2021 Volume 44 Issue 5 Pages 653-658
Alogliptin (ALG), an inhibitor of dipeptidylpeptidase-4, is used in the management of type 2 diabetes mellitus, and has a high absorption rate (>60–71%), despite its low lipophilicity (logP=−1.4). Here, we aimed to clarify the mechanism of its intestinal absorption. ALG uptake into Caco-2 cells was time-, temperature-, and concentration-dependent, but was not saturated at concentrations up to 10 mmol/L. The uptake was significantly inhibited by the organic anion transporting polypeptide (OATP) substrate fexofenadine and by the OATP inhibitor 4,4′-diisothiocyanostilbene-2,2′-disulfonic acid (DIDS), but was not inhibited by organic cation transporter (OCT)/organic cation/carnitine transporter (OCTN) or peptide transporter 1 (PEPT1) substrates. Grapefruit, orange, and apple juices and their constituents, which are known to strongly inhibit intestinal OATPs, significantly inhibited ALG uptake into Caco-2 cells. The pH dependence was bell-shaped, indicating the involvement of a pH-sensitive transporter. However, ALG uptake by HEK293 cells overexpressing OATP2B1, a key intestinal OATP transporter of amphiphilic drugs, was not different from that of mock cells. In a rat in vivo study, apple juice reduced systemic exposure to orally administered ALG without changing the terminal half-life. These observations suggest that intestinal absorption of ALG is carrier-mediated, and involves a fruit-juice-sensitive transporter other than OATP2B1.
Intestinal drug absorption can occur via simple diffusion according to the pH-partition hypothesis and/or via various types of transporters expressed on enterocytes. In the latter case, drug–drug interaction at the transporters or loss of transport function due to genetic polymorphisms can affect drug efficacy and toxicity.1) Therefore, it is important from the clinical viewpoint to clarify the pharmacokinetic determinants of drug absorption.
Alogliptin (ALG), an inhibitor of dipeptidylpeptidase-4, is used in the management of type 2 diabetes mellitus. Clinical studies have demonstrated that ALG dose-dependently improves plasma glucose levels, and reduces glycated hemoglobin levels in type 2 diabetes mellitus patients, suggesting that its efficacy is dependent on its plasma concentration.2) In addition, ALG is generally used in combination with other glucose-lowering agents, which may influence its efficacy.
ALG is a hydrophilic, weakly cationic drug with pKa 8.5 and log P −1.4 (basic product information of Nesina tablets, ver.12). According to the Henderson–Hasselbalch equation, most of the drug is expected to be present in ionic form at physiological pH in the small intestine, suggesting that its intestinal absorption rate might be very low. However, a pharmacokinetic study with healthy subjects demonstrated that 60–70% of orally administered ALG was recovered in urine in the intact form.3) The absorption of ALG is rapid with a short time to maximum concentration (tmax, 1–2 h) and a long half-life (12–21 h).3) The extent of metabolism is minor; the main metabolites, the N-demethylated and N-acetylated forms, are excreted in urine in amounts of up to 1.8 and 5.6% of the dose, respectively. In addition, renal clearance of ALG (160–200 mL/min) exceeds the glomerular filtration rate (120 mL/min), suggesting the occurrence of active tubular secretion.3) These findings suggest that carrier-mediated transport might be involved in the absorption, distribution and excretion of ALG.
In this study, we focused on clarifying the intestinal absorption mechanism of ALG by means of in vitro studies with colorectal carcinoma-derived Caco-2 cells, which are widely used as a model for human enterocytes. We also investigated drug interactions during the uptake of ALG in vitro and during the absorption process of ALG in rats in vivo.
Alogliptin benzoate (ALG) was extracted from Nesina® 25 mg tablets (Takeda Pharmaceutical Company, Ltd., Osaka, Japan). Briefly, 12 tablets were pulverized in a mortar, and the film coating was removed. The powder was dissolved in 18 mL of ethanol, and the solution was filtered at 70 °C. The filtrate was evaporated to dryness under 40 °C beneath a nitrogen gas flow. The residue was washed with chilled ethanol, and subjected to lyophilization. The structure was confirmed by 1H-NMR, LC-MS and melting point measurements. Grapefruit, orange, and apple juices (Tropicana™; 100% pure at normal strength) were purchased from Kirin Holdings Company, Ltd. (Tokyo, Japan). All other reagents were commercial products of analytical grade or better.
