Mizagliflozin

Absorption, disposition, metabolism, and excretion of [14C]mizagliflozin, a novel selective
SGLT1 inhibitor, in rats

Hitoshi Ono, Yasunari Kojima, Hiroshi Harada, Yoshikazu Abe, Takuro Endo &
Mamoru Kobayashi

To cite this article: Hitoshi Ono, Yasunari Kojima, Hiroshi Harada, Yoshikazu Abe, Takuro Endo &
Mamoru Kobayashi (2018): Absorption, disposition, metabolism, and excretion of [14C]mizagliflozin, a novel selective SGLT1 inhibitor, in rats, Xenobiotica, DOI: 10.1080/00498254.2018.1449269
To link to this article: https://doi.org/10.1080/00498254.2018.1449269

Accepted author version posted online: 20 Mar 2018.

Submit your article to this journal

View related articles

View Crossmark data

Full Terms & Conditions of access and use can be found at
http://www.tandfonline.com/action/journalInformation?journalCode=ixen20

Absorption, disposition, metabolism, and excretion of [14C]mizagliflozin, a novel selective SGLT1 inhibitor, in rats
Absorption, disposition, metabolism, and excretion of [14C]mizagliflozin, a novel selective SGLT1 inhibitor, in rats

Hitoshi Ono, Yasunari Kojima, Hiroshi Harada, Yoshikazu Abe, Takuro Endo, and Mamoru Kobayashi

Central Research Laboratories, Kissei Pharmaceutical Co., Ltd., 4365-1 Kashiwabara, Hotaka, Azumino, Nagano, 399-8304, Japan.

Address for correspondence: Hitoshi Ono
Central Research Laboratories, Kissei Pharmaceutical Co. Ltd., 4365-1 Kashiwabara, Hotaka, Azumino, Nagano, 399-8304, Japan.
Tel: +81-(0)263-82-8820
E-mail: [email protected]

Keywords:
SGLT1 selective inhibitor, pharmacokinetic, metabolite profile, mizagliflozin, rat

Running head:
Pharmacokinetics and metabolism of mizagliflozin

Abstract

1.The pharmacokinetic and metabolite profiles of mizagliflozin, a novel selective sodium glucose co-transporter 1 inhibitor designed to act only in the intestine, were investigated in rats.
2.Mizagliflozin administrated intravenously (0.3 mg/kg) and orally (3 mg/kg) declined with a short half-life (0.23 and 1.14 h, respectively). The absolute bioavailability was only 0.02%. Following intravenous administration of [14C]mizagliflozin (0.3 mg/kg), radioactivity in plasma was also rapidly declined. Up to 24 h after oral administration of [14C]mizagliflozin (1 mg/kg), radioactivity was recovered in the faeces (98.4%) and in the urine (0.8%). No remarkable accumulation of radioactivity in tissues was observed using tissue dissection technique and whole body autoradiography.
3.Orally dosed [14C]mizagliflozin was mostly metabolised to its aglycone, KP232, in the intestine. In the plasma, KP232 and its glucuronide were predominant. KP232 glucuronide was also prominent in the bile, and was recovered as KP232 in the faeces possibly because of the deconjugation by gut microflora. Mizagliflozin was observed neither in the urine nor the faeces.
4.These findings suggest that orally administered mizagliflozin is poorly absorbed, contributing to low systemic exposure; if absorbed, mizagliflozin is rapidly cleared from circulation.

Accepted

Introduction

Chronic constipation is a highly common gastrointestinal (GI) disorder. The main symptoms of chronic constipation include infrequent bowel movements, hard stools, straining during defecation, a feeling of incomplete evacuation, abdominal discomfort, and a sensation of bloating (McCallum et al., 2009). Chronic constipation adversely affects the quality of life of patients and increases their economic burden (Sun et al., 2011). Limited therapeutic options are available for treating chronic constipation at present. Commonly used treatment options include saline, stimulant, osmotic, and bulk laxatives (Fukudo et al., 2011; Longstreth et al., 2006). However, approximately 50% of patients with chronic constipation are not satisfied with their current treatment, mostly because of a lack of efficacy (Johanson and Kralstein, 2007). This highlights a
continued medical need for more effective and safer therapeutic agents. Sodium glucose co-transporter (SGLT) 1 and SGLT2 are the most well
characterised and strongly expressed SGLT subtypes. SGLT1 is mainly expressed in the small intestine and plays a critical role in intestinal glucose absorption (Wright et al., 2007; Gorboulev et al., 2012). On the other hand, SGLT2 is the major transporter responsible for the reabsorption of glucose filtered through the renal glomerulus (Kanai et al., 1994; Isaji, 2011). Recent studies have identified SGLTs as attractive therapeutic targets. SGLT2 inhibition is suggested to increase renal glucose excretion and to reduce plasma glucose levels; at present, many SGLT2 inhibitors are clinically used as anti- diabetic drugs (Scheen, 2015). Selective inhibition of SGLT1 decreases glucose uptake from the small intestine and alters postprandial blood glucose excursion (Shibazaki et al., 2012). However, no SGLT1-selective inhibitor is under development as an anti- diabetic drug at present.

Mizagliflozin (3-{[3-(4-{[3-(β-D-glucopyranosyloxy)-5-(propan-2-yl)-1H- pyrazol-4-yl]methyl}-3-methylphenoxy)propyl]amino}-2,2-dimethylpropanamide, Figure 1) is a selective SGLT1 inhibitor that has been developed by Kissei Pharmaceutical Co., Ltd. (Azumino, Japan) with selectivity ratios (IC50 for human SGLT2/IC50 for human SGLT1) of 51-fold. Mizagliflozin was originally used as an anti-diabetic drug that can modify postprandial blood glucose levels. However, the latest report suggested the potential use of the SGLT1-selective inhibitor mizagliflozin for treating chronic constipation (Inoue et al. 2017).
In this study, we investigated the absorption, distribution, metabolism, and excretion profile of 14C-radiolabeled mizagliflozin. The investigation of these profiles in rats helps us understand the pharmacokinetic character of mizagliflozin and also speculate about the pharmacokinetic properties of mizagliflozin in humans.

