WO2002034880A2 - Cadherin peptides for drug delivery and inhibition of tumor metastasis/invasion - Google Patents

Abstract

Peptides which modulate porosity of intercellular junctions by inhibiting E-cadherin-E-cadherin interactions are provided. These peptides are derived from the bulge and groove regions of E-cadherin. For some peptides, a portion of a sequence derived from a groove region is conjugated with a portion of a sequence derived from a bulge region via a linker. By inhibiting E-cadherin-E-cadherin interactions, the transepithelial electrical resistance of cells is decreased, paracellular transport is increased, and adhesion of single cells to cell layers is inhibited. Accordingly, the present invention is useful for inhibiting tumor metastasis, and for delivery of protein drugs across biological barriers.

Description

CADHERIN PEPTIDES FOR DRUG DELIVERY AND INHIBITION OF TUMOR METASTASIS/INVASION

BACKGROUND OF THE INVENTION

Field ofthe Invention

The present invention is concerned with peptides and proteins which can be used as therapeutic agents. More particularly, the present invention is concerned with peptides and proteins which modulate intercellular junctions. More particularly, the present invention is concerned with peptides and proteins derived from the groove and bulge regions of cadherins. Finally, and most particularly, the present invention is concerned with peptide sequences derived from the EC-1 through EC-4 domains of cadherins which inhibit cell-cell adhesion mediated by E-cadherin, thereby modulating the intercellular junctions of cell monolayers and tumor cell metastasis and invasion.

Description ofthe Prior Art

E-cadherin, a member ofthe Cell Adhesion Molecule (CAM) family, plays an important role in the formation and maintenance of tight intercellular junctions in the biological barricades such as the intestinal mucosa and the blood-brain barrier. Tight intercellular junctions are one ofthe barriers drugs encounter during drug permeation via the paracellular route. Thus, modulation of the intercellular junctions may be beneficial in improving paracellular permeation of large hydrophilic drugs (i.e., peptides and proteins) across the biological barriers. In the past, application of peptides and proteins as therapeutic agents has been restricted by their inability to cross the biological barricades and reach the target site. Due to the physicochemical properties of peptides and proteins, they cannot passively diffuse through the biological barriers via the transcellular route. Similarly, they cannot cross via the paracellular route due to the presence of the tight intercellular junctions. Paracellular transport exhibits a minimal porosity of >11 A, enabling only small molecules and ions to cross this path. Thus, any method which modulates the intercellular junctions may impact paracellular permeability of peptides and proteins across the biological barriers. Ultrafine structural studies of the tight intercellular junctions have revealed the existence of complex morphological structures that can be divided into three layers; (a) zonula occludens (tight junctions), (b) the zonula adherens (adherens junctions), and (c) the macula adherens (desmosomes). The zonula occludens are mediated by transmembrane proteins called occludens at the extracellular space and ZO-1, ZO-2 and ZO-3 at the intracellular space. The Ca²⁺-dependent proteins, E-cadherins, are localized within the adherens junctions (zonula adherens) and they are known to primarily stabilize and maintain the cell-cell adhesion at the intercellular junction between the opposing cells. The E-cadherins interact with the a-, a-, and a-catenins in the cytoplasmic domain. Finally, the macula adherens (desmosomes) are mediated by transmembrane adhesion proteins called desmocollins (for example, Dsc-1 and Dsc-2), desmogleins (for example, Dsg-1 and Dsc-2), and the cytoplasmic plaks (for example, plactoglobin and desmoplakin). E-cadherin consists of an extracellular (EC) domain, a membrane spanning region, and a short cytoplasmic domain. The extracellular section of E- cadherin contains five tandemly repeat domains, EC-1 to EC-5. The active conformation of these domains is dependent upon the presence of Ca²⁺ ions. E-cadherins protrude from the same cell surface as a parallel cis-dimex. Additionally, they form an antiparallel trαrø-dimer with E-cadherins from opposing cells. The formation of these tr π-s-dimers to make the adherens junctions is Ca²⁺ binding at the interconnection of

EC repeats. The EC-1 domain is important for the selectivity of E-cadherin in binding with another E-cadherin but not N- or P-cadherin.

Many peptides and proteins, newly synthesized or occurring naturally, have been found and developed as potential therapeutic agents. However, the utilization of such compounds as therapeutic drugs is often restricted by the difficulties of delivering them to target site(s) due to the presence of biological barricades such as the intestinal mucosa and the blood-brain barrier (BBB). These barriers usually consist of cell membranes constructed from cells with intercellular junctions. Peptides and proteins cannot cross these barriers via transcellular pathways due to their size and hydrophilic properties. Alternatively, these molecules may be transported through paracellular pathways. Unfortunately, the paracellular transport of peptides and proteins is limited by the presence of tight intercellular junctions. These tight junctions exhibit a minimal porosity (<11 A), allowing only small molecules and ions to cross. Therefore, there is a need to develop methods which improve paracellular delivery of large hydrophilic molecules such as peptides and proteins. One way to accomplish this goal is by increasing the porosity of the intercellular junctions via the modulation of protein- protein interactions (i.e., those of E-cadherins) in the intercellular junctions. Therefore, what is needed are compounds which modify paracellular transport and which can be delivered to their target site. It is known that His- Ala-Nal (HAN) peptides derived from the EC 1 domain of cadherin can inhibit cell-cell adhesion mediated by E-cadherin and modulate the intercellular junctions of cell monolayers. Mutations of the HAN sequence in E- cadherin have been shown to eliminate cell-cell contact. Flanking residues to the HAN sequence are also important for peptide selectivity as these peptides can modulate the cellular junctions in MDCK and bovine brain microvessel endothelial (BBME) cells.

In addition, other possible interaction regions have been identified using antibodies to E-cadherin.

E-cadherin sequences that were important for cadherin-cadherin interactions were identified using ELISA. An anti-E-cadherin antibody directed against the extracellular domain of E-cadherin has been shown to prevent the resealing of tight junctions; therefore, this anti-E-cadherin antibody may recognize the peptide sequences responsible for homophilic cadherin-cadherin interactions. Thus, the cadherin-cadherin binding region can be identified using an antibody to E-cadherin (anti-E-cadherin or anti-Uvomorulin). By evaluating the binding of anti-E-cadherin antibody to synthetic peptides derived from various sequences of cadherins, the antigenic reactive sequences that may be responsible for cadherin-cadherin interactions can be located.

Table 1. The Peptide Sequences from the EC 1 Domain of N- and E- Cadherin Used for Solution and Immobilized Peptide ELISA.

^aAmcap = ε-aminocaproic acid. Bold letters represent conserved amino acid sequences ontheECl domain. Underlines represent conserved calcium binding sequences.

Using normal ELISA, E-cadherin peptide sequences were recognized by anti-E- cadherin antibody. SEQ ID Νos. 1, 2 and 3 (Table 1; Figure 1) were synthesized by peptide synthesizer and purified by HPLC. SEQ ID Νos. 1 and 2 were derived from the

EC1 region of Ν- and E-cadherins, respectively, and they contained the conserved histidine-alanine-valine (HAV) sequence [Lutz et al., Peptide Research, 9, 233-239 (1996). Figure 1 shows binding of Ν- and E-cadherin peptides to anti-E-cadherin antibody using normal ELISA for the peptides noted as SEQ ID Νos. 1-3 and immobilized peptide ELISA for SEQ ID Νos. 4-8. As shown by Fig. 1 wherein the numbers 1-8 represent SEQ ID Νos. 1-8, respectively, SEQ ID Νos. 1 and 2 were recognized by the anti-E-cadherin antibody, however, the control, peptide, SEQ ID Nos. 3, was not. SEQ ID No. 2 displayed a greater affinity for the anti-E-cadherin antibody than did SEQ ID Nos. 1 because SEQ ID No. 2 contains two conserved sequences, the HAN sequence and the conserved calcium binding sequence Asp-Arg-Glu (DRE), at the Ν-terminus. Also, the higher antigenic reactivity could result from the increase in propensity of SEQ

