2DM are clinically available (e.g. Thiazolidinedione, DPP-4 inhibitor and Sulfonylurea), side effects such as increased nausea, edema [2],

⁎ Corresponding authors.
E-mail addresses: [email protected] (M. Qian), [email protected] (H. Li), [email protected] (X. Shen).
1 T.Z. and J.W. contributed equally to this work.

hypoglycemia and weight gain [3] are restricting their applications. Thus, developing safe and effective medication to treat T2DM based on new strategies is invariably needed.
In the pathogenesis of T2DM, insulin resistance and relatively insuf- ficient insulin secretion result in excessive activation of hepatic glucose output and further hyperglycemia [4]. The liver as an insulin target organ plays a vital role in modulating blood glucose homeostasis, and hepatic gluconeogenesis and glycogen synthesis contribute largely to hepatic glucose output [5]. Hepatic gluconeogenesis is mainly regulated by glucose-6-phosphatase (G6Pase) and phosphoenolpyruvate carboxy kinase (PEPCK) [6], which are modulated by nuclear receptors, adiponectin, free fatty acids and transcription factors [7–11]. Gluconeo- genesis inhibition has been determined a promising therapeutic strat- egy for T2DM [4], as supported by the fact that clinical drugs

metformin and phenformin as well as some preclinical agents (LY-2409021 [12], MK-0893 [13], etc.) repress blood glucose by inhibiting gluconeogenesis.
Apart from hepatic gluconeogenesis, hepatic glycogen synthesis is also contributable to glucose output [14]. Hepatic glycogen synthesis is mainly regulated by glycogen synthase kinase 3 (GSK3β) [15]. GSK3β activation phosphorylates glycogen synthase resulting in enzy- matic activity inhibition and further glycogen synthesis decrease [16]. Insulin suppresses GSK3β activity leading to promotion of glycogen syn- thesis by activating AKT in the liver, and GSK3β inhibitors have been discovered to regulate hepatic glucose metabolism by inducing glyco- gen synthesis [17]. For example, Kenpaullone and CHIR98014 as GSK3β inhibitors improved glucose metabolism by enhancing hepatic glycogen synthesis in T2DM mice [18,19].
Farnesoid X receptor (FXR) belongs to a member of metabolic nu- clear receptor superfamily and plays an important role in glucose me- tabolism [20,21]. FXR is accepted as a potential target for drug design against hyperglycemia. For example, FXR agonists GW4064 and Cholic acid reduced blood glucose by inhibiting gluconeogenesis and promot- ing glycogen synthesis in T2DM mice [22,23], while FXR antagonists

Glycine-β-muricholic acid [24], NDB [25] and HS218 [26] respectively reduced blood glucose in obese mice, reduced gluconeogenesis gene ex- pression in db/db mice and improved glucose metabolism in T2DM mice by suppressing gluconeogenesis. Additionally, miRNAs are small non- coding RNAs implicated in the pathogenesis of various diseases including obesity and T2DM [27,28]. For example, miR-9 [29] and miR-22-3p [30] regulated gluconeogenesis, and miR-19a [31] and miR-20a-5p [32] were linked to glycogen synthesis. It was reported that miR-22-3p could be transcriptionally regulated by FXR through di- rect binding to an invert repeat 1 (IR1) motif located at −1012 to
−1025 bp upstream from miR-22-3p [30].
Here, we reported that antiallergic drug Mebhydrolin (Fig. 1A) as a selective FXR antagonist improved glucose homeostasis in T2DM mice, and the underlying mechanisms were intensively investigated. More- over, molecular docking combined with molecular dynamics simulation analysis was performed to elaborate the molecular basis of Mebhydrolin-mediated FXR antagonism at an atomic level. To our knowledge, our work might be the first report on a new mode for FXR regulation against glucose metabolism and highlight the potential of Mebhydrolin in treating T2DM.

Fig. 1. Mebhydrolin suppressed hepatic glucose production by functioning as a selective FXR antagonist in mouse primary hepatocytes. (A) Chemical structure of Mebhydrolin. (B) Binding affinity of Mebhydrolin to His-FXR-LBD was detected by MST. (C) Effect of Mebhydrolin on FXR-LBD affinity to SCR-1 was detected by AlphaScreen-based protein-peptide interaction assay. (D) Mammalian one-hybrid assay was performed to detect the regulation of Mebhydrolin against FXR. HEK293T cells were co-transfected with plasmids of pCMX-Gal4-FXR- LBD, pUAS-TK-luc and pRL-SV40, and treated with GW4064 (50 nM) and Mebhydrolin (5, 10, 20 μM) for 12 h, and luciferase activity was finally measured. (E) HEK293T cells were transiently transfected with pcDNA3.1a-FXR, pcDNA3.1a-RXRα, pGL3-FXRE-Luc and pRL-SV40. After 6 h, cells were treated with different concentrations of Mebhydrolin (5, 10, 20 μM) and GW4064 (50 nM) for 12 h. Transactivation activity of FXR was then detected by luciferase reporter assay. (F) Mouse primary hepatocytes were pretreated with glucagon (10 nM) and Mebhydrolin (5, 10, 20 μM) for 12 h, and then incubated for another 6 h in glycogenetic medium with glucagon (10 nM) and Mebhydrolin (5, 10, 20 μM). Finally, glucose level was detected. (G) Mouse primary hepatocytes were pretreated with glucagon (10 nM), Mebhydrolin (5, 10, 20 μM) or HS-592 (10 nM) for 12 h, and then incubated for another 6 h in glycogenetic medium with the same compounds. Finally, glucose level was detected. (H) Mouse primary hepatocytes were pretreated with glucagon (10 nM), Mebhydrolin (5, 10, 20 μM) or GS (15 μM) for 12 h, and then incubated for another 6 h in glycogenetic medium with the same compounds. Finally, glucose level was detected. All data were presented as mean ± S.E.M, from three independent experiments (*P < 0.05, **P < 0.01, ***P < 0.001). 2. Research design and methods 2.1. Cell culture Human embryonic kidney HEK293T cells were cultured in Dulbecco's minimum essential medium (DMEM) at 37 °C in 5% humid- ified CO2 incubator. Mouse primary hepatocytes were isolated from 8 to 12 week-old male C57BL/6 mice by two-step collagenase perfusion method [33]. 2.2. Preparation of FXR-LBD wild type and its mutant proteins Residues 244-476 of wild type FXR-LBD (LBD, ligand binding do- main) were cloned into Nde1/XhoI sites of pET-15b vector to get the plasmid expressing N-terminal His-hFXRα-LBD. Plasmids of hFXRα- LBD-L291A, hFXRα-LBD-M332A and hFXRα-LBD-Y373A were con- structed by published methods [34]. Expression and purification of His-hFXRα-LBD and its mutants were based on the published approach [25], and protein concentration was determined by Microvolume Spec- trometer (Colibri, Germany). 2.3. AlphaScreen-based protein-peptide interaction assay The assay was performed by the published approach [26]. Protein- peptide interaction was detected with ﬂuorescence wavelength at 520–620 nm using Histidine Detection Kit (PerkinElmer Life Science, USA). 2.4. Luciferase reporter assay Mammalian transactivation and one-hybrid experiments were con- ducted by the reported approach [30,35]. Luciferase activities were measured by dual-luciferase reporter assay system kit (Promega, USA). 