AU5854096A - Kunitz type protease inhibitors - Google Patents

Description

KUNITZ TYPE PROTEASE INHIBITORS

Background of the Invention The plasma, or serine, proteases of the blood contact system are known to be activated by interaction with negatively charged surfaces. For example, tissue injury during surgery exposes the vascular basement membrane, causing interaction of the blood with collagen, which is negatively charged at physiological pH. This induces a cascade of proteolytic events, leading to production of plasmin, a fibrinolytic protease, and consequent blood loss.

Perioperative blood loss of this type can be particularly severe during cardiopulmonary bypass (CPB) surgery, in which the patient's blood flow is diverted to an artificial heart-lung machine. CPB is an essential component of a number of life-saving surgical procedures. For example, in the United States, it is estimated that 300,000 patients every year undergo coronary artery bypass grafts involving the use of CPB. Although necessary and generally safe, CPB is associated with a significant rate of morbidity, some of which may be attributed to a "whole body inflammatory response" caused by activation of plasma protease systems and blood cells through interactions with the artificial surfaces of the heart-lung machine (Butler et al., Ann. Thorac. Surg. 55:552 (1993); Edmunds et al . , J. Card. Surg. 8:404 (1993)). For example, during extracorporeal circulation, exposure of blood to negatively charged surfaces of the artificial bypass circuit, e.g., plastic surfaces in the heart-lung machine, results in direct activation of plasma factor XII.

Factor XII is a single-chain 80 kDa protein that circulates in plasma as an inactive zymogen. Contact with negatively charged nonendothelial surfaces, like those of the bypass circuit, causes surface-bound factor XII to be autoactivated to the active serine protease factor Xlla. See Colman, Agents Actions Suppl . 42:125 (1993) . Surface-activated factor Xlla then processes prekallikrein (PK) to active kallikrein, which in turn cleaves more Xlla from XII in a reciprocal activation reaction that results in a rapid amplification of the contact pathway. Factor Xlla can also activate the first component of complement Cl, leading to production of the anaphylatoxin C5a through the classical complement pathway.

The CPB-induced inflammatory response includes changes in capillary permeability and interstitial fluid accumulation. Cleavage of high molecular weight kininogen (HK) by activated kallikrein generates the potent vasodilator bradykinin, which is thought to be responsible for increasing vascular permeability, resulting in edema, especially in the lung. The lung is particularly susceptible to damage associated with CPB, with some patients exhibiting what has been called "pump lung syndrome" following bypass, a condition indistinguishable from adult respiratory distress. See Johnson et al . , J. Thorac. Cardiovasc. Surg. 107:1193 (1994).

Post-CPB pulmonary injury includes tissue damage thought to be mediated by neutrophil sequestration and activation in the microvasculature of the lung. (Butler et al . , supra; Johnson, et al., supra) . Activated factor XII can itself stimulate neutrophil aggregation. Factor Xlla-generated kallikrein, and complement protein C5a generated by Factor Xlla activation of the complement cascade, both induce neutrophil chemotaxis, aggregation and degranulation. See Edmunds et al., supra (1993). Activated neutrophils may damage tissue through release of oxygen-derived free-radicals, proteolytic enzymes such as elastase, and metabolites of arachidonic acid. Release of neutrophil products in the lung can cause changes in vascular tone, endothelial injury and loss of vascular integrity.

Intrinsic inhibition of the contact system occurs through inhibition of activated Xlla by Cl-inhibitor (Cl- INH) . See Colman, supra. During CPB, this natural inhibitory mechanism is overwhelmed by massive activation of plasma proteases and consumption of inhibitors. A potential therapeutic strategy for reducing post-bypass pulmonary injury mediated by neutrophil activation would, therefore, be to block the formation and activity of the neutrophil agonists kallikrein, factor Xlla, and C5a by inhibition of proteolytic activation of the contact system.

Protease inhibitor therapywhichpartially attenuates the contact system is currently employed clinically in CPB. Aprotinin, also known as basic pancreatic protease inhibitor (BPPI) , is a small, basic, 58 amino acid polypeptide isolated from bovine lung. It is a broad spectrum serine protease inhibitor of the Kunitz type, and was first used during bypass in an attempt to reduce the inflammatory response to CPB. See Butler et al . , supra. Aprotinin treatment results in a significant reduction in blood loss following bypass, but does not appear to significantly reduce neutrophil activation. Additionally, since aprotinin is of bovine origin, there is concern that repeated administration to patients could lead to the development of an immune response to aprotinin in the patients, precluding its further use.

