WO1998017242A1 - Unique peptides for targeting fibronectin-enriched surfaces and a method for their delivery in the treatment of metastatic cancer - Google Patents

Abstract

A method for the identification of unique fibronectin-targeting peptides derived from the tryptic digestion of purified gelatin is described. Two unique peptides are isolated and shown to have higher affinities for fibronectin than the gelatin alone. The peptides are stabilized and delivered by covalently bonding to the phosphatidylethanolamine existing naturally in phospholipid-stabilized triglyceride emulsions currently employed clinically as injectable nutritional emulsions. Unexpectedly, we discovered that the peptides could be coupled to emulsion droplets in situ. One of the systems identified as the 12-mer Peptide I linked to the phosphatidylethanolamine through its N-terminus (PIN-E), is shown to retain the fibronectin-affinity of the starting material and to inhibit the spreading activity of baby hamster kidney cells, a well-recognized in vitro model of metastatic tumor spreading activity. This particular system is stable at refrigerator temperature for at least a month but demonstrates some decoupling of the peptide and aggregation of the emulsion droplets in the presence of rabbit serum at 37 °C after two days storage. It is anticipated that these systems will have utility by passively blocking metastatic processes that involve fibronection. In addition, they would have advantages as drug delivery systems specifically targeting fibronectin-enriched surfaces, especially for hydrophobic drugs such as paclitaxel (Taxol®).

Description

UNIQUE PEPTIDES FOR TARGETING FIBRONECTIN-ENRICHED SURFACES AND A METHOD FOR THEIR DELIVERY IN THE TREATMENT OF METASTATIC

CANCER

This application claims priority to U.S. Provisional Application 60/029,509 filed October 24, 1996.

FIELD OF THE INVENTION

The present invention relates to reagents and methodologies for preventing the adhesion of a cell to a matrix component such as fibronectin. The invention may be utilized to treat a disease condition such as cancer by inhibiting the interaction of a cell with fibronectin, thus reducing the ability of the cell to migrate to various portions of the body. The present invention further provides the reagents and methodologies for the delivery of drugs to fibronectin-enriched targets.

BACKGROUND OF THE INVENTION

Researchers have placed great emphasis on the development of anticancer therapies that directly inhibit or destroy cancer cell growth; however, most patients die of metastatic tumors following cancer treatment. Metastasis involves the movement of cancer cells throughout the body and adhesion at distant sites. Those skilled in the art have, therefore, attempted to develop methods of interfering with the adhesion process, either directly or indirectly. This process is complex but often involves a major adhesion protein (integrin) such as fibronectin (Akiyama et al., 1995). Fibronectin is a large glycoprotein comprising two dissimilar disulfide-linked subunits, each with molecular masses of 250-280 kDa. On a cell surface the dimer can itself polymerize to form large disulfide-linked complexes and the molecule is found as a prominent component of plasma, other body fluids and connective tissue matrices. It is secreted by a number of tumor cells and has interactive domains for collagen, cell surfaces, heparin, DNA and fibrin (Fidler, 1978). The primary role of fibronectin in the metastatic process remains to be examined in detail but preliminary studies indicate that it is required for transportation of cells across basement membranes and the process can be inhibited by antifibronectin antibodies (Nicolson, 1982). Accordingly, the use of antiadhesive peptides to interfere with the metastatic colonization process has received a great deal of attention. The pentapeptide, GRGDS, which represents the principle cell-binding domain of fibronectin, has demonstrated dose-dependent inhibition of the metastatic process (Humphries et al., 1986; Akiyama et al, 1995) but in humans this peptide has a low specific activity which requires a very high dose (> 10 g/70 kg adult) without the stability required for a realistic pharmaceutical product. Iwamoto et al., (1987) also described the YIGSR-amide as an alternative antimetastatic peptide for melanoma cells since it reacts with the same cell adhesion site in laminin. These methods have not proven effective for treating a disease such as cancer and, therefore, there is a need in the art for reagents and methodologies useful for such a purpose. Collagen contains a major fibronectin recognition site and is commercially available as gelatin. Applicants reasoned that smaller peptides which reacted with fibronectin could be produced by enzymatically breaking down the gelatin molecule to peptidal segments. In addition, by covalently bonding these peptides to phospholipid-stabilized emulsion droplets, a stable fibronectin-binding system could be developed to passively block the metastatic process as well as selectively deliver antitumor compounds (especially hydrophobic drugs such as Taxol^R) to fibronectin-enriched surfaces. This latter targeting ability has the potential for delivery of much smaller doses of otherwise toxic entities used to treat cancer. The present invention provides several benefits to a patient such as a cancer patient including improved effectiveness of treatment with fewer side effects.

BRIEF DESCRIPTION OF THE FIGURES

Figure 1 is an elution profile of limed bovine skin gelatin (type B, Bloom strength 225 g) on Sephadex G-200.

Figure 2 is a size exclusion HPLC profile of limed bovine skin gelatin (type B, Bloom strength 225 g) purified using Sephadex G-200, 20μL, using a Bio-Sil SEC-125 (300 x 7.8 i.d. mm) column, a mobile phase consisting of 0.05 M NaH,P0₄ + 0.05 M Na₂HP0₄ + 0.15 M NaCl (pH 6.8) at a flow rate of 1 mL/min and a 215 nm UV detector.

Figure 3 is a size exclusion HPLC profile of the digested mixture of gelatin by trypsin using a Bio-Sil SEC-125 column (300 x 7.8 i.d. mm) column; the sample gelatin was digested in trypsin (25:1, w/w) at 37°C for 90 min in digestion buffer (0.1 M NH₄HC0₃, 0.1 mM CaCl₂, pH 8.0), lOμL volume; the mobile phase was 0.05 M NaH₂P0₄ + 0.05 M Na^PO, + 0.15 M NaCl

(pH 6.8) at a flow rate of 1 ml/min using a UV detector at 215 nm.

Figure 4 is a size exclusion HPLC profile of fibronectin-binding fragments fractionated by fibronectin-Sepharose affinity chromatography using a Bio-Sil SEC-125 column (300 x 7.8 i.d. mm) column, a sample size of 10 μL, a mobile phase 0.05 M NaH₂PO₄ + 0.05 M Na^PO,, + 0.15 M NaCl (pH 6.8), a flow rate of 1 mL/min and a UV detector 215 nm.

Figures 5 A, 5B, and 5C are plots of binding to fibronectin with increasing fibronectin concentration where 5A is gelatin, 5B is peptide I, and 5C is peptide II (n=3). Figure 6 is a plot of the inhibition of the binding of fibronectin to plastic-bound gelatin by gelatin in the ELISA.

SUMMARY OF THE INVENTION It is an objective of the present invention to provide reagents and methodologies for inhibiting adhesion of a cell to a surface for use in treating a disease condition such as cancer. In one embodiment, a process for isolating a peptide having the ability to bind fibronectin comprising chromatographically fractionating gelatin dissolved in a buffer, followed by digestion with a protease, preferably trypsin, and applying a sample of the digested material to a fibronectin affinity matrix. By elution of the bound material, a peptide having affinity for fibronectin is isolated. In another embodiment, a method of preventing adhesion of a cell to fibronectin comprising coating a fibronectin-enriched surface with a composition comprising a peptide having affinity for fibronectin is provided.

It is another objective of the present invention to provide a method for delivering a peptide to a patient. In one embodiment, a process for stabilizing a peptide for administration to a patient comprising generating a complex comprising said peptide covalently linked through its N or C termini to a phosphatidylethanolamine in a sterile phospholipid-stabilized emulsion is provided.

It is yet another objective of the present invention to provide a method for delivering a bioactive agent to a region of the body comprising a fibronectin-enriched surface. In one embodiment, a method of delivering a bioactive agent to a fibronectin-enriched surface comprising administration of a composition comprising said bioactive agent and a peptide having affinity for fibronectin covalently linked through its N or C termini to a phosphatidylethanolamine in a sterile phospholipid-stabilized emulsion is provided. In a preferred embodiment, the surface is a tumor cell membrane.

DETAILED DESCRIPTION

A clear distinction can be made between altered adhesive behavior that is either a direct or indirect consequence of the neoplastic process and the use of normal adhesive mechanisms by abnormal cells. Tumor cells may need to adhere in order to proliferate or migrate, but this does not necessarily require alterations to their normal adhesive machinery (Humphries et al., 1989). Thus, antiadhesive agents developed through the study of normal cell adhesion may be useful for preventing certain adhesive interactions of tumor cells. The rational development of antiadhesive agents requires identification of the molecules that mediate adhesion and the elucidation of their mechanisms of interaction.