ID No. 2 to form a preferable secondary structure that is recognized by the antibody. An immobilized peptide assay that involves immobilization of peptides on the surface of glass beads (controlled pore glass, CPG) using a covalent bond. This assay is very fast compared to the first assay because the peptide is synthesized directly on the glass beads. To test the validity ofthe immobilized peptide-antibody binding assay, SEQ ID

No. 1, which is known to bind the antibody, was synthesized on CPG beads (SEQ ID No. 7; Table 1). Decapeptides neighboring this N-cadherin sequence were also synthesized (SEQ ID Nos. 4, 5, 6 and 8; Table 1). These immobilized decapeptides, SEQ ID Nos. 4-8, represent a 50 amino acid sequence of N-cadherin from the N- to C- termini. The antibody binding was analyzed using the primary and secondary antibodies. SEQ ID No. 7 exhibits a binding activity similar to that of SEQ ID No. 1 because they both possess a similar sequence (Fig. 1). In addition, anti-E-cadherin antibody also recognized SEQ ID No. 6; this peptide contains a calcium binding sequence. SEQ ID Nos. 4, 5 and 8 did not show antibody binding, and these peptides can be considered as negative controls. The recognition of SEQ ID Nos. 6 and 7 indicates that multiple epitopes in N-cadherin are recognized by the anti-E-cadherin antibody. Only peptides containing conserved sequences displayed significant antigenic reactivity to the anti-E-cadherin antibody.

Table 2. The sequences of immobilized peptides ofthe ECl, EC2 and EC3 domains of E-cadherin.

^aThe number in parenthesis is the amino acid number in E-cadherin. The peptides were synthesized on derivatized controlled pore glass (CPG). Bold letters represent conserved amino acid sequences among cadherins. Underlined letters represent conserved calcium binding sequences.

After observing consistent results among normal and immobilized peptide ELISAs, this study was directed to scanning the first three extracellular domains of mouse E-cadherin. Immobilized decapeptides from the ECl, EC2 and EC3 domains were synthesized (Table 2). The binding between SEQ ID Nos.9-39 (Table 2) and anti- E-cadherin antibody is shown in Fig. 2, wherein the numbers 9-39 represent SEQ ID

Nos. 9-39, respectively.

Immobilized decapeptides, SEQ ID Nos. 9-18 represent a 95 amino acid sequence from the N- to the C-terminal direction of ECl of E-cadherin (Table 2). Two peptides derived from this domain displayed high antigenic reactivity (Fig.2a). SEQ ID No. 16 contains the conserved HAN sequence (ILYSHAVSSΝ), whereas SEQ ID No.

18 (NITVTDQNDN) contains a conserved calcium binding sequence (DQNDN). The high antigenic reactivity of SEQ ID No. 16 correlates with the antigenic reactivity of SEQ ID Nos. 1, 2 and 7 from the initial study (Fig. 1). These conserved His and Val residues of E-cadherin are solvent-exposed amino acids on the surface of ECl, and it has been inferred that this region confers homophilic specificity on cadherins.

The EC2 domain of E-cadherin also displayed two antigenic reactive peptide sequences, SEQ ID Nos. 23 and 27 (Fig. 2b). SEQ ID No. 27 is positionally aligned with the same 10 amino acids as SEQ ID No. 16, which contains the conserved HAV sequence from the ECl domain. SEQ ID No. 27 contains a QAA (Gin- Ala- Ala) sequence instead ofthe HAV sequence. The reactivity of SEQ ID No. 27 suggests that this sequence is similar to the sequence of SEQ ID No. 16 from the ECl domain; the X-ray structure also confirms this finding. However, the calcium-binding sequence of EC2 (SEQ ID No. 29) did not bind to the anti-E-cadherin antibody. It is not clear if this is due to the presence of different residues (Lys and He) in the calcium-binding sequence. SEQ ID No. 23 is a new sequence from the EC2 region that is recognized by the antibody. The EC2 domain has a tertiary structure similar to the EC 1 domain. SEQ ID No.23 (YTIVSQDPEL) can be placed around the homophilic specificity surface of E-cadherin along with the sequence of SEQ ID No. 27.

The peptide mapping ofthe EC3 domain of E-cadherin displays strong binding of SEQ ID No. 35 and moderate binding of SEQ ID No. 31 (Fig. 2c). These two peptides contain no cadherin-conserved sequences. SEQ ID No. 37, which is similar in position to the HAN-peptide, SEQ ID No. 16, showed no antibody binding. SEQ ID No. 37 has an HAR sequence instead of an HAN sequence. The presence of the positively charged residue Arg may suppress antibody binding. E-cadherins work as glue between the cellular junctions of biological barricades, including the intestinal mucosa and the blood-brain barrier. E-cadherins also mediate the intercellular junctions of MDCK cell monolayers. The exact mechanisms of E- cadherin-E-cadherin interactions between opposing cells are not well understood. It has been shown that E-cadherins can form cis- and trans-άimsτ interactions (see above).

In this respect, the ECl domain is critical in forming the trarø-dimer formation. Furthermore, the HAV sequence in ECl is important for trαrø-dimer interactions. Additionally, peptides containing the HAV sequence can bind to and inhibit cell-cell aggregation in BBMEC as well as modulate the intercellular junctions of BBMEC monolayers. Unfortunately, the counter-sequence, in which the HAV sequence can interact in the EC domains ofthe partner E-cadherin is not understood. For example, the HAV sequence of the ECl domain of E-cadherin from one cell may interact with a different sequence in the EC 1 , EC2, EC3 , EC4 or EC5 domain of E-cadherin from the opposing cell.

SUMMARY OF THE INVENTION The present invention modulates the porosity ofthe tight intercellular junctions by inhibiting E-cadherin-E-cadherin interactions. E-cadherin-E-cadherin interactions can be disrupted using synthetic peptides derived from the sequence of contact (bulge and groove) regions of these interactions. E-cadherins are a family of transmembrane glycoproteins found in the zonula adherens (adherens junction), which are sandwiched between tight junctions (zonula adherens) and desmosomes. Homophilic interactions of E-cadherins are the primary force for cell-cell adhesion ofthe opposing cells. The formation of tight junctions is a secondary response to this primary interaction. The structure of E-cadherin contains an extracellular (EC) domain, a single membrane- spanning segment, and a relatively short cytoplasmic domain. The extracellular domain has five tandem repeats called ECl-to-EC5 domains, which bind to calcium ions. These calcium ions are located between the interconnections of EC repeats. For example, three calcium ions are bound to the interface between the ECl and EC2 domains. These calcium ions are important for the structural integrity of cadherins and for cadherin-cadherin interactions. The extracellular domain of E-cadherin forms a parallel cz-s-dimer (strand dimer) between two E-cadherin molecules in the same cells, and an antiparallel trαrø-s-dimer (adhesion dimer) between E-cadherin molecules from opposing cells. Therefore, inhibition of cadherin-cadherin interaction maybe achieved by blocking the recognition sites for cis- or trαn-s-cadherin interactions. The cytoplasmic tail anchors to cytoskeletal actin filaments through catenins. The cytoplasmic cadherin-catenin binding is also important in regulating cadherin-cadherin interactions in the extracellular space.

The present invention identifies a region counter to the HAV sequence that is responsible for tr ns -cadherin interaction. This region, called a bulge region, contains the sequence QGADTPPVGV which interacts with the HAV sequence, called a groove region. This interaction was found by using the X-ray structure of the EC1-EC2 domains of E-cadherins. Several peptides were derived from the bulge region and evaluated for their ability to modulate the intercellular junction of MDCK cell monolayers. The biological activities of these bulge region peptides were compared to those of the groove region peptides (HAV peptides). Perturbation of intercellular junctions by peptides was followed using two different parameters including the ability of peptides to decrease transepithelial electrical resistance (TEER) and the ability of peptides to increase mannitol flux via the MDCK monolayers. In another aspect of the present invention, bulge region peptides were conjugated with groove region peptides using a non-natural amino acid as the linker. A preferred linker in this respect is aminocaproic acid which is a derivative of the amino acid lysine. As is well known in the art, other non-natural amino acid linkers such as analogs of aminocaproic acid will work for purposes ofthe present invention. One preferred analog is tranexamic acid. The ability of FITC-labeled peptides to bind directly to the intercellular junctions of MDCK monolayers was also evaluated.

