Hybrid Proteins Having Cross-Linking and Tissue-Binding Activities

The present invention relates generally toward methods for producing recombinant hybrid proteins, and more specifically, to methods for producing hybrid proteins from host cells through the use of recombinant DNA techniques.

The utilization of tissue sealants to replace or augment the use of mechanical wound closure devices has expanded in recent years in many surgical and trauma applications. Tissue sealants include biological adhesives (e.g. fibrin-based adhesives) and synthetic preparations (e.g. cyanoacrylates). It is widely acknowledged that the use of synthetic preparations of tissue sealants is limited due to their toxicity and limited applications. Biological tissue adhesives have demonstrated utility in cases where the use of mechanical devices to close wounds is insufficient, such as in joining blood vessels, closing holes in the dura, and in surgery on small or delicate tissues such as in the eye or ear.

Fibrin-based biological tissue adhesives generally contain fibrinogen, factor XIII and thrombin as principal ingredients, although in practice biological tissue adhesives are derived from whole blood and contain additional blood proteins. The fibrinogen and factor XIII components of these adhesives are prepared from pooled human plasma by cryoprecipitation (e.g. U.S. Pats. Nos. 4,377,572; 4,362,567; 4,909,251), by ethanol precipitation (e.g. U.S. Pat. No. 4,442,655) or from single donor plasma (e.g. U.S. Pat. No. 4,627,879; Spotnitz et al., Am. Surg. 55: 166-168, 1989). The resultant fibrinogen/factor XIII preparation is mixed with bovine thrombin immediately before use to convert the fibrinogen to fibrin and activate the factor XIII, thus initiating coagulation of the adhesive.

Fibrin-based tissue adhesives, in their current form, have significant drawbacks that include poor standardization, lack of quality control from batch to batch and the possibility of transmission of human immunodeficiency virus (HIV), hepatitis virus and other etiologic agents. While recombinant production of thrombin and factor XIII have been reported, and while these proteins might be used in biological tissue adhesives, the biological tissue adhesives still rely on large amounts of fibrinogen that is obtained from pooled human blood. At present, current fibrin(ogen)-based tissue adhesives are not approved for use in the United States.

There is therefore a need in the art for tissue adhesive components, particularly components that facilitate cross-linking to improve clot strength, that are prepared at high levels with reproducible activity levels and which do not carry the possibility of transmission of viral or other etiologic agents. The present invention addresses these needs by providing recombinant hybrid proteins that provide cross-linking and tissue-adhesive properties and that may be prepared at high levels.

Briefly stated, the present invention provides hybrid proteins having cross-linking and tissue-binding activities, DNA molecules encoding such hybrid proteins and methods for producing hybrid proteins by recombinant means. In one aspect of the invention, the hybrid proteins comprise a tissue-binding domain from a first protein covalently linked to a cross-linking domain from a second protein. Within a related aspect of the invention, the tissue-binding domain of the first protein is a heparin binding domain of thrombospondin, a heparin binding domain of fibronectin, a collagen binding domain of fibronectin or a cell binding domain of fibronectin. Within a preferred embodiment, the tissue-binding domain of the first protein comprises the amino acid sequence of Sequence ID No. 6 from Alanine, amino acid 2 to Glutamic acid, amino acid number 926. Within another related aspect of the invention, the cross-linking domain of the second protein comprises the carboxy-terminal 103 amino acids of loricrin, the ten amino acid repeat beginning with glutamine, amino acid number 496 of involucrin or the 400 amino-terminal amino acids of the fibrinogen .alpha. chain. Within a preferred embodiment of the invention, the cross-linking domain of the second protein comprises the amino acid sequence of Sequence ID No. 6 from Glycine, amino acid number 928 to Proline, amino acid number 1336. Within a particularly preferred embodiment, the hybrid protein comprises the amino acid sequence of Sequence ID No. 6 from alanine, amino acid number 2 to proline, amino acid number 1336.

The present invention provides DNA molecules encoding hybrid proteins of the present invention comprising a first DNA segment encoding a tissue-binding domain from a first protein joined to a second DNA segment encoding a cross-linking domain from a second protein. In one embodiment, the first DNA segment comprises the nucleotide sequence of Sequence ID No. 5 from nucleotide 3 to nucleotide 2780. In another embodiment, the second DNA segment comprises the nucleotide sequence of Sequence ID No. 5 from nucleotide 2784 to nucleotide 4013. In a preferred embodiment, the DNA molecule comprises the nucleotide sequence of Sequence ID Number 5 from nucleotide 3 to nucleotide 4013.

In related embodiments of the invention, DNA constructs are provided which comprise a DNA molecule encoding a hybrid protein, wherein said DNA molecule comprises a first DNA segment encoding a tissue-binding domain from a first protein joined to a second DNA segment encoding a cross-linking domain from a second protein and wherein said DNA molecule is operably linked to other DNA segments required for the expression of the DNA molecule. Other embodiments of the invention concern host cells containing the DNA constructs of the present invention and methods of producing hybrid proteins.

FIG. 1 discloses a representative hybrid protein containing (1) an N-terminal end-to-end inter-chain cross-linking domain, (2) a domain that promotes inter-chain cross-linking; (3) a domain that confers tissue binding activity; and (4) a carboxy-terminal domain that promotes end-to-end inter-chain cross-linking.

FIGS. 2-5 disclose absorbance time courses of representative cross-linking assays carried out in the presence of varying levels of factor XIII (activated to factor XIIIa via thrombin during the assay) or factor XIIIa.

The present invention provides novel hybrid proteins having cross-linking and tissue adhesive activities. The hybrid proteins comprise a cross-linking domain from a first protein covalently linked to a tissue-binding domain from a second protein. The hybrid proteins of the present invention are capable of cross-linking to themselves and to other proteins such as fibrin and fibrinogen and are capable of adhering to cell surfaces and/or extracellular matrix components. While not wishing to be bound by a graphical representation, FIG. 1 shows a representative hybrid protein containing an N-terminal end-to-end inter-chain cross-linking domain; a domain that promotes inter-chain cross-linking; a domain that confers tissue binding activity; and a carboxy-terminal domain that promotes end-to-end inter-chain cross-linking. As used herein, cross-linking refers to the formation of covalent bonds between polypeptides.

The hybrid proteins of the present invention are useful as components of tissue sealant formulations to provide matrix material and to improve clot strength over a wound site, and as components in formulations that promote wound healing. The proteins of the present invention may contain native (i.e. wild-type) protein domains as well as domains that are allelic variants and genetically engineered or synthetic variants of the respective naturally occurring domains. Such variants are characterized by the presence of conservative amino acid substitutions and/or other minor additions, substitutions or deletions of amino acids.