Accepted

Materials and methods

Chemicals

Mizagliflozin sebacate, mizagliflozin aglycone (KP232), and KP232 glucuronide were synthesised at Kissei Pharmaceutical Co., Ltd. (Azumino, Japan). [14C]Mizagliflozin sebacate (2.07 GBq/mmol, radiochemical purity > 98%) were synthesised at GE Healthcare (Chalfont St. Giles, Buckinghamshire, UK).

Animals

Prior to the study, the animal study protocols were reviewed by the Institutional Animal Care and Use Committee, and approved by the director of Pharmacokinetics Research Laboratory in consideration of the animal welfare.
Male Crl:CD Sprague-Dawley (SD) rats (age 7 weeks) were purchased from Charles River Japan (Yokohama, Japan). They were housed under constant temperature (20–26°C) and humidity (30%–70%) conditions with 12 h light and 12 h dark (lights on at 8:00), and fed a conventional laboratory diet and tap water ad libitum for 1 week of acclimatisation.

PK study

Mizagliflozin was administered orally or intravenously to fasted rats at a dose of 3 mg/kg as free-base (0.6 mg/mL solution, 5 mL/kg) or 0.3 mg/kg as free-base (0.3 mg/mL solution, 1 mL/kg), respectively. Blood samples were collected from the jugular vein at 2 (intravenous dosing), 5 (oral dosing), 15, 30, 60, 120, 180, 240, 360, and 1440 min after administration. Plasma samples were obtained after centrifugation (15000 × g, 1 min, 4°C) and stored in a freezer (-20°C) until the day of analysis. The plasma

concentrations of mizagliflozin, KP232, and KP232 glucuronide were determined using a validated method by liquid chromatography/tandem mass spectrometry (LC-MS/MS). These analytes in plasma samples were extracted using OASIS HLB µ Elution Plate (Waters, Milford, MA), and reconstituted with 0.1% formic acid solution/acetonitrile mixture (75:25, v/v). Stable isotope-labelled analytes were used as internal standards. The eluate was analysed with by LC-MS/MS. Separation of metabolites with a high performance liquid chromatography (HPLC) was achieved with an Aqua 5µ C18 column (150 mm × 2.0 mm I.D., 5 µm, Phenomenex, Torrance, CA) at a flow rate of
0.4 mL/min. The temperature of the column was set at 40°C. The mobile phase was a mixture of 10 mmol/L ammonium formate (A) and acetonitrile (B). The gradients were as follows: Solvent B started at 23%, was held for 3 min, and then linearly increased to 70% at 4 min. It was maintained at 70% for 3 min, then decreased to 23% at 7.1 min, and was held at 23% for 7 min. The HPLC eluent was introduced to a mass spectrometer, API3000 (AB/MDS Sciex, USA), with electrospray ionisation in positive ion mode. Mizagliflozin, KP232, and KP232 glucuronide were quantified by multiple reaction monitoring: m/z 565 > 403 for mizagliflozin, m/z 403 > 139 for KP232, and m/z 579 > 403 for KP232 glucuronide. Calibration curves ranged 0.2–200 ng/mL for each analyte. The concentrations were calculated using Analyst 1.4.1 (AB/MDS Sciex,
USA).

PK study with [14C]mizagliflozin was also performed in a similar manner to cold PK study with a minor modification in sampling time. Blood samples were collected from the jugular vein at 2 (intravenous dosing), 5 (oral dosing), 15, 30, 60, 120, 180, 240, 480, and 1440 min after administration. Plasma samples prepared from blood were mixed well with 3 mL of ULTIMA GOLD XR (PerkinElmer), and the radioactivity was

determined by liquid scintillation counting (LSC) with TRI CARB 3100TR (PerkinElmer).

Autoradiography

[14C]Mizagliflozin was orally administered to fasted rats at a dose of 3 mg/kg (297.2 μCi/kg, 5 mL/kg). At 0.5, 4, and 24 h after administration, animals were euthanised by carbon dioxide inhalation.
Whole animals were then immersed in a 4% carboxymethylcellulose sodium (CMC-Na) aqueous solution and then frozen in a dry ice/acetone. Then, 30-μm-thick frozen sections were sliced using a microtome. Frozen sections wereFrozen sections were lyophilised and exposed to an imaging plate (TYPE BAS SR2040, GE Healthcare, Buckinghamshire, UK) After 24 h-exposure, the radioactivity images were scanned using BAS2500 (GE Healthcare) and radioluminograms were prepared.

Distribution study (combustion)

At 0.5, 4, 24, and 72 h after oral administration of [14C]mizagliflozin to fasted rats at a dose of 3 mg/kg (297.2 μCi/kg, 5 mL/kg), rats were sacrificed by exsanguination from the abdominal aorta under anaesthesia.
Plasma was obtained by centrifugation of blood. Aliquots of blood and plasma (200 μL) were transferred onto a combusto-pad (PerkinElmer Life and Analytical Sciences, Waltham, MA) on a combusto-cone (PerkinElmer). The tissues and organs collected were as follows: cerebrum, cerebellum, pituitary, eyeball, thyroid gland, mandibular gland, thymus, heart, lung, liver, kidney, adrenal gland, spleen, pancreas, white fat, skeletal, muscle, skin, bone marrow, bone, mesenchymal lymph nodes, testis, prostate, stomach, small intestine, cecum, and large intestine. These samples (approximately 0.2 g, or whole tissue if less than 0.2 g) were collected on a combusto-

cone and dried at room temperature for several days. The radioactivity was determined by LSC with TRI CARB 3100TR after combustion.

