Novel Approaches for Oral Delivery of Macromolecules

ALESSIO FASANO

Contribution from Division of Pediatric Gastroenterology and Nutrition and Gastrointestinal Pathophysiology Section, Center forVaccine Development, University of Maryland, School of Medicine, Baltimore, Maryland 21201.

Received March 10, 1998. Accepted for publication April 10, 1998.

Abstract 0 Traditional forms of administrations of nonabsorbabledrugs and peptides often rely on their parenteral injection, since theintestinal epithelium is poorly permeable to these therapeutical agents.A number of innovative drug delivery approaches have been recentlydeveloped, including the drug entrapment within small vesicles or theirpassage through the intestinal paracellular pathway. Zonula occludenstoxin, a recently discovered protein elaborated by Vibrio cholerae,provided tools to gain more insights on the pathophysiology of theregulation of intestinal permeability through the paracellular pathwayand to develop alternative approaches for the oral delivery of drugsand macromolecules normally not absorbed through the intestine.

The oral route presents a series of attractive advantagesfor the administration of therapeutical compounds, includ-ing the avoidance of pain and discomfort associated withinjections and the elimination of possible infections causedby the use of needles. Moreover, oral formulations areless expensive to produce, because they do not need to bemanufactured under sterile conditions.

In the past few years we have witnessed an explosion inresearch aimed at creating new oral drug delivery systems.This research has been fueled by unprecedent challenges,such as the need to deliver new, more complex drugs (e.g.proteins, hormones, etc.) that are becoming availablethrough recombinant DNA technology. Thus, considerableattention has been directed at finding ways to increase theintestinal permeability of these compounds. However, theintestinal absorption of these molecules is profoundlylimited by their physicochemical characteristics. Theoreti-cally, three transepithelial pathways are available for thepassage of molecules from the intestinal lumen to thebloodstream (Figure 1): (i) transcellular (i.e. through thecell) carrier-mediated active or facilitated transport; (ii)transcellular passive transport; and (iii) paracellular (i.e.between adjacent cells) transport.

To overcome the intestinal barrier, several strategieshave been developed to target either the transcellular orthe paracellular pathway for drug delivery. The mostpromising techniques currently available will be reviewed,highlighting the advantages and disadvantages of eachsystem.

Transcellular PathwaysThe intestinal epithelium rep-resents the largest interface (more then 200 m2) betweenthe external environment and the internal host milieu andconstitutes a major barrier through which molecules caneither be absorbed or secreted. Conceptually, the phos-pholipid bilayer of the plasma membrane of the epithelialcells that normally line the intestine (the enterocytes) isconsidered to be the major factor restricting the freemovement of substances from the lumen to the bloodstreamthrough the transcellular pathway. The uptake of hydro-phobic molecules usually occurs by passive diffusion,

because the cell membrane behaves like an inert barrierand the molecules enter the cell by endocytosis throughthe apical cell membrane. Facilitated and active transcel-lular transport occurs via specific carriers for smallermolecules, including sugars and amino acids, while theenterocyte membrane is almost impermeable to large andhydrophilic substances, such as proteins. Therefore, newstrategies have been developed that apply the principle ofthe “Trojan horse”: the macromolecules to be delivered arehidden inside hydrophobic, biodegradable microspheresthat can be taken up by endocytosis, by intestinal cells(Figure 1). Even if, theoretically, this seems to be asolution to the problem, several factors may affect theextent of uptake of microparticles, across the gut.

Particle Size, Surface and Intestinal TargetsParticlescurrently used for drug delivery fall into two classes: (i)nanoparticles, ranging in size from 10 to 1000 nm and (ii)microparticles, in the size range 1-1000 µm. For oraldelivery, nanoparticles seem to be more efficiently ab-sorbed, because the uptake of particles within the intestineincreases with decreasing particle size and increasinghydrophobicity.1 Furthermore, the extent and pathway ofnanoparticle uptake is different in different parts of theintestine.2 The M cells of the Peyer’s patches (Figure 1)represent a sort of lymphatic island within the intestinalmucosa and possibly the major gateway through whichparticles can be absorbed.