As used within the context of the present invention, tissue-binding domains include protein domains containing amino acid sequences that facilitate adherence to cell surfaces and/or to extracellular matrix components such as collagen, fibronectin, hyaluronic acid and glycosaminoglycans. Fibronectin, for example, contains the sequence Gly-Arg-Gly-Asp-Ser (from amino acid 1614 through amino acid 1618 of Sequence I.D. No. 2) that has been shown to be central to cell recognition by the fibronectin receptor (for review see Yamada, Current Opinion in Cell Biology 1: 956-963, 1989) The heparin binding domains of fibronectin (Sekiguchi et al., Proc. Natl. Acad. Sci. USA 77: 2661-2665, 1980), and thrombospondin contain sequences that recognize heparin sulfate-containing glycosaminoglycans which are extracellular matrix components. The collagen binding domain of fibronectin (Sekiguchi et al. ibid., 1980) contains amino acid sequences that bind to the extracellular matrix component collagen.

Particularly preferred tissue-binding domains are the heparin binding domain of fibronectin, comprising the sequence of amino acids of Sequence I.D. No. 2 from alanine, amino acid number 1812 to valine, amino acid number 2171; the collagen binding domain of fibronectin, comprising the sequence of amino acids of Sequence I.D. No. 2 from glycine, amino acid number 282 to serine, amino acid number 608; and the amino terminal 229 amino acids of thrombospondin. In this regard, a particularly preferred tissue-binding domain is the cell-binding domain of fibronectin, comprising the sequence of amino acids of Sequence I.D. No. 2 from alanine, amino acid number 1357 to glutamic acid, amino acid number 1903. It will be evident to one skilled in the art that smaller portions of the cell-binding domain of fibronectin may be used within the hybrid proteins of the present invention, more particularly the sequence of amino acids of Sequence I.D. No. 2 from isoleucine, number 1532 through threonine, amino acid number 1631. As noted above, it is generally accepted that the sequence Gly-Arg-Gly-Asp-Ser (Amino acids 1614 to 1618 of Sequence I.D. No. 2) is central to cell recognition by fibronectin.

Cross-linking domains suitable for use in the hybrid proteins of the present invention are protein domains which contain amino acid sequences required for the formation of specific covalent bonds between peptide chains. In a preferred embodiment the inter-chain cross-links are covalent bonds formed by the action of a transglutaminase such as factor XIII, tissue transglutaminase, prostate transglutaminase, keratinocyte transglutaminase, epidermal transglutaminase or placental transglutaminase. Transglutaminases catalyze the formation of .epsilon.-(.gamma.-glutamyl)lysine bonds between specific glutamine and lysine residues. However, other inter-chain cross-links, such as those formed by disulfide bonds, are also suitable cross-links. Suitable cross-linking domains include domains from the fibrinogen .alpha. chain, the glutamine/lysine rich domains of loricrin that are involved in isodipeptide cross-link formation (Hohl et al., J. Biol. Chem. 266: 6626-6636, 1991), and at least one of the 10 amino acid-long repeats of involucrin (Cell 46: 583-589, 1986 and Etoh et al., Biochem. Biophys. Res. Comm. 136: 51-56, 1986). Preferred cross-linking domains are the carboxy-terminal 103 amino acids of loricrin (Hohl et al., ibid.) and the ten-amino acid repeat beginning with glutamine, amino acid number 496 of involucrin (Simon et al. J. Biol. Chem. 263: 18093-18098, 1988). A particularly preferred cross-linking domain comprises the 400 amino-terminal amino acids of the fibrinogen .alpha. chain (Doolittle et al., Nature 280: 464-468, 1979; Rixon et al., Biochemistry 22: 3250-3256, 1983). More particularly, the amino acid sequence of Sequence ID No. 6 from Glycine, amino acid number 928 to Proline, amino acid number 1336 is preferred.

Although the hybrid proteins of the present invention may consist essentially of covalently linked cross-linking and tissue binding domains, they may further contain domains that facilitate end-to-end covalent cross-linking. The .gamma. chain of fibrinogen contains a domain that facilitates end-to-end cross-linking to another .gamma. chain via .epsilon.-(.gamma.-glutamyl)lysine bonds. This domain includes at least the 19 carboxy-terminal amino acids and more preferably includes the amino-terminal 275 amino acids of the fibrinogen .gamma. chain. The .alpha. chain of fibrinogen contains an amino-terminal domain that is involved in interchain disulfide bond formation between .alpha. chains. This domain includes the amino-terminal portion of the .alpha. chain of fibrinogen from glycine, amino acid 36 to glycine, amino acid 67 of Sequence ID Number 4.

As will be evident to one skilled in the art, the hybrid proteins of the present invention may contain domains of human and other animal proteins. Proteins containing domains suitable for use in the present invention from human and other animals and the DNA molecules encoding such proteins have been reported. Involucrin, loricrin, fibrinogen and fibronectin, for example, have been studied in a variety of animals. DNA sequences encoding primate, canine and porcine involucrin have been reported (Djian and Green, Mol. Biol. Evol. 9: 417-432, 1992; Djian and Green, Proc. Natl. Acad. Sci. USA 88: 5321-5325, 1991 and Tseng and Green, Mol. Biol. Evol. 7: 293-302, 1990). Mehrel et al. (Cell 61: 1103-1112, 1990) have reported a DNA sequence encoding mouse loricrin. DNA sequences encoding rat and frog fibrinogen gamma chain have been reported (Haidaris and Courtney, Blood 79: 1218-1224, 1992 and Bhattacharya et al., Mol. Cell. Endocrinol. 72: 213-220, 1990; respectively). DNA sequences encoding chicken and lamprey fibrinogen .alpha. chains have been reported by Weissbach and Greininger (Proc. Natl. Acad. Sci. USA 87: 5198-5202, 1990) and Pan and Doolittle (Proc. Natl. Acad. Sci. USA 89: 2066-2070, 1992), respectively. DNA sequences encoding bovine and rat fibronectin have been reported by Petersen et al. (Proc. Natl. Acad. Sci. USA 80: 137-141, 1983) and Schwarzbauer et al., (Cell 35: 421-431, 1983). In general, it is preferred to prepare proteins that contain component domains from a single species to minimize the possibility of immunogenicity. Thus, the present invention provides hybrid proteins that can be used in human and veterinary medicine.

