In Vitro Evaluation Of P‑gp‑Mediated Drug–Drug Interactions Using The RPTEC/TERT1 Human Renal Cell Model
Mar 01, 2022
Introduction In vitro evaluation of the P-glycoprotein (P-gp) inhibitory potential is an important issue during the drug development process, as it allows the prediction of clinically relevant drug–drug interactions (DDIs) [1–3]. P-gp belongs to the ATP-binding cassette (ABC) transporter superfamily and is encoded by the multidrug resistance gene MDR1 (also known as ABCB1). This membrane transporter is known to be overexpressed in tumoral cells and to cause resistance to many anticancer drugs [4]. Located within all physiological barriers, including intestine, liver, and kidneys, P-gp protects against xenobiotics by limiting the absorption of these substrates from the digestive tract and facilitating their efux into the bile and urine. Thus, P-gp plays a notable role in the pharmacokinetics of various therapeutic classes [5–7]. A plethora of drugs such as anticancer agents, antifungals, and cardiovascular drugs are known to be P-gp substrates and/or inhibitors, and many of them are involved in clinically relevant interactions [8–10]. Therefore, the US Food and Drug Administration (FDA) mandates that the P-gp inhibitory potential must be evaluated during the early stages of drug development. In vitro assays carried out for this purpose are commonly based on drug transport studies using P-gp-expressing cell lines. More precisely, the FDA guidelines recommend the determination of the in vitro halfmaximal inhibitory concentration (IC50) values to assess the risk of clinical DDIs resulting from P-gp inhibition. To this end, many experimental assays have been conducted, and most of them have focused on drug interactions related to intestinal absorption using the Caco-2 or MDCK-MDR1 models [11–13]
CISTANCHE WILL IMPROVE KIDNEY/RENAL DISEASE
Since renal elimination is also a common elimination route for various classes of drugs, active tubular secretion via ABC transporters may also play a key role in these interactions [14]. However, in vitro data on renal DDIs mediated by P-gp are limited. This is partly due to a lack of development and characterization of in vitro models for the prediction of renal drug transport. Among various human cell lines, the RPTEC/TERT1 model appears to show promise for the evaluation of renal drug interactions. This cell line is derived from a healthy human donor and has been generated from immortalized proximal tubule cells. Moreover, the expression and functionality of P-gp have been demonstrated using this model, thus confrming its ability to predict renal drug efuxes [15, 16]. In this context, the present study was designed to investigate P-gp inhibitory potential using the RPTEC/TERT1 model. First, a rhodamine 123 (R123) accumulation screening assay was performed to obtain an inhibition profle for each tested drug. Based on this screening, four drugs were then selected for assessment of their concentration-dependent efects on the intracellular accumulation of two P-gp substrate drugs: apixaban and rivaroxaban
Keywords: renal cell, renal model, renal drug, renal elimination, kidneys.
Materials and Methods
Reagents Apixaban, [2H7 1 3C ] - a p i xa b a n , r i va r oxa b a n ,[13C6]-rivaroxaban, nilotinib, crizotinib, erlotinib, axitinib, idelalisib, warfarin, and dabigatran etexilate were purchased from Alsachim (Illkirch, France). Verapamil, ketoconazole, simvastatin, amiodarone, rhodamine 123, Hank’s balanced salt solution (HBSS), and HEPES solution were purchased from Sigma–Aldrich (Saint-Quentin-Fallavier, France).
Cell Culture RPTEC/TERT1 cells were obtained from the American Type Culture Collection (ATCC, Molsheim, France) and cultured in hormonally defned, serum-free medium consisting of Dulbecco’s Modifed Eagle’s Medium F12 (ATCC) supplemented with a growth factor kit (ATCC) and a 1% antibiotic/antimycotic mixture (penicillin–streptomycin, amphotericin B) at 37 °C and 5% CO 2. For all the experiments, cells were seeded in 96-well plates at a density of 50,000 cells per well and were used to grow after 14 days of culture. The medium was renewed every 2 days and cells were used from passage 27 to passage 34. Caco-2 cells were also purchased from the ATCC. The cells were maintained in a culture medium consisting of Eagle’s Minimal Essential Medium (Sigma–Aldrich, Missouri, USA) supplemented with 10% FBS, 1% nonessential amino acids, and a 1% antibiotic/antimycotic mixture (penicillin–streptomycin, amphotericin B) at 37 °C and 5% CO2. Caco-2 cells were seeded in 96-well plates at a density of 5 × 10 3 cells per well and were used to grow after 14 days of culture. Human proximal tubular epithelial cells (HPTECs) were obtained from BIOPREDIC (Rennes, France) and cultured in DMEM/F12 supplemented with hydrocortisone, EGF, insulin, transferrin, and sodium selenite. Cells were seeded in 96-well collagencoated culture plates at a density of 6600 cells per well and were used to grow after 11 days of culture.
