In Vitro Models of the Intestinal Barrier

The Report and Recommendations of ECVAM Workshop 461,2

Reprinted with minor amendments from ATLA 29: 649-668.

Eric Le Ferrec,3 Christophe Chesne,3 Per Artusson,4 David Brayden,5 Gérard Fabre,6 Pierre Gires,7 François Guillou,6 Monique Rousset,8 Werner Rubas9 and Maria-Laura Scarino10
3BIOPREDIC, Technopole Atalante Villejean, 14-18 rue Jean Pecker, 35000 Rennes, France; 4Department of Pharmaceutics, Biomedical Centre, Uppsala University, 75123 Uppsala, Sweden; 5ELAN Biotechnology Research, Biotechnology Building, Trinity College, Dublin 2, Ireland; 6Sanofi Synthelabo Recherche, Departement Phamacologie & Méabolisme, 371 rue du Pr Blayac, 34184 Montpellier, France; 7Rhone Poulenc Rorer, DMPK, Preclinical Department (in vivo & in vitro), 13 Quai Jules Guesdes, 94403 Vitry Sur Seine, France; 8INSERM Ul78, Hopital Paul Brousse, 16 Avenue Paul Vaillant Couturier, 94807 Villejuif France; 9GENENTECH Inc., 1 DNA Way, South San Francisco, CA 94080, USA; 10Istituto Nazionale di Ricerca per gli Alimente e la Nutrizione, Via Ardeatina 546, 00178 Rome, Italy

1ECVAM - The European Centre for the Validation of Alternative Methods. 2This document represents the agreed report of the participants as individual scientists.

Address for correspondence: BIOPREDIC, Technopole Atalante Villejean, 14-18 rue Jean Pecker, 35000 Rennes, France.

Address for reprints: ECVAM, TP 580, JRC Environment Institute, 21020 Ispra (VA), Italy

Preface:

This is the report of the forty-sixth of a series of workshops organised by the European Centre for the Validation of Alternative methods (ECVAM). ECVAM's main goal, as defined in 1993 by its Scientific Advisory Committee, is to promote the scientific and regulatory acceptance of alternative methods which are of importance to the biosciences and which reduce, refine or replace the use of laboratory animals. One of the first priorities set by ECVAM was the implementation of procedures which would enable it to become well-informed about the state-of-the-art of non-animal test development and validation and the potential for the possible incorporation of alternative tests into regulatory procedures. It was decided that this would be best achieved by the organisation of ECVAM workshops on specific topics, at which small groups of invited experts would review the current status of in vitro tests and their potential uses, and make recommendations about the best ways forward (1). In addition, other topics relevant to the Three Rs (reduction, refinement and replacement) concept of alternatives to animal experiments have been considered in several ECVAM workshops.

The workshop on in vitro models of the intestinal barrier was held in Paris, France, on 4-5 March 1999. The principal aims of the workshop were to seek a consensus on the current models of the intestinal barriers and ways to screen for the movement of drugs across this barrier, and to make useful recommendations for the promotion of the Three Rs in this area. The panel of techniques used for the prediction of the relevant parameters is very large, including in vivo and in situ methods, cell culture techniques, and in vivo studies with human volunteers. It was intended to compare these different models, and to delineate their strengths and weaknesses in the various screening situations encountered in the pharmaceutical industry.

Introduction

The intestinal epithelium is a gatekeeper, i.e. it controls the entry of nutrients and xenobiotics (for example, medicines). Knowledge of the absorption and metabolism of these substances at the intestinal mucosal level is of particular importance, since the oral bioavailability of a drug is defined as the fraction of an oral dose that reaches the systemic circulation.

Drug absorption is considered to be a complex transfer process across the intestinal lining, which includes passive diffusion through the paracellular space and/or membranes of absorptive cells, vesicular uptake (endocytosis/pinocytosis), and release at the basolateral space (transcytosis; 2). This transport may or may not be receptor-mediated (or transporter-mediated), entailing uptake across the apical domain with subsequent passive diffusion into the basolateral space (figure 1; 3). Each transport mechanism depends on the physicochemical properties of the absorbed compound, such as its stereochernistry, partition into membranes, molecular weight and/or size, molecular volume, pKa, solubility, chemical stability and charge distribution. Physiological factors such as gastric emptying, gastrointestinal motility, intestinal pH, blood flow, lymph flow (Figure 2), pathological state, drug interactions, nutrition, and mucus dissolution, also need to be considered when evaluating absorption (4-6).

Figure 1: Routes and mechanisms of transport of molecules across the intestinal epithelium

1) Paracellular; 2) transcellular pssive diffusion; 3) transcytosis; 4) carrier-mediated uptake at the apical domain followed by passive diffusion across the basolateral membrane

Figure 2: Factors influencing intestinal absorption

From (3).

Upon ingestion, compounds have to be liberated from their dosage form, which includes dissolution into a complex medium containing numerous compounds (bile salts, ions, lipids, cholesterol and enzymes; 7). This medium can vary considerably, depending on the individual, the intestinal segment, the diet, etc. (8). Measurements of absolute solubility and dissolution are performed routinely during drug development. However, knowledge of intestinal drug dissolution is limited, due to the non-physiological aspects of in vitro systems; for example, instead of using intestinal fluid, buffer systems are commonly employed, which may result in an overestimation or underestimation of the actual events in vivo (912). Since the intestinal fluid is rich in enzymes derived from dying enterocytes and/or the intestinal flora, gastrointestinal stability includes both physicochemical and enzymic stability.

The intestinal cell lining is covered with a viscous and elastic gel produced by the goblet cells of the villous epithelium. The physicochemical properties of this gel can influence the rate of diffusion from the bulk to the site of absorption. Furthermore, metabolic enzymes could be associated with the mucus layer. Thus, the gel is considered to be the first in a series of absorptive barriers (13). The enzymic barrier to oral absorption could consist of several layers, depending on the properties of a compound. Also, as well as the enzymes within the intestinal lumen, there are additional enzymes located in the absorptive cells themselves, such as several cytochrome P450 (CYP) isoforms (Table I). In combination with P-glycoprotein (P-gp), these enzymes are responsible for the intestinal first-pass effect.

Table I: Examples of CYP expression as a function of their human intestinal localisation

CYP Stomach Duodenum Ascending colon Transverse colon Descending colon
1B1 +++++ + ++++ + +
2E1 +++ ++ ++ + +++
3A4 +++ ++++ ++ + ++
3A5 ++++ +++++ ++++ +++ ++

See reference (60).
The number of + signs indicates the relative amount of CYP expressed.

