CLOT IMAGING AGENTS AND USE THEREOF

Technical Field

The present invention is directed to compositions and methods useful in the in vivo detection and imaging of blood clots. More specifically, the invention is directed to the use of labeled zymogen factor

XIII, factor Xllla and factor XIII derivatives for detecting the presence of blood clots.

Background of the Invention Coagulation-related disorders are a major cause of morbidity and mortality in western populations. Pathogenic blood coagulation may result in deep vein thrombosis, arterial thromboembolism, myocardial infarction, pulmonary embolism and stroke. Early diagnosis of venous thrombosis is difficult because outward symptoms are absent in a majority of patients in the early stage. Such early detection is desirable because venous thrombosis is associated with a high risk of subsequent pulmonary embolism.

Mavor et al. (The Lancet. March 25, 1972, 661- 663), Kakkar et al. (The Lancet. March 14, 1970, 540-542) and Caretta et al. (J. Nucl. Med. 18: 5-10, 1977) disclose the use of ¹²⁵ I-labeled fibrinogen for detecting venous thrombosis. However, the in vivo half-life of fibrinogen is short, the high levels of endogenous fibrinogen lead to a low signal-to-noise ratio and the use of fibrinogen purified from human blood creates a risk of viral transmission. Furthermore, it is difficult to detect thrombi older than about 12 to 24 hours using labeled fibrinogen, making this protein of little use in imaging existing arterial thrombi. Stratton and Ritchie (New Concepts Cardiac. Imaσ. 3.: 139-196, 1987) disclose methods of in vivo imaging using indium-Ill labeled platelets.

More recently, Rosenbrough et al. (Radiology 156: 515-517, 1985) proposed the use of ¹²⁵ I-labeled, fibrin-specific monoclonal antibodies for detecting thrombi. The use of antibodies foreign to the patient, however, may elicit an immune response. Despite its shortcomings, radiolabeled fibrinogen remains the standard for in vivo clot imaging.

There is therefore a need in the art for in vivo clot imaging agents with high specificity, low immunogenicity, and long half-life. The present invention provides such imaging agents and also provides other related advantages.

Summary of the Invention

Briefly stated, the present invention provides methods for detecting the presence of a blood clot in a patient. The methods of the present invention generally comprise the steps of (a) administering to a patient a diagnostic composition comprising a protein selected from the group consisting of factor XIII and factor XIII derivatives, and a physiologically acceptable carrier or diluent, wherein the protein is coupled to a substance capable of detection; and (b) detecting the substance and determining therefrom the presence of a clot. Within certain embodiments of the invention, the substance capable of detection may be a radionuclide, such as a gamma ray-emitting radionuclide, or a paramagnetic compound. Preferred gamma ray-emitting radionuclides include iodine-125, iodine-131, iodine-123, indium-Ill and technetium-99. Within another embodiment, the protein is a null mutant of factor XIII zymogen or factor Xllla. Preferred null mutants include factor XIII proteins characterized by the substitution or deletion of at least one amino acid residue within the Gln-Cys-Trp-Val active site region. Another aspect of the present invention is directed to diagnostic compositions comprising a protein selected from the group consisting of factor XIII and

factor XIII derivatives coupled to a substance capable of detection, and a physiologically acceptable carrier or diluent. Suitable substances capable of detection include radionuclides, particularly gamma ray-emitting radionuclides, and paramagnetic compounds as disclosed above. In one embodiment the factor XIII derivative is a null mutant of factor XIII zymogen or factor Xllla.

In yet another aspect, the present invention is directed to factor XIII derivatives characterized by the substitution or deletion of at least one amino acid acid residue within the Gln-Cys-Trp-Val active site region, wherein said factor XIII derivatives are substantially free of factor XIII biological activity. One such group of factor XIII derivatives is characterized by a cysteine to serine substitution at the active site. Within one embodiment the factor XIII derivative is a human factor XIII null mutant characterized by the substitution or deletion of at least one amino acid residue within the Gln-Cys-Trp-Val active site region. These factor XIII derivatives may be coupled to a substance capable of detection for use within diagnostic compositions and methods for detecting blood clots in patients.

Other aspects of the invention will become apparent upon reference to the following detailed description and attached drawings.

Brief Description of the Drawings

Figure 1 illustrates the amino acid sequences of the a and a' subunits of human factor XIII predicted from the cloned cDNA. The amino acids numbered 1 (serine) to 731 (methionine) represent the amino acids present in the mature a subunit. The amino acids numbered 38 (glycine) to 731 (methionine) represent the amino acids present in the mature a' subunit. The cDNA sequence is shown below the amino acid sequence.

The active site Cys at residue 315 (of the a subunit) is enclosed in parentheses. The arrow identifies

the cleavage site for the conversion of the factor XIII a subunit to the factor XIII a' subunit by thrombin.

Figure 2 illustrates the construction of plasmid pMVRl. Figure 3 illustrates the construction of plasmid

PAT-1.

Figure 4 illustrates the construction of plasmid pTRK4c.

Figure 5 illustrates the construction of plasmids pRS201 and pRS202.

Figure 6 illustrates the construction of the plasmid pRS215.

Figure 7 illustrates the construction of expression vectors for factor XIII. Factor 8 illustrates the construction of a yeast expression vector for factor Xllla.

Detailed Description of the Invention

Prior to setting forth the invention, it may be helpful to an understanding thereof to define certain terms used herein.

Biological activity: A function or set of functions performed by a molecule in a biological system or an in vitro facsimile thereof. In general, biological activities can include effector and catalytic activities. Effector activities include binding of a molecule to other molecules or to cells. Catalytic activities include proteolytic activity and cross-linking activity. Effector activity may enhance or be necessary for catalytic activity. The biological activity of factor XIII is characterized by the ability of the protein in its activated form to catalyze the cross-linking of fibrin polymers through the formation of intermolecular e(γ- glutamyl) lysine bonds, or to catalyze the cross-linking of other molecules in standard factor XIII activity

assays, such as incorporation of ³ H-histamine into N, N- dimethyl casein.

Factor XIII: Factor XIII is a plasma transglutaminase described by Lorand et al. , Prog. Hemost. Thro b. 5 : 245-290, 1980, which is incorporated herein by reference in its entirety. Factor XIII exists in blood and certain tissues in a catalytically inactive (zymogen) form. During the final stages of blood coagulation, thrombin converts the zymogen to an intermediate form, which is subsequently activated in the presence of calcium ions. The active protein catalyzes the cross-linking of, inter alia, fibrin polymers through the formation of intermolecular £(7-glutamyl)lysine bonds. As used herein, unless specifically noted, the term "factor XIII" includes the 2&2 tetrameric form of zymogen factor XIII, the a ₂ dimeric form and the catalytically active form of the protein ("factor Xllla") .

Factor XIII derivative: A protein having substantially the same amino acid sequence as naturally occuring factor XIII, but having one or more amino acid substitutions, deletions, additions or derivatizations. Factor XIII derivatives include proteins having the biological activity of factor XIII as well as proteins that are substantially free of factor XIII biological activity. Factor XIII derivatives are further characterized by the ability to bind fibrin. As used herein, "substantially the same amino acid sequence" indicates a minimum amino acid sequence homology of at least 60%, and preferably greater than 80% to the a or ' subunit of factor XIII.