2.5. MST assay His-FXR-LBD protein was labeled with primary-amine coupling of MO-LO18 dye (Nano Temper), followed by the addition of different con- centrations of Mebhydrolin (0.03 nM to 1 mM). After incubation for 0.5 h, Microscale thermophoresis (MST) assay was performed at room tem- perature on a Monolith NT.115 instrument (Nano Temper) by the man- ufacturer's protocol. Data analysis was performed with Nano Temper Analysis software. 2.6. Glucose output assay Glucose output assay was performed by the published approach [36]. Glucose level was measured by glucose detection kit (Nanjing Jiancheng Bioengineering Institute, China), and protein concentration was measured by BCA detection kit (Beyotime, China) to normalization. 2.7. Glycogen detection assay Mouse primary hepatocytes and liver tissues were lysed in RIPA buffer and supernatants were collected after centrifugation. Glycogen level was determined by bioassay glycogen test kit (Bioassay system, USA), and protein concentration was measured by BCA detection kit to normalization. 2.8. PAS dyeing Mouse primary hepatocytes and liver tissues were fixed with 4% fix- ative solution and stained by using PAS/Hematoxylin stain kit (Solarbio, Beijing, China), and observed under Leica DM1000 Upright microscope (Leica, Germany). 2.9. RNA isolation and quantitative real-time PCR Quantitative real-time PCR (qRT-PCR) was performed by published method [30,37]. The primers sequences are listed in Supplementary material. 2.10. Western blot assay Western blot analysis was performed by published protocols [38]. The antibodies used were given in Supplementary material. 2.11. Cell transfection with FXR-siRNA or miR-22-3p inhibitor Mouse primary hepatocytes were transiently transfected with 25 pmol FXR-siRNA (Thermo Fisher, USA) or 20 nM miR-22-3p inhibitor (Qiagen, Germany). FXR-siRNA or miR-22-3p-inhibitor was dissolved in 75 μl of Opti-MEM and mixed with transfection reagent (Lipofecta- mine RNAi MAX dissolved in 75 μl of Opti-MEM medium) for 15 min, and the mixture was incubated in cells for 6 h, while the medium was changed into MEM and incubated for 24 h. 2.12. Immunoﬂuorescence staining with FoxO1 antibody Mouse liver tissues were fixed with 4% fixative solution for 48 h and embedded in paraffin and sectioned. Tissue sections were incubated with primary FoxO1 antibody (1:300) for 48 h, and then incubated with ﬂuorescent secondary anti-rabbit antibody (1:200) for 2 h and Hochest-33342 was used to stain the nuclear. Leica ﬂuorescence micro- scope equipped was used to image. 2.13. Animals All animal experiments were implemented following the rules and regulations of the Institution Animal Care and Use Committers of Nan- jing University of Chinese Medicine (No. 201812A015). All mice were housed in a controlled temperature and humidity environment and free to drink water and diet under a 12 h light/12 h dark cycle. Male db/db mice (BKS.Cg-Dock7m +/+ Leprdb/J) were purchased from the Model Animal Research Center of Nanjing University (No.201819367), and male C57BL/6 mice were purchased from Beijing Vital River Laboratory Animal Technology Co., Ltd. (No.33000800000510). HFD/STZ-induced T2DM mice (HFD/STZ mice) were modeled by published approach [39]. Blood was ob- tained by tail-tip bleeding, and blood glucose levels were measured using an Accu-Chek Roche® glucometer 1 week after STZ or citrate buffer (control) injection. The animals were considered diabetic when fasting blood glucose levels exceeded 11 mmol/L [26]. 2.14. AAV-induced FXR knockdown model mice Liver (TBGP-EGFP-MCS-SV40 Poly A) adeno-associated-virus (AAV)-FXR-RNAi (5′ to 3′: (RNA)-GCCAUGUACAGAUUCUCGUAGAAU U) and negative control vector (AAV-NC) were purchased from Shang- hai Genechem Co., Ltd. AAV-induced FXR knockdown model mice (HFD/STZ + AAV-FXR-RNAi) were obtained by injection of the virus (1011 titer) into HFD/STZ mice through tail vein [40]. AAV expression was detected after 2 weeks. 2.15. Animal administration Thirty-eight db/db mice were treated with vehicle (4% DMSO and 4% Tween-80, N = 12), Mebhydrolin (15 mg/kg/day, N = 13) or Mebhydrolin (30 mg/kg/day, N = 13) by intraperitoneal injection for 5 weeks. Thirty-eight HFD/STZ were treated with vehicle (4% DMSO and 4% Tween-80, N = 13), Mebhydrolin (15 mg/kg/day, N = 11) or Mebhydrolin (30 mg/kg/day, N = 14) by intraperitoneal injection for 5 weeks. Thirty-six HFD/STZ + AAV-FXR-RNAi mice were treated with vehicle (4% DMSO and 4% Tween-80, N = 11), Mebhydrolin (15 mg/kg/day, N = 13) or Mebhydrolin (30 mg/kg/day, N = 12) by intra- peritoneal injection for 5 weeks. Ten control (C57BL/6 mice) and seven HFD/STZ + AAV-NC mice were administrated with the same volume of the vehicle buffer as that for T2DM mice. For all mice, plasma glucose level after 6 h fasting was measured weekly, and glycated hemoglobin (HbA1c) was investigated by DCA Vantage analysis system (Siemens, Germany) after mice were sacrificed. 2.16. OGTT and PTT For oral glucose tolerance test (OGTT) and pyruvate tolerance test (PTT), mice were fasted overnight, 1 g/kg glucose (for OGTT) or 1.5 g/kg pyruvate (for PTT) solution was orally or intraperitoneally admin- istered, and blood glucose level was measured at 0, 15, 30, 45, 60, 90 and 120 min. 2.17. Molecular dynamics simulation and homology modeling 2.17.1. Systems set up The crystal structure of FXR-LBD complexed with NDB (PDB ID: 4OIV) [25] was selected as the initial structure, which was prepared with the Protein Preparation Wizard available in Maestro. The ligands were prepared and docked into the active site of FXR-LBD with Glide 10.6 [41]. 2.17.2. Molecular dynamics simulations Protonation states of the ionizable residues were computed by using the PROPKA online server [38]. Ligand parameters were generated with Antechamber in Amber 14 [4]. Molecular dynamics (MD) simulations were performed and analyzed with the Amber 14 package. All atom force field ff99SB was applied. Protein-ligand complexes were solvated by using the TIP3P water model. Counter ions were added to neutralize the systems. The systems were minimized with descending restraint forces. The temperature was slowly increased from 0 to 300 K under NVT ensemble conditions and equilibrated under NPT ensemble condi- tions for 500 ps at 300 K. 200 ns unrestrained MD production was per- formed for each system. The time step was set to 2 fs and the atom coordinates were saved every 10 ps for subsequent analysis. 2.17.3. Homology modeling The molecular model of FXR-LBD/RXR complex with DNA binding domain was built by homology modeling using Maestro [42], and PPARγ-LBD/RXRα complex crystal structure (PDB ID: 3DZY) [43] was used as a template (Fig. S7C). As for model evaluation, Verify-3D [44] was regularly used to help identify possible regions where a model might be improved. 2.18. Statistical analysis Data were presented as means ± SEM or means ± SD. Unpaired 2- tailed Student's t-test was used for two-group comparison. One-way ANOVA with post hoc comparisons using Tukey's or Dunnett's post hoc tests was used for at least three groups' comparisons. Significance was defined as P < 0.05. 3. Results 3.1. Mebhydrolin as a selective FXR antagonist suppressed hepatic glucose production in mouse primary hepatocytes 3.1.1. Mebhydrolin bound to FXR-LBD To discover FXR antagonist, we at first screened for FXR-LBD binders against the lab in-house FDA approved drug library by Reichert 4 SPR-based assay system. Mebhydrolin was finally determined to bind FXR-LBD (Fig. S1), and MST assay result further verified the affinity of FXR-LBD to Mebhydrolin by dissociation constant (KD) value of 9.87 ± 5.78 μM (Fig. 1B). 