According to the present invention hybrid proteins having cross-linking and tissue adhesive activities are produced recombinantly from host cells transformed with a DNA construct comprising a DNA segment encoding a cross-linking domain from a first protein joined to a DNA segment encoding a tissue-binding domain from a second protein. As used within the context of the present invention, two or more DNA coding sequences are said to be joined when, as a result of in-frame fusions between the DNA coding sequences or as a result of the removal of intervening sequences by normal cellular processing, the DNA coding sequences can be translated into a polypeptide fusion. Unless otherwise noted, the DNA segments may be joined in any order to result in a DNA coding sequence that can be translated into a polypeptide chain. Thus, the DNA segment encoding the tissue-binding domain may be joined to the 5' or the 3' end of the DNA segment encoding the cross-linking domain. However, as will be evident to one skilled in the art, the production of hybrid proteins that additionally include domains that facilitate end-to-end cross-linking will require that the DNA segments encoding such domains be positioned at the 5' and 3' termini of the molecules.

Thus the present invention also provides isolated DNA molecules encoding hybrid proteins comprising a cross-linking domain from a first protein covalently linked to a tissue-binding domain from a second protein. In general, cDNA sequences are preferred for carrying out the present invention due to their lack of intervening sequences which can lead to aberrant RNA processing and reduced expression levels. DNA molecules encoding human fibronectin (Dufour et al., Exper. Cell Res. 193: 331-338, 1991) and a human fibrinogen .alpha. chain (Rixon et al., Biochemistry 22: 3250-3256, 1983) may be obtained from libraries prepared from liver cells according to standard laboratory procedures. It will be understood however, that suitable DNA sequences can also be obtained from genomic clones or can be synthesized de novo according to conventional procedures. If partial clones are obtained, it is necessary to join them in proper reading frame to produce a full length clone, using such techniques as endonuclease cleavage, ligation, and loop-out mutagenesis.

DNA sequences encoding hybrid proteins of the present invention may be prepared from cloned DNAs using conventional procedures of endonuclease cleavage, exonuclease digestion, ligation and in vitro mutagenesis. Alternatively, DNA sequences encoding the cross-linking and tissue-binding domains, such as those mentioned above, may be synthesized using standard laboratory techniques.

An exemplary DNA molecule encoding a hybrid protein having cross-linking and tissue-binding activities may be prepared by joining a DNA segment encoding at least the cell-binding domain of fibronectin and a DNA segment encoding at least an inter-chain cross-linking domain of fibrinogen at a convenient restriction site using synthetic adapters to facilitate in-frame joining of the DNA segments. Alternatively, such DNA segments encoding hybrid proteins of the present invention may be prepared by joining the two domains at a convenient restriction site followed by loop-out mutagenesis to precisely remove unnecessary sequences and directly join the DNA segment encoding the cell-binding domain of fibronectin with the DNA segment encoding the cross-linking domain of fibrinogen.

DNA segments encoding the hybrid proteins of the instant invention are inserted into DNA constructs. As used within the context of the present invention, a DNA construct is understood to refer to a DNA molecule, or a clone of such a molecule, either single- or double-stranded, which has been modified through human intervention to contain segments of DNA combined and juxtaposed in a manner that would not otherwise exist in nature. DNA constructs of the present invention comprise a first DNA segment encoding a hybrid protein operably linked to additional DNA segments required for the expression of the first DNA segment. Within the context of the present invention, additional DNA segments will generally include promoters and transcription terminators, and may further include enhancers and other elements.

DNA constructs may also contain DNA segments necessary to direct the secretion of a polypeptide or protein of interest. Such DNA segments may include at least one secretory signal sequence. Secretory signal sequences, also called leader sequences, prepro sequences and/or pre sequences, are amino acid sequences that act to direct the secretion of mature polypeptides or proteins from a cell. Such sequences are characterized by a core of hydrophobic amino acids and are typically (but not exclusively) found at the amino termini of newly synthesized proteins. DNA segments encoding secretory signal sequences are placed in-frame and in the correct spatial relationship to the DNA segment encoding the protein of interest in order to direct the secretion of the protein. Very often the secretory peptide is cleaved from the mature protein during secretion. Such secretory peptides contain processing sites that allow cleavage of the secretory peptides from the mature proteins as they pass through the secretory pathway. A preferred processing site is a dibasic cleavage site, such as that recognized by the Saccharomyces cerevisiae KEX2 gene. A particularly preferred processing site is a Lys-Arg processing site. Processing sites may be encoded within the secretory peptide or may be added to the peptide by, for example, in vitro mutagenesis.

Preferred secretory signals include the .alpha. factor signal sequence (pre-pro sequence: Kurjan and Herskowitz, Cell 30: 933-943, 1982; Kurjan et al., U.S. Pat. No. 4,546,082; Brake, U.S. Pat. No. 4,870,008), the PHO5 signal sequence (Beck et al., WO 86/00637), the BAR1 secretory signal sequence (MacKay et al., U.S. Pat. No. 4,613,572; MacKay, WO 87/002670), the SUC2 signal sequence (Carlsen et al., Molecular and Cellular Biology 3: 439-447, 1983). Alternately, a secretory signal sequence may be synthesized according to the rules established, for example, by von Heinje (European Journal of Biochemistry 133: 17-21, 1983; Journal of Molecular Biology 184: 99-105, 1985; Nucleic Acids Research 14: 4683-4690, 1986).

Secretory signal sequences may be used singly or may be combined. For example, a DNA segment encoding a first secretory signal sequence may be used in combination with a DNA segment encoding the third domain of barrier (described in U.S. Pat. No. 5,037,243, which is incorporated by reference herein in its entirety). The DNA segment encoding the third domain of barrier may be positioned in proper reading frame 3' of the DNA segment of interest or 5' to the DNA segment and in proper reading frame with both the DNA segment encoding the secretory signal sequence and the DNA segment of interest.

The choice of suitable promoters, terminators and secretory signals is well within the level of ordinary skill in the art. Methods for expressing cloned genes in Saccharomyces cerevisiae are generally known in the art (see, "Gene Expression Technology," Methods in Enzymology, Vol. 185, Goeddel (ed.), Academic Press, San Diego, Calif., 1990 and "Guide to Yeast Genetics and Molecular Biology," Methods in Enzymology, Guthrie and Fink (eds.), Academic Press, San Diego, Calif., 1991; which are incorporated herein by reference). Transformation systems for other yeasts, including Hansenula polymorphic, Schizosaccharomyces pombe, Kluyveromyces lactis, Kluyveromyces fragilis, Ustilago maydis, Pichia pastoris, Pichia quillermondil and Candida maltosa are known in the art. See, for example, Gleeson et al., J. Gen. Microbiol. 132: 3459-3465, 1986 and Cregg, U.S. Pat. No. 4,882,279.

Proteins of the present invention can also be expressed in filamentous fungi, for example, strains of the fungi Aspergillus (McKnight et al., U.S. Pat. No. 4,935,349, which is incorporated herein by reference) Methods for transforming Acremonium chrysogenum are disclosed by Sumino et al., U.S. Pat. No. 5,162,228, which is incorporated herein by reference.

