*3.3. Substrate Specificity of DWP16001, Dapagliflozin, and Ipragliflozin for OAT1 and OAT3*

The AUC ratios of these three SGLT2 inhibitors were much greater than unity, suggesting that DWP16001, dapagliflozin, and ipragliflozin are distributed in the kidney, in which tissue protein binding ability to the specific protein or drug transporters may be involved. Therefore, firstly, we measured the protein binding of DWP16001, dapagliflozin, and ipragliflozin in the plasma and kidney homogenate.

The plasma and kidney tissue protein binding of DWP16001, dapagliflozin, and ipragliflozin were high and comparable and the logP values of three compound were also comparable (i.e., in the range of 2.27–2.65) (Table 2). It suggests that the large distribution of DWP16001, dapagliflozin, and ipragliflozin in the kidney was not necessarily reliant on plasma/tissue-specific binding ability or lipophilicity of these compounds.

**Table 2.** Protein binding of DWP16001, dapagliflozin, and ipragliflozin in mouse plasma and kidney homogenates.


a logP values were obtained from the partition coefficient of n-octanol/water; <sup>b</sup> Data expressed as mean ± SD of triplicate experiments.

We conducted further investigations into the tissue-specific transport of DWP16001, dapagliflozin, and ipragliflozin by using cell systems overexpressing OAT1 and OAT3 since OAT1 and OAT3 are exclusively distributed in the kidney and contribute to the drug distribution to the kidney [16]. The functionality of the OAT1 and OAT3 transport system was confirmed by the significantly greater uptake rates of the probe substrates in HEK293-OAT1 and -OAT3 cells than that in HEK293-mock cells (i.e., 14.7-fold increase in PAH uptake for OAT1 and 16.4-fold increase in ES uptake for OAT3) and by the decreased uptake rate of each probe substrate by the addition of a representative inhibitor, probenecid (Figure 5A,B) [17]. The uptake of DWP16001 in HEK293-OAT1 and OAT3 cells was significantly higher than that in HEK293-mock cells, and it was significantly decreased by the presence of probenecid (Figure 5C). However, dapagliflozin and ipragliflozin were not substrates for both OAT1 and OAT3 (Figure 5D,E). Next, we investigated the concentration dependency in the OAT1 and OAT3-mediated DWP16001 uptake. As shown in the figure, the OAT1 and OAT3-mediated uptake of DWP16001 showed saturable kinetics, and the kinetic parameters, such as K<sup>m</sup> and Vmax, calculated from the concentration-dependent uptake of DWP16001 in HEK293-OAT1 and –OAT3 cells are shown in Figure 6. Collectively, OAT1 and OAT3-mediated uptake of DWP16001 could contribute to the highest kidney distribution of DWP16001 among three SGLT2 inhibitors. However, these processes were not dominant over the cellular uptake of the three SGLT2 inhibitors into HEK293-mock cells (i.e., passive diffusion), which is evident from two-fold increases in HEK293-OAT1 or -OAT3 cells compared with mock cells (Figure 5C). Therefore, OAT1- and OAT3-mediated active transport and passive diffusion all contributed to the highly permeable absorptive process of DWP16001 in the kidney, and dapagliflozin and ipragliflozin also showed higher uptake levels into HEK293 cells via passive diffusion.

69

**OAT1-mediated uptake** 

**of DWP16001 (pmol/min)**

**Figure 5.** (**A**) Uptake of 0.1 µM [3H]para-aminohippuric acid (PAH) was measured in HEK293-mock and –OAT1 cells in the presence and absence of 20 µM probenecid for 5 min. (**B**) Uptake of 0.1 µM [ <sup>3</sup>H]Estrone-3-sulfate (ES) was measured in HEK293-mock and –OAT3 cells in the presence and absence of 20 µM probenecid. Uptake of (**C**) DWP16001, (**D**) dapagliflozin, and (**E**) ipragliflozin (2 µM each) into HEK293-mock and HEK293 cells expressing OAT1 and OAT3 was measured for 5 min. Each data point represents the mean±standard deviation of triplicate experiments. \*: *p* < 0.05, compared with mock cells; +: *p* < 0.05, compared with control group. **mock OAT1 OAT3 mock OAT1 OAT3 mock OAT1 OAT3**

**DWP16001 ( M)** 0 10 20 30 40 50 60 **DWP16001 ( M)** 0 10 20 30 40 50 60 **Figure 6.** (**A**) OAT1- and (**B**) OAT3-mediated uptake of DWP16001. Concentration dependent uptake of DWP16001 was measured for 5 min into HEK293-mock cells and HEK293 cells expressing OAT1 and OAT3. The transporter-mediated uptake rate was obtained by subtracting the uptake in HEK293-mock cells (passive diffusion) from those in HEK293 cells expressing OAT1 and OAT3 (total uptake). Each data point represents the means ± standard deviation from triplicate experiments.