3.1.2. Mebhydrolin antagonized GW4064-induced promotion on FXR-LBD binding to SRC-1 AlphaScreen-based assay result (Fig. 1C) indicated that Mebhydrolin antagonized the GW4064 (FXR known agonist)-induced promotion on FXR-LBD binding to coactivator SRC-1, thus implying the antagonistic feature of Mebhydrolin against FXR. 3.1.3. Mebhydrolin inhibited FXR transcriptional activity Inhibition of Mebhydrolin against FXR transcriptional activity was evaluated by mammalian one-hybrid and transcriptional activation as- says in HEK293T cells, and the results (Fig. 1D and E) indicated that Mebhydrolin antagonized the GW4064-stimulated reporter gene ex- pressions. Moreover, further transcriptional activity assay also verified the antagonistic selectivity of Mebhydrolin against FXR (Fig. S2). Thus, all results demonstrated that Mebhydrolin was a selective FXR antagonist. 3.1.4. Mebhydrolin suppressed hepatic glucose output by antagonizing FXR We next investigated the potential of Mebhydrolin in suppressing glucose output in primary hepatocytes [26], and the result indicated that Mebhydrolin inhibited glucagon-stimulated glucose output (Fig. 1F). Moreover, given that Mebhydrolin also targeted H1 receptor [45], we inspected whether H1 receptor targeting was also involved in the Mebhydrolin-mediated suppression of hepatic glucose output. As shown in Fig. 1G, H1 known antagonist HS-592 gave no inﬂuence on the Mebhydrolin-mediated reduction in glucagon-stimulated glucose output in mouse primary hepatocytes, which thus demonstrated that H1 receptor targeting was not involved in the Mebhydrolin-mediated suppression against hepatic glucose output. In addition, GS, the antago- nist of FXR, was used to verify whether Mebhydrolin inhibited glucose output through targeting FXR. In this assay, mouse primary hepatocytes were pretreated with GS to silence FXR, followed by addition of Mebhydrolin simultaneously. As shown in Fig. 1H, the Mebhydrolin- mediated reduction in glucagon-stimulated glucose output in mouse primary hepatocytes was abrogated by GS, indicating that silencing FXR by GS could block the inhibition of Mebhydrolin against the glucagon-stimulated glucose output. Thus, all results demonstrated that Mebhydrolin reduced hepatic glucose output by antagonizing FXR. 3.2. Mebhydrolin inhibited hepatic gluconeogenesis through FXR/PI3K/AKT/ FoxO1 pathway in mouse primary hepatocytes 3.2.1. Mebhydrolin inhibited hepatic gluconeogenesis by antagonizing FXR Considering that hepatic gluconeogenesis contributes largely to hepatic glucose output, we investigated the potential of Mebhydrolin in repressing hepatic gluconeogenesis. As shown in Fig. 2A and B, Mebhydrolin efficiently inhibited the glucagon- promoted mRNA level of G6Pase or PEPCK in mouse primary hepato- cytes. Moreover, si-FXR deprived Mebhydrolin of its capability in suppressing the expressions of G6Pase and PEPCK genes (Fig. 2C and D). Thus, all results indicated that Mebhydrolin suppressed glu- coneogenesis through antagonizing FXR. 3.2.2. Mebhydrolin suppressed gluconeogenesis through FXR/PI3K/AKT/ FoxO1 pathway Given that FoxO1 translocates from nucleus to upregulate gene tran- scriptions of G6Pase and PEPCK functioning potently in the regulation of gluconeogenesis [46], we detected the potential of Mebhydrolin in reg- ulating FoxO1 signaling in mouse primary hepatocytes by RT-PCR and Fig. 2. Mebhydrolin suppressed gluconeogenesis by antagonizing FXR in mouse primary hepatocytes. (A, B) Mouse primary hepatocytes were pretreated with Mebhydrolin (5, 10, 20 μM) for 4 h, followed by incubation with glucagon (10 nM) for 4 h. mRNA levels of (A) G6Pase and (B) PEPCK were evaluated by quantitative RT-PCR analysis and normalized to GAPDH. (C, D) After transfected with si-control or si-FXR, (C) G6Pase and (D) PEPCK were measured by RT-PCR analysis. All results were normalized to GAPDH. (E) Mouse primary hepatocytes were pretreated with Mebhydrolin (5, 10, 20 μM) for 4 h, followed by incubation with glucagon (10 nM) for 4 h. mRNA level of FoxO1 was evaluated by quantitative RT-PCR and normalized to GAPDH. (F) Mouse primary hepatocytes were treated by Mebhydrolin (5, 10, 20 μM) for 4 h, and phosphorylated and total protein levels of PI3K, AKT and FoxO1 were determined by Western blot. (G) Quantification results for (F), and results were normalized to total protein. (H, I) Mouse primary hepatocytes were pretreated with Mebhydrolin (20 μM) and FoxO1 inhibitor AS1842856 (100 nM) for 4 h, followed by incubation with glucagon (10 nM) for 4 h. mRNA levels of (H) G6Pase and (I) PEPCK were evaluated by RT-PCR analysis and normalized to GAPDH. (J) Mouse primary hepatocytes were treated with Mebhydrolin (20 μM) and PI3K inhibitor wortmannin (5 μM) for 4 h, and phosphorylated and total protein levels of AKT and FoxO1 were determined by Western blot. (K) Quantification results for (J), and results were normalized to total protein. (L) Mouse primary hepatocytes were transfected with si-control or si-FXR for 24 h, and then incubated with Mebhydrolin (20 μM) for 4 h. Phosphorylated and total protein levels of PI3K, AKT and FoxO1 were detected by Western blot assay. (M) Quantification results for (L), and results were normalized to total protein. All data were presented as mean ± S.E.M from three independent experiments (*P < 0.05, **P < 0.01, ***P < 0.001). Western blot assays. The results demonstrated that Mebhydrolin re- pressed glucagon-stimulated mRNA level of FoxO1 gene (Fig. 2E) and enhanced phosphorylation level FoxO1 (Fig. 2F–G). Moreover, RT-PCR results indicated that FoxO1 inhibitor AS1842856 deprived Mebhydrolin of its capability in suppressing G6Pase and PEPCK genes expression (Fig. 2H–I), thus demonstrating that FoxO1 signaling was re- quired for Mebhydrolin-mediated gluconeogenesis suppression. Additionally, considering that PI3K/AKT signaling is linked to FoxO1 signaling [47], Western blot assay was also performed in mouse primary hepatocytes. The results demonstrated that Mebhydrolin increased phosphorylation levels of PI3K and AKT (Ser 473) (Fig. 2F–G) and PI3K inhibitor wortmannin blocked the activity of Mebhydrolin in regulating either AKT or FoxO1 (Fig. 2J–K). Moreover, si-FXR deprived Mebhydrolin of its capability in regulating PI3K/AKT/FoxO1 pathway (Fig. 2L–M). Thus, all results demonstrated that Mebhydrolin suppressed hepatic gluconeogenesis through FXR/PI3K/AKT/FoxO1 pathway in mouse pri- mary hepatocytes. 