transport of octreotide and evaluation of mechanism of opening the paracellular tight junctions...

Transport of Octreotide and Evaluation of Mechanismof Opening the Paracellular Tight Junctions usingSuperporous Hydrogel Polymers in Caco-2 Cell Monolayers

FARID A. DORKOOSH,1,2 CORINE A.N. BROEKHUIZEN,1 GERRIT BORCHARD,1 MORTEZA RAFIEE-TEHRANI,1

J. COOS VERHOEF,1 HANS E. JUNGINGER1

1Department of Pharmaceutical Technology, Leiden/Amsterdam Center for Drug Research, Leiden University,P.O. Box 9502, 2300 RA Leiden, The Netherlands

2Department of Pharmaceutics, Organon International BV, P.O. Box 20, 5340 BH Oss, The Netherlands

Received 12 June 2003; revised 25 August 2003; accepted 26 August 2003

ABSTRACT: The purpose of this study was to investigate the mechanism of opening oftight junctions in Caco-2 cell monolayers using superporous hydrogel (SPH) and SPHcomposite (SPHC) polymers as permeation enhancers for peptide drug delivery.Moreover, the transport of octreotide across Caco-2 cell monolayers was assessed byapplication of SPH and SPHCpolymers onCaco-2 cell monolayers. In these experiments,N,N,N-trimethyl chitosan chloride with 60% quaternization (TMC60) was used as apositive control for opening of tight junctions. Transepithelial electrical resistance(TEER) studies showed that all three polymers (TMC60, SPH, and SPHC) were able todecrease TEER values to �30% of the initial values, indicating the ability of thesepolymers to open the tight junctions. Recovery TEER studies showed that the effects ofthe polymers on Caco-2 cell monolayers were reversible, indicating viability of the cellsafter incubation with polymers. Both SPH and SPHC (compared with TMC60) were ableto increase the paracellular transport of octreotide by theirmechanical pressures on tightjunctions. The mechanistic studies showed that junctional proteins, including actin,occludin, and claudin-1,were influenced by application of SPHandSPHCpolymers to theCaco-2 cell monolayers. SPH and SPHC induced clear changes in the staining pattern ofall three proteins compared with the control, indicating that the expression of theseproteins in the tight junctions was increased, most likely due to the mechanical pressureof the polymers on the junctional proteins. � 2004 Wiley-Liss, Inc. and the American

Pharmacists Association J Pharm Sci 93:743–752, 2004

Keywords: hydrogels; Caco-2 cells; paracellular transport; tight junction; polymers;proteins; peptides

INTRODUCTION

It is well-known that hydrophilic macromolecularcompounds, such as endogenous peptides orpeptidomimetic drugs, will merely be transported

paracellularly via the tight junctions.1,2 Tightjunctions are one mode of cell-to-cell adhesion inepithelial and endothelial cell sheets. Thesejunctions constitute continuous and circumferen-tial seals around cells and serve as a primarybarrier preventing solutes and water from pas-sing freely through the paracellular pathway.Tight junctions are also thought to function as aboundary between the apical and basolateralplasma membrane domains to generate andmaintain cell polarity. Both the barrier and fence

JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 93, NO. 3, MARCH 2004 743

Correspondence to: Farid A. Dorkoosh (Telephone: 31 412661582; Fax: 31 412 662524;E-mail: farid.dorkoosh@organon.com)

Journal of Pharmaceutical Sciences, Vol. 93, 743–752 (2004)� 2004 Wiley-Liss, Inc. and the American Pharmacists Association

functions of tight junctions are essential formulticellular organisms. These functions nor-mally prevent the transport of hydrophilic macro-molecules across the intestinal cells.2,3 Tightjunctions consist of multiple proteins that aresignaling molecules, such as occludin, claudin(1–18), JAM (junctional-associated membraneprotein), MAGUKs (membrane-associated guany-late kinase; ZO-1, ZO-2, and ZO-3), non-MAGUKs(cingulin, 7H6 antigen, and symplekin). Of these,occludin and claudin are the most importantproteins for the functioning of tight junctions.4–9

Beneath the tight junctions are adherence junc-tions. Actin, an important protein present in theseadherence junctions, has an essential role in thefunctionality of tight junctions.10,11