E-cadherins mediate cell-cell adhesion and play important roles in many biological processes, ranging from embryonal morphogenesis to the maintenance ofthe integrity of epithelial tissues in adult organisms, from bacterial entry to tumor metastasis. Understanding the molecular and atomic levels of E-cadherin-E-cadherin interactions is useful in designing molecules that can modulate various cellular functions mediated by E-cadherins. One application of the modulation of cadherin- cadherin interactions is the improvement of paracellular delivery of peptide and protein drugs via intercellular junctions. The present invention identified and synthesized peptides derived from the groove and bulge regions of ECl domain of E-cadherin and tested these peptides for their ability to modulate cadherin-cadherin interactions in the intercellular junctions in MDCK cell monolayers. Naturally, other groove and bulge sequences (i.e. from other EC domains) having similar repeating units will work in the same fashion due to their similar structures. Another application of these peptides is for the inhibition of tumor metastasis and invasion which E-cadherin has been proven to regulate. After tumor cells are anchored, the E-cadherin functions as an initiator of tumor growth via cell-cell contact. When cadherin-mediated cell-cell contact is inhibited, the tumor cell is disassociated and undergoes apoptosis or other cell death mechanism. The present invention demonstrates that these peptides can dissociate E-cadherin-mediated cell-cell adhesion in BBMEC, MDCK, and CaCo-2 cell monolayers.

Additionally, these peptides may be used as adjuvants which modulate the porosity of intercellular junctions, thereby assisting in drug delivery across biological barriers. Thus, the present invention permits protein drugs to be transported via the paracellular route or intercellular junctions. Although there are many other ways to modulate the intercellular junctions for improving drug delivery, many of these methods are not very selective. The present invention provides a method which modulates a specific protein (E-cadherin) in the intercellular junction, thereby providing a great deal of selectivity. These peptides can be used to carry drugs through the biological barriers (i.e., intestinal mucosa or the blood-brain barriers) via the paracellular route; this can be done by conjugating the peptides with drugs directly or via cleavable linkers. The peptide will open the intercellular junctions and travel through the paracellular route along with the drug that is conjugated with the peptide. As used herein, the following definitions will apply: "Sequence Identity" as it is known in the art refers to a relationship between two or more polypeptide sequences or two or more polynucleotide sequences, namely a reference sequence and a given sequence to be compared with the reference sequence. Sequence identity is determined by comparing the given sequence to the reference sequence after the sequences have been optimally aligned to produce the highest degree of sequence similarity, as determined by the match between strings of such sequences. Upon such alignment, sequence identity is ascertained on a position-by-position basis, e.g., the sequences are "identical" at a particular position if at that position, the nucleotides or amino acid residues are identical. The total number of such position identities is then divided by the total number of nucleotides or residues in the reference sequence to give % sequence identity. Sequence identity can be readily calculated by known methods, including but not limited to, those described in Computational Molecular Biology, Lesk, A. N., ed., Oxford University Press, New York (1988), Biocomputing: Informatics and Genome Projects, Smith, D.W., ed., Academic Press, New York (1993); Computer Analysis of Sequence Data, Part I, Griffin, A.M., and Griffin, H. G., eds., Humana

Press, New Jersey (1994); Sequence Analysis in Molecular Biology, von Heinge, G., Academic Press (1987); Sequence Analysis Primer, Gribskov, M. et al., eds., M. Stockton Press, New York (1991); and Carillo, H., et al. Applied Math., 48:1073 ( 1988), the teachings of which are incorporated herein by reference. Preferred methods to determine the sequence identity are designed to give the largest match between the sequences tested. Methods to determine sequence identity are codified in publicly available computer programs which determine sequence identity between given sequences. Examples of such programs include, but are not limited to, the GCG program package (Devereux, J., et al., Nucleic Acids Research, 12(1):387 (1984)), BLASTP, BLASTN and FASTA (Altschul, S. F. et al., J. Molec. Biol., 215:403-410

(1990). The BLASTX program is publicly available from NCBI and other sources (BLAST Manual, Altschul, S. et al, NCVI NLM NIH Bethesda, MD 20894, Altschul, S. F. etal., J. Molec. Biol., 215:403-410 (1990), the teachings of which are incorporated herein by reference). These programs optimally align sequences using default gap weights in order to produce the highest level of sequence identity between the given and reference sequences. As an illustration, by a polynucleotide having a nucleotide sequence having at least, for example, 95% "sequence identity" to a reference nucleotide sequence, it is intended that the nucleotide sequence of the given polynucleotide is identical to the reference sequence except that the given polynucleotide sequence may include up to 5 point mutations per each 100 nucleotides of the reference nucleotide sequence. In other words, in a polynucleotide having a nucleotide sequence having at least 95%) identity relative to the reference nucleotide sequence, up to 5% of the nucleotides in the reference sequence may be deleted or substituted with another nucleotide, or a number of nucleotides up to 5% ofthe total nucleotides in the reference sequence may be inserted into the reference sequence.

These mutations ofthe reference sequence may occur at the 5' or 3' terminal positions of the reference nucleotide sequence or anywhere between those terminal positions, interspersed either individually among nucleotides in the reference sequence or in one or more contiguous groups within the reference sequence. Analogously, by a polypeptide having a given amino acid sequence having at least, for example, 95% sequence identity to a reference amino acid sequence, it is intended that the given amino acid sequence ofthe polypeptide is identical to the reference sequence except that the given polypeptide sequence may include up to 5 amino acid alterations per each 100 amino acids ofthe reference amino acid sequence. In other words, to obtain a given polypeptide sequence having at least 95% sequence identity with a reference amino acid sequence, up to 5% ofthe amino acid residues in the reference sequence may be deleted or substituted with another amino acid, or a number of amino acids up to 5% ofthe total number of amino acid residues in the reference sequence may be inserted into the reference sequence. These alterations ofthe reference sequence may occur at the amino or the carboxy terminal positions of the reference amino acid sequence or anywhere between those terminal positions, interspersed either individually among residues in the reference sequence or in the one or more contiguous groups within the reference sequence. Preferably, residue positions which are not identical differ by conservative amino acid substitutions. However, conservative substitutions are not included as a match when determining sequence identity.

Similarly, "sequence homology", as used herein, also refers to a method of determining the relatedness of two sequences. To determine sequence homology, two or more sequences are optimally aligned as described above, and gaps are introduced if necessary. However, in contrast to "sequence identity", conservative amino acid substitutions are counted as a match when determining sequence homology. In other words, to obtain a polypeptide or polynucleotide having 95% sequence homology with a reference sequence, 95% ofthe amino acid residues or nucleotides in the reference sequence must match or comprise a conservative substitution with another amino acid or nucleotide, or a number of amino acids or nucleotides up to 5% ofthe total amino acid residues or nucleotides, not including conservative substitutions, in the reference sequence may be inserted into the reference sequence.

A "conservative substitution" refers to the substitution of an amino acid residue or nucleotide with another amino acid residue or nucleotide having similar characteristics or properties including size, charge, hydrophobicity, etc., such that the overall functionality does not change significantly.

Isolated" means altered "by the hand of man" from its natural state., i.e., if it occurs in nature, it has been changed or removed from its original environment, or both. For example, a polynucleotide or polypeptide naturally present in a living organism is not "isolated, " but the same polynucleotide or polypeptide separated from the coexisting materials of its natural state is "isolated", as the term is employed herein. Finally, all references and teachings cited herein which have not been expressly incorporated by reference are hereby incorporated by reference.