The Current Models

The intestinal mucosa is characterised by the presence of villi that constitute the anatomical and functional unit for nutrient and drug absorption (14). The presence of villi and microvilli provides a massive surface area for absorption (appioximately 250m2 in a human). The mucosa consists of the epithelial layer, the lamina propria (collagen matrix containing blood and lymphatic vessels) and the muscularis mucosa. Therefore, any xenobiotic entering the bloodstream has to pass through the epithelial layer, part of the lamina propria, and the wall of the respective vessel. It is crucial to select an appropriate model for understanding the ratelimiting step in the absorption process.

Three groups of methods have emerged for investigating the principal mechanisms of absorption in animals, namely, in vivo, in situ and in vitro methods. The choice of model depends totally on the questions to be answered with respect to the test compound being studied.

In vivo models

The main advantage of in vivo models is the integration of the dynamic components of the mesenteric blood circulation, the mucous layer and all the other factors that can influence drug dissolution.

The most frequently used animal model is the rat, since it better reflects the human situation with respect to paracellular space and metabolism than the dog, which is different to the human, particularly in relation to metabolism, and overestimates the absorption of paracellularly restricted compounds (15). However, oral studies in rats also have limitations, and tend to provide false-positive results. It has therefore been stated that "the only real model for man is man" (16). Techniques such as "cassette dosing" can give global information on drug bioavailability, including intestinal barrier passage (17). This technique is useful when there are a large number of products to test (high throughput screening [HTS]), when products with good bioavailability are tested, and when organotypic in vitro models are not appropriate.

The disadvantage of in vivo models is that it is impossible to separate the variables involved in the process of absorption, i.e. it is not possible to identify individual rate-limiting factors.

In situ models

The development of stable, vascularly perfused preparations of the small intestine has provided a powerful research tool for the investigation of intestinal transport and metabolism. In this approach, the abdominal cavity of an anaesthetised animal is exposed by laparotomy. The intestinal segment into which the drug solution is introduced can be either a closed loop or an open loop.

In situ methods have significant advantages over in vivo models. For example, bypassing the stomach means that acidic compounds are not likely to precipitate, so dissolution rates do not confuse intestinal drug concentrations and therefore plasma levels. Furthermore, in situ instillation allows the experimenter to assess formulation-independent breakdown in the stomach under acidic conditions. Although the animal has been anaesthetised and surgically manipulated, mesenteric blood flow is intact. However, caution must be taken with the choice of anaesthetic, since it has recently been demonstrated that anaesthesia can have strong effects on intestinal drug absorption (18).

An additional consideration when using in situ techniques is the volume of the luminal drug solution, because this may change due to either absorption or secretion of water. This necessitates the use of non-absorbable or low-absorbable volume-marker compounds, such as radiolabelled poly(ethylene glycol) (PEG) 4000, inulin or mannitol, and fluorescent markers, such as lucifer yellow.

It is noteworthy that the disappearance of drug from the perfusate does not always equate with absorption. Thus, it is prudent to include sampling from the portal vein, in addition to monitoring the change in perfusate concentration. By moving the sampling location from the portal vein to the hepatic vein, additional information about liver first-pass effect can be obtained.

In humans, the intestinal passage of drugs can be studied by the balloon technique (Loc-1-Gut), with drug administration by catheter and analysis of blood sampling (2, 19). With this technique, drug passage can be studied at various intestinal levels. This is a reference technique, but it is expensive to perform and difficult to handle; it cannot be used routinely during drug development (20).

With this model, it is also possible to study compound secretion into the intestinal lumen after intravenous administration, i.e. it is feasible to investigate mediation of export of xenobiotics into the intestinal lumen by P-gp, multidrug resistance-associated protein (MRP) and lung cancer-associated resistance protein (LRP).

In vitro models

Organotypic models
All intestinal cell types (for example, enterocytes, caliciform cells and lymphocytes) are present in organotypic models, which are used to study formulation effects (with the possible use of parenteral lipid emulsions, such as Intralipid®, or bovine serum albumin for drugs having low solubility), intestinal metabolism/stability, and regional differences in permeability. Some studies have shown that permeability to various marker molecules varies along the intestinal canal. In general, permeability decreases in the order: jejunum > ileum > colon (21). The half-lives of these models are short (1-3 hours).

Everted gut sac
The everted gut sac of the rat small intestine can be used to determine kinetic parameters with high reliability and reproducibility (19, 22). Oxygenated tissue culture media and specific preparation techniques ensure tissue viability for up to 2 hours. The technique can be used to study drug transport across the intestine and into the epithelial cells, provided that sensitive detection methods are employed (23). Radiolabelled compounds are most appropriate.

This technique was used in the past to study the transport of macromolecules and liposomes but, more recently, it was used mainly to quantify the paracellular transport of hydrophilic molecules, and to estimate the effects of potent enhancers on their absorption (23). The transport of mannitol, a paracellular marker, shows an apparent permeability (P PP) of 1.5 x 10-5 to 1.7 x 10-5 cm/s. This value is the same as those reported with low-molecular weight hydrophilic drugs in human perfusion studies. The toxicities of potential enhancers can be monitored by studying the release of intracellular enzymes or by histological examination. Molecules that cross the epithelial barrier by a transcellular route have a much higher permeability, which can also be accurately quantified by using the everted sac system.

This kind of model is suitable for measuring absorption at different sites in the small intestine (24), and for performing preliminary experiments on the colon (19, 21). It is also useful for estimating the first-pass metabolism of drugs in intestinal epithelial cells. Also, by using this model (everted or not), it is convenient to study the effect of Pgp on xenobiotic transport through the intestinal barrier.

A potential disadvantage of this approach is the presence of the muscularis mucosa, which is not usually removed from everted sac preparations. Therefore, this model does not reflect the actual intestinal barrier, because compounds under investigation pass from the lumen into the lamina propria (where blood and lymph vessels are found) and across the muscularis mucosa. Thus, the transport of compounds with a propensity to bind to muscle cells might be underestimated.

Isolated and perfused intestinal segments
During the last decade, a wide range of isolated organ systems have been developed for biomedical and pharmaceutical research. The availability of sophisticated equipment, increased manual skills, and the routine use and standardisation of models and protocols, have led to the increased reproducibility and validity of experimental results under circumstances that are virtually "true-to-life". These methods contribute to the reduction of animal experimentation. Figure 3 shows the place of isolated perfused organs in biomedical research, compared with cell cultures of either human or animal origin. The results are recognised as predictive of the in vivo situation including absorption at the organ level (19, 20, 25). Isolated perfased organs have the advantage that the scientist works with an intact organ, where physiological cell-cell contacts and normal intracellular matrixes are preserved (25). The major limitation is the short duration of the experiments that are possible, since changes occur rapidly.

Figure 3: The place of isolated perfused organs in biomedical research

= predictive for; = cell/organs derived from.