In vitro plasma protein binding

[14C]Mizagliflozin sebacate was dissolved in water to prepare 1, 10, and 100 μg/mL. Aliquots of these solutions (10 μL) were added to rat and human plasma (1 mL) to prepare 10, 100, and 1000 ng/mL solutions. After incubation for 5 min at 37°C, aliquots of plasma (100 μL) were sampled to determine the total concentration of [14C]mizagliflozin in plasma (Ct). The remaining plasma (800 μL) was transferred to a Centrifree ultrafiltration device (Millipore, Billerica, MA) and centrifuged (1880 × g, 4°C, 20 min). Aliquots of the ultrafiltrate (100 L) were sampled to determine the unbound concentration of [14C]mizagliflozin in plasma (Cu). Samples were mixed well with 3 mL of ULTIMA GOLD XR and the radioactivity was determined by LSC. Plasma protein binding was calculated using the following equation: %Bound = (1 – Cu/Ct) × 100

Excretion study

Urine/Faeces excretion

After [14C]mizagliflozin was administered orally to fasted rats at a dose of 1 mg/kg (99.1 μCi/kg, 5 mL/kg), rats were individually housed in metabolic cages and the urine and faeces were collected 0–8, 8–24, 24–48, 48–72, 72–96, 96–120, 120–144, and 144– 168 h after administration. Faeces were diluted with distilled water to approximately 40 mL. All samples were weighed and stored at -30°C until analysis. After thawing, the urine was shaken well and the faeces were homogenised using a homogeniser. Approximately 0.2 g of samples were collected on a combusto-pad and dried at room

temperature. The radioactivity was determined by LSC after combustion.

Biliary excretion

[14C]Mizagliflozin was administered orally to fasted bile duct-cannulated (BDC) rats at a dose of 1 mg/kg (99.1 μCi/kg, 5 mL/kg). BDC rats were individually housed in Ballman cages. Bile was collected 0–2, 2–4, 4–6, 6–8, 8–24, and 24–48 h after administration. The urine and the faeces were collected 0–24 and 24–48 h after administration. At 48 h after administration, rats were euthanised to sample the GI tract (from stomach to rectum, including contents). All samples were stored at -30°C until analysis. After thawing, bile and urine were weighed and mixed well, and faeces were homogenised. The GI tracts were minced with scissors and homogenised. Approximately 0.02 g of the bile collected at 24 hours and 0.2 g of the other samples were collected on a combusto-pad and dried at room temperature. The radioactivity was determined by LSC after combustion.

Enterohepatic circulation

[14C]Mizagliflozin was administered orally to fasted BDC rats at a dose of 1 mg/kg (99.1 μCi/kg, 5 mL/kg). BDC rats were individually maintained in Ballman cages. Bile was collected up to 8 h after administration. Then, the bile was intraduodenally administered to other BDC rats (0.028 mg/kg, 5 mL/kg). These rats were individually maintained in Ballman cages as well. Bile was collected on ice 0–8 h and 8–24 h after administration, and the urine and faeces were collected 0–24 h after administration. After anaesthesia, GI tracts (from stomach to rectum including contents) were collected. The bile and urine (containing washings) were weighed. The faeces and GI tracts were diluted with distilled water and weighed. After thawing, the bile and urine were shaken well and the faeces were homogenised. The GI tracts were minced with scissors and

homogenised. Approximately 0.2 g of each sample was collected on a combusto-pad, and weighed and dried at room temperature. The radioactivity was determined by LSC after combustion.

Metabolite profiling study

Tissues

At 0.5 and 4 h after administration of [14C]mizagliflozin to fasted rats at a dose of 10 mg/kg (990.8 μCi/kg, 5 mL/kg), rats were sacrificed by exsanguination from the abdominal aorta under anaesthesia. Metabolite profiles were explored in the plasma, liver, and kidney. These tissues were excised and homogenised using the following method. Tissues from three animals were pooled by equal amounts and used for metabolite analysis. These samples were stored at -30°C until the day of analysis.
Plasma samples were diluted with an equal volume of 1% (v/v) acetic acid. Metabolites in plasma samples (1 mL) were extracted using OASIS HLB µ Elution Plate, and reconstituted with 90% (v/v) acetonitrile. The eluates were pooled and evaporated under nitrogen gas at room temperature.
The liver and kidney were minced with scissors, and then diluted with two volumes of ice-cold acetonitrile/1% acetic acid (3:1, v/v) and homogenised. A portion of the tissue homogenate was weighed. The homogenates were centrifuged at 1800 × g for 5 min at 4°C to obtain supernatants. The residues were resuspended with two volumes of ice-cold acetonitrile/1% acetic acid (3:1, v/v) and treated by the same procedure again. The collected supernatants were pooled and evaporated under nitrogen gas at room temperature.

The residues after evaporation were reconstituted in 10% (v/v) acetonitrile and analysed by HPLC coupled with a flow scintillation analyser (FSA, Radiomatic 525TR, PerkinElmer).

GI contents

At 0.25, 2, and 4 h after administration of [14C]mizagliflozin to fasted rats at a dose of 1 mg/kg (99.1 μCi/kg, 5 mL/kg), rats were sacrificed by exsanguination from the abdominal aorta under anaesthesia. After the GI tract was clamped, the stomach, small intestine (upper and lower), cecum, and large intestine were excised. After mincing with scissors in 3 mL of acetonitrile/1% acetic acid (3:1, v/v), each segment (including contents) was homogenised. Homogenates were centrifuged (1800 × g, 5 min, 4°C) to collect supernatants. The residue was resuspended in 3 mL of acetonitrile/1% acetic
acid (3:1, v/v), homogenised, and centrifuged again. The supernatants were mixed together. The supernatant and residue were weighed. SOLUENE-350 (0.5 mL, PerkinElmer) and Hionic-Fluor (5 mL, PerkinElmer) were added to approximately 0.05 g of the supernatant. The residue was dissolved in SOLUENE-350 (0.5 mL) by the incubation at 60°C for 1 h. Following the addition of Hionic-Fluor (5 mL), radioactivity was determined by LSC.
Supernatants (including tissue extracts) obtained from three animals were pooled according to weights of the GI samples. Aliquots of pooled supernatant (100 µ L) were evaporated under nitrogen gas at room temperature, and the residue was reconstituted in the mobile phase and analysed. Samples were mixed well with Hionic-Fluor (5 mL),
and the radioactivity was determined by LSC. The HPLC eluate was mixed with Ultima-Flo M (PerkinElmer) at a flow rate of 1 mL/min and the radioactivity was measured using a FSA. FLO-ONE for Windows Analysis Version 3.61 (PerkinElmer) was used for data analysis.

