The Advantages of the Ussing Chamber in Drug Absorption Studies

10.1177/1087057105276034Gotoh et al.Improvement of Absorption Screening

The Advantages of the Ussing Chamberin Drug Absorption Studies

YASUMASA GOTOH, NOBORU KAMADA, and DENICHI MOMOSE

By adding high concentrations of test drugs to an Ussing chamber with rat jejunum, we established a system that yields veryhigh correlations between the rat absorption percentage and the membrane permeability, and that can accurately predict theabsorption percentage for rats. An advantage of this technique is that, unlike the results obtained using Caco-2, the slope of theabsorption/membrane-permeability curve is gentle, which facilitates a more exact prediction of the absorption percentage. Inaddition, the results obtained with this technique demonstrated that it could be used to evaluate the absorption percentage ofdrugs with an affinity for P-glycoprotein (P-gp), which cannot be assessed using Caco-2. This method also allows for cassettescreening, which would facilitate evaluation of the contribution of P-gp to absorption in the small intestine. Cassette screeningshowed that absorption of fexofenadine was unaffected by combination with the P-gp substrate ketoconazole. Consistent withthis finding, in vivo studies showed that ketoconazole did not affect the FaFg for fexofenadine, a pharmacokinetic parameterthat reflects absorption and bioavailability in the small intestine. This confirms the usefulness of the Ussing chamber for cas-sette screening and also suggests that intestinal P-gp has a minimal contribution to drug absorption. (Journal of BiomolecularScreening 2005:517-523)

Key words: Ussing chamber, absorption, membrane permeability, Caco-2, P-gp, predict

INTRODUCTION

THERE ARE POOR INTERSPECIES CORRELATIONS betweenbioavailability (BA) data obtained from experimental ani-

mals because BA includes both absorption percentage and clear-ance factor and because species differences in excretion and me-tabolism are also contributing factors. Accordingly, there are nointerspecies correlations in BA calculated from in vivo studies, andit is impossible to predict BA in humans based on results in animalmodels. Because it has now become possible to predict metabo-lism in humans, it should also be possible to predict BA in humans,provided that the absorption percentage can be accuratelycalculated.

One method of calculating the absorption percentage in experi-mental animals in the early stages of drug discovery and develop-ment is to calculate biliary and urinary excretion rates after oral ad-ministration and to consider the total excretion by these routes to

be the absorption percentage. Under such circumstances, the iden-tification of metabolites becomes necessary, except in cases inwhich the administered compound is excreted in the urine and bilewithout transformation. Combined with the fact that the determi-nation of metabolites in biological samples is very difficult, this be-comes extremely labor-intensive work to conduct during the initialscreening stages in which unlabeled compounds are used.

Compound absorption is often evaluated during the initialstages of drug discovery by measuring the membrane permeabilityin Caco-2 cells. Although this method can be very useful for deter-mining the relative permeability of various compounds, the pre-dicted value for the absorption percentage obtained from the per-meability coefficient using Caco-2 systems is not reliable becauseof the steepness of the curves.1 Another disadvantage of methodsusing Caco-2 is that the absorption values predicted from the per-meability coefficient are underestimated with compounds havingaffinity for P-gp.2-5 The first objective of this study was to investi-gate the prediction of absorption percentage in rats using an Ussingchamber, which is applicable to the evaluation of drug transportthrough mucosal membranes, with rat small intestine to facilitatemore accurate prediction of absorption percentage in human in vi-tro studies. An additional objective of this study was to obtainuseful information regarding the contribution of intestinal P-gpusing this technique.

© 2005 The Society for Biomolecular Screening www.sbsonline.org 517

Division of Basic Discovery Research, Kissei Pharmaceutical Co., Ltd., Nagano,Japan

Received Nov 26, 2004, and in revised form Feb 22, 2005. Accepted for publica-tion Feb 22, 2005.