Ussing chambers

Ussing chambers were introduced by Ussing & Zehran in 1951 (26) for studying the active transport of sodium as a source of electric current in short-circuited, isolated frog skin. Later on, these chambers were extensively used for the study of ion transport across many types of membrane.

The usefulness of Ussing chambers for intestinal transport studies has long been recognised, and they have also been used to study the intestinal metabolism of xenobiotics (19). In this system, the drug can be exposed at either the mucosal level (apical side of enterocytes) or the serosal level (basolateral side of enterocytes). Furthermore, the simplicity of Ussing chambers makes them an attractive in vitro model system for studying drug transport.

When properly equipped with electrodes, Ussing chambers are useful for studying the effects of compounds on electrophysiological parameters of the intestinal barrier. This type of study may add additional information on the pharmacological behaviour of the test compound (27).

Cell models

The study of absorption mechanisms is best performed in a model that contains only absorptive cells, without the confounding contributions of mucus, the lamina propria and/or the muscularis mucosa. Therefore, much attention is currently paid to the use of epithelial cell cultures for studies of drug transport mechanisms. However, the use of isolated intestinal epithelial cells has been slow to gain popularity, because they are difficult to culture and have limited viability (28-35).

The development of human cell culture systems has been limited by the loss of important in vivo anatomical and biochemical features. Attention has therefore turned to the use of human adenocarcinoma cell lines, such as HT-29 and Caco-2, that reproducibly display a number of properties characteristic of differentiated intestinal cells (36). In addition, the development of these cell line models was paralleled by the development of sensitive and automated measurement techniques (for example, liquid chromatography and mass spectrometry).

The limitations of cell models must not be overlooked, but they offer the advantage of relative simplicity, and are suitable for automated procedures and HTS. These cell lines originated from tumours, and are out of the in vivo physiological environment; therefore, extrapolation of the data to the in vivo situation may be difficult (as is true of most in vitro systems).

Non-intestinal cell systems

Madin Darby canine kidney (MDCK) cells were isolated from a dog kidney by Madin & Darby (37). They are currently used to study the regulation of cell growth, drug metabolism, toxicity and transport at the distal renal tubule epithelial level. MDCK cells have been shown to differentiate into columnar epithelial cells, and to form tight junctions when cultured on semi-permeable membranes.

The use of these cells as a cellular barrier model for assessing intestinal epithelial drug transport was discussed by Cho et al. (38). The results suggested that MDCK cells, like Caco-2 cells, are suitable for molecular-permeability screening studies. Interestingly, these cells do not need 3 weeks in culture before they can be used and, unlike Caco-2 cells, they do not express P-gp.

Small-intestine cell lines from fetal and neonatal rats
Cell lines such as IEC (39) and RIE (40) have been isolated after the repeated cloning of epithelial cells from neonatal rat small intestines. These cell lines show morphological and functional characteristics which suggest that they are derived from crypt cells (41). The IEC line was specifically employed to analyse the role of growth factors in epithelial cell physiology, and for studies on the specific functions of intestinal cells (for example, involving amino acids, glucose and nucleotide transport, or cholesterol synthesis), as well as to perform fundamental studies (36). In contrast, only a few studies have dealt with the passage of test compounds.

Caco-2 cells

Caco-2 cells are the most popular cellular model in studies on passage and transport (Table II). They were derived from a human colorectal adenocarcinoma. In culture, they differentiate spontaneously into polarised intestinal cells possessing an apical brush border and tight junctions between adjacent cells, and they express hydrolases and typical microvillar transporters.

Table II: Characteristics of parental Caco-2 cells

Origin Human colorectal adenocarcinoma
Growth in culture Monolayer epithelial cells
Differentiation 14-21 days after confluence in standard culture medium
Morphology Polarised cells, with tight junctions, apical, brush border
Electrical parameters High electrical resistance
Digestive enzymes Typical membranous peptidases and disaccharidases of the small intestine
Active transport Amino acids, sugars, vitamins, hormones
Membrane ionic transport Na+/K+ ATPase, H+/K+ ATPase, Na+/H+ exchange, NaI/K+/Cl- co-transport, apical Cl- channels
Membrane non-ionic transporters Permeability-glycoprotein, multidrug resistant associated protein, lung cancer associated resistance protein
Receptors Vitamin B12, vitamin D3, epidermal growth factor, sugar transporters (GLUT1, GLUT3, GLUT5, GLUT2, SGLT1)

This cell line was first used as a model for studying differentiation in the intestinal epithelium, and later for estimating the relative contributions of paracellular and transcellular passage in drug absorption.

Caco-2 cells, despite their colonic origin, express in culture the majority of the morphological and functional characteristics of small intestinal absorptive cells, including phase I and phase II enzymes, detected either by measurement of their activities toward specific substrates, or by immunological techniques.

However, CYP3A, which is present in almost all intestinal cells, is very weakly expressed in Caco-2 cells. Treatment with 1a,25-dihydroxyvitamin D3, an inducer of CYP3A4 at the mRNA level, and transfection of CYP3A4 cDNA, are two ways of increasing CYP3A expression levels in Caco-2 cells. However, these expression levels do not reach the levels observed in vivo (42, 43). With regard to phase II enzymes, Caco-2 cells express N-acetyl transferase and glutathione transferase activity.

In addition, many groups have demonstrated the presence of P-gp activity in Caco-2 monolayers, at levels higher than those found in the human colon in vivo (44, 45). The interpretation of transport data is therefore confusing, and is not always in agreement with in vivo observations, even when P-gp is blocked by specific inhibitors. Other membrane transporters, i.e. MRP and LRP, are also expressed (Table II).

Interestingly, TC7 cells, isolated after the exposure of Caco-2 cells to methotrexate express CYP3A at a higher level than their parental counterparts (Table III; 46). TC7 cells offer marked advantages over parental Caco-2 cells, because they express CYP3A, actively transport taurocolic acid, and have lower levels of P-gp compared with the parent Caco-2 cells. Several studies have demonstrated that TC7 cells are a good alternative to the Caco-2 parental line for drug transport studies (49).

Table III: Expression of mRNAs of various CYPs in Caco-2 and TC7 cells

CYP Caco-2 TC7
1A1 + +
2B6 + -
2B6 + +
2C8/19 + +
2D6 + -
2E1 + +
3A4 - +
3A5 + +
3A7 - -

Caco-2 cells grow as a monolayer and differentiate on a semi-permeable membrane. Thus, separating the apical compartment from the basolateral compartment, which correspond to the intestinal lumen side and the serosal side, respectively, is possible (Figure 4). The complete morphological and functional differentiation of Caco-2 requires 3 weeks in culture under the conditions described in Table IV.