Rhodamine 123 Accumulation Screening Assay The P-gp inhibitory potential was determined by measuring the intracellular accumulation of rhodamine 123 in RPTEC/TERT1 cells in the presence and absence of various classes of drugs. Among the 14 tested drugs, cyclosporin A (10 µM), ketoconazole (50 µM), verapamil (100 µM), and amiodarone (50 µM) were used as P-gp inhibitors. Apixaban, rivaroxaban, and dabigatran etexilate—three direct oral anticoagulants (DOACs)—were used as P-gp substrates (10 µM). Warfarin (50 µM) was used as a non-inhibitor, and fve anticancer drugs including nilotinib, crizotinib, erlotinib, and idelalisib were used without knowing their inhibition profles at a concentration of 10 µM. Several conditions were reproduced with the Caco-2 cells, which are approved by the FDA for drug transport studies. In this way, the intracellular retention of rhodamine 123 was determined in Caco-2 cells in the presence and absence of verapamil (100 µM), cyclosporin A (10 µM), nilotinib (10 µM), and DOACs (10 µM). Briefy, after 14 days of culture in 96-wellplates, the cells were pre-incubated for 10 min at 37 °C with each drug dissolved in HBSS with 10 mM HEPES (v/v). Then the cells were incubated with 10 µM rhodamine 123 for 45 min at 37 °C. Finally, after three washes in cold HBSS/ HEPES solution, the cells were lysed at room temperature for 45 min in a solution of sodium dodecyl sulfate (SDS) containing 1% sodium borate. The amount of intracellularrhodamine 123 was quantifed using an Infnite M nanospectrofuorometer (Life Sciences, TECAN, Switzerland), setting the wavelengths to 485/535 nm. Data were expressed as percentages of the rhodamine 123 accumulation in control cells not exposed to any potential P-gp inhibitors and arbitrarily set at 100% accumulation.
DOAC Intracellular Accumulation Assay and IC50 Determination From the previous rhodamine 123 screening assay, four drugs were chosen to investigate their concentration-dependent efects on the intracellular accumulation of apixaban and rivaroxaban (10 µM) in RPTEC/TERT1 cells. Ketoconazole, crizotinib, and nilotinib were selected as P-gp inhibitors, and warfarin was chosen as a non-inhibitor. Cyclosporin A (10 µM) was used as a broad-spectrum inhibitor of transporters. Studies of the interactions between DOACs and nilotinib for determining IC50 values were reproduced in Caco-2 cells in order to compare the two cellular models. All compounds were diluted either alone or with the associated inhibitor in HBSS transport bufer supplemented with 1% HEPES (v/v) and 1% DMSO (v/v). Prior to incubation, all solutions were pre-warmed to 37 °C and the pH was adjusted to 7.4. For the anticancer drugs crizotinib and nilotinib, due to their poor solubility and cytotoxic potential, concentrations ranged from 0.1 to 25 µM. For ketoconazole and warfarin, concentrations ranged from 0.1 to 100 µM. Briefy, after 14 days of culture, all drugs dissolved in HBSS with 1% HEPES (v/v) were pre-incubated for 10 min at 37 °C. Then, cells were incubated with 10 µM of apixaban or rivaroxaban for 60 min at 37 °C. Finally, after three washes in cold HBSS/HEPES solution, cells were lysed at room temperature for 45 min in a 0.2% solution of Triton X-100. The amount of intracellular DOAC was then quantifed by liquid chromatography–mass spectrometry (LC-MS). Data were expressed as percentage increases in DOAC accumulation, and the DOAC intracellular accumulation was arbitrarily set at 100% in control cells. The IC50 values for inhibition of P-gp activity, which correspond to half-maximal efective concentration (EC50) values for increasing DOAC accumulation, were determined from unweighted nonlinear least squares regression modeling of increased accumulation with the 'nls()'function in R software, according to the following equation (Eq. 1):
Liquid Chromatography–Mass Spectrometry Analysis Quantifcation of apixaban (m/z 460.19793) and rivaroxaban (m/z 436.07285) was performed using an Ultimate U3000 liquid chromatography system (Dionex, Sunnyvale, CA, USA) coupled with a Q-Exactive Plus mass spectrometer (ThermoFisher, Bremen, Germany). LC separations were achieved using a Hypersil Gold C18 (3 µm, 50 × 2.1 mm) analytical column (ThermoFisher Scientifc, Waltham, MA, USA) and a fow rate of 0.6 mL/min. Mobile phase A was water with 0.1% formic acid (FA) and mobile phase B was acetonitrile with 0.1% FA. For rivaroxaban, the gradient was: 0–0.3 min, 10% B; 0.3–1 min, linear from 10% to 70% B; 1–1.5 min, 70% B; 1.51 min, return to starting conditions until 3 min. For