Journal of Biomolecular Screening 10(5); 2005DOI: 10.1177/1087057105276034

at TEXAS SOUTHERN UNIVERSITY on December 17, 2014jbx.sagepub.comDownloaded from

http://jbx.sagepub.com/

MATERIALS AND METHODS

Materials

Norfloxacin, terbutaline, bromocriptine, enarapril, cimetidine,and digoxin were purchased from Sigma (St. Louis, MO).Atenolol, verapamil, and ketoconazole were purchased fromWako Pure Chemical Industries, Ltd. (Tokyo, Japan). Quinidinewas purchased from Nacalai Tesque, Inc. (Kyoto, Japan). Fexo-fenadine was purchased from Kemprotec, Ltd. (Middlesbrough,UK), [n-methyl-3H] cimetidine (specific activity [SA]; 8.14 E 11Bq/mmol) and digoxin [3H(G)] (SA; 1.4 E 12 Bq/mmol) were pur-chased from PerkinElmer Life Science, Inc. (Boston, MA), andquinidine [9-3H] (SA; 7.4 E 11 Bq/mmol) and verapamil [3H (G)](SA; 2.96 E 11 Bq/mmol) were purchased from MuromachiYakuhin (Tokyo, Japan). Krebs-Ringer Bicarbonate buffer powderwas purchased from Sigma (St. Louis, MO). Reagents and allcompounds were analytical grade. Male, 7-week-old Sprague-Dawley (SD-IGS) rats were purchased from Charles River Japan,Inc. (Tokyo, Japan). All animals had free access to food and water.The studies reported in this article have been carried out in accor-dance with the Laboratory Animal Committee of Kissei Pharma-ceutical (Nagano, Japan).

Ussing chamber technique

Rats were exsanguinated under ether anesthesia via the abdom-inal aorta. The jejunum was immediately removed and placed inchilled Krebs-Ringer bicarbonate buffer (pH 6.0; Sigma). Eachsection of the small intestine was then cut into 2- to 3-cm pieces,which were rinsed and mounted in a diffusion cell. Peyer patcheswere visually identified and removed to exclude them from the ex-periment. The exposed intestinal surface area was 0.635 cm2. Fourmilliliters of Krebs-Ringer bicarbonate buffer (pH 6.0; Sigma) wasadded to the cell on both sides of the membrane. The tissue wasmaintained at 37° C with continuous bubbling with biogas (O2/CO2, 95:5). The test compound was then dissolved or suspended inthis buffer, which was added to the mucosal side (final concentra-tion of 1.5 and 0.75 mM). Samples (0.1 mL) were taken from thereceiving compartment at specified times (1 and 2 h), and were an-alyzed by high performance liquid chromatography (HPLC) orliquid chromatography–tandem mass spectrometry (LC/MS/MS).When labeled compounds were used, 1-mL samples were takenfrom the serosal side at specified times, and an additional 1 mL ofKrebs-Ringer bicarbonate buffer (pH 6.0; Sigma) was added. Fi-nally, 10 mL of Hionic Fluor (PerkinElmer) was added, and themixture was thoroughly shaken and left to stand before theradioactivity was measured in a liquid scintillation counter(TriCarb 1900CA, PerkinElmer).

In vivo rat study

Rats were intravenously administered test compound at 1 mg/kg after the urinary bladder was catheterized with polyethylene

tubing. Other rats were fasted for 16 h before dosing and were thenadministered 3 mg/kg by oral gavage. Blood samples were col-lected at 1, 5, 15, 30, 60, and 120 min after intravenous administra-tion and at 5, 15, 30, 60, and 120 min after oral administration.Samples (approximately 200 µL) were taken from the jugular veinusing a heparin-treated syringe, and plasma was obtained bycentrifugation. Urine samples were taken every 30 min for 2 h.

High performance liquid chromatography

An HPLC system consisting of a Hitachi 6000 series liquidchromatograph was used in the determination of norfloxacin,terbutaline, bromocriptine, enarapril, and atenolol. Isocratic elu-tion was performed with a Symmetry Shield RP8 (3.9 × 150 mm;Waters, Milford, MA) with a mobile phase of acetonitrile: 0.01%H3PO4 or 20 mM NaHPO4 (pH 7.0) at a flow rate of 1.0 mL/min at50° C. Compounds were detected at 210 nm.