Figure 4: A schematic representation of culture of Caco-2 cells on a microporous filter

Table IV: Culture conditions for use of Caco-2 cells in passage assays

Inactivated serum in the medium on the apical site 20%
Inactivated serum in the medium on the basolateral site 20%
Coating Type 1 collagen (not required with 0.4µM filters)
Not added under basal culture conditions
Additive Non-essential amino acids
Glucose
Glutamine
Antibiotics		 1%
25mM
2mM
Streptomycin (100gg/1) (optional)
Penicillin (100mU/ml) (optional)
C02 10% or 5%
pH Usually 7.4 at both sides or 7.4 in basolateral side and 6.5 in apical side (57)
Cell density at seeding 2.5 x 105 - 4 x 105 cells/cm2
Number of passages 25-100

Before using this model, various controls (summarised in Table V) have to be performed.

Table V: Main controls for cell monolayers

Transepithelial electrical resistance (TEER) Cell monolayer integrity. TEER measurement, depending on the filter area, can reveal a toxicity or an opening of tight junctions induced by the drug. For CaCo-2 cells, TEER values are 260-420 ohms/cm2 (62-71); human intestine: 12-69 (64), rat ileum: 35 ± 4.9 (62); rat colon: 100 ± 26 (62).
Differentiation markers By determination of sucrase-isomaltase, aminopeptidase and alkaline phosphatase activities
Morphological differentiation By electron microscopy
Permeability-glycoprotein (P-gp) expression By Western blotting or measurement of P-gp activity and by using immunohistochemistry
Absence of contamination by mycoplasma Detection by standard microbiological methods
Integrity of the cell monolayer by measuring permeability of test compounds Mannitol, PEG 4000, lucifer yellow

Calculation of the permeability coefficient and interpretation of results

In most cases, three parameters are determined: the apparent permeability (Papp), the flux, and the effective permeability (Peff).

The coefficient of permeability is used in in vitro techniques, and the values are expressed as cm/second and are calculated by the following equation:

Papp =      dQ      =      VdC     
dT A C0 dT A C0

Equation 1

where V = sample volume (ml), dC = concentration variations, dQ = quantity variations, dT = time variations, CO = the initial concentration in the donor compartment, and A = exposed surface (cellular monolayer in cm2). In the above formula, dQ/(dT * A) represents the mass transfer per unit time and unit surface across the monolayer.

The effective permeability is used in in situ and organotypic techniques, and the values are calculated as follows:

Peff =      (Cin - Cout) Qin     
Cout 2πRL

Equation 2

where Cin and Qin = the concentration and the quantity of compound at the entry of the intestinal segment, Cout = the concentration at the exit, R = the radius, and L = the length of the intestinal segment.

The accuracy of measurements naturally depends upon the precision of dQ (Equation 1) or Cout (Equation 2). Moreover, Co (Equation 1) and Cin (Equation 2) are limited by the solubility of the drug, the analytical sensitivity, and the effects of high drug concentrations on epithelial integrity.

The classical method for studying transport from the apical to the basolateral sides relies on the transfer of the cell culture support at different times, as described in Figure 5.

Figure 5: A schematic representation of the method used for measurement of transmembrane passage (apical [A] to basolateral [B]) of a drug in Caco-2 cells

When samples are removed from the receiver chamber and replaced with control medium, a correction must be made to account for the corresponding dilution. The absence of effects of the drug on permeability chardcteristics and/or electrical parameters has to be verified (TEER measurement).

A linear relationship between the A-to-B flux and the B-to-A flux at several concentration levels is an indication of a passive diffusion mode. A tight correlation between the flux (or Papp) and TEER is of significance in implying the paracellular route of drug transport. When the flux can be saturated, the occurrence of an energy-dependent and/or transporter-mediated transport is likely. In this case, the Michaëlis-Menten principles can be applied (for example, Km, Vmax). Many pharmacological tools exist that can explain transport mechanisms, for example, P-gp involvement, paracellular passage (Table VI).

Table VI: Pharmacological agents for use in passage studies

Mechanism Agents Effects
P-gp involvement Inhibitor Verapamil < 0.5mM in apical (A) and basolateral (B) sides Increase the transport from the apical side to the basolateral side (if the drug is added on the apical side)
Quinidine 0.5-1mM in A and B
Cyclosporin A 50µ in A and B
Substrate Rhodamine 123 1mM Efflux of rhodamine at the apical side
MRP involvement Inhibitor As for substrates    
Substrate Leucotriene C4, S-2,4-dinitrophenyl glutathione, PAH, doxorubicin, etoposide, vinblastine, methotrexate
LRP involvement Inhibitor      
Substrate Anthracycline
Paracellular transport (by action on tight junctions) EGTA, cytochalasine Increase the transport, if paracellular

P-gp = permeability glycoprotein, MRP = multidrug resistance associated protein, LRP = lung cancer associated resistance protein.

Comparison of the models

Consistently good absorption is a key factor in selecting new drug candidates for development. In the discovery stage, drug absorption studies can be performed only in laboratory animals and/or in in vitro systems where the absorption process can be characterised both qualitatively and quantitatively.

The various experimental protocols for predicting the fraction absorbed in humans from permeability coefficients have their own advantages and limitations (Table VII). The choice of model depends on compound availability, the stage of the project, and the questions to be answered. Drug absorption consists of passage through the epithelia] layer and uptake into the bloodstream. In some models, such as the Ussing chamber or the everted gut sac, the drug must traverse all the intestinal wall, part of which might be rate-limiting in vitro, but not involved in vivo, thus leading to an underestimation of transport values. Cell lines, such as Caco-2, can be considered to be good models that mimic the physiological situation, as the drug traverses only the epithelial layer. However, it must be noted that one of the great differences between the in vivo situation and cell lines is the absence of mucus, which might limit absorption, especially that of lipophilic drugs. Moreover, cell monolayers contain only one cell type. Co-cultures consisting of cell lines with enterocytic markers (such as Caco-2 cells) and cell lines with mucus secretory functions (such as HT29-MTX) have been proposed (47).