apixaban, the gradient was: 0–0.3 min, 10% B; 0.3–0.7 min, linear from 10 to 90% B; 0.7–1.5, 90% B; 1.51 min, return to starting conditions until 3 min. Detection was performed in electrospray-positive parallel reaction monitoring (PRM) mode at a resolution of 35,000 (at m/z 200). The internal standards (ISs) were [13C,2H7]-apixaban (m/z 468.2452) for apixaban and [ 13C6]-rivaroxaban (m/z 442.09297) for rivaroxaban. For each drug and its respective IS, a target ion (for quantifcation) and a confrming ion were monitored. For rivaroxaban and its IS, the target ion was m/z 144.95125 and the confrming ions were m/z 231.11280 and m/z 237.13298, respectively. For apixaban and its IS, the target ion was m/z 199.08656 and the confrming ions were m/z 282.12387 and m/z 241.06062, respectively.
Results
Rhodamine 123 Accumulation Screening Assay The rhodamine 123 accumulation screening assay was performed in RPTEC/TERT1 cells with 14 drugs with diferent inhibition profles (Fig. 1). As expected, the intracellular accumulation of rhodamine 123 in the presence of cyclosporin A, a broad-spectrum inhibitor, was among the highest, with an increase in accumulation of 75% compared to the control cells. The presence of verapamil, a specifc P-gp inhibitor, led to an increase in accumulation of rhodamine 123 of 57%, demonstrating the involvement of P-gp. In the same way, ketoconazole, described as a strong P-gp inhibitor, led to a greater increase (66%) in the intracellular accumulation of rhodamine 123 compared to verapamil, confrming its inhibition profle. Conversely, the addition of amiodarone, characterized as a moderate P-gp inhibitor, caused an increase in accumulation of 59%, which is similar to that caused by verapamil. Diferent inhibition profles were observed for the anticancer agents nilotinib, axitinib, crizotinib, erlotinib, and idelalisib. The addition of nilotinib caused a strong increase in the accumulation of rhodamine 123 (71%), suggesting signifcant inhibitory potential, followed by axitinib, which caused an increase in retention of 57%. On the other hand, crizotinib and erlotinib demonstrated moderate to low inhibitory potentials, with increases in rhodamine 123 accumulation of 23% and 16%, respectively. Idelalisib did not afect the accumulation of rhodamine 123; the same was true of warfarin, which was
chosen as a non-inhibitor. Among the three DOACs, apixaban and rivaroxaban caused slight increases in the retention of rhodamine 123: 33% and 12%, respectively. Interestingly, dabigatran etexilate, a prodrug of dabigatran and a known substrate of P-gp, caused an increase in the retention of rhodamine 123 of about 50%, although it is not described as a potential P-gp inhibitor. Finally, the addition of simvastatin caused only a slight increase in the accumulation of the fuorescent substrate (32%). Based on this screening, four drugs were selected to assess their concentration-dependent efects on the intracellular accumulation of apixaban and rivaroxaban within the RPTEC/TERT1 model. Ketoconazole was chosen as a positive control condition for P-gp inhibition, and warfarin was selected as a non-inhibitor of P-gp for the negative control condition. Nilotinib and crizotinib were selected as strong and moderate P-gp inhibitors, respectively. Several conditions were reproduced with the reference model, Caco-2. The intracellular accumulation of R123 in cells was determined after combination (or not) with apixaban and rivaroxaban (10 µM), nilotinib (10 µM), and with cyclosporin A (10 µM) and verapamil (100 µM) for the control conditions for inhibition (Fig. 2A). As expected, verapamil and cyclosporin A caused high increases in R123 retention of 297% and 275%, respectively. In the same way, nilotinib caused a high increase in the intracellular accumulation of R123 (229%), confrming its inhibitory potential. Combining R123 with DOACs resulted in minor increases in the accumulation of R123 (40–44%) compared with the other drugs. Moreover, the intracellular accumulation of rhodamine 123 in the absence and presence of cyclosporin A in renal primary human cells was also investigated in this study to check the reliability of the results obtained with RPTEC/TERT1 cells (Fig. 2B). These analyses showed that the presence of cyclosporin A increased the R123 retention by 71%. This value was close to that obtained under the same conditions with RPTEC/TERT1 cells (an increase of 75% with cyclosporin A). Therefore, the P-glycoprotein activity was similar in the RPTEC/TERT1 model and primary human cells.
