NO331726B1 - Fusion peptide comprising a first domain comprising parathyroid hormone (PTH) and a second domain comprising a substrate domain, set and matrix comprising the fusion peptide, and methods for producing a matrix - Google Patents

Abstract

Proteins are incorporated in protein or polysaccharide matrices for use in tissue repair, regeneration and / or reconstruction and / or drug delivery. The proteins can be incorporated to be released by matrix degradation, by enzymatic action and / or diffusion. As shown in the Examples, one method is to bind heparin to matrix by either covalent or non-covalent methods, to form a heparin matrix. The heparin then binds non-covalently heparin-binding growth factors to the protein matrix. Alternatively, a fusion protein may be constructed containing a cross-linking region such as a Factor XIIIa substrate and the native protein sequence. Incorporating degradable links between matrix and the bioactive factors can be particularly useful when long-term drug delivery is desired, e.g. in the case of nerve regeneration, where it is desirable to vary the degree of drug release spatially as a function of regeneration, e.g. fast near the interface to the living tissue and more slowly into the injury zone. Further advantages include the lower total drug dose in the delivery system, a spatial release regulation that allows a larger percentage of the drug to be released at the time of greatest cellular activity.

Description

Title: Fusion peptide comprising a first domain comprising parathyroid hormone (PTH) and a second domain comprising a substrate domain, kit and matrix comprising the fusion peptide, and method for preparing a matrix.

The invention relates to fusion peptides that contain PTH and an amino acid sequence that allows binding interactions to matrices, sets and matrices that comprise the fusion peptide and method for producing matrix with a matrix material and the fusion peptides.

Parathyroid hormone (PTH) is an 84 amino acid peptide produced and secreted by the parathyroid gland. The hormone plays a major role in the control of serum calcium levels through its action on various tissues, including bone. Human studies with different forms of PTH have shown an anabolic effect on bone, which makes them interesting for the treatment of osteoporosis and related bone disorders (US Patent No. 5,747,456 (Chorev et al.) and WO 00/10596 (Eli Lilly & Co.).Parathyroid hormone acts on cells by binding to a cell surface receptor known to be found on osteoblasts, which are the cells responsible for forming new bone.

The N-terminal 34 amino acid domain of the human hormone has been reported to be biologically equivalent to the full-length hormone. PTH 1-34 and its mode of action were first reported in US Patent No. 4,086,196. Since research has been done on PTH 1-34 and other truncated versions of the native human PTH form, such as PTH1-25, PTH1-31 and PTH1-38 (see, e.g., Rixon RH, et al., J. Bone Miner. Res., 9 (8): 1179-89 (Aug. 1994).

The mechanism by which PTH influences bone regeneration is complex, leading to conflicting results and subsequently a significant number of investigations into the exact mechanisms involved. It has been shown that if PTH is administered systemically in a continuous manner, bone density will be reduced. In contrast, it has been reported that if the same molecule is administered in a pulsatile manner, bone density will increase (see, eg, WO 99/31137 (Eli Lilly & Co.)). This apparent inconsistency can be explained by the mechanism by which PTH modulates bone remodeling and subsequently the observable parameters of bone density. In mature bone, the PTH receptor has only been shown to be present on the surface of cells of the osteoblast lineage, but not on osteoclasts. The role that PTH plays in bone regeneration is directed through the osteoblasts as opposed to the osteoclasts. However, the cells at different stages of the osteoblast lineage respond differently when bound to PTH. Therefore, the dramatic differences observed when PTH is administered using different methods can be explained by understanding the different effects that the same molecule has on different cells of the osteoblast lineage.

When PTH binds to a mesenchymal stem cell, the cell is induced to differentiate into a preosteoblast. Thus by adding PTH to the system there is an increase in the preosteoblast population. However, these preosteoblast cells also have the PTH receptor and the subsequent binding of PTH to the receptor on these cells leads to a different reaction. When PTH binds to preosteoblasts it results in two separate consequences leading to bone resorption. First, the further differentiation of preosteoblasts into osteoblasts is inhibited. Second, it increases the secretion of interleukin 6 (IL-6) from the preosteoblasts. IL-6 both inhibits preosteoblast differentiation as well as enhances preosteoclast differentiation into osteoclasts. This dual response of the cells of the osteoblast lineage is what provides the complex reaction between bone remodeling and PTH exposure. If PTH is dosed periodically for short periods of time, the mesenchymal stem cells are induced to differentiate into osteoblasts. The short dosing periods thereby prevent the newly formed preosteoblasts from producing IL-6 which prevents activation of the osteoclasts. Therefore, during the dosing intervals, these newly formed preosteoblasts can further differentiate into osteoblasts, resulting in bone formation. If a constant dose of PTH is applied, however, the preosteoblasts will have the opportunity to start producing IL-6, thus activating the osteoclasts and inhibiting themselves, which gives the opposite effect of bone resorption.

During tissue repair or regeneration, cells must migrate into a wound bed, proliferate, express matrix components or form extracellular matrix and form a final tissue shape. Multicellular cell populations must often participate in this morphogenetic response, which often includes vascular cells and nerve cells. Matrices have been demonstrated to greatly increase, and in some cases have been found to be essential for this to occur. Approaches have been made by developing matrices from natural or synthetic origins or a mixture of both. Natural cell-ingrown matrices are subjected to reconstruction by cellular influence, all based on proteolysis, i.e. by plasmin (degrade fibrin) and matrix metalloproteonases (degrade collagen, elastin, etc). Such degradation is largely localized and only occurs in direct contact with the migrating cell. In addition, the delivery of specific cell signaling proteins such as growth factors is very well regulated. In the natural model, macroporous cell-ingrowing matrices are not used, but rather microporous matrices that the cells can degrade, locally and when necessary, as the cells migrate into the matrix. Due to considerations in relation to immunogenicity, expensive production, limited availability, portion variability and purification, matrices based on synthetic precursor molecules, such as modified polyethylene glycol, have been developed for tissue regeneration in and/or on the body.

While much work has been done in studying the systemic effects of PTH, research has not examined local or topical administration of PTH. As PTH has a direct anabolic effect on the osteoblast cell line, it should have a strong potential to heal bone defects in addition to influencing bone density if it is presented in a suitable way at the site where there is a defect. Once the defect has been filled with preosteoblasts, if the PTH signal is turned off the newly formed preosteoblasts can then differentiate into osteoblasts and begin to convert the wound bed, first into woven bone tissue and then into a mature bone structure.

It is therefore an aim of the present invention to provide PTH in a form that can be bound to a matrix for tissue repair, regeneration and reconstruction.

It is a further object to present PTH in a form suitable for topical or local administration to a patient to heal bone defects.

Fusion peptides containing a parathyroid hormone (PTH) in one domain and a substrate domain capable of being covalently cross-linked to a matrix in another domain, and kits containing such fusion proteins or peptides are described herein. Fusion proteins that covalently bind to natural or synthetic materials to form matrices, which can be used to heal bone defects. Optionally, all the components for formation of matrix are applied to a bone defect and matrix is formed at the application site. The fusion peptide can be incorporated into the matrix so that either the fusion peptide as a whole or only the respective PTH sequence in the first domain is released by degradation of the matrix, enzymatic and/or hydrolytic action. The fusion peptide may also contain a degradable bond between the first and second domains containing hydrolytic or enzymatic cleavage sites. In particular, the fusion peptide contains PTH in one domain, a substrate domain capable of being covalently cross-linked to a matrix in a second domain, and a degradation site between the first and second domains. Preferably, the PTH is human PTH, although PTH from other sources, such as bovine PTH, may be suitable. PTH can be PTH 1-84 (native), PTH 1-38, PTH 1-34, PTH 1-31 or PTH 1-25 or any modified or allelic version of PTH exhibiting traits ie bone formation similar to the above . The degradation site allows the rate of delivery to be varied at different locations in the matrix depending on cellular activity at that location and/or in the matrix. Additional advantages include the lower total drug dose in the delivery system, and spatial regulation of release that allows a greater percentage of the drug to be released at the time of greatest cellular activity.

The state of the art is also described in US 5171670 A, US 6136564 and US 5840837.

Brief description of the figures

Fig. 1 is a fluorescence detection chromatogram of plasmin-degraded peptide-containing fibrin gels and free peptide. Size exclusion chromatography of a degraded fibrin gel with the a.2PIi-7-ATIIIi2i-i34 peptide incorporated (-), with the same peptide freely added to the degraded fibrin gel containing incorporated peptide

(...), and free peptide alone (--), are shown. The N-terminal leucine residue was dansylated (abbreviated dL). The free peptide eluted at longer times, corresponding to a lower molecular weight, than the peptide incorporated into the fibrin gel during coagulation, which shows covalent attachment to degraded fibrin and thus covalent incorporation via the effect of factor XHIa activity. Fig. 2 is a curve of the incorporation of dLNQEQVSPLRGD (SEQ ID NO: 1) into fibrin gels with exogenous factor XIII added. When 1 U/ml was added, the level of incorporation was increased so that more than 25 mol peptide/mol fibrinogen could be obtained. Fig. 3 is a curve of incorporation of the bidomain peptide, dLNQEQVSPLRGD (SEQ ID NO: 1) into undiluted fibrin glue. Three separate sets were tested and in each case a high level of incorporation could be observed, reaching 25 mol peptide/mol fibrinogen. The concentration of exogenous peptide required for maximal incorporation was at least 5 mM, probably due to diffusion limitations in the very dense fibrin matrix that is created. The level of incorporation was very consistent, with each kit providing a similar incorporation profile.