According to the present invention, factor XIII (including zymogen factor XIII, factor Xllla and processing intermediates) and factor XIII derivatives are coupled to a substance capable of detection (a "label"), and the resultant labeled proteins are administered to patients for detecting and imaging occult blood clots.

Particularly preferred imaging agents include factor XIII derivatives that have been made catalytically inactive or non-activatable by deletion or substitution of one or more amino acids at or near the active site. These catalytically inactive factor XIII derivatives are referred to herein as "null mutants", and include null mutants of factor XIII zymogen as well as null mutants of factor Xllla. Null mutants useful within the present invention include factor XIII proteins in which the active site cysteine ^' at amino acid 314 of the factor XIII a subunit (amino acid 277 of the a' subunit; see Figure l) has been deleted or replaced with another amino acid, particularly serine, alanine, glycine, methionine, phenylalanine, leucine, isoleucine or valine. Particularly preferred null mutants include proteins containing factor XIII a subunits or a' subunits with a serine residue in place of the active site cysteine. Null mutants may also be generated by substitution or deletion of one or more amino acids flanking the active site cysteine, preferably within about 6 amino acids of the active site cysteine. In this regard, it is most preferred to substitute or delete at least one amino acid within the Gln-Cys-Trp-Val active site region (amino acids 313-316 as shown in Figure 1) . Absence of catalytic activity may be confirmed by standard in vitro procedures.

Other suitable factor XIII derivatives are produced by insertion, deletion or substitution of amino acids in such a manner as not to disrupt the ability of the molecule to bind fibrin. As will be appreciated by those skilled in the art, minor changes in amino acid sequence are generally preferred, such as conservative amino acid changes, small internal deletions or insertions, and additions or deletions at the ends of the molecule. Null mutants and other factor XIII derivatives having amino acid substitutions, deletions or additions are produced using standard recombinant DNA techniques.

Mutations are introduced into a factor XIII DNA sequence by standard methods of oligonucleotide-directed site- specific mutagenesis (reviewed by Sambrook et al.. Molecular Cloning: A Laboratory Manual. 2nd ed. , Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y., 1989, pages 15.1-15.113).

Factor XIII fragments may also be used to produce imaging agents. Suitable fragments may be generated by proteolysis of factor XIII or by expression of factor XIII DNA fragments. The fragments are then assayed for the ability to bind fibrin, for example by exposing the peptides to fibrin adsorbed to a plate or column and detecting bound peptide by immunological means. Fragments capable of binding to fibrin are labeled and used for clot imaging as described herein. It is preferred that such fragments lack the biological activity of factor XIII.

For use within the present invention, factor XIII may be obtained from a variety of sources, including plasma, tissue (e.g. placenta) , and genetically engineered cells, with the latter being a particularly preferred source. Methods for isolating factor XIII from these sources are known in the art. See, for example, the disclosures of Curtis and Lorand (Methods Enzvmol. 45: 177-191, 1976), Nakamura et al. (J. Bioche . 78: 1247- 1266, 1975), Zwisler et al. (U.S. Patent No. 3,904,751), Bohn et al. (U.S. Patent No. 3,931,399) and Bishop et al. (U.S. Patent Application Serial No. 07/270,714), which are incorporated herein by reference. A particularly preferred source of factor XIII and factor XIII derivatives is genetically engineered fungal cells, especially cells of the yeast Saccharomvces cerevisiae. Methods for expressing cloned DNA sequences in yeast and culturing transformed yeast cells in large scale fermentors are well known in the art. As described in more detail below, the protein may be secreted into the culture media or retained within the cells and recovered

from a cell lysate. Production of factor XIII in recombinant cells, including bacteria, yeast and cultured mammalian cells, has been described by Grundmann et al. (published Australian patent application 69896/87) and Davie et al. (U.S. Serial No. 174,287; EP 268,772), which are incorporated herein by reference. Briefly, a DNA sequence encoding factor XIII is operably linked to suitable promoter and terminator sequences, and this expression unit is inserted into a vector compatible with the chosen host cell. The vector is then inserted into the host cell and the resulting recombinant cells are cultured to produce factor XIII. Depending on the particular host cell and the expression system utilized, the factor XIII may either he- secreted from the cell or retained in the cytoplasm.

When using cells that do not secrete the factor XIII, the cells are removed from the culture medium (e.g., by centrifugation) and treated to produce a lysate. Typically, yeast cells are treated by mechanical disruption using glass beads to produce a crude lysate. Preferably, the crude lysate is centrifuged at low speed (e.g., 2,000 x g) , and the supernatant fraction is recovered. The supernatant is treated to produce a cleared lysate, typically by high-speed centrifugation (e.g., 20,000-30,000 x g) or filtration through a high molecular weight cutoff membrane. It is generally preferred that cell lysis and lysate clarification be carried out in the presence of a protease inhibitor, such as phenyl methyl sulfonyl fluoride. Factor XIII may also be obtained from cells which secrete it into the culture medium. Cells are transformed to express factor XIII subunits with an attached secretory signal sequence, which is removed from the factor XIII protein by proteolysis as it transits the secretory pathway of the host cell. For purification of the factor XIII, the cells are removed by centrifugation,

the medium is fractionated, and the factor XIII is recovered.

In a preferred embodiment, factor XIII is isolated from a biological fluid (such as plasma, cell culture media, cell lysates and tissue extracts) by adjusting the pH of the fluid to about pH 5.5 to 6.5, such as by buffer exchange, to form a precipitate. Preferred buffers for precipitation include low ionic strength solutions of heterocyclic polyamines, such as piperazine, spermidine, cadaverine and derivatives thereof adjusted to the desired pH. Piperazine and piperazine derivatives, such as piperazine sulfate, are particularly preferred. Other suitable buffers include biological buffers having a useful pH range around pH 6.0, including MES, phosphate, ADA and Bis-Tris. Buffers will generally be used at a concentration of between about 10 mM and 100 mM, preferably about 50 mM. Buffers may be obtained from commercial suppliers such as Sigma Chemical Co., St. Louis, MO. As used herein, the term "low ionic strength" includes solutions having a conductance of less than about 150 mS (equivalent to about 200 mM NaCl) .