Liquid chromatography–tandem mass spectrometry

Mass spectra were obtained using a Sciex API 365 mass spec-trometer (PerkinElmer) equipped with a turbo ion-spray source.Mass Chrom (Version 1.1, PerkinElmer) was the software used toanalyze the data acquired. Analysis was performed using aPhenomenex Synergi MAX 80A (4.6 × 50 mm; Phenomenex,Torrance, CA) system with a mobile phase of 50% acetonitrile/0.1% AcOH at 50° C at a flow rate of 0.2 mL/min, and it was con-ducted in positive mode.

Calculation of the membrane permeabilitycoefficient and in vivo kinetic parameters

The final concentrations for the compounds added to themucosal side were established at 1.5 and 0.75 mM. Serosal con-centrations were determined at 1 and 2 h, and the membrane per-meation rate at each concentration was calculated from the slope ofthe concentration-time curve. The membrane permeation rates forthe 2 concentrations were then plotted against the added concen-trations of the compound, and in vitro rat jejunum permeability(the membrane permeability coefficient) was calculated as theslope of this line.

In vivo analyses were calculated using the following equations:

F = (AUCpo/dose)/(AUCiv/dose) (1)

and

F = FaFhFg, (2)

where F is the bioavailability, Fa is the absorption percentage, Fh isthe hepatic bioavailability, and Fg is the bioavailability in the smallintestine;

CLr = (the amount of urine excretion)/AUCiv/RB, (3)

CLtot = dose/AUCiv/RB, (4)

Gotoh et al.

518 www.sbsonline.org Journal of Biomolecular Screening 10(5); 2005



and

CLh = CLtot – CLr, (5)

where CLr is the renal clearance, CLh is the hepatic clearance, CLtot

is the total clearance, and RB is the blood/plasma concentrationratio; and

Fh = (1 – CLh)/Qh, (6)

where Qh is the portal blood flow rate in rats (a value of 55 mL/min/kg was used). FaFg was then calculated from equations 2 and 6 asfollows:

FaFg = F – Fh. (7)

RESULTS

Relationship between the membrane permeability and theabsorption percentage using Ussing chamber technique

A total of 10 drugs were selected for comparison between theabsorption percentage and the membrane permeability. The ab-sorption percentage of these drugs ranged from 15% to 100%, anda variety of physicochemical properties were selected. The finalconcentrations of the drug solution or suspension were 0.75 and1.5 mM in pH 6.0 buffer in the present study, while the membranepermeability was evaluated using µM drug solutions, which corre-sponds to the Km value in many in vitro studies. The in vitro rat je-junum permeability was calculated using the membrane perme-ability rates for the 2 concentrations and the added drugconcentration. The plotted results revealed a very good correlationbetween the absorption percentage and the membrane permeabil-

ity (r = 0.976) (Fig. 1, Table 1). Moreover, the slope of the linecalculated from the membrane permeability and the absorptionpercentage using the Ussing chamber technique was less steepthan that produced using Caco-21 (17.2 in Ussing chamber and51.3 in Caco-2), which suggests that the absorption percentage canbe more accurately predicted (Fig. 1).

There is a very good correlation between the membrane perme-ability and the absorption percentage of drugs evaluated usingCaco-2 cells, provided that compounds with an affinity for P-gpare excluded. A good relationship between in vitro rat jejunumpermeability and the drug absorption percentage, with drugs hav-ing an affinity for P-gp such as digoxin, quinidine, verapamil, andcimetidine, was observed in the Ussing chamber but not in Caco-2studies6 (Fig. 2). Also, the higher concentration of these drugs didnot dissolve in the Ussing chamber system. However, regardless ofwhether the compounds had an affinity for P-gp or whether theywere in solution or suspension, this system allowed the predictionof the rat absorption percentage.