Fibronectin is one of the more important adhesive molecules which plays a key role in certain adhesive aspects of target organ colonization by metastatic cells. It is advantageous to search for the ligands that can bind to fibronectin with high affinity. Such ligands may be utilized to interfere with adhesion of a cell, such as a tumor cell, that could result in inhibition of tumor cell metastasis. In recent years, this area has been the subject of intense study and the progress made has been the subject of several comprehensive reviews (Akiyama et al., 1995; 1990; Humphries, 1990; Yamada, 1991 ; Hynes, 1992).

A number of approaches involving the use of anti-adhesive peptides have been tested by those skilled in the art in an attempt to interfere with metastatic colonization. GRGDS, the pentapeptide representing the principal recognition site in the central cell binding domains of fibronectin and laminin, was used as a prototype antiadhesive synthetic peptide. Premixing melanoma cells with the pentapeptide elicited a dose-dependent inhibition of colonization in mice (Humphries et al., 1986). At appropriate tumor inocula, complete inhibition could be obtained by addition of GRGDS and this resulted in a substantial prolongation of survival of experimental animals (Humphries et al., 1988). Peptide activity was not due to toxicity and was specific since peptides containing minor changes in sequence were inactive. These other peptides were also unable to affect adhesion in vitro, indicating that the antimetastatic activity of GRGDS correlated with its antiadhesive activity. Yet another report showed that YIGSR amide, a synthetic peptide derivative representing one cell adhesion site in laminin, also possessed antimetastatic activity for melanoma cells (Iwamoto et al., 1987).

Although the studies described above point to a potential use for antiadhesive peptides in the prevention of tumor cell metastasis, a number of significant hurdles need to be overcome before this can be considered a reality. A key consideration is how this type of therapy could be utilized efficiently for treating a disease such as cancer. Specific uses might include prevention of tumor cell seeding during autologous bone marrow transplantation or as a prophylactic treatment during surgery. Indeed, in a mouse model for wound contamination, GRGDS washing was reported to reduce subsequent tumor regrowth (Whalen and Ingber, 1989).

The present invention provides a solution to many of the deficiencies in the prior art including a lack of selectivity, potency and stability by providing conformationally restricted peptide derivatives. In their native form, antiadhesive peptides are unusable, have relatively low activity, and must be administered in large amounts (i.e., extrapolation from mouse experiments suggests that a bolus of at least 10 g of GRGDS peptide would be required for a single treatment in humans and repeated treatments might be needed).

Gelatin binds to fibronectin with high affinity (Mosher, 1980; Forastieri and Ingham, 1983; Ingham et al., 1988, 1985; Garcia-Pardo and Gold, 1993; Nakamura et al., 1992; Lou et al, 1995). By specifically binding to fibronectin, gelatin could interfere with the binding of cells to fibronectin. However, gelatin is a complicated mixture of proteinaceous macromolecules with various ranges of molecular weight. In addition, the discovery of severe side effects resulting from decreased plasma fibronectin concentrations following infusion of gelatin-based plasma substitutes in man (Heene et al., 1968; Schόne, 1969; Zekorn, 1969; Brodin et al., 1984; Saddler and Horsey, 1987; Schreier et aL, 1987; Vedrinne et al, 1991; Blumenstock et al., 1993), limits the therapeutic use of gelatin. Using peptides with activity against fibronectin domains would be more appropriate. In order to facilitate the application of a peptide by making it more effective and stable, an adjuvant system, coupled to the peptide, may also be required. An ideal adjuvant or a drug delivery system, in which antineoplastic compounds could be encapsulated or incorporated, would provide this otherwise passive antimetastatic system with an additional function as a direct tumor cell killing agent.

The primary structure of collagen arises from the linkage of α-amino and imino acids by peptide bonds to form a polymer. Although some differences are apparent between collagens from different sources, there are certain features that are in common with, and uniquely characteristic of, collagen in general (Estoe and Leach, 1977). Although they show the general characteristics of collagen, the various collagen forms differ in amino acid composition and molecular weight. Fortunately, for the consideration of commercial gelatin production, a single collagen type "Type I", constitutes practically all the collagen present in bone and tendon and predominates (80% or more) in hide. The collagen unit, or monomer (tropocollagen), consists of a triple helix of three polypeptide chains, each of which has a helically coiled configuration. The three coiled chains are coiled around each other to give a right-handed coil of about 1.4 nm diameter, forming a tropocollagen molecule, which behaves as a rigid rod with a length of 300 nm and a molecular weight of approximately 300 kDa.

The tropocollagen molecules are joined together in an end-to-end fashion to form protofibrils and these, in turn, form fibrils by aggregation. Freely dangling (non-helical), single peptide chains (telopeptides) at one or both ends of the tropocollagen molecule are believed to be responsible for 'cementing' together the basic units (Jones, 1987). The triple helix structure of tropocollagen can be destroyed by the application of heat (Flory and Weaver, 1960) or by the use of compounds which destroy hydrogen bonds (Steven and Tristram, 1962), with resultant conversion to gelatin. The denaturation involves breaking only the hydrogen bonds and those hydrophobic bonds that help to stabilize the collagen helix. This is followed by the disentanglement of the chains and dissociation into smaller components with a random coil configuration.

The products of the denaturation of tropocollagen depend upon whether cross-links remain between the three-component protein chains. The three chains are not identical. Single-chain species (a-type) of approximately 100 kDa molecular weight (Kang et al., 1966; 1969) and two- chain species (b-type) of approximately 200 kDa are the main components, although a disordered form of tropocollagen itself (g-type), having a molecular weight of approximately 300 kDa, has been found (Piez et al., 1961). Gelatin is a generic term for a mixture of purified protein fractions obtained either by partial acid hydrolysis (type A gelatin) or by partial alkaline hydrolysis (type B gelatin) of animal collagen. Commercial gelatin may be a mixture of both types. The average molecular weight and the molecular weight distribution are both functions of the collagen-to-gelatin conversion process and they have important effects on the physical properties of the resulting gelatin. The amino acid composition of gelatin is quite similar to that of the parent collagen, except for the amide group and the organic residues. In such a polydisperse system, the average molecular weight attributed to a gelatin will depend upon the method of determination. The average molecular weight of commercial gelatins may vary from about 20 to 200 kDa, but appreciably higher molecular weights (in excess of 10⁶) have been reported for gelatin fractions (Tomka et al., 1975).

In one aspect of the present invention, the reagents and methodologies necessary for the further purification and characterization of commercial gelatin to identify binding domain(s) for fibronectin are provided. In one embodiment, tryptic digestion of the purified gelatin and fractionation of the resultant peptide components using a fibronectin affinity column results in the the isolation of a peptide having affinity for fibronectin. In a preferred embodiment, the peptides are designated PI and PII and have the following amino acid sequence:

PI: Thr-Leu-Gln-Pro-Val-Tyr-Glu-Tyr-Met-Val-Gly-Val

PII: Thr-Gly-Leu-Pro-Val-Gly-Val-Gly-Tyr-Val-Val-Thr-Val-Leu-Thr

Neither share significant identity with other known peptides that interact or bind with fibronectin. In addition, these sequences do not demonstrate identity with previously reported gelatin or collagen fragments. Since the sequences bear no relation to the glycine repeating unit characteristic of the well defined helical structure associated with collagen or gelatin, applicants hypothesize that the peptides are derived from the freely floating non-helical telopeptide structures. Moreover, Peptides I and II have no homology with other structures reported to bind to fibronectin such as the Antigen 85 protein derived from Mvcobacterium bovis (Leo et al., 1993; Oner et al., 1994). Thus, PI and PII represent previously unrecognized binding sites for the gelatin/collagen binding domain on fibronectin.

In another embodiment of the present invention, an assay for determining the ability of a peptide to bind fibronectin is provided. In one particular embodiment, such an assay is utilized to demonstrate binding of PI and PII to fibronectin-enriched surfaces.

In yet another embodiment, a method for stabilizing a peptide for administration to a patient is provided wherein a composition comprising a peptide and a phospholipid-stabilized nutritional emulsion is provided as a sterile emulsion. A complex of the present invention may be stabilized by covalent or non-covalent forces. Suitable phospholipid-stabilized nutritional emulsions may include but are not limited to "Intralipid" (Pharmacia-Upjohn, Sweden), Liposyn (Abbott Laboratories, Abort Park, IL), and Lipomul (B. Braun, Germany). In a preferred embodiment, emulsions having either Peptide I and Peptide II linked to phosphatidylethanolamine through either their N- or C- termini is provided. In a preferred preparative method of the invention, covalent coupling was preformed using heat-sterilized phospholipid-stabilized emulsions. In one preferred embodiment, Peptide I was coupled through its N-terminus to the intact emulsion, PIN-E. Several important functional properties of the preformed emulsion, such as droplet size and affinity for a fibronectin-rich surface, were unaffected by the methods provided.

