ABSTRACT
P-glycoprotein (ABCB1), an ATP-binding-cassette efflux transporter, limits intestinal absorption of its substrates and is a common site of drug-drug interactions (DDIs). ABCB1 has been suggested to interact with many antivirals used to treat HIV and/or chronic hepatitis C virus (HCV) infections. Using bidirectional transport experiments in Caco-2 cells and a recently established ex vivo model of accumulation in precision-cut intestinal slices (PCIS) prepared from rat ileum or human jejunum, we evaluated the potential of anti-HIV and anti-HCV antivirals to inhibit intestinal ABCB1. Lopinavir, ritonavir, saquinavir, atazanavir, maraviroc, ledipasvir, and daclatasvir inhibited the efflux of a model ABCB1 substrate, rhodamine 123 (RHD123), in Caco-2 cells and rat-derived PCIS. Lopinavir, ritonavir, saquinavir, and atazanavir also significantly inhibited RHD123 efflux in human-derived PCIS, while possible interindividual variability was observed in the inhibition of intestinal ABCB1 by maraviroc, ledipasvir, and daclatasvir. Abacavir, zidovudine, tenofovir disoproxil fumarate, etravirine, and rilpivirine did not inhibit intestinal ABCB1. In conclusion, using recently established ex vivo methods for measuring drug accumulation in rat- and human-derived PCIS, we have demonstrated that some antivirals have a high potential for DDIs on intestinal ABCB1. Our data help clarify the molecular mechanisms responsible for reported increases in the bioavailability of ABCB1 substrates, including antivirals and drugs prescribed to treat comorbidity. These results could help guide the selection of combination pharmacotherapies and/or suitable dosing schemes for patients infected with HIV and/or HCV.
INTRODUCTION
HIV and hepatitis C virus (HCV) infections are a serious global health issue. It has been estimated that over 36 million people worldwide are HIV positive, with approximately 1 million of them dying every year (1). The total number of HCV-infected patients exceeds 80 million; HCV infection is directly associated with approximately 400,000 deaths annually (2). In developed countries such as the United States, the mortality due to HCV currently exceeds that due to HIV (3, 4).
Nucleoside reverse transcriptase inhibitors (NRTIs), nonnucleoside reverse transcriptase inhibitors (NNRTIs), protease inhibitors (PIs), and C-C chemokine receptor type 5 (CCR5) antagonists are among the most widely used anti-HIV drugs. Recently approved pharmacotherapies for treating chronic HCV infections include directly acting inhibitors of HCV proteins controlling RNA replication, NS5A and the RNA-dependent RNA polymerase NS5B (5, 6). Anti-HIV and/or anti-HCV medication is always administered in regimens that combine antivirals with agents having different mechanisms of action (2, 7–9).
Up to 10% of HIV-infected patients refuse to continue antiretroviral therapy because of its side effects (8, 10–12). Intolerability is frequently due to drug-drug interactions (DDIs), which increase the plasma concentrations of antivirals or coadministered medications (8, 9, 13–20). The molecular mechanisms responsible for this increased bioavailability are often unclear.
Antivirals are administered orally and may therefore be subject to DDIs on membrane drug transporters in the intestinal barrier. P-glycoprotein (ABCB1) is an abundantly expressed drug transporter that controls cellular efflux to the intestinal lumen and thus governs the net intestinal uptake of its substrates (21–23). The International Transporter Consortium and the Food and Drug Administration have therefore both suggested that DDIs on ABCB1 may be clinically important (21, 24, 25). Because many antivirals are substrates and/or inhibitors of ABCB1 (15, 26–29), we hypothesized that they may induce or be affected by DDIs on intestinal ABCB1.
DDIs on the ABCB1 transporter cannot be rigorously evaluated directly in animals or humans because it is difficult to assign an observed DDI to a specific transporter (24). Therefore, experimental studies on DDIs involving antivirals on ABCB1 transporters have generally been based on in vitro observations (21, 30–38); there are few relevant data originating from more complex experimental setups in animals or human tissue (39, 40). This deficiency could be addressed using new techniques such as our recently developed ex vivo method for monitoring drug accumulation in rat- and human-derived precision-cut intestinal slices (PCIS). This method offers a testing throughput comparable to that achieved with the Caco-2 cell line, an in vitro model recommended by the Food and Drug Administration and European Medicines Agency (41–43). Importantly, PCIS retain the integrity of the intact fresh tissue as well as the physiological intestinal architecture and native enzymatic activity levels (42–44).
