Advertisement

A Novel In Vitro Membrane Permeability Methodology Using Three-dimensional Caco-2 Tubules in a Microphysiological System Which Better Mimics In Vivo Physiological Conditions

Published:November 25, 2021DOI:https://doi.org/10.1016/j.xphs.2021.11.016

      Abstract

      The aim of this study was to develop an in vitro drug permeability methodology which mimics the gastrointestinal environment more accurately than conventional 2D methodologies through a three-dimensional (3D) Caco-2 tubules using a microphysiological system. Such a system offers significant advantages, including accelerated cellular polarization and more accurate mimicry of the in vivo environment. This methodology was confirmed by measuring the permeability of propranolol as a model compound, and subsequently applied to those of solifenacin and bile acids for a comprehensive understanding of permeability for the drug product in the human gastrointestinal tract. To protect the Caco-2 tubules from bile acid toxicity, a mucus layer was applied on the surface of Caco-2 tubules and it enables to use simulated intestinal fluid. The assessment using propranolol reproduced results equivalent to those obtained from conventional methodology, while that using solifenacin indicated fluctuations in the permeability of solifenacin due to various factors, including interaction with bile acids. We therefore suggest that this model will serve as an alternative testing system for measuring drug absorption in an environment closely resembling that of the human gastrointestinal tract.

      Keywords

      Introduction

      Accurate evaluation of the absorption mechanism of orally administered drugs contributes to predicting the in vivo performance of the drug products, and therefore plays an important role in drug development. To predict the in vivo performance of a drug product through in vitro measurements of permeability and solubility, Amidon et al.
      • Amidon GL
      • Lennernäs H
      • Shah VP
      • Crison JR.
      A theoretical basis for a biopharmaceutic drug classification: the correlation of in vitro drug product dissolution and in vivo bioavailability.
      established a Biopharmaceutics Classification System (BCS) that classifies small molecules according to solubility and permeability. This concept was applied to the BCS-based biowaiver guidelines, established by the United States Food and Drug Administration (FDA) and subsequently harmonized at the International Council for Harmonization (ICH)
      International Council for Harmonisation of Technical Requirements for Pharmaceuticals for Human Use
      Biopharmaceutics Classification System-Based Biowaivers M9.
      , and these guidelines have been widely used in the pharmaceutical industry. The core concept of the guidelines is that in vitro methods based on permeability and solubility may qualify the in vivo pharmacokinetic (PK) performance of drug products. However, the mechanism of the absorption of orally administered drugs is highly complex, comprising multiple factors not only the solubility but also hydrophobicity of components, variations in the expression of transporters, and susceptibility to the environment of the gastrointestinal lumen.
      • Lennernas H
      • Palm K
      • Fagerholm U
      • Artursson P
      Comparison between active and passive drug transport in human intestinal epithelial (Caco-2) cells in vitro and human jejunum in vivo.
      • Ungell AL.
      In vitro absorption studies and their relevance to absorption from the GI tract.

      Ungell AL, Karlsson J Cell Culture in Drug Discovery: An Industrial Perspective. In: Van de Waterbeemd H, Lennernas H, Artursson P, eds. Drug Bioavailability: Estimation of Solubility, Permeability, Absorption and Bioavailability. Wiley-VCH Verlag GmbH & Co. KGaA; 2003:90–131. Mannhold R, Kubinyi H, Folkers G, eds. Methods and Principles in Medicinal Chemistry.