Products and methods for hard tissue repair, regeneration or reconstruction, particularly for bone growth, using natural and synthetic matrices having PTH releasably incorporated therein are described herein. The natural matrices are biocompatible and biodegradable and can be formed in vitro or in vivo at the time of implantation. PTH can be incorporated into the matrices and retain its full bioactivity. PTH can be releasably incorporated, using techniques that provide control over how and when and to what extent PTH is released, such that the matrix can be used for tissue repair directly or indirectly, using the matrix as a controlled release agent.

Definitions

"Biomaterial" as generally used herein refers to a material intended to establish an interface with biological systems to evaluate, treat, augment or replace any tissue, organ or function of the body depending on the material either permanently or temporarily. The terms "biomaterial" and "matrix" are used synonymously herein and mean a cross-linked polymeric network which, depending on the type of matrix, can swell with water but not dissolve in water, i.e. form a hydrogel that remains in the body over a certain period of time and fulfills certain support functions for traumatized or defective hard tissue.

"PTH fusion peptide" as generally used herein refers to a peptide containing at least a first and a second domain. One domain contains a PTH (native or truncated forms, especially PTH 1-34) and the other domain contains a substrate domain to be cross-linked to a matrix. An enzymatic or hydrolytic degradation site may also be present between the first and second domains.

"Strong nucleophile" as generally used herein denotes a molecule capable of donating an electron pair to an electrophile in a polar bond-forming reaction. Preferably, the strong nucleophile is more nucleophilic than water at physiological pH. Examples of strong nucleophiles are thiols and amines.

"Conjugated unsaturated bond" as generally used herein refers to the alternation of carbon-carbon, carbon-heteroatom, or heteroatom-heteroatom multiple bonds with single bonds, or the attachment of a functional group to a macromolecule, such as a synthetic polymer or a protein. Such bonds can undergo addition reactions.

"Conjugated unsaturated group" as generally used herein refers to a molecule or a region of a molecule containing an alternation of carbon-carbon, carbon-heteroatom, or heteroatom-heteroatom multiple bonds with single bonds, which has a multiple bond that can undergo addition reactions. Examples of conjugated unsaturated groups include but are not limited to vinylsulfones, acrylates, acrylamides, quinones and vinylpyridinium, e.g. 2- or 4-vinylpyridium and itaconates.

"Synthetic precursor molecules" as generally used herein refers to molecules that do not exist in nature.

"Naturally occurring precursor components or polymers" as generally used herein refers to molecules that can be found in nature.

"Functionalize" as generally used herein refers to the modification of a molecule in a manner that results in the attachment of a functional group or residue. E.g. a molecule can be functionalized by introducing a molecule that makes the molecule a strong nucleophile or a conjugated unsaturation. Preferably, a molecule, e.g. PEG, functionalized to become a thiol, amine, acrylate or quinone. In particular, the protein can be efficiently functionalized by partial or complete reduction of disulfide bonds to form free thiols.

"Functionality" as generally used herein refers to the number of reactive sites on a molecule.

"Branch point functionality" as generally used herein refers to the number of arms extending from one point in the molecule.

"Adhesion site or cell attachment site" as generally used herein refers to a peptide sequence to which a molecule, e.g. an adhesion-promoting receptor on the surface of a cell binds. Examples of adhesion sites include, but are not limited to, the RGD sequence from fibronectin, and the YIGSR (SEQ ID NO:2) sequence from laminin. Preferably, adhesion sites are incorporated into the biomaterial by including a substrate domain that is crosslinkable to a matrix.

"Biological activity" as generally used herein refers to functional events mediated by a protein of interest. In some embodiments, this includes events assayed by measuring the interactions of a polypeptide with another polypeptide. It also includes assaying the effect that the protein of interest has on cell growth, differentiation, death, migration, attachment, interactions with other proteins, enzymatic activity, protein phosphorylation or dephosphorylation, transcription or translation.

"Sensitive biological molecule" as generally used herein denotes a molecule that is found in a cell, or in a body, or that can be used as a therapeutic agent from a cell or a body, that can react with other molecules in its presence. Examples of sensitive biological molecules include but are not limited to peptides, proteins, nucleic acids and drugs. Biomaterials can be produced in the presence of sensitive biological materials, without adversely affecting the sensitive biological materials.

"Regenerate" as generally used herein means to grow back part or all of something, such as hard tissue, especially bone.

"Multifunctional" as generally used herein denotes more than one electrophilic and/or nucleophilic functional group per molecule (ie monomer, oligo- and polymer).

"Self-selective reaction" as generally used herein means that the first precursor component of a composition reacts much more rapidly with the second precursor component of the composition and vice versa than with other compounds present in a mixture or at the reaction site. As used herein, the nucleophile is preferably attached to an electrophile and an electrophile is preferably attached to a strong nucleophile rather than to other biological compounds.

"Cross-linking" as generally used herein means the formation of covalent bonds. However, it can also denote the formation of non-covalent bonds, such as ionic bonds, or combinations of covalent and non-covalent bonds.

"Polymeric network" as generally used herein means the product of a process in which substantially all monomers, oligo- or polymers are linked by intermolecular covalent bonds through their available functional groups to result in a large molecule.

"Physiological" as generally used herein means conditions such as those found in living vertebrates. In particular, physiological conditions denote the conditions of the human body, such as temperature, pH, etc. Physiological temperatures mean in particular the temperature range between 35°C to 42°C, preferably around 37°C.

"Crosslink density" as generally used herein refers to the average molecular weight between two crosslinks (Mc) of the respective molecules.

"Equivalent weight" as generally used herein denotes mmol of functional group/g of substance.

"Swelling" as generally used herein denotes an increase in volume and mass due to absorption of water by the biomaterial. The terms "water absorption" and "swelling" are used synonymously in the present application.

"Equilibrium state" as generally used herein is the state in which a hydrogel undergoes no mass gain or loss when stored under constant conditions in water.

I. Matrices and PTH

A. Matrix materials

Matrix is formed by cross-linking ionically, covalently or by combinations thereof precursor molecules into a polymeric network or by swelling one or more polymeric materials, i.e. matrices, to form a polymeric network having sufficient interpolymeric spaces to allow ingrowth or migration into in the matrix of cells.

In one design, the matrix is formed from proteins, preferably proteins that are naturally present in the patient into which the matrix is to be implanted. A particularly preferred matrix protein is fibrin, although matrices made from other proteins such as collagen and gelatin can also be used. Polysaccharides and glycoproteins can also be used to form a matrix. It is also possible to use synthetic polymers which can be cross-linked by ionic or covalent bonding.

Fibrin matrices

Fibrin is a natural material that has been reported for several biomedical applications. Fibrin has been described as the material for cell growth matrices in US Patent No. 6,331,422 (Hubbell et al.). Fibrin gels have been used as sealants due to its ability to bind to many tissues and its natural role in wound healing. Some specific applications include use as a sealant for vascular graft attachment, heart valve attachment, fracture bone positioning, and tendon repair (Sierra, D.H., Journal of Biomaterials Applications, 7:309-352 (1993)). In addition, these gels have been used as drug delivery devices and for neuronal regeneration (Williams et al., Journal of Comparative Neurobiology, 264, 284-290 (1987)). Although fibrin provides a firm support for tissue regeneration and cell growth, there are some active sequences in the monomer that directly enhance these processes.

The process by which fibrinogen is polymerized to fibrin has also been characterized. Initially, a protease cleaves the dimeric fibrinogen molecule at two symmetrical sites. There are several possible proteases that can cleave fibrinogen, including thrombin, reptilase, and protease III, each of which destroys the protein at a different site (Francis et al., Blood Cells, 19:291-307, 1993). When the fibrinogen is cleaved, a cell polymerization step occurs in which the fibrinogen monomers come together and form a non-covalently cross-linked polymer gel (Sierra, 1993). This self-assembly occurs because binding sites are exposed after protease cleavage has occurred. When exposed, these binding sites in the center of the molecule can bind to other sites on the fibrogen chains, which are present on the ends of the peptide chains (Stryer, L. In Biochemistry, W.H. Freeman & Company, NY, 1975). In this way, a polymeric network is formed. Factor XHIa, a transglutaminase activated from factor XIII by thrombin proteolysis can then covalently cross-link the polymer network. Other transglutaminases exist and may also be involved in covalent cross-linking and grafting to the fibrin network.

Once a cross-linked fibrin gel is formed, subsequent degradation is tightly controlled. One of the key molecules in controlling the degradation of fibrin is α2-plasmin inhibitor (Aoki, N., Progress in Cardiovascular Disease, 21: 267-286, 1979). This molecule acts on cross-linking of the α-chain in fibrin through the action of factor XHIa (Sakata, et al., Journal of Clinical Investigation, 65: 290-297, 1980). By attaching itself to the gel, a high concentration of inhibitor can be localized to the gel. The inhibitor then acts by preventing the binding of plasminogen to fibrin (Aoki et al., Thrombosis and Haemostasis, 39:22-31, 1978) and inactivating plasmin (Aoki, 1979). The α2-plasmin inhibitor contains a glutamine substrate. The exact sequence has been identified as NQEQVSPL (SEQ ID NO: 12), where the first glutamine is the active amino acid for cross-linking.

It has been shown that bidomain peptides containing a factor XHIa substrate sequence and a bioactive peptide sequence can be cross-linked into fibrin gels and that this bioactive peptide retains its cellular activity in vitro (Schense, J.C., et al. (1999) Bioconj. Chem. 10: 75-81).