Sequences including or having a sequence which has at least about 50% sequence identity with any one of SEQ ID Nos. 40-46 and 50-58 and which exhibit similar modulation of cell-cell adhesion or inhibition of tumor stasis properties are within the scope ofthe present invention. Preferably, such sequences will have at least about 60%) sequence identity with any one of SEQ ID Nos. 40-46 and 50-58 and still more preferably at least about 75% sequence identity. Alternatively, sequences including or having a sequence which has at least about 50% sequence homology with any one of SEQ ID Nos.40-46 and 50-58 and which exhibits similar modulation of cell- cell adhesion or inhibition of tumor stasis properties are embraced in the present invention. More preferably, such sequences will have at least about 60% sequence homology with any one of SEQ ID Nos.40-46 and 50-58 and still more preferably at least about 75 % sequence homology. Additionally, sequences which differ from any one of SEQ ID Nos. 40-46 and 50-58 due to a mutation event or series of mutation events but which still exhibit similar properties are also embraced in the present invention. Such mutation events include but are not limited to point mutations, deletions, insertions and rearrangements. Furthermore, as it is well known in the art, peptidomimetics may be developed which have the same modulation properties as the preferred peptides detailed herein. As these peptidomimetics require no more than routine skill in the art to produce, such peptidomimetics are embraced within the present application. Notably, the side chains of these peptidomimetics will be very similar in structure to the side chains of the preferred peptides herein, however, their peptide backbone may be very different or even entirely dissimilar. If resistance to degradation in vivo or greater conformational stability were desired, the peptides ofthe present invention could be cyclized by any well known method. One such method adds Penicillamine (Pen) and cysteine (Cys) residues to the N- and C-termini to form cyclic peptides via a disulfide bond between the Penl and Cysl2 residues. The formation of this cyclic peptide restricts the peptide conformation to produce a conformational stability, thereby providing better selectivity for cell surface receptors than its linear counterpart.

BRIEF DESCRIPTION OF THE DRAWINGS Figure 1 is a graph showing binding properties of N- and E-cadherin peptides to anti-E-cadherin antibody usmg normal ELISA for SEQ ID Nos. 1-3 and immobilized peptides (SEQ ID Nos. 4-8);

Fig.2a is a graph showing binding properties of immobilized peptides from EC 1 domains of E-cadherin to anti-E-cadherin antibody; Fig.2b is a graph showing binding properties of immobilized peptides from EC2 domains of E-cadherin to anti-E-cadherin antibody;

Fig 2c is a graph showing binding properties of immobilized peptides from EC3 domains of E-cadherin to anti-E-cadherin antibody;

Fig. 3 is a graph illustrating the change of TEER in cell monolayers after administration of HAV- 10;

Fig. 4 is a graph illustrating the change of TEER in cell monolayers after administration of ADT-10;

Fig. 5 is a graph illustrating the change of TEER in cell monolayers after administration of HAV-6; Fig. 6 is a graph illustrating the change of TEER in cell monolayers after administration of ADT-6 and ADK-6;

Fig. 7 is a graph illustrating the change of TEER in cell monolayers after administration of Amcap- 1 ;

Fig. 8 is a graph illustrating the change of TEER in cell monolayers after administration of Amcap-2;

Fig. 9 is a photograph illustrating fluorescence emission of FITC-HAV-10; Fig. 10 is a photograph illustrating fluorescence emission of FITC-ADT-10; Fig. 11 is a photograph illustrating fluorescence emission of FITC-ADT-6; Fig. 12 is a graph of showing the ability of E-cadherin peptides to inhibit Caco-2 single cell binding to Caco-2 monolayers as a model of tumor cell metastasis and invasion.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENT The following examples set forth preferred embodiments of the present invention. It is to be understood, however, that these examples are provided by way of illustration and nothing therein should be taken as a limitation upon the overall scope ofthe invention.

EXAMPLE 1 This example generated the peptide sequences used for other experiments. Materials and Methods:

Peptide Synthesis and Purification: All peptides used in this work (Table 1) were acetylated and amidated at the N- and C-terminus during the solid phase synthetic procedures. They were synthesized using a Rainin PS-3 peptide synthesizer by a solid- phase peptide synthesis (SPPS) method. Fmoc-amino acids (Bachem Biosciences, Inc.,

King of Prussia, PA) andFmoc-4-methoxy-4' -[γ-carboxypropyloxy]-benzhydryl-amine- resin (DOD-resin) (Bachem Biosciences, Inc., King of Prussia, PA) were used for the synthesis. [2-( l -H-benzotriazol- l -yl )- l , l ,3 ,3 -tetramethyl uronium hexafluorophosphate] (HBTU) (Peptides International, Louisville, KY) was added as the amino acid activator during the synthesis . The peptides were cleaved from the resin by trifluoroacetic acid at 30 °C for 2 hours in the presence of 7.5% v/v (volume/volume) phenol as scavenger. The peptide was precipitated into cold ether and filtered out. In some cases, the peptide was extracted into 5-20% acetic acid solution from the ether solution followed by lyophilization. The crude peptides were purified by semi- preparative reversed-phase HPLC using a C18 column (Rainin, 21.4 x 250 mm, 12 i,

300 A) with a gradient of solvent A (5 % acetonitrile in water containing 0.1% TF A) and solvent B (100% acetonitrile). The purity of peptide fractions was determined by an analytical reversed-phase HPLC using C18 column (4.6 x 250 mm, 5 μ, 300 A) with same solvent system as in the preparative HPLC. The pure peptide fractions were combined, lyophilized, and confirmed by proton nuclear magnetic resonance (Η-NMR) and fast atom bombardment mass spectrometry (FAB-MS).

Results:

A listing of peptides generated by the recited methods is given below in Table 3.

Table 3

Ecad, human E-cadherin ^mEcad, murine E-cadherin amcap, a-aminocaproic acid FITC, fluorescein isothiocyanate

The peptide portion ofthe sequences listed above are included herein as SEQ ID Nos. 40-58 respectively.

EXAMPLE 2 This example conjugated a peptide from the groove region, SEQ ID No.42, with a peptide from the bulge region, SEQ ID No. 44.

Materials and Methods:

ADK-6 and HAN-6 were conjugated via an aminocaproic acid o give two different conjugates, Amcap- 1 and Amcap-2 peptides. The aminocaproic acid was used as a spacer between the two peptides. In Amcap- 1 peptide, the ADK-6 and HAN-6 were conjugated to the Ν- and C-terminus ofthe aminocaproic acid, respectively. To obtain Amcap-2 peptide, the ADK-6 and HAN-6 were conjugated to the C- and Ν- terminus ofthe aminocaproic acid, respectively. Thus, the effect of ADK-6 and HAN-6 on the sequence ofthe conjugate could be evaluated. Peptides having the sequences of Amcap- 1 and Amcap-2 were synthesized as described in Example 1 with the non- natural amino acid, aminocaproic acid, providing the link between the two previously described peptides.

Results:

The sequences of Amcap- 1 and Amcap-2 are given above in Table 3 and are included as SEQ ID Νos. 45 and 46, respectively.

EXAMPLE 3 This example produced the cell monolayers used for later experiments.

Materials and Methods:

The MDCK cell line was obtained from American Type Culture Collection (ATCC, Rockville, MD) at serial passage 24. The cell line was subcultured several times and the cell cultures used in this work were from passages 48-67. Cells were grown in 150 cm² tissue culture treated polystyrene flasks (Fisher Scientific, Pittsburgh, PA) in a controlled atmosphere of 5% CO₂ and 95% relative humidity at 37 °C. Culture medium consisted of 0.87 g/L Earle's balanced salt solution (Sigma, St. Louis, MO) supplemented with 0.85 g/L sodium bicarbonate, 0.1 g/L penicillin G, 0.1 g/L streptomycin sulfate, 5 g/L lactalbumin enzymatic hydrolysate (Sigma, St. Louis, MO), and 5% newborn calf serum (Atlanta Biologicals, Νorcross, GA). The culture medium was replaced every other day for the first 6 days, and everyday thereafter. Nearly confluent cell monolayers (80-90% confluent) were trypsinized with 0.25% trypsin in 1 mM EDTA solution to give a cell suspension and one-fourth ofthe cell suspension was subcultured every 6-7 days.

Results:

Almost confluent cell monolayers were produced by the above-recited methods .

EXAMPLE 4 This example determined the effect of the synthesized peptide on TEER measurements of the cell monolayers.