Here, we report the use of in vitro models of bidirectional transport across Caco-2 monolayers and ex vivo accumulation studies in rat- and human-derived PCIS to assess the inhibition of intestinal ABCB1-controlled efflux of rhodamine 123 (RHD123) by selected drugs, including NRTIs (abacavir, zidovudine, and tenofovir disoproxil fumarate [TDF]), NNRTIs (rilpivirine and etravirine), PIs (atazanavir, lopinavir, ritonavir, and saquinavir), CCR5 antagonists (maraviroc), NS5A inhibitors (daclatasvir and ledipasvir), and NS5B inhibitors (sofosbuvir).
RESULTS
Inhibitory effect of antiviral drugs on RHD123 transport in vitro.We initially verified the functional expression of ABCB1 in the Caco-2 cells by testing their capacity for bidirectional RHD123 (1 μM) transport in the presence and absence of CP100356 (2 μM). The rPapp values obtained in these experiments were 3.20 ± 0.32 and 1.02 ± 0.27, respectively, confirming the transporter’s presence and activity. Lopinavir and ledipasvir were the only antivirals that inhibited RHD123 efflux at the lowest tested concentration (5 μM). Ritonavir, saquinavir, atazanavir, and daclatasvir inhibited RHD123 efflux at a concentration of 20 μM. Maraviroc exhibited an inhibitory effect only at the highest tested concentration (100 μM). Abacavir (100 μM), zidovudine (100 μM), TDF (100 μM), rilpivirine (20 μM), etravirine (20 μM), and sofosbuvir (100 μM) did not affect RHD123 transport significantly (Table 1). We then compared the levels of ABCB1 inhibition induced by selected antivirals at the concentrations that caused the strongest inhibitory effects on bidirectional RHD123 transfer. This revealed that lopinavir (5 μM), ritonavir (20 μM), saquinavir (20 μM), and atazanavir (50 μM) inhibited ABCB1 more strongly than maraviroc (100 μM). Additionally, lopinavir (5 μM) and ritonavir (20 μM) were stronger inhibitors than daclatasvir (20 μM) (Fig. 1).
Effect of CP100356 and selected antiviral drugs on the transport ratio of RHD123 (1 μM) across Caco-2 cells
Inhibition caused by antivirals tested at the concentrations causing the most significant effect on bidirectional transport of RHD123 (1 μM) across a monolayer of Caco-2 cells. (A) Calculated values of rPapp for RHD123 transport in the absence of inhibitors (dotted line) and in the presence (columns) of antivirals. (B) Multiple comparison of the inhibitory effects of antivirals. Data are presented as means ± SD (n = 3). Statistical analysis was performed using Student’s two-tailed t test (n = 3): n.s., not significant; *, P < 0.05; **, P < 0.01.
Inhibitory effects of antiviral drugs on RHD123 efflux in PCIS from rat ileum.We next tested the antivirals’ inhibitory effects on intestinal ABCB1 efflux in rat-derived PCIS. Treatment with CP100356 (2 μM) resulted in a 2.31-fold increase in RHD123 (10 μM) tissue accumulation, in accordance with earlier reports (43). This confirmed the functional expression of ABCB1 in the rat-derived PCIS. At concentrations of 5 μM, none of the antivirals inhibited ABCB1-controlled RHD123 efflux. Lopinavir, ritonavir, saquinavir, maraviroc, ledipasvir, and daclatasvir inhibit this efflux at concentrations of 20 μM, increasing the accumulation of RHD123 in the tissue by factors ranging from 1.37 ± 0.07 to 1.82 ± 0.26. Higher concentrations of ritonavir (50 and 100 μM) and ledipasvir (50 μM) resulted in more profound inhibition of RHD123 efflux, but higher concentrations of maraviroc (50 and 100 μM) did not. Atazanavir inhibited RHD123 efflux only at the highest tested concentration (50 μM) (Table 2). In keeping with the in vitro results, abacavir (100 μM), zidovudine (100 μM), TDF (100 μM), rilpivirine (20 μM), etravirine (20 μM), and sofosbuvir (100 μM) did not significantly inhibit RHD123 efflux in the rat PCIS (Table 2). We then compared inhibitory potency of the antivirals at the concentrations that caused the most profound inhibitory effect ex vivo. Only lopinavir (20 μM), ritonavir (100 μM), and daclatasvir (20 μM) inhibited rat intestinal ABCB1 more strongly than ledipasvir (50 μM) (Fig. 2).
