Empagliflozin Attenuates Hyperuricemia by Upregulation of ABCG2 via AMPK/AKT/CREB Signaling Pathway in Type 2 Diabetic Mice

1Hyperuricemia (HUA) is associated with a variety of diseases such as gout and diabetes. There is high prevalence of HUA among patients with type 2 diabetes (T2DM). Various studies have indicated that incidence of HUA in patients with T2DM was respectively 24.1% and 33.8%. Moreover, HUA has proven to be an independent risk factor for kidney disease progression,, and significantly correlated with the severity of retinopathy in patients with T2DM. Accumulating data has indicated that HUA was closely associated with risks of increased prevalence of cardiovascular disease,, carotid intima-media thickness, hypertension and metabolic syndrome in individuals with T2DM. Thus developing a new type of agent to improve hyperglycemia and HUA seems increasingly urgent.

2Uric acid (UA) is a product generated from metabolism of purines by xanthine oxidase (XO). In brief, HUA occurs when the balance between production and complicated course of UA handing in renal tubules and intestines is disrupted. UA transporters are involved in complex processes. Urate transporter 1 (URAT1, SLC22A12) and glucose transporter 9 (GLUT9, SLC2A9) are responsible for renal reabsorption, and ATP-binding cassette subfamily G member 2 (ABCG2), organic anion transporter 1 (OAT1, SLC22A6), organic anion transporter 3 (OAT3, SLC22A8) facilitates renal secretion. In addition, GLUT9 and ABCG2 has been also found to be expressed in intestines, which take part in the homeostasis of UA levels as well.

3The sodium-glucose co-transporter 2 (SGLT2) is primarily located in renal proximal tubules, and responsible for renal glucose reabsorption to maintain blood glucose homeostasis. SGLT2 inhibitors, which exert the effect of blocking SGLT2 to promote urinary glucose excretion, have been developed as a newly approved oral hypoglycemic drug. Empagliflozin, a potent and selective SGLT2 inhibitor, has been reported to reduce cardiovascular mortality, reduce blood pressure, and slower progression of kidney disease in addition to glycemic control,. Notably, it has been reported that empagliflozin helped to reduce serum uric acid (SUA),.

4To date, the underlying mechanisms that empagliflozin reduces SUA remains largely unexplored. This study is designed to assess the role of empagliflozin on SUA in KK-Ay mice with HUA, and the underlying mechanism is involved with a focus on SGLT2 inhibitor on renal and intestinal UA transporters.

5The body weight of mice from KK-Ay group, KK-Ay+HUA group, KK-Ay+HUA+EMP group was significantly increased compared with the body weight of C57BL/6J mice (Supplemental Figure 1A), while there was no significant difference in body weight among the KK-Ay group, KK-Ay+HUA group, KK-Ay+HUA+EMP group (Fig. 1A). As expected, the blood glucose in the KK-Ay group, KK-Ay+HUA group, KK-Ay+HUA+EMP group were significantly higher than that in the C57BL/6J mice (Supplemental Figure 1B), and there were no significant difference in blood glucose levels between KK-Ay group and KK-Ay+HUA group (Fig. 1B, Supplemental Figure 1B), but empagliflozin treatment notably lowered blood glucose levels (Fig. 1B, Supplemental Figure 1B). Next, we observed no notable difference in SUA between C57BL/6J group and KK-Ay group (Supplemental Figure 1C), however, the SUA of mice in KK-Ay+HUA group were dramatically increased (Fig. 1C, Supplemental Figure 1C), while empagliflozin obviously reduced the SUA (Fig. 1C, Supplemental Figure 1C). It was also demonstrated that the urinary UA levels and FEUA were significantly higher for KK-Ay+HUA+EMP group than KK-Ay group and KK-Ay+HUA group (Fig. 1D-E, Supplemental Figure 1D-E). Moreover, we found that empagliflozin increased the urine volume and urinary glucose excretion (Supplemental Figure 1F-G).

6HUA is confirmed as an independent risk factor for kidney diseases -, and chronic kidney disease (CKD) is a common comorbidity in T2DM,. By western blot, we observed SGLT2 was not only expressed in kidney, but also in ileum and the levels of ABCG2 in ileum were notably reduced in ileum compared with kidney (Supplemental Figure 2A-B). As ileum played an important role in UA excretion, we explored the mechanism not only in kidney but also in ileum. Hematoxylin-eosin (HE) and periodic acid-Schiff (PAS) staining showed that mice in the KK-Ay group and KK-Ay+HUA group displayed drastic tubular dilatation, hydropic degeneration and deposited mesangial matrix compared with mice in the C57BL/6J group, however, those pathological features were significantly attenuated in the KK-Ay+HUA+EMP group (Fig. 2A, C, Supplemental Figure 3A, C). We also examined the histological changes in ileum, and found empagliflozin remarkably improved the histology of ileum in the KK-Ay+HUA+EMP group (Fig. 2B, Supplemental Figure 3B).