Table VII: Advantages and limitations of the various models

Techniques Advantages Limitations
In situ technique (rat) Integrates passage and metabolism aspects.
All physiological factors that influence passage are present.
Studies of absorption in particular sites of the intestine are possible.
Studies of direct effects of the drug on intestinal absorption are possible.		 Used in development, but not routinely.
The increase of luminal hydrostatic pressure during the experiment can influence intestinal permeability.
It is an animal model.
Human intestinal perfusion It is a reference model.
All physiological factors that influence passage are present.
Allows studies in humans.
Good correlation with pharmacokinetic studies.		 A difficult technique with local anaesthesia at the time of catheter introduction.
Not used either in development or routinely.
Intestinal gut sac (rat) All cell types and the mucous layer are present.
A relatively fast and inexpensive technique.
Can be used for mechanism of absorption or formulation studies.		 Not a perfused model.
The drug must cross the whole intestinal wall.
It is an animal model.
Ussing chamber (human, rat) Drug absorption and passage at specific intestinal sites are possible.
The test drug can be added on either the apical or the basolateral side.
Metabolism studies are possible.
A human and animal model.		 Measurement techniques must be sensitive, since the drugs are diluted in the diffusion chambers.
Cell viability is limited.
The drug must cross the whole intestinal wall.
Availability of human tissue is limited.
Not used for screening.
Cell culture Caco-2 cells Relatively fast and simple method.
A flexible model.
Can be used for mechanism transport studies.
The test drug can be exposed at the apical or the basolateral side.
Human 		Physiological factors that influence passage are not present (mucous, bile salts, cholesterol).
A static model.
Cells have a tumoral origin.
A model with only one cell type.
Influence of P-gp difficult to estimate
TC7 cells Express CYP3A.
Growth faster than that of Caco-2 cells.
Need less glucose than Caco-2 cells.		  
MDCK cells Fast and simple method.
Can be used for screening testing.
Can be used for measurement of passive diffusion.
Do not express P-gp.		 No an intestinal model.
It is an animal model.

P-gp = permeability glycoprotein; MDCK = Madin Darby canine kidney.

Caco-2 monolayers have emerged as a suitable model for studying drug absorption (Table VIII). Most studies with Caco2 monolayers were performed to determine whether a drug is actively or passively transported across the intestinal epithelium, and to provide new insights into the regulation of drug transport.

Table VII: Examples of mechanistic and drug absorption studies using Caco-2 cells

Route of passage Factor influencing drug absorption Model used References
Paracellular Molecular size
Flexibility of drug geometric structure		 Caco-2
Caco-2		 (65)
(62)
Transcellular Lipophilicity
Hydrogen bond		 Caco-2
Caco-2		 (63,71)
(66,67)

However, the predictions of absorption in humans based on permeability data obtained with Caco-2 monolayers from different laboratories are not satisfactory. For example, Artursson & Karlsson (48) have compared the passage of 20 structurally unrelated drugs in Caco-2 cell monolayers with the extent of absorption in humans after oral administration, and concluded that drugs having a complete absorption in humans were found to have a high permeability coefficient (Papp > 1 x 10-6cm/s) in Caco-2 cells, whereas poorly absorbed drugs had a low permeability coefficient (Papp < 1 x 10-7cm/s). However, similar studies (49, 50) did not lead to the same results (Table IX). The reasons for such a discrepancy between laboratories in the reported permeability values are not clear.

Such results indicate that, although Caco-2 monolayers are a useful model for ranking drugs according to their permeability, they cannot be used to quantitatively predict human absorption in vivo. As an example, Gan et al. (51) reported that the Caco-2 permeability coefficient of ranitidine was 1 x 10-1cm/s (i.e. corresponding to a low permeability), whereas the human bioavailability of this drug is good (50-70%).

Other attempts have been made to compare in vitro and in vivo drug permeability (Table X). Comparison of the permeability coefficients of a series of drugs by using Caco-2 cells and the double-balloon technique in the human jejunum (52, 53), showed that the permeability of drugs with complete absorption differed by twofold to four-fold between the in vitro and in situ models, whereas the permeability of drugs with poor absorption differed as much as 30-fold to 80fold. Thus, measurement of permeability by using Caco-2 cells allows only a qualitative comparison. Moreover, correlations between the different techniques are better when the mode of passage is passive diffusion (6, 25, 50, 54). When a receptor-mediated molecular transport is involved, the calculation of correlation coefficients is difficult, or even impossible.

Table X: Correlations between various techniques

Technique Number of compounds tested Comment Correlation Currently tested compounds
Caco-2 vs MDCK cells (55) 55 The Papp correlation coefficient calculated by these two techniques is 0.79. Moreoever, comparison of the Papp values calculated in MDCK and Caco-2 cells in relation to the human org route is 0.58 and 0.54, respectively, indicating that both cells are suitable. ++ D-Glucose, poly(ethylene glycol) (PEG), furosemide, propranolol, atenolol, metoprolol, terbutaline, enalapril, L-dopa, D-mannitol
TC7 vs Caco-2 cells (68) 20 Based on morphological and biochemical parameters and also on transport characteristics, it appears that TC7 cells are a reliable alternative to Caco-2 parental lines for transport studies. +++
Caco-2 cells vs human oral route (48, 68, 69) 20 Caco-2 and TC7 cells are used for the prediction of passive human passage +++
TC7 cells vs human oral route (68) 20 Caco-2 and TC7 cells are used for the prediction of passive human passage +++
"Rat" Ussing chamber vs "human" Ussing chamber (70) 12 The "rat" Ussing chamber technique is especially useful for screening substances having local pharmacological and transporter-mediated effects. ++
"Rat" Ussing chamber vs oral route in humans (70) 12   +++
"Rat" in situ perfusion vs "human" in situ perfusion (53) 10 High correlation exists between the two models for passively absorbed molecules. The two models can be used to predict absorption in humans. A marker should be included in the "rat model" to follow viability. +++
"Rat" in situ perfusion vs human oral route (50) 16 For small organic molecules +++
"Rat" intestinal gut sac vs human oral route (50) 12 For preliminary screening. +

+ = limited correlation, ++ = median correlation, +++ = high correlation. MDCK = Madin Darby canine kidney. Papp = apparent permeability.

Because the properties of Caco-2 monolayers can vary with time in culture, passage number and culture medium composition, it is important to include a reference drug when screening for the permeability of test drugs.

It should also be borne in mind that monolayers constitute a two-dimensional system, whereas the intestinal mucosa is a three-dimensional one, i.e. it is not flat, but convoluted because of villi and folds. Depending on the actual available surface area for absorption, the estimates are made difficult by the calculation of the permeability coefficient, where the surface area of the tube (the balloon technique) is used. The actual surface might be larger; hence, the reported permeability would be too high.

Since in vitro models cannot give quantitative predictions of drug absorption in humans, another possibility is to use animal models (49). This is based on the assumption that the membrane permeability of drugs is not species-dependent (55). Since the composition of the plasma membrane of intestinal epithelial cells is similar across species, the permeability of drugs (simple diffusion) across the wall of the gastrointestinal tract could be expected to be similar (Table X). Drug absorption in humans can be extrapolated reasonably well from animal data, when information on first-pass metabolism is also available (56). Bioavailability, however, differs substantially among species, presumably as a result of species differences in the scale of first-pass metabolism.

The United States Pharmacopeial Convention (USP) has proposed two schemes for predicting in vivo and in vitro correlations. Figure 6 illustrates in vivo/in vitro correlations for molecules classified according to solubility and permeability. Figure 7 illustrates correlations between in vivo and in vitro data. It shows that highly permeable molecules (classes I and II) exhibit a high in vivo/in vitro correlation, whereas poorly permeable molecules (classes III and IV) show a low in vivo/in vitro correlation.