Contribution of intestinal P-gp in drug absorption

To examine the contribution of the intestinal P-gp to drug ab-sorption in our system, we examined the rat jejunum permeability(the membrane permeability coefficient) in vitro for test drugsalone or in combination with drugs that have an affinity for P-gp.The results for the test compounds (verapamil, quinidine, digoxin,and fexofenadine) revealed no significant differences in the ab-sorption percentage (calculated from in vitro rat jejunum perme-ability), regardless of the concomitant drugs (Table 2). Even when2 or 3 compounds were added simultaneously, the predicted ab-sorption percentage for the test compound was unaffected. There-fore, it appeared that the intestinal P-gp makes a minimalcontribution to drug absorption in this system (Table 2).

Improvement of Absorption Screening

Journal of Biomolecular Screening 10(5); 2005 www.sbsonline.org 519

Table 1. The Membrane Permeability and the Absorption Ratein Rats for the Investigated Drugs

Drug Rat Absorption (%) Permeability (cm/s)

Norfloxacin14 15 4.3E-07Terbutaline9 60 3.8E-06Atenolol15 50 3.7E-06Bromocriptine16 30 3.4E-07Enarapril17 34 7.9E-07Cimetidine9 100 3.2E-05Digoxin18 70 5.7E-06Quinidine19 80 1.5E-05Verapamil20 90 1.6E-05Fexofenadine21 30 7.2E-07

A piece of rat jejunum was mounted in a diffusion cell, Krebs-Ringer bicarbonate buffer(pH 6.0) was added to the cell on both sides of the membrane, and biogas (O2/CO2, 95:5)was passed through at 37° C. The test compound was added to the mucosal side (final con-centrations of 1.5 and 0.75 mM), and the sample was taken from the receiving compartmentat specified times. The membrane permeation rate at each concentration of the mucosalside was calculated from the slope of the concentration of the serosal side and time curve,and then the membrane permeability coefficient was calculated from the correlation of themembrane permeation rate and the concentration of mucosal side. Each value representsthe mean of 2 experiments. Superscripted numerals in table correspond to referencelistings.

y = 17.182Ln(x) + 274.62

R2 = 0.9527

0

20

40

60

80

100

1.0E-07 1.0E-06 1.0E-05 1.0E-04

Permeability (cm/s)

Rat

abs

orpt

ion

(%)

FIG. 1. Relationship between the percentage of oral dose absorptionand permeability across rat jejunum using the Ussing chamber. Each pointrepresents the mean of 2 experiments.



Pharmacokinetic parameters of fexofenadinewith or without ketoconazole in rat in vivo study

Figure 3 shows the plasma concentrations of fexofenadine afterintravenous administration of 1 mg/kg and after oral administra-tion of 3 mg/kg fexofenadine in rats. Also shown are the plasmaconcentrations of fexofenadine when ketoconazole was concomi-tantly administered at the same dose and administration route. Thevarious parameters were calculated from this study (Table 3). TheBA of fexofenadine in rats was 1.4%, with a total clearance of 20mL/min/kg and a renal clearance of 0.3 mL/min/kg. The Fh andFaFg calculated from this data were 0.641 and 0.022, respectively.When ketoconazole was concomitantly administered withfexofenadine, the Fh and FaFg were 0.713 and 0.017, respectively,which was very similar to when fexofenadine was administeredalone. Similarly, Figure 4 shows the plasma concentrations ofketoconazole when administered alone and with concomitantfexofenadine, and the parameters are shown in Table 3. Theplasma concentrations and pharmacokinetic parameters ofketoconazole were almost the same, regardless of the presence offexofenadine.

DISCUSSION

Many recent reports have demonstrated that active transportplays a critical role in membrane permeability. Drug absorptioncan be estimated from membrane permeability in cultured celllines such as Caco-2. Determination of membrane permeability inCaco-2 cells is simple and provides substantial absorption data, butit often does not accurately predict drug absorption in vivo. In theCaco-2 system, the compound is typically added in a dissolved

state at concentrations much lower than the actual concentrationspresent in the gastrointestinal tract, and the actual absorption canbe misjudged when the relationship between solubility and dose isnonlinear. Moreover, the Caco-2 system may underestimate the

Gotoh et al.