Dose and Administration VehiclesSeveral studies haveshown that the intestinal uptake of nanoparticles is dosedependent.3,4 Le Fevre and Joel3 have shown that nano-particles were identified in Peyer’s patches with difficultyafter 1 day of feeding, but were readily identified followingchronic feeding. Peroral drug delivery may be furtherenhanced by addition of mucoadhesive substances to thenanoparticles, with subsequent longer interaction of theparticles with the cell membrane.5 An alternative strategyto increase the interaction of nanoparticles with theirtarget cells is to mix them with lipid delivery vehicles, suchas lecithin.6

Animal Species, Age, and Food IngestionsThe extent ofuptake of nanoparticles in rabbits seems to be at least 1order of magnitude greater than in mice, probably becauseof the much greater abundance of M cells in rabbit Peyer’spatches.7 The age of the animal also seems to affectparticle uptake, with greater absorption observed in olderanimals.8 The presence of food seems to be anotherenhancing factor for particle uptake, possibly because itmay increase the intestinal transit time.4

Limitations of the Transcellular Pathway for DrugDeliverysIt should be pointed out that the term “uptake”of particles for gut tissues may include both adsorbedparticles (i.e. particles that remain on the surface of theintestinal cells) and absorbed particles (i.e. particles thatare actually translocated to the bloodstream and aretherapeutically relevant). This means that the high figurereported in some of the literature for the particle uptake9,10

is perhaps an overestimatation of the levels of actual* Corresponding author. Tel 410-3280812. Fax 410-3281072. E-

mail: afasano@umabnet.ab.umd.edu.

© 1998, American Chemical Society and 10.1021/js980076h CCC: $15.00 Journal of Pharmaceutical Sciences / 1351American Pharmaceutical Association Vol. 87, No. 11, November 1998Published on Web 09/10/1998

absorption through the gut. Furthermore, the macromol-ecules contained within the microspheres, once taken upby the intestinal cell, must escape degradation by cellularlysosomes and then cross the basolateral membrane inorder to reach the bloodstream.

For successful exploitation of particle uptake, it isnecessary that the process be both predictable and repro-ducible. Currently, there are contrasting data availabledescribing the extent of particle uptake following repeatedadministration to the same animal. While some research-ers have reported high levels of uptake,9,11 low levels ofuptake have been reported by others.3,4,12,13 The preferredmethodology to quantify total particle uptake remainsunknown, and many of the studies reviewed here were notdesigned with this objective in mind. For example, Jenkinset al.12 were concerned only with evaluating the relativeextent of uptake of alternative microparticles formulationsof different sizes. This study did not accurately determinethe total extent of particle uptake, because particle count-ing was performed only from lymph and Peyer’s patchessamples.

The variable uptake of particles reported in the afore-mentioned studies make it unlikely that the process couldbe successfully applied to the delivery of a wide range ofdrugs. It may be possible, however, to use this technologyfor the oral delivery of drugs that have a wide “therapeuticwindow”, that is, drugs that are active at very lowconcentrations and show limited toxicity at much higherdoses.

Transcellular Pathway for Vaccine DeliverysApathway responsible for the uptake of small numbers ofparticles is unlikely to be appropriate as a delivery mech-anism for a therapeutic dose of a drug, but it might beadequate as a mechanism for stimulating a significant

immune response to an orally delivered microencapsulatedantigen. The oral route for vaccine delivery offers severaladvantages, including high potential patient acceptanceand compliance, less pain and discomfort, and low costs ofproduction and administration, because trained personnelwould not be required to carry out immunizations. Con-sequently, a number of vaccines would be significantlyimproved if they could be administered orally. Oralimmunization might also result in improvements in vaccineefficacy, because oral immunization can stimulate mucosalimmunity. This might prove to be particularly advanta-geous in the elderly, because unlike systemic immunity,mucosal immunity does not appear to be subject to age-associated dysfunction. Oral immunization might also beattractive in the very young, because mucosal immunityappears to develop earlier than systemic immunity.

The majority of the gut-associated lymphoid tissue isorganized into aggregates of lymphoid follicles calledPeyer’s patches. The major physiological role of the Peyer’spatches is the induction of a secretory immune responseto ingested antigens. In humans, the largest Peyer’spatches are found in the terminal ileum and are coveredwith a specialized epithelium that is adapted to allowantigen sampling from the lumen. On contact, antigensare then delivered into the underlying dome structures ofthe patches through specialized cells called M cells. Thereare two important aspects of the uptake and transport ofantigens by M cells: i. antigens will probably escapedegradation, and (ii) the antigen will be released into anenvironment rich in immunocompetent cells (Figure 1).Thus, uptake by M cells can anable the delivery of intactantigens into the immunoinductive environment of thePeyer’s patches. As mentioned above, the M cell representsthe favored route for nanoparticle uptake. Therefore, a