Other higher eukaryotic cells may also be used as hosts, including insect cells, plant cells and avian cells. Transformation of insect cells and production of foreign proteins therein is disclosed by Guarino et al., U.S. Pat. No. 5,162,222 and Bang et al., U.S. Pat. No. 4,775,624, which are incorporated herein by reference. The use of Agrobacterium rhizogenes as a vector for expressing genes in plant cells has been reviewed by Sinkar et al., J. Biosci. (Bangalore) 11: 47-58, 1987.

Expression of cloned genes in cultured mammalian cells and in E. coli, for example, is discussed in detail in Sambrook et al. (Molecular Cloning: A Laboratory Manual, Second Edition, Cold Spring Harbor, N.Y., 1989; which is incorporated herein by reference). In addition to E. coli, Bacillus and other genera are useful prokaryotic hosts for expressing foreign proteins. As would be evident to one skilled in the art, one could express the proteins of the instant invention in other host cells such as avian, insect and plant cells using regulatory sequences, vectors and methods well established in the literature.

In yeast, suitable vectors for use in the present invention include YRp7 (Struhl et al., Proc. Natl. Acad. Sci. USA 76: 1035-1039, 1978), YEp13 (Broach et al., Gene 8: 121-133, 1979), POT vectors (Kawasaki et al, U.S. Pat. No. 4,931,373, which is incorporated by reference herein), pJDB249 and pJDB219 (Beggs, Nature 275: 104-108, 1978) and derivatives thereof. Preferred promoters for use in yeast include promoters from yeast glycolytic genes (Hitzeman et al., J. Biol. Chem. 255: 12073-12080, 1980; Alber and Kawasaki, J. Mol. Appl. Genet. 1: 419-434, 1982; Kawasaki, U.S. Pat. No. 4,599,311) or alcohol dehydrogenase genes (Young et al., in Genetic Engineering of Microorganisms for Chemicals, Hollaender et al., (eds.), p. 355, Plenum, N.Y., 1982; Ammerer, Meth. Enzymol. 101: 192-201, 1983). In this regard, particularly preferred promoters are the TPI1 promoter (Kawasaki, U.S. Pat. No. 4,599,311, 1986) and the ADH2-4.sup.c promoter (Russell et al., Nature 304: 652-654, 1983; Irani and Kilgore, U.S. patent application Ser. No. 07/631,763, Calif. 1,304,020 and EP 284 044, which are incorporated herein by reference). The expression units may also include a transcriptional terminator. A preferred transcriptional terminator is the TPI1 terminator (Alber and Kawasaki, ibid.).

Host cells containing DNA constructs of the present invention are then cultured to produce the hybrid proteins. The cells are cultured according to standard methods in a culture medium containing nutrients required for growth of the particular host cells. A variety of suitable media are known in the art and generally include a carbon source, a nitrogen source, essential amino acids, vitamins, minerals and growth factors. The growth medium will generally select for cells containing the DNA construct by, for example, drug selection or deficiency in an essential nutrient which is complemented by a selectable marker on the DNA construct or co-transfected with the DNA construct.

Selection of a medium appropriate for the particular host cell used is within the level of ordinary skill in the art. Yeast cells, for example, are preferably cultured in a chemically defined medium, comprising a non-amino acid nitrogen source, inorganic salts, vitamins and essential amino acid supplements. The pH of the medium is preferably maintained at a pH greater than 2 and less than 8, preferably at pH 6.5. Methods for maintaining a stable pH include buffering and constant pH control, preferably through the addition of sodium hydroxide or ammonium hydroxide. Preferred buffering agents include succinic acid and Bis-Tris (Sigma Chemical Co., St. Louis, Mo.). Yeast cells having a defect in a gene required for asparagine-linked glycosylation are preferably grown in a medium containing an osmotic stabilizer. A preferred osmotic stabilizer is sorbitol supplemented into the medium at a concentration between 0.1M and 1.5M, preferably at 0.5M or 1.0M. Cultured mammalian cells are generally cultured in commercially available serum-containing or serum-free media.

The recombinant hybrid proteins expressed using the methods described herein are isolated and purified by conventional procedures, including separating the cells from the medium by centrifugation or filtration, precipitating the proteinaceous components of the supernatant or filtrate by means of a salt, e.g. ammonium sulfate, purification by a variety of chromatographic procedures, e.g. ion exchange chromatography or affinity chromatography, or the like. Methods of protein purification are known in the art (see generally, Scopes, R., Protein Purification, Springer-Verlag, N.Y. (1982), which is incorporated herein by reference) and may be applied to the purification of the recombinant proteins of the present invention.

The hybrid proteins of the present invention may be used as components of tissue adhesives. It is preferred that the tissue adhesives be formulated to provide a concentration of the hybrid proteins of the present invention of between about 5 mg/ml to 100 mg/ml, with concentrations in the range of 35 to 50 mg/ml being particularly preferred. As disclosed above, tissue adhesives generally contain factor XIII and thrombin. Additional components may also be included in the tissue adhesive formulations. These additional components include growth factors such as PDGF, bFGF, TGFa, or EGF and protease inhibitors, such as aprotinin, transexamic acid, alpha-2 plasmin inhibitor, alpha-1-antitrypsin or the Pittsburgh mutant of alpha-1-antitrypsin (Arg-358 alpha-1-antitrypsin). The tissue adhesives may also contain salts, buffering agents, reducing agents, bulking agents, and solubility enhancers. Albumin, NaCl, CaCl.sub.2, citrate and phosphate buffers, for example, may be included. Preferably, the tissue adhesives of the present invention are prepared as lyophilized powders, liquid concentrates of ready-to-use liquids. Lyophilized powders are preferred for ease of handling and storage.

The following examples are offered by way of illustration and not by way of limitation.

Subcloning and Modification of ADH2 Promoters

An ADH2-4.sup.c promoter was constructed as described in co-pending U.S. patent application Ser. No. 07/631,763, Calif. 1,304,020 and EP 284 044, which are incorporated herein by reference. A DNA construct comprising the complete ADH2-4.sup.c promoter mutagenized at the 3' end to place an Eco RI site in place of the translation start codon, designated p410-4.sup.c (deposited with the American Type Culture Collection (12301 Parklawn Dr., Rockville, Md. 20852) under accession number 68861) was used as the source of the ADH2-4.sup.c promoter.