Materials and Methods:

For TEER measurements and paracellular transport studies, almost confluent cell monolayers were trypsinized and seeded on polystyrene filter inserts (0.4 μm pore, 10 mm diameter) inside 12-well Transwell™ plates (Costar, Cambridge, MA) at a density of 50,000 cells/cm². These cells were grown until confluency was reached on day eight. For fluorescence-localization of peptides, the cells were seeded on 48-well plates and used in between days 6 and 8.

TEER values of cell monolayers were measured using an Evom epithelial voltohmmeter equipped with STX-2 chopstick electrode (World Precision Instruments,

Inc., Sarasota, FL). Prior to peptide treatment, confluent MDCK cell monolayers were grown on polystyrene Transwell™ filters. The monolayers were then washed with pH 7.4 Hanks' balanced salt solution (HBSS) (Cellgro Mediatech, Herndon, NA) containing 10 mM Hepes, 1% glucose and 2 mM CaCl₂, followed by incubation in the same solution to let the cells equilibrate with the new medium. During incubation, the

TEER values of the cell monolayers were measured at several time points until they reached a steady state. A stable TEER value was observed at 1-1.5 hours and the measured value was used as an initial TEER. Only cell monolayers with an initial TEER value between 200 and 250 ohm-cm² were included in this experiment. To determine the effect of peptides on TEER values, 1 mM peptide solutions in HBSS pH 7.4, were applied to the apical (AP) or basolateral (BL) sides, or simultaneously from both sides of cell monolayers. The TEER values were measured from 0 to 5 hours at 1 hour intervals after peptide addition. For each series of measurements, background resistance was measured using several unseeded wells each of which had a value of 80 ohm-cm². This value was used as a correction factor for each measurement. Furthermore, each experiment was performed in triplicate to obtain the average and standard deviation.

Results: The effect of groove region peptides, HAN- 10 and HAN-6, on the modulation of tight intercellular junctions was monitored by the decrease in TEER values of MDCK cell monolayers. These results are given in Figs. 3 and 5, respectively. Several different peptides were used as negative controls including L-reverse peptide (Ac- SSNAHS-NH₂) (SEQ ID No.47), D-reverse peptide (Ac-SSNAHS-NH₂) (SEQ ID No. 48), and an unrelated tripeptide (Ac-NNA-NH₂) (SEQ ID No. 49). Figure 3 shows the effect of HAN- 10 peptide on the TEER of MDCK monolayers when added from the AP or BL side alone or from both sides simultaneously. The HAN- 10 peptide was effective when the cells were treated from both sides (AP-BL). The peptide was also effective in reducing the TEER values when the monolayers were treated with HAN- 10 from the BL side. However, HAN- 10 was slightly less effective when treated from the BL side than when simultaneous treatment from both sides was employed. In contrast, the peptide was not at all effective when administered from the AP side. This ineffectiveness is due to the large size ofthe HAN- 10 peptide, which cannot cross the tight junctions. The negative control, L-reverse peptide, was not effective in modulating the tight junctions, and this result was similar to the blank.

HAN- 10 was reduced to a hexapeptide (HAN-6) to test whether reducing the size ofthe peptide would affect its activity and ability to modulate the junctions from the AP side. As shown in Fig. 5, HAN-6 reduced the TEER values when the MDCK cell monolayers were treated from either the AP side or the BL side alone or from both sides simultaneously. Reducing the size of the peptide from a decapeptide to a hexapeptide slightly reduced its ability to modulate the intercellular junctions when the cell monolayers were treated from the BL side alone or from the AP and BL sides simultaneously. However, HAN-6 was effective when applied to the AP side alone while HAN- 10 was not. For negative controls, MDCK monolayers were also treated with L-reverse peptide, D-reverse peptide, and an unrelated tripeptide from all sides

(AP alone, BL alone, and both sides simultaneously). Results showed that these three peptides (SEQ ID Νos. 47, 48, and 49, respectively) were not effective in modulating TEER values.

Modulation of intercellular junctions using peptides from the bulge region of EC 1 (^hADT- 10, ^hADT-6, ^mADK-6) was evaluated on MDCK cell monolayers. ADT- 10 and ADT-6 were deca- and hexa-peptides derived from human E-cadherin sequence and ADK-6 was derived from the mouse E-cadherin sequence. Figure 4 shows the activity of ADT- 10 peptide when administered from the AP and BL sides alone and simultaneously from both sides. As shown in Fig. 4, ADT- 10 was very effective in lowering TEER values when administered from both sides. The activity decreased slightly when it was added from the BL side alone. Even lower activity was observed when ADT- 10 was added from the AP side alone due to the inability ofthe peptide to permeate the tight junctions.

ADT- 10 was reduced to a hexapeptide ADT-6 and a mutant peptide ADK-6 derived from mouse E-cadherin; the activities of these peptides were evaluated from the AP and BL sides, and from both sides simultaneously. These results are give in Fig. 6. ADT-6 was effective in lowering the TEER values of MDCK monolayers. Treatment of the monolayers from BL side had the same effect as treatment from both sides. Administration from the AP side alone still showed activity, but this activity was lower than that with treatment from the BL or both sides simultaneously. ADT-6 was slightly less active than ADT- 10, but ADT-6 was more effective than ADT- 10 when administered from the AP side alone. ADK-6 is similar to ADT-6, however, the Thr3 in ADT-6 was mutated to Lys3 in ADK-6. The activity of ADK-6 was then tested from both sides simultaneously. ADK-6 had a lower activity than ADT-6, thereby suggesting that the selectivity ofthe peptide is sensitive to the peptide sequence.

Figure 7 shows the effect of the Amcap- 1 conjugate on TEER modulation. Treatment from the BL side alone and from both sides simultaneously reduced the TEER of the MDCK cell monolayers. Due to the large size of Amcap-1, it is not effective in modulating the TEER from the AP side. However, as shown in Fig. 8, Amcap-2 was only slightly effective when applied to both AP and BL sides and it was not effective when administered from either the AP or BL side alone. This suggests that the arrangement ofthe conjugate is important on the activity ofthe conjugate.

EXAMPLE 5 This example tested the paracellular flux of [¹⁴C]-mannitol across the cell monolayers created by the methods of Example 3.

Materials and Methods:

This example determined the ability ofthe peptides which were most effective in modulating TEER (HAN- 10, HAN-6, ADT- 10, ADT-6) at perturbing the cadherin- cadherin interactions in the intercellular junctions. The peptide activities in increasing intercellular junction porosity were examined by measuring the enhancement of ¹⁴C- mannitol paracellular transport across MDCK cell monolayers at 37 °C in the AP-to-BL direction. The cell monolayers were washed with pH 7.4 HBSS for 30 minutes, and then treated with 1 mM peptide solution from the AP or BL side or from both sides simultaneously. After 1 hour of peptide incubation, 10 iL of [¹⁴C]-mannitol (ΝDΝ™, Life Science Products, Inc., Boston, MA) was added to the AP side of each well. The accumulation of [¹⁴C]-mannitol in the BL side was measured by counting the radioactivity of samples taken from the BL side using a Beckman LS-5801 liquid scintillation counter (Beckman Instruments, Fullerton, CA). Aliquots of 30 iL were taken from each well at 0, 1, 2, 3 and 4 hour time points after [¹⁴C]-mannitol addition. After sampling, the wells were replenished with the same volume of HBSS to maintain a constant volume. Before scintillation counting, the sampled aliquots were diluted with 10 mL ScintiNerse (Fisher Chemical Co., Pittsburgh, PA). The flux (dQ/dt) of [¹⁴C]-mannitol through MDCK cell monolayers was determined by plotting the concentration of accumulated [¹⁴C] -mannitol in the BL chamber (Q) versus time (t). The apparent permeability coefficient (P_app) was calculated by the following question: P_app = Flux/D₀-A, where D₀ represents the initial [¹⁴C]- mannitol concentration in the AP chamber and A represents the cross-sectional area of the cell monolayers (1 cm²). Triplicate experiments were performed to obtain the average and standard deviation.