Figure 1sSchematic representation of the three transepithelial intestinal pathways: (a) transcellular active transport, (b) transcellular passive transport, (c)paracellular transport. The carrier-mediated, transcellular active transport is limited to small molecules, such as sugars and amino acids, but the other twopathways are theoretically available for oral delivery of drugs and vaccines because they do not require the presence of specific carriers for the transepithelialtransport of molecules. The transcellular passive transport may be enhanced by entrapment of the active components in microspheres that are more efficientlytaken up through the M cells; particles absorbed through intestinal epithelial cells (enterocytes) are subject to degradation by lysosomes and, therefore, lessefficiently absorbed. The paracellular pathway may be used for drug and peptide delivery by modulating the permeability of tight junctions.

1352 / Journal of Pharmaceutical SciencesVol. 87, No. 11, November 1998

great deal of research has been focused on the delivery ofantigens trapped in particles. Oral immunization withfimbriae from Bordetella pertussis entrapped in nanopar-ticles protected mice from intranasal challenge with thepathogen,14 and whole viruses entrapped in nanoparticlesalso induced protective immunity.15,16 Oral immunizationin mice with nanoparticles induced significant serum IgGand secretory IgA antibody responses;17 the secretory IgAresponse was disseminated throughout the common mu-cosal immune system.17 Hence, oral immunization withmicroencapsulated vaccines potentially offers protectionagainst pathogens which infect the gut, the oral cavity, andthe respiratory and genital tracts.

Several alternative approaches to the oral delivery ofvaccines using polymeric delivery systems other thanmicroencapsulation have also been described, including theuse of enteric coated polymers,18 swellable hydrogels,19 andthe encapsulation of antigens in water-soluble polymers.20

Each of these approaches may have potential advantagesover the use of microparticles, but these have yet to bedemonstrated.

Despite the promising results obtained in animal models,there are still major limitations to antigen delivery throughthe transcellular pathway. These limitations are mainlyrelated to the small number of M cells present within theintestinal mucosa (<0.1% of epithelial cells). In the searchfor alternative solutions, several investigators have usedthe B subunit of cholera toxin (produced by Vibrio cholerae)as an adjuvant to deliver antigens orally, via the entero-cytes.21 It has been demonstrated that, under certainconditions, enterocytes themselves can directly presentantigens.22 These observations suggest that the deliveryof oral vaccines might also be enhanced by harnessing thetranscellular pathway of the major enterocyte populationfor antigen delivery and perhaps even initial antigenprocessing. A fascinating, alternative approach has beenrecently proposed by Kerneis et al.23 who described in ananimal model how to stimulate the conversion of entero-cytes to an M-cell lineage, which more efficiently transportsantigens across the intestinal barrier to the underlyingimmune system.

Paracellular PathwaysWith the exception of thosemolecules that are transported by active or facilitatedtranscellular mechanisms, the absorption of large hydro-philic macromolecules is mainly limited to the paracellularpathway.24 Under normal conditions, however, this path-way is restricted to molecules with molecular radii <11 Åand, therefore, is not accessible to large compounds.

There is now a large body of evidences suggesting thattight junctions (tj) play a pivotal role in epithelial perme-ability. However, the utility of the paracellular route fororal drug delivery has remained unexplored due to ourlimited understanding of tj physiology and the lack ofsubstances capable of increasing the tj permeability with-out irreversibly compromising intestinal integrity andfunction.24-27 Indeed, attempts to find ways to increaseparacellular transport by loosening intestinal tj have beenhampered by unacceptable side effects induced by thepotential absorption enhancing agents tested so far.24-27

For the most part, these agents fall within two classes: (1)calcium chelators and (2) surfactants.26 Both types haveproperties which limit their general utility as a means topromote absorption of various molecules. In the case ofcalcium chelators, Ca2+ depletion induces global changesin the cells, including disruption of actin filaments, disrup-tion of adherent junctions, and diminished cell adhesion.27

In the case of surfactants, the potential lytic nature of theseagents may cause exfoliation of the intestinal epithelium,irreversibly compromising its barrier functions.26

Considering these limitations, it was reasonable toexplore whether findings from basic research on tj regula-tion can be applied to developing new approaches toenhancing drug absorption through the paracellular route.Before addressing these issues, it is worth reviewing someof the structural and biochemical features of tj.