Synthetic matrices

Crosslinking reactions to form synthetic matrices for use in the body include (i) free radical polymerization between two or more precursors containing unsaturated double bonds, as described in Hern et al., J. Biomed. Mater. Res. 39:266-276 (1998), (ii) nucleophilic substitution reaction such as e.g. between a precursor including an amine group and a precursor including a succinimidyl group as described in US Patent No. 5,874,500 to Rhee et al., (iii) condensation and addition reactions and (iv) Michael-type addition reaction between a strong nucleophile and a conjugated unsaturated group or bond (as a strong electrophile). Particularly preferred is the reaction between a precursor molecule having a thiol or amine group as the nucleophilic group and a precursor molecule including acrylate or vinyl sulfone group as electrophilic groups. Most preferred as the nucleophilic group is the thiol group. Michael-type addition reactions are described in WO 00/44808 (Hubbell et al.). Michael-type addition reactions allow in situ cross-linking of at least one first and one second precursor component under physiological conditions in a self-selective manner, even in the presence of sensitive biological materials. When one of the precursor components has a functionality of at least two, and at least one of the other precursor components has a functionality greater than 2, the system will self-selectively react to form a cross-linked three-dimensional biomaterial.

Preferably, the conjugated unsaturated groups or conjugated unsaturated bonds are acrylates, vinylsulfones, methacrylates, acrylamides, methacrylamides, acrylonitriles, vinylsulfones, 2- or 4-vinylpyridinium, maleimides or quinones.

The nucleophilic groups are preferably thiol groups, amino groups or hydroxyl groups. Thiol groups are mainly more reactive than unprotonated amine groups. As previously indicated, pH is an important value in the present assessment: the deprotonated thiol is mainly more reactive than the protonated thiol. Therefore, addition reactions involving a conjugated unsaturation, such as an acrylate or quinone, with a thiol to convert two precursor components in a matrix will often be best performed most rapidly and self-selectively at a pH of about 8. At a pH of approx. 8, most thiols of interest are deprotonated (and thus more reactive) and most amines of interest are still protonated (and thus less reactive). When a thiol is used as the first precursor molecule, a conjugate structure that is selective in its reactivity for the thiols relative to amines is highly desirable.

Suitable first and second precursor molecules include proteins, peptides, polyoxyalkylenes, poly(vinyl alcohol), poly(ethylene-co-vinyl alcohol), poly(acrylic acid), poly(ethylene-co-acrylic acid), poly(ethyloxazoline), poly(vinylpyrrolidone), poly(ethylene-co-vinylpyrrolidone), poly(maleic acid), poly(ethylene-co-maleic acid), poly(acrylamide), and poly(ethylene oxide)-co-poly(propylene oxide) block copolymers. A particularly preferred precursor molecule is polyethylene glycol.

Polyethylene glycol (PEG) provides a suitable building block. One can easily purchase or synthesize linear (by which is meant two ends) or branched (by which is meant more than two ends) PEG and then functionalize PEG end groups to introduce either a strong nucleophile, such as a thiol, or a conjugated structure , such as an acrylate or a vinyl sulfone. When these components are either mixed with each other or with a corresponding component in a slightly basic environment, a matrix will be formed by reaction between the first and second precursor components. A PEG component can be reacted with a non-PEG component, and the molecular weight or hydrophilicity of each compound can be controlled to manipulate the mechanical properties, permeability and water content of the resulting biomaterial.

These materials are generally useful in medical implants, as described in more detail below. In the formation of matrices, especially matrices that are desirable to be degraded in vivo, peptides provide a very suitable building block. It is easy to synthesize peptides containing two or more cysteine residues and this component can then easily serve as the first precursor component with nucleophilic groups. E.g. will a peptide with two free cysteine residues readily form a matrix when mixed with a PEG-tri-vinyl sulfone (a PEG that has three arms with vinyl sulfones on each of its arms) at physiological or slightly higher pH (e.g. 8-9 ). Gelation can also proceed well at even higher pH, but at the potential cost of self-selectivity. When the two liquid precursor components are mixed together, they react over a period of a few minutes to form an elastic gel, which consists of a network of PEG chains, which carry the knots of the network with the peptides as connecting links. The peptides can be selected as protease substrates, so that a network is formed that is capable of being infiltrated and degraded by cells, as is done in a protein-based network, such as in a fibrin matrix. Preferably, the sequences in the domains are substrates for the enzymes involved in cell migration (eg as the substrate for enzymes such as collagenase, plasmin, metalloproteinase (MMP) or elastase), although suitable domains should not be limited to these sequences. A particularly useful sequence is a substrate for the enzyme plasmin (see examples). Degradation properties of the gels can be manipulated by changing the details of the peptide that serve as cross-linking nodes. You can make a gel that is degradable by collagenase but not plasmin, or by plasmin but not collagenase. Furthermore, it is possible to make the gel degrade faster or slower in response to such an enzyme, simply by changing the amino acid sequence so that Km or kcat, or both in the enzymatic reaction are changed. One can thus create a biomaterial that is biomimetic, in that it is capable of being rebuilt with the normal rebuilding properties of the cells. E.g. Such a study shows substrate sites for the important protease plasmin. Gelation of PEG with the peptide is self-selective.

Optionally, biofunctional agents can be incorporated into the matrix to provide chemical binding to other species (eg tissue surface). Having protease substrates incorporated into the matrix is important when the matrix is formed from PEG-vinyl sulfone. In addition to matrices formed from the reaction of PEG acrylates and PEG thiols, matrices formed from PEG vinyl sulfones and PEG thiols do not contain hydrolytically degradable bonds. Therefore, incorporation of protease substrates allows the matrix to degrade in the body.

The synthetic matrices are operationally easy to form. Two liquid precursors are mixed; one precursor contains a precursor molecule with nucleophilic groups and the other precursor molecule contains the electrophilic groups. Physiological saline can serve as a solvent. Minimal heat is generated by the reaction. Therefore, the gelation can be carried out in vivo or in vitro, in direct contact with tissue, without adverse toxicity. Thus, polymers other than PEG can be used, either telegeletically modified or modified on their side groups.

For most healing indications, the rate of cell ingrowth or migration of cells into the matrix in combination with an appropriate degradation rate of the matrix is essential to the overall healing response. The potential of hydrolytic non-degradable matrices to be invaded by cells is primarily a function of network density. If the existing space between the branch points or nodes is too small in relation to the size of the cells or if the rate of matrix degradation, which results in the formation of more space in the matrix, is too slow, a very limited healing response will be observed. Healing matrices found in nature, such as e.g. fibrin matrices, which are formed as a response to injury in the body, are known to consist of a very loose network that can be easily invaded by cells. The infiltration is promoted by ligands for cell adhesion which are an integral part of the fibrin network.

Matrices prepared from synthetic hydrophilic precursor molecules, such as polyethylene glycol, swell in aqueous environments after formation of the polymeric network. In order to achieve a sufficiently short gelation time (between 3 and 10 min. at a pH of between 7 and 8 and a temperature in the range 36-38°C) and quantitative reaction during in situ matrix formation in the body, the initial concentration of the precursor molecules must be sufficiently high. Under such conditions, swelling after network formation would not take place, and the required starting concentrations would lead to matrices that were far too dense for cell infiltration when the matrix is not degradable in an aqueous environment. Thus, swelling of the polymeric network is important to enlarge and expand the space between the branch points.

Regardless of the initial concentrations of the precursor molecules, hydrogels made from the same synthetic precursor molecules, such as a four-armed PEG-vinylsulfone and a peptide with SH groups, swell to the same water content in the equilibrium state. This means that the higher the initial concentration of the precursor molecule, the higher the final volume of the hydrogel when it reaches the equilibrium state. If the space available in the body is far too small to allow sufficient swelling and especially if the bond formed from the precursor components is not hydrolytically degradable, the rate of cell infiltration and the healing response will be reduced. As a consequence, the optimum between two opposing requirements for application in the body must be found. Good cell infiltration and subsequent healing responses have been observed with a three-dimensional polymeric network formed from the reaction of a trifunctional branched polymer with at least three arms substantially equal in molecular weight and a second precursor molecule that is at least a bifunctional molecule. The ratio between the equivalent weight of the functional groups of the first and second precursor molecules is between 0.9 and 1.1. The molecular weights of the arms of the first precursor molecule, the molecular weight of the second precursor molecule and the functionality of the branching points are chosen so that the water content of the resulting polymeric network is between the equilibrium percentage and 92% by weight of the total weight of the polymeric network after the water uptake has ended. Termination of water uptake can be achieved either when the equilibrium concentration is reached or when the space available in the biomaterial does not allow further volume increase. It is therefore preferred to choose starting concentrations for the precursor components to be as low as possible. This is important for all swellable matrices, but especially for those matrices which undergo cell-mediated degradation and do not contain hydrolytically degradable links in the polymeric network.

The balance between gelation time and low initial concentration, especially for hydrolytically non-degradable gels, will be optimized based on the structure of the precursor molecules. In particular, the molecular weight of the arms of the first precursor molecule, the molecular weight of the second precursor molecule and the degree of branching, i.e. the functionality of the branching points, must be adjusted accordingly. The real reaction mechanism has little influence on this interlude.