Cell CultureCaco-2 cells were obtained from American Type Culture Collection (Manassas, VA, U.S.A.). Caco-2 cells were cultured at 37 °C in a humidified atmosphere of 5% CO2 in air, in Dulbecco’s modified Eagle’s medium supplemented with 10% fetal bovine serum, 100 units/mL penicillin, 100 mg/mL streptomycin and 1% minimum essential medium (MEM) non-essential amino acids. For uptake studies, Caco-2 cells were seeded onto type I collagen-coated 12-well plates at 5.8 × 104 cells/cm2. The medium was exchanged every two days from four days after seeding. Experiments were conducted between day 14 and day 18. In the present study, Caco-2 cells were used between passages 35 and 48.
Organic anion transporting polypeptide (OATP)2B1-expressing HEK293 cells and mock cells were purchased from GenoMembrane (Yokohama, Japan). The procedures for cell culture and uptake study followed the manufacturer’s instructions.
Uptake StudyCaco-2 cells that had reached confluence were washed and incubated with 0.5 mL of Hanks’ balanced salt solution-2-(N-morpholino)ethanesulfonic acid (HBSS-MES) buffer (pH 6.5) for 5 min at 37 °C. Uptake was initiated by replacing the buffer with 0.3 mL of buffer containing the test compound, and cells were incubated for 5 min. The reaction was terminated by the addition of 1 mL of ice-cold buffer. Cells were washed three times with 2 mL of ice-cold buffer. For the study of pH-dependence, HBSS-MES buffer (pH 6.5) was replaced with buffer of pH 5.0, 5.5, 6.0, 6.5, 7.0, or 7.4, and the same procedure as above was employed. Fruit juice was diluted to 20% with HBSS-MES buffer (pH 6.5) and the pH was adjusted to 6.5.
Animal StudyAll animal study procedures were approved by the Committee on Ethics of Animal Experiments, Tohoku Medical and Pharmaceutical University, and the study was performed according to the Guidelines for the Care and Use of Laboratory Animals at the university. Six-week-old Sprague-Dawley (SD) male rats purchased from Japan SLC, Inc. (Hamamatsu, Japan) were housed in a standard animal maintenance facility for one week before experiments. Seven-week-old rats were fasted for 17 h before absorption experiments. ALG dissolved in water or apple juice was administered orally by gavage at a dose of 20 mg/10 mL/kg. Blood (0.2 mL) samples were collected from the jugular vein into heparinized syringes just before and at 0.5, 1, 2, 3, 4, 5, 6, and 8 h after administration of ALG. Plasma was obtained by centrifugation at 5000 × g (4 °C), and stored at −20 °C until assay.
Quantitative Analysis of ALGFor the in vitro uptake study, cells were lysed with 0.2 mL of 0.5% Triton-X 100. Cell lysates (0.1 mL) were deproteinized with an equal volume of acetonitrile, followed by centrifugation at 5000 × g. The supernatant was diluted 5 times with water for analysis. Protein concentrations in cell lysates were determined by means of the bicinchoninate method.
Plasma samples were deproteinized with 5 volumes of methanol, and the centrifugal supernatant (5000 × g at 4 °C) was filtered through 0.45 µm syringe filters. Each filtrate was subjected to analysis.
ALG concentration was measured using a HPLC-tandem mass spectrometry system consisting of an HPLC system (ACCELA, Thermo Fisher Scientific, Waltham, MA, U.S.A.) and a TSQ Vantage triple quadrupole mass spectrometer (Thermo Scientific, Yokohama, Japan). Aliquots (10 µL) of samples were injected into a HPLC system equipped with a CAPCELL PAK ADME column (Adamantyl, 2.1 × 50 mm, Shiseido, Tokyo, Japan) and eluted isocratically at 0.2 mL/min with 32% methanol containing 0.05% formic acid. ALG was detected in the positive ion, multiple reaction monitoring (MRM) mode, monitoring the m/z transition at 340 > 323, with a tandem quadrupole mass spectrometer fitted with an electrospray ionization source.
Data AnalysisMaximum plasma concentration (Cmax) was obtained directly from the observed data. The area under the plasma concentration–time curve (AUC) from time 0 to 8 h (AUC0–8) was obtained by employing the trapezoidal formula. The terminal half-life was calculated as ln(2)/kel, where kel is the slope of the terminal phase (5–8 h after administration) of the log-linear plasma concentration time curve.