Effect of the model inhibitor and selected antiviral drugs on accumulation of RHD123 (10 μM) over 2 h in the rat PCIS
Inhibition of ABCB1-controled RHD123 efflux caused by antivirals tested at the concentrations causing the most significant increase in RHD123 (10 μM) accumulation over 2 h in rat PCIS. (A) Tissue RHD123 concentrations in the absence of inhibitors (dotted line) and in the presence of antivirals (columns). (B) Multiple comparison of inhibitory effects of antivirals. Data are presented as medians with interquartile ranges (n = 3). Statistical analysis was performed using the nonparametric Mann-Whitney test: n.s., not significant; *, P < 0.05.
Inhibitory effect of antiviral drugs on RHD123 efflux using PCIS from human intestine.Ex vivo experiments were performed using PCIS prepared from the jejunum of five donors to evaluate the antivirals’ inhibitory effects in a clinically relevant model. Antivirals were tested at the highest concentrations used in rat-derived PCIS. Lopinavir (50 μM), ritonavir (100) μM, saquinavir (20 μM), and atazanavir (50 μM) significantly inhibited RHD123 (10 μM) efflux in all of the intestinal samples. Daclatasvir (20 μM) inhibited RHD123 (10 μM) efflux in PCIS prepared from donors 3 and 5 only, while maraviroc (100 μM) and ledipasvir (50 μM) inhibited efflux in PCIS from donor 3 only (Table 3). It thus appears that there is some interindividual variability in the effects of these antivirals on human intestinal ABCB1.
Effect of the model inhibitor and selected antiviral drugs on the accumulation of RHD123 (10 μM) over 2 h in the human PCIS
Viability of rat and human PCIS exposed to the tested antivirals.Because the validity of accumulation studies using PCIS depends strongly on the tissue’s viability over the course of the experiment, we quantified the ATP content of the tissue samples, which is a stable and verified marker of the preservation of vital cellular processes (42, 43). The detected ATP concentrations in the rat PCIS (n = 3) were approximately 4 nmol · mg−1, while those in the human PCIS collected from five donors included in the study were between 3 and 4 nmol · mg−1. These values are consistent with those reported previously (42–44), and there were no statistically significant differences in the ATP contents of the tested samples (Fig. 3). It thus appears that the model inhibitor and antivirals did not affect the viability of the PCIS even at the highest tested concentrations.
ATP content of the rat ileum (n = 3) (A) and human jejunum (n = 5) (B) after 2.5-h incubation in the presence of the studied antivirals at the highest concentrations used. Data are presented as medians with interquartile ranges. Statistical significance was assessed by nonparametric Kruskal-Wallis analysis followed by Dunn’s test. No statistically significant differences were found. TDF, tenofovir disoproxil fumarate.
DISCUSSION
HIV/HCV coinfection, age, the prescribed antiretroviral regimen, and polypharmacy are independent risk factors for DDIs (45, 46). It is therefore unsurprising that the prevalence of clinically relevant DDIs ranges from 14 to 41% (9, 45, 47). ABCB1 is a crucial determinant of intestinal absorption (22–24), but DDIs on this transporter may remain undiscovered during preclinical research and clinical trials (45). Consequently, there is great interest in methods for detecting ABCB1-mediated DDIs whose results can be extrapolated to the clinical environment; such methods are needed for both drugs in development and those that have already been launched commercially.
Rat and human PCIS have recently been put forward as reliable and efficient ex vivo models for studying the activity of transporters localized in enterocytes and for assessing the inhibitory potential of drugs (42–44). PCIS-based methods thus offer a superior capability for DDI detection while rivaling the throughput of in vitro cell-based models (42–44, 48–50).
In this work, we evaluated the ability of antivirals to inhibit intestinal ABCB1 by analyzing bidirectional transport of RHD123 across Caco-2 cells and measuring drug accumulation in rat- and human-derived PCIS (42–44). We selected antiviral drugs with different mechanisms of action and tested their inhibitory potential at concentrations up to the limit imposed by solubility or the concentration of maximal inhibitory effect. However, the tested concentrations of the antiviral drugs were lower than those likely to be achieved in the intestine following oral administration (51, 52). As a model ABCB1 substrate, we used RHD123 (42, 43), which is suitable for accumulation studies because it is easily detected and its efflux from enterocytes is not affected by metabolism (38, 53, 54). Our experimental designs of accumulation in PCIS were based on previously reported RHD123 concentrations (42, 43).