Octreotide, a synthetic octapeptide analog ofsomatostatin, is a hydrophilicmacromoleculewitha molecular weight of 1019.3 Da. It is poorlyabsorbed by the intestinal tract and thereforeexhibits a low bioavailability after oral adminis-tration.12,13 This low bioavailability is mostlydue to inhibition of octroetide absorption via theparacellular pathway by tight junctions. There-fore, coadministration of octreotide with absorp-tion enhancers that are able to reversibly opentight junctions could improve its bioavailabil-ity.14,15 A variety of such absorption enhancers(such as carbomer and chitosan and its deriva-tives) thathavebeen shown tobe able to selectivelyopen the tight junctions have been used toinfluence the opening of tight junctions andimprove the absorption of such peptides. Themechanism of opening of the tight junctions usingdifferent absorption enhancers has also beenassessed, and various theories for this mechanismhave been proposed.15–17 Recently, novel deliverysystems based on superporous hydrogels (SPH)have been developed for delivery of peptides,including octreotide.14 These SPH-based deliverysystems are potential dosage forms for oraldelivery of peptide drugs. In this investigation,the capability of these SPH and SPH com-posite (SPHC) polymers for transport of octreotideare evaluated in Caco-2 cell monolayers, and themechanism of opening of tight junctions usingthese polymers is assessed.

EXPERIMENTAL

Materials

Octreotide was kindly donated by NovartisPharma AG (Basel, Switzerland). Triton-X was

purchased from Aldrich Chemical (Gillingham,Dorset, UK). Rhodamine phalloidine was ob-tained from Molecular Probes (Eugene, OR).Anti-occludin and rabbit anti-claudin-1 primaryantibodies were from Zymed Laboratories (SouthSan Francisco, CA). FITC-conjugated goat anti-mouse IgG and goat anti-rabbit IgG secondaryantibodies were purchased from Jackson Immu-noResearch Laboratories (West Grove, PA). Goatserum was from Sigma Chemicals(St. Louis, MO).Superporous hydrogel (SPH) and SPH composite(SPHC) were synthesized in our laboratory asdescribed previously.18 TMC60 (N,N,N-trimethylchitosan chloride) with 60% quaternization wasalso synthesized in our laboratories.19 All otherchemicals were of analytical grade and used asreceived.

Caco-2 Cell Cultures

Caco-2 cells were obtained from the AmericanType Culture Collection (Rockville, MD). The cellswere cultured in Dulbecco’s Modified EagleMedium (DMEM; Life Technologies, Gibco BRL,UK) containing 10% v/v heat-denaturated fetalcalf serum FC II (Fetal Clone II; Hyclone, Greiner,The Netherlands), 1% v/v nonessential aminoacids, 160 U/mL benzylpenicillin, and 100 U/mLstreptomycin (Sigma Chemical, St. Louis, MO).The cells were maintained at 378C in an atmo-sphere of 95% air and 5% CO2 at 90% relativehumidity. The cells were grown in 25-cm2 cultureflasks, the medium was changed every other day,and the cells were trypsinized once per week.Caco-2 cells of passage numbers 61–70 were usedin these experiments.

Transport of Octreotide across Caco-2Cell Monolayers

For transport studies, the Caco-2 cells were cul-tured on porous polycarbonate filter membraneswith a pore size of 0.4 mm and a surface area of4.7 cm2 in clusters of 6 wells (Costar Transwell1,Badhoevedorp, The Netherlands). When the cellswere trypsinized, they were seeded at a densityof 104 cells/cm2 onto each filter. These cells weremaintained at 378C in an atmosphere as alreadydescribed. Themediumwas replaced every secondday for 3 weeks. On the day of the transportexperiments, the culture medium was replacedwith an equal volume (2 mL, both apically andbasolaterally) of Hank’s balanced salt solution(HBSS; Sigma-Aldrich, Saint Louis, MO) buffered