      Accordingly, the in vitro methodologies now widely used in the pharmaceutical industry have some difficulties.
      In vitro intestinal absorption methodologies that employ the Caco-2 cell membrane, which was derived from a human colon adenocarcinoma, the parallel artificial membrane permeability assay (PAMPA) and the Ussing Chamber model are now used in the pharmaceutical industry.
      • Bujard A
      • Sol M
      • Carrupt PA
      • Martel S.
      Predicting both passive intestinal absorption and the dissociation constant toward albumin using the PAMPA technique.
      ,
      • Marino AM
      • Yarde M
      • Patel H
      • Chong S
      • Balimane PV.
      Validation of the 96 well Caco-2 cell culture model for high throughput permeability assessment of discovery compounds.
      The Ussing Chamber model uses the complete structure of the human small intestine as a permeability membrane, and could therefore be expected to increase the accuracy of predictions of in vivo permeability and its mechanisms.
      • Miyake M
      • Koga T
      • Kondo S
      • Yoda N
      • Emoto C
      • Mukai T
      • Toguchi H.
      Prediction of drug intestinal absorption in human using the Ussing chamber system: A comparison of intestinal tissues from animals and humans.
      • Rozehnal V
      • Nakai D
      • Hoepner U
      • et al.
      Human small intestinal and colonic tissue mounted in the Ussing chamber as a tool for characterizing the intestinal absorption of drugs.
      • Watanabe E
      • Takahashi M
      • Hayashi M.
      A possibility to predict the absorbability of poorly water-soluble drugs in humans based on rat intestinal permeability assessed by an in vitro chamber method.
      However, this method addresses the obvious technical and ethical difficulties associated with using human subjects as a source of gastrointestinal tissues. With the PAMPA, test compounds and buffer solutions are separated using a lipophilic microfilter impregnated with an alkane or a 1–20 w/v% phospholipid alkane solution to form a rate-limiting artificial membrane barrier to measure the diffusion of test compounds, on the other hand, PAMPA only determines permeability through passive diffusion.
      • Bujard A
      • Sol M
      • Carrupt PA
      • Martel S.
      Predicting both passive intestinal absorption and the dissociation constant toward albumin using the PAMPA technique.
      ,
      • Kansy M
      • Senner F
      • Gubernator K.
      Physicochemical high throughput screening: parallel artificial membrane permeation assay in the description of passive absorption processes.
      • Berggren S
      • Hoogstraate J
      • Fagerholm U
      • Lennernäs H.
      Characterization of jejunal absorption and apical efflux of ropivacaine, lidocaine and bupivacaine in the rat using in situ and in vitro absorption models.
      • Reis JM
      • Sinkó B
      • Serra CH.
      Parallel artificial membrane permeability assay (PAMPA) - is it better than Caco-2 for human passive permeability prediction?.
      In vitro methods employing the Caco-2 cell line, which mimics the in vivo phenotype of intestinal epithelial cells, serve as a classical model for understanding the mechanisms of active and passive transport in vitro. The Caco-2 model is able to give morphologic and functional characteristics very similar to those of enterocytes (e.g. tight junctions, apical and basolateral sides, and a brush border with microvilli).
      • Lozoya-Agullo I
      • Araújo F
      • González-Álvarez I
      • et al.
      Usefulness of caco-2/HT29-MTX and Caco-2/HT29-MTX/Raji B coculture models to predict intestinal and colonic permeability compared to Caco-2 monoculture.
      On the other hand, this model lacks a mucus layer and therefore often employs simple aqueous buffers that lack important components present in luminal fluids, such as bile acids, because the intestinal fluids may be cytotoxic to the Caco-2 monolayer.
      • Markopoulos C
      • Thoenen F
      • Preisig D
      • et al.
      Biorelevant media for transport experiments in the Caco-2 model to evaluate drug absorption in the fasted and the fed state and their usefulness.
      ,
      • Antoine D
      • Pellequer Y
      • Tempesta C
      Biorelevant media resistant co-culture model mimicking permeability of human intestine.
      It is known that human intestinal fluids comprise mixed micelles consisting of bile acids, lecithin, and lipolytic products.
      • Clarysse S
      • Brouwers J
      • Tack J
      • Annaert P
      • Augustijns P.
      Intestinal drug solubility estimation based on simulated intestinal fluids: comparison with solubility in human intestinal fluids.
      Additionally, bile acids in the gastrointestinal juice contribute to the complexity of absorption mechanisms via the formation of micelles, multiple aggregates, or colloidal structures, which affect the solubility and permeability of lipophilic drugs,
      • Meyerhoffer SM
      • McGown LB
      Critical micelle concentration behavior of sodium taurocholate in water.
      • Coello A
      • Meijide F
      • Rodriguez NE Taro J.
      Aggregation behavior of sodium cholate in aqueous solution.
      • Elvang PA
      • Hinna AH
      • Brouwers J
      • Hens B
      • Augustijns P
      • Brandl M.
      Bile salt micelles and phospholipid vesicles present in simulated and human intestinal fluids: structural analysis by flow field-flow fractionation/multiangle laser light scattering.
      • Elvang PA
      • Stein PC
      • Bauer-Brandl A
      • Brandl M.
      Characterization of co-existing colloidal structures in fasted state simulated fluids FaSSIF: a comparative study using AF4/MALLS, DLS and DOSY.
      • Miller JM
      • Beig A
      • Krieg BJ
      • et al.
      The solubility-permeability interplay: mechanistic modeling and predictive application of the impact of micellar solubilization on intestinal permeation.
      and by inhibition of the activity of the efflux transporter P-glycoprotein (P-gp) located in the apical membrane of Caco-2 cells.
      • Ingels F
      • Deferme S
      • Destexhe E
      • Oth M
      • Van den Mooter G
      • Augustijns P.
      Simulated intestinal fluid as transport medium in the Caco-2 cell culture model.
      Therefore, the absence of these luminal ingredients may significantly influence a compound's permeability, which in turn challenge the Caco-2 model's ability to function as an intestinal absorption method. To sum up, each existing in vitro method has its own difficulties, and the development of a more suitable in vitro method that comprehensively and accurately evaluates the in vivo absorption mechanism of drugs is required.
      To meet this challenge, we have focused on microphysiological systems (MPS). This technology is widely employed and has contributed to the major progress in in vitro models of the gut, such as three-dimensional (3D) culture technologies
      • Yu J
      • Peng S
      • Luo D
      • March JC.
      In vitro 3D human small intestinal villous model for drug permeability determination.
      ,
      • VanDussen KL
      • Marinshaw JM
      • Shaikh N
      • et al.
      Development of an enhanced human gastrointestinal epithelial culture system to facilitate patient-based assays.
      and microfluidic devices.
      • Kasendra M
      • Tovaglieri A
      • Sontheimer-Phelps A
      • et al.
      Development of a primary human Small Intestine-on-a-Chip using biopsy-derived organoids.
      • Kim HJ
      • Huh D
      • Hamilton G
      • Ingber DE.
      Human gut-on-a-chip inhabited by microbial flora that experiences intestinal peristalsis-like motions and flow.
      • Kim HJ
      • Li H
      • Collins JJ
      • Ingber DE
      Contributions of microbiome and mechanical deformation to intestinal bacterial overgrowth and inflammation in a human gut-on-a-chip.
      For example, Yu et al.
      • Yu J
      • Peng S
      • Luo D
      • March JC.
      In vitro 3D human small intestinal villous model for drug permeability determination.
      established a 3D-intestinal model which employs collagen hydrogel scaffolds that mimic the shape and size of the human small intestinal villi, and showed higher predictivity of drug permeability compared to 2D cultures. Further, Trietsch et al.
      • Trietsch SJ
      • Naumovska E
      • Kurek D
      • et al.
      Membrane-free culture and real-time barrier integrity assessment of perfused intestinal epithelium tubes.
      introduced an intestinal model comprising perfused tubules composed of Caco-2 cells in a microfluidic platform, the OrganoPlate. These gut tubules showed robust and reproducible cellular polarization and tight junction formation. These reports can be expected to be applied to in vitro methods for the evaluation of drug absorption, and the pharmaceutical industry's wide adoption of the readily available Caco-2 cell line has significantly accelerated the use of this model platform.
      • Beaurivage C
      • Naumovska E
      • Chang YX
      • et al.
      Development of a gut-on-a-chip model for high throughput disease modeling and drug discovery.
      ,
      • Moisan A
      • Michielin F
      • Jacob W
      • et al.
      Mechanistic investigations of diarrhea toxicity induced by Anti-HER2/3 combination therapy.
      Despite the availability of gut-on-a-chip models, however, few studies to our knowledge have comprehensively examined the possibility of evaluating drug absorption, and the impact of bile acids and ingredients in drug products using an MPS.
      Here, we conducted a proof-of-concept study of a novel in vitro drug absorption methodology that employs 3D-cultured Caco-2 tubules in the OrganoPlate. To protect the Caco-2 tubules from bile acid toxicity, a mucus coating was applied to the apical surface of the membrane. The mucus lining of the gastrointestinal tract contains 10–50 mg/mL of mucus glycoproteins, with a thickness ranging from 50 μm to 450 μm,
      • Matsui H
      • Wagner VE
      • Hill DB
      • et al.
      A physical linkage between cystic fibrosis airway surface dehydration and Pseudomonas aeruginosa biofilms.
      • Rozee KR
      • Cooper D
      • Lam K
      • Costerton JW.
      Microbial flora of the mouse ileum mucous layer and epithelial surface.
      • Sakata T
      • von Engelhardt W.
      Luminal mucin in the large intestine of mice, rats and guinea pigs.
      • Sellers LA
      • Carroll NJ
      • Allen A.
      Misoprostol-induced increases in adherent gastric mucus thickness and luminal mucus output.
      • Strous GJ
      • Dekker J.
      Mucin-type glycoproteins.
      and serves as a barrier to protect cells from toxic concentrations of endogenous compounds and foreign substances.
      • Khanvilkar K
      • Donovan MD
      • Flanagan DR.
      Drug transfer through mucus.
      We used propranolol and solifenacin as model compounds. Propranolol is classified as a BSC class I compound
      and is often used as a reference compound in Caco-2 permeability studies. Solifenacin is also classified as a BSC class I compound
      • Doroshyenko O
      • Fuhr U.
      Clinical pharmacokinetics and pharmacodynamics of solifenacin.
      , but its passive absorption is reportedly prevented by interaction with aggregates of bile acids in the FaSSIF
      • Yamamoto Y
      • Kumagai H
      • Haneda M
      • et al.
      The mechanism of solifenacin release from a pH-responsive ion-complex oral suspension in the fasted upper gastrointestinal lumen.
      , which is the biorelevant media simulating fasted state intestinal fluid. Yamamoto et al.
      • Yamamoto Y
      • Kumagai H
      • Haneda M
      • et al.
      The mechanism of solifenacin release from a pH-responsive ion-complex oral suspension in the fasted upper gastrointestinal lumen.
      suggested that solifenacin is released and then absorbed when bile acids are absorbed and recycled in the lower small intestine after disaggregation, however, no study has investigated this possibility using an in vitro methodology employing Caco-2 membranes. Therefore, solifenacin was selected with the expectation for not only a proof-of-concept study of a novel in vitro drug absorption methodology but also experimental-based findings of the mechanism.
      Moreover, Gijzen et al.
      • Gijzen L
      • Marescotti D
      • Raineri E
      • et al.
      An intestine-on-a-chip model of plug-and-play modularity to study inflammatory processes.
      developed an intestine-on-a-chip model which incorporated a mucus layer by coculturing Caco-2 cells with HT29-MTX cells, which are mucus-producing cells commonly called goblet cells, derived from a colorectal adenocarcinoma. In this study, we also evaluated this model to gain a comprehensive understanding of the mechanism of drug absorption, including the effects of bile acids and pharmaceutical excipients.

      Materials and Methods

       Materials

      Solifenacin succinate was prepared at Astellas Ireland (Dublin, Ireland), and a solifenacin succinate oral suspension 1mg/mL, was supplied by Astellas Pharma Europe (Leiden, the Netherlands). In this study, the simple substance of solifenacin succinate was defined as drug substance. A solifenacin succinate oral suspension 1mg/mL contained the following excipients in addition to solifenacin succinate; polacrilin potassium (adsorbent), methylparaben and propylparaben (antimicrobial preservation), propylene glycol (solubilizer), simethicone emulsion (antifoam agent), Carbopol 974P (thickener), xylitol and acesulfame potassium (sweetener), orange flavor (flavor), sodium hydroxide (pH adjuster) and water (solvent). Propranolol (P0884) was purchased from Sigma-Aldrich (St. Louis, MO), and simulated intestinal fluid (SIF) powder was purchased from biorelevant.com (FaSSIF/FeSSIF/FaSSGF; London, UK). The OrganoPlate three-lane 40 (4004-400-B), whose schematic image is shown in Figure 1, was supplied by Mimetas (Leiden, the Netherlands). Caco-2 cells (86010202), HT29-MTX-E12 cells (12040401), mucus (mucin from porcine stomach Type III, M1778), penicillin-streptomycin (P4333), Dulbecco's modified Eagle's medium (DMEM, D6546), Hank's balanced salt solution (HBSS, H4641), sodium hydrogen carbonate (NaHCO3, S5761), formaldehyde (252549), Triton X-100 (T8787), bovine serum albumin (BSA, A2153), and Tween-20 (P9616) were obtained from Sigma-Aldrich.
      Fig. 1
      Fig. 1Schematic image of OrganoPlate three-lane.
      Fetal bovine serum (FBS, 16140-071 or A13450) were obtained from Thermo Scientific (Waltham, MA) or the American Type Culture Collection (ATCC, Manassas, VA). GlutaMAX (35050-061), HEPES (15630-122), trypsin/ethylenediaminetetraacetic acid (EDTA) solution (15400-154), and 0.25% trypsin (15290-046), EDTA (AM9260G) were obtained from Thermo Scientific. Eagle's minimum essential medium (EMEM, 30-2003) was obtained from ATCC. Nonessential amino acids (NEAA, 11140-050) and phosphate-buffered saline (PBS, 20012068) were obtained from Life Technologies (Carlsbad, CA). Collagen-I (Cultrex 3D collagen-I Rat Tail, 5 mg/mL, 3447-020-01) was obtained from AMS Biotechnology (Abingdon, UK). Hoechst 33342 (1:2000; H3570) and rabbit anti-occludin (OCLN) antibody (1:100; 71-1500) were obtained from Thermo Scientific, a mouse anti-ezrin (EZR) antibody (1:50; 610602) was from BD Biosciences (San Jose, CA), a rabbit anti-ZO-1 antibody (1:125; 617300) was from Invitrogen (Carlsbad, CA), and a mouse anti-MUC2 antibody (1:100; MA512345), mouse anti-MUC5AC antibody (1:100; MA1-38223), mouse IgG isotype control (86599), rabbit IgG isotype control (86199), goat anti-rabbit AlexaFluor 488 (A11008) and donkey anti-mouse AlexaFluor 647 (A31571) were from Thermo Scientific.