All statistical analyses were performed using ystat 2013 for EXCEL. The significance of difference was evaluated using Student’s t-test or non-repeated measures ANOVA followed by Dunnett’s post hoc test. p < 0.05 was considered statistically significant.
ALG uptake by Caco-2 cells increased linearly for five minutes (Fig. 1). Therefore, all subsequent uptake experiments were performed for 5 min. The initial uptake rate of ALG increased concentration-dependently, up to at least 10 mmol/L (Fig. 1). When the incubation temperature was decreased to 4 °C, the initial uptake rate of ALG was drastically reduced (Fig. 1). These results suggest the involvement of a low-affinity, high-capacity transport system in the uptake of ALG into Caco-2 cells.
ALG uptake was measured at 37 °C (open circle) and 4 °C (closed circle) for 5 min. Inset: Time dependence of ALG uptake into Caco-2 cells. The concentration of ALG was 0.1 mmol/L. Values are mean ± standard deviation (S.D.) (n = 3). Asterisks indicate significant differences from the uptake at 4 °C (* p < 0.05; ** p < 0.01).
In order to identify the transporter involved in the uptake of ALG, the effects of substrates and inhibitors of several transporters were investigated (Fig. 2). The transporters tested were reported to be present and functional in Caco-2 cells.4–8) The concentration of ALG was fixed at 0.1 mmol/L based on the estimated gastric concentration of ALG after administration at the clinical dose (Package insert of Nesina®).
Values are mean ± S.D. (n = 3). The concentrations of alogliptin and inhibitors were 0.1 mmol/L and 1 mmol/L, respectively. #, Cells were pretreated with DIDS for 15 min. L-Car, L-carnitine; TEA, tetraethylammonium; Gly-Sar, glycylsarcosine; Gly-Gly, glycylglycine; MTX, methotrexate; Fexo, fexofenadine; DIDS, 4,4′-diisothiocyanostilbene-2,2′-disulfonic acid; Asterisks indicate significant differences from the control (** p < 0.01).
Although ALG is a cationic drug, the organic cation transporter (OCT/OCTN) inhibitor tetraethyl ammonium (TEA) and the OCTN1 and OCTN2 substrate L-carnitine had no effect on its uptake. However, the OATP substrate fexofenadine (p < 0.01) and the OATP inhibitor 4,4′-diisothiocyanostilbene-2,2′-disulfonic acid (DIDS) (p < 0.01) significantly inhibited ALG uptake. Although the peptide transporter 1 (PEPT1) substrate glycylsarcosine was weakly inhibitory, and another PEPT1 substrate, glycylglycine, enhanced the uptake, these effects were not statistically significant (Fig. 2). These results suggested the involvement of OATPs in ALG uptake into Caco-2 cells at least in part. Therefore, we next examined the effect of fruit juices (orange, apple, grapefruit (OJ, AJ, GFJ)) and their components, which are known to strongly inhibit intestinal OATPs. Fruit juices were diluted to 20% and adjusted to pH 6.5. OJ, AJ and GFJ as well as their constituents, naringin, hesperetin and hesperidin, all drastically reduced the uptake of ALG (p < 0.01 each, Fig. 3). To further investigate the contribution of OATPs, we evaluated the pH dependence of ALG uptake into Caco-2 cells. The pH dependence of ALG uptake showed a bell-shaped profile with a maximum at pH 7.0 (Fig. 4), suggesting the involvement of a pH-sensitive transporter, such as OATPs.
The concentrations of ALG and inhibitors were 0.1 mmol/L and 1 mmol/L, respectively. Fruit juice was diluted to 20% with HBSS-MES buffer (pH 6.5) and the pH was adjusted to 6.5. OJ, orange juice; AJ, apple juice; GFJ, grapefruit juice. Values are mean ± S.D. (n = 3). Asterisks indicate significant differences from the control (** p < 0.01).
Values are mean ± S.D. (n = 3). The concentration of alogliptin was 0.1 mmol/L.