A PAP-I cDNA (U.S. Pat. No. 4,937,324) was joined with the ADH2-4.sup.c promoter. Plasmid pAP1.7, comprising the 1.7 kb cDNA in pUC18, was cut with Nco I and Bam HI, and the linearized plasmid was isolated through two rounds of gel purification. The ADH2-4.sup.c promoter from p410-4.sup.c was joined to the 5' end of the PAP-I cDNA via an Eco RI-Nco I adapter. The 1.2 kb Bam HI-Eco RI promoter fragment from p410-4.sup.c, Eco RI-Nco I adapter and the Nco I-Bam HI linearized pAP1.7 plasmid were ligated. The resultant plasmid was designed pPR1. The presence of the correct promoter fusion was confirmed by DNA sequencing.

A yeast expression vector comprising the ADH2-4.sup.c promoter, the PAP-I cDNA and the TPI1 terminator was constructed. Plasmid pZUC13 (comprising the S. cerevisiae chromosomal LEU2 gene and the origin of replication from S. cerevisiae 2 micron plasmid inserted into pUC13 and constructed in a manner analogous to pZUC12, described in published EP 195,691, using the plasmid pMT212, which is described in published EP 163 529) was cut with Bam HI. Plasmid pPR1 was digested completely digested with Bam HI and partially digested with Sac I to isolate the 2.1 kb ADH2-4.sup.c promoter-PAP-I cDNA fragment. Plasmid pTT1 (described in detail below) was digested with Sac I and Bam HI to isolate the 0.69 bp TPI1 terminator fragment. The Bam HI-Sac I fragment from pPR1 and the Sac I-Bam HI fragment from pTT1 were ligated with the Bam HI-linearized pZUC13. A plasmid containing the expression unit was designated pZ3.

Subcloning of the TPI1 terminator

The yeast TPI1 terminator fragment was obtained from plasmid p270 described by Murray and Kelly (U.S. Pat. 4,766,073, which is incorporated by reference herein in its entirety). Plasmid p270 contains the TPI1 terminator inserted as and Xba I-Bam HI fragment into YEp13. Alternatively, the TPI1 terminator may be obtained from plasmid pM220 (deposited with American Type Culture Collection as an E. coli RR1 transformant under accession number 39853) by digesting the plasmid with Xba I, and Bam HI and purifying the TPI1 terminator fragment (.about.700 bp).

The TPI1 terminator was removed from plasmid p270 as a Xba I-Bam HI fragment. This fragment was cloned into pUC19 along with another fragment containing the TPI1 promoter fused to the CAT (chloramphenicol acetyl transferase) gene to obtain a TPI1 terminator fragment with an Eco RV end. The resultant plasmid was designated pCAT. The TPI1 terminator was then cut from pCAT as an Eco RV-Bam HI fragment and cloned into pIC19H (Marsh et al., Gene 32: 481-486, 1984) which had been cut with the same enzymes, to obtain pTT1 (disclosed in U.S. Pat. No. 4,937,324, which is incorporated herein by reference).

Construction of Yeast Vectors pDPOT and pRPOT

Plasmid pDPOT was derived from plasmid pCPOT (ATCC No. 39685) by replacing the 750 bp Sph I-Bam HI fragment of pCPOT containing 2 micron and pBR322 sequences with a 186 bp Sph I-Bam HI fragment derived from the pBR322 tetracycline resistance gene.

Plasmid pRPOT was derived from plasmid pDPOT by replacing the Sph I-Bam HI fragment with a polylinker. Plasmid pDPOT was digested with Sph I and Bam HI to isolate the 10.8 kb fragment. Oligonucleotides ZC1551 and ZC1552 (Sequence ID Nos. 7 and 8) were designed to form an adapter with a Bam HI adhesive end and an Sph I adhesive end flanking Sma I, Sst I and Xho I restriction sites. Oligonucleotides ZC1551 and ZC1552 (Sequence ID Nos. 7 and 8) were kinased and annealed to form the Bam HI-Sph I adapter. The 10.8 kb pDPOT fragment was circularized by ligation with the ZC1551/ZC1552 adapter (Sequence ID Nos. 7 and 8). The resultant plasmid was termed pRPOT.

A. Construction of pFN14A

A DNA construct containing a DNA segment encoding the fibronectin cell-binding domain operably linked to the ADH2-4.sup.c promoter in plasmid pUC19 was constructed. The fibronectin coding sequence was obtained from plasmid pFH103 (Dufour et al., Exper. Cell Res. 193: 331-338, 1991). Plasmid pFH103 was digested with Nco I and Xba I to isolate the 4 kb fragment containing the fibronectin coding sequence. Oligonucleotides ZC2052 and ZC2053 (Sequence ID Nos. 9 and 10) were designed to provide, upon annealing, an adapter containing a 5' Eco RI adhesive end, an internal Nco I site, a DNA segment encoding a methionine and amino acids 979 through 981 of Sequence ID Number 2 and a 3' Nco I adhesive end that destroys the Nco I site. Oligonucleotides ZC2052 and ZC2053 (Sequence ID Nos. 9 and 10) were annealed and ligated with the 4 kb Nco I-Xba I fibronectin fragment into Eco RI-Xba I linearized pUC19. The resultant plasmid was designated pFN4.

Plasmid pFN4 was digested with Hind III and Apa I to isolate the 3.3 kb fibronectin fragment. Oligonucleotides ZC2493 and ZC2491 (Sequence ID Nos. 12 and 11) were designed to provide, when annealed, an Apa I-Xba I adapter encoding the amino acids Pro and Phe followed by a stop codon. The oligonucleotides were annealed and combined with the 3.3 kb Hind III-Apa I fragment and Hind III-Xba I linearized pUC19 to form plasmid pFN7. Plasmid pFN7 comprises a DNA segment encoding amino acids 1273-2186 of Sequence ID Number 2 followed by an in-frame stop codon.

The ADH2-4.sup.c promoter was joined to the 5' end of the fibronectin cDNA in plasmid pFN5. Plasmid pFN4 was digested with Nco I and Hind III to isolate the 0.89 kb fibronectin coding sequence. Plasmid pZ3 (described in detail above) was digested with Bam HI and Nco I to isolate the 1.25 kb ADH2-4.sup.c promoter fragment. The 1.25 kb Bam HI-Nco I promoter fragment and the Nco I-Hind III fibronectin coding sequence fragment were ligated to Bam HI-Hind III linearized pUC19 to form plasmid pFN5.

Plasmid pFN5 was digested with Bam HI and Hind III to isolate the 2.1 kb promoter-fibronectin fragment. Plasmid pFN7 was digested with Hind III and Xba I to isolate the 2.8 kb fibronectin fragment that was modified to encode a stop codon following the Pro-Phe sequence. The TPI1 terminator sequence was obtained from pTT1 as a 0.7 kb Xba I-Sal I fragment. The 2.1 kb Bam HI-Hind III promoter-fibronectin fragment, the 2.8 kb Hind III-Xba I fibronectin fragment and the 0.7 kb TPI1 terminator fragment were joined in a four-part ligation with Bam HI-Xho I linearized pRPOT. A plasmid containing the fibronectin expression unit in the pRPOT vector was designated pR1.