ABCB1 seems to be the transporter with the dominant influence on the pharmacokinetics of RHD123 (38, 54, 55). Other transporters such as the organic anion transporter peptide 1A2 and organic cation transporter 1 also reportedly contribute to RHD123 membrane transfer (56). However, their activity is unlikely to have influenced our results because the anion transporter peptide 1A2 exhibits little or no functional expression in the intestine (56) and none of the tested antivirals inhibits organic cation transporter 1 (26). Importantly, another crucial intestinal ABC transporter, breast cancer resistance protein, is not involved in RHD123 efflux (57).
Quenching effects on RHD123 fluorescence have been observed for atazanavir, lopinavir, saquinavir, and abacavir (38). Surprisingly, we observed significant increases in fluorescence in Hanks’ balanced salt solution (HBSS) buffer containing RHD123 (1 μM) upon adding 50 μM atazanavir (9%), 20 μM saquinavir (12%), 100 μM TDF (14%), 100 μM maraviroc (14%), or 100 μM sofosbuvir (15%). Only ledipasvir decreased the fluorescence (by 13% [data not shown]). In the bidirectional transport experiments, the same concentrations of the antiviral drugs were applied to the apical and basolateral compartments, and the concentrations of RHD123 in the receiving compartment were determined using a calibration curve based on reference solutions of the drugs. In RHD123 solutions containing an antiviral drug, the drug’s concentration had no detectable effect on the measured fluorescence. Therefore, the fluorescence measurements should not be significantly affected by potential changes in the drug concentrations of the apical and basolateral compartments due to ABC transporter activity (27, 32, 58, 59). In contrast to the results obtained in HBSS buffer, the fluorescence of RHD123 solutions in acetonitrile was independent of the presence of antivirals (data not shown). Therefore, a simple six-point calibration curve based on reference solutions without added antiviral drugs was used to determine RHD123 concentrations in PCIS-based experiments.
Using bidirectional RHD123 (1 μM) transport across Caco-2 cells, we demonstrated significant ABCB1-controlled transport that was abolished by a model ABCB1 inhibitor (CP100356). This confirmed the functional expression of ABCB1 in our cell line (Table 1). Subsequent experiments showed that lopinavir, ritonavir, saquinavir, atazanavir, daclatasvir, ledipasvir, and maraviroc all inhibit RHD123 efflux in Caco-2 cells, in keeping with previously reported in vitro studies (38, 53, 60–65). Of the tested antivirals, maraviroc (100 μM) and daclatasvir (20 μM) were the weakest inhibitors of RHD123 transport (Fig. 1).
Ex vivo accumulation studies were conducted in PCIS isolated from rat ileum and human jejunum. It has been shown that the functional expression of rat ABCB1 increases from the jejunum to the colon (61, 66), and ABCB1 model inhibitors affected RHD123 more strongly in the rat ileum than in other intestinal sections (42). In the human intestine, protein- and mRNA-level ABCB1 expression occurs predominantly in the jejunum and ileum (67, 68). ABCB1 activity seems to be higher in the ileum than the jejunum (43), but it was impossible to collect healthy segments of human ileum. The functional expression of ABCB1 in both rat and human experimental models was validated using CP100356, revealing that both models exhibited increases in RHD123 (10 μM) accumulation (Fig. 2 and Table 3) consistent with recently published data (42, 43).
Lopinavir, ritonavir, saquinavir, atazanavir, daclatasvir, ledipasvir, and maraviroc all strongly inhibited ABCB1 in rat-derived PCIS (Table 2). Of the tested antivirals, ritonavir (100 μM), lopinavir (20 μM), and daclatasvir (20 μM) had the greatest effects (Fig. 2). The ABCB1-inhibitory effect of ritonavir was previously demonstrated by in situ perfusion of murine intestines (39) and in murine whole-body pharmacokinetic studies (69, 70). We were unable to find any reports documenting the inhibition of animal-derived intestinal ABCB1 by any of the other tested drugs, however. Importantly, accumulation assays using rat-derived PCIS yielded similar results as in vitro bidirectional studies despite the previous finding that the inhibition of rat ABCB1 by antivirals can be affected to at least some extent by metabolic biodegradation in rat enterocytes, a phenomenon that does not occur in Caco-2 cells (71).