Caco-2 Ussing Chamber

R = 0.9098

0

20

40

60

80

100

1.0E-07 1.0E-06 1.0E-05 1.0E-04

Permeability (cm/s)

Ra

t ab

sorp

tion

(%

)

verapamil cimetidine

digoxin quinidine

R2 = 0.0033

0

20

40

60

80

100

120

1.0E-07 1.0E-06 1.0E-05 1.0E-04

Permeability (cm/s)

Rat

abs

orpt

ion

(%) verapamil

cimetidine

digoxinquinidine

2

FIG. 2. Relationship between the percentage of oral dose absorption and permeability across Caco-2 monolayers (left) and rat jejunum using theUssing chamber (right) with reference drugs for P-gp. Caco-2 data were quoted from Adachi et al.6 Briefly, the concentrations of compounds were as fol-lows: cimetidine (64.5 nM), digoxin (52.6 nM), quinidine (50 nM), and verapamil (11.8 nM). Transcellular transport experiments were carried out inHank’s balanced salt solution. The transported amount was determined at 1 and 2 h. Each point represents the mean of 2 to 3 experiments.

Table 2. Permeability Coefficients Using the Ussing Chamberand Predicted Percentage Absorbed with

or without Additive Drugs in theReference Drugs for an Affinity for P-glycoprotein

Rat Permeability PredictedReference Additive Absorption (%) (cm/s) Absorption (%)

Verapamil 9020 1.6E-05 84.9Verapamil Digoxin 90 2.0E-05 88.3Quinidine 8019 1.5E-05 83.8Quinidine Cimetidine 80 1.3E-05 81.5Cimetidine 1009 3.2E-05 97.0Digoxin 7018 5.7E-06 67.2Digoxin Quinidine 70 9.6E-06 76.1Digoxin Cimetidine 70 4.9E-06 64.6Digoxin Fexofenadine 70 7.0E-06 70.6Digoxin Ketoconazole 70 6.2E-06 68.5Fexofenadine 3021 7.2E-07 31.5Fexofenadine Quinidine 30 4.3E-07 22.9Fexofenadine Cimetidine 30 9.5E-07 36.4Fexofenadine Verapamil 30 6.6E-07 30.0Fexofenadine Ketoconazole 30 8.7E-07 34.9Fexofenadine Cimetidine, digoxin 30 5.5E-07 27.0

The test and concomitant drugs were added simultaneously at the same concentration in themucosal side of the Ussing chamber, and the membrane permeability coefficients were cal-culated. Each value represents the mean of 2 experiments. Superscripted numerals in tablecorrespond to reference listings.



absorption of compounds with an affinity for P-gp, which is ex-pressed both in the gastrointestinal tract and in Caco-2 cells. Con-sequently, it has not been possible to accurately predict absorption

percentage for all compounds.2-6 Thus, there has not been a univer-sal method that can be used to accurately predict the absorptionpercentage during the early stages of drug discovery.



10

100

1000

10000

0 0.5 1 1.5 2

Time (hr)

Fexo

fena

dine

con

cent

ratio

n(n

g/m

L)

alonewith Ketoconazole

1

10

100

1000

0 0.5 1 1.5 2Time (hr)

Fex

ofen

adin

e co

ncen

trat

ion

(ng/

mL)

alone

with Ketoconazole

FIG. 3. Plasma concentration of fexofenadine after oral (right) and intravenous (left) administration of fexofenadine with or without ketoconazole tothe rats. Each point and vertical bar represent the mean ± standard deviation of 3 experiments.