       Cell culture

      Caco-2 was cultured in T75 flasks in EMEM supplemented with 10% FBS, 1% NEAA, and 1% penicillin-streptomycin. HT29-MTX-E12 was cultured in T75 flasks in DMEM supplemented with 10% FBS, 1% NEAA, 1% GlutaMAX, and 1% penicillin-streptomycin.

       Caco-2 Cell Culture in the OrganoPlate

      Caco-2 cells were cultured in the OrganoPlate three-lane containing 40 chips per plate. Each chip comprises 3 channels (400 μm × 220 μm (w × h) each) and 2 phaseguides (100 μm × 55 μm (w × h) each) (Figure 1). First, 50 μL HBSS was added to each observation window to provide clear visibility of the culture. Next, 2 μL of ECM composed of 4 mg/mL collagen-I, 100 mM HEPES and 3.7 mg/mL NaHCO3 was loaded into the ECM channel from the gel inlet followed by 15 min static incubation at 37°C. To keep the ECM hydrated after the incubation period, 30 μL HBSS was added to the gel inlet and the plate was incubated overnight at 37°C. Caco-2 cells trypsinized using a solution containing 0.25% trypsin and 0.53 mM EDTA were pelleted and resuspended in EMEM-supplemented Caco-2 medium (concentration, 1 × 107 cells/mL). Subsequently, a 2 μL cell suspension was applied to the inlet of the top perfusion channel followed by 50 µl of medium to the same well. The plate was placed on its side for 3–4 h at 37°C to allow the cells to settle to the ECM. Next, an additional 50 μL of culture medium was applied to each of the remaining inlets and outlets of the apical (top) and basal (bottom) perfusion channels. The plate was placed horizontally on an interval rocker in an incubator (37°C, 5% CO2/95% air). The rocker was set to switch between a +7° and –7° inclination every 8 min (OrganoFlow; Mimetas), thereby creating bidirectional flow. These settings achieved a mean flow rate of 2.02 μL/mL, resulting in a mean shear stress of 0.13 dyne/cm2. This closely mimics physiological levels of intestinal epithelial shear stress, which range between approximately 0.002 dyne/cm2 and 0.08 dyne/cm2.
      • Vormann MK
      • Gijzen L
      • Hutter S
      • et al.
      Nephrotoxicity and kidney transport assessment on 3d perfused proximal tubules.
      • Ishikawa T
      • Sato T
      • Mohit G
      • Imai Y
      • Yamaguchi T.
      Transport phenomena of microbial flora in the small intestine with peristalsis.
      • Lentle RG
      • Janssen PW.
      Physical characteristics of digesta and their influence on flow and mixing in the mammalian intestine: a review.
      The medium was replaced every 2–3 days, and tubules cultured for 4 or 5 days were used for subsequent analyses.

       Coating Caco-2 Mono-Culture Tubules in the OrganoPlate with an Artificial Mucus

      Media in the inlets and outlets on the apical and basal perfusion channels of each chip was removed completely. Next, 50 μL of 50 mg/mL mucus in HBSS was added to the inlets and outlets on the apical perfusion channel, where the Caco-2 cells were cultured in a tubular shape, and 50 μL HBSS was added to the inlets and outlets of the basal perfusion channel. Subsequently, the plate was placed horizontally in a humidified incubator (37°C; 5% CO2/95% air) on an interval rocker, switching between a +7° and –7° inclination every 8 min (OrganoFlow) for 15 min. After incubation, the HBSS that filled the inlets and outlets of the basal perfusion channel was replaced with Caco-2 medium, and the mucus in the inlets and outlets of the top medium channel was replaced with the test solution.

       Coculture of Caco-2 and HT29-MTX-E12 Cells in the OrganoPlate

      Caco-2 and HT29-MTX-E12 cells trypsinized using a solution of 0.25% trypsin and 0.53 mM EDTA were pelleted, and resuspended in EMEM-supplemented Caco-2 medium (total concentration, 1 × 107 cells/mL). The ratio of Caco-2 to HT29-MTX-E12 cells in the suspension was 6:1. Subsequent steps were performed as described in the above section, Caco-2 cell culture in the OrganoPlate.

       Confocal Imaging of Immunofluorescence

      OrganoPlate cultures were fixed with 3.7% formaldehyde in PBS for 10 min and washed twice for 5 min with PBS. Subsequently, the cultures were permeabilized with 0.3% Triton X-100 in PBS for 10 min, washed with 4% FBS in PBS for 5 min, and incubated with blocking solution (2% FBS, 2% BSA, and 0.1% Tween-20 in PBS) for 40 min at room temperature (RT). After blocking, the cells were incubated with primary antibodies for 60 min at RT, washed twice, incubated with secondary antibodies for 30 min at RT, and washed twice with 4% FBS in PBS.
      Antibodies use for immunohistochemical analysis were as follows: rabbit anti-OCLN, mouse anti-EZR, rabbit anti-ZO-1, mouse anti-MUC2 antibody and mouse anti-MUC5AC antibody, mouse isotype control, rabbit isotype control, donkey anti-mouse AlexaFluor 647 and goat anti-rabbit AlexaFluor 488. Nuclei were stained with Hoechst 33342 and the cells were stored in PBS.
      The images of cultures were taken using an ImageXpress Micro XLS-High Content Imaging System (Molecular Devices, San Jose, CA). Image analysis was performed using FiJi.
      • Schindelin J
      • Arganda-Carreras I
      • Frise E
      • et al.
      Fiji: an open-source platform for biological-image analysis.

       Transepithelial Electrical Resistance (TEER) Measurement

      The barrier integrity of the cultures in the OrganoPlate was determined by measuring TEER (n = 4–9) using an automated multichannel impedance spectrometer (OrganoTEER; Mimetas). Prior to exposure, TEER measurements were performed to serve as a quality control check. For single-point measurements, the test solutions were added to the chips and the plate was placed on an interval rocker in a humidified incubator (37°C, 5% C02). The plate was taken out of the incubator and left at RT to equilibrate for 30 minutes before starting the measurement. After the measurement, the plate was placed back in the incubator, on the interval rocker. For time-lapse TEER measurements, the solutions were added to the chips and the plate was placed in the OrganoTEER instrument, on an interval rocker, in a humidified incubator (37°C, 5% C02). The time-lapse measurement was started directly, with measurements taken every 10 min for a total of 120 min. The software for OrganoTEER generated the TEER values of each chip in ohms (Ω), then converted them to ohm-squared centimeters by multiplying them by the surface area of the tubule–ECM interface (0.0057 cm2). All data are expressed as the mean ± standard deviation (SD) (n = 4-9). The Smirnov–Grubbs’ test was applied for evaluating outliers. Data analysis was performed using Microsoft Excel (Microsoft Corp., Redmond, WA) and R software for Windows (version 4.0.3; R Core Team, 2020).

