**2. Materials and Methods**

#### *2.1. Chemicals*

Metformin hydrochloride was donated from Daelim Pharmaceutical Company (Seoul, Korea). Verapamil hydrochloride, norverapamil, ipriflavone (internal standard (IS) for the high-performance liquid chromatography (HPLC) analysis of metformin) and propranolol (IS for the HPLC analysis of verapamil and norverapamil), the reduced form of β-nicotinamide adenine dinucleotide phosphate, tris(hydroxymethyl)-aminomethane (Tris) buffer and ethylenediamine tetraacetic acid were purchased from Sigma–Aldrich Corporation (St. Louis, MO, USA). HEK-293 cells overexpressing OCT1 (SLC22A1) and OCT2 (SLC22A2) were purchased from Corning Life Sciences (Corning, NY, USA). Other chemicals were of reagent or HPLC grade.

#### *2.2. Animals*

The protocol for the animal studieswas approved by the Animal Care and Use Committee of the College of Pharmacy, Dongguk University-Seoul, Korea (approval no. IACUC-2013-006, 15 December 2013). Male Sprague–Dawley rats (5–7 weeks old, weighing 190–260 g) were purchased from Taconic Farms Inc. (Samtako Bio Korea, O-San, Korea). The rats were housed and handled similarly based on the reported methods [17,33,34].

### *2.3. Pharmacokinetic Studies of Metformin, Verapamil and Both Drugs in Rats*

Early in the morning, the carotid artery (for blood sampling in the intravenous and oral studies) and the jugular vein (for intravenous drug administration only in the intravenous study) of the rats were cannulated, similar to previously reported methods [17,33,34]. The rats were not restrained during the experimental period.

For the intravenous study, 30 mg (2 mL)/kg of metformin (as metformin hydrochloride dissolved in 0.9% NaCl-injectable solution; *n* = 6), 20 mg (2 mL)/kg of verapamil (as verapamil hydrochloride dissolved in 0.9% NaCl-injectable solution; *n* = 6) and both drugs together (*n* = 6) were administered to the rats. Blood samples (approximately 0.12 or 0.22 mL for each drug alone or both drugs together, respectively) were collected via the carotid artery at 0, 1, 5, 15, 30, 60, 90, 120, 180, 240, 300 or 360 min after the start of the drug administration. After each blood sampling, 0.3 mL of 0.9% NaCl-injectable solution containing heparin (20 U/mL) was immediately flushed into each cannula to prevent blood clotting. The blood samples were immediately centrifuged and a 50 µL (or two 50 µL for both drugs) of plasma sample was stored at −80 ◦C (Revco ULT 1490 D-N-S; Western Mednics, Asheville, NC, USA). The 24-h urine sample (Ae0–24 h) and the gastrointestinal tract (including its contents and feces) sample at 24 h (GI24 h) were prepared following previously reported methods [17,33,34], and the samples were also stored at −80 ◦C.

For the oral study, after overnight fasting with free access to water, the same doses of metformin (*n* = 5), verapamil (*n* = 5) and both drugs together (*n* = 5) were orally administered (total oral volume of 6 mL/kg) to the rats using a gastric gavage tube. Blood samples were collected via the carotid artery at 0, 5, 15, 30, 60, 90, 120, 180, 240, 360, 480 or 600 min after the drug administration. Other procedures were similar to those for the intravenous study.

#### *2.4. E*ff*ect of Verapamil on Metformin Uptake in HEK-293 Cells Overexpressing OCT1 or OCT2*

To investigate whether OCT1 and OCT2, as the main transporters affecting metformin pharmacokinetics, were changed by verapamil, the effect of verapamil on OCT1- or OCT2-mediated metformin uptake was investigated in HEK-293 cells overexpressing OCT1 or OCT2 following previously published procedures [33,34]. Briefly, a density of 4.0 × 10<sup>5</sup> cells/well of HEK-293 cells overexpressing either OCT1 or OCT2 were seeded into 24-well plates coated with poly-D-lysine (Corning Incorporated, Corning, NY, USA) and incubated with Dulbecco's Modified Eagle Medium supplemented with 10% fetal bovine serum for 24 h. After washing twice with pre-warmed Hank's Balanced Salt Solution with Ca2<sup>+</sup> and Mg2<sup>+</sup> (Hank's buffer) and pre-incubating the cells with Hank's buffer for 10 min, Hank's buffer was replaced to Hank's buffer containing 10 µM metformin with verapamil added as an inhibitor (0–100 µM). The concentrations of metformin and verapamil were chosen in the ranges of concentrations used in HEK-293 cells overexpressing OCT1 or OCT2 in the previous reports [33,34] and a protocol of in vitro screening method for human SLC uptake transporter inhibition (OCT1 and OCT2) by Cyprotex. Metformin uptake was initiated at this time and the cells were incubated at 37 ◦C for 10 min. At 10 min after starting metformin uptake, Hank's buffer was removed and the cells were immediately washing twice with ice-cold Hank's buffer to stop metformin uptake. The cells were lysed with distilled water and harvested by scraping them off in 200 µL distilled water followed by ultra-sonification at 4 ◦C for 10 s. After centrifuging the cells at 15,000× *g* for 10 min at 4 ◦C, the supernatant was determined by LC/MS/MS analysis of metformin [33,34]. The half-maximal inhibitory constant (IC50) values of verapamil for the inhibition of OCT1 or OCT2-mediated metformin uptake are expressed as % of metformin uptake without verapamil (exposed to vehicle instead of verapamil) in HEK-293 cells overexpressing OCT1 or OCT2, arbitrarily set at 100% (*n* = 2 for each dose). From percentages of metformin uptake versus inhibitor concentrations, a sigmoid shaped curve was fitted to the data and IC<sup>50</sup> was calculated by fitting Hill equation to the data using GraphPad Prism 5 (GraphPad Software Inc., San Diego, CA, USA).