Excreta

After oral administration of [14C]mizagliflozin to fasted rats at a dose of 1 mg/kg (99.1 μCi/kg, 5 mL/kg), rats were individually placed in metabolic cages. The urine and faeces were collected on ice until 24 h after administration. The faeces were diluted with distilled water to approximately 45 mL. For bile samples, BDC rats were individually placed in Ballman cages after administration. Excreted bile was collected on ice until 8 h after administration. These samples were stored at -30°C until analysis.
The urine and bile samples were pooled according to weight. Faecal homogenates were pooled equally. Urine (1 g) and bile samples (0.05 g) were diluted with two volumes of acetonitrile. Faecal samples (0.05 g) were diluted with four volumes of 1% (v/v) acetic acid and eight volumes of acetonitrile, and sonicated for 10 min. These mixtures were centrifuged (15000 × g for 2 min at room temperature) to obtain supernatants. The residues were treated in the same manner again to obtain supernatants. Supernatants were mixed together and evaporated under nitrogen gas at room temperature. The residue was reconstituted in 10% (v/v) acetonitrile and analysed by HPLC with an FSA and TSQ7000 (Finnigan, San Jose, CA).

Radio-HPLC-MS analysis for metabolite profiling

Reconstituted extracts were analysed using a liquid chromatography system (HP1100, Agilent Technologies, Santa Clara, CA) coupled to an FSA. Separation of metabolites was achieved with Capcell Pak C18 MG (3 × 250 mm, 5 µm, Shiseido, Tokyo, Japan), and the mobile phase was a mixture of 10 mmol/L ammonium hydrogen carbonate (A) and acetonitrile (B). The mobile phase was passed over the column at 0.5 mL/min in the flowing linear gradient mode starting with 10% B composition and maintained for 10 min, increasing to 30% for 10 to 20 min, increasing to 34% for 20 to 40 min, increasing

to 60% for 40.1 min, maintained at 60% for 40.1 to 45 min, decreasing to 10% for 45.1 min, and finally maintained at 10% for 45.1 to 60 min. The column was maintained at 40°C. The HPLC eluate was mixed with Ultima-Flo M at a flow rate of 1 mL/min and the radioactivity was measured with an FSA by 6-sec integration. Peak integration of the UV and radioactivity were performed using FLO-ONE for Windows Analysis, version 3.6.5 (PerkinElmer). Metabolites were identified by the retention times, m/zs, and fragmentation patterns.

Determination of radioactivity in samples

The dried samples were combusted with an Oxidizer system 387 (PerkinElmer). The generated [14C]CO2 gas was trapped into 8 mL of CARBO-SORB E (PerkinElmer), and 10 mL of Permafluor E+ (PerkinElmer) were added. Liquid samples, such as urine and bile, were mixed well with Ultima-Gold or Hionic Fluor. The radioactivity of these samples was counted by LSC. The counting efficiency was corrected using external standards.

Pharmacokinetic Analysis

The pharmacokinetic analysis was performed using WinNonlin Professional version 5.0 (Pharsight, Mountain View, CA). The area under the plasma concentration-time curve (AUC), maximum concentration (Cmax), and time to reach Cmax (Tmax) were calculated according to the non-compartmental model. Total body clearance (CLtot) and volume of distribution at steady state (Vdss) were calculated from the data of the intravenous administration. In case a parameter could not be calculated, such as a lack
of plasma concentration data, the parameter was expressed as N.A. (not applicable). The absolute bioavailability (BA) value was calculated using the following equation: BA
(%) = (AUCp.o./Dosep.o.)/(AUCi.v/Dosei.v) × 100.

Results

Pharmacokinetics of mizagliflozin

Plasma concentrations of mizagliflozin and its two metabolites, mizagliflozin aglycone (KP232) and KP232 glucuronide, were determined and the pharmacokinetic parameters are summarised in Table 1. Plasma concentration-time profiles are shown in Figure 2. Following intravenous administration of mizagliflozin to fasted rats, mizagliflozin was rapidly eliminated with quite a short t1/2 of 0.23 h and the high CLtot was calculated at 1186 mL/h/kg, and the AUC0-24 and Vdss of mizagliflozin were 255.78 ngh/mL and 111.8 mL/kg, respectively. After oral administration of mizagliflozin (3 mg/kg) to fasted rats, the plasma concentrations of mizagliflozin reached Cmax at 0.23 h after administration and then rapidly declined with quite a short half-life of 1.14 h, andabsolute bioavailability of mizagliflozin calculated from AUC0-24 was only 0.02%. The AUC0-24 of KP232 and KP232 glucuronide were markedly higher than that of mizagliflozin by 17.7-fold and 156.6-fold, respectively. The plasma concentration-time profiles of KP232 and KP232 glucuronide had two peaks, and the second ones reached Cmax at 6.00 h and 4.63 h for KP232 and KP232 glucuronide, respectively.
Plasma concentration-time profiles and pharmacokinetic parameters after intravenous administration of [14C]mizagliflozin are shown in Figure 3 and Table2, respectively. The AUC0-24 of radioactivity was comparable with the total AUC0-24 of mizagliflozin, KP232, and KP232 glucuronide.