       Permeability Study

      Drug permeability was determined using the OrganoPlate three-lane with Caco-2 mono-culture tubules without an artificial mucus layer (Caco2-Mono), Caco-2 mono-culture tubules with an artificial mucus layer (Caco2-MUC), or Caco-2/HT29-MTX-E12 coculture tubules (Caco2-HT29). All conditions of the drug permeability studies are summarized in Table 1. After cell culture or mucus coating, the mucus, HBSS and the culture medium were aspirated from the apical and basal perfusion channels. Subsequently, 40 μL of culture medium was added to the inlet and outlet wells of the gel and the basal perfusion channel; and 60 μL and 50 μL of the test solutions were added to inlet and outlet wells of the apical perfusion channel, respectively. The OrganoPlate was horizontally placed for 60 or 120 min in an incubator (37°C; 5% CO2/95% air) on an interval rocker, switching between +7° and –7° inclination every 8 min (OrganoFlow; Mimetas).
      Table 1Testing Conditions for Permeability Study.
      Condition No.CompoundCompound conc. (μg/mL)TissueMediumBile acid conc
      Sodium taurocholate, 3 mM; lecithin, 0.75 mM; for 100% of Bile acid.
      Artificial Mucus LayerExcipients
      “-” indicates that the simple substance “solifenacin succinate” was used, and “+” indicates that the commercial product “solifenacin succinate oral suspension 1mg/mL” was used.
      Papp of Propranolol (× 10−6 cm/s)
      1Propranolol50Caco-2Caco-2 medium-+-27.38
      2Propranolol50Caco-2FaSSIF100%+-2.97
      3Solifenacin15Caco-2Caco-2 medium----
      4Solifenacin15Caco-2Caco-2 medium-+--
      5Solifenacin5Caco-2FaSSIF0%+--
      6Solifenacin15Caco-2FaSSIF0%+--
      7Solifenacin5Caco-2FaSSIF20%+--
      8Solifenacin15Caco-2FaSSIF20%+--
      9Solifenacin5Caco-2FaSSIF100%+--
      10Solifenacin15Caco-2FaSSIF100%+--
      11Solifenacin15Caco-2FaSSIF0%++-
      12Solifenacin15Caco-2FaSSIF20%++-
      13Solifenacin15Caco-2FaSSIF100%++-
      14Solifenacin5Caco-2/ goblet cellFaSSIF100%---
      15Solifenacin15Caco-2/ goblet cellFaSSIF100%---
      *1 Sodium taurocholate, 3 mM; lecithin, 0.75 mM; for 100% of Bile acid.
      low asterisk2 “-” indicates that the simple substance “solifenacin succinate” was used, and “+” indicates that the commercial product “solifenacin succinate oral suspension 1mg/mL” was used.
      Different chips containing the culture model were used for different times. At each time, the culture medium and test solutions were sampled from the OrganoPlate and collected in V-bottom 96-well plates. The amounts of solifenacin and bile acids (taurocholic acid) in these solutions were determined using an UPLC-MS/MS system. Data analysis was performed using Microsoft Excel. All data are expressed as the mean ± SD of four individual chips.

       LC-MS Measurements

      The concentrations of permeated solifenacin and bile acids (taurocholic acid) were determined using UPLC-MS/MS. Chromatographic separation was achieved using an Agilent 1290 Infinity II LC System (Agilent Technologies, Santa Clara, CA) equipped with a Phenomenex Synergi Fusion RP LC column (dp = 4 μm, 2 × 150 mm; Phenomenex, Torrance, CA), flow rate = 0.4 mL/min over an 8 min gradient. The samples were kept at 4°C until injection. The UPLC was coupled to an electrospray ionization triple quadrupole mass spectrometer (AB SCIEX Qtrap 6500; AB Sciex, Framingham, MA). Analytes were detected in positive ion mode with Multiple Reaction Monitoring (MRM) using nominal mass resolution. The data were evaluated by integration of assigned MRM peaks using MultiQuant Software for Quantitative Analysis (AB SCIEX, Version 3.0.2; AB Sciex).

       Apparent Permeability Coefficient

      The apparent permeability coefficient (Papp) was calculated as follows:
      Papp(cm/s)=ΔQ/Δt×1/(A×C0),


      where ΔQ/Δt is the quantity amount of measurement target permeated from apical to basal per unit time, A is the surface area of the gel-medium interface (0.5 mm2) of the OrganoPlate, and C0 is the initial concentration of the measuring target added to the top perfusion.

       Measurement of Barrier Integrity in the OrganoPlate

      The barrier integrity of Caco-2 cells cultured in the OrganoPlate was determined using a leak test that detected the fluorescence of tetramethylrhodamine isothiocyanate (TRITC)–dextran (4.4 kDa; T1037; Sigma-Aldrich). To this end, all medium was aspirated from channels. Subsequently, 20 μL of culture medium without TRITC–dextran was added in the inlet and outlet of the basal perfusion channel and ECM channel. Next, 40 μL of culture medium containing 0.5 mg/mL of TRITC–dextran was added to the inlet and 30 µl to the outlet on the apical perfusion channels. The leakage of TRITC–dextran from the apical to the basal side through the ECM channel was determined by imaging using an ImageXpress XLS Micro (Molecular Devices, San Jose, CA).

      Results and Discussion

       Immunofluorescence Analysis of Marker Expression in Caco2-MUC

      Caco2-MUC, namely Caco-2 mono-culture tubules with an artificial mucus layer, was established in the OrganoPlate three-lane (Figure 1). We used an immunofluorescence-based approach to analyze the distribution of nuclei, ZO-1 and MUC2 in Caco2-MUC before and after exposure to solifenacin (15 µg/mL) in FaSSIF (Figure 2). Stained nuclei covered the tubules, suggesting Caco-2 cells spread and formed a tubular structure on the surface of the ECM gel and on the inside wall of the perfusion channel (Figure 2a, b). The max projections of ZO-1 fluorescence, shown in green, covered the entire Caco-2 tube (Figure 2b), indicating the proper expression and structural integrity of the tight junctions, consistent with the findings of Trietsch et al.
      • Trietsch SJ
      • Naumovska E
      • Kurek D
      • et al.
      Membrane-free culture and real-time barrier integrity assessment of perfused intestinal epithelium tubes.
      MUC2 immunofluorescence was localized to the tubular structures formed by Caco-2 cells, suggesting that the artificial mucus layer coated the entire surface of the Caco-2 membrane tube (Figure 2a, b). The max projections of stained nuclei, ZO-1 and MUC2 were maintained after introducing solifenacin in FaSSIF (Figure 2c).
      Fig. 2
      Fig. 2Confocal imaging of Caco2-MUC before or after exposure to solifenacin in FaSSIF. (a) 3D-overlay of images of nuclei and MUC2 before exposure; (b) Max projections of nuclei, ZO-1, and MUC2 before exposure; (c) Max projections of nuclei, ZO-1, and MUC2 after exposure.

       Barrier Integrity of Caco2-MUC Tubules in the OrganoPlate

      To confirm the barrier integrity of Caco2-MUC, we next assessed temporal changes in the TEER of Caco2-MUC. The mean TEER values of Caco2-MUC using Caco-2 medium and solifenacin in FaSSIF were 527 Ω·cm2 and 582 Ω·cm2 at t = 0 min, respectively, and the barrier integrity of each membrane was tightly maintained after each solution had been introduced through the tubules (Figure 3a).
      Fig. 3
      Fig. 3(a) Time-dependent TEER values of Caco2-MUC during exposure to Solifenacin in FaSSIF; (b) TEER value ratio after 120 min compared to each mean initial value. Mean ± SD (n = 6) for Caco2-MUC, ●; (n = 9) for Caco2-MUC in exposure condition (condition 10), .
      The membrane integrity of Caco2-Mono was also evaluated using a 4.4 kDa TRICT-dextran probe. Leakage of the fluorescent reagent from the apical to basal channels was observed after introducing FaSSIF (data not shown), consistent with studies showing that bile acids included in FaSSIF were toxic to Caco-2 cells
      • Markopoulos C
      • Thoenen F
      • Preisig D
      • et al.
      Biorelevant media for transport experiments in the Caco-2 model to evaluate drug absorption in the fasted and the fed state and their usefulness.
      ,
      • Antoine D
      • Pellequer Y
      • Tempesta C
      Biorelevant media resistant co-culture model mimicking permeability of human intestine.
      and that a mucus layer prevented this damage.
      • Antoine D
      • Pellequer Y
      • Tempesta C
      Biorelevant media resistant co-culture model mimicking permeability of human intestine.
      ,
      • Wuyts B
      • Riethorst D
      • Brouwers J
      • Tack J
      • Annaert P
      • Augustijns P.
      Evaluation of fasted state human intestinal fluid as apical solvent system in the Caco-2 absorption model and comparison with FaSSIF.
      ,
      • Meaney C
      • O'Driscoll C.
      Mucus as a barrier to the permeability of hydrophilic and lipophilic compounds in the absence and presence of sodium taurocholate micellar systems using cell culture models.
      Therefore, these findings indicate that the tubules were sufficiently coated and protected by the artificial mucus layer. Further, no leakage was confirmed in the case of solifenacin in Caco-2 medium (data not shown), indicating that the impact from solifenacin is not necessarily to be considered in this model. Moreover, according to the results of TEER measurement of Caco2-Mono using Caco-2 medium with solifenacin (condition 3), the barrier integrity of this condition was maintained (data not shown), corroborating the idea that no need to consider the impact from solifenacin in this model.
      Among the measurements under condition 10 (n = 9), one measurement showed a significantly lower TEER value after 120-min incubation with a P-value of 0.048 by Smirnov–Grubbs’ test (Figure 3b). These findings indicate that bile acids may damage Caco-2 cells if the coating of the tube by the artificial mucus layer is insufficient. Therefore, the TEER value of each tubule was assessed before and after testing, confirming the validity of all permeability data presented here.