When working with biological fluids containing complex mixtures of proteins it is generally preferred to first fractionate the biological fluid by anion exchange chromatography to produce an enriched fraction. Factor XIII is then precipitated from the resulting enriched fraction by buffer exchange. Typically, the biological fluid is concentrated (e.g. by dialysis) and passed over a column of an anion exchange medium and eluted using a suitable elution buffer. Suitable anion exchange media include derivatized dextrans, agarose, cellulose, polyaery1amide, specialty silicas, etc. PEI, DEAE.and QAE derivatives are preferred, with DEAE Sepharose (Pharmacia, Piscataway, NJ) being particularly preferred. As will be appreciated by those skilled in the art, fractionation can also be carried out in a batch process. Peak fractions are pooled for subsequent purification of factor XIII. It is

generally preferred at this stage to reduce the volume of the pooled fractions by concentration. A preferred method of concentration is precipitation of the factor XIII with 40% saturated (NH4)2S04- Factor XIII is then precipitated by adjusting the pH of the factor XIII preparation as described above. In a preferred embodiment, an enriched fraction, prepared by ammonium sulfate precipitation, is dissolved in a small volume of buffer at a pH between about 7.0 and 8.0. The resulting solution is then dialyzed against 50 mM piperazine, pH adjusted to 5.8 with HCl, containing 10 mM EDTA, 5 mM 2-merceptoethanol (2-ME) , 0.02% NaN ₃ at 4"C to produce a crystalline precipitate. The precipitate is recovered by centrifugation, redissolved and dialyzed a second time against piperazine buffer.

As will be evident to those skilled in the art, precipitation of factor XIII from other biological fluids will be carried out in substantilly the same manner, i.e., by dialyzing the fluid against the precipitation buffer to produce a precipitate.

If desired, additional purification may be achieved through the use of conventional chemical separation techniques, including ion exchange chromatography, size-exclusion chromatography, etc. In many instances it will be desirable to remove residual proteases from the preparation. In one embodiment, this is achieved by dissolving the precipitated factor XIII to produce a solution, typically in a low ionic strength buffer at slightly alkaline pH, then fractionating the solution by size exclusion chromatography. Suitable chromatographic media in this regard include cross-linked dextran, polyacrylamide and specialty silicas with hydrophillic coatings or bonded phases. Sephacryl S-400 (Pharmacia) is a particularly preferred polyacrylamide- based medium. In a preferred embodiment, the piperazine precipitate is redissolved in 25 mM Tris-HCl, pH 7.4, 100 mM NaCl, 5 mM EDTA, 10 mM 2-mercaptoethanol, 0.19 M

glycine, 0.02% NaNβ. The resulting solution is fractionated on a Sephacryl S-400 column. Peak fractions are pooled and dialyzed against an appropriate buffer (e.g. 2% sucrose in 0.19 M glycine pH 7.4) and lyophilized for storage. Factor XIII prepared in this way is typically greater than 99% pure and pyrogen-free.

The purified factor XIII or factor XIII derivative is then coupled to a substance capable of detection. Particularly preferred substances capable of detection include radiosotopes, preferably gamma ray emitters such as iodine-125, iodine-131, iodine-123, technetium-99, and indium-Ill. High energy emitters (e.g. ⁹⁹ Tc, ¹²³ I) are generally preferred for imaging of arterial clots. Other suitable substances capable of detection include paramagnetic compounds, including spin labels such as sterically hindered free radical nitroxide compounds that may be used in conjunction with magnetic resonance imaging techniques; electron-dense materials that absorb sufficient X radiation to allow detection by tomography; and electron-dense elements (e.g. nickel) that scatter X-rays anomalously.

Methods of coupling proteins to a variety of such labels are known in the art. For example, radioiodination may be accomplished by the lactoperoxidase method of Greenberg and Shuman (J. Biol. Chem. 257; 6096- 6101, 1982) or by the use of N-chloro-benzenesulfonamide derivatized polystyrene beads (U.S. Patents 4,436,718 and 4,448,764). Factor XIII may be labeled with technetiu by art-recognized procedures, such as the method of Fritzberg et al. (Proc. Natl. Acad. Sci. USA 85: 4025-4029, 1988) or the method disclosed in U.S. Patent No. 4,830,847. Methods for attaching magnetic resonance contrast agents to targetting molecules are disclosed in U.S. Patents 4,656,026, 4,735,210 and 4,827,945, which are incorporated herein by reference. The use of magnetic resonance imaging in medical studies is reviewed by Budinger and Lauterbur (Science 226: 288-298, 1984).

For use in detection of thrombosis, the compounds of the present invention are combined with a physiologically acceptable diluent and administered to a patient at risk for thrombosis or having an existing thrombus. Suitable diluents include sterile, pyrogen-free water, saline and physiologically compatible buffers. In general, the diagnostic compositions will be formulated to contain the labeled factor XIII or factor XIII derivative at a concentration of about 0.1 mg/ml to about 10 mg/ml, with concentrations in the range of about 0.5 to 2.0 mg/ml particularly preferred. It will be understood, however, that the labeled molecules can be packaged in lyophilized form and reconstituted with a suitable diluent immediately before administration. Alternatively, the lyophilized factor XIII or factor XIII derivative may be provided in unlabeled form, then reconstituted and labeled shortly before administration. The latter method is preferred when working with radioisotope labels having short half- lives. Methods of formulation are reviewed in Remingtons Pharmaceutical Sciences. 16th edition. Mack Publishing Co., 1982, which is incorporated herein by reference.

Diagnostic compositions prepared according to the present invention are administered by intravenous injection. The actual amount administered will be determined by such factors as the weight of the patient, condition to be diagnosed, type of label (e.g. radioisotope or other label, energy of emission if a radioisotope, etc.), and amount of label per unit of protein. Such determination is within the level of ordinary skill in the art. When using radioisotope labels, the compositions are typically formulated to provide a total of about 10 ⁶ -10 ¹⁰ cpm, more preferably about 10 ⁷ -10 ⁸ cpm, with a dose of between about 0.1-10 μg of protein per kg of body weight, with doses in the range of 1-3 μg/kg particularly preferred. For example, a 70 kg patient typically will be given an intravenous injection

of about 100-200 μg of protein labeled with 5 x 10 ⁷ cpm of ¹²⁵ I.

As noted above, these diagnostic compositions are particularly useful in detecting thrombosis in patients previously identified as being at risk for this condition, although it will be evident that the compostions can be used any time that thrombosis is suspected of occuring or is expected to occur. Due to the long in vivo half-life of factor XIII relative to fibrinogen, the compositions of the present invention are particularly well-suited for administration prior to surgery, thus permitting ongoing monitoring of the patient during surgery and for two weeks or more after surgery with only a single dose. For ongoing use in screening at- risk patients, the use of zymogen factor XIII or derivatives having a long in vivo half-life is preferred.

In addition to those undergoing surgery

(particularly hip surgery) , patients at risk for thrombosis include those with malignant disease, a history of thrombosis or acquired thrombotic disease and the elderly.

Following administration of the diagnostic composition, the patient is screened regularly (e.g. daily) to detect newly formed thrombi. Detection is achieved by scanning the patient's body or a portion thereof with an instrument capable of detecting the particular label employed. Suitable detection instruments are known in the art and include gamma cameras, gamma counters and magnetic resonance imaging equipment. In a typical application, the protein will be labeled with a gamma-emitting radionuclide and administered intravenously to a patient at risk for thrombosis. The patient's legs will periodically be scanned with a gamma counter. Sequential sites on both legs are counted, and the results are compared between the two limbs and between adjacent sites on one limb. An increase in radiation of 20% as compared to an adjacent site or the same site on the other

limb is generally indicative of the presence of a thrombus. For imaging of existing thrombi, the diagnostic composition, preferably comprising factor Xllla or a null mutant thereof, is administered to the patient, and the patient's body is scanned using a gamma camera.