Table 3. Pharmacokinetic Parameters of Fexofenadine and Ketoconazole in Rats

F Fh FaFg CLtot (mL/min/kg) CLr (mL/min/kg) CLh (mL/min/kg)

Fexofenadine 0.014 0.641 0.022 20.0 0.3 19.8Fexofenadine with ketoconazole 0.012 0.713 0.017 16.0 0.2 15.8Ketoconazole 0.151 0.656 0.230 18.9 0.0 18.9Ketoconazole with fexofenadine 0.156 0.774 0.201 12.4 0.0 12.4

Rats received an intravenous dose of fexofenadine (1 mg/kg) and ketoconazole (1 mg/kg) and an oral dose of fexofenadine (3 mg/kg) and ketoconazole (3 mg/kg). The plasma and urine concen-trations were determined using liquid chromatography–tandem mass spectrometry. F = bioavailability; Fh = hepatic bioavailability; Fa = absorption rate; Fg = bioavailability in the small intestine;CLtot = total clearance; CLr = renal clearance; CLh = hepatic clearance. Each value represents the mean of 3 experiments.

10

100

1000

10000

0 0.5 1 1.5 2

Time (hr)

Ket

ocon

azol

e co

ncen

trat

ion

(ng/

mL)

alonewith Fexofenadine

1

10

100

1000

10000

0 0.5 1 1.5 2

Time (hr)

Ket

ocon

azol

e co

ncen

trat

ion

(ng/

mL)

alonewith Ketoconazole

FIG. 4. Plasma concentration of ketoconazole after oral (right) and intravenous (left) administration of ketoconazole with or without fexofenadine tothe rats. Each point and vertical bar represent the mean ± standard deviation of 3 experiments.



Here, we propose a relatively simple in vitro method for esti-mating the absorption percentage. An Ussing chamber with a pH6.0 buffer is used in this technique, the concentration of the addedcompound is around 1 mM (solubilization is not required), and thesystem is monitored for 2 h. This method has several advantages: avery good correlation between the membrane permeability and theabsorption percentage even for compounds that have an affinity forP-gp; a more gradual slope for the absorption percentage versus invitro rat jejunum permeability plot than in studies using Caco-2, al-lowing a more accurate prediction of the absorption percentage;and, in contrast to the Caco-2 assay, the ability to measureabsorption of compounds that are incompletely dissolved.

The suitability of the concentration of the added compound inthis technique corresponds to the state (maximum concentration)of the drug in the gastrointestinal tract after oral administration inrats (3 mg per 5 mL/kg for a compound having a molecular weightof 500). Provided that the compound is absorbed by simple diffu-sion and that a linear relationship between the concentration andthe membrane permeability rate can be shown, the membrane per-meability coefficient can be easily estimated. On the other hand,the Km value for the gastrointestinal transporter will be in therange of several µM if the affinity for the transporter(s) is high. Forconcentrations at which the Km value is exceeded (mM), it is be-lieved that active transport will reach the point of saturation, andthe contribution of passive transport will rise as the concentrationis increased. When we evaluated the relationship between mem-brane permeability and added compound concentration, a good re-lationship was observed using compounds which are absorbed bypassive diffusion, but some compounds exhibited a saturationcurve (data not shown). In addition, Chiou et al2 described thedominantly passive diffusion process in the case of P-gp saturationin the intestine as oral dose increases.

The solubility of the drug is an important factor in the analysisof its permeability and is therefore an important factor in its ab-sorption. Although higher solubility is better for the membranepermeability, we believe that solubility is not a major influence onthe permeability because the solubility is the maximum solubleconcentration, not the rate of the dissolution. When the added con-centration of the drug in the Ussing chamber cell corresponded tothe saturation of P-gp transport, some drugs exhibited precipitationin the buffer, indicating that the actual dissolved concentration be-came lower than the added concentration. This technique may thusallow evaluation of not only the contribution of P-gp, but also thesolubility and membrane permeability under the notion that onlythe dissolved drug is absorbed. Accordingly, regardless of thephysicochemical properties or the affinity for the intestinal trans-porter(s), this technique can predict the absorption percentage inthe linear dose-dependent state. Furthermore, the solubility did notaffect prediction of the absorption percentage in this system, but itmight be useful to compare the absorption between suspension andsolutions generated by the addition of a solubilizing reagent.