Figure 6: Biopharmaceutical classification system (57)

See ref. 57.

Figure 7: In vitro/in vivo correlation

aIf the dissolution rate is slower than the gastric emptying rate, otherwise limited or no correlation
bIf in vivo and in vitro dissolution rates are similar (57).

The USP proposes the selection of reference molecules with known Papp values (methotrexate: 1.2 x 10-6cm/s; propranolol hydrochloride: 28 x 10-6cm/s; testosterone: 73 x 10-6cm/s). If the Papp values obtained with Caco-2 cells are the same as the Papp values given by the USP (± 20%), the model can be considered to be valid (57). However, because the USP defines rigid limits, such recommendations are not accepted by all users. Many researchers prefer to use a collection of laboratory values, with internal specifications and criteria of acceptance for each assav system. Another approach chosen by some researchers is to delineate the physicochemical properties that favour intestinal absorption, and they have developed computational methods for their prediction. The best-known method is probably the "rule of five" designed by Lipinski and coworkers (58) from an analysis of 2245 drugs. As implemented in their system, the "rule of five" generates an alert (an indication of possible absorption problems) for compounds, when: 1) there are more than five hydrogenbond donors (expressed as the sum of hydroxyl and primary amine groups); 2) the molecular mass is over 50; 3) the Log P is over 5 (or MlogP is over 4.15); 4) there are more than ten hydrogen-bond acceptors (expressed as the sum of nitrogen and oxygen atoms); and 5) compound classes that are substrates for biological transporters are exceptions to the rule.

The "rule of five" method was originally proposed because many of the large numbers of hit compounds selected by HTS did not possess "drug-like" properties. Obviously, the growing number of publications in this field indicate that methods for predicting "drug-likeness" are already having a major impact on the design and selection of compounds in pharmaceutical companies. The routine use of experimental absorption systems in the pre-screening of compounds is providing valuable data, and should permit the development of improved models for drug absorption. The value of such experimental and theoretical systems will become apparent when drug development times are reduced.

 

Conclusions and Recommendations

  1. Knowledge on P-gp expression in human intestine and cell lines should be improved.
  2. Cell lines other than Caco-2 (for example, HT29, TC7, 2/4/A1) should be more fully investigated.
  3. The absence of mucus in cell lines is a problem. Co-culturing cell lines expressing enterocytic markers (for example, Caco-2 cells) with cell lines exhibiting mucus secretory properties (for example, the HT-29 cell line [591) could permit this limitation to be overcome.
  4. Internet sites and software should be created and developed to facilitate communication among researchers concerning the intestinal barrier, and also to improve in silico predictions.
  5. Simple and powerful cell-culture systems should be commercialised (for example, dynamic systems, three-dimensional models, automated systems).
  6. Suitable conditions permitting the safe and ethical preservation, transport and use of human intestinal fragments (from tissue banks) should be established.
  7. Comparison of data from different laboratories on drug passage across the intestinal barrier remains unstandardised. A consensus should therefore be sought on the optimal experimental conditions (medium, reference substances, etc.), and harmonised with the FDA and USP texts.