If the first precursor molecule is a three- or four-armed polymer with a functional group at the end of each arm and if the second precursor molecule is a linear bifunctional molecule, preferably a peptide containing at least two cysteine groups, then the molecular weight of the arms of the first precursor molecule is and the molecular weight of the second precursor molecule preferably chosen so that the links between the branch points after formation of the network have a molecular weight in the range of between 10 and 13 kD (under the conditions that the links are linear, not branched), preferably between 11 and 12 kD. This allows an initial concentration of the sum of the first and second precursor molecules in a range of between 8-12% by weight, preferably between 9 and 10% by weight of the total weight of the first and second precursor molecules in solution (before network formation). In the event that the degree of branching of the first precursor molecule is increased to 8 and the second precursor molecule is still a linear bifunctional molecule, the molecular weight of the bond between the branch points is preferably increased to a molecular weight of between 18 and 24 kDa. In the case of the degree of branching of the second precursor molecule being increased from linear to a three- or four-armed precursor component, the molecular weight, i.e. the length of the bonds, increases accordingly. In a preferred embodiment of the present invention, a composition is chosen which includes as the first precursor molecule a trifunctional three-armed 15 kD polymer, i.e. each arm has a molecular weight of 5 kD, and as the second precursor molecule a bifunctional linear molecule with a molecular weight in the range from between 0.5-1.5 kD, more preferably around 1 kD. Preferably, the first and second precursor components are a polyethylene glycol.

In a preferred embodiment, the first precursor component includes as functional groups conjugated unsaturated groups or bonds, most preferably an acrylate or a vinyl sulfone and the functional groups of the second precursor molecule include a nucleophilic group, preferably a thiol or amino groups. In another preferred embodiment of the present invention, the first precursor molecule is a four-armed 20 kD

(each arm a molecular weight of 5 kDa) polymer which has functional groups at the end of each arm and the second precursor molecule is a bifunctional linear molecule with a molecular weight in the range of between 1-3 kD, preferably between 1.5 and 2 kD. Preferably, the first precursor molecule is a polyethylene glycol having vinyl sulfone groups and the second precursor molecule is a peptide having cysteine groups. In both preferred embodiments, the initial concentration of the sum of the first and second precursor molecules varies from 8-11% by weight, preferably between 9 and 10% by weight of the total weight of the first and second precursor molecules and water (before the formation of polymeric networks), preferably between 5 and 8% by weight % to achieve a gelation time of less than 10 min. These compositions have a gelation time at pH 8.0 and 37°C of approx. 3-10 min. after mixing.

When the matrix contains hydrolytically degradable links, formed e.g. in the preferred reactions between acrylates and thiols, network density with respect to cell infiltration is particularly important initially, but in aqueous environments the linkages will hydrolyze and the network will become looser, to allow cell infiltration. With an increase in the total degree of branching of the polymeric network, the molecular weight of the intermediate links, i.e. the length of the links, must increase.

B. Cell attachment sites

Cells interact with their environment through protein-protein, protein-oligosaccharide and protein-polysaccharide interactions on the cell surface. Extracellular matrix proteins provide a host with bioactive signals to the cell. This dense network is necessary to subdivide the cells; and many matrix proteins have been shown to control cell adhesion, proliferation, migration and differentiation (Carey, Annual Review of Physiology, 53:161-177, 1991). Some of the specific proteins that have been shown to be particularly active include laminin, vitronectin, fibronectin, fibrin, fibrinogen and collagen (Lander, Journal of Trends in Neurological Science, 12:189-195, 1989). Many studies of laminin have been carried out and it has been shown that laminin plays a vital role in the development and regeneration of nerves in vivo and in neurons in vitro (Williams, Neurochemical Research, 12:851-869, 1987), as well as in angiogenesis.

Some of the specific sequences that directly interact with cellular receptors and cause either adhesion, spreading or signal transduction have been identified.

Laminin, a large multidomain protein (Martin, Annual Review of Cellular Biology, 3:57-85, 1987) has been shown to consist of three chains with multiple receptor binding domains. These receptor binding domains include the YIGSR (SEQ ID NO:2) sequence of laminin B1 chain (Graf et al., Cell, 48:989-996, 1987; Kleineman et al., Archives of Biochemistry andBiophysics, 272:39-45 , 1989; and Massia et al., J. of Biol. Chem., 268:8053-8059, 1993), LRGDN (SEQ ID NO: 3) to laminin A chain (Ignatius et al.,./, of Cell Biology, 111:709-720, 1990) and PDGSR (SEQ ID NO: 4) to laminin B1 chain (Kleinman et al., 1989). Several other recognition sequences for cells have also been identified. These include IKVAV (SEQ ID NO: 5) of the laminin A chain (Tashiro, et al., J. of Biol. Chem., 264:16174-16182, 1989) and the sequence RNIAEIIKDI (SEQ ID NO: 6) to the laminin B2 chain (Liesi, et al., FEBS Letters, 244:141-148, 1989). The receptors that bind to these specific sequences have also often been identified. A subset of cellular receptors that has been shown to be responsible for much of the binding is the integrin superfamily (Rouslahti, E., J. of Clin. Investigation, 87:1-5, 1991). Integrins are protein heterodimers consisting of α and β subunits. Previous work has shown that the tripeptide RGD binds to several pi and P3 integrins (Hynes, R.O., Cell, 69:1-25, 1992; Yamada, KM., J. of Biol. Chem., 266:12809-12812, 1991) , IKVAV (SEQ ID NO: 5) binds to a 110 kDa receptor (Tashiro, et al., J. of Biol. Chem., 264: 16174-16182, 1989); Luckenbill-Edds, et al., Cell Tissue Research, 279:371-377, 1995), YIGSR (SEQ ID NO: 2) binds to a 67 kDa receptor (Graf, et al., 1987) and DGEA (SEQ ID NO: 2) ID No. 7), a collagen sequence binds to α2, Pi integrin (Zutter & Santaro, Amer. J. of Pathology, 137:113-120, 1990). The receptor for the RNIAEIIKDI (SEQ ID NO: 6) sequence has not been reported.

In a further preferred design, the peptide sites for cell adhesion are incorporated in the matrix, namely the peptides that bind to adhesion-promoting receptors on the cell surfaces in the biomaterials of the present invention. Such adhesion-promoting peptides can be selected from the group as described above. Particularly preferred is the RGD sequence from fibronectin, the YIGSR (SEQ ID NO: 2) sequence from laminin. Incorporation of cell attachment sites is a particularly preferred design with synthetic matrices, but may also be included with some of the natural matrices. The incorporation can be done e.g. simply by mixing a cysteine-containing cell attachment peptide with the precursor molecule including the conjugated unsaturated group, such as PEG-acrylate, PEG-acrylamide, or PEG-vinylsulfone a few minutes before mixing with the remaining precursor component that includes the nucleophilic group, such as a thioline-containing precursor component. If the cell attachment site does not include a cysteine it can be chemically synthesized to include one. During this first step, the adhesion-promoting peptide will be incorporated at one end of the precursor mostly functionalized with a conjugated unsaturation; when the remaining multithiol is added to the system, a cross-linked network will form. Another important implication of the manner in which the network is produced here is the efficiency of incorporation of pendant bioactive ligands such as adhesion signals. This step must be quantitative since e.g. unbound ligands (eg adhesion sites) could inhibit the interaction of cells with matrix. As described later, derivatization of the precursor with such pedantic oligopeptides is carried out in a first step in stoichiometrically large excess (minimum: 40-fold) of multi-armed electrophilic precursors over thiols and are therefore definitively quantitative. Apart from preventing unwanted inhibition, this implementation is biologically even more significant: cell behavior is extremely sensitive to small changes in ligand densities and precise knowledge of incorporated ligands helps to engineer and understand cell-matrix interactions. In summary, the concentration of adhesion sites covalently bound in the matrix has a significant influence on the degree of cell infiltration. E.g. for a given hydrogel, an RGD concentration range can be incorporated into the matrix that supports cell growth and cell migration in an optimal way. The optimal concentration range for adhesion sites such as RGD is between 0.04 and 0.05 mM and even more preferably 0.05 mM especially for a matrix that has a water content between equilibrium concentration and 92% by weight after completion of water uptake.

Excellent bone healing results can be achieved by keeping the rate of cell migration and the rate of matrix degradation constant. With regard to matrix construction (especially PTH 1-34 covalently bound), a four-armed polyethylene glycol with a molecular weight of approx. 20000 Da cross-linked with a protease degradation site CGRPQGIWGQDRC (SEQ ID NO: 8) and 0.050 mM GRGDSP (SEQ ID NO: 9) particularly good cell growth results and bone defect healing. The starting concentration for PEG and peptide below 10% by weight of the total weight of the molecules and water (before swelling). The gels have an unsealable consistency and allow osteoblasts and precursor cells to easily infiltrate the matrix.

The matrix material is preferably biodegradable by naturally occurring enzymes. The rate of degradation can be manipulated by the degree of crosslinking and inclusion of protease inhibitors in the matrix.

C. Crosslinkable substrate domains

The PTH fusion peptide can be cross-linked and covalently attached to matrices through the cross-linkable substrate domain of the PTH fusion peptide. The type of substrate domain depends on the type of matrix. For incorporation into fibrin matrices, transglutaminase substrate domains are particularly preferred. The transglutaminase substrate domain may be a Factor XHIa substrate domain. This factor XHIa substrate domain may include GAKDV (SEQ ID NO: 10), KKKK (SEQ ID NO: 11), or NQEQVSPL (SEQ ID NO: 12). The linkage between PTH and the transglutaminase substrate domain can be accomplished by chemical synthesis.

The transglutaminase substrate domain may be a substrate for a transglutaminase other than Factor XHIa. The most preferred Factor XHIa substrate domain has an amino acid sequence such as NQEQVSPL (SEQ ID NO: 12) (herein referred to as "TG"). Other proteins that transglutaminase recognizes, such as fibronectin, could be linked to the transglutaminase substrate peptide.