744 DORKOOSH ET AL.

JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 93, NO. 3, MARCH 2004

with 30 mM n-(2-hydroxyethyl) piperazine-N-(2-ethanosulfonic acid) (HEPES) at pH 7.2 (trans-port medium), and the cells were allowed toequilibrate for 1 h. The transepithelial electricalresistance (TEER) was measured, following equi-libration, with a Milicell1 ERS meter (MilliporeCorp., Bedford, MA) connected to a pair of chop-stick electrodes to ensure the integrity of themonolayers formed on the filters. TEER measure-ments were also performed during the experimentto check the effect of polymers on opening of thetight junctions at time intervals of �60, �30, 0(i.e., 10 min after adding the polymers), 30, 90,180, and 240 min. For the examination of thetransport of octreotide across the Caco-2 cellmonolayers, dose–response experiments wereinitially performed using different concentrationsof octreotide (0.1, 0.5, 1.0, and 2.0 mg/mL). Afteradding 10 mg of SPH, 20 mg of SPHC, and 1 mL ofTMC 60 (1.0% w/v) polymers for 10 min, 1.0 mL ofthe different octreotide solutions were added tothe apical chamber. Because the swelling ratioof SPH is higher than that of SPHC, it wasnecessary to apply the minimum amount of SPHCpolymer (i.e., 20 mg of SPHC versus 10 mg of SPHpolymer) to cover the entire surface of the Caco-2cell monolayers. In all experiments, 1 mL ofTMC60 (1.0% w/v) was used as a positive control.No polymer was applied to the monolayers as anegative control. Samples of 200 mL were with-drawn from the basolateral chamber at predeter-mined time intervals of 0, 5, 15, 30, 60, 90, 120,180, and 240 min and replaced with equalvolumes of fresh HBSS–HEPES. After comple-tion of the transport studies, the polymers wereremoved carefully, monolayers were rinsed withHBSS–HEPES, and then culture medium wasapplied on the monolayers. The monolayers wereallowed to regenerate for 1 day at 378C in anatmosphere of 95% air and 5% CO2 at 90%relative humidity. TEER was monitored at 5, 6,and 24 h during the recovery period.

Apparent permeability (Papp) for each sub-stance was calculated according to the followingequation:

Papp ¼ dQ

dt

1

AC060ð1Þ

where Papp is the apparent permeability (cm/s),dQ/dt is the permeability rate (amount permeatedper minute during the whole period of 240 min),A is the diffusion area of the monolayers (cm2),and C0 is the initial concentration of the com-

pound studied. Enhancement ratio (ER) wasdetermined with the following equation:

ER ¼PappðpolymerÞPappðcontrolÞ

ð2Þ

Statistical differences were calculated usingStudent’s paired t-test at a significance level ofp< 0.05.

Analytical Procedures

The octreotide samples were analysed with a high-performance liquid chromatography–ultraviolet(HPLC–UV) system at a wavelength of 218 nm.The system consisted of a pump (Spectra-Physics,Fremont, CA), an autosampler (Gilson, Middle-ton, WI), and a UV detector (Spectra-Physics,Fremont, CA). The stationary phase was a 125�4.0 mm Prontosil-eurobond column (Leonberg,Germany), with a particle size of 5.0 mm. Thecolumn was connected to an Alitima C18 7.5�4.6 mm guard column (Alltech, Breda, TheNetherlands). The flow rate was 1 mL/min.Maximum pressure was adjusted to 150 bar, andthe injection volume was 50 mL. The mobile phaseconsisted of 67% ammonium acetate (pH 8.2) and33% acetonitrile (ACN).

Mechanism of Opening the Tight Junctions usingSPH, SPHC, and TMC 60

Caco-2 cells were seeded on tissue culture-treatedmedium polycarbonate filters as already de-scribed; however, in these experiments, 12-wellplates with a surface area of 1 cm2/insert wereused. Two hours prior to the experiment, theculture medium was changed to the experimentalmedium (HBSS–HEPES), and 0.5 and 1.5 mLwere added to the apical and basolateral sides,respectively. The cells were left in the incubator toequilibrate. Then, the polymers (4 mg of SPH,8 mg of SPHC, and 0.5 mL of 1% TMC60 as apositive control) were added to the apical side, andthe cells were left in the incubator for 2 h to checkthe influence of these polymers on opening of tightjunctions. Thereafter, the following tests wereperformed to study the mechanism of opening oftight junctions by staining actin, occludin, andclaudin-1. For all three tests, the polymers wereremoved and the cells washed once with phos-phate buffered saline (PBS; 0.5 mL apically and1.5 mL basolaterally). The cells were fixed withparaformaldehyde (200 mL apically and 500 mL

TIGHT JUNCTION OPENING AND OCTREOTIDE TRANSPORT 745

JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 93, NO. 3, MARCH 2004

basolaterally) in PBS for 30 min. Thereafter, thecells were washed three times with PBS. Thecells were permeabilized using Triton-X in H2O(200 mL apically and 500 mL basolaterally) for10 min. The following steps for each stainingprocedure are described next.

Actin Visualization

The cells were rinsed twice with PBS. The filterswere left to dry at the air. When the filters weredried, they were incubated apically, with 100 mLof freshly prepared rhodamine phalloidin, in thedark (using aluminium foil) for 20 min at ambienttemperature. The cells were rinsed three timeswith PBS and once with glycerol/PBS (1:1, v/v).The samples were kept under glycerol/PBS andtransferred carefully to the confocal laser-scan-ning microscope to visualize the actin. Aninverted Leica confocal laser-scanning microscope(TCS SP2; Leica, Mannheim, Germany), with anexcitation of 550 nm, emission of 580 nm, andmagnification of 63� oil immersion, was used forvisualization of the actin.