References

  1. Anon. (1997). ECVAM News and Views. ATLA 22: 7-11.
  2. Lennernas, H. (1998). Human intestinal permeability. Journal of Pharmaceutical Sciences 87: 403-410.
  3. Tsuji, A. & Tamai, 1. (1996). Carrier-mediated intestinal transport of drugs. Pharmaceutical Research 13: 963-977.
  4. Pade, V. & Stavchansky, S. (1998). Link between drug absorption solubility and permeability measurements in Caco-2 cells. Journal of Pharmaceutical Sciences 87: 1604-1607.
  5. Levin, R.J. (1982). Assessing small intestinal function in health and disease in vivo and in vitro. Scandinavian Journal of Gastroenterology 74: 31S-51S.
  6. Wikman, A., Karlsson, J., Carlstedt, I. & Artursson, P. (1993). A drug absorption model based on the mucus layer producing human intestinal goblet cell line HT29-H. Pharmaceutical Research 10: 843-852.
  7. Welling, P.G. (1984). Interactions affecting drug absorption. Clinical Pharmacokinetics 9: 404-434.
  8. Tietze, K.J. & Putcha, L. (1994). Factors affecting drug bioavailability in space. Journal of Clinical Pharmacology 34: 671-676.
  9. Galia, E., Nicolaides, E., Horter, D., Lobenberg, R., Reppas, C. & Dressman, J.B. (1998). Evaluation of various dissolution media for predicting in vivo performance of class I and II drugs. Pharmaceutical Research 15: 698-705.
  10. Ansoborlo, E., Henge-Napoli, M.H., Chazel, V., Gibert, R. & Guilmette, R.A. (1999). Review and critical analysis of available in vitro dissolution tests. Health Physics 77: 638-645.
  11. Dressman, J.B., Amidon, G.L., Reppas, C. & Shah, V.P. (1998). Dissolution testing as a prognostic tool for oral drug absorption: immediate release dosage forms. Pharmaceutical Research 15: 11-22.
  12. Macheras, P. & Argyrakis, P. (1997). Gastrointestinal drug absorption: is it time to consider heterogeneity as well as homogeneity? Pharmaceutical Research 14: 842-847.
  13. Winne, D. (1979). Rat jejunum perfused in situ: effect of perfusion rate and intraluminal radius on absorption rate and effective unstirred layer thickness. Naunyn Schmiedeberg's Archives of Pharmacology 307: 265-274.
  14. Pacha, J. (2000). Development of intestinal transport function in mammals. Physiological Reviews 80: 1633-1667.
  15. Kararli, T.T. (1995). Comparison of the gastrointestinal anatomy, physiology, and biochemistry of humans and commonly used laboratory animals. Biopharmaceutics and Drug Disposition 16: 351-380.
  16. Wilding, I.R., Kenyon, C.J. & Hooper, G. (2000). Gastrointestinal spread of oral prolonged-release mesalazine microgranules (Pentasa) dosed as either tablets or sachet. Alimentary Pharmacology and Therapeutics 14: 163-169.
  17. White, R.E. & Manitpisitkul, P. (2001). Pharmacokinetic theory of cassette dosing in drug discovery screening. Drug Metabolism and Disposition 29: 957-966.
  18. Yuasa, H., Matsuda, K. & Watanabe, J. (1993). Influence of anesthetic regimens on intestinal absorption in rats. Pharmaceutical Research 10: 884-888.
  19. Barthe, L., Woodley, J. & Houin, G. (1999). Gastrointestinal absorption of drugs: methods and studies. Fundamental and Clinical Pharmacology 13: 154-168.
  20. Acra, S.A. & Ghishan, F.K. (1991). Methods of investigating intestinal transport. Journal of Parenteral and Enteral Nutrition 15: 93S-98S.
  21. Davis, G.R., Santa Ana, C.A., Morawski, S.G. & Fordtran, J.S. (1982). Permeability characteristics of human jejunum, ileum, proximal colon and distal colon: results of potential difference measurements and unidirectional fluxes. Gastroenterology 83: 844-850.
  22. Barthe, L., Woodley, J.E, Kenworthy, S. & Houin, G. (1998). An improved everted gut sac as a simple and accurate technique to measure paracellular transport across the small intestine. European Journal of Drug Metabolism and Pharmacokineties 23: 313-323.
  23. Leppert, P.S. & Fix, J.A. (1994). Use of everted intestinal rings for in vitro examination of oral absorption potential. Journal of Pharmaceutical Sciences 83: 976-998.
  24. Chowhan, Z.T. & Amaro, A.A (1977). Everted rat intestinal sacs as an in vitro model for assessing absorptivity of new drugs. Journal of Pharmaceutical Sciences 66: 1249-1253.
  25. Levet-Trafit, B., Gruver, M.S., Marjanovic, M. & Chou, R.C. (1996). Estimation of oral drug absorption in man based on intestine permeability in rats. Life Sciences 58: L359-363.
  26. Ussing, H.H. & Zerahn, K. (1951). Active transport of sodium as the source of electric current in the short-circuited isolated frog skin. Acta Physiologica Scandinavia 23: 110-127, 1951.
  27. Soderholm, J.D., Hedman, L., Artursson, P., Franzen, L., Larsson, J., Pantzar, N., Permert, J. & Olaison, G. (1998). Integrity and metabolism of human ileal mucosa in vitro in the Ussing chamber. Acta Physiologica Scandinavia 162: 47-56.
  28. Friedman, E.A. (1981). Differential response of premalignant epithelial cell classes to phorbol ester tumor promoters and to deoxycholic acid. Cancer Research 41: 4588-4599.
  29. Roediger, W.E. & Truelove, S.C. (1979). Method of preparing isolated colonic epithelial cells (colonocytes) for metabolic studies. Gut 20: 484-488.
  30. Ishii, S., Steele, G., Ford, R., Paliotti, G., Thomas, P., Andrews, C., Hansen, H.J., Goldenberg, D.M. & Jessup, J.M. (1994). Normal colonic epithelium adheres to carcinoembryonic antigen and type IV collagen. Gastroenterology 106: 1242-1250.
  31. Goldstein, J.L., Sahi, J., Bhuva, M., Layden, T.J. & Rao, M.C. (1994). Escherichia coli heat-stable enterotoxin-mediated colonic Cl- secretion is absent in cystic fibrosis. Gastroenterology 107: 950-956.
  32. Clausen, M.R. & Mortensen, P.B. (1995). Kinetic studies on colonocyte metabolism of short chain fatty acids and glucose in ulcerative colitis. Gut 37: 684-689.
  33. Gibson, P. & Rosella, 0. (1995). Interleukin 8 secretion by colonic crypt cells in vitro: response to injury suppressed by butyrate and enhanced in inflammatory bowel disease. Gut 37: 536-543.
  34. Mahida Y.R., Makh, S., Hyde, S., Gray, T. & Borriello, S.P. (1996). Effect of Clostridium difficile toxin A on human intestinal epithelial cells: induction of interleukin 8 production and apoptosis after cell detachment. Gut 38: 337-347.
  35. Branka, J.E., Vallette, G., Jarry, A., Bou-Hanna, C., Lemarre, P., Van, P.N. & Laboisse, C.L. (1997). Early functional effects of Clostridium difficile toxin A on human colonocytes. Gastroenterology 112: 1887-1894.
  36. Kedinger, M., Haffen, K. & Simon-Assmann, P. (1987). Intestinal tissue and cell cultures. Differentiation 36: 71-85.
  37. Gaush, C.R., Hard, W.L. & Smith, T.E. (1966). Characterization of an established line of canine kidney cells (MDCK). Proceedings of the Society for Experimental Biology and Medicine 122: 931-935.
  38. Cho, M.J., Thompson, D.P., Cramer, C.T., Vidmar, T.J. & Scieszka, J.F (1989). The Madin Darby canine kidney (MDCK) epithelial cell monolayer as a model cellular transport barrier. Pharmaceutical Research 6: 71-77.
  39. Quaroni, A., Wands, J., Trelstad, R.L. & Isselbacher, K.J. (1979). Epithelioid cell cultures from rat small intestine: characterization by morphologic and immunologic criteria. Journal of Cell Biology 80: 248-265.
  40. Blay, J. & Brown, K.D. (1984). Characterization of an epithelioid cell line derived from rat small intestine: demonstration of cytokeratin filaments. Cell Biology International Reports 8: 551-560.