#### *2.5. E*ff*ect of Verapamil on Metformin Concentration in the Liver and Kidneys*

To investigate whether verapamil changes the metformin concentration in the liver and kidneys, a tissue distribution study of metformin with and without verapamil was conducted following a previously reported method [17,33,34]. At 0.5, 1, 3 and 6 h after intravenous or oral administration of metformin with and without verapamil at the same doses as in the pharmacokinetic study, as much blood as possible was collected and each rat was then sacrificed by lethal blood loss. The liver

and kidneys were excised, weighed and homogenized in a 4-fold volume of 0.9% NaCl-injectable solution (*n* = 4 for each organ). After centrifuging each homogenate at 15,000× *g* for 10 min at 4 ◦C, the supernatant was collected and stored at −80 ◦C.

### *2.6. E*ff*ect of Metformin on Verapamil Metabolism in Rat Hepatic and Intestinal Microsomes*

To investigate the effect of metformin on verapamil metabolism, the measurement of kinetic constants, such as *V*max (maximum velocity) and *K*<sup>m</sup> (apparent Michaelis–Menten constant; the concentration at which the rate is one half of the *V*max) for verapamil metabolism, with and without metformin, in hepatic and intestinal microsomes were conducted. Hepatic and intestinal microsomes were prepared by the previously reported method as followings [17,37,38]: freshly excised livers were cut in pieces, washed extensively with ice-cold solution (KCl 0.15 M) to remove remaining blood and were homogenized with Tris-HCl buffer (pH 7.5) containing 0.15 M KCl + 50 mM Tris, 1 mM EDTA in a Potter-Elvehjem glass homogenizer for 30 s. The homogenate was centrifuged for 10 min at 10,000× *g* and 4 ◦C and then followed by ultracentrifugation of the remaining supernatant for 1 h at 100,000× *g* and 4 ◦C. Microsomal pellets were then re-suspended in the same buffer with a hand homogenizer and re-centrifuged for 1 h at 100,000× *g* and 4 ◦C. The supernatant was discarded, and the microsomal pellets were carefully overlaid with 0.15 mol KCl buffer and stored at −80 ◦C. Intestinal microsomes were prepared using freshly excised proximal and middle sections of the small intestine. This part of intestine was excised, rinsed with ice-cold 0.01 M potassium phosphate buffer with 1.15% KCl (pH 7.4), filled with solution A (1.5 mM KCl + 96 mM NaCl + 27 mM sodium citrate + 8 mM KH2PO<sup>4</sup> + 5.6 mM Na2HPO<sup>4</sup> + 40 µg/mL PMSF). The intestine filled with solution A was incubated in a 37 ◦C water bath for 15 min. After discarding solution A, the intestine was filled with ice-cold solution B (phosphate-buffered saline + 1.5 mM EDTA + 0.5 mM dithiothreitol + 40 µg/mL PMSF), wound around a middle finger and tapped against the finger three times. The upper villus cells were released into solution B during this process, and the collected cells were pooled. The pooled solution was centrifuged at 10× *g* and 4 ◦C for 5 min. After discarding the supernatant, approximately 15 mL of ice-cold solution C (5 mM histidine + 0.25 M sucrose + 0.5 mM EDTA + 40 µg/mL PMSF) was added into each centrifuge tube, which was inverted twice. Following the discard of the supernatant, the cells were resuspended in fresh ice-cold solution C, homogenized with a Pyrex glass Potter-Elvehjem homogenize and centrifuged at 10,000× *g* and 4 ◦C for 20 min. The supernatant was then centrifuged at 100,000× *g* and 4 ◦C for 65 min. The pellet of intestinal microsome was resuspended in 0.2 mM EDTA/20% glycerol/80% 0.1 M phosphate buffer (pH 7.4), homogenized and stored at −80 ◦C. Protein contents in hepatic and intestinal microsomes were measured by the reported method [39].

In hepatic microsomes (equivalent to 1 mg protein), 2.5 µL of 0.9% NaCl-injectable solution containing 2.5, 5, 10, 20 or 50 µM verapamil (the substrate) as final concentrations, 2.5 µL of 0.9% NaCl-injectable solution containing 10 µM metformin (the inhibitor) as a final concentration and 50 µL of 1 mM of NADPH dissolved in 0.1 M phosphate buffer of pH 7.4 were added. The total volume, 0.5 mL, was adjusted by adding 0.1 M phosphate buffer (pH 7.4), and then the components were incubated at 37 ◦C using a thermomixer at 500 opm. In intestinal microsomes, verapamil (as the substrate) concentrations of 5, 10, 20, 50 or 200 µM were used and the other conditions were the same as in the hepatic microsome study. After incubation for 15 or 30 min of the hepatic and intestinal microsomes, respectively, 1 mL of diethylether containing 1 µg/mL of propranolol, as an IS, was added to terminate the reaction. The verapamil concentration in each sample was determined by HPLC analysis [29].

The *K*<sup>m</sup> and *V*max for the verapamil metabolism were calculated using a non-linear regression method [40]. The unweighted kinetic data from the hepatic and intestinal microsomes were fitted using a single-site Michaelis−Menten equation; *V* = *V*max × [S]/(*K*<sup>m</sup> + [S]), in which [S] was the substrate concentration. The intrinsic clearance (CLint) was calculated by dividing the *V*max by the *K*m.