*3.4. Inhibition Potential of DWP16001, Dapagliflozin, Ipragliflozin on SGLT1 and SGLT2 Activities*

3.4.1. Comparison of IC<sup>50</sup> Values of DWP16001, Dapagliflozin, and Ipragliflozin on SGLT1 and SGLT2

To confirm the functionality of CHO-SGLT1 cells, we measured the time-dependent uptake of AMG, a representative substrate for SGLT1 [11,18]. As shown in Figure 7A, AMG uptake was increased with incubation time, and the optimal incubation time was selected as 2 h from the linear phase of the slope. AMG uptake into CHO-SGLT1 cells was 42-fold greater than that into CHO-mock cells and significantly inhibited by the presence of known inhibitors, such as dapagliflozin and ipragliflozin [3,5] (Figure 7B). With the same experimental condition, the inhibitory effect of DWP16001, dapagliflozin, and ipragliflozin on the AMG uptake into CHO-SGLT1 cells was measured in a concentration range of 1–50,000 nM and IC<sup>50</sup> values were calculated (Figure 7C–E; Table 3).

**Figure 7.** (**A**) Uptake of 10 µM [14C]methyl-a-D-glucopyranoside (AMG) was measured in CHOsodium-glucose cotransporter 1 (SGLT1) cells with a various incubation time (0.5–3 h). (**B**) Inhibitory effect of dapagliflozin and ipragliflozin (1, 10 µM) on the uptake of 10 µM [14C]AMG in CHO-mock and -SGLT1 cells was measured for 2 h. Concentration dependent inhibition of DWP16001 (**C**), dapagliflozin (**D**), and ipragliflozin (**E**) on the SGLT1-mediated uptake of [14C]AMG. SGLT1-mediated uptake of [14C]AMG was calculated by subtracting the uptake of 10 µM [14C]AMG for 2 h in CHO-mock cells from that in CHO-SGLT1 cells in a concentration range of 1–50,000 nM of DWP16001, dapagliflozin, and ipragliflozin. Each data point represents the mean ± standard deviation of three independent experiments. \*: *p* < 0.05, compared with control group.


**Table 3.** Inhibition potential of DWP16001, dapagliflozin, and ipragliflozin on SGLT1 and SGLT2 activities.

Data were expressed as mean ± SD from triplicate measurement.

Similarly, the functionality of CHO-SGLT2 cells was confirmed from the time-dependent uptake of AMG, a representative substrate for SGLT2 [3,5]. As shown in Figure 8A, AMG uptake was increased with incubation time, and the optimal incubation time was selected as 2 h from the linear phase of the slope. AMG uptake into CHO-SGLT1 cells was 9.8-fold greater than that into CHO-mock cells and significantly inhibited by the presence of known inhibitors, such as dapagliflozin and ipragliflozin [3] (Figure 8B). With the same experimental conditions, the inhibitory effect of DWP16001, dapagliflozin, and ipragliflozin on the AMG uptake into CHO-SGLT2 cells was measured in a concentration range of 0.001–100 nM, and IC<sup>50</sup> values were calculated (Figure 8C–E; Table 3).

DWP16001 showed greater affinity for SGLT2 compared with dapagliflozin and ipragliflozin. Moreover, the affinity to SGLT1 was also greater with DWP16001 compared with dapagliflozin and ipragliflozin. When compared with selectivity, which was calculated from the SGLT1/SGLT2 IC<sup>50</sup> ratio, the selectivity of DWP16001 was high relative to that of dapagliflozin and ipragliflozin (Table 3).

**Figure 8.** (**A**) The uptake of 10 µM [14C]methyl-a-D-glucopyranoside (AMG) was measured in CHO-SGLT2 cells with a various incubation time (0.5, 1, 1.5, 2, 3 h). (**B**) Inhibitory effect of dapagliflozin and ipragliflozin (10, 100 nM) on the uptake of 10 µM [14C]AMG in CHO-mock and –SGLT2 cells was measured for 2 h. Representative concentration dependent inhibition of DWP16001 (**C**), dapagliflozin (**D**), and ipragliflozin (**E**) on the SGLT2-mediated uptake of [14C]AMG. SGLT2-mediated uptake of [14C]AMG was calculated by subtracting the uptake of 10 µM [14C]AMG for 2 h in CHO-mock cells from that in CHO-SGLT2 cells in a concentration range of 1–100 nM of DWP16001, dapagliflozin, and ipragliflozin. Each data point represents the mean±standard deviation of three independent experiments. \*: *p* < 0.05, compared with control group.