3.3. Mebhydrolin promoted glycogen synthesis through FXR/PI3K/AKT/ GSK3β pathway in mouse primary hepatocytes 3.3.1. Mebhydrolin promoted glycogen synthesis by antagonizing FXR Next, the potential of Mebhydrolin in regulating glycogen synthesis was inspected in mouse primary hepatocytes. Assay result by commer- cial kits demonstrated that Mebhydrolin enhanced glycogen level (Fig. 3A), which was in agreement with PAS staining assay result that Mebhydrolin incubation increased red positive dewpoints (indicative of the increased glycogen) (Fig. 3B). Moreover, si-FXR deprived Mebhydrolin of its capability in promoting glycogen synthesis Fig. 3. Mebhydrolin promoted glycogen synthesis by antagonizing FXR in mouse primary hepatocytes. (A) Mouse primary hepatocytes were incubated with Mebhydrolin (5, 10, 20 μM) for 24 h. Glycogen level was determined using a Bioassay glycogen test kit. (B) Immunohistochemistry assay for PAS staining in mouse primary hepatocytes treated with Mebhydrolin (5, 10, 20 μM). Blue: Nucleus, Fuchsia: Glycogen (Scalebars, 200 μM). (C) Mouse primary hepatocytes were transfected with si-control or si-FXR for 24 h, and then incubated with Mebhydrolin (5, 10 and 20 μM) for 24 h. Glycogen level was determined using a Bioassay glycogen test kit. (D) Mouse primary hepatocytes were treated with Mebhydrolin (5, 10 and 20 μM) for 4 h, and phosphorylated and total protein levels of GSK3β were determined by Western blot. (E) Quantification results for (D), and results were normalized to total protein. (F) Mouse primary hepatocytes were pretreated with Mebhydrolin (20 μM) and LiCl (20 μM) for 24 h. Glycogen level was determined using a Bioassay glycogen test kit. (G) Mouse primary hepatocytes were treated with Mebhydrolin (20 μM) and PI3K inhibitor wortmannin (5 μM) for 4 h, and phosphorylated and total protein levels of GSK3β were determined by Western blot. (H) Quantification results for (G), and results were normalized to total protein. (I) Mouse primary hepatocytes were transfected with si-control or si-FXR for 24 h, and then incubated with Mebhydrolin (20 μM) for 4 h. Phosphorylated and total protein levels of GSK3β were determined by Western blot. (J) Quantification results for (I), and results were normalized to total protein. All data were presented as mean ± S.E.M from three independent experiments (*P < 0.05, **P < 0.01, ***P < 0.001). (Fig. 3C). Thus, all results verified that Mebhydrolin promoted glycogen synthesis by antagonizing FXR. 3.3.2. Mebhydrolin promoted glycogen synthesis through FXR/PI3K/AKT/ GSK3β pathway Considering that glycogen synthesis is mainly regulated by GSK3β, we investigated the potential of Mebhydrolin in regulating against GSK3β in mouse primary hepatocytes. Western blot result (Fig. 3D and E) indicated that Mebhydrolin promoted phosphorylated GSK3β (i.e. inhibited GSK3β enzyme activity). Moreover, the result that GSK3β known inhibitor LiCl treatment antagonized the Mebhydrolin- enhanced glycogen synthesis (Fig. 3F) implied that Mebhydrolin stimu- lated hepatic glycogen synthesis by inhibiting GSK3β. As PI3K/AKT signaling was involved in GSK3β regulation [47] and Mebhydrolin has been determined to regulate PI3K/AKT signaling, we investigated whether PI3K/AKT signaling was also linked to the regula- tion of Mebhydrolin against GSK3β signaling in mouse primary hepatocytes. As indicated in the Western blot result (Fig. 3G–H), PI3K inhibitor wortmannin blocked the Mebhydrolin-induced enhancement on GSK3β phosphorylation, thus implying that PI3K/AKT signaling was responsible for the regulation of Mebhydrolin against GSK3β. Moreover, si-FXR deprived Mebhydrolin of its capability in activating GSK3β phos- phorylation (Fig. 3I and J). Thus, combining the determined result of Mebhydrolin-mediated FXR/PI3K/AKT signaling (Fig. 2L and M), it was concluded that Mebhydrolin promoted glycogen synthesis through FXR/PI3K/AKT/GSK3β pathway in mouse primary hepatocytes. 3.4. miR-22-3p was required for Mebhydrolin-mediated FXR regulation against PI3K/AKT signaling in mouse primary hepatocytes Given that FXR directly binds to invert repeat 1 (IR1) motif of miR- 22-3p in cancer cells [30], we examined whether Mebhydrolin regu- lated miR-22-3p through antagonizing FXR. To our expect, either co- transfection of PGL3-IR1 with FXR and RXRα or FXR agonist CDCA in Fig. 4. miR-22-3p was responsible for Mebhydrolin-mediated FXR regulation against PI3K/AKT signaling. (A) FXR, RXRα and SV40 plasmids were co-transfected with PGL3-IR1 or PGL3-NC plasmids for 6 h in HEK293T cells. After 6 h post-transfection, cells were treated with CDCA (100 μM) or DMSO for 24 h. The fireﬂy and renilla luciferase activities were measured by Dual- Luciferase Reporter Assay System kit. (B) PGL3-IR1, FXR and RXRα were co-transfected in HEK293T cells for 6 h, and cells were treated with CDCA (100 μM) or Mebhydrolin for 24 h. The fireﬂy and renilla luciferase activities were measured by Dual-Luciferase Reporter Assay System kit. (C) Mouse primary hepatocytes were treated with Mebhydrolin (20 μM) for 48 h, and miR-22-3p expression level was detected by RT-PCR and normalized to U6. (D) Mouse primary hepatocytes were transfected with si-control or si-FXR for 24 h, followed by incubation with Mebhydrolin (20 μM) for 48 h. miR-22-3p level was detected by RT-PCR and normalized to U6. (E) Mouse primary hepatocytes were transfected with si-control or miR-22-3p inhibitor for 24 h, and then treated with Mebhydrolin (20 μM) for 4 h. Phosphorylated and total protein levels of PI3K, AKT, FoxO1 and GSK3β were detected by Western blot. (F) Quantification results for (E), and results were normalized to total protein. All data were presented as mean ± S.E.M from three independent experiments (*P < 0.05, **P < 0.01, ***P < 0.001). HEK293T cells increased luciferase activity (Fig. 4A), and Mebhydrolin decreased the CDCA-induced luciferase activity (Fig. 4B). These re- sults thus implied that Mebhydrolin antagonized FXR association to miR-22-3p. In addition, Mebhydrolin reduced miR-22-3p expression (Fig. 4C) and si-FXR deprived Mebhydrolin of the ability in regulat- ing miR-22-3p (Fig. 4D) in mouse primary hepatocytes. Thus, all re- sults indicated that Mebhydrolin reduced miR-22-3p level by antagonizing FXR. Moreover, Western blot result demonstrated that miR-22-3p in- hibitor blocked the regulation of Mebhydrolin against PI3K/AKT/ FoxO1 or PI3K/AKT/GSK3β pathway in mouse primary hepatocytes (Fig. 4E–F). Thus, all data demonstrated that miR-22-3p was required for Mebhydrolin-mediated FXR regulation against PI3K/AKT signaling in mouse primary hepatocytes. 