       Proof of Concept of the Utility of Caco2-MUC

      Propranolol is generally used as a reference high permeable compound in in vitro Caco-2 permeability studies,
      and thus the permeability of propranolol (conditions 1 and 2, Table 1) was determined (Figure 4a). The propranolol concentrations in the basal channel increased with time under condition 1 (Caco-2 medium) and condition 2 (FaSSIF), and the concentrations of permeated propranolol under the latter condition after 60 and 120 min were lower than those of the former. The Papp values of propranolol, calculated using the propranolol concentration at 120 min, were 27.4 × 10−6 cm/sec and 3.0 × 10−6 cm/sec under conditions 1 and 2, respectively.
      Fig. 4
      Fig. 4Concentrations of permeated propranolol, solifenacin, and bile acid in the basal channel after 60 and 120 min. (a) Propranolol concentrations (conditions 1 and 2); (b) Solifenacin concentrations (Conditions 3-13); (c) Bile acid concentrations (Conditions 7-10, 12 and 13). Data are expressed as the mean ± SD (n = 4).
      The Papp of propranolol for the Caco-2 membrane treated with Caco-2 medium in the conventional 2D-culture method ranged from 39.4 × 10−6 to 47.2 × 10−6 cm/sec.
      • Li C
      • Liu T
      • Cui X
      • Uss AS
      • Cheng KC.
      Development of in vitro pharmacokinetic screens using Caco-2, human hepatocyte, and Caco-2/human hepatocyte hybrid systems for the prediction of oral bioavailability in humans.
      • Zhu C
      • Jiang L
      • Chen TM
      • Hwang KK.
      A comparative study of artificial membrane permeability assay for high throughput profiling of drug absorption potential.
      • Alsenz J
      • Haenel E.
      Development of a 7-day, 96-well Caco-2 permeability assay with high-throughput direct UV compound analysis.
      Further, Ingels et al.
      • Ingels F
      • Beck B
      • Oth M
      • Augustijns P.
      Effect of simulated intestinal fluid on drug permeability estimation across Caco-2 monolayers.
      reported that the Papp of propranolol with FaSSIF, in contrast to Caco-2 medium, was 4.5 × 10−6 cm/sec and that the use of FaSSIF tended to decrease drug permeability compared with Caco-2 medium because of the interaction with bile acids contained in FaSSIF or through being incorporated into micelles formed by bile acids. Here, an artificial mucus layer was applied to the surface of the Caco-2 membrane as an additional barrier, resulting in a slightly lower Papp. However, Caco2-MUC reproduced Papp values equivalent to those previously published concerning the decrease in Papp values caused by FaSSIF,
      • Ingels F
      • Beck B
      • Oth M
      • Augustijns P.
      Effect of simulated intestinal fluid on drug permeability estimation across Caco-2 monolayers.
      supporting the idea that the present model serves as an alternative method for assessing in vitro drug permeability.

       Permeability Assessments

       a. Solifenacin Permeability

      Solifenacin is classified as a BSC class I compound and absorbed by passive diffusion.
      • Doroshyenko O
      • Fuhr U.
      Clinical pharmacokinetics and pharmacodynamics of solifenacin.
      We therefore assessed the permeability of solifenacin in conditions 3–13 shown in Table 1 (Figure 4b). The solifenacin concentrations in the basal side increased after 120 min incubation under all conditions. In the presence of FaSSIF, the amounts of permeated solifenacin were lower compared with those of Caco-2 medium (conditions 4 vs. 10), as in the case of propranolol (Figure 4a). Further, the permeabilities of solifenacin in the presence of the artificial mucus layer were lower than in its absence (condition 3 vs 4).
      To determine the concentration-dependency of solifenacin permeability (5 µg/mL and 15 µg/mL) as well as the impact of excipients added to the apical channel, the solifenacin concentrations in the basal side after 60 min incubation were evaluated. Permeated solifenacin was detected when its concentration was at 15 µg/mL (conditions 5, 7, and 9) but not at 5 µg/mL (conditions 6, 8, and 10). Moreover, solifenacin was detected in the basal side after 60 min when it was applied as drug substance (without excipients; conditions 6, 8, and 10) but not when it was applied as drug product (including excipients; conditions 11, 12, and 13). Since solifenacin in the drug product forms an ionic complex with a resin and is gradually released in FaSSIF,
      • Yamamoto Y
      • Kumagai H
      • Haneda M
      • et al.
      The mechanism of solifenacin release from a pH-responsive ion-complex oral suspension in the fasted upper gastrointestinal lumen.
      Wang et al.
      • Wang Y
      • Cao J
      • Wang X
      • Zeng S.
      Stereoselective transport and uptake of propranolol across human intestinal Caco-2 cell monolayers.
      found that the permeability rate of a drug correlated with its concentration and that the concentration gradient of a drug between the apical and basal sides critically affects active and passive transport. Therefore, our results indicate that the concentration of free solifenacin might have impact on its permeation.

       b. Bile Acids Permeability

      Bile acids are components of FaSSIF and are taken up at the lower small intestine.
      • Marin JJ
      • Macias RI
      • Briz O
      • Banales JM
      • Monte MJ
      Bile acids in physiology, pathology and pharmacology.
      ,
      • Slijepcevic D
      • van de Graaf SF.
      Bile acid uptake transporters as targets for therapy.
      Accordingly, we determined the permeabilities of bile acids under conditions 7–10, 12, and 13 (Table 1, Figure 4c). Bile acids concentrations in the basal side increased after 120 min incubation under these conditions. Permeability of these bile acids depended on factors such as bile acids concentration (conditions 7 and 8 vs conditions 9 and 10) and the presence or absence of excipients (conditions 8 and 10 vs conditions 12 and 13). These results indicate that the permeabilities of bile acids at the least depend on factors in this system, likely because the bile acids-absorbing transporter (apical sodium-dependent bile acids transporter), which functions as an ileal bile acids transporter, is activated or inhibited by the excipients and bile acids itself.
      • Marin JJ
      • Macias RI
      • Briz O
      • Banales JM
      • Monte MJ
      Bile acids in physiology, pathology and pharmacology.
      ,
      • Slijepcevic D
      • van de Graaf SF.
      Bile acid uptake transporters as targets for therapy.
      Further, it is reported that the addition of a highly viscous compound (e.g. β-glucan) increases the viscosity of a solution and thereby reduces the permeability of bile acids
      • Naumann S
      • Haller D
      • Eisner P
      • Schweiggert-Weisz U
      Mechanisms of interactions between bile acids and plant compounds-a review.
      , and Carbopol 974P was indeed included as a thickener in the excipients of this study. Therefore, the excipients might have increased the viscosity, decreased the fluidity and reduced the permeability of bile acids in our study.

       c. Effect of Bile Acids on Solifenacin Permeability

      We also evaluated the effect of bile acids on the permeability of solifenacin and the effect of solifenacin on the permeability of bile acids. When the bile acids concentration in the apical channel was varied from 20% to 100%, the amount of permeated bile acids increased approximately 5-fold in the presence of 5 µg/mL of drug substance solution (conditions 7 and 9, Figure 4c), approximately 3-fold in the presence of 15 µg/mL of drug product suspension (conditions 12 and 13, Figure 4c), and approximately 15-fold in the presence of 15 µg/mL of drug substance solution (conditions 8 and 10, Figure 4c). Similar differences were observed in the permeability study of solifenacin: the amount of permeated solifenacin decreased with increasing bile acids concentrations in the presence of 5 µg/mL of drug substance solution (conditions 5, 7, and 9; Figure 4b) and 15 µg/mL of drug product suspension (conditions 11, 12, and 13; Figure 4b), but increased with increasing bile acids concentrations in the presence of 5 µg/mL of drug substance solution (conditions 6, 8, and 10; Figure 4b). Therefore, we conclude that solifenacin and bile acids exerted reciprocal effects on permeability through the Caco-2 membrane.
      Yamamoto et al.
      • Yamamoto Y
      • Kumagai H
      • Haneda M
      • et al.
      The mechanism of solifenacin release from a pH-responsive ion-complex oral suspension in the fasted upper gastrointestinal lumen.
      found that solifenacin was not absorbed promptly into the small intestine because of the formation of aggregates with bile acids and was subsequently released and gradually absorbed as the bile acids were absorbed in the lower small intestine. Thus, not only the concentrations of solifenacin and bile acids but also their existing forms may impact their permeation. Our present results support their conclusion and indicate the importance of simultaneously measuring the permeabilities of a drug of interest and bile acids to understand the mechanism of absorption of the drug.

