catherine only

Contents lists available at ScienceDirect

Biomedicine & Pharmacotherapy

journal homepage: www.elsevier.com/locate/biopha

Effects of the SGLT2 inhibitor ipragliflozin on food intake, appetite- regulating hormones, and arteriovenous differences in postprandial glucose levels in type 2 diabetic rats

Atsuo Taharaa,⁎, Yoshinori Kondob, Toshiyuki Takasua, Hiroshi Tomiyamab

a Candidate Discovery Science Laboratories, Astellas Pharma Inc., Tsukuba, Ibaraki, Japan b Research and Development Department, Kotobuki Pharmaceutical Co., Ltd., Hanishina-gun, Nagano, Japan

A R T I C L E I N F O

Keywords: Food intake Hyperglycemia Ipragliflozin SGLT 2 Type 2 diabetes

A B S T R A C T

Aims: The sodium-glucose cotransporter (SGLT) 2 inhibitor, ipragliflozin, improves not only hyperglycemia but also obesity in type 2 diabetic animals and patients; however, there have been concerns that it may also cause an increase in compensatory food intake. Appetite is regulated by complex mechanisms involving the central nervous system, part of which involves appetite-related hormones and arteriovenous differences in postprandial glucose levels. We evaluated the effect of ipragliflozin in type 2 diabetic rats on food intake, appetite-related hormones and arteriovenous differences in postprandial glucose levels, and their correlation with food intake. Main methods: Ipragliflozin and several antidiabetic drugs were administered to type 2 diabetic rats and various parameters concerning food intake were measured. Key findings: Ipragliflozin significantly increased urinary glucose excretion and reduced postprandial hy- perglycemia. Compared to normal rats, diabetic rats exhibited hyperphagia and elevated plasma levels of the appetite-stimulating hormones neuropeptide Y and ghrelin. Ipragliflozin induced significant weight loss and reduced plasma levels of appetite-stimulating hormones without affecting food intake. Diabetic rats exhibited a significantly reduced arteriovenous difference in postprandial glucose levels due to insulin insufficiency; this was improved by ipragliflozin, which increased renal arteriovenous differences in glucose levels by increasing ur- inary glucose excretion. Significance: These results indicate that the SGLT2 inhibitor, ipragliflozin, exerts anti- hyperglycemic actions by increasing urinary glucose excretion, and induces weight loss without a compensatory increase in food intake in type 2 diabetic mice. The mechanisms underlying these effects can be attributed, in part, to an increased arteriovenous difference in postprandial glucose levels and improved regulation of appetite- related hormones in the diabetic animal model. While this study was conducted in rodents and the results may be distinct from those in humans, it is possible that some of the pharmacological mechanisms, including the reg- ulation of appetite-related hormones, can be extrapolated to clinical settings and may be valuable for further studies including clinical investigations.

1. Introduction

Food intake is regulated by opposing neural activities in the satiety center (the ventromedial hypothalamic nucleus) and the hunger center (the lateral hypothalamic area) [1]. Regulation of food intake by the hypothalamus is highly sensitive to changes in plasma glucose con- centrations. In the 1950 s, Mayer proposed the glucostatic theory, which states that the glucose utilization rate in the hypothalamus acts as a stimulus for sensations of hunger and satiety: a small arteriovenous difference in glucose levels stimulates hunger, while a large arter- iovenous difference stimulates satiety [2]. Later studies revealed the

highly complex regulatory mechanisms governing food intake, which involve not only the central nervous system, but also various hormones and gastrointestinal peptides in the periphery [3]. In particular, leptin, which is secreted by adipose tissue [4], acts on the hypothalamus and the sympathetic nervous system to suppress appetite and promote en- ergy consumption, and is strongly associated with the pathogenesis of obesity and type 2 diabetes [5].

In recent years, sodium-glucose cotransporter (SGLT) 2 inhibitors, which inhibit the reabsorption of filtered glucose in the kidney to in- crease urinary glucose excretion, have been developed and proposed as novel antihyperglycemic agents for treating type 2 diabetes [6]. In

https://doi.org/10.1016/j.biopha.2018.06.062 Received 5 March 2018; Received in revised form 13 June 2018; Accepted 13 June 2018

⁎ Corresponding author at: Candidate Discovery Science Laboratories, Astellas Pharma Inc., 21 Miyukigaoka, Tsukuba, Ibaraki 305-8585, Japan. E-mail address: atsuo.tahara@jp.astellas.com (A. Tahara).

Biomedicine & Pharmacotherapy 105 (2018) 1033–1041

0753-3322/ © 2018 Published by Elsevier Masson SAS.

T

addition to antihyperglycemic actions, SGLT2 inhibitors also have an- tiobesity effects in type 2 diabetic animals and patients. Since the an- tiobesity effects of SGLT2 inhibitors are thought to stem from increased glucose (energy) excretion and subsequent improvement in glucose metabolism [7,8], concerns have been raised regarding a potential concomitant compensatory increase in food intake. Several preclinical and clinical studies have reported small to moderate increases in compensatory food intake resulting from SGLT2 inhibitors [9]. How- ever, these increases were not sufficient to cancel the improvement in obesity. We previously evaluated the pharmacological effect of the SGLT2 inhibitor, ipragliflozin, in type 2 diabetic mice, and showed that ipragliflozin exerts antihyperglycemic and antiobesity effects by in- creasing urinary glucose excretion without obvious increases in com- pensatory food intake [10]. However, the effects of SGLT2 inhibitors on various appetite-regulating mechanisms have not been studied in detail.

Here, we compared the effects of ipragliflozin and other antidiabetic drugs on food intake in type 2 diabetic rats. We focused on appetite control mechanisms by analyzing the arteriovenous difference in post- prandial glucose levels as a representative central regulatory me- chanism, and plasma levels of appetite-related hormones as a periph- eral regulatory mechanism.