3.5. Mebhydrolin treatment improved glucose homeostasis in T2DM mice by antagonizing FXR As expected, Mebhydrolin reduced fasting blood glucose and HbA1c levels in both db/db (Fig. 5A and B) and HFD/STZ mice (Fig. 5C and D) but had no effects on body weight of T2DM mice (Fig. S3A–C). Next, HFD/STZ + AAV-FXR-RNAi mice related assay was performed. The results in Fig. S4A–C demonstrated the efficiency of AAV-FXR-RNAi in inducing FXR knockdown in the liver tissue of HFD/STZ mice, and the results in Fig. S5A–I indicated that the viral vector treatment had no ef- fects on gluconeogenesis or glycogen-related indicators in HFD/STZ mice. Notably, injection of AAV-FXR-RNAi rendered no effects on the level of blood glucose or HbA1c in HFD/STZ mice (Fig. 5E and F), in ac- cordance with the published results in diet-induced T2DM mice with liver FXR−/− [48]. Moreover, Mebhydrolin failed to regulate blood glu- cose (Fig. 5E) or HbA1c (Fig. 5F) in HFD/STZ + AAV-FXR-RNAi mice. In addition, Mebhydrolin treatment improved OGTT and PTT in T2DM mice (Fig. 5G–J; Fig. 5K–N), and injection of AAV-FXR-RNAi rendered no effects on OGTT and PTT levels in HFD/STZ mice, while Mebhydrolin treatment failed to regulate OGTT and PTT levels in HFD/STZ + AAV- FXR-RNAi mice (Fig. 5O–R). Thus, all data indicated that Mebhydrolin improved blood glucose homeostasis in T2DM mice by targeting FXR. 3.6. Mebhydrolin treatment suppressed gluconeogenesis through FXR/miR- 22-3p/PI3K/AKT/FoxO1 pathway in T2DM mice As indicated, Mebhydrolin treatment reduced mRNA levels of G6Pase and PEPCK in T2DM mice (Fig. 6A–D) but failed to regulate these genes expression in HFD/STZ + AAV-FXR-RNAi mice (Fig. 6E and F). Western blot and RT-PCR results indicated that Mebhydrolin treatment enhanced phosphorylated PI3K, AKT (Ser473) and FoxO1 (Ser256) levels and decreased miR-22-3p expression in db/db (Fig. 6G–I) and HFD/STZ mice (Fig. 6J–L). Moreover, Mebhydrolin failed to regulate miR-22-3p/PI3K/AKT/FoxO1 signaling in HFD/STZ + AAV- FXR-RNAi mice (Fig. 6M–O). Similarly, Mebhydrolin treatment reduced mRNA level of FoxO1 (Fig. 6P–Q) and promoted FoxO1 nuclear export (Fig. 6S–T) in T2DM mice, but failed to regulate FoxO1 signaling in HFD/STZ + AAV-FXR- RNAi mice (Fig. 6R and U). Together, all data demonstrated that Mebhydrolin treatment sup- pressed gluconeogenesis through FXR/miR-22-3p/PI3K/AKT/FoxO1 pathway in T2DM mice. 3.7. Mebhydrolin treatment promoted glycogen synthesis through FXR/ miR-22-3p/PI3K/AKT/GSK3β pathway in T2DM mice As indicated in Fig. 7A and B determined by commercial kit, Mebhydrolin treatment increased glycogen level of the liver tissue in T2DM mice, in agreement with PAS staining assay result (Fig. 7C and D for db/db mice, Fig. 7E and F for HFD/STZ mice). Moreover, Mebhydrolin treatment rendered no inﬂuence on glycogen level in HFD/STZ + AAV-FXR-RNAi mice (Fig. 7G for commercial kit result, Fig. 7H and I for PAS staining assay result). Thus, all data suggested that Mebhydrolin promoted hepatic glycogen synthesis in T2DM mice through antagonizing FXR. As mentioned above, we have already determined Mebhydrolin- mediated FXR/miR-22-3p/PI3K/AKT signaling in T2DM mice. Here, we also determined that Mebhydrolin treatment increased p-GSK3β (Ser9) in T2DM mice (Fig. 7J–M) but had no effects on p-GSK3β (Ser9) in HFD/STZ + AAV-FXR-RNAi mice (Fig. 7N–O). Together, all results suggested that Mebhydrolin promoted hepatic glycogen synthesis in T2DM mice through FXR/miR-22-3p/PI3K/AKT/ GSK3β pathway. 3.8. Structural basis for Mebhydrolin antagonism against FXR Previously, we reported that FXR antagonist HS218 suppressed gluconeogenesis by inhibiting FXR binding to PGC-1α promoter [26]. Interestingly, Mebhydrolin rendered no effects on PGC-1α (Fig. S6) thus exhibiting totally different mechanism from that of HS218. To investigate the structural basis for Mebhydrolin- mediated antagonism against FXR at an atomic level, molecular docking and molecular dynamics (MD) technology-based study was performed. Alanine scanning (Fig. S7A) and site-directed mutation combined with MST assay (Fig. 8A) were at first carried out and the results demon- strated that residues L291, M332 and Y373 of FXR were required for Mebhydrolin binding to FXR-LBD. Next, we compared the binding modes of HS218 and Mebhydrolin with FXR-LBD. In FXR-LBD/HS218 complex structure, HS218 (Fig. 8B, yellow sticks) had a similar binding mode to NDB (Fig. 8B, white sticks). Mebhydrolin (Fig. 8C, green sticks) bound to the same region of the ligand-binding pocket without polar interaction. Then, MD simulations for 200 ns were performed for FXR-LBD/ Mebhydrolin and FXR-LBD/HS218. As shown in Fig. 8D, compared with FXR-LBD/HS218 (black), FXR-LBD/Mebhydrolin (red) had smaller RMSD values. Moreover, the RMSDs of the ligands (Fig. S7B) in two sys- tems demonstrated that HS218 was more unstable than Mebhydrolin. These results thus indicated that different small molecules might affect the structural stability of the whole protein leading to different protein conformations. Through comparing the values of RMSF of the two systems (Fig. 8E), we found that the main differences of protein conforma- tions were reﬂected in H2, H6, and H7. Further RMSD analysis for these helical structures indicated that the conformational variations between H2 and H6 in FXR-LBD were more evident after small mol- ecule binding (Fig. 8F). Moreover, via comparing the average struc- tures of H2 and H6 regions for the two complexes, certain changes were observed for H2 and H6, that is, compared with the apo struc- ture (Fig. 8G), when HS218 bound to FXR-LBD, H2 shifted −0.087° and H6 shifted −26.077°, while H2 shifted −8.296° and H6 shifted 17.875° after Mebhydrolin binding. Therefore, we speculated that the conformational changes of H2 and H6 induced by binding of Mebhydrolin or HS218 to FXR-LBD may affect the downstream regu- lation of FXR. 4. Discussion and conclusion In the current work, we determined that Mebhydrolin as a selec- tive FXR antagonist ameliorated glucose homeostasis in T2DM mice by both suppressing hepatic gluconeogenesis and promoting glyco- gen synthesis. To our knowledge, our work might be the first report on the regulation of FXR antagonist against both gluconeogenesis and glycogen synthesis in T2DM mice. In addition, Mebhydrolin is Fig. 5. Mebhydrolin treatment improved glucose homeostasis in T2DM mice by antagonizing FXR. (A, B) Fasting blood glucose level (N = 12) and plasma HbA1c level (N= 7) of vehicle or Mebhydrolin (15, 30 mg/kg)-treated db/db mice (Veh or Mebhydrolin). (C, D) Fasting blood glucose level (N ≥ 10) and plasma HbA1c level (N = 7) of control mice (Control, age-matched C57BL/6 mice as control in the related assay for HFD/STZ mice) and vehicle or Mebhydrolin (15, 30 mg/kg)-treated HFD/STZ mice (Veh or Mebhydrolin). (E, F) Fasting blood glucose level and plasma HbA1c level (N = 7) of HFD/STZ mice (HFD/STZ), AAV-NC (N = 7) or AAV-FXR-RNAi injected HFD/STZ mice (HFD/STZ + AAV-NC or HFD/STZ + AAV-FXR-RNAi) and Mebhydrolin (15, 30 mg/kg)-treated AAV-FXR-RNAi injected HFD/STZ mice (HFD/STZ + AAV + Mebhydrolin) (N ≥ 11). (G) OGTT results in vehicle or Mebhydrolin (15, 30 mg/kg)- treated db/db mice (Veh or Mebhydrolin) and (H) AUC results of OGTT (N = 12). (I) PTT results in vehicle or Mebhydrolin (15, 30 mg/kg)-treated db/db mice (Veh or Mebhydrolin) and (J) AUC results of PTT (N = 12). (K). OGTT results in control mice (Control) and vehicle or Mebhydrolin (15, 30 mg/kg)-treated HFD/STZ mice (Veh or Mebhydrolin), (L) AUC results of OGTT (N > 7). (M) PTT results in control mice (Control) and vehicle or Mebhydrolin (15, 30 mg/kg)-treated HFD/STZ mice (Veh or Mebhydrolin), (N) AUC results of PTT (N > 7). (O) OGTT results in HFD/STZ mice (HFD/STZ), AAV-NC or AAV-FXR-RNAi injected HFD/STZ mice (HFD/STZ + AAV-NC or HFD/STZ + AAV-FXR-RNAi) and Mebhydrolin (15, 30 mg/kg)-treated AAV- FXR-RNAi injected HFD/STZ mice (HFD/STZ + AAV + Mebhydrolin), (P) AUC results of OGTT (N ≥ 7). (Q) PTT results in HFD/STZ mice (HFD/STZ), AAV-NC or AAV-FXR-RNAi injected HFD/ STZ mice (HFD/STZ + AAV-NC or HFD/STZ + AAV-FXR-RNAi) and Mebhydrolin (15, 30 mg/kg)-treated AAV-FXR-RNAi injected HFD/STZ mice (HFD/STZ + AAV + Mebhydrolin), (R) AUC results of PTT (N ≥ 7). Values were showed as mean ± S.D. (*P < 0.05, **P < 0.01 and ***P < 0.001). an antiallergic drug, whose obtained preclinical and clinical data should no doubt provide valuable references for the subsequent de- velopment of anti-T2DM drug based on this “old drug”. miR-22-3p as a highly abundant hepatic microRNA is abnor- mally upregulated by insulin resistance in T2DM mice [49], and miR-22-3p repression ameliorated glucose tolerance and insulin sensitivity [50] although the underlying mechanism was obscure. Here, we determined the association of FXR-mediated miR-22-3p/ PI3K signaling with gluconeogenesis and glycogen synthesis by Mebhydrolin as a probe. Our findings have provided new evidence on the role of FXR antagonism in amelioration of blood glucose in T2DM mice. Previously, we reported that compound HS218 as an FXR antago- nist suppressed gluconeogenesis by inhibiting FXR binding to PGC- 1α promoter but failed to regulate glycogen synthesis [26]. In the current work, Mebhydrolin ameliorated both hepatic gluconeogene- sis and glycogen synthesis exhibiting different mechanisms in regu- lating glucose homeostasis of T2DM compared with HS218. Thus, molecular docking and molecular dynamics technology-based study was performed to investigate the difference in structural basis to bind FXR-LBD between Mebhydrolin and HS218 at atomic level. It was found that Mebhydrolin or HS218 as an FXR antagonist could induce H2 and H6 shifting, which might result in different structural variations in FXR/RXR DNA binding domain, eventually af- fecting the regulation of downstream target genes. Our molecular modeling results thus proposed a novel mode for Mebhydrolin- mediated FXR antagonism, which may helped better understand the potency of H2 and H6 conformation in designing new drug lead compound for treatment of T2DM. In conclusion, we reported that Mebhydrolin as a selective FXR an- tagonist efficiently ameliorated blood glucose homeostasis in db/db and HFD/STZ-induced T2DM mice, and the underlying mechanisms have been intensively investigated. Mebhydrolin suppressed hepatic gluconeogenesis through FXR/miR-22-3p/PI3K/AKT/FoxO1 pathway and promoted glycogen synthesis through FXR/miR-22-3p/PI3K/ AKT/GSK3β pathway (Fig. 8H). In addition, Molecular docking com- binedwithmoleculardynamicssimulationandsite-directedmutation analysis further provided a novel molecular basis for Mebhydrolin- mediated FXR antagonism. Our work has provided a new mode for FXR antagonism in the regulation of glucose metabolism of T2DM mice and highlighted the potential of Mebhydrolin in the treatment of T2DM. Supplementary data to this article can be found online at https://doi. org/10.1016/j.metabol.2021.154771. CRediT authorship contribution statement Shen Xu and Zhao Tong designed the study. Shen Xu reviewed the manuscript. Wang Jie and Li Shiliang performed molecular docking and molecular dynamics simulation experiments. Zhao Tong, Chen Yidi, Lu Jian, and Lv Jianlu. performed the animal and cell experiments. Zhao Tong, Wang Jiaying. and Qian Minyi analyzed and interpreted data. Zhao Tong wrote the manuscript. Wang Shan helped revise the manuscript. Li Honglin and Shen Xu are the guarantors of this work and, as such, have full access to all data in the study and take responsi- bility for the integrity of the data and the accuracy of the data analysis. All authors approved the manuscript. Declaration of competing interest The authors declare that they have no conﬂict of interest. All institu- tional and national guidelines for the care and use of laboratory animals were followed. Acknowledgements The authors thank Xin Xu for expert technical assistance. Funding This work was supported by National Science & Technology Major Project “Key New Drug Creation and Manufacturing Program”, China (No. 2018ZX09711002), National Natural Science Foundation for Young Scientists of China (No. 81703806), Postgraduate Research & Practice Innovation Program of Jiangsu Province (No. KYCX18- 1600), Open Project Program of Jiangsu Key Laboratory for Pharma- cology and Safety Evaluation of Chinese Materia Medica (No. JKLPSE201801), Project of the Priority Academic Program Develop- ment of Jiangsu Higher Education Institutions (PAPD) and Innovative Research Team of Six Talent Peaks Project in Jiangsu Province (No. TD-SWYY-013) and National Natural Science Foundation of China (No. 81825020). Data and resource availability 1) Data availability statements. The datasets generated during and/or analyzed during the current study are available from the corre- sponding author upon reasonable request. Fig. 6. Mebhydrolin suppressed gluconeogenesis through FXR/miR-22-3p/PI3K/AKT/FoxO1 pathway in T2DM mice. (A–F) mRNA levels of G6Pase and PEPCK in liver tissue of mice were detected by RT-PCR. (A, B) for vehicle or Mebhydrolin (15, 30 mg/kg)-treated db/db mice (N = 5–6), (C, D) for vehicle or Mebhydrolin (15, 30 mg/kg)-treated HFD/STZ mice (N = 6) and (E, F) for AAV-NC or AAV-FXR-RNAi injected HFD/STZ (HFD/STZ + AAV-NC or HFD/STZ + AAV-FXR-RNAi, N = 6) and Mebhydrolin (15 or 30 mg/kg)-treated AAV-FXR-RNAi injected HFD/ STZ mice (HFD/STZ + AAV-FXR-RNAi + Mebhydrolin, N = 6). (G) Phosphorylated and total protein levels of PI3K, AKT and FoxO1 in the liver tissue of vehicle or Mebhydrolin (15, 30 mg/kg)-treated db/db mice were detected by Western blot, N = 3/group. (H) Quantification results for (G), and results were normalized to total protein. (I) Level of miR-22-3p in vehicle or Mebhydrolin (15, 30 mg/kg)-treated db/db mice was detected by RT-PCR (N = 8). (J) Phosphorylated and total protein levels of PI3K, AKT and FoxO1 were detected by Western blot in the liver tissue of vehicle or Mebhydrolin (15, 30 mg/kg)-treated HFD/STZ mice, N = 3/group. (K) Quantification results for (J), and results were normalized to total protein. (L) Level of miR- 22-3p expression in vehicle or Mebhydrolin (15, 30 mg/kg)-treated HFD/STZ mice was detected by RT-PCR (N = 5–6). (M) Phosphorylated and total protein levels of PI3K, AKT and FoxO1 in the liver tissue of AV-NC or AAV-FXR-RNAi injected HFD/STZ mice (HFD/STZ + AAV-NC or HFD/STZ + AAV-FXR-RNAi) and Mebhydrolin (15 or 30 mg/kg)-treated AAV-FXR-RNAi injected HFD/STZ mice (HFD/STZ + AAV-FXR-RNAi + Mebhydrolin) were detected by Western blot, N = 3/group. (N) Quantification results for (M), and results were normalized to total protein. (O) Level of miR-22-3p was detected by RT-PCR in AAV-NC or AAV-FXR-RNAi injected HFD/STZ (HFD/STZ + AAV-NC or HFD/STZ + AAV-FXR-RNAi, N = 4) and Mebhydrolin (15 or 30 mg/kg)-treated AAV-FXR-RNAi injected HFD/STZ (HFD/STZ + AAV-FXR-RNAi + Mebhydrolin, N = 4) mice. (P-R) mRNA level of FoxO1 in the liver tissue of mice was detected by RT- PCR. (P) for vehicle or Mebhydrolin (15, 30 mg/kg)-treated db/db mice (N = 5–6), (Q) for vehicle or Mebhydrolin (15, 30 mg/kg)-treated HFD/STZ mice (N = 6) and (R) for AAV-NC or AAV-FXR-RNAi injected HFD/STZ mice (HFD/STZ + AAV-NC or HFD/STZ + AAV-FXR-RNAi, N = 6) and Mebhydrolin (15 or 30 mg/kg)-treated AAV-FXR-RNAi injected HFD/STZ mice (HFD/ STZ + AAV-FXR-RNAi+Mebhydrolin, N = 6). (S, T) Representative images of immunoﬂuorescence staining of liver