Distribution

The tissue distribution of radioactivity was investigated after oral administration of [14C]mizagliflozin to rats. The radioactivity concentrations were reached the maximum at 0.5 h after administration in the plasma, kidney, stomach, and small intestine, and 4 h

after administration in the other tissues (Table 3). Radioactivity concentrations in the liver, kidney, and mesenteric lymph nodes were relatively higher than that in plasma up to 4 h after administration but declined over time. In the GI tract, radioactivity concentrations were also higher than those in other tissues, although radioactivity in the GI tract was near the lower limit of detection 72 h after administration. Radioactivity was rapidly eliminated from the tissues and marginally detected only in liver, skin, and GI tract 72 h after administration.
To further explore the tissue distribution of mizagliflozin, whole-body autoradiography was performed after single oral administration of [14C]mizagliflozin. The representative luminograms are presented in Figure 4. At 0.5 h after administration, the radioactivity was detected in the small intestinal contents, gastric contents, urine in the bladder, and liver. At 4 h after administration, radioactivity was detected in the large and small intestinal contents, gastric contents, urine in the bladder, and liver. At 24 and 72 h after administration, radioactivity was detected only in the large intestinal contents. These results suggest that the absorption of mizagliflozin was quite low and the limited amount of radioactivity was distributed to the tissues.
The plasma protein binding of mizagliflozin was also investigated. The binding of [14C]mizagliflozin in rats and human plasma was nearly constant in the range of 10– 1000 ng/mL (46.2–47.8% and 29.1–33.6%, respectively) (Supplemental table 1). Additionally, the plasma proteins to which [14C]mizagliflozin bound were examined using purified human plasma proteins. The binding of [14C]mizagliflozin to purified human plasma proteins was as follows: 9.1–14.5% to albumin, 8.5–15.8% to γ-globulin, and 10.3–12.6% to α1-acid glycoprotein. No marked changes were observed in protein binding dependent on the concentration of [14C]mizagliflozin.

Excretion

The cumulative radioactivity in urine and faeces is summarised in Table 4 after oral administration of [14C]mizagliflozin to rats. Up to 24 h after administration, 98.4% of the oral dose was excreted in the faeces. The mean overall recovery of the total radioactivity was 100.7%, of which urinary excretion was only 0.9%. In BDC rats, the mean biliary excretion of radioactivity up to 48 h after administration was 31.9%. The urinary and faecal excretion accounted for 2.4% and 55.2% of the radioactivity, respectively. Enterohepatic circulation of mizagliflozin was investigated. The cumulative excretion of radioactivity up to 24 h after administration was 15.1% of dosed radioactivity in the bile, 6.4% in the urine, 65.9% in the faeces, and 12.4% in the GI tracts including the contents. This result suggests at least 21.5% of radioactivity excreted via the bile was reabsorbed.

Metabolism

The components of radioactivity in the GI tract, including contents, were investigated after oral administration of [14C]mizagliflozin because mizagliflozin was expected to exhibit efficacy in the gastrointestine. Movement of radioactivity and metabolite composition in the gastrointestine are shown in Figure 5. At 0.25 h after administration, the radioactivity was detected mainly in the upper small intestine (56.5%) followed by the lower small intestine (20.5%) and stomach (10.6%). The amount of KP232 was comparable to mizagliflozin in the small intestine. After 2 h, the majority of radioactivity moved to the lower small intestine where KP232 was the most abundant metabolite. Then, most radioactivity shifted to the cecum and large intestine, and almost the entire radioactivity was explained by KP232.

Radioactivity compositions in the plasma, liver, and kidney after oral administration of [14C]mizagliflozin were examined (Table 5). In the plasma at 0.5 h after administration, N-dealkylated KP232 (acid type, M2), KP232 glucuronide, and N- dealkylated KP232 glucuronide (amine type, M3) were detected and accounted for 23.1%, 22.7%, and 13.5% of radioactivity, respectively. Mizagliflozin and KP232 were also detected, but to a lesser extent. At 4 h after administration, the proportions of KP232 and KP232 glucuronide in tissue radioactivity increased to 25.8% and 57.7%, respectively, while M3 and mizagliflozin were undetectable. In the liver at 0.5 h after dosing, KP232 and KP232 glucuronide accounted for 37.9% and 25.7% of the radioactivity, respectively. At 4 h after administration, KP232 and KP232 glucuronide remained major components. The composition of M4 increased to 26.2%, and mizagliflozin decreased to below the detection limit. In the kidney at 0.5 h after administration, mizagliflozin and M1 were relatively abundant and covered 23.6% and 21.0% of the tissue radioactivity, respectively. The composition of KP232 and KP232 glucuronide increased gradually from 9.9% to 35.0% and from 10.1% to 24.8%, respectively.
After oral administration of [14C]mizagliflozin to rats, excreta were collected and the metabolite profile was explored. Metabolite compositions in excreta are summarised in Table 6. In the urine, KP232, KP232 glucuronide, and M1 were mainly detected, and mizagliflozin was not observed. Although KP232 glucuronide and, to a lesser extent, mizagliflozin and KP232 were also detected in bile, which accounted for 85.2%, 3.7%, and 7.6% of the biliary radioactivity, respectively. The only metabolite that appeared in faeces was KP232; no other metabolites were detected. No marked
differences were observed in the radioactivity compositions in the excreta between male and female rats (unpublished data).