2. Materials and methods

2.1. Materials

Ipragliflozin and sitagliptin were synthesized at Astellas Pharma Inc. (Ibaraki, Japan). Streptozotocin, metformin, and glibenclamide were purchased from Sigma-Aldrich Co. (St. Louis, MO, USA), and in- sulin (Novolin®R 100) was purchased from Novo Nordisk Pharma, Ltd. (Tokyo, Japan). Voglibose (BASEN®) was purchased from Takeda Pharmaceutical Company, Ltd. (Osaka, Japan) and purified at Astellas Pharma Inc. Insulin was diluted in physiological saline and adminis- tered intraperitoneally. All other drugs were dissolved or suspended in 0.5% methylcellulose (MC) solution and administered orally via a sto- mach tube. The vehicle-treated group received oral administration of 0.5% MC solution. Doses of all drugs were expressed as the free base form.

2.2. Animal models

Male Wistar rats (aged 6 weeks) were obtained from Charles River Japan (Shizuoka, Japan) and used at 7 weeks of age. Streptozotocin- nicotinamide-induced mildly diabetic rats were generated by fasting rats overnight and intraperitoneally administering a nicotinamide so- lution (100 mg/kg) followed by a streptozotocin solution (50 mg/kg) 20 min later [11]. Normal control rats were administered physiological saline. Two weeks later, blood glucose levels were measured and dia- betic rats were grouped such that blood glucose levels were uniform among the groups (normal: 129 ± 4 mg/dL, diabetes: 274 ± 4 mg/ dL). All rats were housed under conventional conditions with controlled temperature, humidity, and light exposure (12-h light-dark cycle), with access to standard commercial diet and water (ad libitum). Animals were handled and cared for in accordance with the Guide for the Care and Use of Laboratory Animals, and all procedures were approved by the Animal Ethical Committee of Astellas Pharma Inc.

2.3. Effect of ipragliflozin on urinary glucose excretion

Ipragliflozin (0.03–3 mg/kg) was administered orally to diabetic rats, and spontaneously voided urine was collected for 24 h while the animals were kept in metabolic cages. After measuring urine volume, urine glucose concentration was measured using Glucose CII test re- agent (Wako Pure Chemical Industries, Ltd., Osaka, Japan).

2.4. Effect of ipragliflozin on blood glucose and plasma insulin levels during the liquid meal tolerance test

After measuring baseline overnight fasting blood glucose and plasma insulin levels, liquid meal (ENSURE® H, 20 mL/kg, containing 206 mg/mL carbohydrates, 53 mg/mL fat, and 53 mg/mL protein; Abbott, Osaka, Japan) was orally loaded 0.5 h after ipragliflozin ad- ministration (0.03–3 mg/kg). Blood samples were obtained from a tail vein immediately before and 0.5, 1, and 2 h after liquid meal loading, and blood glucose concentrations were measured as described above. Plasma insulin levels were measured using an enzyme-linked im- munosorbent assay (ELISA) kit (Morinaga Institute of Biological Science, Inc., Kanagawa).

2.5. Effects of ipragliflozin and various antidiabetic drugs on body weight and food intake

Ipragliflozin (1 mg/kg), metformin (300 mg/kg), voglibose (0.3 mg/ kg), glibenclamide (3 mg/kg), sitagliptin (1 mg/kg), and insulin (0.2 IU/kg) were administered to diabetic rats. Doses of antidiabetic drugs were chosen based on previous studies [11]. Ipragliflozin and si- tagliptin were administered once a day, while the other drugs were administered twice a day for two weeks, with body weight and food intake measured weekly. After the final drug administration on Day 14, blood samples were collected under nonfasting conditions. Hemoglobin A1c (HbA1c) levels were measured using a DCA2000 System (Bayer Medical, Tokyo); plasma insulin levels were measured as described above; and plasma levels of appetite regulating hormones, neuropep- tide Y (Phoenix Pharmaceuticals; Burlingame, CA, USA), ghrelin (LSI Medience Corporation, Tokyo, Japan), leptin (R&D systems; Minnea- polis, MN, USA), and peptide YY (Phoenix Pharmaceuticals), were measured using commercial ELISA kits.

2.6. Effects of ipragliflozin and various antidiabetic drugs on arteriovenous differences in plasma glucose levels

After overnight fasting, blood was collected from rats under iso- flurane anesthesia for measuring fasting plasma glucose levels. Ipragliflozin and several antidiabetic drugs were administered to fasted rats. Thirty-minutes later, the rats were fed, and blood was collected 1–1.5 h later under anesthesia for measuring postprandial plasma glu- cose levels. Blood collection from arteries and veins (approximately 0.5 mL from each site) was performed as follows: blood was first col- lected from the renal vein and then the renal artery (at the branch point from the abdominal artery) and the abdominal vein. Bleeding was stopped by clipping. Blood was then collected from the portal vein and bleeding was stopped before the last collection from the thoracic artery. The arteriovenous difference in glucose levels was calculated from the glucose concentrations in blood collected from the thoracic artery and abdominal vein. The difference in glucose levels between the portal and abdominal veins, and the renal arteriovenous difference in glucose were also calculated to evaluate liver and kidney function, respectively. Plasma levels of glucose, insulin, and appetite-related hormones were measured from the blood collected from the abdominal vein.

2.7. Statistical analysis

The results were expressed as mean ± standard error of the mean (S.E.M.). The areas under the curve (AUCs) were calculated from blood glucose and plasma insulin concentrations measured over time. Differences between two groups were determined using the Student’s t- test. Differences between multiple groups were assessed using Dunnett’s multiple comparisons test. The vehicle-treated group received an oral dose of 0.5% MC solution. All drug-treated groups, including the in- sulin-treated group (diluted in physiological saline and administered intraperitoneally), were compared with this vehicle group. P < 0.05

A. Tahara et al. Biomedicine & Pharmacotherapy 105 (2018) 1033–1041

1034

indicated statistical significance. Statistical and data analyses were conducted using GraphPad Prism 5 (GraphPad Software, La Jolla, CA, USA).