For incorporation of PTH into a matrix formed from synthetic precursor components, the PTH fusion peptide or any other peptide to be incorporated must be synthesized with at least one additional cysteine group (-SH) preferably in the N-terminus of PTH as the crosslinkable substrate domain. The cysteine can either be directly attached to PTH or through a linker sequence. The linking sequence can also include an enzymatically degradable amino acid sequence, so that PTH can be cleaved from the matrix by enzymes in mainly native form. The free cysteine group reacts with the conjugated unsaturated group of the precursor component in a Michael-type addition reaction. In the case of PTH 1-34, the binding to a synthetic matrix for PTH 1-34 is made possible by attaching an additional amino acid sequence to the N-terminus of PTHi-34 containing at least one cysteine. The thiol group of the cysteine can react with a conjugated unsaturated bond on the synthetic polymer to form a covalent bond. In option (a) only a cysteine is attached to the peptide, in option (b) an enzymatically degradable, especially a plasmin degradable sequence is attached as a link between the cysteine and the peptide. The sequence GYKNR (SEQ ID NO: 15) between the first domain and the second domain, the cysteine, makes the linker plasmin degradable.

Thus, PTH fusion peptides can be further modified to contain a degradable site between the attachment site, i.e. the second domain (i.e. the Factor XHIa substrate domain or cysteine) and PTH, i.e. the first domain. These sites may be degradable either by non-specific hydrolysis (ie, an ester bond) or they may be substrates for specific enzymatic (either proteolytic or polysaccharide-degrading) degradation. These degradable sites allow the engineering of more specific release of PTH from matrices such as fibrin gels. E.g. degradation based on enzymatic activity allows the release of PTH to be controlled by a cellular process rather than by diffusion of the factor through the gel. The degradable site or bond is cleaved by the enzyme released from the cells that invaded the matrix.

The degradation sites allow PTH to be released with little or no modification to the primary peptide sequence, which may result in higher activity of the factor. In addition, it allows the release of the factor to be controlled by cell-specific processes, such as localized proteolysis rather than diffusion from some porous materials. This allows factors to be released at different rates in the same material depending on the location of cells in the material. This also reduces the amount of total PTH needed, since its release is controlled by cellular processes. Conservation of PTH and its bioavailability are clear advantages of exploiting cell proteolytic activity over the use of diffusion-controlled release devices. In one possible explanation for the strong healing of a bone defect with PTH covalently bound to a matrix, it is considered important that PTH is administered locally over an extended period of time (not just a simple pulsed dose) but not in a continuous manner. This is carried out by a slow degradation, through either enzymatic cleavage or hydrolytic cleavage of the matrix. In this way, the molecule is delivered through a pseudo-pulsed effect that occurs over an extended period of time. When a precursor cell infiltrates the matrix, it will encounter a PTH molecule and can differentiate into a preosteoblast. However, if that particular cell does not continue to release bound PTH from the matrix, it will effectively convert to an osteoblast and begin producing bone matrix. Finally, the therapeutic effects of peptides are localized to the defective region and are then enhanced.

Enzymes that could be used for proteolytic degradation are numerous. Proteolytic degradable sites could include substrates for collagenase, plasmin, elastase, stromelysin or plasminogen activators. Examples of substrates are listed below. P1-P5 indicate amino acids 1-5 positions towards the amino terminus in the protein from the site where proteolysis occurs. Pl'-P4' indicate amino acids 1-4 positions towards the carboxy terminus of the protein from the site where proteolysis occurs.

In another preferred design, an oligoester domain could be inserted between the first and second domains. This could be done by using an oligoester such as oligomers of lactic acid.

Non-enzymatic degradation substrates can consist of any bond that undergoes hydrolysis by an acid- or base-catalyzed mechanism. These substrates may include oligoesters such as oligomers of lactic acid or glycolic acid. The rate of degradation of these materials can be controlled through the choice of oligomer.

D. PTH

The term "PTH" as used herein includes the human sequence of PTH 1-84 and all truncated, modified and allelic versions of PTH that exhibit bone formation properties when covalently bound to biodegradable natural or synthetic matrices. Preferred truncated versions of PTH are PTH 1-38, PTH 1-34, PTH 1-31 or PTH 1-25. Most preferred is PTH 1-34. Preferably, the PTH is human PTH, although PTH from other sources, such as bovine PTH, may be suitable.

Methods for incorporating and/or releasing bioactive factors

In one preferred embodiment for incorporating a PTH into a matrix, the matrix includes fibrin formed from fibrinogen, a calcium source and thrombin, and the PTH fusion peptide will be incorporated into the fibrin during coagulation. PTH fusion peptide is constructed as a fusion peptide that includes two domains, a first and a second, of which one domain, the second, is a substrate for a cross-linking enzyme such as Factor XHIa. Factor XHIa is a transglutaminase that is active during coagulation. This enzyme, formed naturally from factor XIII by cleavage with thrombin, functions so that it attaches fibrin chains to each other via amide bonds, formed between glutamine side chains and lysine side chains. The enzyme also functions by attaching other peptides to fibrin during coagulation, e.g. cell attachment sites provided they also include a factor XHIa. Specifically, the sequence NQEQVSP (SEQ ID NO: 16) has been shown to function as an efficient substrate for Factor XHIa. As described herein before, it is either directly linked to PTH or it may include a degradation site between PTH (first domain) and the NQEQVSP (SEQ ID NO: 16) sequence (second domain). As such, the PTH fusion peptide can be incorporated into the fibrin during coagulation via a Factor XHIa substrate.

Construction of fusion proteins for incorporation

PTH fusion peptides that include a first domain that includes PTH, a second domain that includes a substrate domain for a cross-linking enzyme, and optionally a degradation site between the first and second domains can be incorporated into the fibrin gels using several different schemes. Preferably, the second domain includes a transglutaminase substrate domain and even more preferably it includes a Factor XHIa substrate domain. Most preferably, Factor XHIa includes the substrate domain NQEQVSP (SEQ ID NO: 16). When this PTH fusion peptide is present during the polymerization of fibrinogen, i.e. during the formation of fibrin matrix, it is directly incorporated into the matrix.

The degradation site between the first and second domains of the PTH fusion peptide may be an enzymatic degradation site as described previously. Preferably, the degradation site is cleavable with an enzyme selected from the group consisting of plasmin and matrix metalloproteinase. By careful selection of the Km and kcat for this enzymatic degradation site, degradation could be controlled to occur either before or after the protein matrix and/or by utilizing the same or different enzymes to degrade the matrix, with the location of the degradation site tailored to each type of protein and application . This PTH fusion peptide could be directly cross-linked into fibrin matrix as described above. However, incorporation of an enzymatic degradation site alters the release of PTH during proteolysis. When the cell-derived proteases reach the sequenced fusion peptide they can cleave the engineered protein and the newly formed degradation site. The resulting degradation product will include the released PTH, which will now be nearly free of any engineered fusion sequence, as well as any degradable fibrin.

II. Procedures for application

The matrices can be used for tissue repair, regeneration or reconstruction, and/or release of PTH, before or at the time of implantation. In some cases, it will be desirable to induce cross-linking at the site of administration in order to conform the matrix to the tissue at the site of implantation. In other cases, it will be suitable to prepare a matrix before implantation.

Cells can also be added to the matrix before or at the same time as implantation, or even subsequent implantation, either by or subsequent cross-linking of the polymer to form the matrix. This may be in addition to or instead of matrix cross-linking to produce interstitial spaces designed to promote cell proliferation or ingrowth.

Although in most cases it will be desirable to implant matrix to promote cell growth or proliferation, in some cases the bioactive factors will be used to inhibit the rate of cell proliferation. A specific application is to inhibit the formation of adhesions after surgery.

III. Application procedure

In the preferred design, the material is gelled in situ in or on the body. In another design, the matrix can be formed outside the body and then applied in the preformed shape. The matrix material can be prepared from synthetic or natural precursor components. Regardless of the type of precursor component used, the precursor components should be separated prior to application of the mixture to the body to prevent combination or contact with each other under conditions that allow polymerization or gelation of the components. To prevent contact before administration, a kit that separates the compositions from each other can be used. By mixing under conditions that allow polymerization, the compositions form a bioactive factor supplemented three-dimensional network. Depending on the precursor components and their concentrations, gelation can occur almost instantaneously after mixing. Such rapid gelation makes injection, i.e. squeezing the gelled material through an injection needle, almost impossible.

In one embodiment, matrix is formed from fibrinogen. Fibrinogen is gelled to form a matrix through a cascade of different reactions when brought into contact with thrombin and a calcium source at a suitable temperature and pH. The three components, fibrin, thrombin and the calcium source, should be stored separately. However, as long as at least one of the three components is kept separate, the other two components can be combined before administration.

In a first embodiment, fibrinogen is dissolved (which may contain additional aprotinin to increase stability) in a buffer solution at physiological pH (in a range from pH 6.5-8.0, preferably from pH 7.0-7.5) and stored separately from a solution of thrombin in a calcium chloride buffer (eg concentration range from 40-50 mM). The buffer solution for fibrinogen can be a histidine buffer solution at a preferred concentration of 50 mM including additional NaCl at a preferred concentration of 150 mM or TRIS buffer saline (preferably at a concentration of 33 mM).

In a preferred embodiment, a kit containing a fusion protein, fibrinogen, thrombin and a calcium source is provided. Optionally, the kit may contain a cross-linking enzyme, such as Factor XHIa. The fusion protein contains a bioactive factor, a substrate domain for a crosslinking enzyme and a degradation site between the substrate domain and bioactive factor. The fusion protein can be present in either fibrinogen or thrombin solution. In a preferred embodiment, the fibrinogen solution contains the fusion protein.