Results: Table 4 contains the results of this example. As shown in Table 4, BL and AP-

BL treatment ofthe monolayers with HAN- 10 peptide increased the transport of ¹⁴C- mannitol across the monolayers. Treatment ofthe MDCK monolayers with HAN- 10 peptide from the AP side alone was less effective than combined AP-BL and BL side alone treatments. Treatment of the monolayers from combined AP-BL and BL side alone using HAN-6 produced a result similar to that of HAN- 10. However, AP treatment ofthe monolayers with HAN-6 produced a higher mannitol transport than did treatment with HAN- 10. These results were consistent with the results of TEER modulations and suggest that reducing the size ofthe peptides reduces the selectivity ofthe HAN-6 peptide. Peptides (ADT- 10 and ADT-6) from the bulge region of the EC 1 domain were also examined for their ability to increase the porosity of the intercellular junctions. ADT- 10 improved paracellular mannitol transport 1.6 x, 6.1 x, and 6.6 x relative to the control by treating the monolayers from the AP side alone, the BL side alone, and the AP-BL sides concurrently. Results were consistent with those from the TEER measurements. Relative to the control, the smaller ADT-6 peptide also improved the

¹⁴C-mannitol transport when administered from the AP side alone (3.2 x) or BL side alone (4.8 x) or both sides simultaneously (5.1 x). Similar to the results from the TEER measurements, ADT-6 was more effective than ADT- 10 when administered from the AP side alone. This increased effectiveness is due to the smaller size of these peptides. Overall, the mannitol flux results support the findings from the TEER measurements.

Table 4

ADT-6 1.12 ± 0.26 [3.2 x] 1.70 ± 0.14 [4.9 x] 1.80 ± 0.17 [5.1 x] ^a unrelated peptide used as negative control ^b No peptide added to the system, only HBSS * Relative activity of peptide when applied from both sides

EXAMPLE 6 This example tested localization of peptide binding in intercellular junctions.

Materials and Methods: Peptide localization and binding in the intercellular junctions was achieved using fluorescein isothiocyanate (FITC)-labeled peptide. FITC was conjugated to the N-terminal of LFSHANSSNG-NH₂ (SEQ ID No. 50), QGADTPPNGV-NH₂ (SEQ ID No. 51), and ADTPPN-NH₂ (SEQ ID No. 52) peptides by treating the peptides with FITC at apH 9. The crude FITC-peptides were purified and analyzed by reversed-phase HPLC using the method described in Example 1. The identities of FITC-

LFSHANSSNG-NH₂, FITC-QGADTPPNGV-NH₂, and FITC- ADTPPV-NH₂ were all confirmed by mass spectrometry.

Prior to localization studies, the confluent cell monolayers grown on 48-well plates were washed 3 times for 10 minutes with HBSS pH 7.4 containing 2 mM Ca²⁺. Following this washing, the cells were incubated with 3% bovine serum albumin

(Sigma, St. Louis, MO) in HBSS (Cellgro Mediatech, Herndon, NA) for one hour to block the non-specific binding. The cell monolayers were washed again with HBSS and incubated with 0.1 mM solution of FITC-labeled peptide for one hour at 37 °C. Finally, the cells were thoroughly washed with HBSS and then observed under a fluorescence microscope.

Results:

Intercellular Junctions Localization ofthe FITC-labeled Peptides: The ability ofthe FITC-labeled peptides (FITC-HAN-10,FITC-ADT-10, andFITC- ADT-6) to bind E-cadherins in the intercellular junctions was investigated by incubating the MDCK monolayers with the labeled peptide and observing the fluorescence emission from the FITC group. Photographs of these results are given in Figs. 9-11, respectively. Fluorescence microscopy studies showed that the intercellular junctions of MDCKs monolayers were decorated by these peptides. The FITC-labeled peptides showed up as punctate fluorescence spots at the cell borders. These results were distinctly different than the results from the control where MDCK monolayers were treated with FITC alone wherein the FITC reacted with the cell surface proteins and decorated the entire cells (data not shown).

EXAMPLE 7 This example evaluates the ability of E-cadherin peptides from EC 1 , EC2, EC3 and EC4 domains to inhibit Caco-2 single cell adhesion to Caco-2 monolayers. This assay was used to evaluate the inhibition of tumor invasion and metastasis by E- cadherin peptides.

Materials and Methods:

Epithelial adenocarcinoma clone, Caco-2 cells were purchased from the ATCC (Manassas, NA). The cells were maintained in 10% fetal bovine serum (FBS) contained DMEM (Sigma, MO). Anti-E-cadherin Monoclonal antibody (U3254) and anti-rat IgG FITC conjugate antibody (F1763) were purchased from SIGMA(St. Louis, MO).

Another anti-E-cadherin Monoclonal antibody (SHE78-7) was purchased from PanNera Co. (Madison, WI). The fluorescence markers, 2',7'-bis(2-carboxyethyl)-5(and-6)- carboxyfluorescein acetoxymethyl ester (BCECF-AM) and Calcein-AM were purchased from Molecular Probes (Eugene, OR). These amino acid sequences were decided according to their binding energies and

CellTiter 96 ™ AQueous was purchased from Promega (Madison, WI). Dimethyl sulfoxide (DMSO) and Triton X-100 were purchased from SIGMA. Several peptides (Ac-HSASNA-ΝH₂ (Provided herein as SEQ ID No, 59); Ac-LFSHANSSNG-NH₂ (SEQ ID No. 40); Ac-YTALIIATDN-NH₂ (SEQ ID No. 58); Ac- DRERIATYTLFSHANSSNGNAVED-NH₂ (SEQ ID No. 2)) were used for this study.

SEQ ID No. 59 served as a negative control and anti-E-cadherin antibody served as a positive control.

Caco-2 cells were maintained in 10% FBS contained DMEM (FBS/DMEM) on T-75 plastic flasks at 37 °C in a humidified 5% CO₂ atmosphere. Culture medium was changed every other day. Cells were subcultured once a week so that the cells might not reach to the point of confluent layers. Basically Caco-2 cells, which were not passaged more than 60 times, were used.

Caco-2 cells were seeded on 48-well culture dishes. When the cells reached a confluent layer, media was replaced by B SA/DMEM, and the layers were used as Caco- 2 cell monolayers. On the other hand, the other Caco-2 cell layer, which was cultured on T-75 flask, was treated with Ca²⁺- and Mg²⁺-free Hank's balanced saline solution (HBSS-) for 2 hours to obtain Caco-2 single cells. The isolated Caco-2 single cells were incubated with 5 ug/ml BCECF-AM containing FBS/DMEM for 2 hours at 37 °C in a humidified 5% CO₂ atmosphere in order to label the cells with the fluorescent marker. After incubation, cells were washed extensively with PBS- (80mM Na₂HPO₄, 20mM

KH₂PO₄, 140mM NaCI, lOmM KC1, pH 7.4) to remove free marker. The labeled cells were re-suspended in BSA/DMEM at a density of 4 10⁵ cells/ml. BCECF-labeled Caco-2 cells were mixed with peptides added by 1:1 dilution, and then the mixtures were added on Caco-2 cell monolayers. The cells were incubated for 2 hours in the dark at 37 °C under a humidified 5% CO₂ atmosphere. Caco-2 cell layers were washed three times by PBS+ (0.63mM CaCl₂, 0J4mMgCl₂, and 1.0 mg/ml glucose in PBS-). Next, cells were lysed with 350 ul/well of 3.0% Triton X-100 solution. Lysates were collected in micro-centrifuge tubes, and centrifuged at 10,000 rpm for 10 minutes at room temperature. Supernatants were transferred (150 ul/well) in duplicate to black 96-well plates for measuring the fluorescence intensities. Excitation was measured at 485 nm and Emission at 530 nm, using a Microplate Fluorescence Reader (FL600, Bio-Tek).