In yet another embodiment, a method for delivering a bioactive agent to a fibronectin- enriched surface of a patient's body is provided. A composition comprising a peptide having affinity for fibronectin, a phospholipid emulsion and the bioactive agent is provided. In a preferred embodiment, such a composition is administered to a patient. In another preferred embodiment, the fibronectin-enriched surface is a tumor cell. A bioactive agent may comprise any compound having a direct or indirect effect upon a cell following contact with the cell or a factor influencing the activity of the cell.

It will be evident to one skilled in the art that this invention is applicable to targeted liposomal products (i.e., a peptide having affinity for fibronectin complexed with a bioactive agent and a liposomal formulation) with appropriate modifications. The preferred embodiment of the present invention comprises a phospholipid-stabilized emulsion because such formulation is more stable and offers the ability to deliver hydrophobic bioactive agents to a targeted site in the body (i.e., fibronectin-enriched surfaces). In such a manner, the effective dose of bioactive agent may be reduced, resulting in fewer side effects and increased efficacy. Liposomes may provide some utility in delivering hydrophilic bioactive agents to targeted sites in the body when prepared as a composition comprising a peptide of the present invention. It is more difficult, however, to ensure the appropriate size and stability of the composition using liposomes.

Many different compounds which are bioactive agents may be utilized in practicing the present invention. Suitable bioactive agents may include but are not limited to steroidal compounds, an adjuvant, paclitaxel (Taxol™), vinblastine, carmustine, procarbazine (base), etoposide, teniposide, cisplatin, methotrexate (base), vindesine, mechloroethamine (nitrogen mustard base), fluorouracil. Other such compounds include but are not limited to antibacterial compounds such as gentamycin, antiviral agents such as rifampacin, antifungal compounds such as amphoteracin B, anti-parasitic compounds such as antimony derivatives, tumoricidal compounds such as adriamycin, anti-metabolites, neurotransmitter antagonists, and anti-inflammatories such as indomethacin. Other suitable bioactive agents may include anti-asthmatics such as melairoterenol, aminophylline, theophylline, terbutaline, norepinephrine, ephedrine, isoproternol, adrenalin.

The compositions of the present invention may be administered parenterally, by inhalation spray, rectally, or topically in dosage unit formulations containing conventional pharmaceutically acceptable carriers, adjuvants, and vehicles. The term parenteral as used herein includes, subcutaneous, intravenous, intramuscular, intrasternal, infusion techniques or intraperitoneally. In a preferred embodiment of the present invention, the emulsion is administered as an injectable preparation. The dosage regimen for the compositions of this invention is based on a variety of factors, including the type of disease, the age, weight, sex, medical condition of the patient, the severity of the condition, the route of administration, and the particular compound employed. Thus, the dosage regimen may vary widely, but can be determined routinely using standard methods. The compositions of the present invention can be processed in accordance with conventional methods of pharmacy to produce medicinal agents for administration to patients, including humans and animals.

While the compounds of the invention can be administered as the sole active pharmaceutical agent, they can also be used in combination with one or more compounds of the invention or other agents. When administered as a combination, the therapeutic agents can be formulated as separate compositions that are given at the same time or different times, or the therapeutic agents can be given as a single composition. The following Examples are for illustrative purposes only and are not intended, nor should they be construed as limiting the invention in any manner. Those skilled in the art will appreciate that variations and modifications can be made without violating the spirit or scope of the invention. EXAMPLES

Example 1 Preparation of purified gelatin Gelatin is a generic term for a mixture of purified protein fractions obtained either by partial acid hydrolysis (type A gelatin) or by partial alkaline hydrolysis (type B gelatin) of animal collagen. Commercial gelatin may be a mixture of both types. The protein fractions consist almost entirely of amino acids joined together by amide linkages to form linear polymers. The average molecular weight and the molecular weight distribution are functions of the collagen-to- gelatin conversion process and they have an important effect on the physical properties of the gelatin. Bovine skin gelatin (lime cured, type B, Bloom strength 225 g, Sigma Chemical, St.

Louios, MO), dissolved in a Tris buffer (0.02 M Tris + 0.025 M KC1) was applied to the Sephadex^® G-200 (superfine, Pharmacia Fine Chemicals, Uppsala, Sweden) column (50 x 1.5 i.d. cm) and eluted with the Tris buffer (~ rate 3 mL/hr). Each 2 mL fraction was collected with a Bio-Rad Model 2110 Fraction Collector (Richard, CA). The absorbance of the fractions were monitored at UV 215 nm (Backman DU-65 Spectrophotometer, Beckman Instruments, Fullerton, CA). Absorbance of each fraction was monitored at 215 nm and showed a major peak between fraction numbers 11 and 17, Fig. 1. These fractions were combined and characterized by sodium dodecylsulfate polyacrylamide gel electrophoresis (SDS-PAGE), running against protein molecular weight standards obtained from Bio-Rad (Hercules, CA). Using a 5% polyacrylamide gel and staining with Coomassie Brilliant Blue demonstrated that the purified gelatin had a molecular weight range of 55-646 kDa. The major components were identified according to the scheme of Jones 1987 as the α single chain species of approximate molecular weight 100 kDa, the disordered tropocollagen of approximately 300 kDa and smaller amounts of lower molecular weight species, together with oligomers of the α-chains. This purified gelatin was used in subsequent work. Trypsin catalyses the hydrolysis of lysyl and arginyl bonds, except where the following residue is proline. The mode of action of trypsin on gelatin is believed to involve attack of the terminal peptides, leading to the breaking of intra- and intermolecular cross-links, leaving the helical region intact (Jones, 1987). Purified gelatin was digested with TPCK trypsin (Sigma Chemical, St. Louis, MO) for 90 mins at 37°C in digestion buffer (0.1 M NH₄HC0₃, 0.1 mM CaCl₂ at pH 8.0). Size-exclusion HPLC profiles of the gelatin and the cleaved gelatin mixture are shown in Figs. 2 and 3. It is apparent that the gelatin had been cleaved into a series of smaller molecular fragments ranging from approximately 1.0 to 10.0 kDa after trypsin digestion, while the initial starting material contained a group of materials with molecular sizes ranging from 33.7 to 413.3 kDa, estimated from the calibration of the column with standards of known molecular weight.

The wide range of molecular weight of gelatin is readily explained by its method of preparation on the commercial scale, being the product of denaturation of tropocollagen (the basic collagen unit) and depending upon whether or not cross-links remain between the three chains. Single-chain species (α-type) of approximately 100 kDa and two-chain species (β-type) of approximately 200 kDa, are the main components. A disordered form of tropocollagen itself (γ-type), having a molecular weight of approximately 300 kDa, may also be found (Jones, 1987). The results here show that the gelatin contains all types of denatured forms and confirms the complex constitution.

Example 2

Isolation of peptide I (PI) and peptide II (PII) using a fibronectin-affinity column To identify those peptides having affinity for fibronectin, a fibronectin-affinity column was generated. The methodology followed in this example is illustrated below, (a) Preparation of the affinity matrix

1.36 g of CNBr-activated Sepharose^® 4B (Pharmacia LKB Biotechnology AB, Uppsala, Sweden) was completely swollen in 1 mM HC1 and washed extensively with ImM HC1 (about 400 mL). Well drained, the swollen gel was washed with lOOmL of coupling buffer (0.1 M NaHC0₃, 0.5 M NaCl, pH 8.3), and finally equilibrated with 8mL of the coupling buffer. 20 mg of human plasma fibronectin (Gibco BRL Life Technologies™, Grand Island, NY) was covalently coupled to ~ 4 mL of the swollen CNBr-activated Sepharose^® 4B at 3.33 mg/mL in the coupling buffer. The coupling process was carried out on a Hematology/Chemistry Mixer (Mixer Model 346, Fisher Scientific, Fair Lawn, NY) for 20 hrs. After the coupling was completed, the matrix was again washing extensively with the coupling buffer (120 mL), and blocked with 15 mL blocking buffer (0.2 M Tris-HCl, 0.2 M glycine, pH 8) at 4°C for 2 hours on the Hematology/Chemistry Mixer. The blocked matrix was washed alternatively with the coupling buffer and acetate buffer (0.1 M acetate buffer (pH 4.0) + 0.5 M NaCl) for three times. The matrix was finally equilibrated with 0.01 M phosphate buffer (PB) and loaded into a Poly- Prep Chromatography Column (4 x 0.8 cm i.d., Bio-Rad, Hercules, CA) before use. After each use, the fibronectin-Sepharose was reequilibrated by washing with 4 M urea, distilled water and finally 0.01 M PB. The concentration of fibronectin before and after the coupling was measured using the BCA Protein Assay Reagent and albumin standard (Pierce, Rockford, IL), following the manufacture's instruction, (b) Fractionation of fibronectin-binding fragments of gelatin after trypsin digestion Gelatin purified by Sephadex G-200 was digested with TPCK-trypsin (Sigma, St. Louis,