The solutions are preferably mixed using a two-way syringe device, in which mixing takes place by pushing the contents of both syringes through a mixing chamber and/or needle and/or static mixing device.

In a preferred design, both fibrinogen and thrombin are stored separately in lyophilized form. Either can contain the fusion protein. Before use, tris or histidine buffer is added to the fibrinogen, the buffer may additionally contain aprotinin. The lyophilized thrombin is dissolved in the calcium chloride solution. Then, fibrinogen and thrombin solutions are placed in separate containers, ampoules, syringe bodies and mixed by a two-way coupling device, such as a two-way syringe. Optionally, the containers/ampoules/syringe bodies are two-parted to thus have two chambers separated by an adjustable separation that is perpendicular to the syringe body wall. One of the chambers contains the lyophilized fibrinogen or thrombin, while the other chamber contains a suitable buffer solution. When the plunger is pushed down, the partition moves and releases the buffer in the fibrinogen chamber to dissolve the fibrinogen. When both fibrinogen and thrombin are dissolved, both bipartite syringe bodies are attached to a two-way coupling device and the contents are mixed by forcing them through the injection needle attached to the coupling device. Optionally, the coupling device contains a static mixing device to improve the mixing of the contents.

In a preferred embodiment, the fibrinogen is diluted eightfold and the thrombin is diluted twentyfold prior to mixing. This ratio results in a gelation time of approx.

1 min.

In another preferred embodiment, the matrix is formed from synthetic precursor components capable of undergoing a Michael addition reaction. Since the nucleophilic precursor component (multithiol) only reacts with the multiacceptor component (the conjugated unsaturated group) at basic pH, the three components that must be stored separately before mixing are: the base, the nucleophilic component, and the multiacceptor component. Both the multiacceptor component and the multithio component are stored as solutions in buffers. Both compositions may include the cell attachment site and further the bioactive molecule. Thus, the first composition in the system can e.g. include the resolution of the nucleophilic component and the second composition of the system may include the resolution of the multiacceptor component. Either one or both of the two compositions may include the base. In another embodiment, the multiacceptor and multithiol may be included as a solution in the first composition and the second composition may include the base. Linking and mixing takes place in the same way as previously described for fibrinogen. The two-piece syringe body is similarly suitable for the synthetic precursor components. Instead of fibrinogen and thrombin, the multiacceptor and multithiol components are stored in powdered form in one of the chambers, the other chamber contains the basic buffer.

In summary, the present invention thus relates to a fusion peptide which is characterized in that it comprises a first domain comprising parathyroid hormone

(PTH) and a second domain comprising a substrate domain, which is covalently cross-linkable to a fibrin matrix material or to a synthetic matrix material, so that the fusion peptide can be linked to the matrix through the second domain.

The invention further relates to kits comprising the fusion peptide according to the invention.

In addition, the invention relates to a matrix which is suitable for cellular growth or ingrowth where this comprises the fusion peptide according to the invention, in which the fusion peptide is covalently bound to the matrix.

Furthermore, the invention relates to a method for producing a matrix, where the method comprises

providing at least one matrix material capable of forming a cross-linked matrix, wherein the matrix material is selected from the group consisting of proteins and synthetic materials,

adding the fusion peptide as set forth in any of claims 1-10 to the matrix material, and

crosslink the matrix material, so that the fusion peptide is connected to the matrix through the second domain.

The following examples are included to show preferred embodiments of the invention.

Example 1: Matrices containing covalently bound TGPTH.

Synthesis of TGPTH

PTH 1-34-mer peptide that shows equal activity to the entire protein, and proteins of this length can be synthesized by standard solid-state peptide synthesis methods.

All peptides were synthesized on solid resin using an automated peptide synthesizer using standard 9-fluoroenylmethyloxycarbonyl chemistry. Peptides were purified by cl8 chromatography and analyzed using reverse phase chromatography via HPLC to determine purity as well as mass spectroscopy (MALDI) to identify the molecular weight of each product. Using this method, the following peptide (hereafter referred to as "TGPTH") was synthesized: NH3-Asn-Gln-Glu-Gln-Val-Ser-Pro-Leu-Ser-Val-Ser-Glu-Ile-Gln -Leu-Met-His-Asn-Leu-Gly-Lys-His-Leu-Asn-Ser-Met-Glu-Arg-Val-Glu-Trp-Leu-Arg-Lys-Lys-Leu-Gln-Asp-Val -His-Asn-Phe-COOH (SEQ ID NO: 17).

In vivo results

The activity of TGPTH to enhance bone regeneration was tested in a TISSUCOL<®> matrix in a drilled hole defect in a sheep. Eight mm holes 12 mm deep were made in the proximal and distal femur and humerus of sheep. These holes were filled with an in situ polymerization fibrin gel. The defects were left empty, filled with TISSUCOL<®>or TGPTH added to TISSUCOL<®>fibrin at 400 ug/ml before polymerization. In each example in which TISSUCOL<®> was used, it was diluted fourfold from an available standard concentration, resulting in a fibrinogen concentration of 12.5 mg/ml.

The defects were allowed to heal for 8 weeks. After this healing process, the animals were euthanized, the bone samples were removed and analyzed by microcomputer-assisted topography (u.CT). The percentage of defect volume filled with calcified bone tissue was then determined. When the defects were left empty there was no formation of calcified tissue inside the fibrin matrix. When only a fibrin gel was added there was virtually no bone healing. However, with the addition of 400 µg/ml of TGPTH, the level of healing increased dramatically where the defect was 35% with calcified bone.

Example 2: Healing response with modified PTH 1-34 attached to a fibrin matrix

Materials

The modified version of PTH1-34 that can be incorporated into a fibrin matrix was investigated for healing response in critical size rat skull defect.

A fibrin gel was prepared from the TISSUCOL<®> kit (Baxter AG, CH-8604 Volketswil/ZH) fibrin sealant precursor components. The fibrinogen was diluted in sterile 0.03 M Tris-buffered solution (TBS, pH 7.4) to form an approx. 8 mg/ml solution and the thrombin was diluted in sterile 50 mM CaCl solution to form a 2 U/ml solution. The final concentration of fibrinogen was 1:8 original TISSUCOL<®->formulation (about 100 mg/ml) and 1:160 original TISSUCOL<®->thrombin concentration (about 500 IU/ml). A predetermined amount of TG-pl-PTHi.34 or TGPTH1.34 was then added to the thrombin, and mixed to form a homogeneous concentration.

To form the fibrin gel, the diluted precursors were mixed together by injecting the fibrinogen into the tube containing the thrombin. In the case of the sheep drill defect (as described below), this mixture was then injected immediately into a drill defect formed in sheep porous bone, where a fibrin gel formed within 1-5 min. In the first series of animal experiments, the effectiveness of a fusion protein containing PTH1.34 as the bioactive factor

(NQEQVSPL YKNRSVSEIQLMHNLGKHLNSMERVEWLRKKLQDVHNF,

SEQ. ID no. 18) in cortical bone healing tested in a small animal model. The sequence YKNR (SEQ ID NO: 19) renders the binding plasmin degradable ("TG-pl-PTHi.34). TG-pl-PTHi-34 was prepared by chemical synthesis. Purification was performed through reverse phase HPLC (C18 column) by using a TFA as counterion and which resulted in a final product which was a TFA salt The purity of TG-pl-PTHi.34 was determined to be 95%.

The healing response was examined at both short (3 weeks) and long healing times (7 weeks) to determine if an improvement in healing could be observed.

Rat cranial defect of critical size

Rats were anesthetized and the cranial bone was exposed. The periosteum on the outer surface of the skull was retracted so that it would play no role in the healing process, and a simple 8 mm round defect was created. The defect size was chosen as it has previously been determined that defects of 8 mm or larger do not spontaneously heal by themselves, and are defects of critical size. The defect was then patched with a preformed fibrin matrix and the animal was allowed to heal for 3 and 7 weeks. The defect region was then explanted and analyzed by radiology as well as histology.

Results

When TG-pl-PTHi-34 was examined at the 3-week time point, the level of healing was very similar to that observed with a fibrin matrix alone. The 3 week time point was chosen as an early time point as other potent morphogens, including rhBMP-2 showed a strong healing effect at 3 weeks. In contrast, the healing effect for TG-pl-PTHi.34 was not observable at this early time point. However, when the long time point (7 weeks) was analyzed, a moderate dose-dependent improvement in healing was observed in the critical size rat skull defect with the addition of modified PTH1.34 to the fibrin matrices. The result is shown in table 3. When high doses of modified PTH1.34 were used, the healing response increased by 65%.

These results show that when PTH1.34 is incorporated into a fibrin matrix it retains some activity as shown by the modest increase in bone formation.

Boring defect in sheep bone

TG-pl-PTHi-34 was also tested in a long bone defect model to test the effect of this hormone on bone healing. In the drilling defect model in sheep, 8 mm cylindrical drilling defects that were approx. 15 mm deep located in both the proximal and distal regions of the femur and humerus bone. Since the defect was located in the epiphysis of the long bones, the defect was surrounded by trabecular bone with a thin layer of cortical bone at the edge of the defect. These defects were then filled with an in situ polymerization fibrin (approximately 750 µl) containing different doses of TG-pl-PTHi-34 or TGPTH1-34. The animals were allowed to heal for 8 weeks and then sacrificed. The defect was analyzed with u-CT and histology.