Occludin Visualization

The cells were washed three times with PBS andtreated (apically) for 30 min with 200 mL of ablocking solution comprised of 5% normal goatserum and 5% bovine serum albumin (BSA) inPBS. The blocking solution was removed, and thecells were incubated apically, with 200 mL of anti-occludin at a dilution of 1:50, at 48C overnight.Next, the cells were washed once with PBS(0.5 mL apically and 1.5 mL basolaterally) andtreated with a blocking solution as before. After30 min of incubation, the blocking solution wasremoved and the secondary antibody (200 mL ofFITC-conjugated goat anti-mouse IgG at a dilu-tion of 1:100 in PBS) was applied apically for 1 h.The cells were washed twice with 0.5% Tween-20in PBS (0.5 mL apically and 1.5 mL basolaterally).

Thereafter, the monolayers were visualized foroccludin by confocal laser-scanning microscopy asalready explained, but with an excitation of 520and emission of 455.

Claudin-1 Visualization

The cells were washed three times using PBS. Thecells were soaked in PBS containing 1% BSA for10 min (0.5 apically and 1.5 basolaterally). Then,the BSA–PBS solution was removed and the cellswere treated for 30 min apically with 200 mL ofthe primary antibody (rabbit anti-claudin-1 at adilution of 1:50 in PBS). The cells were washedthree times with PBS and again incubated for30 min with the secondary antibody (200 mL ofgoat anti-rabbit IgG at a dilution of 1:100 in PBS).Finally, the cells were washed three times withPBS and immersed in 0.5% Tween-20 (0.5 mLapically and 1.5 mL basolaterally). The mono-layers were visualized for claudin-1 by con-focal laser-scanning microscopy as they were foroccludin.

RESULTS

Transport of Octreotide across Caco-2Cell Monolayers

The effects of SPH, SPHC, and TMC60 polymerson octreotide transport at four different con-centrations of this peptide (0.1, 0.5, 1.0, and2.0 mg/mL) are shown in Table 1. The apparentpermeabilities (Papp; cm/s) of octreotide increasedwhen the peptide dose was increased from 0.1 to0.5 mg/mL. However, when the dose was furtherincreased to 1.0 and 2.0 mg/mL, the permeabilityappeared to decrease. Based on these results, thefollowing transport experiments were performedusing 0.5 mg/mL octreotide.

The TEER values of the Caco-2 cell monolayers,measured during the transport experiments with

Table 1. Apparent Permeability (Papp) for Octreotide Transport across Caco-2 Cell Monolayersa

Octreotide Concentration(mg/mL)

SPH Papp

(�107) (ER)SPHC Papp

(�107) (ER)TMC 60 Papp

(�107); (ER)Control Papp

(�107) (ER)

2 10.4� 0.2 (5.2) 7.6� 2.1 (3.8) 2.9� 2.0 (1.5) 2.0� 0.2 (1)1 16.3� 1.2 (14.8) 9.4� 1.4 (8.5) 5.0� 0.06 (4.5) 1.1� 0.002 (1)0.5 20.1� 2.0 (11.8) 15.2� 3.3 (8.9) 7.0� 1.1 (4.2) 1.7� 0.2 (1)0.1 18.8� 1.6 (14.5) 13.3� 0.5 (10.2) 5.8� 3.8 (4.5) 1.2� 1.1 (1)

aResults are expressed as mean�SD of three experiments.SPHandSPHCare significantly different fromTMC60andnegative control (p<0.05).Papp valuesare expressedas cm/s; transport

enhancement ratios (ER) are shown in brackets.

746 DORKOOSH ET AL.

JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 93, NO. 3, MARCH 2004

the 0.5-mg/L octreotide solution, are presented aspercentage of the initial values at t¼ 0 min inFigure 1. A significant decrease (�30% of theinitial value; p< 0.05) was observed in TEERvalues for each of the three polymers (SPH, SPHC,and TMC60) compared with the control (i.e.,without any polymer).

Recovery of TEER values after removal of thepolymers is shown inTable 2.A slight recoverywasobserved at 5 and 6 h for all three polymers. Thecontrol shows a slight decrease in TEER values at5 and 6 h, which was probably due to the extrastresses applied to the cells by replacing theexperimental medium with nutritional medium.After 24 h of incubation of the monolayers withnutritionalmedium, TEER recovered to 48.3, 61.8,and 64.4% of the initial values for SPH, SPHC, andTMC60, respectively.