MO) (25:1, w/w) for 90 minutes at 37°C in a digestion buffer (0.1 M NH₄HCO₃, 0.1 mM CaCl₂, pH 8.0). The digested mixture was applied to the fibronectin-Sepharose column, prepared above, at 5 mg protein/mL of swollen gel. Unbound fragments were removed by washing the column with the digestion buffer and the fragments that bound with high affinity were eluted with 0.5 M Tris-HCl (pH 7.2) buffer containing 8 M urea. These fragment-containing fractions were fast desalted by passing through a Sephadex^® G-25 M column (PD-10 column, Pharmacia Biotech AB, Uppsala, Sweden) and stored at 4°C before further purification.

2. Size exclusion HPLC analysis of the fibronectin-binding fragment-containing fractions

Instrument — Waters HPLC system (Waters, Millipore, Beford, MA), including Waters 501 HPLC pump, Waters U6K injector, Waters 486 Tunable Absorbance Detector, Waters Pump Control Module and Millennium™ 2010 software, version 2.0. Samples — Purified gelatin by Sephadex G-200, digested mixture and fibronectin-binding fragments(s)-containing fraction.

Method — 10 μL of each of the samples was loaded to the column (Bio-Sil™ SEC-125 column, 300 x 7.8 i.d. mm, Bio-Rad, Hercules, CA) and eluted with the mobile phase (0.05 M NaH₂PO₄ + 0.05 M Na^PO, + 0.15 M NaCl, pH 6.8) at a flow rate of 1.0 mL/min. The detector wavelength was set at 215 nm. Peaks with various molecular weight were estimated with the calibration of the Gel Filtration Molecular Weights Standards (Bio-Rad, Hercules, CA). Peaks in the fibronectin-binding fragment(s)-containing fraction were separately collected manually and the same peak from each run was pooled and concentrated by Speed- Vac.

3. Further purification and N-terminal amino acid sequencing

The concentrated fragment solutions were further purified by reverse-phase HPLC (Waters system, as described above) on the Macrosphere 300 C18 7μ column (250 x 4.6 i.d. mm, Alltech, Deerfield, IL) eluted with 0.1% trifluoroacetic acid (TFA) in acetonitrile/water (25:75, v/v). The fractions thus obtained were pooled and concentrated by Speed-Vac. The concentration of the concentrated fractions were estimated by the BCA Protein Assay Reagent (Pierce, Rockford, IL).

Two main components were identified as being interactive with fibronectin, with molecular weights of 0.65 and 1.07 kDa (Table I).

RT: Retention Time a: kDa was estimated according to calibration curve using molecular weight standards b : % fraction was based on the % of each peak area These fragments were further purified and sequenced on an Applied Biosystems Model 477A Protein Sequence and determined to have the following amino acid sequences: 1. Thr-Leu-Gln-Pro-Val-Tyr-Glu-Tyr-Met-X-Gly-Val

2. Thr-Gly-Leu-Pro-Val-Gly-Val-Gly-Tyr-Val-Val-Thr-Val-Leu-Thr Substituting Val for the unknown in Peptide 1 enabled us to obtain sufficient quantities of high quality (» 98% purity) of peptide for further investigation from a commercial supplier (Genosys Biotechnologies, Woodland, TX). Thus, the sequences of Peptide I (PI) and Peptide II (PII) are illustrated below:

Peptide I (PI; SEQ ID NO. 1) Thr-Leu-Gln-Pro-Val-Tyr-Glu-Tyr-Met-Val-Gly-Val

Peptide II ("PII"; SEQ ID NO. 2)

Thr-Gly-Leu-Pro-Val-Gly-Val-Gly-Tyr-Val-Val-Thr-Val-Leu-Thr

The lack of a three-peptide repeating sequence involving glycine suggests that neither of these peptides is derived from the helical structure characteristic of collagen or gelatin but rather originates from the terminal tropopeptidal chains that are attached to the helical region. These sequences suggest that they originate from the non-helical region of the gelatin, since, as described elsewhere (Jackson and Grant, 1974; Knight and Hunt, 1974; Rucker and Murray,

1978; Mechanic, 1988; Jones, 1987), the collagen unit (tropocollagen) consists of a triple helix of three, not identical, polypeptide chains, each of which has a helically coiled configuration. Each helix is built up of amino-acid triplets of glycine together with two other amino acids. However, freely dangling single peptide chains at one or both ends of the tropocollagen molecule do not show this triplet sequence. These may cement together the basic units. It also is possible that these peptides contain the fibronectin-binding domain(s) of gelatin. The specific location and function of these peptides in the overall molecular structure may make them more accessible to the outside environment and, therefore, readily available for binding.

Example 3 Affinity of PI and PII for fibronectin

ELISA is known for its sensitivity, specificity, accuracy, reproducibility and requirment for small quantities of sample (Engvall, 1977; van Weemen and Schuurs, 1971; Belanger, 1978). Although ELISA has been used to quantitate fibronectin concentration in the plasma (Damas et al., 1987; Vincent et al., 1988; Daudi et al., 1991), those skilled in the art have not been able to utilize the assay to measure the binding affinity between fibronectin and its bound components. Determined by this ELISA, the binding of gelatin and the two fibronectin-binding peptides to fibronectin exhibited the typical binding saturation curves with an increasing concentration of fibronectin (Figure 5). The negative control, obtained by replacing the antifibronectin (human plasma) antiserum (rabbit) with normal rabbit serum, did not result in any subsequent absorbance reading, indicating that there was no binding between normal rabbit serum and human plasma fibronectin. Accordingly, color development in the assay was based on binding between human plasma fibronectin on the wells and the anti-fibronectin (human plasma) antiserum (rabbit). Moreover, the binding of fibronectin to the plastic-bound gelatin could be inhibited in a dose- dependent fashion by pre-incubating the fibronectin solutions with gelatin (Figure 6), which further indicated that the ELISA assay was solely based on the specific interaction between fibronectin and gelatin. By combining binding data measured by ELISA with advanced nonlinear regression analysis, binding models and binding constants could be obtained accurately. The binding of gelatin and fibronectin-binding peptides to fibronectin fitted a two-site binding model (Table II). The binding constant of gelatin at the higher affinity site (Kd, = 5.38xl0^"9 M) showed similar values to those in the literature (Mosher, 1980; Forastieri and Ingham, 1983; Garcia-Pado and Gold, 1993; Nakamura et al., 1992; Lou et al., 1995), further suggesting the reliability of the method.

Table II. Non-linear regression result of binding of gelatin and fibronectin-binding peptides to fibronectin, determined by the ELISA and analyzed by GraphPad Prism™, version 2.0.

The two fibronectin-binding peptides fractionated from gelatin all possessed significant higher binding affinities to fibronectin than gelatin alone (Table III; t = 5.616, df= 4, p < 0.01 for peptide I; t = 4.476, df = 4, p < 0.01 for peptide II). However, the binding affinity to fibronectin between peptide I and peptide II did not show any significant difference (t = 0.08956, df = 4, p > 0.1). In conclusion, two peptides, apparently derived from the non-helical region of the gelatin structure, were fractionated and sequenced to demonstrate previosuly unknown fibronectin- binding regions. Both peptides possess significantly higher binding affinities to fibronectin than their parent gelatin.

Table III: Published binding constants of collagen or gelatin and peptides derived from gelatin to fibronectin³.