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

Example 1 Preparation of Recombinant Factor XIII A. Cloning of a cDNA Encoding the a Subunit of Human Factor XIII

A λgtll expression library containing cDNAs prepared from human placenta mRNA was screened for the a subunit of human factor XIII. An ¹²⁵ I-labeled affinity- purified rabbit antibody (specific activity = 6 X 10 ⁶ cpm/μg) was used to screen filters containing phage plated at a density of 1.5 X 10 ⁵ plaques per 150 mm plate. Six positive clones were isolated by screening approximately 3 X 10 ⁶ phage, and each positive phage was plaque-purified.

Plaque-purified clones were then screened with a ³² P-labeled oligonucleotide probe ( ⁵ ' CTC CAC GGT GGG CAG GTC GTC CTC G ^J ) which codes for the ammo acid sequence of Ala-Glu-Asp-Asp-Leu-Pro-Thr-Val-Glu that is present in the activation peptide of the a subunit (Takagi and Doolittle, Biochemistry 13.: 750-756, 1974). The nucleotide sequence for the probe was selected by employing the most common codon usage for amino acids for a number of different human proteins (Chen and Barker, Trends in Genetics ι 221-223, 1985). The oligonucleotide was labeled with ³² P to a specific activity of 1.1 X 10 ⁸ cpm/μg.

Phage DNA was prepared from positive clones by the liquid culture lysis method (Siljavy et al., in Experiments with Gene Fusions. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y., pp. 140-141, 1984), followed by centrifugation and banding on a cesium chloride step gradient. cDNA inserts were isolated by digestion of the phage DNA with Eco RI endonuclease, and the 5' and 3' ends of each insert were sequenced. One of the clones with a large cDNA insert (λHFXIIIa3.77, deposited with ATCC under Accession No. 40261) was selected for further sequence analysis. This insert contained three internal Eco RI sites, giving rise to four

cDNA fragments upon digestion of the phage DNA with Eco RI. These fragments and several additional restriction fragments were subcloned into plasmids pUC9 and pUC19. Additional restriction fragments from other cDNA inserts were subcloned into M13mpl0, M13mpll, M13mpl8, or M13mpl9 in order to obtain overlapping sequences. The cDNA inserts were then sequenced by the dideoxy method (Sanger et al., Proc. Natl. Acad. Sci. USA 74: 5463-5467, 1977) using [α ³⁵ S]dATP and buffer gradient gels (Biggin et al., Proc. Natl. Acad. Sci. USA 80: 3963-3965, 1983). Digestions with nuclease BAL-31 were performed to generate five additional fragments that provided overlapping sequences with the Eco RI restriction fragments.

The cloned sequence of 3831 base pairs was found to code for the entire amino acid sequence of the mature a subunit of human factor XIII that circulates in blood (Figure 1). The a subunit is composed of 731 amino acids, starting with an amino-terminal sequence of Ser-Glu-Thr- Ser, as reported by Takagi and Doolittle (ibid.). The carboxyl-terminal Met (nucleotides 2281-2283) is followed by a stop codon (TGA) , 1535 base pairs of noncoding sequence, and a potential polyadenylation or processing signal of AATAAA. The polyadenylation sequence was located 14 nucleotides upstream from the poly (A) tail of 10 nucleotides. The poly (A) tail was present only in a second cDNA clone, designated λHFXIIIa3.82.

B. Cloning of a cDNA Encoding the b Subunit of Human Factor XIII A λgtll expression library containing cDNAs prepared from human liver mRNA was screened for the b subunit of human factor XIII with an ¹²⁵ I-labeled, affinity-purified rabbit antibody. The purified antibody was labeled with Na ¹²⁵ I to a specific activity of 4 X ιo ⁶ cpm/μg, and was used to screen filters containing phage plated at a density of 1.5 X 10 ⁵ plaques per 150 mm plate.

Nine positive clones were isolated by screening 2 X 10 ⁶ phage, and each was plaque purified.

Phage DNA was prepared from positive clones by the liquid culture lysis method (Silhavy et al., ibid.), followed by centrifugation and banding on a cesium chloride step gradient. The clone with the largest cDNA insert (approximately 2.2 kilobases) was designated λHFXIIIb2.2 and was selected for further study. The phage DNA from λHFXIIIb2.2 was cut with Eco RI to isolate the 2.2 kb cDNA insert. This fragment was subcloned into plasmid pUC9, which had been linearized by digestion with Eco RI, to construct plasmid pUC9b2.2 (deposited with ATCC under Accession No. 40260) . Appropriate restriction fragments from the insert were then subloned into M13mpl0 or M13mpl8 for sequencing by the dideoxy method (Sanger et al, ibid.) using [α ³⁵ S]dATP and buffer gradient gels (Biggin et al., ibid.). Controlled digestions with nuclease BAL-31 were performed to generate suitable fragments, which provided overlapping sequences with the restriction fragments. All sequence determinations were performed on both strands of DNA at least three times. The cDNA insert was found to be composed of 2180 base pairs coding for the entire amino acid sequence for the b subunit of Factor XIII that circulates in blood.

C. Construction of Yeast Expression Vectors

Factor XIII subunits and subunit derivatives were expressed in yeast through the use of expression vectors based on the yeast 2-miσron plasmid. These vectors expressed factor XIII using the ADH2-4c promoter and TPI terminator, and further contained the Schizosaccharomvces pombe triose phosphate isomerase (POT i) gene (Kawasaki and Bell, EP 171,142; Murray et al., U.S. Patent No. 4,801,542) as a selectable marker. Yeast cells transformed with these vectors can be cultured in glucose-containing culture media without the need for

additional selective conditions. Vector construction is illustrated in Figures 2-7.

Plasmids pDPOT and pSPOT were derived from plasmid pCPOT (ATCC No. 39685) , which comprises the entire 2-micron plasmid, the leu2-d gene, pBR322 vector sequences and the Schizosaccharomyces pombe POT1 gene as a selectable marker. Plasmid pCPOT was digested with Sph I and Bam HI to isolate the 10.8 kb fragment. Plasmid pBR322 (Boliver et al., Gene 2_: 95-113, 1977) was digested with Sph I and Bam HI to isolate the 186 bp fragment. The 186 bp Sph I-Bam HI fragment was joined with the 10.8 kb pCPOT fragment by ligation. The resultant plasmid was designated pDPOT (shown in Figure 7) . Plasmid pDPOT was further modified by the insertion of a polylinker. Plasmid pDPOT was digested with Bam HI, and a linker constructed from oligonucleotides ZC1887 (5' GCC TCG AGG AGC TCG AGA TCT GCA TG 3') and ZC1888 (5' CAG ATC TCG AGC TCC TCG AGG CCA TG 3') was inserted. . The resultant vector, designated pSPOT, contains Sph I, Xho I, Sac I and Bgl II sites in the inserted polylinker.