7To find the mechanism responsible for anti-hyperuricemic effect of empagliflozin, we evaluated the expression of the major UA transporters by RT-PCR, western blot and immunohistochemistry in both kidneys and small intestines. As shown in Fig. 3A and Supplemental figure 4A, empagliflozin notably increased the mRNA levels of ABCG2 and significantly reduced the URAT1 mRNA compared with KK-Ay+HUA group in kidney, but had no effect on mRNA of OAT1, OAT3 and GLUT9 in kidney. In small intestines, we observed that ABCG2 is mainly expressed in the ileum (Fig. 3B) and GLUT9 has similar expression trends in the three groups (Fig. 3C). In ileum, empagliflozin remarkably up-regulated the mRNA levels of ABCG2, but had no effect on GLUT9 mRNA expression (Fig. 3C, Supplemental figure 4B). Interestingly, results from western blot and immunohistochemistry demonstrated that the protein levels of ABCG2, but not OAT1, OAT3, URAT1 or GLUT9 were obviously altered in the kidneys after empagliflozin treatment (Fig. 3D-E, H-J, Supplemental Figure 4C-D, G, I, Supplemental Figure 5A-B). Similarly, in ileum, empagliflozin notably increased the protein levels of ABCG2, while did not significantly affect the expression of GLUT9 (Fig. 3F-G, I-J, Supplemental figure 4E-F, H-I).

8As demonstrated in Fig. 4A-D, renal and ileal tissues were homogenized, and the levels of p-AMPK, p-AKT and p-CREB were examined. It turned out that no significant changes in the levels of p-AMPK, p-AKT and p-CREB were observed in kidney or ileum in mice from the KK-Ay group and KK-Ay+HUA group. In contrast, the levels of p-AMPK, p-AKT and p-CREB were sharply enhanced in the KK-Ay +HUA+EMP group.

9To confirm the mechanism of empagliflozin, AMPK inhibitor (Compound C) was applied to cultured human tubular epithelial (HK-2) cells prior to empagliflozin treatment. Western blot and immunofluorescence showed that ABCG2 was significantly increased in the group of HG+UA+EMP. However, ABCG2 was notably decreased in the HG+UA +Compound C +EMP group (Fig. 5A-B, E-F). Moreover, we found that the levels of p-AMPK, p-AKT, p-CREB were obviously increased in the HG+UA+EMP group, while significantly decreased after treatment with Compound C (Fig. 5C-D). Our observation further suggested that empagliflozin promoted ABCG2 expression through AMPK/AKT/CREB pathway in vitro.

10To further investigate the potential mechanism between CREB and ABCG2, we conducted ChIP assays using HK-2 cells. According to the JASPAR database, we predicted a motif that CREB facilitates ABCG2 expression (Fig. 6A). Then we designed 2 pairs of primers based on ABCG2 promoter region (Table. 1). As shown in Fig. 6B-C, ChIP-PCR results indicated that CREB may bind to ABCG2 promoter and the signal relative to input in anti-CREB group is dramatically higher than the negative anti-IgG group. We next examined whether CREB physically interacts with the ABCG2 promoter by ChIP experiments using HEK-293A cells transfected with CREB plasmid and results showed CREB activated ABCG2 via binding to the promoter of ABCG2 to induce transcription. (Fig. 6D).

11Next we examined the transfection efficiency of ABCG2 promoter plasmid and CREB plasmid by PCR and western blot, respectively. As shown in Fig. 7A-D, ABCG2 and CREB were overexpressed compared with their blank negative control. We also assessed whether CREB binding to the ABCG2 promoter by Gaussia luciferase reporter assays. HEK-293A cells were next transiently co-transfected with the following four groups: ABCG2NC+CREBNC, ABCG2+CREBNC, ABCG2NC+CREB, ABCG2+CREB. As expected, the luciferase activity in the group of ABCG2+CREB was significantly increased (Fig. 7E). The plasmid maps can be obtained in the Supplemental figure 6. Then, we mutated the 511 to 518 sites, 833 to 840 sites or both of them in the ABCG2 promoter region and co-transfected with CREB plasmid respectively. Results showed that 833 to 840 mutant-sites group showed significantly reduced luciferase activity but it was not obvious for 511 to 518 mutant-sites group (Fig. 7F-G).