sections in vehicle or Mebhydrolin (30 mg/kg/day)-treated (S) db/db or (T) HFD/STZ mice (N = 4). FoxO1 nuclear export was then observed by laser scanning confocal microscopy (LSCM). Scale bars: 25 μm; Red: FoxO1; Blue: nuclear. (U) Representative images of immunoﬂuorescence staining of liver sections in AAV-NC or AAV-FXR-RNAi injected HFD/STZ mice (HFD/STZ + AAV-NC or HFD/STZ + AAV-FXR-RNAi, N = 4) and Mebhydrolin (30 mg/kg/day)-treated AAV-FXR-RNAi injected HFD/STZ mice (HFD/STZ + AAV-FXR-RNAi + Mebhydrolin, N = 4). FoxO1 nuclear export was observed by LSCM. Scale bars: 25 μm; Red: FoxO1; Blue: nuclear. All data were presented as mean ± S.E.M from three independent experiments (*P < 0.05, **P < 0.01, ***P < 0.001). Fig. 7. Mebhydrolin promoted glycogen synthesis through FXR/miR-22-3p/PI3K/AKT/GSK3β pathway in T2DM mice. (A, B) Glycogen level of vehicle or Mebhydrolin (15, 30 mg/kg/day)- treated (A) db/db or (B) HFD/STZ mice was measured with commercial kit (N = 6). (C) Immunohistochemistry assay for PAS staining liver sections in vehicle or Mebhydrolin (30 mg/kg/day)-treated db/db mice (N = 4, scale bars, 100 μm (top), 50 μm (down)). Blue: Nucleus, Fuchsia: glycogen. (D) Quantification results for (C). (E) Immunohistochemistry assay for PAS staining liver sections in vehicle or Mebhydrolin (30 mg/kg/day)-treated HFD/STZ mice (N = 4, scale bars, 100 μm (top), 50 μm (down)). Blue: Nucleus, Fuchsia: glycogen. (F) Quantification results for (E). (G) Glycogen level of AAV-NC or AAV-FXR-RNAi injected HFD/STZ mice (HFD/STZ + AAV-NC or HFD/STZ + AAV-FXR-RNAi) and Mebhydrolin (15 or 30 mg/kg)-treated AAV-FXR-RNAi injected HFD/STZ mice (HFD/STZ + AAV-FXR-RNAi+Mebhydrolin) were measured with commercial kit (N = 8). (H) Immunohistochemistry assay for PAS staining liver sections in AAV-NC or AAV-FXR-RNAi injected HFD/STZ-induced T2DM mice (HFD/STZ + AAV-NC or HFD/STZ + AAV-FXR-RNAi) and Mebhydrolin (30 mg/kg)- treated AAV-FXR-RNAi injected HFD/STZ mice (HFD/STZ + AAV-FXR-RNAi + Mebhydrolin) (N = 4, scale bars, 100 μm (top), 50 μm (down)). Blue: Nucleus, Fuchsia: glycogen. (I) Quantification results for (H) (J, L) Phosphorylated and total protein levels of GSK3β in vehicle or Mebhydrolin (15, 30 mg/kg)-treated (J) db/db or (L) HFD/STZ mice were detected by Western blot, N = 3/group. (K, M) Quantification results for (J, L), and results were normalized to total protein. (N) Phosphorylated and total protein levels of GSK3β in AAV-NC or AAV-FXR-RNAi injected HFD/STZ mice (HFD/STZ + AAV-NC or HFD/STZ + AAV-FXR-RNAi) and Mebhydrolin (30 mg/kg)-treated AAV-FXR-RNAi injected HFD/STZ mice (HFD/STZ + AAV-FXR-RNAi + Mebhydrolin) were detected by Western blot, N = 3/group. (O) Quantification results for (N), and results were normalized to total protein. All data were presented as mean ± S.E.M from three independent experiments (*P < 0.05, **P < 0.01, ***P < 0.001). Fig. 8. Molecular docking with molecular dynamics simulation and site-directed mutation analysis provided a novel molecular basis for Mebhydrolin-mediated FXR antagonism against glucose homeostasis in T2DM mice. (A) Binding affinities of Mebhydrolin to FXR-LBD-L291A, FXR-LBD-M332A, FXR-LBD-Y373A were detected by MST. (B, C) Superposition of the antagonist NDB (B, white sticks) in crystal structure of (B, yellow sticks) HS218 and (C, green sticks) Mebhydrolin docked with FXR. (D) The RMSD and (E) RMSF values were calculated to examine the variations of the atoms on the protein backbone with respect to the initial structure frame. FXR/Mebhydrolin system was colored red and FXR/HS218 colored black. (F) The RMSD values were calculated to examine the variations of the atoms on the H2 and H6 backbone with respect to the initial structure frame. FXR/Mebhydrolin system was colored dark blue and FXR/HS218 colored light blue. (G) Superposition of the last 10 ns average structures of MD simulations among FXR/Mebhydrolin complex (pink), FXR/HS218 complex (blue) and the apo structure (white). (H) Mebhydrolin as a FXR antagonist bound to residues L291, M332 and Y373 of FXR-LBD, then transcriptionally regulated the expression of miR-22-3p that is required for the Mebhydrolin-mediated FXR regulation against PI3K/AKT signaling. Next, Mebhydrolin promoted FoxO1 translocation from nucleus through FXR/miR-22-3p/PI3K/AKT pathway to reduce the expressions of gluconeogenesis genes G6Pase and PEPCK, while increased the phosphorylation level of GSK3β (Ser 9) through FXR/miR-22-3p/PI3K/AKT pathway to promote glycogen synthesis. Finally, Mebhydrolin functioned in reducing glucose output and improving blood glucose balance. 2) Resource availability statements. The resource generated during and/or analyzed during the current study is available from the corre- sponding author upon reasonable request. References [1] Chatterjee S, Davies MJ, Heller S, Speight J, Snoek FJ, Khunti K. Diabetes structured self-management education programmes: a narrative review and current innova- tions. Lancet Diabetes Endocrinol. 2018;6(2):130–42. [2] Nyenwe EA, Jerkins TW, Umpierrez GE, Kitabchi AE. Management of type 2 diabetes: evolving strategies for the treatment of patients with type 2 diabetes. Metabolism. 2011;60(1):1–23. [3] Holstein A, Plaschke A, Egberts EH. Lower incidence of severe hypoglycaemia in pa- tients with type 2 diabetes treated with glimepiride versus glibenclamide. Diabetes Metab Res Rev. 2001;17(6):467–73. [4] Petersen MC, Vatner DF, Shulman GI. Regulation of hepatic glucose metabolism in health and disease. Nat Rev Endocrinol. 2017;13(10):572–87. [5] Han HS, Kang G, Kim JS, Choi BH, Koo SH. Regulation of glucose metabolism from a liver-centric perspective. Exp Mol Med. 2016;48:e218. [6] Pilkis SJ, Granner DK. Molecular physiology of the regulation of hepatic gluconeo- genesis and glycolysis. Annu Rev Physiol. 1992;54:885–909. [7] Lee JM, Han HS, Jung YS, Harris RA, Koo SH, Choi HS. The SMILE transcriptional core- pressor inhibits cAMP response element-binding protein (CREB)-mediated transactivation of gluconeogenic genes. J Biol Chem. 2018;293(34):13125–33. [8] Liu Q, Zhang L, Zhang W, Hao Q, Qiu W, Wen Y, et al. Inhibition of NF-kappaB re- duces renal inﬂammation and expression of PEPCK in type 2 diabetic mice. Inﬂam- mation. 2018;41(6):2018–29. [9] Yan H, Yang W, Zhou F, Li X, Pan Q, Shen Z, et al. Estrogen improves insulin sensitiv- ity and suppresses gluconeogenesis via the transcription factor Foxo1. Diabetes. 2019;68(2):291–304. [10] Barooti A, Kamran M, Kharazmi F, Eftakhar E, Malekzadeh K, Talebi A, et al. Effect of oral magnesium sulfate administration on blood glucose hemostasis via inhibition of gluconeogenesis and FOXO1 gene expression in liver and muscle in diabetic rats. Biomed Pharmacother. 2019;109:1819–25. [11] Zhang P, Tu B, Wang H, Cao Z, Tang M, Zhang C, et al. Tumor suppressor p53 coop- erates with SIRT6 to regulate gluconeogenesis by promoting FoxO1 nuclear exclu- sion. Proc Natl Acad Sci U S A. 2014;111(29):10684–9. [12] Kelly RP, Garhyan P, Raddad E, Fu H, Lim CN, Prince MJ, et al. Short-term administra- tion of the glucagon receptor antagonist LY2409021 lowers blood glucose in healthy people and in those with type 2 diabetes. Diabetes Obes Metab. 