Molecular Composition of the Intestinal Tight JunctionssThere is now substantial evidence that intestinal tj play amajor role in regulating epithelial permeability by influ-encing paracellular flow of fluid and solute. Moreover,structural features of occluding junctions such as strandsnumber often correlate inversely with the permeability ofepithelia as measured electrophysiologically.28 A centuryago, tj were thought to be a secreted extracellular cementforming an absolute and unregulated barrier within theparacellular space.29 There is now abundant evidence thattj are dynamic structures that readily adapt to a varietyof developmental,30,31 physiological,32-34 and pathological35-37

circumstances.The assembly of tj is the result of cellular interactions

that trigger a complex cascade of biochemical events thatultimately lead to the formation and modulation of anorganized network of tj elements, the composition of whichhas been only partially characterized38 (Figure 1). Acandidate for the transmembrane protein strands, occludin,has recently been identified.39 Six proteins have beenidentified in a cytoplasmic submembranous plaque under-lying membrane contacts, but their function remains to beestablished.38 ZO-1 and ZO-2 exist as a heterodimer40 ina detergent-stable complex with an uncharacterized 130kD protein (ZO-3). Most immunoelectron microscopicstudies have localized ZO-1 to precisely beneath membranecontacts.41 Two other proteins, cingulin42 and the 7H6antigen,43 are localized further from the membrane andhave not yet been cloned. Rab 13, a small GTP bindingprotein, has also recently been localized to the junctionregion.44 Other small GTP-binding proteins are known toregulate the cortical cytoskeleton, i.e., rho regulates actin-membrane attachment in focal contacts,45 and rac regulatesgrowth factor-induced membrane ruffling.46 On the basisof analogy with the known functions of plaque proteins inthe better characterized cell junctions, focal contacts,47 andadherens junctions,48 it has been hypothesized that tj-associated plaque proteins are involved in transducingsignals in both directions across the cell membrane, andin regulating links to the cortical actin cytoskeleton.

Regulation of Intestinal Tight JunctionssTo meetthe many diverse physiological and pathological challengesto which epithelia are subjected, tj must be capable of rapidand coordinated responses that require the presence of acomplex regulatory system. The precise characterizationof the mechanisms involved in the assembly and regulationof tj is an area of current active investigation. Thediscovery of Zonula occludens toxin (Zot) elaborated byVibrio cholerae49,50 sheds some light on the intricatemechanisms involved in the regulation of tj permeability.Zot seems to activate a complex intracellular cascade ofevents that regulate the intestinal permeability.51 Themolecule induces a dose- and time-dependent PKCR-relatedpolymerization of actin filaments strategically localized toregulate the paracellular pathway.51 These changes are aprerequisite to opening of tj and are evident at a toxinconcentration as low as 1.1 × 10-13 M.52 The toxin exertsits effect by interacting with a specific surface receptor thatis present on mature cells of small intestinal villi, but notin the colon.52 The regional distribution of Zot receptor(s)coincides with the different permeabilizing effect of thetoxin on the various tracts of intestine tested.52 Both invivo52,53 and in vitro 49,52,53 studies demonstrated that theeffect of Zot on tissue permeability occurs within 20 min

Journal of Pharmaceutical Sciences / 1353Vol. 87, No. 11, November 1998

the addition of the protein to the intestinal mucosa and isreadily reversible once the toxin is removed.

Use of Zot as a Tool for Oral Drug DeliverysZotdisplays multiple properties that make it the most promis-ing tool currently available to enhance drug and peptidetransport through the intestinal mucosa. Zot (a) is notcytotoxic and does not affect the viability of the intestinalepithelium ex vivo,49,51 (b) fails to completely abolish theintestinal transepithelial resistance,49,51,53 (c) interacts witha specific intestinal receptor whose regional distributionwithin the intestine varies,52 (d) is not effective in the largeintestine where the presence of the colonic microflora couldbe potentially harmful if the mucosal barrier was compro-mised,52,53 (e) does not induce acute systemic side-effects(for at least 80-90 h) when orally administered,53 and (f)induces a reversible increase of tissue permeability.49,51,53