#### 3.4.2. Mode of Inhibition

To investigate whether the inhibition of SGLT2 was time-dependent, the inhibitory effects of DWP16001, dapagliflozin, and ipragliflozin on SGLT2-mediated uptake of AMG were measured with or without pretreatment with DWP16001, dapagliflozin, and ipragliflozin, respectively. Figure 9 shows that the inhibitory potential of DWP16001, dapagliflozin, and ipragliflozin on SGLT2 did not change with different pretreatment times, suggesting that all the DWP16001, dapagliflozin, and ipragliflozin inhibited SGLT2 function in a time-independent manner.

To determine the mode of inhibition on SGLT2, the inhibitory effects of DWP16001, dapagliflozin, and ipragliflozin on the AMG uptake into CHO-SGLT2 cells were measured with different substrate concentrations, and the results were expressed as Dixon plots (Figure 10A–C) and a replot of the Dixon slopes (Figure 10D–F) [13,19,20]. Dixon transformation of inhibition profile by DWP16001 (0.001–250 nM) for different concentrations of AMG indicated competitive inhibition of DWP16001 on SGLT2 activity (Figure 10A). A replot of the Dixon slopes vs 1/S produced a straight line converging on the zero (Figure 10D), suggesting competitive and reversible inhibition of DWP16001. Similar results were shown in case of dapagliflozin and ipragliflozin (Figure 10B,C,E,F). The results collectively

suggested that the inhibition of SGLT2 by DWP16001, dapagliflozin, and ipragliflozin is reversible and competitive.

**Figure 9.** Inhibitory effect of (**A**) DWP16001, (**B**) dapagliflozin, and (**C**) ipragliflozin on the SGLT2-mediated uptake of 10 µM [14C]methyl-a-D-glucopyranoside (AMG) was measured with preincubation time for 1 and 2 h or without preincubation of DWP16001, dapagliflozin, and ipragliflozin, respectively, in a concentration range of 0.001–100 nM. Each data point represents the mean±standard deviation of three independent experiments.

● ○ ▼ **Figure 10.** Dixon plot for the inhibitory effect of (**A**) DWP16001, (**B**) dapagliflozin, and (**C**) ipragliflozin on the uptake of [14C]methyl-a-D-glucopyranoside (AMG) in CHO-SGLT2 cells. Each symbol represents the concentration of [14C]AMG: •, 1 µM; #, 2.5 µM; H, 5 µM; △, 50 µM. Replot of the slopes of Dixon plot (slopes vs 1/[S]) was shown for (**D**) DWP16001, (**E**) dapagliflozin, and (**F**) ipragliflozin. V indicates the SGLT2-mediated uptake rate of [14C]AMG and S indicates the concentration of [14C]AMG. The data are the means±standard deviations of triplicate measurements.

(A) (B) (C) 3.4.3. Retained Inhibition Potential of DWP16001 Compared to Dapagliflozin and Ipragliflozin

AMG uptake (% activity) 100 150 0.2 nM 2 nM 20 nM 200 nM 100 150 100 150 The recovery of SGLT2 activities depending on the washout period after 24 h exposure of DWP16001, ipragliflozin, and dapagliflozin in CHO-SGLT2 cells was investigated. As shown in Figure 11, the activity of reduced SGLT2 by treatment with DWP16001, dapagliflozin, and ipragliflozin for 24 h in CHO-SGLT2 cells revealed differences in the recovery of SGLT2 depending on the time of drug removal from the medium, as well as SGLT2 inhibitors. In the case of DWP16001, the activity

0

50

Time (h) 02468

0.2 nM 2 nM 20 nM 200 nM

0

50

Time (h) 02468

0.2 nM 2 nM 20 nM 200 nM

Time (h) 02468

0

50


**(D)**

**DWP16001 (nM)** 0 50 100 150 200 250 300

<sup>400</sup> **(B)**

<sup>4</sup> **(E)**

**1/V**


**SlopeDixon**

0

1

2

3

0

100

200

300

**(A)**

**1/V**


**SlopeDixon**

0

2

3

0

100

200

300

tended to gradually recover as the washout time increased. When treated at a high concentration (200 nM), the activity did not recover even 8 h after the drug was removed from the medium (Figure 11A). However, when dapagliflozin and ipragliflozin were used, SGLT2 activity was recovered to the control level 4 h after the removal of dapagliflozin and ipragliflozin (Figure 11B,C). The results indicate that the dissociation of DWP16001 to SGLT2 is slower and more incomplete than that of dapagliflozin and ipragliflozin and that the binding affinity to SGLT2 is stronger than that of dapagliflozin and ipragliflozin. **1/S** -1 **1/S** -1 **1/S** -1 ● ○ ▼


**Dapagliflozin (nM)** 0 50 100 150 200 250 300

<sup>400</sup> **(C)**

<sup>4</sup> **(F)**

**1/V**


**SlopeDixon**

0

1

2

3

4

0

100

200

300

400

**Ipragliflozin (nM)** 0 50 100 150 200 250 300


**Figure 11.** Recovery of SGLT2 transport function with different washout time following 24 h of incubation with (**A**) DWP16001, (**B**) dapagliflozin, and (**C**) ipragliflozin (0.2–200 nM) was measured through the uptake of [14C]methyl-a-D-glucopyranoside (AMG) in CHO-SGLT2 cells. Each data point represents the mean±standard deviation of three independent experiments.