2015;17(4):414–22. [13] Xiong Y, Guo J, Candelore MR, Liang R, Miller C, Dallas-Yang Q, et al. Discovery of a novel glucagon receptor antagonist N-[(4-{(1S)-1-[3-(3, 5-dichlorophenyl)-5-(6- methoxynaphthalen-2-yl)-1H-pyrazol-1-yl]ethyl}phenyl)carbo nyl]-beta-alanine (MK-0893) for the treatment of type II diabetes. J Med Chem. 2012;55(13):6137–48. [14] Liu TY, Shi CX, Gao R, Sun HJ, Xiong XQ, Ding L, et al. Irisin inhibits hepatic gluconeo- genesis and increases glycogen synthesis via the PI3K/Akt pathway in type 2 diabetic mice and hepatocytes. Clin Sci (Lond). 2015;129(10):839–50. [15] Seo YH, Jung HJ, Shin HT, Kim YM, Yim H, Chung HY, et al. Enhanced glycogenesis is involved in cellular senescence via GSK3/GS modulation. Aging Cell. 2008;7(6): 894–907. [16] Shieh JM, Wu HT, Cheng KC, Cheng JT. Melatonin ameliorates high fat diet-induced diabetes and stimulates glycogen synthesis via a PKCzeta-Akt-GSK3beta pathway in hepatic cells. J Pineal Res. 2009;47(4):339–44. [17] Eldar-Finkelman H, Ilouz R. Challenges and opportunities with glycogen synthase ki- nase-3 inhibitors for insulin resistance and Type 2 diabetes treatment. Expert Opin Investig Drugs. 2003;12(9):1511–9. [18] Maqbool M, Hoda N. GSK3 inhibitors in the therapeutic development of diabetes, cancer and neurodegeneration: past, present and future. Curr Pharm Des. 2017;23 (29):4332–50. [19] Feng ZC, Donnelly L, Li J, Krishnamurthy M, Riopel M, Wang R. Inhibition of Gsk3beta activity improves beta-cell function in c-KitWv/+ male mice. Lab Investig. 2012;92 (4):543–55. [20] Markham A, Keam SJ. Obeticholic acid: first global approval. Drugs. 2016;76(12): 1221–6. [21] Lee FY, Lee H, Hubbert ML, Edwards PA, Zhang Y. FXR, a multipurpose nuclear recep- tor. Trends Biochem Sci. 2006;31(10):572–80. [22] Zhang Y, Lee FY, Barrera G, Lee H, Vales C, Gonzalez FJ, et al. Activation of the nuclear receptor FXR improves hyperglycemia and hyperlipidemia in diabetic mice. Proc Natl Acad Sci U S A. 2006;103(4):1006–11. [23] Ma K, Saha PK, Chan L, Moore DD. Farnesoid X receptor is essential for normal glu- cose homeostasis. J Clin Invest. 2006;116(4):1102–9. [24] Jiang C, Xie C, Lv Y, Li J, Krausz KW, Shi J, et al. Intestine-selective farnesoid X recep- tor inhibition improves obesity-related metabolic dysfunction. Nat Commun. 2015; 6:10166. [25] Xu X, Xu X, Liu P, Zhu ZY, Chen J, Fu HA, et al. Structural basis for small molecule NDB (N-benzyl-N-(3-(tert-butyl)-4-hydroxyphenyl)-2,6-dichloro-4-(dimethylamino) benzamide) as a selective antagonist of farnesoid X receptor alpha (FXRalpha) in stabilizing the homodimerization of the receptor. J Biol Chem. 2015;290(32): 19888–99. [26] Xu X, Shi X, Chen Y, Zhou T, Wang J, Xu X, et al. HS218 as an FXR antagonist sup- presses gluconeogenesis by inhibiting FXR binding to PGC-1alpha promoter. Metab- olism. 2018;85:126–38. [27] Soifer HS, Rossi JJ, Saetrom P. MicroRNAs in disease and potential therapeutic appli- cations. Mol Ther. 2007;15(12):2070–9. [28] Li S, Chen X, Zhang H, Liang X, Xiang Y, Yu C, et al. Differential expression of microRNAs in mouse liver under aberrant energy metabolic status. J Lipid Res. 2009;50(9):1756–65. [29] Yan C, Chen J, Li M, Xuan W, Su D, You H, et al. A decrease in hepatic microRNA-9 expression impairs gluconeogenesis by targeting FOXO1 in obese mice. Diabetologia. 2016;59(7):1524–32. [30] Yang F, Hu Y, Liu HX, Wan YJ. MiR-22-silenced cyclin A expression in colon and liver cancer cells is regulated by bile acid receptor. J Biol Chem. 2015;290(10):6507–15. [31] Dou L, Meng X, Sui X, Wang S, Shen T, Huang X, et al. MiR-19a regulates PTEN ex- pression to mediate glycogen synthesis in hepatocytes. Sci Rep. 2015;5:11602. [32] Fang W, Guo J, Cao Y, Wang S, Pang C, Li M, et al. MicroRNA-20a-5p contributes to hepatic glycogen synthesis through targeting p63 to regulate p53 and PTEN expres- sion. J Cell Mol Med. 2016;20(8):1467–80. [33] Kao CY, Factor VM, Thorgeirsson SS. Reduced growth capacity of hepatocytes from c- myc and c-myc/TGF-alpha transgenic mice in primary culture. Biochem Biophys Res Commun. 1996;222(1):64–70. [34] Liu H, Naismith JH. An efficient one-step site-directed deletion, insertion, single and multiple-site plasmid mutagenesis protocol. BMC Biotechnol. 2008;8:91. [35] Xu X, Lu Y, Chen L, Chen J, Luo X, Shen X. Identification of 15d-PGJ2 as an antagonist of farnesoid X receptor: molecular modeling with biological evaluation. Steroids. 2013;78(9):813–22. [36] Qiu BY, Turner N, Li YY, Gu M, Huang MW, Wu F, et al. High-throughput assay for modulators of mitochondrial membrane potential identifies a novel compound with beneficial effects on db/db mice. Diabetes. 2010;59(1):256–65. [37] Zhou TT, Ma F, Shi XF, Xu X, Du T, Guo XD, et al. DMT efficiently inhibits hepatic glu- coneogenesis by regulating the Galphaq signaling pathway. J Mol Endocrinol. 2017; 59(2):151–69. [38] Zhou TT, Zhao T, Ma F, Zhang YN, Jiang J, Ruan Y, et al. Small molecule IVQ, as a prodrug of gluconeogenesis inhibitor QVO, efficiently ameliorates glucose homeo- stasis in type 2 diabetic mice. Acta Pharmacol Sin. 2019;60(9):1193–204. https:// doi.org/10.1038/s41401-018-0208-2. [39] Li H, Li Y, Xiang L, Zhang J, Zhu B, Xiang L, et al. GDF11 attenuates development of type 2 diabetes via improvement of islet beta-cell function and survival. Diabetes. 2017;66(7):1914–27. [40] Sands MS. AAV-mediated liver-directed gene therapy. Methods Mol Biol. 2011;807: 141–57. [41] Glideversion 10.6 ed.. New York, NY: Schrödinger, LLC; 2016. [42] Maestro2016-1 ed.. New York, NY: Schrödinger, LLC; 2016. [43] Xu X, Xu X, Liu P, Zhu ZY, Chen J, Fu HA, et al. Structural basis for small molecule NDB (N-benzyl-N-(3-(tert-butyl)-4-hydroxyphenyl)-2,6-dichloro-4-(dimethylamino) benzamide) as a selective antagonist of farnesoid X receptor alpha (FXR alpha) in stabilizing the homodimerization of the receptor. J Biol Chem. 2015;290(32): 19888–99. [44] Eisenberg D, Lüthy R, Bowie JU. VERIFY3D: assessment of protein models with three- dimensional profiles. . 1992;356(6364):83–5. [45] O’Neill PA, Morton PB, Stark RD. Chlorpromazine—a specific effect on breathless- ness? Br J Clin Pharmacol. 1985;19(6):793–7. [46] Lu M, Wan M, Leavens KF, Chu Q, Monks BR, Fernandez S, et al. Insulin regulates liver metabolism in vivo in the absence of hepatic Akt and Foxo1. Nat Med. 2012;18(3): 388–95. [47] Zhao Y, Wang Y, Zhu WG. Applications of post-translational modifications of FoxO family proteins in biological functions. J Mol Cell Biol. 2011;3(5):276–82. [48] Prawitt J, Abdelkarim M, Stroeve JH, Popescu I, Duez H, Velagapudi VR, et al. Farnesoid X receptor deficiency improves glucose homeostasis in mouse models of obesity. Diabetes. 2011;60(7):1861–71. [49] Diniz GP, Huang ZP, Liu J, Chen J, Ding J, Fonseca RI, et al. Loss of microRNA-22 pre- vents high-fat diet induced dyslipidemia and increases energy expenditure without affecting cardiac hypertrophy. Clin Sci (Lond). 2017;131(24):2885–900. [50] Kaur K, Vig S, Srivastava R, Mishra A, Singh VP, Srivastava AK, et al. Elevated hepatic miR-22-3p expression impairs gluconeogenesis by silencing the Wnt-responsive transcription factor Tcf7. Diabetes. 2015;64(11):3659–69.PI3K Inhibitor Library