Consequently, if in vitro rat jejunum permeability can be accu-rately calculated using this technique, it also becomes possible to

estimate the absorption percentage in rats. Because it is known thatabsorption percentage in rats and humans is virtually the same,7 theaccurate calculation of absorption percentage in rats is consideredto be extremely useful.8-10

Clinical interactions are a concern for compounds that havean affinity for P-gp and other transporters such as the multidrugresistance-associated protein. Although P-gp is expressed in manytissues, including the brain, small intestine, and liver, the contribu-tion of this transporter in the small intestine and liver (which play amajor role in pharmacokinetics) has yet to be established. There-fore, even when drug-drug interactions are observed between suchcompounds, it is difficult to determine or even estimate whetherthe interaction is occurring in absorption, uptake, and/or excretion.The guiding hypothesis is that drugs with an affinity for P-gp haveintrinsically good absorption properties11,12 and that even if intesti-nal P-gp is inhibited, any change in this contribution would not besignificant. However, no theories have accurately described thecontribution of P-gp in the small intestine.

We investigated the membrane permeability using drugs withan affinity for P-gp and observed the same in vitro rat jejunum per-meability, regardless of whether the drug was added alone or con-comitantly with another drug to the same cell of the Ussing cham-ber. If the same examination was performed using Caco-2, wewould only observe an inhibition effect, resulting in changes in themembrane permeability rates in both the direction of absorptionand efflux. However, the Caco-2 system does not allow evaluationof the contribution of the P-gp. In our Ussing chamber system, wewere able to show that drugs with an affinity for P-gp have thesame membrane permeability regardless of whether they wereadded alone or in combination with other drugs, suggesting thatthe P-gp has a minimal contribution to intestinal absorption. Vari-ous transporters have been evaluated in previous in vitro studies,but there is no method that evaluates the final clinical effects of ab-sorption percentage and drug interactions. The present techniqueis believed to obtain useful information regarding the contributionof intestinal transporter(s).

The absorption percentage predicted by this method was almostthe same, even when multiple types of compound were simulta-neously added to the same well in the Ussing chamber or when acompound was added alone. This suggests that simultaneous eval-uation using cassette dosing is possible with this method during theinitial screening stages of drug development.

Lastly, we investigated the contribution of the intestinal P-gp invivo. We found that the inhibition of P-gp (which functions as anefflux transporter in the small intestine) with ketoconazole had noeffect on the FaFg of fexofenadine. This agreed with the finding thatthe absorption percentage for fexofenadine was unaffected by theaddition of ketoconazole in the Ussing chamber. Similarly, it is re-ported that ketoconazole had no effect on the absorption offexofenadine in clinical trials,13 thus suggesting that the contribu-tion of intestinal P-gp is minimal. Thus, both clinical and in vivofindings indicate that our Ussing chamber technique canaccurately predict drug-drug interactions.

Gotoh et al.




In summary, this method allows the in vitro estimation of theabsolute absorption percentage of compounds, which includes ac-tive transport associated with absorption and excretion via trans-porters expressed in the small intestine, and passive diffusion in-cluding transport through intercellular spaces and paracellulartransport, regardless of whether the tested compounds are in solu-tion or suspension. This method is also believed to be very usefulas a means for considering the effects of various drug-drug interac-tions on absorption.

ACKNOWLEDGMENTS

We are grateful to Prof. Yuichi Sugiyama and Asst. Prof.Hiroyuki Kusuhara of the Graduate School of Pharmaceutical Sci-ences at the University of Tokyo for giving us valuable advice.

REFERENCES

1. Artursson P, Karlsson J: Correlation between oral drug absorption in humansand apparent drug permeability coefficients in human intestinal epithelial(Caco-2) cells. Biochem Biophys Res Commun 1991;175:880-885.

2. Chiou WL, Chung SM, Wu TC, Ma CA: Comprehensive account on the roleof efflux transporters in the gastrointestinal absorption of 13 commonly usedsubstrate drugs in humans. Int J Clin Pharmacol Ther 2001;39:93-101.

3. Troutman MD, Thakker DR: Novel experimental parameters to quantify themodulation of absorptive and secretory transport of compounds by P-glycoprotein in cell culture models of intestinal epithelium. Pharm Res2003;20:1210-1224.