Discussion

Mizagliflozin is a novel SGLT1 inhibitor that shows remarkable therapeutic efficacy for chronic constipation (Inoue et al. 2017). To date, several SGLT inhibitors have been developed, and most of them are reported to inhibit SGLT2. SGLT2 is one of the sodium glucose transporters and plays an important role in renal glucose reabsorption. These SGLT2 inhibitors have a C--glucoside moiety (Zhang and Liu, 2016) and are more resistant to the degradation to aglycones. Therefore, they can reach the kidney and inhibit renal SGLT2. On the other hand, SGLT1 functions as a glucose uptake transporter in the intestinal brush border membranes and also as a glucose reuptake transporter in the renal proximal tubule. SGLT1 inhibitors can exert an inhibitory effect on SGLT1 activity inside the GI tract without being absorbed into the body. Hence, a non-absorbable property is an advantage in safety by avoiding both systemic exposure and tissue accumulation which may cause adverse effects. We developed mizagliflozin with an O--glucoside moiety, not C--glucoside moiety, to be easily degraded to aglycone. This is the first report of pharmacokinetic data of mizagliflozin in a non- clinical stage. In the present study, the absorption, distribution, metabolism, and excretion properties of mizagliflozin were investigated in rats. The investigation of
these properties in rats helps us predict the pharmacokinetic properties of mizagliflozin in humans.
Mizagliflozin was revealed to have the desired pharmacokinetic properties as an SGLT1 inhibitor to minimise systemic exposure. After oral administration of mizagliflozin to rats, the plasma concentrations of mizagliflozin reached the maximum at 0.23 h and declined with quite a short t1/2 of 1.14 h. Pharmacokinetic analysis showed low absolute bioavailability (0.02%), Vdss (111.8 mL/kg), and systemic exposure (0.49 nghr/mL as AUC0-24), representing a limited amount of mizagliflozin absorbed and

distributed to tissues. The rapid decrease of radioactivity in plasma was also observed after intravenous administration of [14C]mizagliflozin. Although t1/2 was longer than
that of mizagliflozin, based on the result from cold PK study, the plasma radioactivity in the latter period can be mainly attributed to metabolites, not mizagliflozin. These results suggested that the systemic exposure to mizagliflozin could be limited in clinical use. The distribution of radioactivity in most of the tissues was lower than 10 ng eq./g (Table 3). With the exception of the gastrointestine, the high distribution of radioactivity was observed in the liver and kidney. Radioactivity declined over time and undetectable at
72 h after dosing in most of the tissues. Overall, no tissues exhibited high accumulation of radioactivity through study periods. Liver and kidney were considered to play an important role in the excretion of mizagliflozin and its metabolites. The radioactivity in kidney was explained by mizagliflozin, KP232, KP232 glucuronide, and M1 (Table 5). On the other hand, Table 6 shows that metabolites in urine were mainly composed of KP232, KP232 glucuronide, and M1, but not mizagliflozin. Therefore, uptake transporters could be involved in the disposition of mizagliflozin and also its metabolites. SGLT1 is expressed in not only small intestine but also other organs including kidney and liver. Therefore, mizagliflozin may be distributed to these organs mediated by SGLT1. However based on these findings, it is suggested that
accumulation of mizagliflozin in tissues will be limited in humans and therefore that the distribution property of mizagliflozin is less likely to be a concern related to toxicities caused by high accumulation of mizagliflozin in tissues.
In general, the systemic concentration and the distribution of drug may be altered if the extent of plasma protein binding is changed. Guidelines from European Medicines Agency (EMA) and Ministry of Health, Labour and Welfare (MHLW) state that highly bound drugs with the following properties should have the risk of clinically

relevant interactions caused by displacement from plasma protein binding site: (1) a narrow therapeutic window; (2) a high hepatic extraction ratio (if administered intravenously); and/or (3) a high renal extraction ratio (EMA 2012, MHLW 2014). In addition, highly bound drugs with small Vdss are also included in MHLW guidelines. As pharmacokinetic analysis showed, mizagliflozin has quite a small Vdss of 111.8 mL/kg, which was smaller than 297 mL/kg, the volume of extracellular fluid of rat (Davies and Morris. 1993). Thus, plasma protein binding study on mizagliflozin was performed to clarify the extent of protein binding and to investigate the susceptibility to protein binding displacement. This study revealed that mizagliflozin moderately bound to plasma proteins in rat and human plasma and that there is no remarkable species differences in plasma protein binding among mice, rats, monkeys, and humans. Therefore, although mizagliflozin has a relatively small Vdss, drug–drug interaction via the competition of plasma protein binding is less likely to occur in clinical settings. KP232 (aglycone of mizagliflozin) and KP232 glucuronide were also circulated, but the extent of plasma protein binding has not been determined. Plasma protein bindings of KP232 and KP232 glucuronide may need to be considered if these metabolites are also abundant in humans. Excretion study showed the rapid excretion of the radioactivity. Almost the orally administered dose of [14C]mizagliflozin was excreted via the faeces up to 24 h. Less than 1% of administered radioactivity was excreted via the urine. There were no marked sex differences in excretion of [14C]mizagliflozin-related radioactivity
in rats (unpublished observations). In addition, the preliminary study on excretion routes of non-radiolabeled mizagliflozin in BDC rats after intravenous dosing (0.3 mg/kg) demonstrated the cumulative excretion (0–8 h) was 36.5±6.0% and 23.5±27.2% in urine and bile, respectively (unpublished observations). This indicates that mizagliflozin in plasma was eliminated by the excretion into both urine and bile. Mizagliflozin is