To establish the efficacy of Zot as an intestinal absorptionenhancer, we selected insulin and immunoglobulin G (IgG).This choice was based on the relative size and structure,biological activities, and the therapeutic relevance of theseproteins. In vitro experiments in the rabbit ileum mountedin Ussing chambers demonstrated that Zot (1.1 × 10-10 M)reversibly increases the intestinal absorption of bothinsulin (by 72%) and IgG (by 52%) in a time-dependentmanner.53 Zot permeabilizing effect peaked at 80 min andwas completely reversible within 20 min of the withdrawalof the toxin from the Ussing chambers (Figure 2). ThisZot-induced increase in absorption coincided with a reduc-tion in tissue resistance (Rt) (Figure 2). When tested inthe intact host by using the rabbit in vivo perfusion assay,Zot (1.1 × 10-10 M) increased the passage of insulin acrossboth the jejunum and distal ileum 10-fold, whereas nosubstantial changes were observed in the colon53 (Figure

3). The increased absorption of insulin was reciprocal witha shift of water absorption toward secretion (Figure 3), achange that has been related to the permeabilizing effectof Zot on the paracellular pathway in vivo.52 This effectwas detectable as soon as 20 min after Zot perfusion inthe small intestine and was completely reversible within60 min of its withdrawal (Figure 3). Zot also reversiblyincreased the serum concentration of both insulin and thenonabsorbable marker [14C]poly(ethylene glycol) (PEG)-4000 from jejunum and ileum, but not from colon.53

Similar results were obtained with IgG, whereby Zot (1.1× 10-10 M) induced 2-fold and 6-fold increases of IgGabsorption in the jejunum and ileum, respectively. Again,no increases in absorption were detected in the colon.53

To evaluate the bioactivity of insulin after enteralcoadministration with Zot, the hormone was orally admin-istered to acute type 1 diabetic male BB/Wor rats with orwithout Zot, and the blood glucose levels of the rats wereserially measured. After oral administration of insulinalone, given at doses between 5 and 30 IU, blood glucoselevels of treated animals were not appreciably lowered.53

In contrast, when insulin at doses as low as 10 IU wasorally coadministered with Zot (1.1 × 10-10 mol, 5 mg), asignificant reduction in blood glucose concentration wasobserved (Figure 4). This decrement was comparable tothat seen with a conventional dose of SQ insulin andreturned to baseline by 6 h postadministration (Figure 4).None of the animals treated with insulin+Zot experiencedfever or other systemic symptoms, and no structuralchanges could be demonstrated in the small intestine onhistological examination.53 Furthermore, Zot administra-tion did not induce diarrhea, despite the secretory effectof the toxin. The lack of diarrhea is probably related to

Figure 2sReversible effect of purified Zot on tissue resistance (A and B) and transepithelial transport of insulin (C) and IgG (D) in rabbit ileum in vitro. Pairedtissues, matched on the basis of their resistance, were exposed luminally to either [125I]insulin 10-11 M (2 mCi ) 10-12 M) (left panel) or [125I]IgG 156.25 ng (1mCi ) 83.3 ng) (right panel), alone (O) or in the presence of 1.1 × 10-10 M Zot (0). After 80 min of incubation, the Ringer’s solutions were replaced withsolutions of identical composition but without Zot. Zot reversibly increased the transepithelial absorption of both insulin and IgG. These effects paralleled the Rtdecrement induced by the toxin. No. of animals ) 4.

1354 / Journal of Pharmaceutical SciencesVol. 87, No. 11, November 1998

the distribution of the Zot receptors within the intestine.52

Following the activation of the Zot receptors in the smallintestine, tj are reversibly opened and fluid leaks into theintestinal lumen driven by the osmotic gradient. Theexcess of fluid that accumulates in the small intestine iscompletely reabsorbed in the colon (where tj regulation isnot operative because of the lack of Zot receptors), prevent-ing intestinal fluid loss and, therefore, diarrhea.

Taken together, these results demonstrate that coad-ministration of Zot with biologically active ingredientsenhances intestinal absorption of the active molecule, andthat this enhancement is effective for both relatively small(5733 Da: insulin) and large molecules (140-160 kDa:IgG). Furthermore, the experiments in BB/Wor diabeticrats demonstrate that orally delivered insulin can retainits biological activity without provoking severe hypoglyce-mia within the range of the insulin administered, i.e., upto 15 times more than the effective parenteral insulin dose.

These findings have important practical implications, sincethe insulin therapeutic index (i.e. the ratio between themedian toxic dose and the median therapeutic dose) isrelatively low.