Collagen or gelatin type K_d (M) Publication

Nonhelical fragment of collagen -2.5 x lO^"9 Mosher, 1980 Gelatin from bovine bone^b 2.5 x 10^"9 Nakamura et al, 1992 Rat tail collagen 1.3 x l0⁸ Ingham et a]., 1988

Bovine skin gelatin, type B formulated into microspheres 4.3 x 10^"8 Lou et al., 1995 Bovine skin gelatin, type B 5.4 x lO^"9 In this present study Peptide I derived from bovine skin gelatin, type B 6.6 x lO^"10 In this present study Peptide II derived from bovine skin gelatin, type B 7.7 x 10 ' In this present study

^aonly the higher binding constants are listed, in the case of more than one class of binding sites existing. ^bno information about the gelatin type. Using an enzyme-linked immunosorbent assay (ELISA) developed in this laboratory applicants' measured the affinity of fibronectin for the purified gelatin and the peptides isolated above. Using an improved method for calculating affinities we have demonstrated that there are two binding sites in fibronectin, one more dominant than the other. These values for the stronger binding are shown in Table III, together with previously published values for the gelatin/fibronectin affinities.

Example 4 Coupling of Peptides I and II to a phospholipid-stabilized emulsion It is known that commercially available phospholipid-stabilized nutritional emulsions such as Tntralipid' (Pharmacia-Upjohn, Clayton, NC) contain approximately 160 mg/100 mL product of phosphatidylethanolamine (PE) (Herman, 1992). PE provides a primary amine group available for covalently bonding with peptides using established techniques through either the C- or N- termini. Tolylene-2, 4-diisocyanate (TDIC) was utilized as a cross-linking agent between PE and the N- terminal group of the peptides. l-ethyl-3-(3-dimethylaminopropylcarbodiimide (EDCI) was utilized to link the primary amine of the PE and the carboxylic acid terminus of the peptides, in a one-step reaction. A. Coupling of N-termini of the peptides to the ^' Intralipid'

To 1 mL of the 'Intralipid' was added 20 μL of a 2% solution of TDIC in p-dioxane. The reaction mixture was incubated at 17°C for 2 hr with gentle shaking. Polymerized material derived from the TDIC was removed from the mixture by centrifugation (Sorvall^® RC-5B Refrigerated

Superspeed Centrifuge, DuPont Instrument, New Torn, CT) at 400 x g for 5 min. The resulting supernatant was then mixed with 1 mL of the peptide solution (2 mg/mL) and incubated at 37°C for 2 hr to produce peptide-coupled ^Λ Intralipid'. The peptide-coupled 'Intralipid' droplets were separated from non-covalently bound peptide and free peptide by adding 8 mL of 0.9% NaCl containing 0.02% EDTA to the coupled mixture with gentle agitation, followed by chromatography on a Bio-Gel A- 1.5m column (1.5 x 30 cm) with 0.9%) NaCl as the elution buffer. B. Coupling of C-termini of the peptides to the phospholipid-stabilized emulsion

To a mixture of Tntralipid' 20%> (1 mL) and solution of the peptide (1 mL, 2 mg/mL), was added 0.2 mL of a 20% solution of EDCI. After adjusting the pH to 4.7 with 0.3 N HC1 and standing for 2 hr at 24°C, this mixture was fractionated to give covalently coupled peptide- Tntralipid' with the separation procedure described above. C. Orientated coupling of peptides into the phospholipid-stabilized emulsion

Peptides are also able to associate with phospholipids to form complexes which are stabilized by non-covalent forces such as hydrogen bonds. Both phosphatidylcholine (PC) and phosphatidylethanolamine (PE) associate with, for example, GRDGS at an approximate reatio of 1 mol peptide:2-3 mols of phospholipid. The peptide molecules are associated with the phospholipids in the lipid emulsion mesophase, most probably oriented at random to the outer surface. By covalently bonding the peptide to PE externally, the peptide is orientated allowing the peptide to project into the aqueous phase. In such a manner, the availability of the peptide for binding to fibronectin is increased. Following the external reaction of the peptide and PE, the covalently bonded compound is prepared in ethanol, and sterile-filtered into a pre-sterilized emulsion. The ethanol is removed with a sterile stream of nitrogen. The strength of this reaction depends upon the length and conformation of the projecting peptide. A preferred emulsion comprises PUCE; a more preferred emulsion is PICE; an even more preferred emulsion is PILNE; and, a most preferred emulsion is PINE.

D. Coupling kinetics of Peptide I onto the phospholipid-stabilized emulsion

Various amounts of Peptide I were coupled to the Tntralipid' (i.e. TDIC and EDCI). Peptide I that was not coupled to the Tntralipid' was separated from the coupled material. In the absence of coupling agents, the Tntralipid' was simply incubated with Peptide I at room temperature for 2 hr and then chromatographed on the Bio-Gel A- 1.5m column with or without the addition of 0.02% EDTA. The amount of Peptide I remaining coupled or adsorbed (no coupling agents) to Tntralipid' was collected after three cycles of trichloracetic acid precipitation and measured by the Modified Lowry Protein Assay Reagent with Peptide I alone as standard. The size parameters of the Tntralipid' before and after the coupling are summarized in Table IV. TLC analysis of the PE alone and PE in 'Intralipid' after coupling to Peptide I was performed. After the coupling process, the intact PE content in the extract was decreased significantly and new spots, with shifter R_f values were present in larger quantities. The new spots were apparently the products of the coupling processes.

Table IV: Size parameters of ^"Intralipid' before and after coupling to fibronectin-binding peptides at either terminus.

Note: PLN-E: Peptide I coupled to Tntralipid' at its N-terminus. PIC-E: Peptide I coupled to Tntralipid' at its C-terminus. PIIN-E: Peptide II coupled to Tntralipid' at its N-terminus.

PIIC-E: Peptide II coupled to Tntralipid' at its C-terminus. d „: Geometric mean diameter on a number basis. s_g: Geometric standard deviation on a number basis. Peptide I coupled significantly to the 'Intralipid' with the cross-linking agents, TDIC or EDCI. Non-specific binding of the peptide to the Tntralipid' was also observed in the absence of either of the cross- linking agents. The coupling amount of Peptide I onto the Tntralipid' was related to the amount added in the reaction. The amount of Peptide I coupled to the Tntralipid' increased almost linearly, until a saturation limit was reached at ~ 0.43 μ mole/μ mole PE (at N- terminus) and 0.48 μ mole/μ mole PE (at C-terminus).

Example 5 Binding of Peptide I and Peptide II, alone and coupled to a phospholipid-stabilized emulsion to fibronectin-enriched surfaces An ELISA method was established in this study to measure the binding affinities of the fractionated peptides to fibronectin. Briefly, Microtiter wells (96-well vinyl EIA/RIA plates,

Gibco, Grand Island, NY) were first coated with gelatin or each fractionated peptide and incubated at 4°C for 2 days. After the removal of the supernatants, the wells were washed with

PBS-T (0.01 M phosphate buffer saline pH 7.4 (PBS) supplemented with 0.05% Tween-20 and 0.02% sodium azide). The wells were then blocked with blocking buffer (0.25% bovine serum albumin [BSA]) for 1 hour at 37°C. After a serial dilution of fibronectin was added at 37°C for 2 hours, anti-fibronectin antiserum (Rabbit, Calbiochem, San Diego, CA) in PBS-T (1 :1000) was added to the wells and incubated at 37°C for 1.5 hours. The wells were then further incubated with the Goat-anti-Rabbit Ig G (H+L) alkaline phosphatase conjugate (Bio-Rad, Hercules, CA) diluted 1000 times with conjugate buffer (0.05 M Tris + 1% BSA + 0.02% sodium azide, pH 8.0) at 37°C for 30 minutes. The plates were incubated with the substrate buffer (0.05 M glycine +

1.5 mM magnesium chloride, pH 10.5) and substrate (p-Nitrophenyl phosphate disodium, Sigma, St. Louis, MO) for 15 minutes at 37°C. The enzymatic reaction was terminated by the addition of 1M NaOH and absorbance at 410 nm measured using the Dynatech plate reader (MR 300, Dynatech Laboratories, Chantilly, VA). The wells coated with coating buffer served as controls. The negative control was provided by replacing the antifibronectin (human plasma) antiserum (rabbit) with normal rabbit serum. The inhibition experiment was conducted by pre-incubating a concentration of fibronectin (10 μg/mL) with various amounts of gelatin.

The binding data were analyzed by non-linear regression using GraphPad Prism™, version 2.0 (San Diego, CA). Data were fitted into both one-site binding and two-site binding models and compared. The best fit model was selected by the program and the binding constants calculated according to the model.

One-site binding: A = A„_ιaχ.[L] {K_d + [L]} (1)

Two-site binding: A = ^...[L] {K_dl + [L]} + B _max2. [L] {K_d2 + [L]} (2) where, A is the absorbance at 410 nm, A_^ is the maximum absorbance at saturation; (L) is the molar concentration of the free fibronectin and K_d is the dissociation constant. Binding affinities were measured using an improved ELISA method, data being analyzed by software provided by GraphPad Prism™ (Version 2.0) (GraphPad Software, San Diego, CA). Binding activities fitted a two-site binding model and data are shown in Table 3.