designed to be poorly absorbed due to gut metabolism. If absorbed, mizagliflozin can be excreted into both urine and bile.
Metabolic pathways were postulated based on findings from in vivo metabolite profiling studies (Figure 6). Metabolite composition studies with the GI tracts (including the contents) showed the metabolic clearance of mizagliflozin in the gastrointestine. No metabolite was detected in the stomach, and KP232 was the only metabolite that appeared in the small intestine. At 0.25 h after administration, KP232 was already formed in the upper small intestine, and the amount of KP232 was nearly equal to that of mizagliflozin. As the radioactivity moved to the lower gastrointestine over time, the ratio of KP232 increased. In the large intestine, almost all radioactivity consisted of KP232. This implies that mizagliflozin is resistant to hydrolysis at a low pH condition. Thus, it is suggested that KP232 is formed enzymatically in the gastrointestine, and that the degradation of mizagliflozin in the gastrointestine may contribute to low absolute bioavailability.
The most abundant metabolite in the plasma at 4 h was KP232 O-glucuronide followed by KP232. Mizagliflozin was thought to be metabolised to KP232 in the gastrointestine. KP232 and, to a lesser extent, mizagliflozin might be absorbed into the body and then undergo further metabolism such as glucuronidation of KP232. KP232 underwent N-dealkylation to form M2, which was relatively abundant in the plasma. M1 and M3 were also detected at 0.5 h. M1 was formed by O-dealkylation of KP232 followed by glucuronidation. M3 was N-dealkylated KP232 followed by glucuronidation. The radioactivity in the plasma at 4 h was mainly explained by KP232 and KP232 glucuronide. KP232 and KP232 glucuronide were major metabolites in other tissues and excreta as well. Therefore, it is supposed that these compounds will not cause excessive pharmacological or adverse effects resulting from SGLT1 inhibition

because neither KP232 nor KP232 glucuronide have inhibitory effects on SGLT1 (unpublished observations).
Mizagliflozin has a low absolute bioavailability; thus, it was hypothesised that the unabsorbed fraction of mizagliflozin should be in the faeces. Additionally, the biliary excretion study with BDC rats demonstrated that approximately 31.9% of orally
administered radioactivity was eliminated via bile, and 85.2% of the radioactivity in bile consisted of KP232 O-glucuronide. Therefore, it seemed reasonable to expect that KP232 O-glucuronide would be also detected in the lower gastrointestine and the
faeces. As PK studies showed, the time-concentration curves of KP232 and KP232 glucuronide had second peaks, which indicated these underwent enterohepatic circulation. However, only a small amount of mizagliflozin and KP232 glucuronide was detected in the lower gastrointestine. The radioactivity in the faeces was predominantly composed of KP232.
One possible explanation of these results is that mizagliflozin and KP232 O- glucuronide were considered to be metabolised to KP232 in the GI tract. In terms of the chemical structures of mizagliflozin and KP232 O-glucuronide, these compounds are likely to be the substrate of -glycosidase because both compounds have the -O- glycoside moiety. This is supported by the result of metabolite profiling studies using GI tracts. As the radioactivity moved from the upper small intestine to the lower small intestine after administration of [14C]mizagliflozin, [14C]mizagliflozin gradually disappeared and instead KP232 increased. In the cecum, the most abundant component of the radioactivity was found to be KP232. After oral administration of [14C]mizagliflozin (10 mg/kg), composition of metabolites were explored in small intestine tissue without contents at 0.5 h post-dose. KP232 and KP232 glucuronide accounted for 51.1% and 9.6% of the total radioactivity, respectively. Hence,

approximately 60% of mizagliflozin was metabolized to aglycone within 0.5 h in the small intestine (unpublished observations). This suggests that gut metabolism has an important role in first pass gut metabolism.
In the degradation of mizagliflozin, there are at least two possible players. One is bacterial flora in the GI lumen. Enterobacteria hydrolyse glycosides to aglycone (Hidalgo et al. 2012; Ozdal et al. 2016). In general, enterobacteria are more abundant in the lower segment of the GI tract (Donaldson et al. 2016). Therefore, hydrolysis of glycosides by enterobacteria mainly takes place in the lower gastrointestine. Some drugs and endogenous metabolites (e.g., bile acids) are conjugated with glucuronide in the liver and then excreted into the bile (Smith, 1966; Peris-Ribera et al. 1991, Atsumi et al. 1991). In the gastrointestine, these are deconjugated by enterobacteria to become lipophilic and absorbed into the body again. Thus, these metabolites undergo enterohepatic circulation. It is possible that mizagliflozin and KP232 glucuronide are also deconjugated to KP232 by enterobacteria in the lower gastrointestine and undergo enterohepatic circulation.

Enzymes with -glycosidase activity expressed in epithelial cells of the small intestine also play a role in the degradation of glycosides as well. Usually, these enzymes are responsible for the degradation of compounds such as sugar and flavonoid glycosides contained in dietary foods and supplements. Several kinds of -glycosidase were reported to have -glucosidase activity including lactase phlorizin hydrolase (LPH), cytosolic -glycosidase, glucocerebrosidase, and so on (Ketudat Cairns and Esen, 2010). Among these, LPH is known to hydrolyse phlorizin that is known to inhibit SGLT1 and SGLT2 non-selectively (Ehrenkranz et al. 2005). Other flavonoid or isoflavonoid glycosides, such as quercetin 3-O--glucoside, are also metabolised by LPH (Day et al. 2000; Németh et al. 2003). A recent study demonstrated that broad-specific -glucosidase metabolise calycosin 7-O--

glucoside (Shi et al. 2016).It is possible for these enzymes to metabolise mizagliflozin to its aglycone KP232. To summarise, mizagliflozin has a glucoside moiety, which is probably cleaved to form aglycone KP232 by glucosidase in the gut wall and/or that of gut microflora. KP232 is suggested to be absorbed and conjugated with glucuronic acid. KP232 glucuronide can be excreted into bile and then metabolized to KP232 in the gut again. Thus, it is possible that secondary peak occurred in rat is observed also in humans if there is no difference between rats and humans in glucosidase activity of mizagliflozin and also possibly in UDP-glucuronosyltransferase activity and biliary excretion.Conclusion

The pharmacokinetics and metabolism of mizagliflozin were investigated in rats. After oral administration of mizagliflozin, quite a small amount of mizagliflozin was absorbed. Mizagliflozin was mostly metabolised to the pharmacologically inactive aglycone KP232 in the gastrointestine, where bacterial flora or digestive enzymes in the brush border membrane may play a central role in mizagliflozin metabolism. Hence, the major component in plasma was not mizagliflozin but KP232 glucuronide, which is also a pharmacologically inactive metabolite. Absorbed fractions of mizagliflozin and its metabolites were eliminated from the body mainly via the bile. These results confirmed that the pharmacokinetic character of mizagliflozin matches the concept of non- absorbable SGLT1 inhibitor, and these findings should be useful for the accurate prediction of mizagliflozin pharmacokinetics in clinical settings.