Results:

Figure 12 shows the inhibition of Caco-2 single cell adhesion to Caco-2 monolayers by peptides. In this study SEQ ID No. 59 (Ac-HSASNA-NH₂) was used as a negative control. This peptide has a randomly scrambled sequence ofthe HAV- peptide (Ac-SHANSS-ΝH₂) (SEQ ID No. 42.). SEQ ID No. 59 and SEQ ID No. 40 (Ac-LFSHANSSNG-NH₂) only slightly inhibit the adhesion of single cells to the cell monolayers. However, SEQ ID No. 58 (Ac-YTALIIATDN-NH₂) and SEQ ID No. 2 (Ac-DRERIATYTLFSHANSSNGNAVED-NH₂) each significantly inhibit the single cell adhesion to the Caco-2 monolayers. Similarly, anti-E-cadherin antibody can inhibit the adhesion of single cells to Caco-2 monolayers indicating that this adhesion is mediated by E-cadherin. In conclusion, E-cadherin peptides can be used to inhibit E- cadherin-mediated single cell adhesion to Caco-2 monolayers. This indicates that E- cadherin peptides can inhibit tumor cell invasion and metastasis.

DISCUSSION The X-ray structure ofthe EC 1 -EC2 domains of E-cadherin was used to find the counter sequence of HAN peptide. Peptides derived from the bulge (ADT- 10 and ADT-6) and groove (HAN- 10 and HAN-6) regions were synthesized and evaluated for their ability to modulate the E-cadherin-mediated intercellular junctions. The bulge (ADT- 10) and groove (HAV- 10) decapeptides lowered the TEER values ofthe MDCK monolayers when administered from the BL side alone and when administered simultaneously from the AP and BL sides compared to control peptides (L-reverse, D- reverse, and unrelated tripeptide) which did not lower the TEER values. However, these decapeptides were not very effective in modulating the intercellular junctions when administered from the AP side alone. This ineffectiveness is due to the size of the decapeptides which were too large to permeate via the tight junctions (zonula occluden) and reach the adherens junctions (zonula adherens) where E-cadherin is located, contrast, these decapeptides were effective when administered only from the

BL side due to the absence of tight junctions which keep the peptides from reaching the E-cadherins at the adherens junctions. The highest activity was found when the monolayers were treated simultaneously from both sides. These results suggest that decapeptides introduced from the AP side permeate more effectively to help the modulation of the intercellular junctions after some junction opening by the peptides from the BL side. The loosening ofthe intercellular junctions shown here was not due to cell death or damage as shown by cell viability which was higher than 95% after incubation with peptide solution for 6 hours. Thus, these results confirm the ability of the peptides to increase the porosity ofthe intercellular junctions. To confirm the increase in porosity of the intercellular junctions by these decapeptides, the enhancement of the paracellular transport of ¹ C-mannitol was examined via the decapeptide-treated MDCK cell monolayers. Interestingly, the decapeptides improved the mannitol flux when the peptides were added from the AP side alone, BL side alone and when administered from both sides simultaneously. As with the TEER modulation experiments, these decapeptides were very effective when used from BL side alone or both sides simultaneously and caused an increase in the mannitol flux of around 3.7 to 6.6 times compared to the control peptide. Although the decapeptides (HAN- 10 and ADT- 10) were not effective in lowering the TEER when administered from the AP side alone, they were able to improve the mannitol flux 2 x in comparison to the control peptide. Therefore, the TEER value modulation results were congruent with the mannitol flux measurement results.

The inability of decapeptides delivered from the AP side alone to modulate the intercellular junctions may not be useful in achieving the goal of improving the paracellular drug delivery of peptide and protein drugs in vivo because the delivery of these therapeutic molecules is primarily in the AP-to-BL direction. Thus, the cadherin peptides have to overcome the tight junctions prior to working on the cadherin-cadherin interactions in the zonula adherens. This is ineffective due to the size of the decapeptides. Accordingly, the decapeptides were reduced to hexapeptides (ADT-6 to HAN-6) to test whether the decrease in peptide size would increase the modulation of intercellular junction porosity when administered solely from the AP side of the monolayers. Interestingly, both hexapeptides were effective in modulating the TEER values when applied from AP side alone, the BL side alone, or from both sides simultaneously. Thus, these hexapeptides were able to penetrate the AP tight junctions (zonula occludens) and effect TEER values when applied from the AP side alone. However, the hexapeptides (ADT-6 and HAN-6) showed a lower activity compared to the corresponding decapeptides when applied from the BL side alone or from both sides simultaneously. This may be due to the contribution of residues which surround the active sequence to the peptide binding. Furthermore, the decapeptides may have better conformational stability than the hexapeptides. To evaluate the effect of sequence on peptide activity, the ADK-6 peptide derived from mouse E-cadherin was synthesized.

This peptide differs from the ADT-6 peptide by having a Lys3 instead of a Thr3. The ADK-6 peptide showed lower activity than the ADT-6 peptide, thereby suggesting that the activity of these peptides is sequence specific. In other words, the intercellular junctions of E-cadherins recognize Thr3 better than Lys3 in the hexapeptides. This also suggests that the E-cadherins in the MDCK monolayers may have a higher homology to human than to the mouse E-cadherins, particularly at the bulge region.

The increase in paracellular porosity produced by the hexapeptides (ADT-6 and HAN-6) was also examined using ¹⁴C-mannitol administered from the AP side alone, the BL side alone, and from both sides simultaneously. From the AP side, the hexapeptides improved the mannitol flux about three-fold over the control peptide. This is also about 1.8 times higher than the effect exhibited by the tested decapeptides. When administered from the BL side alone, the hexapeptides caused a greater increase in the mannitol flux than did the decapeptides. This result may be due to the ability of the hexapeptides to permeate the desmosomes and adherens junctions more effectively than the decapeptides. However, when the hexapeptides were administered simultaneously from both the AP and BL sides, the decapeptides were more effective than the hexapeptides in improving the mannitol flux. These results suggest that the higher selectivity of the decapeptides overcomes the rate of permeation of both hexapeptides and decapeptides. In order to get more information on peptide-cadherin interactions, the activities of L-reversed (Ac-SSNAHS-NH₂) and D-reversed (Ac-SSVAHS-NH₂) hexapeptides were examined. Besides serving as negative controls, these peptides gave additional information on the nature of binding between the HAN peptides (i.e., HAN-6 = Ac- SHANSS-ΝH₂). If the D-reverse peptide had bound to the receptor, than it is probable that the peptide binds to the receptors via the side chain only without the involvement ofthe backbone interaction. However, the D-reversed peptide did not lower the TEER value of MDCK monolayers. Thus, the D-reversed peptide did not bind to E-cadherin and modulate cadherin-cadherin interactions. This indicates that binding between the HAN-6 peptide and E-cadherins may involve both side chain and backbone interactions (i.e., hydrogen bond formations). The L-reversed peptide did not have any activity, thereby suggesting the importance of sequence selectivity ofthe E-cadherin.

To evaluate the potential synergistic effect of both the groove and bulge regions, the conjugated peptides, Amcap- 1 and Amcap-2, each of which contain both of the important sequences from the bulge and groove regions, were evaluated. The conjugated peptides were synthesized by linking SHANS S and ADKPPN sequences via an ε-aminocaproic acid. The choice of an ε-aminocaproic acid linker was based on the approximately 6-10 A distance between the SHANS S and ADKPPN sequences in the structure of the ECl domain. The distance between the Ν- and C-tennini of ε- aminocaproic acid is about 7 A. Amcap- 1 lowered the TEER values of MDCK cell monolayers treated from the BL and AP-BL sides but not from the AP side.

Amcap-2 was not effective in modulating the intercellular junctions regardless ofthe side of administration (Table 4). This suggests that the conjugation position of ADKPPN and SHANS S in the linker affects the binding ofthe conjugate to E-cadherin. These data indicate that the peptides bound to E-cadherin molecules in an antiparallel manner, as in native E-cadherin-cadherin tra-ras-interaction. When the position of these sequences was reversed in Amcap-2, the peptide activity was completely abolished. Similar to HAN- 10, Amcap- 1 was not effective when administered from the AP side alone due to the size of this molecule. Amcap- 1 was less effective than the decapeptides (ADT- 10 and ADT- 10) and hexapeptides (ADT-6 and ADT-6) when incubated up to 5 hours. However, Amcap- 1 modulated the TEER values in a manner similar to HAN-6 when incubated about 7 hours. This result has several possible explanations. First, both ofthe SHANSS and ADKPPN sequences may not work in a synergistic manner. Second, the Amcap-1 conjugate is too large to be effective in percolating through the desmosomes and zonula adherens. Third, the SHANSS and ADKPPN sequences produced intramolecular peptide-peptide interactions which require energy to dissociate the intramolecular interaction and produce an intermolecular interaction with the E-cadherin. ADTPPN will be used in the future conjugates.