ConclusionsCurrent knowledge on the regulation of intestinal tight

junctions by Zonula occludens toxin were applied to en-hance the absorption of macromolecules normally notabsorbed through the intestine. The promising resultsobtained in the animal model, both in vitro and in vivo,represent an encouraging basis for further studies toestablish the possible clinical applications of this systemfor the treatment of human diseases that currently requirefrequent and long-life parenteral drug administration.

Figures 2-4 were reproduced from J. Clin. Invest. 1997,99, 1158-64, by copyright permission of the AmericanSociety for Clinical Investigation. This work was supportedby grants from the National Institute of Health, DK-48373and AI-35740.

References and Notes1. Kreuter, J. Nanoparticles and microparticles for drug and

vaccine delivery. J. Anat. 1996, 189, 503-505.2. Michel, C.; Aprahamian, M.; Defontaine, L.; Couvreur, P.;

Damge′, C. The effect of site of administration in thegastrointestinal tract on the absorption of insulin fromnanoparticles in diabetic rats. J. Pharm. Pharmacol. 1991,43, 1-5.

3. Le Fevre, M. E.; Joel, D. D. In Intestinal toxicology; Schiller,C. M., Ed.; Raven Press: New York, 1984; pp 45-56.

4. Ebel, J. P. A method to quantify particle absorption from thesmall intestine of the mouse. Pharm. Res. 1990, 7, 848-851.

5. Kreuter, J. Nanoparticles. In Colloidal Drug Delivery Sys-tems; Kreuter, J., Ed.; Marcel Dekker: New York, 1994; pp219-342.

6. Thomas, N. W.; Jenkins, P. G.; Howard, K. A.; Smith, M.W.; Lavelle, E. C.; Holland, J.; Davis, S. S. Particle uptakeand translocation across epithelial membranes. J. Anat.1996, 189, 487-490.

Figure 3sEffect of purified Zot on water (9) and insulin (0) transport, asdetermined by the in vivo perfusion assay, in rabbit jejunum, distal ileum, andcolon. Note the reversible increment of insulin absorption that Zot induced inthe small, but not in the large intestine. This effect coincided with the decreasedabsorption of water evoked by the toxin.

Figure 4sEffect of oral insulin 10 IU, alone (9) or in the presence of 5 µgof Zot (0) on serum glucose decrement in BB/Wor diabetic rats. Thecoadministration of Zot induced a reduction in blood glucose concentrationcomparable to that seen with a conventional dose of SQ insulin (b) andreturned to baseline by 6 h postadministration. Blood glucose decrement ofuntreated animals (O) and animal treated with oral Zot alone (2) are shownfor comparison. No. of observations ) 3 (*) p ) 0.005; (‡) p ) 0.003; (§) p) 0.009, as compared to oral insulin alone.

Journal of Pharmaceutical Sciences / 1355Vol. 87, No. 11, November 1998

7. Pappo, J.; Ermak, T. H. Uptake and translocation of fluo-rescent latex particles by rabbit Peyer’s patch follicle epi-thelium: a quantitative model for M cell uptake. Clin. Exp.Immunol. 1989, 76, 144-148.

8. Le Fevre, M. E.; Boccio, A. M.; Joel, D. D. Proc. Soc. Exp.Biol. Med. 1989, 190, 23-27.

9. Jani, P. U.; Halbert, G. W.; Langridge, J.; Florence, A. T.Nanoparticle uptake by the rat gastrointestinal mucosa:quantitation and particle size dependency. J. Pharm. Phar-macol. 1990, 42, 821-826.

10. Jani, P. U.; Florene, A. T.; McCarthy, D. E. Further histologi-cal evidence of the gastrointestinal absorption of polystyrenenanoparticles in the rat. Intern. J. Pharm. 1992 84, 245-252.

11. Eyles, J.; Alpar, H. O.; Field, W. N.; Lewis D. A.; Keswick,M. J. Pharm. Pharmacol. 1992, 47, 561-565.

12. Jenkins, P. G.; Howard, K. A.; Blackhall, N. W.; Thomas, N.W.; Davis, S. S.; O’Hogan D. T. J. Controlled Rel. 1994, 29,339-350.

13. O’Hagan, D. T. The intestinal uptake of particles and theimplications for drug and antgen delivery. J. Anat. 1996, 189,477-482.

14. Jones, D. H. et al. Protection of mice from Bordetella pertussisrespiratory infection using orally administered microincap-sulated pertussis fimbriae. Infect. Immun. 1996, 64, 489-494.