The cumulative transport of octreotide (0.5 mg/mL) across Caco-2 cell monolayers is shown inFigure 2. In the presence of SPH and SPHCpolymers, this transport was 9.1� 1.5% and5.0� 2.4% of the apical concentrations, respec-tively. These values are comparable to that ofTMC60 as a positive control (4.1� 0.7%) andmuch higher than that of the negative control

(0.3� 0.4%). Thus, the application of SPH, SPHC,and TMC60 polymers resulted in transportenhancements of 30.3-, 16.6-, and 13.6-fold, re-spectively, compared with the negative control.

Mechanism of Opening the Tight Junctionswith SPH, SPHC, and TMC60

Actin Visualization

As shown in Figure 3A, the actin filaments of theCaco-2 cell monolayers in the control group arelocalized at the apical perijunctional area andappear as a continuous band encircling the cells atthe cellular borders (shown by broken arrows).The actin pattern for the SPH polymer (Fig. 3B)is different than that of the control. The fila-ments are no longer a continuous band; rather,the perijunctional actin filaments are progres-sively disrupted as evidenced by their breakage,displacement, and clumping. Inside the cells,stress fibers (arrow) are observed, indicating thatthe cells underwent a certain kind of stress, suchas opening of the tight junctions. The SPHCpolymer, shown in Figure 3C, demonstrates asimilar pattern as that found with SPH. TMC60(Fig. 3D) also shows accumulation of actin at thecell borders, but no apparent stress fibers areevident. This result is probably due to the dif-ferent mode of interaction of TMC60 (chemicalinteraction with actin) compared with that of SPHand SPHC polymers (mechanical interaction).

Occludin Visualization

The control cells after occludin staining are shownin Figure 4A. A clear homologous staining isobserved, indicating that occludin was presentin the tight junctions between the Caco-2 cells.

Figure 1. Effects of polymers on TEER of Caco-2 cellmonolayers. Key: (^) SPH; (&) SPHC; (~) TMC60; (�)negative control. Data are expressed as mean�SD of3 experiments.

Table 2. TEER Recovery of Caco-2 Cell Monolayersafter Removing The Polymersa

Time(h)

TEER (% of Initial Value)

SPH SPHC TMC60 Control

4 27.6� 17.7 27.6� 22.6 39.4� 6.0 111.4� 21.25 29.3� 5.6 29.2� 14.1 49.6� 7.1 77.4� 13.46 37.8� 19.1 45.9� 9.9 55.2� 12.7 80.2� 4.924 48.3� 4.7 61.8� 4.8 64.4� 0.7 141.2� 4.9

aData are expressed as mean�SD of three experiments.

Figure 2. Cumulative transport of octreotide acrossCaco-2 cell monolayers. Key: (^) SPH; (&) SPHC; (~)TMC60; (�) negative control. Data are expressed asmean�SD of 3 experiments.

TIGHT JUNCTION OPENING AND OCTREOTIDE TRANSPORT 747

JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 93, NO. 3, MARCH 2004

Cells pretreated with SPH polymer show accu-mulation of occludin between the contact pointsof the Caco-2 cells (Fig. 4B). The same resultswere observed for SPHC polymers (data are notshown). These results indicate greater expressionof occludin when the tight junctions were openedwith SPH and SPHC polymers. Cells pretreatedwith TMC60 (Fig. 4C) showed the same pattern asfound for SPH and SPHC. However, the occludinaccumulation between the contact points of theCaco-2 cells was less intensive, which might bedue to a less pronounced effect of TMC60 on tightjunctions compared with that of SPH and SPHCpolymers.

Claudin-1 Visualization

The control cells stained for claudin-1 show ahomologous pattern of claudin-1 in the planes ofthe tight junctions, which indicates that claudin-1was located in the tight junctions (Fig. 5A). Thebands observed for the cells pretreated with SPH(Fig. 5B) were more expressed (shown by arrow)than the control cells and also ruptured (shown bybroken arrow). These results suggest that open-ing of the tight junctions alters the position ofclaudin-1 and that a higher expression of claudin-1 occurs. The same pattern was observed forSPHC (data are not shown). In the case of TMC60

Figure 3. Staining of the tight junction protein actin: (A) cells without any polymertreatment, (B) cells incubated with SPH, (C) cells incubated with SPHC, and (D) cellsincubated with TMC60.

748 DORKOOSH ET AL.

JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 93, NO. 3, MARCH 2004

Figure 4. Staining of the tight junction proteinoccludin: (A) cells without any polymer treatment, (B)cells incubated with SPH, and (C) cells incubated withTMC60.