The inhibitory potential of selected antivirals was also tested in human-derived PCIS prepared from jejunum samples collected from five donors. Lopinavir (20 μM), ritonavir (100 μM), saquinavir (20 μM), and atazanavir (50 μM) exhibited significant inhibition of ABCB1 in intestinal samples from all five donors. Daclatasvir (20 μM) exhibited significant ABCB1 inhibition in PCIS from two donors, while maraviroc (100 μM) and ledipasvir (50 μM) abolished ABCB1 activity only in PCIS from one donor (Table 3). None of the tested compounds significantly affected the viability of both rat- and human-derived PCIS (Fig. 3).
It is important to note that ABCB1 activity is probably highest in the human ileum (43), and the antivirals were tested at concentrations that are probably lower than those occurring in the intestines of patients taking these drugs orally. The drugs’ inhibitory effects in clinical settings may therefore be even more pronounced than these results suggest. In accordance with our findings, a ritonavir-boosted atazanavir regimen reportedly increased the bioavailability of tenofovir by 91% when tenofovir was administered as the prodrug tenofovir alafenamide. Similarly, lopinavir boosted with ritonavir or atazanavir monotherapy increased the bioavailability of tenofovir by up to 32% and 37%, respectively, when tenofovir was administered as TDF (8). Tenofovir alafenamide and TDF are substrates of ABCB1 (8, 27, 51, 72), but tenofovir itself is not transported by either ABCB1 or other important drug efflux transporters such as breast cancer resistance protein and multidrug resistance-associated protein 2 (27). Moreover, tenofovir is not metabolized by cytochrome P450 (CYP450) isoenzymes, which are inhibited by lopinavir, atazanavir, or ritonavir (72, 73). Therefore, it can be hypothesized that intestinal ABCB1 inhibition is a key determinant of elevated plasma concentrations of tenofovir. Ledipasvir, daclatasvir, and maraviroc exhibited inconsistent inhibitory effects in the intestinal samples from the five donors (Table 3), while abacavir, TDF, and zidovudine exhibited no detectable inhibitory effects whatsoever (Tables 1 and 2). It is therefore unsurprising that (to our knowledge) there have been no reports of clinically relevant DDIs involving ledipasvir, daclatasvir, maraviroc, abacavir, TDF, or zidovudine that affect the bioavailability of coadministered ABCB1 substrates (8).
The variation in the reported ABCB1-inhibiting activity of certain drugs may be due to the use of different substrates in accumulation experiments as well as the use of different experimental techniques (74). We are therefore conducting follow-up experiments to test the effects of antivirals on intestinal efflux of digoxin, which is regarded as the standard ABCB1 probe for in vitro inhibitory studies (21, 24, 75). Lopinavir, atazanavir, daclatasvir, and ledipasvir reduced the efflux ratio of digoxin in bidirectional transport assays using Caco-2 cells, while abacavir had no significant effect, which is fully consistent with the results obtained using RHD123. In contrast to the findings obtained with RHD123, rilpivirine did inhibit ABCB1-mediated digoxin transport (data not shown). The experiments with digoxin are ongoing, and their results will be published in full elsewhere upon completion.
In conclusion, we have conducted the first complex study combining bidirectional transport assays and ex vivo accumulation in rat- and human-derived PCIS to evaluate DDIs involving clinically relevant drugs on intestinal ABCB1. Our results show that intestinal ABCB1 activity is not significantly inhibited by the anti-HIV and anti-HCV antiviral abacavir, zidovudine, TDF, etravirine, or rilpivirine but is inhibited by lopinavir, ritonavir, saquinavir, atazanavir, maraviroc, ledipasvir, and daclatasvir. Drugs in the latter group thus have a high potential to induce DDIs when administered orally. Our data may help explain the molecular mechanisms underpinning the elevated bioavailability of ABCB1 substrates, including antivirals and drugs prescribed to treat comorbidity following their oral administration. These results may also help guide the selection of combination pharmacotherapies and/or optimal dosing schemes for patients infected with HIV and/or HCV.