In this series of experiments, three types of compositions were tested. First, TG-pl-PTHi-34 was tested over a large concentration range. then, another modified PTHi_34, TGPTHi_34

No. 17) was used which only had a transglutaminase sequence at the amino terminus without a degradation site. Thus, TGPTH1.34 could only be released by degradation of fibrin matrix itself. TGPTH1-34 was prepared and purified in the same manner as TG-pl-PTHi-34. The purity was determined to be 95%. TGPTH1.34 was tested at several

concentrations that were similar to those of TG-pl-PTHi.34 to compare efficacy. Finally, matrices were prepared in the presence of granular material, with either TGPTH1.34 or TG-pl-PTHi.34. The granular material was a standard tricalcium phosphate/hydroxyapatite mixture that was baked into the matrix during gelation.

The effect of adding these granules on the effectiveness of PTH 1.34 was investigated.

Results

As a control, unmodified fibrin was tested. When each of the modified PTH1.34 molecules was placed into the defect in the long bones, a significant improvement in the healing response was observed over the use of fibrin matrices (control) alone. Application of fibrin alone resulted in little healing, only 20% of the original defect was filled with newly formed bone.

TG-pl-PTHi-34 was tested in a concentration range from 20-1000 µg/ml. For each dose tested, a significant increase in the healing response was observed. For example when 100 µg/ml TG-pl-PTHi-34 was used, the healing rate was increased to almost 60%.

In a second series of experiments, TGPTH].34 was tested. Application of TGPTH].34 also increased bone healing. E.g. application of 400 µg/ml improved the healing response to 40% and 1000 µg/ml increased bone healing to 65%. Thus, the addition of each modified PTHi-34 sequence resulted in a stronger healing response than the control.

Finally, when each modified PTHi-34 molecule was bound to matrix and a granule/matrix mixture was used, the efficacy of PTH].34 was maintained. This was tested for both TG-pl-PTHi.34 (see Table 4) as well as for TGPTH1-34 (see Table 5). Histological evaluation showed high infiltration in the original defect by spindle and osteoblast precursor cells supported on an extracellular matrix. Active osteoids with large round osteoblasts were common and signs of endochondral bone formation (chondrocytes) were observed. At 8 weeks, osteoclasts and fresh signs of remodeling could be found. However, unlike the results obtained from continuous exposure to systemic PTH, no reaction from osteoclasts was observed and new bone formation was significantly greater than absorption in and around the defect area. No foreign body inflammatory reaction was detected (ie, no giant cells and only mild monocyte presence). Granules were still present in samples with added mineral particles.

Example 3: Preparation of precursor components for synthetic matrices Preparation of PEG-vinylsulfones

Commercially available branched PEGs (4-arm PEG, molecular weight 14800, 4-arm PEG, molecular weight 10000, and 8-arm PEG, molecular weight 20000; Shearwater Polymers, Huntsville, AL, USA) were functionalized on the OH ends.

PEG-vinylsulfones were prepared under an argon atmosphere by reacting a dichloromethane solution of precursor polymers (previously dried over molecular sieves) with NaH and then, after hydrogen evolution, with divinylsulfone (molar ratio: OH 1:NaH 5:divinylsulfone 50). The reaction was carried out at room temperature for 3 days under argon with constant stirring. After the neutralization of the reaction solution with concentrated acetic acid, the solution was filtered through paper until clear. The derived polymer was isolated by precipitation in ice-cold diethyl ether. The product was redissolved in dichloromethane and reprecipitated in diethyl ether (with extensive washing) twice to remove any excess divinyl sulfone. Finally, the product was dried under vacuum. The derivation was confirmed by<]>H NMR. The product showed characteristic vinyl sulfone peaks at 6.21 ppm (two hydrogens) and 6.97 ppm (one hydrogen). The degree of end group conversion was found to be 100%.

Production of PEG acrylates

PEG acrylates were prepared under an argon atmosphere by reacting an azeotropically dried toluene solution of precursor polymers with acryloyl chloride, in the presence of triethylamine (molar ratio: OH 1: acryloyl chloride 2: triethylamine 2.2). The reaction was continued with stirring overnight in the dark at room temperature. The resulting pale yellow solution was filtered through a neutral alumina bed; after evaporation of the solvent, the reaction product was dissolved in dichloromethane, washed with water, dried over sodium sulfate and precipitated in cold diethyl ether. Utilization: 88%; conversion of OH to acrylate: 100% (from<!>H NMR analysis).

<]>H NMR (CDC13): 3.6 (341H (14800 4arm: 337H theor.), 230 (10000 4arm: 227H theor.), or 210H (20000 8 arm: 227H theor.), PEG chain protons), 4.3 (t, 2H, -CH2-CH^-0-CO-CH=CH2), 5.8 (dd, 1H, CH2=CH-COO-), 6.1 and 6.4 (dd, 1H , CH^CH-COO-) ppm.

FT-IR (film on ATR plate): 2990-2790 (v C-H), 1724 (v C=0), 1460 (vsCH2), 1344, 1281, 1242, 1097 (vas C-O-C), 952, 842 (vsC- O-C) cm'<1.>

Peptide synthesis

All peptides were synthesized on solid resin using an automated peptide synthesizer (9050 Pep Plus Synthesizer, Millipore, Framingham, USA) with standard 9-fluorenylmethyloxycarbonyl chemistry. Hydrophobic "renovators"

(scavengers) and cleaved protective groups were removed by precipitation of the peptide in cold diethyl ether and dissolution in deionized water. After lyophilization, the peptides were redissolved in 0.03 M Tris-buffered saline (TBS, pH 7.0) and purified using

HPLC (Waters; Milford, USA) on a size exclusion column with TBS, pH 7.0 as running buffer.

Matrix formation by conjugate addition reactions

MMP-sensitive gels formed by conjugate addition with a peptide-linked nucleophile and a PEG-linked conjugated unsaturation allowing proteolytic cell migration.

The synthesis of gels is carried out entirely through Michael-type addition reaction of thiol-PEG on vinylsulfone-functionalized PEG. In a first step, adhesion peptides were attached pendant (e.g. peptide Ac-GCGYGRGDSPG-NH2(SEQ ID NO: 20)) to a multi-armed PEG-vinyl sulfone and then this precursor was cross-linked with a dithioline-containing peptide (e.g. MMP -the substrate Ac-GCRDGPQGIAGFDRCG-NH2 (SEQ ID NO: 21)). In a typical gel preparation for three-dimensional in vitro studies, 4-arm-PEG-vinyl sulfone (mol weight 15000) was dissolved in a TEOA buffer (0.3 M, pH 8.0) to give a 10% (w/w) solution. To make the gels cell-adhesive, the dissolved peptide Ac-GCGYGRGDSPG-NH2 (SEQ ID NO: 20) (same buffer) was added to this solution. The adhesion peptide was allowed to react for 30 min. at 37°C. Then, the crosslinker peptide Ac-GCRDGPQGIQGQDRCG-NH2 (SEQ ID NO: 21) was mixed with the above solution and the gels were synthesized. Gelation occurred within a few minutes while the cross-linking reaction was however carried out for 1 hour at 37°C to guarantee complete reaction.

MMP-insensitive gels formed by conjugate addition with a PEG-linked nucleophile and a PEG-linked conjugate unsaturation allowing non-proteolytic cell migration.

The synthesis of gels was also completely completed through Michael-type addition reaction of thiol-PEG onto vinyl sulfone-functionalized PEG. In a first step, adhesion peptides were prepared pendant (e.g. the peptide Ac-GCGYGRGDSPG-NH2(SEQ. ID no. 20)) to a multi-armed PEG-vinyl sulfone and then this precursor was cross-linked with a PEG-dithiol (m.v. 3,4 kD). In a typical gel preparation for three-dimensional in vitro studies, 4-arm-PEG-vinylsulfone (mol weight 15000) was dissolved in a TEOA buffer 0.3 M, pH 8.0) to give a 10% (w/w) solution. To make the gels cell-adhesive, the dissolved peptide AC-GCGYGRGDSPG-NH2 (SEQ ID NO: 20) (in the same buffer) was added to this solution. The adhesion peptide was allowed to react for 30 min. at 37°C. Then the PEG-dithiol precursor was mixed with the above solution and the gels were synthesized. The gelation occurred within a few minutes, while the cross-linking reaction was however carried out for 1 hour at 37°C to guarantee complete reaction.

Matrix formation by condensation reaction

MMP-sensitive gels formed by condensation reactions with a peptide X-linker containing multiple amines and an electrophilic active PEG that allows proteolytic cell migration.

MMP-sensitive hydrogels were also formed by performing a condensation reaction between MMP-sensitive oligopeptide containing 2 MMP substrates and 3 Lys (Ac-GKGPQGIAGQKGPQGIAGQKG-NH2(SEQ ID NO: 22)) and a commercially available (Shearwater polymers) difunctional double ester PEG-N-Hydroxysuccinimide (NHS-HBS-CM-PEG-CM-HBA-NHS) In a first step, an adhesion peptide (e.g. the peptide Ac-GCGYGRGDSPG-NH2) (SEQ ID NO: 20) was reacted with a small fraction of NHS-HBS-CM-PEG-CM-HBA-NHS and then this precursor was cross-linked into a network by mixing with the peptide Ac-GKGPQGIAGQKGPQGIAGQKG-NH2 (SEQ ID NO: 22) carrying three ε-amines (and one primary amine).In a typical gel preparation for 3-dimensional in vitro studies, both components were dissolved in 10 mM PBS at pH 7.4 to give a 10% (w/w) solution and hydrogels were formed in less than one hour.