References
Atsumi R, Suzuki W, Hakusui H. (1991). Identification of the metabolites of irinotecan, a new derivative of camptothecin, in rat bile and its biliary excretion. Xenobiotica 21: 1159-69.
Davies and Morris. (1993). Physiological parameters in laboratory animals and humans. Pharm Res. 10:1093-5
Day AJ, Cañada FJ, Díaz JC, et al. (2000). Dietary flavonoid and isoflavone glycosides are hydrolysed by the lactase site of lactase phlorizin hydrolase. FEBS Lett 468: 166-70.
Donaldson GP, Lee SM, Mazmanian SK. (2016). Gut biogeography of the bacterial microbiota. Nat Rev Microbiol 14: 20-32.
Ehrenkranz JR, Lewis NG, Kahn CR. et al. (2005). Phlorizin: a review. Diabetes Metab Res Rev 21: 31-8.
EMA, EMA Guideline on the investigation of drug interactions, July 2012. http://www.ema.europa.eu/docs/en_GB/document_library/Scientific_guideline/2 012/07/WC500129606.pdf
Fukudo S, Hongo M, Kaneko H, et al. (2011). Efficacy and safety of oral lubiprostone in constipated patients with or without irritable bowel syndrome: a randomized, placebo-controlled and dose-finding study. Neurogastroenterol Motil 23: 544- e205.
Gorboulev V, Schürmann A, Vallon V, et al. (2012). Na(+)-D-glucose cotransporter SGLT1 is pivotal for intestinal glucose absorption and glucose-dependent incretin secretion. Diabetes 61: 187-96.
Hidalgo M, Oruna-Concha MJ, Kolida S, et al. (2012). Metabolism of anthocyanins by human gut microflora and their influence on gut bacterial growth. J Agric Food Chem 60: 3882-90.
Inoue T, Takemura M, Fushimi N, et al. (2017). Mizagliflozin, a novel selective SGLT1 inhibitor, exhibits potential in the amelioration of chronic constipation. Eur J Pharmacol 806: 25-31.
Isaji, M. (2011). SGLT2 inhibitors: molecular design and potential differences in effect. Kidney Int Suppl: S14-9.
Johanson JF and Kralstein J. (2007). Chronic constipation: a survey of the patient perspective. Aliment Pharmacol Ther 25: 599-608.
Kanai Y, Lee WS, You G, et al. (1994). The human kidney low affinity Na+/glucose cotransporter SGLT2. Delineation of the major renal reabsorptive mechanism for D-glucose. J Clin Invest 93: 397-404.
Ketudat Cairns JR and Esen A. (2010). β-Glucosidases. Cell Mol Life Sci 67: 3389-405. Longstreth G F, Thompson WG, Chey WD, et al. (2006). Functional bowel disorders.
Gastroenterology 130: 1480-91.
McCallum IJ, Ong S, Mercer-Jones M. (2009). Chronic constipation in adults. BMJ 338: b831.
MHLW, Drug interaction guideline for drug development and labeling recommendations (final draft), July 2014. https://www.pmda.go.jp/files/000206158.pdf
Németh K, Plumb GW, Berrin JG, et al. (2003). Deglycosylation by small intestinal epithelial cell beta-glucosidases is a critical step in the absorption and metabolism of dietary flavonoid glycosides in humans. Eur J Nutr 42: 29-42.
Ozdal T, Sela DA, Xiao J, et al. (2016). The reciprocal interactions between polyphenols and gut microbiota and effects on bioaccessibility. Nutrients 8: 78.

Peris-Ribera JE, Torres-Molina F, Garcia-Carbonell MC, et al. (1991). Pharmacokinetics and bioavailability of diclofenac in the rat. J Pharmacokinet Biopharm 19: 647-65.
Scheen AJ. (2015). Pharmacodynamics, efficacy and safety of sodium-glucose co- transporter type 2 (SGLT2) inhibitors for the treatment of type 2 diabetes mellitus. Drugs 75: 33-59.
Shi J, Zheng H, Yu J, et al. (2016). SGLT-1 Transport and Deglycosylation inside Intestinal Cells Are Key Steps in the Absorption and Disposition of Calycosin- 7-O-β-d-Glucoside in Rats. Drug Metab Dispos 44: 283-96.
Shibazaki T, Tomae M, Ishikawa-Takemura Y, et al. (2012). KGA-2727, a novel selective inhibitor of a high-affinity sodium glucose cotransporter (SGLT1), exhibits antidiabetic efficacy in rodent models. J Pharmacol Exp Ther 342: 288- 96.
Smith RL. (1966). The biliary excretion and enterohepatic circulation of drugs and other organic compounds. Fortschr Arzneimittelforsch 9: 299-360.
Sun SX, Dibonaventura M, Purayidathil FW, et al. (2011). Impact of chronic constipation on health-related quality of life, work productivity, and healthcare resource use: an analysis of the National Health and Wellness Survey. Dig Dis Sci 56: 2688-95.
Wright EM, Hirayama BA, Loo DF. (2007). Active sugar transport in health and disease. J Intern Med 261: 32-43.
Zhang Y, Liu ZP. (2016). Recent developments of C-aryl glucoside SGLT2 inhibitors. Curr Med Chem 23: 832-49.

Accepted

Manuscript
Accepted

Manuscript
Accepted

Manuscript
Accepted