3. Results

3.1. Urinary glucose excretion and glucose tolerance

Type 2 diabetic rats showed glucose intolerance due to impaired early insulin secretion, hyperglycemia, glycosuria, and hyperphagia (Figs. 1–3). Single administration of ipragliflozin (0.03–3 mg/kg) dose- dependently increased urinary glucose excretion and urine volume, and these effects were significant at doses of 1 mg/kg or higher (Fig. 1A, B), while urinary excretion of electrolytes was unaffected (Fig. 1C–E). Further, during the liquid meal tolerance test, single administration of

ipragliflozin (0.03–3 mg/kg) dose-dependently improved glucose tol- erance and decreased plasma insulin levels, and these effects were significant at doses of 0.3 mg/kg or higher (Fig. 2).

3.2. Repeated administration study

In the two-week repeated administration study, ipragliflozin (1 mg/ kg) and various antidiabetic drugs, metformin (300 mg/kg), voglibose (0.3 mg/kg), glibenclamide (3 mg/kg), sitagliptin (1 mg/kg), and in- sulin (0.2 IU/kg), significantly reduced HbA1c levels (Fig. 3A). Insulin and the insulin secretagogues, glibenclamide and sitagliptin, sig- nificantly increased plasma insulin levels, while the remaining drugs showed a trend towards significantly (P < 0.1) decreasing plasma in- sulin levels (Fig. 3B). Metformin and voglibose significantly reduced food intake, while the other drugs had no affect (Fig. 3C). Metformin

Fig. 1. Effects of ipragliflozin on (A) urinary glucose excretion, (B) urine volume and urinary electrolyte excretion, (C) Na+, (D) K+, and (E) Cl− in type 2 diabetic rats. Ipragliflozin was orally administered to rats, and spontaneously voided urine was collected for 24 h. Values indicate mean ± S.E.M. for six animals per group. *P < 0.05 vs. normal group, #P < 0.05 vs. vehicle group. N: normal, V: diabetic vehicle.

A. Tahara et al. Biomedicine & Pharmacotherapy 105 (2018) 1033–1041

1035

and voglibose also significantly reduced body weight gain, while ipra- gliflozin induced weight loss regardless of food intake (Fig. 3D), and the remaining drugs had no effect on body weight. Compared to normal rats, diabetic rats had significantly elevated plasma levels of the ap- petite-stimulating hormones, neuropeptide Y and ghrelin, while levels of the appetite-suppressing hormones, leptin and peptide YY, were unchanged (Fig. 4). All of the antidiabetic drugs tested, except vogli- bose, significantly reduced or showed a trend towards significantly (P < 0.1) reducing the plasma levels of appetite-stimulating hormones (Fig. 4A, B). In contrast, these drugs did not significantly alter the plasma levels of appetite-suppressing hormones, although a trend to- wards a significant (P < 0.1) increase was observed with insulin and insulin secretagogues (Fig. 4C, D).

3.3. Arteriovenous differences in plasma glucose levels

Diabetic rats exhibited similar arteriovenous differences in fasting glucose levels but significantly lower arteriovenous differences in

postprandial glucose levels compared to normal rats (normal: 10 ± 2 mg/dL, diabetic: 3 ± 2 mg/dL; p < 0.05) (Fig. 5A, Table 1). Diabetic rats also had a significantly lower difference in postprandial glucose levels between portal and abdominal veins compared to normal rats (normal: 20 ± 3 mg/dL, diabetic: 8 ± 4 mg/dL; p < 0.05) (Fig. 5B), while the renal arteriovenous difference in postprandial glucose levels was comparable (normal: 1 ± 2 mg/dL, diabetic: -4 ± 9 mg/dL) (Fig. 5C). In diabetic rats, ipragliflozin, glibenclamide, sitagliptin, and insulin significantly increased the arteriovenous dif- ference in postprandial glucose levels; metformin showed a trend to- wards significantly (P < 0.1) increasing the arteriovenous difference in postprandial glucose levels, while voglibose had no effect (Fig. 6A). Similarly, ipragliflozin, glibenclamide, sitagliptin, insulin, as well as metformin significantly increased the difference in postprandial glucose levels between the portal and abdominal veins (Fig. 6B). Ipragliflozin also induced a significant and marked increase in the renal arter- iovenous difference in postprandial glucose levels, while the other drugs had no effect (Fig. 6C). Under fasting conditions, diabetic rats

Fig. 2. Effects of ipragliflozin on (A) blood glucose and (B) plasma insulin levels during the liquid meal tolerance test in type 2 diabetic rats. Ipragliflozin was orally administered to overnight fasted rats, liquid meal (ENSURE-H, 20 mL/kg) was orally loaded 0.5 h post-dose, and blood glucose and plasma insulin levels were measured. Values indicate mean ± S.E.M. for six animals per group. *P < 0.05 vs. normal group, #P < 0.05 vs. vehicle group. N: normal, V: diabetic vehicle.

Fig. 3. Effects of two-week repeated adminis- tration of ipragliflozin and antidiabetic drugs on levels of (A) HbA1c and (B) plasma insulin, (C) food intake and (D) body weight in type 2 diabetic rats. Values indicate mean ± S.E.M. for six animals per group. N: normal, V: dia- betic vehicle, Ipra: ipragliflozin, Met: met- formin, Gli: glibenclamide, Sita: sitagliptin, Ins: insulin. *P < 0.05 vs. normal group, #P < 0.05 vs. vehicle group.

A. Tahara et al. Biomedicine & Pharmacotherapy 105 (2018) 1033–1041

1036

exhibited significantly elevated plasma levels of ghrelin compared to normal rats (Fig. 7). Under postprandial conditions, diabetic rats ex- hibited significantly elevated plasma levels of appetite-stimulating hormones (neuropeptide Y and ghrelin), while appetite-suppressing hormones (leptin and peptide YY) were unaffected compared to normal rats. These results resembled the non-fasting plasma levels in the re- peated administration study described above. All drugs, except vogli- bose, significantly reduced postprandial plasma levels of appetite-sti- mulating hormones, but had no significant affect on the plasma levels of appetite-suppressing hormones.