12The present study was firstly conducted in the model of T2DM mice with HUA, and demonstrated that empagliflozin treatment improved HUA through enhancing ABCG2 expression via AMPK/AKT/CREB pathway in both kidney and ileum. A remarkable improvement in biochemical parameters and histopathologic changes in vivo were observed after empagliflozin treatment, indicating the SUA-lowering effect of empagliflozin. Notably, our data from HK-2 cells after empagliflozin and AMPK inhibitor Compound C treatment further confirmed the mechanism that empagliflozin exerted anti-hyperglycemic effect in vitro.

13HUA carries a significant risk for developing various diseases in T2DM -, therefore, exploring the metabolism of UA and associated diseases has attracted researchers' much attention. However, it is a challenge to create hyperuricemic models in mice, because in rodents, uricase oxidizes UA to allantoin which is highly soluble in water and excreted unchanged in the urine, but not in humans. As potassium oxonate (PO) can block uricase, and hypoxanthine (HX) is the precursor of UA which is finally oxidized into UA by xanthine oxidase, combination of PO and HX has been widely used to induce HUA in rodents -. As our data indicated, we successfully induced HUA in KK-Ay mice by combination of peritoneal injection of PO and intragastric administration of HX. Meanwhile, empagliflozin administration markedly suppressed SUA in KK-Ay mice with HUA, suggesting the hypouricemic effect of empagliflozin in T2DM mice.

14It is well known that kidney is the main organ for excretion of approximately two thirds of the total UA. UA handling involves several classes of UA transporters responsible for UA handing. In brief, after glomerular filtration, UA is reabsorbed through URAT1 and GLUT9, and secreted through ABCG2, OAT1, and OAT3. According to previous studies, about one third of the total UA was excreted into the intestinal tract, and decreased extra-renal UA excretion is a common cause of HUA. Intestinal transporters play significant roles in mediating UA mobilization as well. ABCG2 is expressed on the apical membrane in intestines as well as kidneys. ABCG2 regulates SUA via physiologically important roles in both renal and intestinal UA excretion. Increased SUA in patients with ABCG2 dysfunction could be explained by the decreased excretion of UA from the intestine. Previous studies suggested that GLUT9 which was localized to the apical and basolateral enterocyte membranes, was abundantly expressed in intestine,, especially in the jejunum and ileum. Furthermore, it functioned to regulate enterocyte UA clearance, and enterocyte GLUT9-deficient mice developed HUA, hyperuricosuria and early-onset metabolic syndrome. Consistently, we detected low expression of ABCG2 in the KK-Ay group and KK-Ay+HUA group. In contrast, after empagliflozin administration, the SUA were decreased and the ABCG2 expression was up-regulated in both renal and ileal tissues in KK-Ay mice. This finding suggested that empagliflozin might exert anti-hyperuricemic effects by facilitating UA excretion through promoting ABCG2 expression in kidneys and ilea.

15SGLT2 inhibitors have recently emerged as a promising novel class of glucose lowering drugs to facilitate glucose excretion in urine independent of insulin. Accumulating data indicated their anti- hyperuricemic effects in both clinical,, and basic studies,. One of basic studies performed in vitro using Xenopus oocytes indicated that luseogliflozin lowered the SUA level due to inhibition of UA reabsorption mediated by GLUT9 isoform 2 at the collecting duct of the renal tubule. Another study conducted in STZ-induced diabetic rats suggested that empagliflozin may exert anti-hyperuricemic effects via regulating URAT1 and ABCG2. However, our study demonstrated that empagliflozin promoted ABCG2 expression in kidney and ileum in KK-Ay mice with HUA, but has little impact on the other transporters. The difference of the animal model may explain the difference to a certain extent. Although SGLT2 transporters are primarily located in the kidney, they are also found in other tissues, such as the small intestine, especially ileum. In our investigation, we also confirmed that SGLT2 was expressed in ileum (Supplemental figure. 2A-B).

16Empagliflozin has been previously reported to trigger AMPK activation, which increased phosphorylation of AKT, resulting in CREB phosphorylation. Furthermore, a previous study implied that the CREB facilitated ABCG2 expression through promoter activation. The present study showed that empagliflozin treatment significantly promoted CREB directly bind to ABCG2 promoter through activating AMPK/AKT/CREB pathway. Moreover, with the application of AMPK inhibitor (Compound C), our data further confirmed that empagliflozin enhanced ABCG2 expression by promoting phosphorylation of AMPK, AKT and CREB.