Figure 5. Staining of the tight junction proteinclaudin-1: (A) cells without any polymer treatment, (B)cells incubated with SPH, and (C) cells incubated withTMC60.

TIGHT JUNCTION OPENING AND OCTREOTIDE TRANSPORT 749

JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 93, NO. 3, MARCH 2004

polymer (Fig. 5C), the bands are less expressedthan those of SPH and SPHC polymers, but thequantitatively similar pattern indicates the open-ing of tight junctions.

DISCUSSION

The peptide drug octreotide is poorly absorbedby the intestinal mucosa and therefore exhibitslow bioavailability after oral administration. Co-administration of octreotide with absorption en-hancers such as chitosan and its derivatives couldimprove its bioavailability.13 TMC60 has beenshown to enhance the peroral absorption ofoctreotide in vivo in pigs.20 Therefore, in thepresent research TMC60 was selected as a posi-tive control to compare the effect of SPH andSPHC polymers on opening of tight junctions andtransport of octreotide across Caco-2 cell mono-layers. Both SPH and SPHCwere able to decreaseTEER as effectively as TMC60, suggesting thehigh potential of these polymers for opening oftight junctions. The observed recovery of TEERusing SPH, SPHC, and TMC60 is indicative of thesafety of these polymers as absorption enhancers.

Transport of octreotide across Caco-2 cells atdifferent concentrations was also evaluated usingTMC 60, SPH, and SPHC polymers. The Papp

values of octreotideweremuchhigher for polymer-treated cells than for the control group cells, andthis variation in Papp values might indicate apartial active transport of octreotide that can besaturated at higher peptide concentrations. It hasalready been established that SPH and SPHCpolymers are able to enhance the transport ofhydrophilic macromolecules across porcine intes-tine ex vivo by exertingmechanical pressure on theintestinal epithelium.21Moreover, the transport offluorescein isothiocyanate-labeled dextrans withmolecularweights of 4400Da (FD4) and 19,600Da(FD20) across Caco-2 cell monolayers using SPHand SPHC polymers have been shown to occur viathe paracellular pathway.22 Therefore, transportof octreotide, which is a hydrophilic molecule witha molecular weight of 1019 Da, is also expected tobe enhanced via the paracellular pathway by theapplied mechanical pressure of SPH and SPHC onintercellular tight junctions of the intestinalepithelium. It is known that SPH and SPHC bothapplymechanical pressure on the cells by their fastswelling properties and suck water from inter-stitial spaces between cells,which cause shrinkageof the cells and result in opening of the tight

junctions.21However, TMC60 triggers the openingof tight junctions by physicochemical interactionwith proteins associated with epithelial tightjunctions. It is known that TMC60 binds to theepithelial cell membrane by electrostatic interac-tion, resulting in F-actin depolymerization anddisbandment of the tight junction protein ZO-1and, finally, opening of tight junctions.11

The mechanism of opening of tight junctionsusing SPH and SPHC polymers was studied bystaining the junctional proteins actin, occludin,and claudin-1. The actin cytoskeleton plays a rolein facilitating the process that brings opposingmembranes together and stabilizes them whenjunction formation has been initiated.10 The pat-terns of the actin ring are continuous at the site ofplasma membrane (areas of cell–cell contact).11

Untreated cells exhibit a smooth distribution ofthe F-actin cytoskeleton around the cells. Thiscontinuous and smooth distribution of the F-actinring was also observed in the control cells in thepresent studies. Within these cells (Fig. 3A), F-actin is localized at junctions between the plasmamembranes. When the cells were pretreated withSPH and SPHC polymers (Fig. 3B and C, respec-tively), clear differences in F-actin staining wereobserved in comparison with the control cells.Stress fibers were visible inside the cells, indicat-ing that the cells had undergone some kind ofstress stimulated by the applied mechanicalpressure of the polymers on the cells. These stressfibers were also noted by Thanou et al.15 duringincubation with TMC60. However, TMC60 in thepresent studies did not show these stress fibers tosuch a high extent as SPH and SPHC because ofthe different mode of interaction with the cells.This difference would indicate that SPH andSPHC polymers have a greater effect on the tightjunctions than TMC60. TMC60 has already shownto result in accumulation of F-actin at cell–cellcontacts,15 and this was also observed for SPH andSPHC in the present studies. Nevertheless, moreinvestigations are necessary to evaluate the rever-sibility of stress fiberswhen the tight junctions areclosed after absorption of active compounds.