15. Moldovenwanu, Z.; Novak, M.; Huang W.-Q. Oral immuniza-tion with influenza virus in biodegradable microspheres. J.Infect. Dis. 1993, 167, 84-90.

16. Ray, R. et al. Microincapsulated human parainfluenza virusinduces a protective immune response. J. Infect. Dis. 1993,167, 752-755.

17. Challacombe, S. J.; Rahman, D.; Jeffery, H., Davis, S. S.;O’Hagan, D. T. Enhanced secretory IgA and systemic IgGantibody responses after oral immunization with biodegrad-able microparticles containing antigen. Immunology 1992,76, 164-168.

18. Klipstein, F. A.; Engbert, R. F.; Sherman, W. T. Peroralimmunization with Escherichia coli heat-labile enterotoxindelivered by microspheres. Infect. Immun. 1983, 39, 1000-1003.

19. Bowerstock, T. L. et al. The potential use of poly (methacrylicacid) hydrogels for oral administration of drugs and vaccinesin ruminants. J. Contolled Rel. 1994, 31, 245-254.

20. Offit, P. A. et al. Enhancement of rotavirus immunogenicityby microincapsulation. Virology 1994, 203, 134-143.

21. Elson, C. O.; Ealding, W. Generalized systemic and mucosalimmunity in mice after mucosal stimulation with choleratoxin. J. Immunol. 1984, 132, 2736-2741.

22. Mayer, L.; Schlien F. Evidence for function of Ia moleculeson gut epithelial cells in man. J. Exp. Med. 1987, 166, 1471-1476.

23. Kerneis, S.; Bodganova, A.; Kraehenbuhl, J.-P.; Pringault,E. Conversion by Peyer’s patch lymphocytes of humanenterocytes into M cells that transport bacteria. Science 1997,277, 949-952.

24. Lee, V. H. L.; Yamamoto A.; Kompella V. B. Mucosalpenetration enhancers for facilitation of peptide and proteindrug absorption. Crit. Rev. Ther. Drug Carrier Syst. 1991,8(2), 91-192.

25. Muranishi, S. Absorption enhancers. Crit. Rev. Ther. DrugCarrier Syst. 1990, 7(1), 1-33.

26. Hochman, J.; Artursson, P. Mechanisms of absorption en-hancement and tight junction regulation. J. Controlled Rel.1994, 29, 253-267.

27. Citi, S. Protein kinase inhibitors prevent junction dissociationinduced by low extracellular calcium in MDCK epithelialcells. J. Cell Biol. 1992, 117(1), 169-178.

28. Madara, J. L.; Dharmsathaphorn. Occluding junction struc-ture-function relationship in a cultured epithelial mono-layer. J. Cell Biol. 1985, 101, 2124-2133.

29. Cereijido, M. Evolution of ideas on the tight junction. In TightJunctions; CRC: Boca Raton, FL, 1992; pp 1-13.

30. Magnuson, T.; Jacobson, J. B.; Stackpole, C. W. Relationshipbetween intercellular permeability and junction organizationin the preimplantation mouse embryo. Dev. Biol. 1978, 67,214-224.

31. Schneeberger, E. E.; Walters, D. V.; Olver, R. E. Developmentof intercellular junctions in the pulmonary epithelium of thefoetal lamb. J. Cell Sci. 1978, 32, 307-324.

32. Madara, J. L., Pappenheimer, J. R. Structural basis forphysiological regulations of paracellular pathways in intes-tinal epithelia. J. Membr. Biol. 1987, 100, 149-164.

33. Mazariegos, M. R.; Tice, L. W.; Hand, A. R. Alteration of tightjunctional permeability in the rat parotid gland after iso-proteranol stimulation. J. Cell. Biol. 1984, 98, 1865-1877.

34. Sardet, C.; Pisam, M.; Maetz, J. The surface epithelium ofteleostean fish gills. Cellular and junctional adaptations of

the chloride cell in relation to salt adaptation. J. Cell Biol.1979, 80, 86-117.

35. Milks, L. C.; Conyers, G. P.; Cramer, E. B. The effect ofneutrophil migration on epithelial permeability. J. Cell Biol.1986, 103, 2729-2738.

36. Nash, S.; Stafford, J.; Madara, J. L. The selective andsuperoxide-independent disruption of intestinal epithelialtight junctions during leukocyte transmigration. Lab. Invest.1988, 59, 531-537.