17As summarized in Figure 8 our study demonstrated that empagliflozin treatment possessed anti-hyperglycemic effects and was attributed to UA excretion promotion through up-regulating ABCG2 expression in kidney and ileum in KK-Ay mice with HUA and in HK-2 cells. We also found that empagliflozin promoted the phosphorylation of AMPK, AKT and CREB and further activated ABCG2 by facilitating CREB binding to the promoter of ABCG2 to induce transcription. The findings in our studies lead to a new understanding of the mechanism about the effects of empagliflozin on improving HUA, and provided novel insights into the targeted therapies for HUA. Further studies are needed to explore whether other mechanisms are also involved in the SUA-lowering effect of empagliflozin in T2DM with HUA.

18Six-week-old male C57BL/6J (24±0.5g) and KK-Ay mice (25 ± 0.4g) were obtained from Beijing HFK Bioscience Co. Ltd. (Beijing, China). They were housed at 24 ± 2°C. C57BL/6J mice were fed regular chow and there were considered as control group. KK-Ay mice were a kind of T2DM mouse model, and allowed free access to high-fat diet which consisted of 48.5% carbohydrates, 17.5% protein, 17.9% fat (Beijing HFK Bio-Technology Co. Ltd). KK-Ay mice were randomly assigned to 3 groups: diabetic control group (KK-Ay group, n=10), T2DM with HUA group (KK-Ay+HUA group, n=10), empagliflozin-treated KK-Ay mice with HUA group (KK-Ay+HUA+EMP group, n=10). The hyperuricemic model was induced by combination of peritoneal injection of PO at dose of 250mg/kg and intragastric administration of HX at dose of 300mg/kg daily, and mice in the KK-Ay+HUA+EMP group were orally administered with empagliflozin (10 mg/kg) in KK-Ay mice with HUA daily for 8 weeks. The dose of empagliflozin was determined based on previous studies. The study was approved by the ethical committee of Tianjin Medical University, and all procedures involving mice were conducted according to the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health as well as the guidelines of the Animal Welfare Act.

19At the end of treatment, 10 mice from each group were placed in individual metabolic cages for 24 hours to collect the urine. The samples of blood were obtained from the retroorbital venous plexus. The blood and the urine were then used for analysis of SUA, urine UA with an automatic biochemistry analyzer (Roche). The fractional excretion of uric acid (FEUA) was calculated according to the formula: FEUA= (urine UA× serum creatinine) / (serum UA × urine creatinine) × 100, expressed as percentage. Mice were sacrificed by intraperitoneal injection of chloral hydrate.

20Kidneys were fixed in 4% paraformaldehyde and were then made into paraffin blocks. After deparaffinization and dehydration, HE staining and PAS staining of renal tissues were performed according to the manufacturer's instructions (Leagene Biotechnology, Beijing).

21For Immunohistochemistry staining, kidney sections were heated in sodium citrate buffer (pH 6.0) for antigen retrieval. After they were blocked with 3% H2O2 and 5% BSA, they were incubated with primary antibodies ABCG2 (1:200, Affinity), GLUT9 (1:100, Novus), URAT1 (1:100, Abbiotec), OAT1 (1:100, Bioworld), OAT3 (1:100, Bioworld) overnight at 4 °C. Secondary antibodies were subsequently used to incubate for 30 minutes at 37°C and a diaminobenzidine (DAB) kit was performed to visualize the positive expression. After that, sections counterstained with haematoxylin. Stained sections were captured with a light microscopy in 10 randomly cortical sections (×400) and quantified with Image Pro Plus 6.0 software.

22HK-2 cells were cultured with DMEM/F12 medium containing 10% fetal bovine serum (FBS) and 1% penicillin/streptomycin in the presence of 5% CO2. The cells were divided into four groups: (1) high glucose (HG) group incubated in DMEM/HG medium, (2) HG+UA group incubated in UA (10mg/dl) added to DMEM/HG medium (HG+ UA) group, (3) HG+UA+EMP group incubated in DMEM/HG medium with UA (10mg/dl) and treated with 50 μM empagliflozin, (4) HG+UA+ Compound C group incubated in DMEM/HG medium with UA (10mg/dl), AMPK inhibitor Compound C (10μM) and empagliflozin (50 μM) for 48 hours.