In contrast to the present hydrogels formed by Michael-type reaction, the desired self-selectivity by this approach is not guaranteed since amines present in biological materials such as cells or tissues would also react with difunctionally activated double esters. This is also true for other PEGs that carry electrophilic functionalities such as PEG-oxycarbonylimidazole (CDI-PEG), or PEG-nitrophenyl carbonate.

MMP-insensitive hydrogels formed by condensation reactions with a PEG-amine cross-linker and an electrophilically active PEG that allow non-proteolytic cell migration. Hydrogels were also formed by performing a condensation reaction between commercially available branched PEG-amines (Jeffamines) and the same difunctional double ester PEG-N-hydroxysuccinimide (NHS-HBS-CM-PEG-CM-HBA-NHS). In a first step, the adhesion peptides (e.g. the peptide Ac-GCGYGRGDSPG-NH2) (SEQ ID NO: 20) were reacted with a small fraction of NHS-HBS-CM-PEG-CM-HBA-NHS and then this precursor cross-linked into a network by mixing with the multi-armed PEG-amine. In a typical gel preparation for three-dimensional in vitro studies, both components were dissolved in 10 mM PBS at pH 7.4 to give a 10% (w/w) solution and hydrogels were formed in less than 1 h.

Again in contrast to the present hydrogels formed by Michael-type reaction, the desired self-selectivity in this approach is not guaranteed since amines present in biological materials such as cells or tissues will also react with difunctional activated double esters. This is also true for other PEGs that carry electrophilic functionalities such as PEG-oxycarbonylimidazole (CDI-PEG), or PEG-nitrophenyl carbonate.

Example 4: Bone regeneration with synthetic enzymatically degradable matrices

Two different starting concentrations of the enzymatically degradable gels were used. In each of these, the concentration of RGD and the active factor (CplPTH at 100 µg/ml) was kept constant. The polymeric network was formed from a 4-armed branched PEG functionalized with four vinyl sulfone end groups with molecular weight o of 20 kD (molecular weight for each of the arms 5 kD) and dithiol peptide with the following sequence Gly-Cys-Arg-Asp-(Gly-Pro-Gln -Gly-Ile-Trp-Gly-Gln)-Asp-Arg-Cys-Gly (SEQ ID NO: 21). Both precursor components were dissolved in 0.3 M triethanolamine. The initial concentration of the functionalized PEG (first precursor molecule) and the dithiol peptide (second precursor molecule) was varied. In one case, the concentration was 12.6% by weight of the total weight of the composition (first and second precursor components plus triethanolamine solution). 12.6% by weight corresponds to a 10% by weight solution calculated on the basis of only the first precursor component. (100 mg/ml first precursor molecule). The second starting concentration was 9.5% by weight of the total weight of the composition (first and second precursor component + triethanolamine solution) which corresponds to 7.5% by weight on the basis of only the first precursor molecule (75 mg/ml first precursor molecule) of the total weight. This has the consequence that the amount of dithiol peptide was changed so that the molar ratio between vinyl sulphones and thiols was maintained.

The gel, which started from an initial concentration of 12.6% by weight, swelled to a concentration of 8.9% by weight of the total weight of the polymeric network + water, thus the matrix had a water content of 91.1. The gel which started from an initial concentration of 9.5% by weight swelled to a final concentration of 7.4% by weight of the total weight of the polymeric network plus water, and thus had a water content of 92.6.

To explore the impact of this change, these materials were tested in the sheep bore defect. Here, a 750 ul defect was placed in the porous bone of the diaphyses of the sheep femur and humerus and filled with an in situ relating enzymatic gel. The following amount of calcified tissue was obtained, determined by ^CT, in which each group was N=2:

By making the gels less dense and easier for cell penetration, the resulting healing reaction with the addition of an active factor was stronger. The effect of having final solids concentrations below 8.5% by weight is clear from these results. It is therefore clear that the construction of the matrix is essential to allow healing of wound defects. Each of these hydrogels was composed of large chains of polyethylene glycol, end-linked to form a matrix. However, the details of how they were linked, via enzymatic degradation sites, the density of the linkers and several other variables were essential to allow a functional healing response. These differences very clearly observed in the sheep borer defect model.

Example 5: Bone Formation with Synthetic Hydrolytically Degradable Matrices PTH 1-34 fusion peptide was tested in a synthetic gel as well as in the sheep bore defect model exactly as described for the fibrin matrices. A hydrogel network was formed by mixing together acrylated 4-arm polyethylene glycol, MV 15000 (Peg 4<*>15Acr) with a linear polyethylene glycol dithiol of MV 3500. Through a Michael-type reaction, where the two components are mixed in a buffer of 0 .3 M triethanolamine at pH 8.0, the resulting thiolates that are formed at this pH can then react with the conjugated unsaturation of the acrylate to form a covalent bond. By mixing together multifunctional precursors such that the combined multifunctionality is greater than or equal to 5, a hydrogel is formed. In addition, bioactive factors can be added to the matrix through an identical reaction scheme. In this case, bioactive factors, including cell adhesion motifs or morphogenic or mitogenic factors could be bound to the matrix by the addition of a cysteine, the thioline-containing amino acid to the sequence. Here we have added a cysteine to the cell adhesion sequence, RGD, and more specifically, RGDSP (SEQ ID NO: 23) as well as to the PTHi-34 sequence and both were bound to the matrix via the acrylates in the crosslinker. Subsequently, these newly formed hydrogels will have numerous esters near a thiol that have been shown to be hydrolytically unstable. This instability allows the gels to slowly degrade and be replaced by newly formed tissue.

These special gels, hydrolytically degradable with RGD and PTH covalently bound to matrix, were tested in the sheep borer defect model to test the ability of matrix-bound C-PTH1.34 to enhance bone growth. To determine the amount of amplification, 0, 100 and 400 µg/ml PTHi.34 fusion peptide was added to the matrix. When this was done, an increase in bone formation was observed with the addition of PTH 1.34.1 each test, the healing reaction was measured at the same time of 8 weeks. This was compared to defects that were left empty. When the synthetically gradable matrix without PTH fusion peptide was used, the healing reaction was measured at approx. 40%. This means that 40% of the original defective volume was filled with newly formed bone tissue. Then, when 400 µg/ml modified PTHi-34 fusion peptide was used, the healing response increased to approx. 60%. In comparison, when the defects were left empty, approx. 10% filled with calcined tissue. These data are shown in table 6 below.

In comparison to one that was defective, addition of the hydrolytically degradable PEG gel alone had a large effect on bone healing, with the amount of calcined tissue increasing by 300%. When PTH 1-34 was coupled to this matrix, healing was even further increased with levels up to 50% higher than when matrix was used alone and 500% higher than the level of healing achieved when the defect was left empty.

Claims (19)

1. Fusion peptide, characterized in that it comprises a first domain comprising parathyroid hormone (PTH) and a second domain comprising a substrate domain, which is covalently cross-linkable to a fibrin matrix material or to a synthetic matrix material, so that the fusion peptide can be connected to the matrix through the second domain. 2. Fusion peptide as stated in claim 1, characterized in that the matrix is fibrin and the second domain comprises a transglutaminase substrate domain. 3. Fusion peptide as stated in claim 2, characterized in that the second domain comprises a Factor XIIIa substrate domain. 4. Fusion peptide as stated in claim 3, characterized in that the Factor XIIIa substrate domain comprises SEQ. ID. No. 12. 5. Fusion peptide as stated in claim 1, characterized in that the matrix is a synthetic matrix and the second domain comprises at least one cysteine. 6. Fusion peptide as set forth in any of claims 1-5, characterized in that PTH is selected from the group consisting of PTH 1-84, PTH 1-38, PTH 1-34, PTH 1-31 and PTH 1-25. 7. Fusion peptide as stated in claim 6, characterized in that PTH is PTH 1-34. 8. Fusion peptide as set forth in any of claims 1-7, characterized in that it further comprises a degradation site between the first and second domains. 9. Fusion peptide as stated in claim 8, characterized in that the degradation site is an enzymatic or hydrolytic degradation site. 10. Fusion peptide as stated in claim 8, characterized in that the degradation site is an enzymatic degradation site, which is cleaved by an enzyme selected from the group consisting of plasmin and matrix metalloproteinase. 11. Set, characterized in that it comprises the fusion peptide as stated in any of claims 1-10. 12. Set as stated in claim 11, characterized in that it further comprises fibrinogen, thrombin and a calcium source. 13. Set as stated in claim 11, characterized in that the kit further comprises a cross-linking enzyme. 14. Matrix suitable for cellular growth or ingrowth, characterized in that it comprises the fusion peptide as stated in any of claims 1-10, in which the fusion peptide is covalently bound to the matrix. 15. Matrix as specified in claim 14, characterized in that the matrix is fibrin. 16. Matrix as stated in claim 14, characterized in that the matrix is formed by a Michael-type addition reaction between a first precursor molecule comprising n nucleophilic groups and a second precursor molecule comprising m electrophilic groups in which n and m are at least two and the sum n+m is at least five. 17. Matrix as specified in claim 16, characterized in that the electrophilic groups are conjugated unsaturated groups and the nucleophilic groups are selected from the group consisting of thiols and amines. 18. Matrix as specified in claim 14, characterized in that it comprises polyethylene glycol. 19. Method for producing a matrix, characterized in that it includes providing at least one matrix material capable of forming a cross-linked matrix, wherein the matrix material is selected from the group consisting of proteins and synthetic materials, adding the fusion peptide as set forth in any of claims 1-10 to the matrix material, and crosslink the matrix material, so that the fusion peptide is connected to the matrix through the second domain.