Since Peptide II coupled to the emulsion appeared to affect the state of aggregation of the dispersed droplets, Table 2, we preferred the N- coupled Peptide I system (PIN-E) as the embodiment of our invention since the strong binding site activity was higher than that of the unbound peptides and significantly higher than that of the starting gelatin (Tables 1 and 3) whilst, at the same time, the physical properties of the parent emulsion system were only slightly affected (Table 2).

Table 3: Summary of the binding parameters of Peptide I, Peptide II and after coupling to "Intralipid'.

PLN-E: Peptide I coupled to 'Intralipid' at its N-terminus. PIC-E: Peptide I coupled to Tntralipid' at its C-terminus. PLTN-E:Peptide II coupled to Tntralipid' at its N-terminus. PIIC-E Peptide II coupled to Tntralipid' at its C-terminus.

EXAMPLE 6

The effect of PLN-E against in vitro fibronectin-mediated models of cell spreading The murine SI 80 sarcoma cell line is known to express fibronectin and gelatin microparticles were shown to inhibit in vitro adherence of the SI 80 cells to a polystyrene substrate (Lou et al., 1995). This methodology was adapted to determine if PIN-E had the same activity. In addition, the baby hamster kidney (BHK) is a well recognized model of cell spreading mediated by fibronectin (Grunnel et al., 1977; Yamada and Kennedy, 1984) and is amenable to demonstrating effectiveness of fibronectin antagonists. A. Materials and methods Materials. SI 80 murine sarcoma cells were obtained from the American Type Culture Collection (Rockville, MD). The cells were maintained in vitro at this Institute in CMEM-E (Eagle's minimum essential medium with non-essential amino acids, Earle's basal salts, 5% calf serum, 100 units/mL penicillin and 100 μg/mL streptomycin) in an atmosphere of 5% C0₂, 95% relative humidity at 37°C. Baby hamster kidney cells (BHK-21) were obtained from the American Type Culture Collection (Rockville, MD) and maintained in minimum essential medium (Eagle) with non-essential amino acids, 90% Earle's BSS, 10% fetal bovine serum. Human plasma fibronectin were obtained from Gibco BRL (Grand Island). Bovine serum albumin (BSA) was from Sigma (St. Louis, MO). PLN-E was prepared as in Example 4. Cluster plates (24-well) were from Coster (Cambridge, MA).

B. Methods

Targeting activity of PLN-E to fibronectin-bearing tumor cells

Viable SI 80 cells, 1 x 10⁶ per well, in MEM-E (Eagle's medium with serum) (0.9 mL) plus 0.1 mL of different concentrations of PIN-E (concentration at zero as control) were distributed among the center eight wells of the cluster plates. After 5 hours incubation at 37°C, non-adherent cells were resuspended by gentle aspiration three times with phosphate-buffered saline (PBS) from a Pasteur pipet and removed. Adherent cells were harvested and counted in a hemocytometer. Inhibition of the targeting of PIN-E to SI 80 cells by fibronectin was carried out by pre-incubating PLN-E (equivalent to 7.77 μg/well of Peptide I) with a series of different fibronectin concentrations. The rest of the experiment was carried out in exactly the same way as above.

C. Inhibition of fibronectin-mediated cell spreading

1. Inhibition of fibronectin-mediated cell spreading by PLN-E

Fibronectin-mediated cell spreading of baby hamster kidney (BHK) cells was assayed as described (Grinnel et al., 1977; Yamada and Kennedy, 1984). Tissue culture clusters, 24-well (Coster, Cambridge, MA), were preincubated with 3 μg/mL of fibronectin in a adhesion medium (150 mM NaCl, 3 mM KC1, 1 mM CaCl₂, 0.5 mM MgCl₂, 6 mM Na_^HPO,, 1 mM KH₂P0₄, pH 7.3) at room temperature for 60 min and the non-specific adsorption sites blocked with 10 mg/mL heat-denatured (80°C for 30 min) BSA for 30 min. BHK cells were trypsinized, washed three times with PBS and incubated in adhesion medium with or without added various concentrations of PIN-E for 45 min at 37°C. After this attachment period, cells were fixed with 2.5% glutaraldehyde in PBS for 1 hr, and cell spreading was quantitated as described (Yamada and Kennedy, 1984). Controls for the background spreading were wells that had not been coated with fibronectin; controls for the possible non-specific affect of protein present during the cell-spreading assay included incubation with various concentrations of BSA.

2. Competitivity of the inhibition effect of PIN-E by fibronectin The competitivity of the inhibition by fibronectin was studied by preincubation of 24-well tissue culture clusters with various concentrations of fibronectin in the adhesion medium for 60 min at room temperature. The following non-specific site blocking with BSA and incubation of BHK cells with or without PLN-E were carried out as described above.

3. Reversibility of the inhibition effect of PLN-E After the inhibition of BHK cell spreading by PIN-E, the cells were washed three times with adhesion medium and incubated in the adhesion medium only for another 45 min at 37°C. The cell spreading was again quantitated under the same condition.

D. Results

1. In vitro targeting PIN-E induced maximally approximately 23.7% decrease of adherent cells. The decrease of adherent cells was believed to be due to the binding of the fibronectin on the surface of the cells to the Peptide I coupled to Tntralipid' droplets, which were in suspension.

2. Inhibition of the targeting by fibronectin

Pre-incubation of PIN-E with fibronectin significantly reduced its ability to inhibit SI 80 cell adherence. At the fibronectin concentration of 250 μg/well, the inhibition of the adherence of

SI 80 cell by PIN-E was totally abolished. This result further indicates that the interference of SI 80 cell adherence by PIN-E is entirely due to the interaction between Peptide I and fibronectin on the surfaces of SI 80 cells.

3. Inhibition of fibronectin-mediated cell spreading by PLN-E

BHK cells spread nearly completely on tissue culture substrates procoated with 3 μg/mL human plasma fibronectin, while the spread percentage approached zero on the substrates not precoated with fibronectin. This fibronectin-mediated spreading was progressively inhibited by increased concentrations of PIN-E added to the adhesion medium. Inhibition appeared maximal at an equivalent of added Peptide I to 932 μg/well (calculated from coupling density data). Although spreading was severely inhibited, some cells appeared to display abortive spreading, with increased phase density of cytoplasm according to phase-contrast microscopy, but with poor elaboration of peripheral lamellae.

BSA that was tested at various concentrations as controls for non-specific effect of protein added to the assay system did not affect fibronectin-mediated spreading of these cells. The inhibition of BHK cells spreading by PIN-E was competitive because the inhibitory effect was diminished by preadsorbed fibronectin in a dose-dependent fashion. Removal of the PIN-E by extensive washing resulted in the BHK cell recovering their spreading activity, confirming that the activity of PIN-E was not due to toxicity but rather competition with fibronectin.

Example 7

Generation of stable PIN-E complexes

Cleavage of the associated peptide was measured in the PIN-E emulsion systems by storage at refrigerator or ambient room temperatures or at 37°C in distilled water or in rabbit serum at

37°C. Aliquots were removed at timed intervals and the cleaved peptide separated from the emulsion coupled material by centrifuging through 5 kDa molecular weight cut off membrane filters (Alltech Associated, Deerfield, IL) followed by two further water washes. The filtrate were pooled and assayed for free peptide using the modified Lowry protein assay reagent (Pierce, Rockford, IL) with Peptide I as standard. In the case of the serum experiments the separation was achieved using a Bio-Gel A- 1.5m column, collecting the uncoupled peptide after three cycles of trichloracetic acid precipitation. In addition sample particle size was determined after separation by photon correlation spectroscopy (Malvern Instruments, Malvern, UK).

Results showed that the PIN-E system was stable for at least 22 days at refrigerator temperature with no changes in free peptide or particle size being found. At ambient room temperature no change was observed for at least six days but by the 22nd day the peptide remaining coupled had dropped to 55% of the starting value and the measured droplet size had also increased under these conditions. Incubation at 37°C with rabbit serum showed only an immediate but small drop in bound Peptide I, with only minor changes over the first two days but aggregation or increased size over six days. It is concluded that the peptide-covalently bonded to phospholipid- stabilized emulsions are substantially stable at refrigerated temperatures for at least one month and for at least 1 or 2 days at ambient room temperature. These conditions indicate that the system is stable enough for storage purposes prior to administration and for all practical purposes for the administration process if given by addition to an intravenous administration line.