DOACs Intracellular Accumulation Assay: IC50 and I1/IC50 Ratio Determination Intracellular accumulation studies were performed to assess the P-gp-mediated transport of apixaban and rivaroxaban at a constant concentration of 10 µM in RPTEC/TERT1 cells in vitro and in the presence of increasing concentrations of nilotinib, crizotinib, ketoconazole, and warfarin (Figs. 2, 3). The modeling of increased apixaban and rivaroxaban retention was assessed to determine IC50 values. Combining DOACs with nilotinib led to the lowest IC50 values: 0.85 µM and 1.37 µM for rivaroxaban and apixaban, respectively (Figs. 4, 5). Combination with crizotinib yielded IC50 values of 10.1 µM and 12.2 µM for rivaroxaban and apixaban, respectively. Surprisingly, combining DOACs with ketoconazole did not produce the lowest IC50 values (16.5 µM and 16.9 µM for rivaroxaban and apixaban, respectively). Finally, as expected, combination with warfarin, which was
chosen as a non-inhibitor, did not afect the intracellular retentions of the two DOACs. The intracellular accumulation of both apixaban and rivaroxaban within Caco-2 cells in the presence of increasing concentrations of nilotinib was therefore investigated for comparison with the cellular models. Interestingly, as observed with the RPTEC/TERT1 model, the IC50 value for rivaroxaban (4.16 µM) was lower than that obtained for apixaban (9.35 µM). It is also interesting to note that the IC50 values observed in Caco-2 cells were higher than those obtained in RPTEC/TERT1 cells for the same inhibitor. The clinical relevance of DDIs can be predicted from in vitro data. According to the FDA guidelines, [I1]/IC50 ratios were calculated for each combination of drugs. This ratio allows in vivo concentrations to be compared to those known to produce a relevant efect in vitro. For apixaban in RPTEC/TERT1 cells, the ratios were 3.1, 0.06, and 17.4 with nilotinib, crizotinib, and ketoconazole, respectively (Table 1). In Caco-2 cells, the [I1]/IC50 ratio was 0.46 for the interaction between apixaban and nilotinib (Table 1). For rivaroxaban, the ratios were slightly higher than those obtained for apixaban in RPTEC/TERT1 cells: 5.1, 0.08, and 17.7, respectively, for nilotinib, crizotinib, and ketoconazole (Table 2). In the same way, the [I1]/IC50 ratio observed for the interaction between rivaroxaban and nilotinib in Caco-2 cells was 1.03, which was higher than that obtained for apixaban (Table 2).
Discussion
Numerous studies carried out in recent decades have reported a notable role of P-gp in drug pharmacokinetics [17–19]. In view of the increasing regulatory interest in drug interactions mediated by P-gp, in vitro assays to investigate the inhibitory potential of drugs are an importantaspect of drug development and clinical practice. To this end, many in vitro assays have been conducted, and most of the associated data on predicted intestinal absorption have been generated from the Caco-2 and MDCK-MDR1 cell lines [20–22]. However, few data are available on the active tubular secretion of drugs, which also plays a key role in drug disposition and depends on the presence of ABC transporters. This observation is clearly linked to the lack of characterized renal cell lines for predicting drug renal efux. This goal requires an in vitro renal model that closely mimics the physiological barrier. The human cell line RPTEC/ TERT1, which expresses several ABC transporters (particularly P-gp), appears to be a good alternative, as demonstrated in a previous study [16]. In this context, the present work investigated the application of the RPTEC/TERT1 model to assess the P-gp inhibitory potential. To the best of our knowledge, this work is the frst to provide data from P-gp inhibition studies using human renal cells.