Plasmid pMVRl, used as the source of the TPIl promoter in subsequent vector constructions, comprises the TPIl promoter, an alpha-1-antitrypsin (AAT) cDNA and the TPIl terminator in the vector pIC7Rl . Plasmid pMVRl was constructed as shown in Figure 2. Plasmid pIC7 (Marsh et al., Gene 32: 481-486, 1984) was digested with Eco RI, the fragment ends blunted with DNA polymerase I (Klenow fragment) , and the linear DNA recircularized using T4 DNA ligase. The resulting plasmid was used to transform E. coli strain RR1. Plasmid DNA was prepared from the transformants and screened for the loss of the Eco RI site. A plasmid having the correct restriction pattern was designated pIC7RI . The TPIl promoter fragment was obtained from plasmid pTPICIO (Alber and Kawasaki, J. Mol. APPI. Genet. χι 419, 1982). This plasmid was cut at the unique Kpn I site within the TPIl gene and the TPIl coding region was removed by treatment with nuclease BAL-31.

Kinased Eco RI linkers (GGAATTCC) were added to .the fragment, which was then digested with Bgl II and Eco RI to yield a 0.9 kb TPIl promoter fragment. This fragment was joined to plasmid YRp7' (Stinchcomb et al., Nature 282: 39-43, 1979) which had been cut with Bgl II and Eco

RI. The resultant plasmid, pTE32, was cleaved with Eco RI and Bam HI to remove a portion of the tetracycline resistance gene. The linearized plasmid was then recircularized by the addition of a kinased Eco RI-Bam HI oligonucleotide adapter (5' AAT TCA TGG AG 3' and 5' GAT

CCT CCA TG 3'). The resultant plasmid, pTEA32, was digested with Bgl II and Eco RI to isolate the 900 bp TPIl promoter fragment. This fragment was joined with ^' pICl9H

(Marsh et al., ibid) which had been linearized by digestion with Bgl II and Eco RI. The resultant plasmid was designated pICTPI. Plasmid pFATPOT (deposited with

ATCC as an S^. cerevisiae transformant in strain E18 under

Accession No. 20699) was digested with Sph I and Hind III to isolate the 1750 bp fragment comprising the partial TPIl promoter, a cDNA encoding human alpha-1-antitrypsin, and the TPIl terminator. Plasmid pICTPI was digested with

Nar I and Sph I to isolate the 1.1 kb fragment comprising the partial TPIl promoter and lacZ' coding sequence.

Plasmid pIC7RI*, which had been digested with Hind III and Nar I to isolate the 2.5 kb vector fragment, was joined with the l.l kb fragment derived from plasmid pICTPI and the 1.75 kb fragment derived from pFATPOT in a three-part ligation to construct pMVRl.

The ADH2-4— promoter was derived from plasmids pBR322-ADR2-BSa (Williamson et al., Cell 3: 605-614, 1981) and YRp7-ADR3-4 ^c (Russell et al Nature 304: 652-654, 1983). An Eco RI site was placed just 3' to the translation start codon of the ADH2 promoter (derived from plasmid pBR322-ADR2-BSa) by in vitro mutagenesis. Following the mutagenesis, 5' flanking sequences which confer the ADH2-4- phenotype were used to replace the analogous sequences of the ADH2 promoter. The 2.2 kb Bam

HI fragment from pBR322-ADR2-BSa, containing the ADH2 structural gene and 5' flanking sequences, was ligated with M13mpl9 which had been liniearized with Bam HI. The orientation of the isert was determined by restriction analysis. single-stranded template DNA was made from the resultant phage clone. Site-specific in vitro mutagenesis (Zoller and Smith, DNA 3.: 479-488, 1984) was carried out on the template using oligonucleotides ZC87 ( ⁵ TCC CAG TCA CGA CGT ³ ') and ZC237 ( ⁵ ' GCC AGT GAA TTC CAT TGT GTA TTA ³ ) to loop out the structural portion of the ADH2 gene, fusing the 5' flanking sequence, including the translation start signal, with the Eco RI site of the M13mpl9 polylinker. The replicative form DNA of the mutagenized phage was made and cut with Bam HI and Eco RI to isolate the 1.2 kb promoter fragment. This fragment was ligated into pUC13, which had been linearized by digestion with Bam HI and Eco RI, to generate plasmid p237WT. To change the p237WT promoter to the "promoter up" mutant ADH2-4—. a 1.1 kb Bam Hl-Sph I partial promoter fragment from YRp7-ADR3-4 ^c was subcloned into the vector fragment of p237WT cut with Bam HI and Sph I. The resultant plasmid was designated p237-4 ^c .

Referring to Figure 3, the ADH2 promoter from p237WT was modified to create a "universal" ADH2 promoter. The promoter was first subcloned into pCPOT. Plasmid pCPOT was digested to completion with Bam HI and Sal I to isolate the approximately 10 kb linear vector fragment. Plasmid pMVRl was cut with Eco RI and Xho I to isolate the 1.5 kb AAT cDNA-TPIl terminator fragment. The 1.2 kb ADH2 promoter fragment was isolated from plasmid p237WT as a Bam HI-Eco RI fragment and ligated with the 1.5 kb AAT cDNA-TPIl terminator fragment and the linearized pCPOT in a three-part ligation to yield the plasmid designated pAT- 1. The ADH2 promter present in plasmid pAT-1 was then modified to create a "universal" promoter by removing the translation start codon and fusing the promoter to an

Eco RI site. Plasmid pAT-1 was digested with Sph I and Bam HI to isolate the 190 bp fragment comprising the ADH2 promoter from the Sph I site through the Bam HI site of the AAT cDNA. This fragment was ligated into M13mpl8 which had been linearized by digestion with Bam HI and Sph I. A positive clone was confirmed by restriction analysis. Template DNA was made from the positive clone. Oligonucleotide ZC410 (5' CGT AAT ACA GAA TTC CCG GG 3') was designed to replace the ADH2 translation start signal and pUC18 polylinker sequences with a single Eco RI site fused to the M13mpl8 polylinker at the Sma I site. The template was subjected to in vitro mutagenesis using the two-primer method (Zoller and Smith, ibid., 1984) using oligonucleotides ZC410 and ZC87. Positive clones were confirmed by dideoxy sequencing through the fusion point. For ease of manipulation, the 175 bp Sph I-Eco RI mutagenized promoter fragment was ligated into pUCl9 which had been linearized with Sph I and Eco RI. The resultant plasmid, designated p410ES, comprises the 3' portion of a "universal" ADH2 promoter.

The full "universal" ADH2-4— promoter was then constructed using the mutagenized ADH2 promoter fragment from p410ES and the ADH2-4°- promoter fragment from p237- 4 ^C . Plasmid p410ES was digested with Sph I and Eco RI to isolate the 175 bp partial ADH2 promoter fragment. Plasmid p237-4c was cut with Bam HI and Sph I to isolate the 1 kb partial ADH2-4- promoter fragment. The ADH2-4- promoter was reconstructed in a three-part ligation with the Bam HI-Sph I promoter fragment from p237-4 ^c , the Sphl- Eco RI promoter fragment from p410ES, and pUC13 which had been linearized by digestion with Bam HI and Eco RI. The resultant plasmid, p410-4 ^c , comprised a "universal" ADH2- 4.— promoter.