Example 8 Phospholipid-stabilized emulsions having affinity for fibronectin containing a bioactive agent To deliver a bioactive agent to a fibronectin-enriched surface, such as a tumor cell, a phospholipid-stabilized emulsion having affinity for fibronectin and containing a bioactive agent is generated. One such emulsion includes the anti-tumor cell bioactive agent paclitaxol. To generate an emulsion containing paclitaxol, 100 mg of paclitaxol is dissolved with gentle heat in 20 g winterized soybean oil. 1.2 g purified egg lecithin is then dissolved in the solution with gentle agitation. The oil solution is then added with stirring to a solution of 2.5 g glycerol USP in 75 ml water. The crude emulsion so obtained is then repeatedly passed through a homogenizer until the mean droplet size is at least 200 nm in diameter. The targeting ligand (peptide covalently bonded to PE) is slowly added in warm ethanolic solution with slow stirring and the ethanol removed in a stream of sterile-filtered nitrogen. The composition is then sterile- filtered through a 400 nm pore size filter in series with 800 nm and 1.2 μm filters to remove unnecessary particulate matter and bacteria. After analytical characterization, the volume of the emulsion is adjusted to 100 ml with sterile water for injections and the pH adjusted to 8.0 with sterile sodium hydroxide solution. This provides a 1 mg/ml slow infusion product suitable for dilution with regular intravenous nutritional emulsion products as required on an individual basis.

Example 9 Administration of phospholipid-stabilzed emulsions having affinity for fibronectin following tumor resection

Following surgical resection of a primary tumor in a cancer patient, it is common for metastases to develop and ultimately lead to a poor prognosis for the patient. The present invention provides the skilled artisan with the methodologies and reagents for preventing or decreasing the likelihood of post-resection metastatic tumor development. A patient having a tumor is treated surgically to remove the tumor (i.e., surgical resection of the tumor). Prior to or immediately following surgery and continuing for a period of time that the skilled artisan believes is sufficient, a composition of the present invention comprising a bioactive agent having anti-metastatic properties, a peptide having affinity for fibronectin, and a phospholipid-stabilized nutritional emulsion is administered to the patient. In such a manner, cells in the region of a fibronectin-enriched surface of the patient's body are contacted with the composition, and in particular the bioactive agent having anti-metastatic properties. Metastatic growth of any remaining tumor cells is prevented or significantly slowed by such treatment.

While a preferred form of the invention has been shown in the drawings and described, since variations in the preferred form will be apparent to those skilled in the art, the invention should not be construed as limited to the specific form shown and described, but instead is as set forth in the claims.

SEQUENCE LISTING

(1) GENERAL INFORMATION:

(i) APPLICANT: Board of Trustees of the University of Illinois

(ii) TITLE OF INVENTION: Unique Peptides For Targeting Fibronectin-Enriched Surfaces And A Method For Their Delivery In The Treatment Of Metastatic Cancer

(iii) NUMBER OF SEQUENCES: 2

(iv) CORRESPONDENCE ADDRESS:

(A) ADDRESSEE: McDonnell, Boehnen, Hulbert & Berghoff (B) STREET: 300 S. acker Drive

(C)CITY: Chicago

(D) STATE: IL

(E) COUNTRY: USA

(F)ZIP: 60606

(v) COMPUTER READABLE FORM:

(A) MEDIUM TYPE: Floppy disk

(B) COMPUTER: IBM PC compatible

(C) OPERATING SYSTEM: PC-DOS/MS-DOS (D) SOFTWARE: ASCII

(vi) CURRENT APPLICATION DATA: (A) APPLICATION NUMBER: (B) FILING DATE: (C) CLASSIFICATION:

(viii) ATTORNEY/AGENT INFORMATION: (A) NAME: Hughes, A. Blair (B) REGISTRATION NUMBER: (C) REFERENCE/DOCKET NUMBER: 96,2088-A

(ix) TELECOMNUNICATION INFORMATION: (A) TELEPHONE : 312-913-0001 (B)TELEFAX: 312-913-0002

(2) INFORMATION FOR SEQ ID NO : 1

(i) SEQUENCE CHARACTERISTICS: (A) LENGTH: 12 amino acids (B)TYPE: peptide (D) TOPOLOGY: linear

(ii) MOLECULE TYPE: amino acid sequence (ix) FEATURE: (A) NAME/KEY: Misc_feature

(B) LOCATION: 1..12

(D) OTHER INFORMATION: Peptide I (PI)

(xi) SEQUENCE DESCRIPTION: SEQ ID NO : 1 :

ThrLeuGlnProValTyrGluTyrMetValGlyVal 12

(3) INFORMATION FOR SEQ ID NO : 2 : (i) SEQUENCE CHARACTERISTICS: (A) LENGTH: 15 amino acids (B)TYPE: peptide (D) TOPOLOGY: linear (ii) MOLECULE TYPE: peptide

(ix) FEATURE:

(A) NAME/KEY: misc_feature

(B) LOCATION: 1..15

(D) OTHER INFORMATION: Peptide II (PII) (xi) SEQUENCE DESCRIPTION: SEQ ID NO : 2 :

ThrGlyLeuProValGlyValGlyTyrValValThrValLeuThr 15

REFERENCES

Akiyama, S.K., Olden, K., and Yamada, K.M.: Fibronectin and integrins in invasion and metastasis, Cancer and Metastasis Review, 14(3), 173-189, 1995.

Akiyama, S.K., Nagata, K., Yamada, K.M.: Cell surface receptors for extracellular matrix components, Biochim. Biophys. Acta, 1031, 91-110, 1990.

Blumenstock, F.A., Celle, P.L., Herrmannsfoerfer, A., Giunta, C, Minnear, F.L., Cho, E., and Saba, T.M.: Hepatic removal of ¹²⁵I-DLT gelatin after burn injury: a model of soluble collagenous debris that interacts with plasma fibronectin, J. Leukocyte Biol., 54, 56-64, 1993.

Brodin, B., Hesselvik, F., and von Schenck, H.: Decrease of plama fibronectin concentration following infusion of a gelatin-based plasma substitute in man, Scand. J. Clin. Lab. Invest., 44, 529-533, 1984.

Eastoe, J.E. and Leach, A.A., In: The Science and Technology of gelatin, A.G. Ward and A. Gourts (Ed.), pp. 77-85, London, Academic Press, 1977.

Fidler, I.J., General considerations for studies of experimental metastasis, Methods Cancer Res., 15, 399, 1978.

Flory, P.J. and Weaver, E.S., Helix — > coil transitions in dilute aqueous collagen solutions, J. Am. Chem. Soc, 82, 4518-4525, 1960. Forastieri, H. and Ingham, K.C., Fluid-phase interaction between human plasma fibronectin and gelatin determined by fluorescence polarization assay, Arch. Biochem. Biophys., 227, 358-366, 1983.

Forastieri, H., and Ingham, K.C., Interaction of gelatin with a fluorescein-labeled 42-kDa chymotryptic fragment of fibronectin, J. Biol. Chem., 260(19), 10546-10550, 1985.

Garcia-Pardo, A. and Gold, L.I., Further characterization of the binding of fibronectin to gelatin reveals the presence of different binding interactions, Arch. Biochem. Biophys., 304(1), 181-188, 1993.

Grinnell, F., Hays, D.G., Minter, D., Cell adhesion and spreading factor: partial purification and properties, Exp. Cell Res., 110, 175-190, 1977.

Heene, D. L. Zekorn, D., Lasch, H.G., Gelatin plasma volume expanders: chemistry, biological activities and clinical experiences, Proc. 11^th Congr. Int. Soc. Blood Transf. Sydney 1966; Bibl. No. 29, Part 3, pp. 907-913 (Karger, Basel/New York 1968).

Herman, C.J., The influence of thermal stress on the properties of phospholipid stabilized emulsions, Ph.D. thesis, University of Illinois at Chicago, Chicago, 1992.

Hunphries, M.J., The molecular basis and specificity of integrin-ligand interactions, J. Cell Sci., 97, 585-592, 1990. Hunphries, M.J., Olden, K., and Yamada, K.M., Fibronectin and cancer: implications of cell adhesion to fibronectin for tumor metastasis. In: Fibronectin in Health and Disease, ed. S.E. Carsons, pp. 161-199. Florida, CRC Press, Inc., 1989.

Humphries, M.J., Olden, K., Yamada, K.M., A synthetic peptide from fibronectin inhibits experimental metastasis of murine melanoma cells, Science, 233, 467-470, 1986.

Humphries, M.J., Yamada, K.M., Olden, K., Investigation of the biological effects of anti-cell adhesive synthetic peptides that inhibit experimental metastasis of B16-F 10 murine melanoma cells, J. Clin,. Invest., 81, 782-790, 1988.