Occludin plays an important role in the elec-trical barrier function of the tight junction,possibly by the formation of aqueous pores withintight junction strands.9 Triggering the phosphor-ylation of occludin on its serine–threonine resi-dues has been demonstrated to influence thepermeability of the tight junctions.23,24 SPH,SPHC, and TMC60 resulted in accumulation ofoccludin and co-localization at cell–cell borders

750 DORKOOSH ET AL.

JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 93, NO. 3, MARCH 2004

(Fig. 4B and C), which was also observed afterover-expression of occludin.6 The increased ex-pression of occludin after treating the cells withSPH or SPHC could be due to the phosphorylationof the serine–threonine residues by uptake of theextracellular Ca2þ and stimulating the cascade ofphosphorylation. Occludin is mainly displayed asa continuous band around the cell periphery15

(as also shown in Fig. 4A), and if tight junctionsare opened, the cells show a disrupted patternof occludin (as observed for SPH, SPHC, andTMC60).

The word claudin for the integral membraneprotein claudin comes from the Latin word‘‘claudere’’ which means ‘close to’, because claudinis closely related to occludin. There are two types ofclaudin (claudin-1 and -2) that are directly incor-porated into tight junction strands. These proteinsare concentrated at cell contact sites and can beobserved as a homologous ring around the cells.8

This observation is also evident for claudin-1 inFigure 5.

The results of the studies presented heredemonstrate that occludin and claudin-1 are bothfound at cell–cell contact sites in a homologousmanner. The SPH and SPHC polymers showed areduction in the homogenous ring around the cells.The pattern appeared to be disrupted, and accu-mulation was observed at places where three cellscome together. Therefore, SPH and SPHC poly-mers clearly altered the claudin-1 pattern in thetight junctions by their mechanical pressure onthese junctions.9

In conclusion, SPH and SPHC, new polymers inperoral peptide drug delivery systems,25 are able toenhance the paracellular transport of hydrophilicand macromolecular peptide drugs such as octreo-tide across Caco-2 cell monolayers. Moreover,these polymers reversibly decrease the TEERvalues of themonolayers, indicating the capabilityof SPH and SPHC for opening of tight junctions.The mechanism of opening of tight junctions bySPH and SPHC polymers was studied by stainingactin, occludin, and claudin-1 (junctional pro-teins). The results demonstrate the expression ofthese proteins by the applied mechanical pressureof the polymers on the cells, thereby promoting theopening of the tight junctions.

ACKNOWLEDGMENTS

The authors thank Jan Slats from the Depart-ment of Molecular Cell Biology of LUMC (Leiden

University Medical Center) for his assistancewith the CLSM studies.

REFERENCES

1. Sakai M, Imai T, Ohtake H, Azuma H, Otagiri M.1998. Effects of absorption enhancers on cytoske-letal actin filaments in Caco-2 cell monolayers. LifeSci 63:45–54.

2. Artursson P. 1990. Epithelial transport of drugsin cell culture: A model for studying the passivediffusion of drugs over intestinal absorptive (Caco-2) cells. J Pharm Sci 79:476–482.

3. Tsukita S, Furuse M, Itoh M. 1999. Structural andsignalling molecules come together at tight junc-tions. Curr Opin Cell Biol 11:628–633.

4. Sambati SA, Zarif A, Basson MD. 1996. Octreotidedifferentially modulates Human Caco-2 intestinalepithelial cell proliferation and differentiation bydecreasing intracellular camp. Regul Peptides61:219–227.

5. Bamforth SD, Kniesel U,WolburgH, Engelhardt B,Risau W. 1999. A dominant mutant of occludindisrupts tight junction structure and function.J Cell Sci 112:1879–1888.

6. Furuse M, Fujimoto K, Sato N, Hirase T, Tsukita S,Tsukita S. 1996. Overexpression of occludin, a tightjunction-associated integral membrane protein, in-duces the formation of intracellular multilamellarbodies bearing tight junction-like structures. J CellSci 109:429–435.

7. Matter K, Balda MS. 1998. Biogenesis of tightjunctions: the C-terminal domain of occludin med-iates basolateral targeting. J Cell Sci 111:511–519.

8. Furuse M, Fujita K, Hiiragi T, Fujimoto K, TsukitaS. 1998. Claudin-1 and -2: Novel integral mem-brane proteins localizing at tight junctions with nosequence similarity to occludin. J Cell Biol 141:1539–1550.

9. Furuse M, Sasaki H, Fujimoto K, Tsukita S. 1998.A single gene product, claudin-1 or -2, reconstitutestight junction strands and recruits occludin infibroblasts. J Cell Biol 143:391–401.