The original clone pFH103 contained a frame-shift mutation in the EIIIB region of the fibronectin cDNA. The mutation was corrected by the replacement of the region with an analogous region from the plasmid pFH.DELTA.3 (obtained from Jean Paul Thiery, Laboratoire de Physiopathologie du Developpement, CNRS URA 1337, Ecole Normale Superiure, 46 rue d'Ulm, 75230 Paris Cedex 05, France). Plasmid pFH.DELTA.3 was derived from pFH103 by excising the 3211 bp Xba I-Asp 718I fragment of fibronectin, blunting of the resultant adhesive ends and religating. Plasmid pFH.DELTA.3 contains a DNA segment encoding the signal and propeptides, the first three and one half type I repeats, and the carboxy-terminal half of human fibronectin from the middle of the EIIIB segment.

Plasmid pR1 was digested with Bam HI and Kpn I to isolate the 2.2 kb promoter-fibronectin fragment. Plasmid pFH.DELTA.3 was digested with Kpn I and Apa I to isolate the internal fibronectin fragment that corrects the frame-shift mutation present in the parent cDNA from pFH103. Plasmid pR1 was digested with Apa I and Bam HI to isolate the TPI1 terminator fragment. The 2.2 kb Bam HI-Kpn I promoter-fibronectin fragment, the 2.75 kb Kpn I-Apa I internal fibronectin fragment and the 0.69 kb Apa I-Bam HI TPI1 terminator fragment were joined in a four-part ligation with Bam HI-linearized pDPOT. The resulting construction was designated pD32.

A DNA segment encoding the ADH2-4.sup.c promoter and initiation methionine from plasmid pD32 was subcloned into pIC19H (Marsh et al., Gene 32: 481-486, 1984) as a 1.25 kb Bam HI-Nco I fragment. Plasmid pD32 was also digested with Nco I and Bgl II to isolate the 3 kb fibronectin cDNA fragment encoding amino acids 979-1972 of Sequence ID Number 2. The 1.25 kb Bam HI-Nco I fragment and the Nco I-Bgl II fragment were ligated with Bam HI-linearized pIC19H. A plasmid containing a Bam HI site proximal to the ADH2-4.sup.c promoter was designated pFN14A.

B. Construction of Plasmid pD38

An expression vector comprising a DNA segment encoding a fibronectin-fibrinogen hybrid protein operably linked to the ADH2-4.sup.c promoter and the TPI1 terminator was constructed. To assemble the DNA sequence encoding the hybrid protein, a DNA segment encoding approximately the carboxy-terminal 409 amino acids of the .alpha. chain of fibrinogen was first subcloned.

A fibrinogen .alpha. chain cDNA was obtained from Dominic W. Chung (Department of Biochemistry, University of Washington, Seattle, Wash.) in plasmid pHI.alpha.3 (Rixon et al., Biochemistry 22: 3250-3256, 1983). Sequence analysis of the cDNA insert in plasmid pHI.alpha.-3 revealed a deletion of codons 1348-1350 of the published sequence resulting in the deletion of Serine, amino acid 417.

The DNA segment encoding the carboxy-terminus of the fibrinogen a chain was subcloned into plasmid pUC19. Plasmid pHI.alpha.-3 was digested with Asp 718 and Ssp I to isolate the approximately 2 kb fragment encoding the carboxy-terminus of the fibrinogen a chain from amino acid 244 to amino acid 643 and some 3' untranslated sequence of Sequence ID Number 4. Plasmid pTT1 was digested with Eco RV and Sal I to isolate the approximately 700 bp TPI1 terminator fragment. The 2 kb fibrinogen .alpha. chain sequence and the TPI1 terminator sequence were ligated with pUC19 that had been linearized with Asp 718 and Sal I. The ligation mixture was transformed into E. coli, and plasmid DNA was prepared and analyzed by restriction endonuclease and DNA sequence analysis. DNA sequence analysis of a candidate clone revealed that the Sal I site joining the TPI1 terminator sequence and the pUC19 polylinker site was not present. Plasmid DNA from the candidate clone was digested with Asp 718 and Bam HI to liberate the approximately 1.9 kb fibrinogen-TPI1 terminator fragment.

To join the fibronectin coding sequence with the fibrinogen .alpha. chain sequence, synthetic oligonucleotides were synthesized to provide, when annealed, a Sal I-Asp 718 adapter encoding an internal Afl II restriction site, and a sequence encoding amino acids 1886 through 1903 of fibronectin (Sequence ID Number 2), a glycine residue and amino acids 235 through 243 of the fibrinogen a chain (Sequence ID Number 4). Oligonucleotides ZC3521 and ZC3522 (Sequence ID Nos. 13 and 14) were annealed. The 1.9 kb Asp 718-Bam HI fibrinogen-TPI1 terminator fragment and the Sal I-Asp 718 ZC3521/ZC3522 adapter (Sequence ID Nos. 13 and 14) were ligated with pUC19 that had been linearized with Sal I and Bam HI. The resultant plasmid was designated pFG4.

The DNA segment encoding the fibronectin-fibrinogen .alpha. chain sequence in plasmid pFG4 was joined with the DNA segment encoding the amino-terminal fibronectin sequence (from amino acid 989 to amino acid 1885 of Sequence ID Number 2) in plasmid pFN14A to construct plasmid pD37. Plasmid pFN14A was digested with Bgl II and Afl II to isolate the approximately 3.9 kb ADH2-4.sup.c promoter-fibronectin fragment. Plasmid pFG4 was digested with Afl II and Bam HI to isolate the approximately 2 kb fibronectin-fibrinogen-TPI1 terminator fragment. The 3.9 kb Bgl II-Afl II fragment and the 2 kb Afl II-Bam HI fragment were ligated with Bam HI-linearized pDPOT. A plasmid with the expression unit inserted with the direction of transcription in the same direction as the POT1 gene in the pDPOT vector was designated pD37.