23HEK-293A cells were cultured with DMEM/HG medium containing 10% FBS and 1% penicillin/ streptomycin. Transient transfection was performed using Lipofectamine 2000 (Invitrogen, USA) according to the manufacturer's instructions. To study whether CREB activated ABCG2 expression by facilitating CREB binding to the promoter of ABCG2, HEK-293A cells were divided to the following groups: (1) empty plasmid as control for ABCG2 promoter plasmid (ABCG2NC), (2) ABCG2 promoter plasmid (ABCG2), (3) empty plasmid as control for CREB plasmid (CREBNC), (4) CREB plasmid (CREB), (5) ABCG2 promoter plasmid and empty plasmid as control for CREB plasmid (ABCG2+CREBNC), (7) empty plasmid as control for ABCG2 promoter plasmid and CREB plasmid (ABCG2NC+CREB), (8) ABCG2+CREB, ABCG2 promoter plasmid and CREB plasmid, (9) mutation of the promoter at 511 to 518 sites in ABCG2 promoter region and CREB plasmid (ABCG2mut511+CREB); (10) mutation of the promoter at 833 to 840 sites and CREB plasmid (ABCG2mut833+CREB); (11) mutation of the promoter at 511 to 518 sites and 833 to 840 sites and CREB plasmid (ABCG2 double mut511/833+CREB).

24Total RNA was extracted using E.Z.N.A. Total RNA Kits (OMEGA, GA, USA). 1 μg of RNA was reverse-transcribed with Revert Aid First Strand cDNA Synthesis Kits (Thermo, MA, USA) according to the manufacturer's instructions. Reverse transcription polymerase chain reaction (RT-PCR) was performed using the CFX96 real-time PCR system (Bio-Rad, USA) with the SYBR Green PCR Kit (Takara, Otsu, Japan). The primers were listed in the table 2.

25After protein extracts from kidney, ileum and cells were boiled in lysis buffer, and they were subsequently loaded in 10% SDS-PAGE gels for electrophoresis. Then the protein was transferred to a PVDF membrane (Millipore, MA, USA). After blocked in 5% non-fat milk in TBST for 2 hours at room temperature, the membrane was incubated with primary antibody overnight at 4 °C. The primary antibodies included ABCG2 (1:100, Santa Cruz Biotechnology), GLUT9 (1:1000, Novus), URAT1 (1:200, Abbiotec), OAT1 (1:500, Bioworld), OAT3 (1:500, Bioworld), p-AMPK (1:1000, Cell Signaling), AMPK (1:1000, Cell Signaling), p-CREB (1:1000, Cell Signaling), CREB (1:1000, Cell Signaling), p-AKT (1:1000, Cell Signaling), AKT (1:1000, Cell Signaling). After that, they were incubated with the appropriate secondary antibody for 2 hours at room temperature. ECL kit (Advansta, USA) was used to visualize the bands and ImageJ software was performed to quantify the band intensity.

26After 48 hours of intervention, different groups of HK-2 cells cultured in 24-well plates were fixed with 4% paraformaldehyde. After that, cells were blocked with 1% BSA and then incubated with anti-ABCG2 (1:25, Affinity) at 4°C overnight. Then, the cells were incubated with secondary antibodies at 37 °C for 1 hour and cell nucleus was stained by DAPI for 2 minus in the dark.

27Cells were first cross-linked with 1% formaldehyde, and then lysed by Micrococcal Nuclease (Cell Signaling). Chromatin was harvested and subjected to immunoprecipitation with IgG antibody and p-CREB antibody (1:50, Cell Signaling). The protein-DNA complexes were reversed and the DNA was purified using the Simple ChIP Kit (Cell Signaling). The enrichment of the particular DNA sequence was detected by PCR (Bio-Rad, USA).

28A luciferase reporter gene assay was performed using a Secrete-Pair Gaussia Luciferase Assay kit (Gene Copoeia, Inc.). HEK-293A cells were seeded at a density of 1.5x104/ml into 96-well plates. After 12 hours, the cells were transiently co-transfected with 0.1 μg plasmid per well. 48 hours after transfection, supernatants of transfected cells were collected, and then the luciferase activity in the supernatants was measured using SynergyMx (BioTek, USA) according to the manufacturer's instructions.

29All the data were presented as mean ± standard error of mean (SEM). Two tailed, unpaired Student's t-test was applied in statistical analysis between two groups. One-way analysis of variance (ANOVA) was performed to determine statistical significance among more than three groups, and Tukey's post hoc test was used to compare any two groups among them. P < 0.05 was considered significant.