4. Discussion

Type 2 diabetes is caused by a combination of genetic factors related

to impaired insulin secretion and insulin resistance, and environmental factors such as obesity, overeating, lack of exercise, stress, drinking, smoking, and aging [12]. In particular, obesity is strongly associated with the pathogenesis of not only type 2 diabetes, but also dyslipidemia and hypertension. Concurrent occurrence of these conditions is en- compassed by the term “metabolic syndrome”, which greatly increases the risk of atherosclerotic disorders, such as cardiovascular disease and stroke, leading to an increased mortality rate [13]. The main causes of obesity are overeating, a poor diet containing high levels of fat and sugar, physical inactivity, and genetics. Therefore, appropriate dietary interventions are established methods for improving obesity [14]. Many studies have shown that SGLT2 inhibitors improve not only hypergly- cemia but also dyslipidemia and the symptoms of obesity [10,15]. However, concerns have been raised regarding a potential concomitant

Fig. 4. Effects of two-week repeated adminis- tration of ipragliflozin and antidiabetic drugs on plasma levels of (A) neuropeptide Y, (B) ghrelin, (C) leptin, and (D) peptide YY in type 2 diabetic rats. Values indicate mean ± S.E.M. for six animals per group. N: normal, V: dia- betic vehicle, Ipra: ipragliflozin, Met: met- formin, Gli: glibenclamide, Sita: sitagliptin, Ins: insulin. *P < 0.05 vs. normal group, #P < 0.05 vs. vehicle group.

Fig. 5. Plasma glucose differences between the (A) thoracic artery and abdominal vein, (B) portal vein and abdominal vein, and (C) renal artery and renal vein in type 2 diabetic rats. Values are mean ± S.E.M. for six animals per group. N: normal, D: diabetes. *P < 0.05 vs. normal group.

Table 1 Effects of antidiabetic drugs on plasma glucose levels in various vascular compartments in type 2 diabetic rats.

Fasting condition

Rats Treatment Portal vein

Thoracic artery

Abdominal vein

Renal artery

Renal vein

Fasting Normal Vehicle 79.7 ± 2.2 87.2 ± 2.9 84.1 ± 3.3 84.8 ± 4.0 84.7 ± 3.8 Diabetes Vehicle 99.7 ± 2.7* 112.5 ± 3.2* 108.9 ± 3.3* 108.5 ± 3.7* 106.3 ± 3.5*

Nonfasting Normal Vehicle 150.9 ± 5.5 140.8 ± 4.4 130.9 ± 4.3 135.2 ± 4.4 134.0 ± 3.5 Diabetes Vehicle 248.1 ± 9.6* 243.0 ± 8.0* 239.7 ± 7.8* 240.4 ± 8.9* 244.2 ± 11.4*

Ipragliflozin 203.1 ± 6.4# 195.9 ± 5.2# 169.1 ± 5.7# 202.8 ± 6.3# 160.4 ± 4.5#

Metformin 177.9 ± 5.8# 163.5 ± 5.0# 154.4 ± 5.1# 161.0 ± 9.1# 162.5 ± 7.2#

Voglibose 169.6 ± 6.3# 162.9 ± 5.1# 160.5 ± 4.7# 162.9 ± 7.1# 163.4 ± 6.2#

Glibenclamide 203.2 ± 6.4# 187.3 ± 4.8# 170.1 ± 5.4# 180.5 ± 7.8# 180.5 ± 6.3#

Sitagliptin 202.2 ± 4.8# 194.8 ± 4.9# 172.8 ± 4.5# 182.6 ± 6.3# 183.8 ± 7.3#

Insulin 197.2 ± 7.9# 179.4 ± 5.5# 162.0 ± 6.3# 167.9 ± 5.7# 167.2 ± 6.8#

Values indicate mean ± S.E.M. for six animals per group. *P < 0.05 vs. normal group, #P < 0.05 vs. vehicle group.

A. Tahara et al. Biomedicine & Pharmacotherapy 105 (2018) 1033–1041

1037

increase in compensatory food intake because SGLT2 inhibitors induce these beneficial outcomes by promoting urinary glucose excretion. To date, several preclinical and clinical studies have shown small to moderate increases in compensatory food intake resulting from SGLT2 inhibitors. However, these increases were not sufficient to cancel the improvement in obesity [9,16]. In addition, the effects of SGLT2 in- hibitors on various appetite-regulating mechanisms have not been studied in detail. We compared the effects of the SGLT2 inhibitor, ipragliflozin, and several other antidiabetic drugs on food intake. We also measured plasma levels of appetite-related hormones and arter- iovenous differences in postprandial glucose levels, which are strongly associated with the control of food intake, to examine the pharmaco- logical effects of ipragliflozin.

Type 2 diabetic rats show glucose intolerance associated with im- paired early insulin secretion, hyperglycemia, and glycosuria accom- panied by polyuria, but do not exhibit symptoms of obesity or insulin resistance [17]. Administration of the antidiabetic drugs, except the anti-insulin resistance drugs, peroxisome proliferator activated receptor (PPAR)-γ agonists, improved the pathological condition in these rats. Use of this type 2 diabetic rat model therefore allowed us to evaluate the effects of various antidiabetic drugs, in addition to the SGLT2 in- hibitor. Single administration of ipragliflozin increased urinary glucose excretion in a dose-dependent manner. Urine volume was also in- creased, while urinary excretion of electrolytes was unaffected,

indicating that ipragliflozin caused osmotic (isotonic) diuresis asso- ciated with glucose excretion. These results are consistent with findings in normal mice in our previous study [18]. Further, during the liquid meal tolerance test, single administration of ipragliflozin improved glucose tolerance in a dose-dependent manner. Compared to normal rats, diabetic rats showed impaired early insulin secretion after liquid meal loading, and ipragliflozin administration led to a further sig- nificant reduction in insulin levels. Since previous studies have shown that ipragliflozin does not directly affect insulin secretion [18], the decrease in insulin levels after liquid meal loading is likely to be a secondary effect of postprandial hyperglycemia improvement.