In vitro assays for determining the P-gp inhibitory potential are commonly based on the use of a specifc reference P-gp substrate, such as digoxin or rhodamine 123 [23–25]. Due to their ease of use, fuorescent probes are advantageous for high-throughput assays. Moreover, rhodamine 123 has been widely applied to the detection of P-gp activity in a wide range of studies [26–28]. In this way, assays of rhodamine 123 accumulation in RPTEC/TERT1 cells were performed in this study to identify the inhibition profles of 14 drugs. Among these drugs, cyclosporin A, ketoconazole, and verapamil were chosen as strong P-gp inhibitors. These inhibitors have been extensively characterized for their signifcant P-gp inhibitory potential, which leads to the modulation of both digoxin and rhodamine 123 transport [27, 29, 30]. As expected, these drugs caused the highest rhodamine 123 retentions in RPTEC/TERT1 cells. In contrast, no increase in the retention of R123 was observed with warfarin, which was used as a negative control, confrming the reliability of the RPTEC/TERT1 model. Interestingly, among all the tested drugs, DOACs—including dabigatran etexilate, apixaban, and rivaroxaban—caused distinct increases in the retention of R123, even though they were not previously characterized as inhibitors, only P-gp substrates [31, 32]. This observation may be due to the existence of so-called “competitive” inhibition, where the drugs can interact with the same binding sites on P-gp. The most well-characterized sites for P-gp are the H site (for binding Hoechst 33342) and the R site (for binding rhodamine 123). However, multiple other unknown drug-binding sites could play a role in these interactions, which could explain the diferent efects of the DOACs on the accumulation of R123 [33, 34]. The P-gp inhibition observed for a given drug is therefore dependent on the substrate used during in vitro studies [28, 35]. The use of diferent P-gp substrates that interact with diferent drug-binding sites should therefore be considered to accurately describe the putative P-gp inhibitory potencies of drugs. It is also interesting to note that several conditions from the rhodamine 123 screening were also performed in Caco-2 cells in order to compare the RPTEC/ TERT1 cells with a reference model in drug transport studies. As expected, cyclosporin A, verapamil, and nilotinib caused the highest increases in rhodamine retention. The same inhibition profles were observed with both Caco-2 and RPTEC/TERT1 cells. However, the increases in rhodamine 123 retention generated by the interactions in the Caco-2 model were greater than those observed in RPTEC/TERT1 cells, suggesting a potential diference in the expression of P-glycoprotein between models Therefore, in the present study, a second method was used to assess and confrm the potential inhibition of P-gp by drugs.
CISTANCHE WILL IMPROVE KIDNEY/RENAL DIALYSIS
Based on the rhodamine 123 accumulation assay, four drugs were chosen to determine their concentration dependent efects on the intracellular accumulation of apixaban and rivaroxaban. P-gp has been shown to play a main role in the efux of apixaban and rivaroxaban [32, 36]. The IC50 values of nilotinib, crizotinib, and ketoconazole were therefore assessed, with DOACs selected as the “victim” drugs. To the best of our knowledge, this study is the frst to perform intracellular accumulation studies with DOACs in human renal cells. The IC50 values obtained for nilotinib and crizotinib were in accordance with their inhibition profles obtained from the rhodamine 123 accumulation assay. Nilotinib, which caused a strong increase in the retention of R123, presented the lowest IC50 value for the two DOACs, confrming its high inhibition potential. This result is also in accordance with a previous study that showed a concentration-dependent increase in the intracellular accumulation of [3 H]-paclitaxel in MDR1-transfected cells, suggesting that it has an inhibitor profle [37]. For crizotinib, the R123 assay showed moderate inhibitory potential, with less of an increase in R123 retention than that caused by nilotinib. This observation is supported by a previous study where crizotinib increased the intracellular accumulation of R123 and doxorubicin in MDR1-transfected cells [38]. This result is also in accordance with the IC50 values determined with DOACs, which were higher than the corresponding values for nilotinib, confrming its moderate inhibition profle. However, it is interesting to note that a concentration of 10 µM of either nilotinib or crizotinib did not cause the same increases in the retention of rhodamine 123 and DOACs. In rhodamine