As illustrated in Figure 4, the "universal" promoter from plasmid p410-4 ^c was used to replace the TPIl promoter present in plasmid pMVRl. Plasmid p410-4 ^c was cut with Bam HI and Eco RI to isolate the 1.2 kb ADH2-4-

promoter fragment. Plasmid pMVRl was digested to completion with Eco RI and partially digested with Bgl II to isolate the 4.2 kb AAT-TPI1 terminator-vector fragment. These two fragments were ligated to form the plasmid designated pTRK4c.

Plasmid pTRK4c was linearized by digestion with Eco RI and Pst I. The vector fragment was recovered and ligated with a linker constructed from oligonucleotides ZC1056 (5' AAT TAG ATC TGC A 3') and ZC1057 (5' GATCT 3'). The resultant plasmid was designated pRS185 (shown in Figure 8) .

D. Construction of Factor XIII a Subunit Expression Units The Factor XIII a subunit (asFXIII) cDNA was modified by in vitro mutagenesis to replace the 3' noncoding region with an Xho I site (Figure 5) . The phage clone λHFXIIIa3.82 was digested to completion with Pst I to isolate the 2.3 kb fragment comprising the asFXIII cDNA. This ^' fragment was ligated with pUC18, which had been linearized by digestion with Pst I. The resultant plasmid, pUClδ #9, was determined to have the 2.3 kb Pst I insert in the anti-sense orientation. The asFXIII cDNA insert present in pUClδ #9 comprised 19 bp 5' to the translation start, the asFXIII coding region and 120 bp 3' to the translation stop.

The asFXIII cDNA insert was then isolated and subcloned into the vector pUCllδ (obtained from J. Vierira and J. Messing, Waksman Institute of Microbiology, Rutgers, Piscataway, N.J.). The 2.3 kb asFXIII insert was isolated from a Pst I digest of pUC18 #9 and subcloned into the pUC118 Pst I site. The resulting plasmid was transformed into E. coli strain JM109. A positive clone, which contained the 2.3 kb asFXIII insert in the anti- sense orientation, was designated pRS201.

The 120 bp 3' untranslated region of the asFXIII cDNA was removed, and an Xho I site was inserted 3' to the

translation stop codon by site directed mutagenesis. Plasmid pRS201 was transformed into £. coli strain MV1193, which had been infected with phage K07, and single- stranded template DNA was isolated after 18 hours. Oligonucleotide ZC1113 ( ⁵ ' CGA CCT TCC ATG TGA TAA CTC GAG AAG CTG AGA TGA AC ³ ) was designed to remove the 3' untranslated region following the asFXIII coding sequence and to introduce an Xho I site immediately 3' to the translation stop. The single-stranded template of pRS201 was subjected .to in vitro mutagenesis by the method of Zoller and Smith (ibid., 1984) using the mutagenic oligonucleotide ZC1113. A positive clone, confirmed by restriction analysis, was designated pRS202 (Figure 5) .

As shown in Figure 6, the 2.2 kb asFXIII cDNA fragment from plasmid pRS202 and the ADH2-4— promoter were placed in plasmid pIC7RI . Plasmid pTRK4c was digested to completion with Eco RI and Sal I to isolate the 4 kb fragment comprising the ADH2-4— promoter, the TPIl terminator and the pIC7RI vector sequences. Oligonucleotides ZC1056 ( ⁵ ' AAT TAG ATC TGC A ³ ') and ZC1057 ( ⁵ GAT CT ³ ) were kinased and annealed to from an adapter with an Eco RI adhesive end, a Bgl II site and a Pst I adhesive end. The Eco RI adhesive end of the adapter, upon ligation to another Eco RI adhesive end, destroys the Eco RI site. The 2.2 kb Pst I-Xho I asFXIII fragment, isolated from plasmid pRS202, was joined with the ZC1056/ZC1057 adapter and the 4. Icb pTRK4c fragment in a three-part ligation. The resultant plasmid was designated pRS215.

E. Expression of Factor XIII 2 Dimer

The expression unit in plasmid pRS215 was inserted into the yeast/E. coli shuttle vector pDPOT. Plasmid pDPOT was linearized by digestion with Bam HI and treated with calf alkaline phosphatase to prevent recircularization. Plasmid pRS215 was digested with Sst I and Pst I to isolate the 0.65 kb fragment comprising a

portion of the ADH2-4 ^S promoter and the ZC1056/ZC1057 adapter. Plasmid pRS215 was also digested with Pst I and Bgl II to isolate the 2.3 kb fragment comprising the asFXIII cDNA and TPIl terminator. Plasmid p410-4 ^c was digested with Bam HI and Sst I to isolate the 0.55 kb fragment comprising a 5' portion of the ADH2-4— promoter. The 0.55 kb Bam HI-Sst I fragment derived from p410-4 ^c , the 0.65 kb Sst I-Pst I fragment and the 3.1 kb Pst I-Bgl II fragment from pRS215 were joined with the Bam HI- linearized pDPOT in a four part ligation. Two plasmids were identified as having the correct insert, in opposite orientations. Plasmid pD15 contained the expression unit with the ADH2-4 - promoter distal to the POT1 gene in the vector. Plasmid pD16 contained the expression unit with the ADH2-4 ^C - promoter proximal to the pOTl gene in the vector (Figure 7) .

Plasmids pD15 and pD16 were transformed into Saccharomyces cerevisiae strain ZM118, and the transformants were grown in YEPD (containing, per liter, 20 g glucose, 10 g yeast extract and 20 g peptone) at 30*C. Assays of cell lysates (using the activity assay of Curtis and Lorand, Meth. Enzymology 45: 177-191, 1976) showed that both pD15 and pD16 transformants produced factor XIII.

Example 2—Preparation of Recombinant Factor Xllla

The activated form of factor XIII was expressed directly in yeast cytosol by deletion of the activation peptide coding sequence in vitro and substitution by an intiator methionine codon. The Pst I - Sma I sequence encompassing the activation peptide sequence and upstream non-coding sequence was excised and replaced by complementary oligonucleotides compatible with the restriction sites at the respective ends. The sequence positions a methionine in phase with the initial glycine residue of the factor Xllla molecule and replaces the G residue of the gly codon lost by restriction digestion.

The -3 position is an A residue for optimal expression, and a Bam HI site was introduced into the adapter for screening. Oligonucleotides ZC2315 (5' GGA TCC ATC GAC TAA GAT GG 3') and ZC2316 (5' CCA TCT TAG TCG ATG GAT CCT GCA) were annealed to form the 5'-most portion of the coding sequence. The adapter was ligated to a Sma I + Pst I fragment from pRS218 (constructed by inserting the expression unit from pRS215 into pIC19R as a Hind III - Xho I fragment) , which contained the partial factor XIII a subunit and TPIl terminator sequences, and Pst I-cut pRS185. A plasmid with the correct insert orientation was designated pRS271. Plasmid pRS271 was digested with Xho I and Sst I, and the expression cassette was subcloned into the yeast expression vector pSPOT, which had been digested with Xho I and Sac I, to yield the plasmid pRS274 (Figure 8).