MATERIALS AND METHODS
Reagents and chemicals.RHD123 was purchased from Sigma-Aldrich (St. Louis, MO, USA). Lopinavir, ritonavir, abacavir, zidovudine, TDF, rilpivirine, saquinavir, atazanavir, maraviroc, and etravirine were obtained from the NIH AIDS Reagent Program. Ledipasvir, daclatasvir, simeprevir, and sofosbuvir were obtained from Scintila (Jihlava, Czech Republic). The ABCB1 inhibitor CP100356 monohydrochloride (CP100356) (76), Dulbecco’s modified Eagle’s medium (DMEM), nonessential amino acid solution (NEAA), penicillin-streptomycin solutions, dimethyl sulfoxide (DMSO), fetal bovine serum (FBS), and the ATP bioluminescence assay kit were purchased from Sigma-Aldrich (St. Louis, MO, USA). Hanks’ balanced salt solution (HBSS), William’s medium E containing l-glutamine (WME), and the bicinchoninic acid (BCA) protein assay kit were obtained from Thermo Fisher Scientific (Waltham, MA, USA). Krebs‐Henseleit buffer was prepared as described by de Graaf et al. (44). All other reagents were of analytical grade.
Stock solution and test solution.CP100356 and all the antivirals were dissolved in DMSO, while RHD123 was dissolved in 99.9% ethyl alcohol (EtOH). The final concentration of DMSO or EtOH was 0.1% in all experiments. The highest tested concentrations of antivirals in the experimental settings were determined by their solubility in the chosen medium (38) and/or the criteria specified in the relevant paragraphs/tables. Stock solutions of test compounds were prepared in accordance with the manufacturer’s instructions; concentrations of antivirals in test solutions were not analytically determined (35, 38, 77).
Cell culture and growth condition.The colorectal adenocarcinoma (Caco-2) cell line (ATCC HTB-37) was obtained from the American Type Culture Collection and cultured in DMEM-complete high-glucose medium with l-glutamine, supplemented with 10% FBS and 1% NEAA. Cells were routinely cultured in antibiotic-free medium and incubated in a humidified incubator at 37°C in an atmosphere containing 5% CO2. Cells from passages 10 to 40 were used in all in vitro experiments.
Animals.Wistar male rats (340 to 560 g) were purchased from Velaz (Prague, Czech Republic) and maintained under standard conditions (12-h/12-h day/night cycles, water and pellets ad libitum). All experiments were approved by the Ministry of Education, Youth and Sport, Czech Republic (approval no. MSMT-8358/2016-13), and the procedures used were consistent with the recommendations given in both the Guide for the Care and Use of Laboratory Animals (78) and the European Convention for the Protection of Vertebrate Animals Used for Experimental and Other Scientific Purposes (79).
Human tissue samples.Intestinal samples (jejunum) were collected based on anamnesis from five donors, all nonsmokers (summarized in Table 4). After receiving written informed patient consent as approved by the local research ethics committee (approval no. 201511 S26P), the samples were collected at the University Hospital in Hradec Kralove, Czech Republic during pancreaticoduodenectomy procedures. The donors had not been administered drugs known to cause upregulation of ABCB1.
Characteristics of the human intestine donors
In vitro bidirectional permeability assays.Transport assays were performed using microporous polycarbonate membrane filters (0.4-μm pore size, 12-mm diameter; Transwell 3401; Costar, Corning, NY) together with Caco-2 cells as previously described (40). The cells were seeded at a density of 3 × 105 cells per insert and cultured for 21 days in standard cultivation medium containing 1% penicillin-streptomycin. The medium was changed three times a week, during which transepithelial electrical resistance (TEER) across cell monolayers was routinely measured using a Millicell-ERS instrument (Millipore Corporation, Bedford, MA) (40). TEER values before the start of the experiment ranged from 1,100 to 1,900 Ω · cm2, in keeping with previous reports (53, 80). For the bidirectional permeability assay, an HBSS buffer was used. The pH in the apical (A) compartment was adjusted to 6.5 using a methanesulfonic acid solution, while that of the buffer in the basolateral (B) compartment was adjusted to 7.4 using HEPES (81). To improve the reproducibility of the results, the receiver compartment always contained 1% bovine serum albumin as previously recommended (54, 81). All wells were preincubated for 30 min with the appropriate transport buffer containing the model inhibitor CP100356 or a tested antiviral drug (40, 81–83). The assay was initiated by adding fresh buffer containing RHD123 (1 μM) and the appropriate inhibitor or tested antiviral drug into the donor compartment (compartment A for A→B transport and compartment B for B→A transport). Samples (200 μl) were collected after 1 and 2 h from the receiver compartment; after this was done the first time, fresh receiver solution was added to the receiver compartment to maintain the original volume (81, 84). The A→B and B→A transports were evaluated in terms of an apparent permeability coefficient (Papp) calculated using the equation Papp = (dC/dt) × Vr/(A × C0), where dC/dt is the change in concentration over time measured during the linear phase of transport over 1 h, Vr is the volume of the receiver well in milliliters, A is the area of the membrane in square centimeters, and C0 is the initial concentration in the donor compartment. The efflux ratio (rPapp) was then calculated using the equation rPapp = (Papp B→A)/(Papp A→B) (51).