37. Shasby, D. M.; Winter, M.; Shasby, S. S. Oxidants andconductance of cultured epithelial cell monolayers: inositolphospholipid hydrolysis. Am. J. Physiol. 1988, 255 (CellPhysiol. 24), C781-C788.

38. Anderson, J. M.; Balda, M. S.; Fanning, A. S. The structureand regulation of tight junctions. Cell. Biol. 1993, 5, 772-778.

39. Furuse, M.; Hirase, T.; Itoh, M.; Nagafuchhi, A.; Yonemura,S.; Tsukita, S.; Tsukita, S.; Ocludin: a novel integralmembrane protein localizing at tight junctions. J. Cell Biol.1992, 123, 1777-1788.

40. Gumbiner, B.; Lowenkopf, T.; Apatira, L. Identification of160-kDa polypeptide that binds to the tight junction proteinZO-1. Proc. Natl. Acad. Sci. U.S.A. 1991, 88, 3460-3464.

41. Stevenson, B. R.; Anderson, J. M.; Bullivant, S. The epithelialtight junction: structure, function, and preliminar biochemi-cal characterization. Mol. Cell Biochem. 1988, 83, 129-145.

42. Citi, S.; Sabannay, H.; Jakes, R.; Geiger, B.; Kendrich-Jones,J. Cingulin, a new peripheral component of tight junctions.Nature 1988, 333, 272-275.

43. Zhong, Y.; Saitoh, T.; Minase, T.; Sawada, N.; Enomoto, K.Monoclonal antibodies 7H6 reacts with a novel tight-junctionassociated protein distinct from ZO-1, cingulin and ZO-2. J.Cell Biol. 1993, 120, 477-483.

44. Zahraoui, A.; Joberty, G.; Arpin, M.; Fontaine, J. J.; Hellio,R.; Tavitian, A.; Louvard, D. A small rab GTPase is distrib-uted in cytoplasmatic vesicles in non polarized cells but co-localized with the tight junction marker ZO-1 in polarizedepithelial cells. J. Cell Biol. 1994, 124, 101-115.

45. Ridley, A. J.; Hall, A. The small GTP-binding protein rhoregulates the assembly of focal adhesions and actin stressfibers in response to growth factors. Cell 1992, 70, 389-399.

46. Ridley, A. J.; Paterson, H. F.; Johnston, C. L.; Diekmann,D.; Hall, A. The small GTP-binding protein rac regulatesgrowth factor-induced membrane ruffling. Cell 1992, 70,401-410.

47. Guan, J. L.; Shalloway, D. Regulation of focal adhesion-activated protein tyrosin kinase by both cellular adhesionand oncogenic transformation. Nature 1992, 358, 690-692.

48. Tsukita, S.; Itoh, M.; Nagafuchi, A.; Yonemura, S.; Tsukita,S. Submembraneous junctional plaque proteins includepotential tumor suppressor molecules. J. Cell Biol. 1993, 123,1049-1053.

49. Fasano, A.; Baudry, B.; Pumplin, D. W.; Wasserman, S.S.;Tall, B. D.; Ketley, J. M.; Kaper, J. B. Vibrio choleraeproduces a second enterotoxin, which affects intestinal tightjunctions. Proc. Natl. Acad. Sci. U.S.A. 1991, 88(22), 5242-5246.

50. Baudry, B.; Fasano, A.; Ketley, J. M.; Kaper, J. B. Cloningof a gene (zot) encoding a new toxin produced by Vibriocholerae. Infect. Immun. 1992, 60(2), 428-434.

51. Fasano, A.; Fiorentini, C.; Donelli, G.; Uzzau, S.; Kaper, J.B.; Margaretten, K.; Ding, X.; Guandalini, S.; Comstock, L.;Goldblum, S. E. Zonula occludens toxin modulates tightjunctions through protein kinase C-dependent actin reorga-nization, in vitro. J. Clin. Invest. 1995, 96, 710-720.

52. Fasano, A.; Uzzau, S.; Fiore, C.; Margaretten, K. Theenterotoxic effect of Zonula occludens toxin on rabbit smallintestine involves the paracellular pathway. Gastroenterology1997, 112, 839-846.

53. Fasano, A.; Uzzau, S. Modulation of intestinal tight junctionsby Zonula occludens toxin permits enteral administration ofinsulin and other macromolecules in an animal model. J.Clin. Invest. 1997, 99, 1158-1164.

JS980076H

1356 / Journal of Pharmaceutical SciencesVol. 87, No. 11, November 1998