       Brief Summary of Caco2-MUC

      The suitability of Caco2-MUC was evaluated by measuring the permeabilities of propranolol and solifenacin in FaSSIF. Both compounds showed increases in concentration in the basal channel over time, indicating permeability over the Caco-2 layer. Previous studies showed that the in vivo permeabilities of many drugs could be evaluated in Caco-2 medium,
      • Ingels F
      • Beck B
      • Oth M
      • Augustijns P.
      Effect of simulated intestinal fluid on drug permeability estimation across Caco-2 monolayers.
      ,
      • Fossati L
      • Dechaume R
      • Hardillier E
      • et al.
      Use of simulated intestinal fluid for Caco-2 permeability assay of lipophilic drugs.
      but that in contrast, those of drugs such as chlorothiazine and sulfasalazine were difficult to evaluate in Caco-2 medium and require FaSSIF. The reason for this is that bile acids in FaSSIF improve drug solubility, prevent drug adsorption and suppress the transporter activity of P-glycoprotein, and these effects may be crucial for certain compounds that are actively transported, poorly soluble, poorly permeable, easy to adsorb, or difficult to recover.
      • Ingels F
      • Beck B
      • Oth M
      • Augustijns P.
      Effect of simulated intestinal fluid on drug permeability estimation across Caco-2 monolayers.
      ,
      • Fossati L
      • Dechaume R
      • Hardillier E
      • et al.
      Use of simulated intestinal fluid for Caco-2 permeability assay of lipophilic drugs.
      To sum up, these findings suggest that mimicking in vivo conditions will help to accurately predict drug permeability. Further, not only bile acids but also mucus, which forms a highly viscous water layer near the intestinal epithelium, affect the absorption of drugs.
      • Sigurdsson HH
      • Kirch J
      • Lehr CM.
      Mucus as a barrier to lipophilic drugs.
      ,
      • Miyazaki K
      • Kishimoto H
      • Muratani M
      • Kobayashi H
      • Shirasaka Y
      • Inoue K.
      Mucins are involved in the intestinal permeation of lipophilic drugs in the proximal region of rat small intestine.
      Leal et al.
      • Leal J
      • Smyth HDC
      • Ghosh D
      Physicochemical properties of mucus and their impact on transmucosal drug delivery.
      also found that physicochemical factors such as viscoelasticity, pH, and ionic strength contribute to the barrier function of mucus. Our finding that the permeability of solifenacin was delayed in the presence of an artificial mucus layer also supports these findings. Taken together, these findings suggest that mimicking in vivo conditions will help to accurately predict drug permeability.
      Our results showed that the introduction of an artificial mucus layer to the Caco-2 tubules made it possible to use simulated intestinal fluids to mimic in vivo conditions and that solifenacin and bile acids exerted reciprocal effects on their permeability. Further, 3D cell cultures may express more proteins representative of an authentic epithelial barrier, which represents a significant advantage compared with the 2D-model.
      • Ravi M
      • Paramesh V
      • Kaviya SR
      • Anuradha E
      • Solomon FD.
      3D cell culture systems: advantages and applications.
      Thus, Caco2-MUC is suitable for use as a test system for measuring drug permeability, and the introduction of an artificial mucus layer facilitates the simulation of drug permeability in vivo.

       Feasibility Study of Caco2-HT29

      To investigate the suitability of a 3D-coculture model, we generated 3D tubules using Caco-2 cells cocultured with the mucus-secreting goblet cell line HT29-MTX-E12 in the OrganoPlate three-lane (Figure 1). Immunofluorescence analysis of Caco2-HT29 was employed to determine the expressions of ZO-1, MUC2, OCLN, and MUC5AC (Figure 5). OCLN and MUC2/MUC5AC were used as markers of tight junctions and mucus, respectively. Fluorescence signals emitted by nuclei, ZO-1, MUC2, OCLN, and MUC5AC were evenly distributed throughout the apical perfusion channel (Figure 5a, b), and nuclei, OCLN and MUC5AC fluorescence covered the tubules (Figure 5c), indicating that Caco-2 cells were distributed over entire tubules and that tight junctions were properly formed. Further, these observations also suggest that goblet cells spread with Caco-2 cells over entire tubules and that MUC2 and MUC5AC secreted from goblet cells covered the apical surface of the tubules, consistent with the findings of others.
      • Gijzen L
      • Marescotti D
      • Raineri E
      • et al.
      An intestine-on-a-chip model of plug-and-play modularity to study inflammatory processes.
      The max projections of fluorescent nuclei, OCLN, and MUC5AC of Caco2-MUC were obtained for comparison (Figure 5d). Stained nuclei and OCLN were distributed over entire tubules, but MUC5AC was not detected in Caco2-MUC.
      Fig. 5
      Fig. 5Confocal imaging of the Caco2-HT29 and Caco2-MUC. (a) Max projections of nuclei, ZO-1, and MUC2 of Caco2-HT29; (b) Max projections of nuclei, OCLN, and MUC5AC of Caco2-HT29; (c) 3D-overlay image of Caco2-HT29 coculture tubules showing nuclei, OCLN, and MUC5AC; (d) Max projections of nuclei, OCLN, and MUC5AC of Caco2-MUC.
      The changes in the TEER of Caco2-HT29 during the introduction of Caco-2 medium and solifenacin in FaSSIF were determined (Figure 6). All Caco2-HT29 tightly maintained their barrier integrities in Caco-2 medium, although their TEER values (mean 331 Ω・cm
      International Council for Harmonisation of Technical Requirements for Pharmaceuticals for Human Use
      Biopharmaceutics Classification System-Based Biowaivers M9.
      , t0) were lower compared with those of Caco2-MUC (mean 527 Ω・cm
      International Council for Harmonisation of Technical Requirements for Pharmaceuticals for Human Use
      Biopharmaceutics Classification System-Based Biowaivers M9.
      , t0) (Figure 6a). These findings are consistent with those of Meaney et al.
      • Meaney C
      • O'Driscoll C.
      Mucus as a barrier to the permeability of hydrophilic and lipophilic compounds in the absence and presence of sodium taurocholate micellar systems using cell culture models.
      and Lozoya-Agullo et al.,
      • Lozoya-Agullo I
      • Araújo F
      • González-Álvarez I
      • et al.
      Usefulness of caco-2/HT29-MTX and Caco-2/HT29-MTX/Raji B coculture models to predict intestinal and colonic permeability compared to Caco-2 monoculture.
      who found that the TEER value of the coculture model was lower compared with that of the monoculture model because of structural variability and high ion permeability of tight junctions formed by goblet cells. Caco2-HT29 tightly maintained their barrier integrities after exposure to solifenacin in FaSSIF as Caco2-MUC (Figure 6b), indicating that the mucus layer secreted from goblet cells protected the coculture membrane tubules from the cytotoxicity of bile acids in FaSSIF.
      Fig. 6
      Fig. 6(a) Time-dependent TEER values of Caco2-MUC and Caco2-HT29; (b) TEER value ratios after 120 min compared to each mean initial value. Mean ± SD (n = 6) for Caco2-MUC, ●; (n = 8) for Caco2-HT29, ; (n = 9) for Caco2-MUC in exposure condition (condition 10), ; (n = 4) for Caco2-MUC in exposure condition (condition 15), .
      The permeabilities of solifenacin dissolved in FaSSIF using Caco2-HT29 (conditions 14 and 15) and Caco2-MUC (conditions 9 and 10) are shown in Figure 7. The mean amounts of permeated solifenacin in the former were equivalent to those in the latter. Meaney et al.
      • Meaney C
      • O'Driscoll C.
      Mucus as a barrier to the permeability of hydrophilic and lipophilic compounds in the absence and presence of sodium taurocholate micellar systems using cell culture models.
      and Lozoya-Agullo et al.
      • Lozoya-Agullo I
      • Araújo F
      • González-Álvarez I
      • et al.
      Usefulness of caco-2/HT29-MTX and Caco-2/HT29-MTX/Raji B coculture models to predict intestinal and colonic permeability compared to Caco-2 monoculture.
      found that the coculture model exhibits lower TEER values and higher drug permeabilities compared with those of the mono-culture model. Our TEER measurement (Figure 6) also indicated that the TEER values of Caco2-HT29 were lower compared with those of Caco2-MUC. On the other hand, Sigurdsson et al.
      • Sigurdsson HH
      • Kirch J
      • Lehr CM.
      Mucus as a barrier to lipophilic drugs.
      and Miyazaki et al.
      • Miyazaki K
      • Kishimoto H
      • Muratani M
      • Kobayashi H
      • Shirasaka Y
      • Inoue K.
      Mucins are involved in the intestinal permeation of lipophilic drugs in the proximal region of rat small intestine.
      reported that the mucus layer may function as an absorption barrier for drugs. Our immunofluorescence analysis (Figures 2 and 5) showed that the Caco2-HT29 expressed not only MUC2 but also MUC5AC, which was not detected in Caco2-MUC. In other words, these results indicate that Caco2-HT29 had richer amount of mucus compared than Caco2-MUC and that the barrier function of Caco2-HT29 may therefore exceed that of Caco2-MUC. Thus, the equivalence of the mean permeated amounts of solifenacin between Caco2-HT29 and Caco2-MUC may be explained by the decrease in drug permeability caused by the abundant mucus layer, which is offset by the increase in drug permeability of the coculture.
      Fig. 7
      Fig. 7Concentrations permeated solifenacin in the basal channel through Caco2-MUC or Caco2-HT29 (conditions 9, 10, 14 and 15). Mean ± SD (n = 4 of solifenacin concentrations in the basal channel)
      Although the mean amounts of permeated solifenacin were equivalent between Caco2-HT29 and Caco2-MUC, variation with Caco2-HT29 was smaller compared with that for Caco2-MUC. The variation in Caco2-MUC may be expected to decrease after the procedure used to apply artificial mucus coating on Caco2-MUC is improved (e.g. optimization of mucus viscosity, concentration, exposure time, and application procedure), but in any case, this finding indicates that the utility of Caco2-HT29 is derived from its ability to reduce technical errors and variability. Further, cell-cell interaction arising from coculture may contribute to protein expression, differentiation, or maturation representative of these interactions in vivo.
      • Goers L
      • Freemont P
      • Polizzi KM.
      Co-culture systems and technologies: taking synthetic biology to the next level.
      Thus, Caco2-HT29 more accurately mimics physiological conditions and substantially advances the prediction of in vivo drug permeability.