The inhibition and pH dependence study suggested the involvement of OATPs in ALG uptake into Caco-2 cells. Among the OATP family, OATP2B1 is the most abundantly expressed in both Caco-2 cells and the human intestine, and plays a predominant role in the uptake of amphiphilic drugs.8,9) Therefore, we next examined the involvement of OATP2B1 in ALG uptake. ALG uptake rose linearly for 5 min in both HEK293/OATP2B1 cells and mock cells (Supplementary Fig. 1). In the time dependence study, ALG uptake at the concentration of 0.1 mM was slightly but significantly higher in OATP2B1/HEK293 cells than in mock cells. However, uptake rates of ALG were not significantly different between OATP2B1-expressing HEK293 cells and mock cells over the concentration range of 0.1 to 10 mmol/L (Fig. 5).
The uptake study was conducted at 37 °C for 5 min. Values are mean ± S.D. (n = 3). Open and closed circles indicate OATP2B1 / HEK293 and mock cells, respectively.
In order to establish whether fruit juices affect the gastrointestinal absorption of ALG in vivo, the plasma concentration–time profiles of ALG after a single oral dose of 20 mg/kg ALG dissolved in 10 mL of water or AJ were compared in rats (Fig. 6).
Rats received a single oral dose of 20 mg/kg ALG dissolved in 10 mL/kg of water (open circle) or apple juice (closed circle). Values are mean ± standard error of the mean (S.E.M.) (n = 4). Asterisks indicate significant differences from the water-administered control (* p < 0.05).
AJ significantly reduced the AUC0–8 of ALG by 38% compared to ALG administered in water (12.57 ± 1.40 vs. 7.80 ± 0.75 µmol‧h/L; p < 0.05). However, Cmax was not changed significantly by AJ (Control vs. AJ, 2.90 ± 1.76 vs. 1.76 ± 0.31 µmol/L). The values of terminal half-life, estimated from the data at 5–8 h after administration, in the two groups were not significantly different (Control vs. AJ, 2.41 ± 0.94 vs. 1.39 ± 0.19 h, respectively).
Our results indicate that ALG uptake into Caco-2 cells is carrier-mediated, and that a fruit-juice-sensitive transporter other than OATP2B1 might play a role in ALG absorption in the human intestine. In addition, fruit juice appears to inhibit the uptake of ALG in vitro and its intestinal absorption in vivo. Thus, the possibility of drug–food interaction should be taken into consideration for patients receiving ALG.
ALG is a weak cationic compound that shows a high intestinal absorption rate (>70%)3) despite its low lipophilicity (basic product information of Nesina tablets, ver.12). We studied the intestinal absorption mechanism of ALG by using Caco-2 cells. The initial uptake rate of ALG into Caco-2 cells showed time, temperature, and bell-shaped pH dependences, and was inhibited by several inhibitors of active transport, suggesting the involvement of a carrier-mediated transport system (Fig. 1). In the evaluation of concentration dependence, saturation of the initial uptake rate of ALG could not be observed because measurement could not be performed at concentrations above 10 mM due to limited solubility. ALG uptake was not inhibited by substrates of OCT, OCTN, and PEPT1, whereas it was inhibited by organic anionic drugs (fexofenadine and methotrexate), fruit juices and their components, and the anion exchanger inhibitor DIDS, all of which are prototypical OATP substrates and inhibitors (Figs. 2, 3). Moreover, ALG transport into Caco-2 cells was pH-sensitive, and the optimum pH was near neutral (Fig. 4). These results are all consistent with the involvement of OATPs. OATPs are known to be pH-sensitive transporters that generally have the optimum pH in the weakly acidic region.9–12) However, their optimum pH is substrate-dependent.9–12) Bromosulfophthalein and tebipenem pivoxil showed higher uptakes at neutral or weakly alkaline pH than those at an acidic pH.11,12) Furthermore, the pH profile of ALG uptake was very similar to that of tebipenem pivoxil, a cationic OATP2B1 substrate.11) Among OATPs expressed in the intestine, OATP2B1 has the highest expression level, and is considered to play a major role in the intestinal absorption of amphiphilic drugs.9) Consequently, we next directly examined whether OATP2B1 is involved in ALG uptake by using HEK293/OATP2B1 cells, but we found that ALG uptake was not significantly different between HEK293/OATP2B1 cells and mock cells over the concentration range from 0.1 mM to 10 mM (Fig. 5). Thus, a fruit-juice-sensitive transporter other than OATP2B1 might play a role in the uptake of ALG in Caco-2 cells.