10. Vasioukhin V, Fuchs E. 2001. Actin dynamics andcell-cell adhesion in epithelia. Curr Opin Cell Biol13:76–84.

11. Banan A, Fitzpatrick L, Zhang Y, Keshavarzian A.2001. OPC-compounds prevent oxidant-inducedcarbonylation and depolymerization of the F-actincytoskeleton and intestinal barrier hyperperme-ability. Free Radical Biol Med 30:287–298.

12. Michael S, Thole M, Dillmann R, Fahr A, Drewe J,Fricker G. 2000. Improvement of intestinal peptideabsorption by a synthetic bile acid derivative,cholylsarcosine. Eur J Pharm Sci 10:133–140.

13. Thanou M, Verhoef JC, Marbach P, JungingerHE. 2000. Intestinal absorption of octreotide:

TIGHT JUNCTION OPENING AND OCTREOTIDE TRANSPORT 751

JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 93, NO. 3, MARCH 2004

N-trimethyl chitosan chloride (TMC) amelioratesthe permeability and absorption properties of thesomatostatin analogue in vitro and in vivo. J PharmSci 89:951–957.

14. Dorkoosh FA, Verhoef JC, Verheijden JHM, Rafiee-Tehrani M, Borchard G, Junginger HE. 2002.Peroral absorption of octreotide using deliverysystems based on superporous hydrogel polymers.Pharm Res 19:1530–1534.

15. Thanou M, Verhoef JC, Romeijn SG, NagelkerkeJF, Merkus FWHM, Junginger HE. 1999. Effects ofN-trimethyl chitosan chloride, a novel absorptionenhancer, on Caco-2 intestinal epithelia and theciliary beat frequency of chicken embryo trachea.Int J Pharm 186:73–82.

16. Borchard G, Lueßen HL, Boer de AG, Verhoef JC,Lehr C, Junginger HE. 1996. The potential ofmucoadhesive polymers in enhancing intestinalpeptide drug absorption. III: Effects of chitosan-glutamate and carbomer on epithelial tight junc-tions in vitro. J Controlled Release 39:131–138.

17. Schipper NGM, Olsson S, Hoogstraate JA, Boer deAG, Varum KM, Artursson P. 1997. Chitosans asabsorption enhancers for poorly absorbable drugs2: Mechanism of absorption enhancement. PharmRes 14:923–929.

18. Dorkoosh FA, Brussee J, Verhoef JC, Borchard G,Rafiee-Tehrani M, Junginger HE. 2000. Prepara-tion and NMR characterization of superporoushydrogels (SPH) and SPH composites. Polymer41:8213–8220.

19. Thanou M, Sieval AB, Kotze AF, VerhoefJC, Junginger HE. 1998. Preparation and NMR

characterization of highly substituted N-trimethylchitosan chloride. Carbohydr Polym 36:157–165.

20. Thanou M, Verhoef JC, Verheijden JHM, Jungin-ger HE. 2001. Intestinal absorption of octreotideusing trimethyl chitosan chloride: studies in pigs.Pharm Res 18:823–828.

21. Dorkoosh FA, Borchard G, Rafiee-Tehrani M,Verhoef JC, Junginger HE. 2002. Evaluation ofsuperporous hydrogel (SPH) and SPH composite inporcine intestine ex-vivo: assessment of drug trans-port, morphology effect, and mechanical fixation tointestinal wall. Eur J Pharm Biopharm 53:161–166.

22. Dorkoosh FA, Setyaningsih D, Borchard G, Rafiee-Tehrani M, Verhoef JC, Junginger HE. 2002.Effects of superporous hydrogels on paracellulardrug permeability and cytotoxicity studies in Caco-2 cell monolayers. Int J Pharm 241:35–45.

23. Hirase T, Kawashima S, Wong EYM, Ueyama T,Rikitake Y, Tsukita S, Yokoyama M, Staddon JM.2001. Regulation of tight junction permeability andoccludin phosphorylation by RhoA-p160ROCK-dependent and -independent mechanisms. J BiolChem 276:10423–10431.

24. Sakakibara A, Furuse M, Saitou M, Ando-Akat-suka Y, Tsukita S. 1997. Possible involvement ofphosphorylation of occludin in tight junction for-mation. J Cell Biol 137:1393–1401.

25. Dorkoosh FA, Verhoef JC, Borchard G, Rafiee-Tehrani M, Junginger HE. 2001. Developmentand characterization of a novel peroral peptidedrug delivery system. J Controlled Release 71:307–318.

752 DORKOOSH ET AL.

JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 93, NO. 3, MARCH 2004