Repeated administration of ipragliflozin and antidiabetic drugs (metformin, voglibose, glibenclamide, sitagliptin, and insulin) sig- nificantly reduced HbA1c levels in diabetic rats. Administration of in- sulin and insulin secretagogues (glibenclamide and sitagliptin) sig- nificantly elevated plasma insulin levels, which is the likely cause of the antihyperglycemic effect of these drugs. In contrast, ipragliflozin, metformin, and voglibose showed a trend towards reducing plasma insulin levels, which is likely to be secondary to the antihyperglycemic effects of these drugs, and likely occur via different mechanisms, namely increasing urinary glucose excretion, inhibiting gluconeogen- esis in the liver, and inhibiting disaccharide hydrolysis in the small intestine, respectively [19]. Diabetic rats exhibited hyperphagia as a result of impaired insulin secretion and other diabetic conditions. Ipragliflozin, glibenclamide, sitagliptin, and insulin had no effect, while metformin and voglibose significantly reduced food intake. Appetite suppression is a physiological effect of insulin in the central nervous system [20]. However, insulin and insulin secretagogues did not sup- press food intake in this study, suggesting that moderate increases in plasma insulin levels caused by these drugs do not affect appetite. Our results are consistent with several reports indicating that sulfonylurea, dipeptidyl peptidase (DPP) 4 inhibitor, and insulin do not have obvious effects on food intake [21–23]. In contrast, metformin is well-docu- mented to suppress food intake, which is considered one of its benefits in clinical use [24,25]. Although the detailed mechanisms underlying metformin-induced suppression of food intake remain unclear, reduced expression of the appetite-stimulating hormone, neuropeptide Y, in the central nervous system might play a role [26]. Consistent with our re- sults, voglibose reportedly reduces food intake due to side effects, such as abdominal distention, associated with the accumulation of dis- accharides in the gastrointestinal tract [27,28]. In contrast, we found that ipragliflozin had no significant effect on food intake in type 2 diabetic rats. While some studies have shown that SGLT2 inhibitors have no significant effect on food intake [10,29], many others have reported that SGLT2 inhibitors induce small to moderate compensatory increases in food intake associated with calorie loss due to increased urinary glucose excretion [30,31]. The discrepancy among these studies may be attributed to differences in the pathology of animal models, experimental conditions, and/or the efficacy of individual SGLT2 in- hibitors. Importantly, the reported compensatory increases in food in- take were not sufficient to cancel the improvement in obesity resulting from SGLT2 inhibitors, although many issues including the pharmaco- logical mechanisms remain unclear. In addition, food intake is regu- lated by both hunger and satiety mechanisms; however, we only ex- amined some of these feeding mechanisms such as levels of appetite- regulating hormones and arteriovenous differences in postprandial glucose. Therefore, additional and detailed examinations including clinical studies are required to further investigate the underlying me- chanisms.

Compared to normal rats, diabetic rats showed a significantly lower body weight gain, presumably resulting from reduced glucose uptake in the liver, muscle, and adipose tissues due to impaired insulin secretion. The drugs that had no influence on food intake (insulin, glibenclamide, and sitagliptin) likewise had no effect on body weight gain. In contrast, those that reduced food intake (metformin and voglibose) caused sig- nificant weight loss. Weight loss by metformin was likely due to

Fig. 6. Effects of ipragliflozin and antidiabetic drugs on postprandial glucose level differences between the (A) thoracic artery and abdominal vein, (B) portal vein and abdominal vein, and (C) renal artery and renal vein in type 2 diabetic rats. Values indicate mean ± S.E.M. for six animals per group. N: normal, V: diabetic vehicle, Ipra: ipragliflozin, Met: metformin, Gli: glibenclamide, Sita: sitagliptin, Ins: insulin. *P < 0.05 vs. normal group, #P < 0.05 vs. vehicle group.

A. Tahara et al. Biomedicine & Pharmacotherapy 105 (2018) 1033–1041

1038

reduced food intake, with some weight loss having been reported in type 2 diabetic patients [32]. The voglibose-induced weight loss likely resulted from not only the reduction in food intake but also inhibition of the digestion and absorption of dietary carbohydrates and im- provement in the intestinal microflora environment [27,28]. Ipragli- flozin induced significant weight loss regardless of the amount of food intake, which likely resulted from calorie loss and diuresis due to in- creased urinary glucose excretion. Administration of ipragliflozin (1 mg/kg) resulted in urinary excretion of 1380 mg of glucose per day, which is equivalent to a loss of 4.6 kcal per day or 65 kcal in two weeks. This calorie loss is equivalent of 9 g of fat, which roughly corresponds to the observed weight loss (−11 g) in this study, considering the con- comitant diuresis. A recent study reported a discrepancy between the hypoglycemic and antiobesity effects of SGLT2 inhibitors [33]. The ipragliflozin-induced decrease in body weight noted in the present study may however be explained by the hypoglycemic effects due to

calorie loss induced by urinary glucose excretion. Although we did not measure lipid parameters or fat weights in this study, ipragliflozin has been shown to reduce body weight and fat mass through calorie loss and increasing fatty acid oxidation due to increased urinary glucose excretion [8]. In addition, ipragliflozin reportedly not only improves hyperglycemia but also reduces body and fat weights and lipid para- meters in type 2 diabetic patients [34]. These findings support our hypothesis that the ipragliflozin-induced weight loss mainly results from calorie loss via the increase in urinary glucose excretion.

We also measured the plasma levels of appetite-stimulating hor- mones, neuropeptide Y and ghrelin, and the appetite-suppressing hor- mones, leptin and peptide YY. Non-fasting plasma levels of these ap- petite-stimulating hormones were significantly elevated in diabetic rats compared to normal rats, while appetite-suppressing hormones were unaffected. Changes in hormone levels are likely correlated with hy- perphagia under diabetic conditions, as shown in a previous report

Fig. 7. Effects of ipragliflozin and antidiabetic drugs on plasma levels of appetite regulating hormones (A) neuropeptide Y, (B) ghrelin, (C) leptin, and (D) peptide YY in type 2 diabetic rats. Values indicate mean ± S.E.M. for six animals per group. N: normal, V: diabetic vehicle, Ipra: ipragliflozin, Met: metformin, Gli: glib- enclamide, Sita: sitagliptin, Ins: insulin. *P < 0.05 vs. normal group, #P < 0.05 vs. vehicle group.