#### **4. Discussion**

DWP16001 is a candidate SGLT2 inhibitor that is currently under development. As a first step, the pharmacokinetic properties and in vitro SGLT2 inhibition were compared with currently used SGLT2 inhibitors. Dapagliflozin and ipragliflozin were selected based on their high kidney distributions and long elimination half-lives (t1/2) in the kidney [5], which are thought to be important for the efficacy and duration of action of SGLT2 inhibitors. DWP16001 showed higher kidney distributions compared with dapagliflozin and ipragliflozin. DWP16001 also had longer t1/<sup>2</sup> in the kidney than dapagliflozin and ipragliflozin, as well as a comparable plasma profile with ipragliflozin (Table 1). Moreover, the kidney concentration of DWP16001 was maintained over 72 h following oral administration of 1 mg/kg DWP16001 (Figure 4). Taken together, these pharmacokinetic results show that the duration of action of DWP16001 is greater than that of dapagliflozin and ipragliflozin and that the oral therapeutic dose of DWP16001 could be reduced compared with both dapagliflozin and ipragliflozin.

To investigate the underlying mechanisms of highest kidney distribution and maintained concentration of DWP16001 in the kidney, we measured the kidney tissue binding and involvement of OAT1 and OAT3 transporters of the three SGLT2 inhibitors. All three SGLT2 inhibitors showed high protein binding, but the kidney tissue binding performances of these three SGLT2 inhibitors were not different from their plasma protein binding capabilities (Table 2). However, DWP16001 was a substrate for both OAT1 and OAT3, which are dominantly expressed in the kidney, whereas dapagliflozin and ipragliflozin were not (Figures 5 and 6). Although this could not solely explain the high kidney distribution, OAT1- and OAT3-mediated transport process may contribute to the high kidney distribution.

Next, we compared the in vitro SGLT2 inhibition and selectivity of SGLT2 inhibition over SGLT1. All three SGLT2 inhibitors inhibited SGLT2 and SGLT1 activity in a concentration-dependent manner and IC<sup>50</sup> values of dapagliflozin and ipragliflozin were in the range of previous reports (1.0–1.3 nM for dapagliflozin; 6.75–8.07 nM for ipragliflozin) [21,22]. IC<sup>50</sup> values of DWP16001 to SGLT2 and SGLT1 were lower than those of dapagliflozin and ipragliflozin, suggesting a greater affinity to SGLT2

inhibition for DWP16001 with a higher selectivity over SGLT1 than dapagliflozin and ipragliflozin. The mode of inhibition of DWP16001 was not different from the other SGLT2 inhibitors. They all showed reversible and competitive inhibition (Figure 10), which is consistent with other SGLT2 inhibitors [23,24]. However, the affinity to SGLT2 inhibition seemed to be different among the three SGLT2 inhibitors (Table 3). In addition, the recovery of SGLT2 transport activity following the pretreatment of DWP16001, dapagliflozin, and ipragliflozin for 24 h was retained at a higher concentration (200 nM) of DWP16001 compared with dapagliflozin and ipragliflozin. These results suggested that DWP16001 had the highest SGLT2 inhibition potential and that this inhibition potential retained for a longer time compared with dapagliflozin and ipragliflozin. Combined with the higher kidney distribution of DWP16001, retained SGLT2 inhibition with a high concentration of DWP16001 could also potentiate the efficacy of DWP16001 compared with dapagliflozin and ipragliflozin.

The comparative pharmacokinetics and in vitro SGLT2 inhibition findings suggest that DWP16001 might be a superior alternative to dapagliflozin and ipragliflozin; however, we should note that comparisons of the in vivo pharmacologic properties of these agents using therapeutic doses in animals and humans need to be further undertaken.

**Author Contributions:** Conceptualization, M.-K.C., H.-Y.J and I.-S.S.; Investigation, M.-K.C., S.J.N., H.-Y.J., M.J.P., J.-S.C. and I.-S.S.; Writing-Original Draft Preparation, M.-K.C.; Supervision, I.-S.S.; Writing-Review & Editing, I.-S.S. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Conflicts of Interest:** The authors declare no conflict of interest.