To place the expression unit present in pD37 in the opposite orientation, such that the direction of transcription of the expression unit was in the opposite direction to that of the POT1 gene, plasmid pD37 was digested with Nco I and Xba I to isolate the approximately 4 kb fibronectin-fibrinogen .alpha. chain fragment. Plasmid pFN14A was digested with Bam HI and Nco I to isolate the approximately 1.3 kb ADH2-4.sup.c promoter fragment. Plasmid pTT1 was digested with Bam HI and Xba I to isolate the approximately 700 bp TPI1 terminator fragment. The Bam HI-Nco I ADH2-4.sup.c promoter fragment, the Nco I-Xba I fibronectin-fibrinogen .alpha. chain fragment and the Xba I-Bam HI TPI1 terminator fragment were ligated with Bam HI-linearized pDPOT that had been treated with calf alkaline phosphatase to prevent recircularization. A plasmid containing the expression unit in the opposite orientation relative to the POT1 gene was designated pD38. The nucleotide sequence and deduced amino acid sequence of the DNA segment encoding the fibronectin-fibrinogen hybrid of plasmid pD38 is shown in Sequence ID Number 5. Plasmid pD38 was deposited on Dec. 15, 1992 with the American Type Culture Collection (12301 Parklawn Drive, Rockville, Md.) as an E. coli transformant under accession Number 69164.

Expression of a Fibronectin-Fibrinogen Hybrid Protein in Yeast

Plasmid pD38 was transformed into the Saccharomyces cerevisiae host strain ZM118 (MATa/MAT.alpha. ura3/ura3 .DELTA.tpil::URA3/.DELTA.tpil::URA3 leu2-3,112/leu2-3,112 bar1/bar1 pep4:: URA3/pep4::URA3 ›cir.sup.o !) using essentially the method described by Hinnen et al. (Proc. Natl. Acad. Sci. USA 75: 1929-1933, 1978). Transformants were selected for their ability to grow on medium containing glucose as the sole carbon source.

The ZM118›pD38! transformant was scaled up in a 60 liter fermenter to facilitate purification of the hybrid protein. A single ZM118›pD38! colony was selected from a YEPD+Ade+Leu plate (Table 1) and inoculated into -LeuTrpThrD medium (Table 1). The culture was incubated for approximately 52 hours after which the cells were harvested. The cells were washed in T. E. buffer (Sambrook et al., ibid.), resuspended in T. E. buffer+30% glycerol, and aliquotted into 1 ml seed vials. The seed vials were stored at -80.degree. C. One seed vial was used to inoculate 100 ml of YEPD+Ade+Leu (Table 1). The culture was grown for approximately 28 hours to a final A.sub.660 of 7.7. The 100 ml culture of ZM118›pD38! was inoculated into a 10 liter fermenter with a final volume of 6.0 liters of medium containing 10 g/L (NH.sub.4).sub.2 SO.sub.4, 5 g/L KH.sub.2 PO.sub.4, 5 g/L MgSO.sub.4. 7H.sub.2 O, 1 g/L NaCl, 0.5 g/L CaCl.sub.2. 2H.sub.2 O, 3.68 g/L A.A.I. (Table 1), 4.2 g/L citric acid, 60 g/L glucose, 10 ml/L Trace Metal Solution (Table 1), 0.4 ml/L PPG-2025 (Polypropylene glycol, MW 2025, Union Carbide Corp, Danbury, Conn.) that had been pH adjusted to pH 5.0 with NaOH. In addition to the inoculation culture, 30 ml of Vitamin solution was added (Table 1). The culture was grown for 23 hours at 30.degree. C. with the addition of 2M NaOH to maintain pH of approximately 5.

A 60 liter fermenter with a final volume of 50 liters of medium containing 60 g/L yeast extract (Universal Foods, Milwaukee, Wis.), 2.5 g/L MgSO.sub.4. 7H.sub.2 O (Mallinkrodt Inc., St. Louis, Mo.), 1 g/L CaCl.sub.2. 2H.sub.2 O (Mallinkrodt, Inc.), 1 g/L KCl (Mallinkrodt, Inc.), 10 ml/L of Trace Metal Solution (Table 1), 0.5 ml/L PPG-2025 (Union Carbide) that had been adjusted to a pH of 5.0 with H.sub.3 PO.sub.4 was prepared, and the medium was sterilized. After sterilization, 5.0 liters of the 23 hour fermentation culture and 500 ml of Vitamin Solution (Table 1) were inoculated into the medium. During the fermentation, a solution of 50% glucose, 5% (NH.sub.4).sub.2 SO.sub.4, 0.05% citric acid was fed into the fermenter at a rate of 150 ml/hour, and the pH was maintained at approximately pH 5 by the addition of 2M NH.sub.4 OH. PPG-2025 was added as needed to control foaming. At approximately 49 hours post inoculation, an ethanol feed was begun by the addition of ethanol to the fermenter at a rate of 150 ml/min. The culture was grown for a total of 67.25 hours at 30.degree. C.

At the end of the fermentation, 50 liters of the culture was diluted to 100 liters with water. The cells were removed from the spent medium by centrifuging 50 liters at a time through a Westfalia CSA 19 centrifuge (Westfalia, Oelde, Germany) at a flow rate of 4 liters/min. The cells were rinsed with water. From the centrifugation, approximately 20 liters of cell slurry containing approximately 35% cells was obtained. Salts were added to the slurry to achieve a final concentration of the following salts: 50 mM NaCl, 10 mM Na.sub.2 HPO.sub.4, 5 mM EDTA. The cell slurry was passed through a Dynomill bead mill using 0.5 mm lead-free glass beads (Willy A Bachofen AG MashinenFabrik, Basle, Switzerland) at a rate of 4 liters per minute. The Dynomill was rinsed with Lysis buffer (50 mM NaCl, 10 mM Na.sub.2 HPO.sub.4, 5 mM EDTA, pH 7.2) to a final volume of 80 liters. The final slurry had a pH of 6.8, a temperature of approximately 10.degree. C. and a conductivity of 5 ms/cm.

The cell slurry was subjected to centrifugation as described above, and the cell pellet was rinsed with lysis buffer. After centrifugation approximately 20 liters of cell slurry was obtained. The cell slurry was extracted by first adjusting the concentration of the cell debris to approximately 40-50% with lysis buffer. Solid urea, NaCl and EDTA were added to the cell slurry to achieve a final concentration of approximately 8M urea, 0.3M NaCl and 10 mM EDTA. The approximate salt concentrations were obtained by the addition of 450 g/L of urea, 18 g/L of NaCl and 4.2 g/L of EDTA. The cell slurry was adjusted to pH 7.8 with 0.5M NaOH. The solids were dissolved into the slurry and the pellets were extracted for a total of 50 minutes. Following extraction, the mixture was diluted 1 to 4 with water, adjusted to a conductivity of 12.5 ms/cm with NaCl and adjusted to a pH of 9.5 with 0.5M NaOH.