A. Tahara et al. Biomedicine & Pharmacotherapy 105 (2018) 1033–1041

1039

[35]. While none of the drugs tested caused significant changes in plasma levels of appetite-suppressing hormones, insulin and insulin secretagogues showed a trend towards increasing levels. In contrast, all drugs, particularly metformin and insulin, except voglibose, showed a trend towards reducing plasma levels of appetite-stimulating hormones. The reduction in appetite-simulating hormones by metformin could be due, in part, to the above-mentioned pharmacological effect on food intake [32]. Similarly, insulin suppresses neuropeptide Y gene expres- sion in the hypothalamus [36]; such central actions of insulin are likely to be involved in this effect.

In diabetic rats, the plasma levels of appetite-suppressing hormones were not significantly elevated, which correlates with the absence of obesity and insulin resistance in this model. Indeed, SGLT2 inhibitors, including ipragliflozin, significantly reduce plasma leptin levels in type 2 diabetic mice with obesity, insulin resistance, and leptinemia [37]. Therefore, the effect of SGLT2 inhibitors and other antidiabetic drugs on appetite-suppressing hormones requires further evaluation in an- other type 2 diabetic animal model that exhibits obesity and insulin resistance. It is unlikely that plasma electrolyte imbalance directly af- fects food intake. However, changes in body fluid composition due to a chronic electrolyte imbalance could influence blood pressure and the regulation of appetite-related hormones, causing secondary changes in food intake [38]. While ipragliflozin has a diuretic effect associated with increasing urinary glucose excretion, it did not affect urinary ex- cretion of electrolytes. Further, ipragliflozin did not affect the renal arteriovenous difference in electrolyte concentrations (data not shown). These results suggest that ipragliflozin does not affect renal excretion or reabsorption of electrolytes and is unlikely to cause electrolyte im- balance, indicating that its effects on food intake do not occur through these mechanisms.

Diabetic rats had similar arteriovenous differences in fasting glucose levels, while the postprandial difference was significantly smaller, compared to normal rats, which is consistent with a previous report [2]. Further, diabetic rats showed similar renal arteriovenous differences in postprandial glucose levels but significantly smaller differences in postprandial glucose levels between the portal and abdominal veins compared to normal rats. Administration of the antidiabetic drugs, except for metformin and voglibose, significantly increased the arter- iovenous difference in postprandial glucose levels. Further, all anti- diabetic drugs including metformin, except voglibose, significantly in- creased the postprandial glucose level difference between the portal and abdominal veins. Ipragliflozin significantly and markedly increased the renal arteriovenous difference in postprandial glucose levels. The insulin-, glibenclamide-, and sitagliptin-induced increase in arter- iovenous difference in postprandial glucose levels was likely due to the stimulation of peripheral glucose uptake by insulin. While metformin only slightly increased the arteriovenous difference in postprandial glucose levels, it markedly increased the difference between portal and abdominal veins. These effects are thought to be mediated by the in- hibition of gluconeogenesis in the liver and enhanced glucose uptake in muscle and adipose tissues. Ipragliflozin appears to have similar effects on the arteriovenous and renal arteriovenous differences in post- prandial glucose levels, the latter being the result of inhibition of glu- cose reabsorption in the kidneys. In contrast, voglibose had no effect on arteriovenous differences in postprandial glucose levels, while glucose levels in the portal vein were markedly reduced. This effect is likely attributed to the fact that the only pharmacological action of voglibose is to inhibit disaccharide hydrolysis in the small intestine.

Administration of all antidiabetic drugs, except voglibose, reduced the elevated plasma levels of appetite-stimulating hormones in diabetic rats. This effect was strongly correlated with the increased arter- iovenous difference in postprandial glucose levels caused by each drug, suggesting that an interplay between these factors is required for postprandial regulation of appetite. Our data suggest that antidiabetic drugs, including ipragliflozin, directly or secondarily improved the regulation of food intake by affecting postprandial arteriovenous

differences in postprandial glucose levels and regulating appetite-re- lated hormones, which is presumably one of the reasons why SGLT2 inhibitors do not cause a compensatory increase in food intake. In contrast, voglibose did not affect the arteriovenous difference in post- prandial glucose levels or the regulation of appetite-related hormones, suggesting that these mechanisms do not underlie its role in regulating food intake. That voglibose did not induce a compensatory increase in food intake is likely due to potent abdominal distension associated with accumulation of disaccharides in the intestine. Previous reports have shown that voglibose reduces food intake and body weight in high-fat diet-fed obese mice [39] and that it affects negative regulators of food intake, such as leptin, suggesting that other mechanisms underlie its regulation of food intake. This is the first study to compare the effects of a SGLT2 inhibitor and various antidiabetic drugs on food intake and their association with appetite-related hormones and postprandial ar- teriovenous differences in glucose levels in type 2 diabetic rats. To in- vestigate the more direct pharmacological effect of a SGLT2 inhibitor and various antidiabetic drugs on feeding, we evaluated the effects after a single and two-week repeated administration. Further studies are needed to evaluate the effects after long-term administration, the de- tailed parameters of food intake/appetite regulation, and the effects of coadministration of various antidiabetic drugs using various type 2 diabetic models. In addition, this study was conducted in rodents and the results may be distinct from those in humans. However, it is pos- sible that some of the pharmacological mechanisms, including the regulation of appetite-related hormones, can be extrapolated to clinical settings and may be valuable for further studies including clinical in- vestigations.

5. Conclusion

Our data provide evidence that the SGLT2 inhibitor, ipragliflozin, has antihyperglycemic effects associated with increased urinary glucose excretion. Ipragliflozin induces weight loss without a compensatory increase in food intake through increasing the postprandial arter- iovenous difference in glucose levels and improving the regulation of appetite-related hormones in type 2 diabetic mice.

Conflict of interest

The authors have no conflict of interest other than being employees of Astellas Pharma Inc.

Acknowledgments

The authors thank Drs. Eiji Kurosaki, Masakazu Imamura, Masanori Yokono, Yuichi Tomura, Hideaki Minoura, Yuka Hayashizaki, Shoji Takakura, Seiji Kaku, and Wataru Uchida (Astellas Pharma Inc.) and Drs. Akira Tomiyama and Yoshihiko Haino (Kotobuki Pharmaceutical Co., Ltd.) for their valuable comments and continuing encouragement.