screening, nilotinib increased the retention of the substrate by about 71%, whereas it was about 40% for DOACs. In contrast, crizotinib caused a small increase in the retention of rhodamine (23%) but greater increases in the retention of DOACs (about 50%) at the same concentration. As discussed previously, this observation could be explained by diferences in the binding sites on P-gp. It is known that many drugs interact and compete with the H binding site rather than the R binding site (which is where rhodamine 123 binds), and vice versa [39]. Interestingly, ketoconazole, which is known to be a strong P-gp inhibitor, caused a slightly smaller increase in the retention of rhodamine 123 than nilotinib did (66% versus 71%, respectively). Nevertheless, a similar value was observed with Caco-2 cells, where the presence of ketoconazole caused an increase of 60% in the accumulation of R123 [40]. The intracellular accumulation of DOACs for determining IC50 values also showed higher values compared to nilotinib and crizotinib. Interestingly, most of the IC50 values found in the LLC-PK1 or Caco-2 models using digoxin as a P-gp substrate were between 3 and 4 µM for ketoconazole [41, 42]. These values are quite diferent from those found in the present study (16.5 µM and 16.9 µM with apixaban and rivaroxaban, respectively). This observation confrms that the choices of the substrate and the model used are crucial infuences when determining the P-gp inhibitory potential. Additionally, a recent study reported that the expression levels of ABC transporters in in-vitro cellular models have an impact on drug transport assays and therefore on the evaluation of DDIs relating to P-gp [43]. Indeed, the determination of IC50 values for verapamil using rivaroxaban as a P-gp substrate revealed a heterogeneity between the MDCKMDR1 and Caco-2 cell models, with IC50 values of 6.94 µM and 21.2 µM, respectively [43]. Furthermore, the concentration-dependent efect of nilotinib on the intracellular accumulation of apixaban and rivaroxaban was also investigated in Caco-2 cells in this study. Interestingly, IC50 values for nilotinib were signifcantly higher in Caco-2 cells than in RPTEC/TERT1 cells. This observation can be explained by the diference in P-gp expression between these models. In addition, it is known that the P-gp distribution depends on the tissue considered. Fallon et al. demonstrated that the level of P-gp was higher in kidney than in liver tissue in humans [45]. On the other hand, the expression level of P-gp seems to be higher in intestine than in kidney tissue [46]. The use of human cells such as the RPTEC/TERT1 model, which do not overexpress transporters, could therefore provide additional data. These data could be used and imputed in physiologically based pharmacokinetic modeling (PBPK) by integrating several parameters, such as the quantity of P-gp in a tissue or in vitro model or IC50 values.
Although the IC50 values found for ketoconazole in the RPTEC/TERT1 model were higher than those observed in the literature, the [I1]/IC50 ratio—validated as a predictor of potential clinically relevant DDIs for orally administered drugs—showed high values that were above the threshold of 0.1 defned by the FDA. This result is in accordance with a clinical study that showed a twofold increase in apixaban exposure with co-administration of ketoconazole [44]. Taken together, all of these results demonstrate that the RPTEC/ TERT1 model is a promising tool for assessing P-gp inhibitory potential.
CISTANCHE WILL IMPROVE KIDNEY/RENAL PAIN
Conclusion
Our study demonstrated that the application of the RPTEC/TERT1 model is convenient for evaluating the P-gp inhibitory potentials of various classes of drugs. Rhodamine 123 accumulation assays allowed an initial drug screening to be performed. However, the P-gp inhibitory potentials of drugs that do not interact with the R site of P-gp must also be investigated using additional substrates to confrm the predictions. The IC50 values determined from the intracellular accumulation of apixaban and rivaroxaban are in accordance with the inhibition profles observed with rhodamine 123. Moreover, the use of ketoconazole and warfarin as a strong inhibitor and a non-inhibitor of P-gp, respectively, confrmed the reliability of the RPTEC/TER1 model when it is used to obtain in vitro data about the inhibitory potentials of drugstowards P-gp. Finally, a comparison of the results obtained using the RPTEC/TERT1 model with those obtained using Caco-2 cells highlighted the importance of conducting in vitro studies in diferent cellular models to confrm the inhibitory profle for a given drug.