S. cerevisiae strain ZM118 was transformed with pRS274 and grown for 48 hours in YEPD. The cell lysate was examined for factor Xllla protein by Western blot analysis and was found to contain the protein.

Example 3—Preparation of Factor XIII Null Mutant

A factor XIII null mutant sequence was constructed in vitro by using site-directed mutagenesis to change the active site cysteine-314 was to a serine residue. Oligonucleotide ZC1760 (5' AGC TCT GAG AAT CCA GTT CGA TAC GGC CAA TCT TGG GTT TTT GCT GGT GTC 3') was prepared and used to change the TGC Cys-314 codon to TCT, encoding serine. A second base change introduced a Taq I site for screening recombinants. Mutagenesis was carried out using a single-stranded pRS202 template and the one primer method. Single-stranded DNA was elongated in vitro. transformed into E.. coli MV1193, and colonies were screened by hybridization against the labeled ZC1760 oligonucleotide. Positive colonies were further screened by restriction with Taq I. The mutation and the integrity of the entire region were confirmed by DNA sequencing, and

a correct plasmid was selected and designated pRS1760. The Stu I-Bam HI fragment of pRS1760 was subcloned into the Stu I-Bam HI site of pRS215, and the expression cassette was subcloned into the yeast vector pSPOT at the Xho I site to yield pRS276.

S. cerevisiae strain ZM118 was transformed with pRS276 and grown for 48 hours. Cell lysates were examined for factor XIII protein by Western blot analysis and were demonstrated to contain high levels of the protein. Assay of transglutaminase activity by fluorometric assay (Hornyak et al.. Biochemistry 28: 7326-7332, 1989; incorporated herein by reference) revealed no enzymatic activity.

Example 4—Purification of Factor XIII

Saccharomyces cerevisiae strain ZM118 (a MATa/MATα diploid homozygous for leu2-3, 112 ura3 tpil: :URA3 ⁺ [cir*]) was transformed with pD16. The transformed cells were inoculated at approximately 0.1 g/1 and cultured in a pH 5.5 medium containing 22.7 g/1 yeast extract, 22.5 g/1 (NH ₄ ) ₂ S0 , 6.5 g/1 KH ₂ P0 , 3 g/1 MgSθ4 ^* 7H2θ, 0.5% glucose, trace elements and vitamins with a glucose feed for 39 hours. After 39 hours, 3.75 g/1 ethanol was added over 1 hour, followed by an ethanol feed beginning at 2.5 g/l/hr. and increasing over 23 hours to a final rate of 3.75 g/l/hr. The pH of the culture was maintained at approximately 5.5 by the addition of 2M NaOH. The culture (approximately 60 liters) was grown at 30*C for 63 hours to a final cell density of approximately 50 g/1.

Cell cultures were harvested by concentration using a 0.2 μ cellulose ester hollow fiber cartridge (Microgon, Laguna Hills, CA) . The final concentrate typically contained 600-3000 g wet weight of yeast cells (concentration > 50% wet weight) in deionized H2O.

The concentrated cells were then lysed. A maximum of 400 g (wet weight) of cells was diluted to 40%

wet weight in lysis buffer (50 mM Tris HCl, pH 7.0, 150 mM NaCl, 5 mM EDTA, 10 mM 2-ME) . 0.5 M PMSF in absolute ethanol was added to the cell slurry to a final concentration of 1 mM. The cells were lysed using a Dynomill (Glen Mills, Inc., Maywood, NJ) in continuous flow mode. The Dynomill was pre-cooled to 0 ^* C or less, and all solutions were at 0-8*C. The cell suspension was combined with 0.5 liter of acid-washed 500 μ glass beads in a 0.6 liter container and lysed at 3000 rpm using a flow rate of 150 ml/min. An additional one liter of lysis buffer was pumped through the container and added to the cell lysate. 0.5 M PMSF was added to a final concentration of 1 mM, and the pH of the lysate was adjusted to 7.8 with 2 M NaOH. The lysate was then clarified by centrifugation.

The pH was adjusted to 7.0 with 2M HCl. The lysate was then centrifuged at 3895 x g at 4'C for at least 40 minutes, and the pellets were discarded. The supernatant fractions were then concentrated and dialyzed against three volumes of lx equilibration buffer (50 mM Tris pH 7.4, 10 mM 2-ME, 5 mM EDTA) to a conductivity of less than 50 mS using a tangential flow system (Pellicon, Millipore, Bedford, MA) and 10 ft ² of polysulfone membrane (PTHK, Millipore) with a 100 kD nominal molecular weight cutoff. Dialysis was carried out at 10*C using an inlet pressure of 20-25 psi, an average transmembrane pressure of 4 psi, a flux of 400-500 ml/minute and a crossflow rate of approximately 20 liters/minute.

The concentrated, dialyzed lysate was then fractionated by chromatography on a column of DEAE Sepharose. A 23.5 cm high x 5.25 cm radius (2.0 liter) DEAE column was equilibrated with equilibration buffer until the conductivity was less than 45 S using a flow rate of approximately 45 ml/minute. The sample was loaded on the column at a flow rate of 16 ml/minute, then the column was washed with equilibration buffer until the absorbance of the eluate at 280 n was less than 10% of

the absorbance at full scale. Factor XIII was eluted from the column with 0.12 M imidazole pH 5.8 containing 5 mM PMSF, 10 mM 2-ME and 5 mM EDTA. Peak fractions were pooled, adjusted to 5 mM PMSF and kept at 4'C. Pooled factor XHI-containing fractions were precipitated by addition of (NH4)2Sθ4 to 40% of saturation.

Factor XIII was then precipitated using piperazine buffer. The (NH4)2S04 mixture was centrifuged at 3958 x g, and the supernatant fraction was discarded. The pellet was -dissolved in a minimal volume of cold 25 mM Tris pH 7.4, 100 mM NaCl, 5 mM EDTA, 10 mM 2-ME, 0.19 M Glycine, 0.02% NaN ₃ (TAGS). The pH of the solution was maintained between 7.0 and 8.0 by the addition of 2 M Tris pH 7.5. The solution was centrifuged at 7649 x g to remove insoluble material. The supernatant was then dialyzed against 50 mM piperazine pH 5.8, 5 mM EDTA, 10 mM 2-ME, 0.02% NaN ₃ overnight at 4*C using 50,000 kD molecular weight cutoff dialysis tubing (Spectra/Por) . The dialyzed solution was then centrifuged at 7649 x g for thirty minutes. The resulting pellet was redissolved in a minimal volume of cold TAGS, maintaining the pH as necessary by addition of 2M Tris pH 7.8. The solution was again centrifuged, and the resulting supernatant was recovered and dialyzed against piperazine buffer overnight at 4'C. The solution was again centrifuged, and the pellet was dissolved in a minimal volume of TAGS as above.