The direct localization ofthe hexapeptides and decapeptides in the intercellular junctions was examined using the FITC-labeled peptides (FITC-HAN- 10, FITC-ADT-

10, and FITC- ADT-6). These peptides bound in the intercellular junction plaque, as shown by the punctate fluorescence intensity around the MDCK cell borders. These results suggest that the peptides bind in the intercellular junctions.

E-cadherin peptides can also inhibit Caco-2 single cell adhesion to Caco-2 cell monolayers (Fig. 12). These results indicate that the peptides from the groove and bulge regions of ECl, EC2, EC3, and EC4 can be used to inhibit tumor cell invasion and metastasis.

In conclusion, the present invention discovered another recognition site for E- cadherin-E-cadherin interactions, located in the bulge region ofthe ECl domain of E- cadherin which has an ADTPPN sequence. This sequence appears to recognize the

SHANSS sequence located in the groove region of the same domain in another E- cadherin molecule. Modulation of TEER values by the peptides is consistent with an increase in the paracellular transport of ¹⁴C-mannitol. Furthermore, the peptides from the groove and bulge regions of ECl, EC2, EC3, and EC4 domains can be used to inhibit tumor cells adhesion mediated by E-cadherin during tumor cell metastasis and invasion.

Claims

We claim:

1. A modulator peptide operable for inhibiting E-cadherin mediated cell to cell adhesion, said modulator peptide being isolated and purified and having at least about 6 amino acid residues and having at least about 50% sequence homology with a first peptide derived from E-cadherin, said modulator peptide characterized by the property of decreasing transepithelial electrical resistance when applied to cell monolayers, in the electrical resistance experiment of Example 4.

2. The modulator peptide of claim 1, said peptide being further characterized by the property of increasing paracellular flux of [¹⁴C]-mannitol across cell monolayers, in the paracellular flux experiment of Example 5.

3. The modulator peptide of claim 1 , said E-cadherin peptide being selected from the group consisting of SEQ ID Nos. 40-46 and 50-58.

4. The modulator peptide of claim 1, said peptide having at least about 60%) sequence homology with a peptide derived from E-cadherin.

5. The modulator peptide of claim 1 , said E-cadherin peptide being derived from the contact regions of E-cadherin to E-cadherin interactions.

6. The modulator peptide of claim 1, said peptide being further characterized by the property of inhibiting single cell adhesion to cell monolayers in the assay of Example 7.

7. The modulator peptide of claim 1, said peptide being selected from the group consisting of SEQ ID Nos. 40-46 and 50-58 and peptidomimetics thereof.

8. The modulator peptide of claim 7, said peptide being selected from the group consisting of SEQ ID Nos. 40-46.

9. The modulator peptide of claim 1 , said peptide being SEQ 3D No. 41.

10. The modulator peptide of claim 1, said peptide further having at least about 50%> sequence homology with a second peptide derived from E-cadherin.

11. The modulator peptide of claim 10, said peptide further including a linker operable for linking said first peptide derived from E-cadherin to said second peptide derived from E-cadherin.

12. The modulator peptide of claim 11, said linker comprising a non- natural amino acid.

13. The modulator peptide of claim 11, said linker comprising aminocaproic acid.

14. The modulator peptide of claim 1, said peptide having a cyclic structure.

15. A purified and isolated peptide sequence having at least 50% sequence homology with a sequence selected from the group consisting of SEQ ID Nos. 40-46 and 50-58 and peptidomimetics thereof.

16. The peptide sequence of claim 15, said sequence having at least about 60%) sequence homology with a sequence selected from the group consisting of SEQ ID Nos. 40-46 and 50-58.

17. The peptide sequence of claim 15, said sequence having at least about 75%o sequence homology with a sequence selected from the group consisting of

SEQ ID Nos. 40-46 and 50-58.

18. The peptide sequence of claim 15, said sequence being selected from the group consisting of SEQ ID Nos 40-46.

19. A modulator peptide operable for inhibiting E-cadherin mediated cell to cell adhesion, said modulator peptide being isolated and purified and having at least about 6 amino acid residues and having at least about 50% sequence homology with a peptide derived from E-cadherin, said modulator peptide characterized by the property of increasing paracellular flux of [^I4C]-mannitol across cell monolayers, in the paracellular flux experiment of Example 5.

20. A modulator peptide operable for inhibiting E-cadherin mediated cell to cell adhesion, said modulator peptide being isolated and purified and having at least about 6 amino acid residues and having at least about 50% sequence homology with a peptide derived from E-cadherin, said modulator peptide characterized by the property of inhibiting single cell adhesion to cell monolayers in the assay of Example 7.

21. A method of altering transepithelial electrical resistance of cell layers comprising the steps of contacting said cell layer with a peptide sequence, said peptide sequence having at least about 50%> sequence homology to a peptide sequence selected from the group consisting of SEQ ID Nos. 40-46 and 50-58.

22. The method of claim 21 , said cell layer comprising MDCK cells.

23. The method of claim 21 , said peptide sequence having at least about 60%) sequence homology to apeptide sequence selected from the group consisting of SEQ ID Nos. 40-46 and 50-58 and peptidomimetics thereof.

24. The method of claim 21 , said peptide sequence having at least about 75% sequence homology to apeptide sequence selected from the group consisting of SEQ ID Nos. 40-46 and 50-58.

25. The method of claim 21 , said contacting step comprising the step of administering said peptide sequence on the basolateral side of said cell layer.

26. Themethodof claim 21, said contacting step comprising the step of administering said peptide sequence on the apical side of said cell layer.

27. Themethodof claim 21, said contacting step comprisingthe step of administering said peptide sequence on both the apical side and the basolateral side of said cell layer.

28. The method of claim 21 , said peptide sequence being in solution.

29. The method of claim 28, said solution including Hanks ' balanced salt solution.

30. A method of altering paracellular transport by peptides comprising the steps of contacting a cell layer with a peptide sequence having at least about 50% sequence homology to a sequence selected from the group consisting of SEQ ID Nos. 40-46 and 50-58 and peptidomimetics thereof.

31. The method of claim 30, said paracellular transport including [¹⁴C] mannitol.

32. The method of claim 30, said contacting step occurring on the basolateral side of said cell layer.

33. The method of claim 30, said contacting step taking place at the apical side of said cell layer.

34. The method of claim 30, said contacting step occurring on the apical and the basolateral sides of said cell layer.

35. A method of altering adhesion between a first cell and a second cell, comprising the step of contacting at least one of said first or second cells with a peptide having at least about 50% sequence homology with a peptide selected from the group consisting of SEQ ID Nos 40-46 and 50-58 and peptidomimetics thereof.

36. The method of claim 35, said contacting step occurring on the basolateral side of said cell.

37. The method of claim 35 , said contacting step taking place at the apical side of said cell.

38. The method of claim 35, said contacting step occurring on the apical and the basolateral sides of said cell.

39. A method of inhibiting single cell adhesion to cell monolayers comprising the step of contacting said single cell with a peptide having at least about 50% sequence homology with a peptide selected from the group consisting of SEQ ID

Nos 40-46 and 50-58 and peptidomimetics thereof.

40. The method of claim 39, said cells including tumor cells.

41. The method of claim 39, said cell monolayers including tumor cells.

42. A method of delivering protein drugs across a biological barrier comprising the steps of: conjugating said protein drug with a peptide having at least about 50% sequence homology with a peptide derived from E-cadherin to form a peptide-drug conjugate; and contacting said biological barrier with said peptide-drug conjugate.

43. The method of claim 42, said biological barrier being selected form the group consisting ofthe intestinal mucosa and the blood-brain barrier.

44. The method of claim 42, said peptide being selected from the group consisting of SEQ ID Nos. 40-46 and 50-58 and peptidomimetics thereof.

45. The method of claim 42, said conjugating step including the step of combining said peptide with said protein drug via a linker.

46. The method of claim 45, said linker being cleavable.