      Conclusion

      Caco-2 mono-culture tubules with an artificial mucus layer in an MPS served as an alternative tool for studying in vitro permeability. The artificial mucus layer enabled the use of simulated intestinal fluids, even though components of the fluids were cytotoxic to Caco-2 cells. Further, the OrganoPlate three-lane did high throughput analyses, because forty leak-tight 3D tubules each plate were able to be generated within 5 days. Moreover, our data support the validity of the in vitro Caco-2 membrane permeability system for mimicking in vivo conditions not only from 3D culture reproducibility but also gastrointestinal tract environment perspectives (e.g. introduction of mucus layer and use of simulated intestinal fluid). Therefore, this model will likely facilitate studies to elucidate the mechanisms of drug absorption, including the impact of bile acids, formulation, and the effects of food components and alcohol consumption. Moreover, our assessment of the performance of Caco-2/HT29-MTX coculture tubules supports its utility and expected ability to better mimic in vivo physiological conditions.

      Declaration of Interests

      None. The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Yuki Hagiwara, Harumi Kumagai, Yoshifumi Katakawa, Kei Motonaga and Tomokazu Tajiri are employees of Astellas Pharma Inc.

      Funding Source

      This study and associated task (e.g. Conception or design of this work, Acquisition of data, Analysis of data, Interpretation of data and Writing the report) were funded by Astellas Pharma Inc. Dr. Guy Harris and Dr. Steve Tronick of DMC Corp assistance for English editing and this was funded by Astellas Pharma Inc.``

      Acknowledgments

      The authors thank Dr. Kazuhiro Tetsuka (Astellas Pharma), Dr. Paul Vulto and Dr. Karlijn Wilschut (Mimetas) for providing scientific advices, Prof. Dr. Thomas Hankemeier and Dr. Amy Harms (LACDR-SBP/Analytical BioSciences and Metabolomics, the Netherlands) for obtaining LC-MS assay data, and Dr. Guy Harris and Dr. Steve Tronick of DMC Corp. (www.dmed.co.jp) for editing a draft of this manuscript.

      References

        • Amidon GL
        • Lennernäs H
        • Shah VP
        • Crison JR.
        A theoretical basis for a biopharmaceutic drug classification: the correlation of in vitro drug product dissolution and in vivo bioavailability.
        Pharm Res. 1995; 12: 413-420
        • International Council for Harmonisation of Technical Requirements for Pharmaceuticals for Human Use
        Biopharmaceutics Classification System-Based Biowaivers M9.
        2019 (Available at:)
        • Lennernas H
        • Palm K
        • Fagerholm U
        • Artursson P
        Comparison between active and passive drug transport in human intestinal epithelial (Caco-2) cells in vitro and human jejunum in vivo.
        Int J Pharm. 1996; 127: 103-107
        • Ungell AL.
        In vitro absorption studies and their relevance to absorption from the GI tract.
        Drug Dev Ind Pharm. 1997; 23: 879-892
      1. Ungell AL, Karlsson J Cell Culture in Drug Discovery: An Industrial Perspective. In: Van de Waterbeemd H, Lennernas H, Artursson P, eds. Drug Bioavailability: Estimation of Solubility, Permeability, Absorption and Bioavailability. Wiley-VCH Verlag GmbH & Co. KGaA; 2003:90–131. Mannhold R, Kubinyi H, Folkers G, eds. Methods and Principles in Medicinal Chemistry.