References

[1] M.W. Schwartz, S.C. Woods, D. Porte, Jr, R.J. Seeley, D.G. Baskin, Central nervous system control of food intake, Nature 404 (2000) 661–671.

[2] J. Mayer, Regulation of energy intake and the body weight: the glucostatic theory and the lipostatic hypothesis, Ann. NY Amad. Sci. 63 (1955) 15–43.

[3] S.P. Kalra, M.G. Dube, S. Pu, B. Xu, T.L. Horvath, P.S. Kalra, Interacting appetite- regulating pathways in the hypothalamic regulation of body weight, Endocr. Rev. 20 (1999) 68–100.

[4] Y. Zhang, R. Proenca, M. Maffei, M. Barone, L. Leopold, J.M. Friedman, Positional cloning of the mouse obese gene and its human homologue, Nature 372 (1994) 425–432.

[5] Z. Al-Hamodi, M. Al-Habori, A. Al-Meeri, R. Saif-Ali, Association of adipokines, leptin/adiponectin ratio and c-reactive protein with obesity and type 2 diabetes mellitus, Diabetol. Metab. Syndr. 6 (2014) 99.

[6] E.C. Chao, SGLT-2 inhibitors: a new mechanism for glycemic control, Clin. Diabet. 32 (2014) 4–11.

[7] L. Zhang, Y. Feng, J. List, S. Kasichayanula, M. Pfister, Dapagliflozin treatment in

A. Tahara et al. Biomedicine & Pharmacotherapy 105 (2018) 1033–1041

1040

patients with different stages of type 2 diabetes mellitus: effects on glycaemic control and body weight, Diabetes Obes. Metab. 12 (2010) 510–516.

[8] M. Yokono, T. Takasu, Y. Hayashizaki, K. Mitsuoka, R. Kihara, Y. Muramatsu, S. Miyoshi, A. Tahara, E. Kurosaki, Q. Li, H. Tomiyama, M. Sasamata, M. Shibasaki, Y. Uchiyama, SGLT2 selective inhibitor ipragliflozin reduces body fat mass by in- creasing fatty acid oxidation in high-fat diet-induced obese rats, Eur. J. Pharmacol. 727 (2014) 66–74.

[9] A. Napolitano, S. Miller, P.R. Murgatroyd, E. Hussey, R.L. Dobbins, E.T. Bullmore, D.J.R. Nunez, Exploring glycosuria as a mechanism for weight and fat mass re- duction. A pilot study with remogliflozin etabonate and sergliflozin etabonate in healthy obese subjects, J. Clin. Trans. Endocrinol. 1 (2014) 3–8.

[10] A. Tahara, E. Kurosaki, M. Yokono, D. Yamajuku, R. Kihara, Y. Hayashizaki, T. Takasu, M. Imamura, Q. Li, H. Tomiyama, Y. Kobayashi, A. Noda, M. Sasamata, M. Shibasaki, Effects of SGLT2 selective inhibitor ipragliflozin on hyperglycemia, hyperlipidemia, hepatic steatosis, oxidative stress, inflammation, and obesity in type 2 diabetic mice, Eur. J. Pharmacol. 715 (2013) 246–255.

[11] A. Tahara, A. Matsuyama-Yokono, R. Nakano, Y. Someya, M. Shibasaki, Hypoglycaemic effects of antidiabetic drugs in streptozotocin-nicotinamide-induced mildly diabetic and streptozotocin-induced severely diabetic rats, Basic. Clin. Pharmacol. Toxicol. 103 (2008) 560–568.

[12] K. Kaku, Pathophysiology of type 2 diabetes and its treatment policy, JMAJ 53 (2010) 41–46.

[13] S.M. Grundy, Obesity, metabolic syndrome, and coronary atherosclerosis, Circulation 105 (2002) 2696–2698.

[14] K. Yamanouchi, T. Shinozaki, K. Chikada, T. Nishikawa, K. Ito, S. Shimizu, N. Ozawa, Y. Suzuki, H. Maeno, K. Kato, Y. Oshida, Y. Sato, Daily walking combined with diet therapy is a useful means for obese NIDDM patients not only to reduce body weight but also to improve insulin sensitivity, Diabetes Care 18 (1995) 775–778.

[15] S.P. Vickers, S.C. Cheetham, K.R. Headland, K. Dickinson, R. Grempler, E. Mayoux, M. Mark, T. Klein, Combination of the sodium-glucose cotransporter-2 inhibitor empagliflozin with orlistat or sibutramine further improves the body-weight re- duction and glucose homeostasis of obese rats fed a cafeteria diet, Diabetes Metab. Syndr. Obes. 7 (2014) 265–275.

[16] Y. Watanabe, K. Nakayama, N. Taniuchi, Y. Horai, C. Kuriyama, K. Ueta, K. Arakawa, T. Senbonmatsu, M. Shiotani, Beneficial effects of canagliflozin in combination with pioglitazone on insulin sensitivity in rodent models of obese type 2 diabetes, PLoS One 10 (2015) e0116851.

[17] A. Tahara, A. Matsuyama-Yokono, R. Nakano, Y. Someya, M. Shibasaki, Hypoglycaemic effects of antidiabetic drugs in streptozotocin-nicotinamide-induced mildly diabetic and streptozotocin-induced severely diabetic rats, Basic. Clin. Pharmacol. Toxicol. 103 (2008) 560–568.

[18] A. Tahara, E. Kurosaki, M. Yokono, D. Yamajuku, R. Kihara, Y. Hayashizaki, T. Takasu, M. Imamura, L. Qun, H. Tomiyama, Y. Kobayashi, A. Noda, M. Sasamata, M. Shibasaki, Pharmacological profile of ipragliflozin (ASP1941), a novel selective SGLT2 inhibitor, in vitro and in vivo, Naunyn-schmiedeberg’s Arch. Pharmacol. 385 (2012) 423–436.