Hynes, R.O., Integrins: versatility, modulation and signaling in cell adhesions, Cell, 69, 11-25, 1992.

Ingham, K.C., Brew, S.A. and Isaacs, B.S., Interaction of fibronectin and its gelatin-binding domains with fluorescent-labeled chains of type I collagen, J. Biol. Chem., 263(10), 4624-4628, 1988.

Iwamoto, Y., Robey, F.A., Graf. J., Saaki, M., Kleinman, H.K., Yamada, Y., and Martin, G.R., YIGSR, a synthetic laminin pentapeptide, inhibits experimental metastasis formation, Science, 238, 1132, 1987.

Jones, R.T., Gelatin: structure and manufacture. In: Hard Gelatin Development and Technologv, Ridgeway, K (ed), pp 15, Pharma. Press, London, 1987. Kang, A.H., Fails, B., and Franzblau, C, Intramolecular cross-link of chick skin collagen, Biochem. Biophys. Res. Commun., 36„ 345-349, 1969.

Kang, A.H., Nagai, Y., Piez, K.A., and Gross, J., Studies on the structure of collagen utilizing a collagenolytic enzyme from tad pole, Biochemistry, N.Y., 5, 509-515, 1966.

Klegerman, M.E., Zeunert, P.L., Lou, Y., Devadoss, P.O., Groves, M.J., Inhibition of muring sarcoma cell adherence to polystyrene substrata by Bacillus Calmette-Guerin: evidence for fibronectin-mediated direct anti-tumor activity of BCG, Cancer Invest., 11(6), 660-666, 1993.

Leao, S.C., Lopes, J.D. and Patarroyo, M.E., Immunological and functional characterization of proteins of Mycobacterium tuberculosis antigen 85 complex using synthetic peptides, J. General Microbiol., 139, 1543-1549, 1993.

Lou, Y., Olson, W.P., Tian, X.X., Klegerman, M.E., and Groves, M.J., Interaction between fibronectin-bearing surfaces and Bacillus Calmette-Guerin (BCG) or gelatin microparticles, J. Pharm. Pharmacol., 47, 177-181, 1995.

Mosher, D.F., Fibronectin, Prog. Hemostasis Thromb., 5, 111-151, 1980.

Nakamura, K., Kashiwagi, S. and Takeo, K., Characterization of the interaction between human plasma fibronectin and collagen by means of affinity electrophoresis, J. Chromat., 597, 351-356, 1992.

Nicolson, G.L., Cancer metastasis. Organ colonization and the cell-surface properties of malignant cells, Biochim. Biophys. Acta, 695, 113, 1982. Oner, F., Tian, X.X., Son, K., Klegrerman, M.E. and Groves, M.J., Characterization of the antigen 85 complex fibronectin-binding proteins derived from aged culture filtrates of Mycobacterium bovis BCG, Tice® substrain, Microbios., 78, 69-81, 1994.

Piez, K.A., Lewis, M.S., Martin, G.R., and Cross, J., Subunits of the collagen, Biochim. Biophys. Acta, 53, 596-598, 1961.

Saddler, J.M. and Horsey, P.J., The new generation gelatins: a review of their history, manufacture and properties, Anaesthesia, 42, 998-1004, 1987.

Schδne, H.H., Chemistry and physicochemical characterization of gelatin plasma substitutes, Modified Gelatin as Plasma Substitutes, Biol. haemat., No. 33, pp. 78-90 (Karger, Basel/New York 1969).

Schreier, H., Abra, R.M., Kaplan, J.E., and Hunt, C.A., Murine plasma fibronectin depletion after intravenous injection of liposomes, Intl. J. Pharm., 37, 233-238, 1987.

Steven, F.S., and Tristram, G.R., The denaturation of acetic acifd-soluble calf-skin collagen, Biochem. J., 85, 207-210, 1962.

Tomka, I., Bohonek, J., Spiihler, A., and Ribeaud, M., Structure and formation of a gelatin gel, J. Photogr. Sci., 23, 97-103, 1975.

Vedrinne, J.M., Hoen, J.P., Bussery, D., Veyssere, C, Richard, M., Motin, J., Plasma fibronectin and complement following infusion of colloidal solutions after spinal anaesthesia, Intensive Care Med., 17, 83-86, 1991. Whalen, G.F. and Ingber, D.E., Inhibition of tumor-cell attachment to extracellular matrix as a method for preventing tumor recurrence in a surgical wound, Ann Surg., 210, 758, 1989.

Yamada, K.M., and Kennedy, D.W., Dualistic nature of adhesive protein function: fibronectin and its biologically active peptide fragments can autoinhibit fibronectin function, J. Cell Biol., 99, 29-36, 1984.

Yamada, K.M., Adhesive recognition sequences, J. Biol. Chem., 266, 12809-12812, 1991.

Zekorn, D., Intravascular retention, dispersal, excretion and break-down of gelatin plasma substitutes, Modified Gelatin as Plasma Substitutes, Biol. haemat., No. 33, pp. 131-140 (Karger, Basel

Claims

We claim:

1. A process for isolating a peptide having the ability to bind fibronectin comprising:

a) chromatographically fractionating gelatin dissolved in a buffer;

b) digesting a sample of step a with a protease;

c) applying a sample of step b to a fibronectin affinity matrix; and,

d) eluting bound material from step c;

whereby a peptide having affinity for fibronectin is isolated.

2. A peptide isolated by a method of claim 1.

3. A peptide comprising the amino acid sequence Thr-Leu-Gln-Pro-Val-Tyr-Glu-Tyr-Met-Val- Gly-Val.

4. A peptide comprising the sequence Thr-Gly-Leu-Pro-Val-Gly-Val-Gly-Tyr-Val-Val-Thr- Val-Leu-Thr.

5. A process for stabilizing a peptide for administration to a patient comprising generating a complex comprising said peptide covalently linked through its N or C termini to a phosphatidylethanolamine in a sterile phospholipid-stabilized emulsion.

6. A process of claim 5 wherein said peptide comprises the sequence Thr-Leu-Gln-Pro-Val-Tyr- Glu-Tyr-Met-Val-Gly-Val linked through the N-terminal Thr residue to said phosphatidylethanolamine in an emulsion such that the complex is able to bind a surface comprising fibronectin.

7. A process of claim 5 wherein said peptide comprises the sequence Thr-Gly-Leu-Pro-Val-Gly- Val-Gly-Tyr-Val-Val-Thr-Val-Leu-Thr linked through the N-terminal Thr residue to said phosphatidylethanolamine in an emulsion such that the complex is able to bind a surface comprising fibronectin.

8. A complex comprising a peptide covalently linked through its N or C termini to a phosphatidylethanolamine in a sterile phospholipid-stabilized emulsion.

9. A complex of claim 8 wherein said peptide comprises the sequence Thr-Leu-Gln-Pro-Val- Tyr-Glu-Tyr-Met-Val-Gly-Val.

10. A complex of claim 8 wherein said peptide comprises the sequence Thr-Gly-Leu-Pro-Val- Gly-Val-Gly-Tyr-Val-Val-Thr-Val-Leu-Thr.

11. A method of preventing adhesion of a cell to fibronectin comprising coating a fibronectin- enriched surface with a composition comprising a peptide having affinity for fibronectin.

12. A method of claim 11 wherein said peptide comprises an amino acid sequence selected from the group consisting of Thr-Leu-Gln-Pro-Val-Tyr-Glu-Tyr-Met-Val-Gly-Val and Thr-Gly- Leu-Pro-Val-Gly-Val-Gly-Tyr-Val-Val-Thr-Val-Leu-Thr.

13. A method of claim 11 wherein said peptide is covalently linked through its N or C termini to a phosphatidylethanolamine in a sterile phospholipid-stabilized emulsion.

14. A method of claim 11 wherein said cell is a tumor cell.

15. A method of delivering a drug to a fibronectin-enriched surface comprising administration of a composition comprising said drug and a peptide having affinity for fibronectin covalently linked through its N or C termini to a phosphatidylethanolamine in a sterile phospholipid- stabilized emulsion.

16. A method of claim 15 wherein said peptide comprises an amino acid sequence selected from the group consisting of Thr-Leu-Gln-Pro-Val-Tyr-Glu-Tyr-Met-Val-Gly-Val and Thr-Gly- Leu-Pro-Val-Gly-Val-Gly-Tyr-Val-Val-Thr-Val-Leu-Thr.

17. A method of claim 16 wherein said fibronectin-enriched surface is a tumor cell membrane.