4. Hunter J, Hirst BH, Simmons NL: Drug absorption limited by P-glycoprotein-mediated secretory drug transport in human intestinal epithelialCaco-2 cell layers. Pharm Res 1993;10:743-749.

5. Bohets H, Annaert P, Mannens G, Beijsterveldt LV, Anciaux K, Verboven P,et al: Strategies for absorption screening in drug discovery and development.Curr Top Med Chem 2001;1:367-383.

6. Adachi Y, Suzuki H, Sugiyama Y: Comparative studies on in vitro methodsfor evaluating in vivo function of MDR1 P-glycoprotein. Pharm Res2001;18:1660-1668.

7. Swaan PW, Stehouwer MC, Tukker JJ: Molecular mechanism for the relativebinding affinity to the intestinal peptide carrier. Comparison of three ACE-inhibitors: enalapril, enalaprilat, and lisinopril. Biochim Biophys Acta1995;1236:31-38.

8. Lennernas H: Related Articles, Links Human jejunal effective permeabilityand its correlation with preclinical drug absorption models. J Pharm

Pharmacol 1997;49:627-638.

9. Chiou WL, Barve A: Linear correlation of the fraction of oral dose absorbedof 64 drugs between humans and rats. Pharm Res 1998;15:1792-1795.

10. Fagerholm U, Johansson M, Lennernas H: Comparison between permeabilitycoefficients in rat and human jejunum. Pharm Res 1996;13:1336-1342.

11. Sakaeda T, Okamura N, Nagata S, Yagami T, Horinouchi M, Okumura K,et al: Molecular and pharmacokinetic properties of 222 commercially avail-able oral drugs in humans. Biol Pharm Bull 2001;24:935-940.

12. Lin JH, Yamazaki M: Clinical relevance of P-glycoprotein in drug therapy.Drug Metab Rev 2003;35:417-454.

13. Tannergren C, Knutson T, Knutson L, Lennernas H: The effect ofketoconazole on the in vivo intestinal permeability of fexofenadine using a re-gional perfusion technique. Br J Clin Pharmacol 2003;55:182-190.

14. Nagatsu Y, Endo K, Irikura T: Studies on the fate of 14C-Labeled AM-715(norfloxiacine). Chemotherapy 1981;29:105-118.

15. Reeves PR, Barnfield DJ, Longshaw S, McIntosh DA, Winrow MJ: Disposi-tion and metabolism of atenolol in animals. Xenobiotica 1978;8:305-311.

16. Maurer G, Schreier E, Delaborde S, Nufer R, Shukla AP: Fate and dispositionof bromocriptine in animals and man, II: absorption, elimination and metabo-lism. Eur J Drug Metab Pharmacokinet 1983;8:51-62.

17. Tocco DJ, deLuna FA, Duncan AE, Vassil TC, Ulm EH: The physiologicaldisposition and metabolism of enalapril maleate in laboratory animals. Drug

Metab Dispos 1982;10:15-19.

18. Abshagen U, Bergmann KV, Rietbrock N: Evaluation of the enterohepaticcirculation after 3H-digoxin administration in the rat. Naunyn SchmiedebergsArch Pharmacol 1972;275:1-10.

19. Greenblatt DJ, Pfeifer HJ, Ochs HR, Franke K, MacLaughlin DS, Smith TW,et al: Pharmacokinetics of quinidine in humans after intravenous, intramuscu-lar and oral administration. J Pharmacol Exp Ther 1977;202:365-378.

20. The Japanese Pharmacopoeia. 14th ed. Drug information.

21. Cvetkovic M, Leake B, Fromm MF, Wilkinson GR, Kim RB: OATP and P-glycoprotein transporters mediate the cellular uptake and excretion offexofenadine. Drug Metab Dispos 1999;27:866-871.

Address reprint requests to:Yasumasa Gotoh

4365-1 KashiwabaraHotaka, Minamiazumi

Nagano 399-8304, Japan

E-mail: [email protected]





Documents

The Advantages of the Ussing Chamber in Drug Absorption Studies