        • Bujard A
        • Sol M
        • Carrupt PA
        • Martel S.
        Predicting both passive intestinal absorption and the dissociation constant toward albumin using the PAMPA technique.
        Eur J Pharm Sci. 2014; 63: 36-44
        • Marino AM
        • Yarde M
        • Patel H
        • Chong S
        • Balimane PV.
        Validation of the 96 well Caco-2 cell culture model for high throughput permeability assessment of discovery compounds.
        Int J Pharm. 2005; 297: 235-241
        • Miyake M
        • Koga T
        • Kondo S
        • Yoda N
        • Emoto C
        • Mukai T
        • Toguchi H.
        Prediction of drug intestinal absorption in human using the Ussing chamber system: A comparison of intestinal tissues from animals and humans.
        Eur J Pharm Sci. 2017; 96: 373-380
        • Rozehnal V
        • Nakai D
        • Hoepner U
        • et al.
        Human small intestinal and colonic tissue mounted in the Ussing chamber as a tool for characterizing the intestinal absorption of drugs.
        Eur J Pharm Sci. 2012; 46: 367-373
        • Watanabe E
        • Takahashi M
        • Hayashi M.
        A possibility to predict the absorbability of poorly water-soluble drugs in humans based on rat intestinal permeability assessed by an in vitro chamber method.
        Eur J Pharm Biopharm. 2004; 58: 659-665
        • Kansy M
        • Senner F
        • Gubernator K.
        Physicochemical high throughput screening: parallel artificial membrane permeation assay in the description of passive absorption processes.
        J Med Chem. 1998; 41: 1007-1010
        • Berggren S
        • Hoogstraate J
        • Fagerholm U
        • Lennernäs H.
        Characterization of jejunal absorption and apical efflux of ropivacaine, lidocaine and bupivacaine in the rat using in situ and in vitro absorption models.
        Eur J Pharm Sci. 2004; 21: 553-560
        • Reis JM
        • Sinkó B
        • Serra CH.
        Parallel artificial membrane permeability assay (PAMPA) - is it better than Caco-2 for human passive permeability prediction?.
        Mini Rev Med Chem. 2010; 10: 1071-1076
        • Lozoya-Agullo I
        • Araújo F
        • González-Álvarez I
        • et al.
        Usefulness of caco-2/HT29-MTX and Caco-2/HT29-MTX/Raji B coculture models to predict intestinal and colonic permeability compared to Caco-2 monoculture.
        Mol Pharm. 2017; 14: 1264-1270
        • Markopoulos C
        • Thoenen F
        • Preisig D
        • et al.
        Biorelevant media for transport experiments in the Caco-2 model to evaluate drug absorption in the fasted and the fed state and their usefulness.
        Eur J Pharm Biopharm. 2014; 86: 438-448
        • Antoine D
        • Pellequer Y
        • Tempesta C
        Biorelevant media resistant co-culture model mimicking permeability of human intestine.
        Int J Pharm. 2015; 481: 27-36
        • Clarysse S
        • Brouwers J
        • Tack J
        • Annaert P
        • Augustijns P.
        Intestinal drug solubility estimation based on simulated intestinal fluids: comparison with solubility in human intestinal fluids.
        Eur J Pharm Sci. 2011; 43: 260-269
        • Meyerhoffer SM
        • McGown LB
        Critical micelle concentration behavior of sodium taurocholate in water.
        Langmuir. 1990; 6: 187-191
        • Coello A
        • Meijide F
        • Rodriguez NE Taro J.
        Aggregation behavior of sodium cholate in aqueous solution.
        Phys Chem. 1993; 97: 10186-10191
        • Elvang PA
        • Hinna AH
        • Brouwers J
        • Hens B
        • Augustijns P
        • Brandl M.
        Bile salt micelles and phospholipid vesicles present in simulated and human intestinal fluids: structural analysis by flow field-flow fractionation/multiangle laser light scattering.
        J Pharm Sci. 2016; 105: 2832-2839
        • Elvang PA
        • Stein PC
        • Bauer-Brandl A
        • Brandl M.
        Characterization of co-existing colloidal structures in fasted state simulated fluids FaSSIF: a comparative study using AF4/MALLS, DLS and DOSY.
        J Pharm Biomed Anal. 2017; 145: 531-536
        • Miller JM
        • Beig A
        • Krieg BJ
        • et al.
        The solubility-permeability interplay: mechanistic modeling and predictive application of the impact of micellar solubilization on intestinal permeation.
        Mol Pharm. 2011; 8: 1848-1856
        • Ingels F
        • Deferme S
        • Destexhe E
        • Oth M
        • Van den Mooter G
        • Augustijns P.
        Simulated intestinal fluid as transport medium in the Caco-2 cell culture model.
        Int J Pharm. 2002; 232: 183-192
        • Yu J
        • Peng S
        • Luo D
        • March JC.
        In vitro 3D human small intestinal villous model for drug permeability determination.
        Biotechnol Bioeng. 2012; 109: 2173-2178
        • VanDussen KL
        • Marinshaw JM
        • Shaikh N
        • et al.
        Development of an enhanced human gastrointestinal epithelial culture system to facilitate patient-based assays.
        Gut. 2015; 64: 911-920
        • Kasendra M
        • Tovaglieri A
        • Sontheimer-Phelps A
        • et al.
        Development of a primary human Small Intestine-on-a-Chip using biopsy-derived organoids.
        Sci Rep. 2018; 8: 2871
        • Kim HJ
        • Huh D
        • Hamilton G
        • Ingber DE.
        Human gut-on-a-chip inhabited by microbial flora that experiences intestinal peristalsis-like motions and flow.
        Lab Chip. 2012; 12: 2165-2174
        • Kim HJ
        • Li H
        • Collins JJ
        • Ingber DE
        Contributions of microbiome and mechanical deformation to intestinal bacterial overgrowth and inflammation in a human gut-on-a-chip.
        Proc Natl Acad Sci U S A. 2016; 113: E7-15
        • Trietsch SJ
        • Naumovska E
        • Kurek D
        • et al.
        Membrane-free culture and real-time barrier integrity assessment of perfused intestinal epithelium tubes.
        Nat Commun. 2017; 8: 262
        • Beaurivage C
        • Naumovska E
        • Chang YX
        • et al.
        Development of a gut-on-a-chip model for high throughput disease modeling and drug discovery.
        Int J Mol Sci. 2019; 20: 5661
        • Moisan A
        • Michielin F
        • Jacob W
        • et al.
        Mechanistic investigations of diarrhea toxicity induced by Anti-HER2/3 combination therapy.
        Mol Cancer Ther. 2018; 17: 1464-1474
        • Matsui H
        • Wagner VE
        • Hill DB
        • et al.
        A physical linkage between cystic fibrosis airway surface dehydration and Pseudomonas aeruginosa biofilms.
        Proc Natl Acad Sci U S A. 2006; 103: 18131-18136
        • Rozee KR
        • Cooper D
        • Lam K
        • Costerton JW.
        Microbial flora of the mouse ileum mucous layer and epithelial surface.
        Appl Environ Microbiol. 1982; 43: 1451-1463
        • Sakata T
        • von Engelhardt W.
        Luminal mucin in the large intestine of mice, rats and guinea pigs.
        Cell Tissue Res. 1981; 219: 629-635
        • Sellers LA
        • Carroll NJ
        • Allen A.
        Misoprostol-induced increases in adherent gastric mucus thickness and luminal mucus output.
        Dig Dis Sci. 1986; 31: 91S-95S
        • Strous GJ
        • Dekker J.
        Mucin-type glycoproteins.
        Crit Rev Biochem Mol Biol. 1992; 27: 57-92
        • Khanvilkar K
        • Donovan MD
        • Flanagan DR.
        Drug transfer through mucus.
        Adv Drug Deliv Rev. 2001; 48: 173-193
      2. Verhoeckx K Cotter P López-Expósito I The Impact of Food Bioactives on Health: In Vitro and Ex Vivo Models [Internet]. Springer, Cham (CH)2015
        • Doroshyenko O
        • Fuhr U.
        Clinical pharmacokinetics and pharmacodynamics of solifenacin.
        Clin Pharmacokinet. 2009; 48: 281-302
        • Yamamoto Y
        • Kumagai H
        • Haneda M
        • et al.
        The mechanism of solifenacin release from a pH-responsive ion-complex oral suspension in the fasted upper gastrointestinal lumen.
        Eur J Pharm Sci. 2020; 142105107
        • Gijzen L
        • Marescotti D
        • Raineri E
        • et al.
        An intestine-on-a-chip model of plug-and-play modularity to study inflammatory processes.
        SLAS Technol. 2020; 25: 585-597
        • Vormann MK
        • Gijzen L
        • Hutter S
        • et al.
        Nephrotoxicity and kidney transport assessment on 3d perfused proximal tubules.
        AAPS J. 2018; 20: 90
        • Ishikawa T
        • Sato T
        • Mohit G
        • Imai Y
        • Yamaguchi T.
        Transport phenomena of microbial flora in the small intestine with peristalsis.
        J Theor Biol. 2011; 279: 63-73
        • Lentle RG
        • Janssen PW.
        Physical characteristics of digesta and their influence on flow and mixing in the mammalian intestine: a review.
        J Comp Physiol B. 2008; 178: 673-690
        • Schindelin J
        • Arganda-Carreras I
        • Frise E
        • et al.
        Fiji: an open-source platform for biological-image analysis.
        Nat Methods. 2012; 9: 676-682
        • Wuyts B
        • Riethorst D
        • Brouwers J
        • Tack J
        • Annaert P
        • Augustijns P.
        Evaluation of fasted state human intestinal fluid as apical solvent system in the Caco-2 absorption model and comparison with FaSSIF.
        Eur J Pharm Sci. 2015; 67: 126-135
        • Meaney C
        • O'Driscoll C.
        Mucus as a barrier to the permeability of hydrophilic and lipophilic compounds in the absence and presence of sodium taurocholate micellar systems using cell culture models.
        Eur J Pharm Sci. 1999; 8: 167-175
        • Li C
        • Liu T
        • Cui X
        • Uss AS
        • Cheng KC.
        Development of in vitro pharmacokinetic screens using Caco-2, human hepatocyte, and Caco-2/human hepatocyte hybrid systems for the prediction of oral bioavailability in humans.
        J Biomol Screen. 2007; 12: 1084-1091
        • Zhu C
        • Jiang L
        • Chen TM
        • Hwang KK.
        A comparative study of artificial membrane permeability assay for high throughput profiling of drug absorption potential.
        Eur J Med Chem. 2002; 37: 399-407
        • Alsenz J
        • Haenel E.
        Development of a 7-day, 96-well Caco-2 permeability assay with high-throughput direct UV compound analysis.
        Pharm Res. 2003; 20: 1961-1969
        • Ingels F
        • Beck B
        • Oth M
        • Augustijns P.
        Effect of simulated intestinal fluid on drug permeability estimation across Caco-2 monolayers.
        Int J Pharm. 2004; 274: 221-232
        • Wang Y
        • Cao J
        • Wang X
        • Zeng S.
        Stereoselective transport and uptake of propranolol across human intestinal Caco-2 cell monolayers.
        Chirality. 2010; 22: 361-368
        • Marin JJ
        • Macias RI
        • Briz O
        • Banales JM
        • Monte MJ
        Bile acids in physiology, pathology and pharmacology.
        Curr Drug Metab. 2015; 17: 4-29
        • Slijepcevic D
        • van de Graaf SF.
        Bile acid uptake transporters as targets for therapy.
        Dig Dis. 2017; 35: 251-258
        • Naumann S
        • Haller D
        • Eisner P
        • Schweiggert-Weisz U
        Mechanisms of interactions between bile acids and plant compounds-a review.
        Int J Mol Sci. 2020; 21: 6495
        • Fossati L
        • Dechaume R
        • Hardillier E
        • et al.
        Use of simulated intestinal fluid for Caco-2 permeability assay of lipophilic drugs.
        Int J Pharm. 2008; 360: 148-155
        • Sigurdsson HH
        • Kirch J
        • Lehr CM.
        Mucus as a barrier to lipophilic drugs.
        Int J Pharm. 2013; 453: 56-64
        • Miyazaki K
        • Kishimoto H
        • Muratani M
        • Kobayashi H
        • Shirasaka Y
        • Inoue K.
        Mucins are involved in the intestinal permeation of lipophilic drugs in the proximal region of rat small intestine.
        Pharm Res. 2019; 36: 162
        • Leal J
        • Smyth HDC
        • Ghosh D
        Physicochemical properties of mucus and their impact on transmucosal drug delivery.
        Int J Pharm. 2017; 532: 555-572
        • Ravi M
        • Paramesh V
        • Kaviya SR
        • Anuradha E
        • Solomon FD.
        3D cell culture systems: advantages and applications.
        J Cell Physiol. 2015; 230: 16-26
        • Goers L
        • Freemont P
        • Polizzi KM.
        Co-culture systems and technologies: taking synthetic biology to the next level.
        J R Soc Interface. 2014; 1120140065