[19] A. Tahara, T. Takasu, M. Yokono, M. Imamura, E. Kurosaki, Effects of the combi- nation of SGLT2 selective inhibitor ipragliflozin and various antidiabetic drugs in type 2 diabetic mice, Arch. Pharm. Res. 39 (2016) 259–270.

[20] K. Gerozissis, Brain insulin and feeding: a bi-directional communication, Eur. J. Pharmacol. 490 (2004) 59–70.

[21] C.T. Musabayane, P.T. Bwititi, J.A. Ojewole, Effects of oral administration of some herbal extracts on food consumption and blood glucose levels in normal and streptozotocin-treated diabetic rats, Methods Find Exp. Clin. Pharmacol. 28 (2006) 223–228.

[22] S.J. Kim, C. Nian, D.J. Doudet, C.H. McIntosh, Inhibition of dipeptidyl peptidase IV with sitagliptin (MK0431) prolongs islet graft survival in streptozotocin-induced diabetic mice, Diabetes 57 (2008) 1331–1339.

[23] J.M. Rojas, R.L. Printz, K.D. Niswender, Insulin detemir attenuates food intake, body weight gain and fat mass gain in diet-induced obese sprague-Dawley rats, Nutr. Diabetes 1 (2011) e10.

[24] W.S. Lv, J.P. Wen, L. Li, R.X. Sun, J. Wang, Y.X. Xian, C.X. Cao, Y.L. Wang, Y.Y. Gao, The effect of metformin on food intake and its potential role in hypothalamic regulation in obese diabetic rats, Brain Res. 1444 (2012) 11–19.

[25] G. Paolisso, L. Amato, R. Eccellente, A. Gambardella, M.R. Tagliamonte, G. Varricchio, C. Carella, D. Giugliano, F. D’Onofrio, Effect of metformin on food intake in obese subjects, Eur. J. Clin. Invest. 28 (1998) 416–441.

[26] Y. Duan, R. Zhang, M. Zhang, L. Sun, S. Dong, G. Wang, J. Zhang, Z. Zhao, Metformin inhibits food intake and neuropeptide y gene expression in the hy- pothalamus, Neural Regen. Res. 8 (2013) 2379–2388.

[27] A.S. Dabhi, N.R. Bhatt, M.J. Shah, Voglibose: an alpha glucosidase inhibitor, J. Clin. Diagn. Res. 7 (2013) 3023–3027.

[28] H.J. Do, Y.S. Lee, M.J. Ha, Y. Cho, H. Yi, Y.J. Hwang, G.S. Hwang, M.J. Shin, Beneficial effects of voglibose administration on body weight and lipid metabolism via gastrointestinal bile acid modification, Endocr. J. 63 (2016) 691–702.

[29] H.M. Salim, D. Fukuda, S. Yagi, T. Soeki, M. Shimabukuro, M. Sata, Glycemic control with ipragliflozin, a novel selective SGLT2 inhibitor, ameliorated en- dothelial dysfunction in streptozotocin-induced diabetic mouse, Front. Cardiovasc. Med. 26 (2016) 43.

[30] N. Terami, D. Ogawa, H. Tachibana, T. Hatanaka, J. Wada, A. Nakatsuka, J. Eguchi, C.S. Horiguchi, N. Nishii, H. Yamada, K. Takei, H. Makino, Long-term treatment with the sodium glucose cotransporter 2 inhibitor, dapagliflozin, ameliorates glu- cose homeostasis and diabetic nephropathy in db/db mice, PLoS One 9 (2014) e100777.

[31] T. Nagata, T. Fukuzawa, M. Takeda, M. Fukazawa, T. Mori, T. Nihei, K. Honda, Y. Suzuki, Y. Kawabe, Tofogliflozin, a novel sodium-glucose co-transporter 2 in- hibitor, improves renal and pancreatic function in db/db mice, Br. J. Pharmacol. 170 (2013) 519–531.

[32] A. Golay, Metformin and body weight, Int. J. Obes. 32 (2008) 61–72. [33] E. Kutoh, T. Murayama, A. Wada, M. Hirate, Distinct glucose-lowering mechanisms

of ipragliflozin depending on body weight changes, Drugs R. D. 16 (2016) 369–376. [34] T. Fukuda, R. Bouchi, M. Terashima, Y. Sasahara, M. Asakawa, T. Takeuchi,

Y. Nakano, M. Murakami, I. Minami, H. Izumiyama, K. Hashimoto, T. Yoshimoto, Y. Ogawa, Ipragliflozin reduces epicardial fat accumulation in non-obese type 2 diabetic patients with visceral obesity: a pilot study, Diabetes Ther. 8 (2017) 851–861.

[35] S. Ishii, J. Kamegai, H. Tamura, T. Shimizu, H. Sugihara, S. Oikawa, Role of ghrelin in streptozotocin-induced diabetic hyperphagia, Endocrinology 143 (2002) 4934–4937.

[36] I. Sato, H. Arima, N. Ozaki, M. Watanabe, M. Goto, M. Hayashi, R. Banno, H. Nagasaki, Y. Oiso, Insulin inhibits neuropeptide Y gene expression in the arcuate nucleus through GABAergic systems, J. Neurosci. 25 (2005) 8657–8664.

[37] A. Tahara, T. Takasu, M. Yokono, M. Imamura, E. Kurosaki, Characterization and comparison of sodium-glucose cotransporter 2 inhibitors: part 2. Antidiabetic ef- fects in type 2 diabetic mice, J. Pharmacol. Sci. 131 (2016) 198–208.

[38] I. Odar-Cederlöf, F. Ericsson, E. Theodorsson, C.M. Kjellstrand, Is Neuropeptide Y a contributor to volume-induced hypertension? Am, J. Kidney Dis. 31 (1998) 803–808.

[39] H.J. Do, T. Jin, J.H. Chung, J.W. Hwang, M.J. Shin, Voglibose administration regulates body weight and energy intake in high fat-induced obese mice, Biochem. Biophys. Res. Commun. 443 (2014) 1110–1117.

A. Tahara et al. Biomedicine & Pharmacotherapy 105 (2018) 1033–1041

1041