Ex vivo accumulation experiments in rat PCIS prepared from the ileum.PCIS assays were performed as previously described (42, 44). Briefly, Wistar male rats were anesthetized with ether and then sacrificed by cervical spine dislocation, after which the intestine was removed and placed in a cold oxygenated (carbogen gas, 95% O2 and 5% CO2) Krebs-Henseleit buffer as previously described (44). After cleaning and preparing the tissue, the intestine was filled with an agarose solution (3% [wt/vol] agarose in 0.9% NaCl solution, 37°C) and embedded into a block of agarose using a tissue embedding unit (44). Tissue-containing slices approximately 250 μm thick were then cut from this block using a Krumdieck tissue slicer (Alabama R&D, Munford, AL, USA). The slices were preincubated for 30 min in a 24-well plate with each well containing 500 μl of WME supplemented with glucose (25 mM) and the appropriate antiviral drugs. After preincubation, the slices were transferred into the same medium containing a cocktail of RHD123 (10 μM) with the model inhibitor or one of the chosen drugs and then incubated for 2 h. The length of incubation was chosen based on previous measurements of the model inhibitor’s effect on RHD123 uptake over incubations of 1 and 2 h (data not shown). Both incubation steps were performed in a humidified atmosphere of 80% O2 and 5% CO2 at 37°C (42, 44). Accumulation was halted after 2 h by washing the slices twice with cold (4°C) Krebs‐Henseleit buffer.
Ex vivo accumulation experiments in human PCIS prepared from the jejunum.The ex vivo accumulation assays were performed as described previously (43, 44). Immediately after surgery, the resected morphologically healthy part of the human intestine was placed into a cold (4°C) Krebs‐Henseleit buffer oxygenated with carbogen gas as previously reported (44). Mucosa was separated from the muscularis layer, dissected into fragments measuring approximately 5 by 20 mm, and then embedded in a 3% agarose solution (3% [wt/vol] in 0.9% NaCl, 37°C). PCIS of approximately 300-μm thickness were cut using a Krumdieck tissue slicer (Alabama R&D, Munford, AL, USA) (43). The slices were preincubated for 30 min in the presence of the antiviral drugs in WME (43) and then transferred into a WME incubation medium containing RHD123 at a concentration of 10 μM (43) with the appropriate inhibitors. Both incubation steps were performed in a humidified atmosphere of 80% O2 and 5% CO2 at 37°C (43, 44). Accumulation was halted after 2 h by washing the slices twice with cold (4°C) Krebs‐Henseleit buffer.
RHD123 quantification.The concentration of RHD123 was quantified with a Tecan Infinite 200 M plate reader (Tecan Group, Männedorf, Switzerland), using excitation and emission wavelengths of 485 nm and 530 nm, respectively. RHD123 concentrations in samples collected from Caco-2 experiments were determined directly in HBSS buffer. Those in PCIS samples were assessed in supernatants prepared by adding 600 μl of acetonitrile solution (acetonitrile/water ratio, 2:1) and approximately 300 mg of glass minibeads (diameter, 1.25 to 1.65 mm; Carl Roth, Karlsruhe, Germany) to each slice and then homogenizing the tissue with a FastPrep24 5G minibead beater (MP Biomedicals, Santa Ana, CA, USA; 6.0 m/s, 2 times for 45 s each). Samples were then centrifuged (10 min; 7,800 × g), and the supernatant was collected for the analysis. Concentrations of RHD123 in samples collected from PCIS-based experiments were determined using a six-point calibration curve (concentrations, 0.000 μM, 0.0625 μM, 0.125 μM, 0.250 μM, 0.500 μM, and 1.000 μM) in acetonitrile solution. Six-point calibration curves for use with the bidirectional transport data were generated by analyzing samples prepared in HBSS buffer containing a tested antiviral drug at the appropriate concentration. RHD123 concentrations determined in PCIS were normalized against the protein content of the pellets obtained during centrifugation. The pellets were dried overnight at 37°C and then solubilized in 200 μl of 5 M NaOH for 24 h. Milli-Q water (800 μl) was then added to the samples to achieve an NaOH concentration of 1 M. The protein content of the processed samples was determined using a BCA protein kit (Thermo Fisher Scientific, Waltham, MA, USA).