The most plausible candidate transporters for ALG uptake are intestinal OATPs other than OATP2B1. Among twelve OATP family members, OATP2A1, 2B1, 3A1, 4A1 are expressed in the small intestine at the mRNA or protein level.8,9,13) OATP1A2 is known to be a fruit-juice-sensitive fexofenadine transporter,14) but OATP1A2 expression is below the limit of detection in the intestine and in Caco-2 cells at both the mRNA and protein level.8,13) Little information is available on the roles of OATP2A1, OATP3A1 and OATP4A1 in drug transport in the intestine. However, OATP4A1 can probably be ruled out as a candidate for the ALG transporter, because OATP4A1 protein is below the limit of detection in Caco-2 cells.15) The OATP4C1 gene is expressed in the small intestine,16) and the DPP-IV inhibitor sitagliptin is a substrate of OATP4C1, but not OATP2B1.17) However, sitagliptin uptake into Caco-2 cells was not inhibited by ALG or by fruit juices (data not shown), so ALG and sitagliptin appear to be taken up by different transporters. The possible involvement of OATP2A1 or OATP3A1 in ALG uptake remains to be explored.
Although fruit juices and their constituents, which are known to inhibit OATP1A2 and OATP2B1 activities, markedly inhibited ALG uptake into Caco-2 cells (Fig. 3), their inhibitory potency towards other intestinal OATPs is unclear at present. GFJ and OJ contain naringin and hesperidin, respectively, at a concentration of 100 or more times the IC50 for OATP2B1.18) Further, a mixture of phloridin, phloretin, hesperidin and quercetin, components of AJ, has a strong inhibitory effect, although the major inhibitory component of AJ has not been identified. In addition, an in vitro experiment showed that hyperosmolarity corresponding to that of 50% fruit juice did not suppress the transport activity of OATP2B1. Therefore, we think it likely that the juice components (1 mmol/L) and 20% juice used in this study would have almost completely inhibited intestinal OATPs. Since fruit juices are usually taken without dilution, they might potentially inhibit OATPs in the human intestine. To test this idea in vivo, we conducted an absorption experiment in rats using AJ, which had the greatest inhibitory effect on ALG uptake. Indeed, AJ significantly decreased systemic exposure to ALG without changing the terminal half-life in rats, suggesting that it reduced the bioavailability of ALG, presumably due to inhibition of uptake via intestinal Oatp(s) (Fig. 6). However, given that we did not adjust the pH of drug solutions, not only the juice constituents, but also the acidic pH of AJ, which might not be fully neutralized in the intestine, might have affected the intestinal absorption of ALG via Oatps. Such food–drug interactions could directly affect the therapeutic efficacy of ALG in humans. However, since in vitro vs. in vivo and rat vs. human discrepancies have been reported for juice–drug interactions,19) the actual clinical significance of the ALG–juice interaction needs to be confirmed.
ALG can be used in combination with a wide variety of drugs. However, little information is available on drug–drug interactions during its intestinal absorption. In our study, fexofenadine and methotrexate, which are substrates of OATP1A2 and OATP2B1, reduced ALG uptake by 50% (p < 0.05) and 30% (not significant), respectively, in Caco-2 cells. However, it is unlikely that these drugs would greatly inhibit the intestinal absorption of ALG in clinical settings, since the inhibitor concentration used (1 mmol/L) is higher than the expected gastric concentration at the highest dosage taken with 200 mL of water (0.6 mmol/L for 60 mg fexofenadine and 0.02 mmol/L for 2 mg methotrexate). So far, there are no reports of clinical drug–drug interaction between ALG and OATP substrate drugs.20) But, even though no example of drug–drug interaction at intestinal OATPs is currently known, caution seems advisable when ALG is administered concomitantly with fruit-juice-sensitive drugs that might interact with the ALG transporter.
In conclusion, our results indicate that a carrier-mediated system contributes to ALG uptake in Caco-2 cells, and intestinal OATPs other than OATP2B1 are candidates for the ALG transporter. Further study is needed to identify the transporter that carries ALG into Caco-2 cells and enterocytes. ALG transport was inhibited not only by a variety of fruit juices, but also by fexofenadine and methotrexate. Therefore, careful consideration of the possibility of clinically significant drug interactions seems necessary in patients prescribed ALG.
We thank Dr. K. Watanabe and Dr. M. Kumagai for technical assistance.
The authors declare no conflict of interest.
The online version of this article contains supplementary materials.