The extracted slurry was centrifuged as described above with the lysis buffer rinse. The pH of the supernatant was adjusted to pH 9.5 with 0.5M NaOH. The supernatant was analyzed by SDS polyacrylamide gel electrophoresis (SDS-PAGE) using the PHAST System Separation and Control Unit (Pharmacia LKB Biotechnology Inc., Piscataway, N.J.), and the protein was visualized using Coomassie Blue staining. A 2 liter Q-sepharose column (Pharmacia) was equilibrated at 5 liters/hour with successive washes of the following solutions: 8 liters of 3M urea, 1M NaCl, 50 mM glycine, pH 11.5; 5 liters of 0.5M NaOH; 1.5 liters of water; 5 liters of 0.1M HCl; and 6.0 liters of Wash buffer (50 mM glycine, 90 mM NaCl, pH 9.5 with a conductivity of 12.5 ms/cm). The supernatant (110 liters) was then applied to the column at 5 liters per hour.

The column ran dry after loading the supernatant. The gel was resuspended in Wash buffer and repacked. The repacked column was washed with 4 liters of 50 mM glycine, 90 mM NaCl, 5 mM EDTA, pH 10.0. The material was eluted with elution buffer (50 mM glycine, 5 mM EDTA (pH 9.9) with a final concentration of NaCl giving a conductivity of 30.2 cm/ms (approximately 270 mM NaCl)) at 100 ml per minute. The approximately 600 ml fractions were collected after the conductivity of the eluant reached the conductivity of the elution buffer. Fractions were analyzed by SDS-PAGE analysis as described above and fractions 1 through 10 were pooled.

The pooled fractions were then applied to a 2 liter phenyl Sepharose column (Pharmacia) that had been equilibrated by successive washes at 5 liters per hour with the following solutions: 3 liters of 0.5M NaOH; 3 liters of water; 3 liters of 2M urea, 50 mM glycine, pH 10.5; 1.5 liters of water; 3 liters of 0.1M HCl; and 3 liters of Equilibration buffer (50 mM glycine, 2.5M NaCl, 2 mM EDTA (pH 10.0) with a conductivity of 180 ms/cm). The pooled peak fractions, which had been adjusted to a conductivity of 180 ms/cm with NaCl and a pH of 10.0 with 0.5M NaOH, were loaded onto the phenyl sepharose column. Following the loading of the peak fractions, the column was washed with Equilibration buffer. The column was eluted with 6 liters of 50 mM glycine, 2 mM EDTA (pH 10.25) with a NaCl concentration giving the solution a conductivity of 96 ms/cm. The conductivity of the eluant was measured throughout the elution. The conductivity of the eluant upon starting the elution was 180 ms/cm. In the third fraction, the conductivity of the eluant dropped to 96 ms/cm. At this point, the elution buffer was changed to a buffer having the conductivity of 42 ms/cm. The eluant was collected through fraction number 8.

Cross-Linking Assay Using the Hybrid Fibrinogen-Fibronectin Protein

The ability of the purified fibrinogen-fibronectin hybrid protein to form transglutminase-catalyzed interchain cross links was assessed. The transglutaminase activity was provided by the addition of recombinant factor XIII and thrombin or by the addition of recombinant factor XIIIa.

A. Preparation of Factor XIII

Recombinant factor XIII was prepared essentially as described in co-pending U.S. patent application Ser. No. 07/927,196, which is incorporated by reference herein in its entirety. Briefly, factor XIII was isolated from a strain of the yeast Saccharomyces cerevisiae that had been transformed with an expression vector capable of directing the expression of factor XIII. The factor XIII-producing cells were harvested and lysed, and a cleared lysate was prepared. The lysate was fractionated by anion exchange chromatography at neutral to slightly alkaline pH using a column of derivatized agarose, such as DEAE FAST-FLOW SEPHAROSE (Pharmacia LKB Biotechnology, Piscataway, N.J.) or the like. Factor XIII was then precipitated from the column eluate by concentrating the eluate and adjusting the pH to between 5.2 and 5.5, such as by diafiltration against ammonium succinate buffer. The precipitate was then dissolved and further purified using conventional chromatographic techniques, such as gel filtration and hydrophobic interaction chromatography. The purified factor XIII was dialyzed, filtered, aliquotted and lyophilized. The factor XIIIa content was determined (Bishop et al., Biochemistry 29: 1861-1869, 1990, which is incorporated by reference herein in its entirety) by fluorometric assay of the dissolved, thrombin-activated material.

Factor XIII was activated to factor XIIIa by adding 2 U of thrombin per 100 mg of factor XIII. The factor XIII was dissolved in buffer (20 mM sodium borate (pH 8.3), 1 mM CaCl.sub.2). The thrombin was added, and the reaction was incubated at room temperature for twenty minutes.

B. Cross-Linking Assays

The level of cross-linking between the hybrid proteins was measured as a rise in the absorbance at 350 nm over time in reaction mixtures containing the hybrid protein, factor XIII and thrombin or the hybrid protein and factor XIIIa. Control reactions were prepared containing factor XIII and thrombin or factor XIIIa alone. Cross-linking reactions were carried out in 1 ml cuvettes. For cross-linking reactions containing factor XIII and thrombin, each reaction mixture was set up by placing 110 .mu.l containing 40 Units of factor XIII, 36.7 .mu.l containing 13 Units of factor XIII or 12.2 .mu.l containing 4 Units of factor XIII (described above) in one corner of the cuvette and 20 .mu.l containing 4 Units of thrombin (Sigma) in the opposite corner such that the solutions were not mixed. The reaction was initiated by the addition of 1 ml of 2 mg/ml hybrid protein in buffer (10 mM Tris (pH 7.6), 20 mM sodium borate, 140 mM NaCl, 10 mM CaCl.sub.2). The absorbance of each reaction was read at 350 nm with the addition of protein being the first absorbance point. For cross-linking reactions containing factor XIIIa, each reaction was set up by placing 110 .mu.l containing 40 Units of factor XIIIa, 36.7 .mu.l containing 13 Units of factor XIIIa or 12.2 .mu.l containing 4 Units of factor XIIIa in the cuvette and adding 1 ml of 2 mg/ml hybrid in buffer (10 mM Tris (pH 7.6), 140 mM NaCl, 10 mM CaCl.sub.2). The absorbance of the solution was read at 350 nm as described above. Analysis of the data generated from the absorbance time courses showed a sharp increase in absorbance in the presence of hybrid protein and the active transglutaminase relative to the rise in absorbance in the absence of hybrid protein (FIGS. 2-5). The results indicated that the hybrid protein is capable of transglutaminase-induced cross-linking.

From the foregoing it will be appreciated that, although specific embodiments of the invention have been described herein for the purpose of illustration, various modifications may be made without deviation from the spirit and scope of the invention. Accordingly, the invention is not to be limited except as by the following claims.