  41. Quaroni, A. & Beaulieu, J.F. (1997). Cell dynamics and differentiation of conditionally immortalized human intestinal epithelial cells. Gastroenterology 113: 1198-1213.
  42. Fisher, J.M., Wrighton, S.A., Calamia, J.C., Shen, D.D., Kunze, K.L. & Thummel, K.E. (1999). Midazolarn metabolism by modified Caco-2 monolayers: effects of extracellular protein binding. Journal of Pharmacology and Experimental Therapeutics 289: 1143-1150.
  43. Hu, M., Li, Y., Davitt, C.M., Huang, S.M., Thummel, K., Penman, B.W. & Crespi, C.L. (1999) Transport and metabolic characterization of Caco-2 cells expressing CYP3A4 and CYP3A4 plus oxidoreductase. Pharmaceutical Research 16: 1352-1359.
  44. Hunter, J., Hirst, B.H. & Simmons, N.L. (1993). Drug absorption limited by P-glycoprotein-mediated secretory drug transport in human intestinal epithelial Caco-2 cell layers. Pharmaceutical Research 10: 743-749.
  45. Burton, P.S., Conradi, R.A., Hilgers, A.R. & Ho, N.F. (1993). Evidence for a polarized efflux system for peptides in the apical membrane of Caco-2 cells. Biochemical and Biophysical Research Communications 190: 760-766.
  46. Carriere, V., Lesuffleur, T., Barbat, A., Rousset, M., Dussaulx, E., Costet, P., de Waziers, I., Beaune, P. & Zweibaum, A. (1994). Expression of cytochrome P-450 3A in HT29-MTX cells and Caco-2 clone TC7 (published erratum appears in FEBS Letters 362: 99 [1995]). FEBS Letters 355: 247-250.
  47. Hilgendorf, C., Spahn-Langguth, H., Regardh, C.G., Lipka, E., Amidon, G.L. & Langguth, P. (2000). Caco-2 versus Caco-2/HT29-MTX co-cultured cell lines: permeabilities via diffusion, inside- and outside-directed carrier-mediated transport. Journal of Pharmaceutical Sciences 89: 63-75.
  48. Artursson, P. & Karlsson, J. (1991). Correlation between oral drug absorption in humans and apparent drug permeability coefficients in human intestinal epithelial (Caco-2) cells. Biochemical and Biophvical Research Communications 175: 880-885.
  49. Rubas, W., Cromwell, M.E., Shahrokh, Z., Villagran, J., Nguyen, T.N., Wellton, M., Nguyen, T.H. & Mrsny, R.J. (1996). Flux measurements across Caco-2 monolayers may predict transport in human large intestinal tissue. Journal of Pharmaceutical Sciences 85: 165-169.
  50. Stewart, B.H., Chan, O.H., Lu, R.H., Reyner, E.L., Schmid, H.L., Hamilton, H.W., Steinbaugh, B.A. & Taylor, M.D. (1995). Comparison of intestinal permeabilities determined in multiple in vitro and in situ models: relationship to absorption in humans. Pharmaceutical Research 12: 693-699.
  51. Gan, L.S., Hsyu, P.H., Pritchard, J.F. & Thakker, D. (1993). Mechanism of intestinal absorption of ranitidine and ondansetron: transport across Caco-2 cell monolayers. Pharmaceutical Research 10: 1722-1725.
  52. Lindahl, A., Sandstrom, R., Ungell, A.L., Abrahamsson, B., Knutson, T.W., Knutson, L. & Lennernas, H. (1996). Jejunal permeability and hepatic extraction of fluvastatin in humans. Clinical Pharmacology and Therapeutics 60: 493-503.
  53. Fagerholm, U., Johansson, M. & Lennernas, H. (1996). Comparison between permeability coefficients in rat and human jejunum. Pharmaceutical Research 13: 1336-1342.
  54. Fine, K.D., Santa Ana, C.A., Porter, J.L. & Fordtran, J.S. (1993). Effect of D-glucose on intestinal permeability and its passive absorption in human small intestine in vivo. Gastroenterology 105: 1117-1125.
  55. Irvine, J.D., Takahashi, L., Lockhart, K., Cheong, J., Tolan, J.W., Selick, H.E & Grove, J.R. (1999). MDCK (Madin Darby canine kidney) cells: a tool for membrane permeability screening. Journal of Pharmaceutical Sciences 88: 28-33.
  56. Lin, J.H. & Lu, A.Y. (1997). Role of pharmacokinetics and metabolism in drug discovery and development. Pharmacological Reviews 49: 403-449.
  57. USP (1998). USP Pharmacopeial forum. US Pharmaeopeia 24: 6015-6023.
  58. Lipinski, C.A., Lombardo, F., Dominy, B.W. & Feeney, P.J. (2001). Experimental and computational approaches to estimate solubility and permeability in drug discovery and development settings. Advanced Drug Delivery Reviews 46: 3-26.
  59. Walter, E., Janich, S., Roessler, B.J., Hilfinger, J.M. & Amidon, G.L. (1996). HT29-MTX/Caco-2 co-cultures as an in vitro model for the intestinal epithelium: in vitro-in vivo correlation with permeability data from rats and humans. Journal of Pharmaceutical Sciences 85: 1070-1076.
  60. Finnström, N., Thörn, M., Lööf, L. & Rane, A. (1998). Gene expression of cytochromes P450 in different parts of the human gastrointestinal tract. 12th International Symposium "Microsomes and Drug Oxidation", Montpellier (France), p. 343. Stockholm, Sweden: Alpha Visa IST/D098.
  61. Raeissi, S.D., Hidalgo, I.J., Segura-Aguilar, J. & Artursson, P (1999). Interplay between CY-P3A-mediated metabolism and polarized efflux of terfenadine and its metabolites in intestinal epithelial Caco-2 (TC7) cell monolayers. Pharmaceutical Research 16: 625-632.
  62. Artursson, P., Ungell, A.L. & Lofroth, J.E. (1993). Selective paracellular permeability in two models of intestinal absorption: cultured monolayers of human intestinal epithelial cells and rat intesfinal segments. Pharmaceutical Research 10: 1123-1129.
  63. Artursson, P. (1990). Epithelial transport of drugs in cell culture. I. A model for studying the passive diffusion of drugs over intestinal absorptive (Caco-2) cells. Journal of Pharmaceutical Sciences 79: 476-482.
  64. Rubas, W., Jezyk, N. & Grass, G.M. (1993). Comparison of the permeability characteristics of a human colonic epithelial (Caco-2) cell line to colon of rabbit, monkey, and dog intestine and human drug absorption. Pharmaceutical Research 10: 113-118.
  65. Chadwick, V.S., Phillips, S.F. & Hofmann, A.F. (1977). Measurements of intestinal permeability using low molecular weight polyethylene glycols (PEG 400). II. Application to normal and abnormal permeability states in man and animals. Gastroenterology 73: 247-251.
  66. Conradi, R.A., Hilgers, A.R., Ho, N.F. & Burton, P.S. (1991). The influence of peptide structure on transport across Caco-2 cells [see comments]. Pharmaceutical Research 8: 1453-1460.
  67. Conradi, R.A., Hilgers, A.R., Ho, N.F. & Burton, P.S. (1992). The influence of peptide structure on transport across Caco-2 cells. II. Peptide bond modification which results in improved permeability. Pharmaceutical Research 9: 435-439.
  68. Gres, M.C., Julian, B., Bourrie, M., Meunier, V., Roques, C., Berger, M., Boulenc, X., Berger, Y. & Fabre, G. (1998). Correlation between oral drug absorption in humans, and apparent drug permeability in TC-7 cells, a human epithelial intestinal cell line: comparison with the parental Caco-2 cell line. Pharmaceutical Research 15: 726-733.
  69. Yamashita, S., Tanaka, Y., Endoh, Y., Taki, Y., Sakane, T., Nadai, T. & Sezaki, H. (1997). Analysis of drug permeation across Caco-2 monolayer: implication for predicting in vivo drug absorption. Pharmaceutical Research 14: 486-491.
  70. Lennernas, H., Nylander, S. & Ungell, A.L. (1997). Jejunal permeability: a comparison between the Ussing chamber technique and the single-pass perfusion in humans. Pharmaceutical Research 14: 667-671.
  71. Artursson, P. & Magnusson, C. (1990). Epithelial transport of drugs in cell culture. II. Effect of extracellular calcium concentration on the paracellular transport of drugs of different lipophilicities across monolayers of intestinal epithelial (Caco-2) cells. Journal of Pharmaceutical Sciences 79: 595-600.