Final purification was achieved by gel filtration. Factor XIII in TAGS (30 ml) was loaded onto a 1.5 liter (95 cm high x 2.25 cm radius) Sephacryl S-400 (Pharmacia) column. Separation was achieved using TAGS as the running buffer at a flow rate of 3.0 ml/min. Absorbance of the column eluate was monitored at 280 n . The main factor XIII peak eluted between 1100-1350 ml. The peak fractions were pooled and concentrated by multiple centrifugations in a membrane concentrator (Centriprep, Amicon, Danvers, MA) to a final concentration of approximately 20 mg/ml. The concentrate was then

dialyzed overnight at 4°C against 0.19 M glycine pH 7.4 containing 2% sucrose using 50,000 kD molecular weight cutoff dialysis tubing (Spectra/Por) .

For storage, the dialyzed material was put through a 0.2 μ filter and aliquoted into vials. The samples were frozen quickly on a sheet of dry ice and stored at -80*C. Lyophilization of the frozen samples was carried out at -20*C for 48 hours. The lyophilized factor XIII was sealed under argon and stored dessicated at -20*C.

Example 5—Labeling of Factor XIII

Five mg of lyophilized, purified factor XIII was resuspended in 1 ml sterile H2O. Four μl of this solution (20 μg factor XIII) was diluted to 500 μl with phosphate buffered saline, and the solution was placed on ice. To the factor XIII solution were added 2 μl ¹³¹ ι (sodium iodide, 2.4 mCi, obtained from ICN, Costa Mesa, CA) and 2 Iodobeads ^™ (Pierce Chemical Co., Rockford, II.). The mixture was incubated on ice for 23 minutes, then fractionated by gel filtration on a G-25 Sepharose column (Pharmacia) that had been previously equilibrated with RIP buffer (20 M Tris-HCl pH 8.0, 100 mM NaCl, 1 mM EDTA, 0.5% NP-40, 0.5% Na deoxycholate, 10 mM Nal, 1% BSA) . The column was washed with RIP buffer, and peak fractions (determined by radioactivity) were pooled (2 ml total volume) . 91% of the counts in the pooled peak were found to be precipitable with trichloracetic acid. Subsequent characterization of the preparation by ELISA and activity assay indicated a factor XIII concentration of approximately 1 μg/ml in the pooled peak. 1.8 ml of this preparation (2.49 x 10 ⁸ cpm) was taken for use in in vivo clot detection studies.

Example 6—In Vivo Labeling of Clots

The ability of labeled factor XIII to specifically bind to clots was assayed in a dog. A 17 kg

male dog was given fourteen drops of potassium iodide daily in food for three days prior to the study. Methylprednisolone sodium succinate (Solu-Medrol; 62.5 mg) was given intramuscularly 24 hours before and on the day of the study. The dog was sedated with -500 mg of sodium thiamyl, intubated and anesthetized with 1-2% halothane. An arterial line was inserted in the left femoral artery for blood gas measurement. Blood gases were drawn at 30- minute intervals, and blood pressure was monitored throughout the study.

The right carotid artery and right femoral artery were isolated, -5 cm of each was tied off, and crush injury was performed with forceps -20 times on each. 100-200 units of thrombin was injected into each injured site, the arteries were allowed to fill with blood from the proximal end, then they were again tied off and maintained for two hours. The left carotid artery was isolated and lifted to the skin surface, but left uninjured to serve as a control. 1.8 ml of ¹³ ^-labeled factor XIII

(approximately 2 μg) was injected through an indwelling catheter in a limb vein. Thirty seconds post-injection the ligatures on the occluded arteries were removed. Two hours post-injection the animal was sacrificed with KC1. The injured artery segments were excised, as was a similar length of the left (control) carotid artery segment. Proximal (closest to the heart) and distal (farthest from the heart) ends of the vessel segments were marked.

The vessel segments were opened longitudinally and fixed in formaldehyde overnight. The fixed vessel walls were cut into one-half cm segments. Clots were isolated from the wall segments and sequentially numbered starting with 1 at the proximal end. Each segment of artery and clot was weighed and counted in a gamma counter. Tissue samples were also obtained from skeletal muscle, liver, spleen, kidney and urine. Biodistribution data, shown in Tables 1-3, indicate that labeled factor

XIII binds preferentialy to blood clots, making it useful for use in detecting and localizing clot formation.

Table 1 Biodistribution of ¹³¹ I-Factor XIII in Arteries and Clots

(CPS/gm of sample)

Sam le Tissue X 2.

Left Carotid Artery 117 206 Right Carotid Artery 560 625 Right Carotid Clot 1960 1797 Right Femoral Artery 445 1885 Right Femoral Clot 1419 3173

Table 2 Biodistribution of ¹³¹ I-Factor XIII in Tissue

Average Tissue CPS/gm

Blood (120'Sample) 561

Muscle 93

Liver 620

Spleen 589 Kidney 539

Urine 1609

Left Carotid Artery 113

Right Carotid Artery 380

Carotid Clot 1139 Femoral Artery 1810

Femoral Clot 2604

Table 3 Biodistribution of ¹³¹ I-Factor XIII: Tissue Ratios

Tissue Ratio

Ave. Left Carotid/Blood 0.20

Ave. Right Carotid/Blood 0.68

Ave. Carotid Clot/Blood 2.03

Ave. Femoral Artery/Blood 3.23

Ave. Femoral Clot/Blood 4.64

The ability of labeled activated factor XIII (factor Xllla) to bind a clot was evaluated in a rabbit model of arterial thrombosis.

Factor Xllla was first labeled with ¹²⁵ ι using an enzymobead (Bio-Rad, Richmond, CA) . 1.79 mg of labeled factor XIII, 0.5 mg/ml in TANEP (0.1 M Tris-Acetate pH 7.5, 0.15 M NaCl, 1 mM EDTA, 0.1% PEG), was adjusted to 14 mM CaCl2 and activated by the addition of 0.0334 ml of 150 U/ml thrombin. Following incubation at 37*C for 30 minutes, the thrombin was inhibited by the addition of 0.0134 ml Shaw peptide (D-phenylalanyl-L-prolyl-L-arginine chloromethyl ketone, 87 μM in 1 mM HCl; Calbioche , La Jolla, CA) , and the mixture was serially dialyzed against TNE (50 mM Tris pH 7.4, 100 mM NaCl, 1 mM EDTA). 0.01 mg of iodinated factor Xllla (2.16 x 10 ⁶ cpm) in 0.4 ml of 30 mM sodium citrate pH 7.48 containing 50 mM NaCl, 0.5 mM EDTA and 30% glycerol was injected intravenously into an anaesthetised, ventilated rabbit which had undergone percutaneous catheter placement of a clot forming coil into the femoral artery 2 hours earlier.

Two hours post injection the animal was sacrificed and the clot-containing coil was removed. The clot was weighed and counted in a gamma counter. An aliquot of blood was also counted. The counts per gram of clot were some nine times higher than the counts per gram of blood at sacrifice as shown in Table 4 below. This

observation demonstrates the ability of the activated form of Factor XIII to localize a clot.

Table 4 Tissue - % of dose iniected/gram

Clot 1.59

Blood 0.17

Although specific embodiments of the invention have been described herein for purposes of illustration, various modifications can be made without deviating from the spirit and scope of the invention. Accordingly, the invention is not limited except as by the appended claims.