Analysis of rat and human PCIS viability.The viability of the rat and human PCIS was evaluated by intracellular ATP concentration analysis as described previously (42, 43, 85). ATP concentrations were measured using the ATP bioluminescence assay kit CLS II (Roche, Mannheim, Germany). Measurements on PCIS were performed after a 2.5-h incubation with RHD123 (10 μM) and also after incubation with RHD123 (10 μM) and the model inhibitor or an antiviral drug.
Statistical analyses.The statistical significance of the data from the in vitro and human ex vivo experiments was assessed using Student’s two-tailed t test. Statistical analysis of the rat ex vivo experiments was performed using nonparametric two-tailed unpaired Mann-Whitney tests. The statistical significance of differences in the measured ATP concentrations was assessed using nonparametric Kruskal-Wallis analysis followed by Dunn’s test. All data were processed using GraphPad Prism 7.04 (GraphPad Software, Inc., San Diego, CA, USA).
ACKNOWLEDGMENTS
This research was financially supported by the Czech Science Foundation (GACR 18-07281Y), by the grant agency of Charles University (GAUK 1600317 and SVV 260 414), and by the project EFSA-CDN (no. CZ.02.1.01/0.0/0.0/16_019/0000841) cofunded by ERDF.
Ondrej Martinec contributed to the experimental design, performed and analyzed the in vitro cell-based experiments and the ex vivo accumulation experiments in PCIS, and helped write the manuscript. Martin Huliciak participated in the execution of the experiments and data analysis. Frantisek Staud contributed to the study’s conception and manuscript revision. Filip Cecka selected patients, performed small intestine resections, and contributed to manuscript revision. Ivan Vokral contributed to experimental design, performed and analyzed the in vitro cell-based experiments and ex vivo accumulation experiments in PCIS, helped write the manuscript, and contributed to the acquisition of financial support. Lukas Cerveny contributed to the experimental design, data analysis, and acquisition of financial support and contributed substantially to the writing of the manuscript. All authors have read and approved the final version of the manuscript.
FOOTNOTES
- Received 3 May 2019.
- Returned for modification 24 May 2019.
- Accepted 30 August 2019.
- Accepted manuscript posted online 3 September 2019.
- Copyright © 2019 American Society for Microbiology.
REFERENCES
- 1.↵
- 2.↵
- 3.↵
- 4.↵
- 5.↵
- 6.↵
- 7.↵
- 8.↵
- 9.↵
- 10.↵
- 11.↵
- 12.↵
- 13.↵
- 14.↵
- 15.↵
- 16.↵
- 17.↵
- 18.↵
- 19.↵
- 20.↵
- 21.↵
- 22.↵
- 23.↵
- 24.↵
- 25.↵
- 26.↵
- 27.↵
- 28.↵
- 29.↵
- 30.↵
- 31.↵
- 32.↵
- 33.↵
- 34.↵
- 35.↵
- 36.↵
- 37.↵
- 38.↵
- 39.↵
- 40.↵
- 41.↵
- 42.↵
- 43.↵
- 44.↵
- 45.↵
- 46.↵
- 47.↵
- 48.↵
- 49.↵
- 50.↵
- 51.↵
- 52.↵
- 53.↵
- 54.↵
- 55.↵
- 56.↵
- 57.↵
- 58.↵
- 59.↵
- 60.↵
- 61.↵
- 62.↵
- 63.↵
- 64.↵
- 65.↵
- 66.↵
- 67.↵
- 68.↵
- 69.↵
- 70.↵
- 71.↵
- 72.↵
- 73.↵
- 74.↵
- 75.↵
- 76.↵
- 77.↵
- 78.↵
- 79.↵
- 80.↵
- 81.↵
- 82.↵
- 83.↵
- 84.↵
- 85.↵