The proteases inhibited by aprotinin during CPB appear to include plasma kallikrein and plasmin. (See, e.g., Scott, e al., Blood 69:1431 (1987)). Aprotinin is an inhibitor of plasmin (K, of 0.23nM), and the observed reduction in blood loss may be due to inhibition of fibrinolysis through the blocking of plasmin action. Although aprotinin inhibits plasma kallikrein, (K; of 20nM) , it does not inhibit activated factor XII, and consequently only partially blocks the contact system during CPB.

Another attractive protease target for use of protease inhibitors, such as those of the present invention, is factor Xlla, situated at the very first step of contact activation. By inhibiting the proteolytic activity of factor Xlla, kallikrein production would be prevented, blocking amplification of the contact system, neutrophil activation and bradykinin release. Inhibition of Xlla would also prevent complement activation and production of C5a. More complete inhibition of the contact system during CPB could, therefore, be achieved through the use of a better Xlla inhibitor.

Protein inhibitors of factor Xlla are known. For example, active site mutants of ovantitrypsin that inhibit factor Xlla have been shown to inhibit contact activation in human plasma. See Patston et al . , J. Biol . Chem. 265:10786 (1990). The large size and complexity (greater than 400 amino acid residues) of these proteins present a significant challenge for recombinant protein production, since large doses will almost certainly be required during CPB. For example, although it is a potent inhibitor of both kallikrein and plasmin, nearly 1 gram of aprotinin must be infused into a patient to inhibit the massive activation of the kallikrein-kinin and fibrinolytic systems during CPB. The use of smaller, more potent Xlla inhibitors such as the corn and pumpkin trypsin inhibitors (Wen, et al . , Protein Exp. & Purif . 4:215 (1993); Pedersen, e al., J. Mol . Biol . 236:385 (1994)) could be more cost-effective than the large c-,-antitrypsins, but the infusion of high doses of these non-mammalian inhibitors could result in immunologic reactions in patients undergoing repeat bypass operations. The ideal protein Xlla inhibitor is, therefore, preferably, small, potent, and of human sequence origin. One candidate for an inhibitor of human origin is found in circulating isoforms of the human amyloid β- protein precursor (APPI) , also known as protease nexin-2. APPI contains a Kunitz serine protease inhibitor domain known as KPI (Kunitz Protease Inhibitor) . See Ponte et al., Nature, 331:525 (1988); Tanzi etal., Nature 331:528 (1988) ; Johnstone et al. , Biochem. Biophys . Res. Commun. 163:1248 (1989); Oltersdorf et al . , Nature 341:144 (1989) . Human KPI shares about 45% amino acid sequence identity with aprotinin. The isolated KPI domain has been prepared by recombinant expression in a variety of systems, and has been shown to be an active serine protease inhibitor. See, for example, Sinha, et al . , J. Biol . Chem. 265:8983 (1990) . The measured in vi tro K, of KPI against plasma kallikrein is 45nM, compared to 20nM for aprotinin.

Aprotinin, KPI, and other Kunitz-type serine protease inhibitors have been engineered by site-directed mutagenesis to improve inhibitory activity or specificity. Thus, substitution of Lys¹⁵ of aprotinin with arginine resulted in an inhibitor with a K_; of 0.32nM toward plasma kallikrein, a 100-fold improvement over natural aprotinin. See PCT application No. 89/10374. See also Norris et al., Biol . Chem. Hoppe Seyler 371:3742 (1990) . Alternatively, substitution of position 15 of aprotinin with valine or substitution of position 13 of KPI with valine resulted in elastase inhibitors with K_;s in the 100 pM range, although neither native aprotinin nor native KPI significantly inhibits elastase. See Wenzel et al., in: Chemistry of Peptides and Proteins, Vol . 3 , (Walter de Gruyter, Berlin, New York, 1986); Sinha et al . , supra. Methods for substituting residues 13, 15, 37, and 50 of KPI are shown in general terms in European Patent Application No. 0 393 431, but no specific sequences are disclosed, and no protease inhibition data are given.

Phage display methods have been recently used for preparing and screening derivatives of Kunitz-type protease inhibitors. See PCT Application No. 92/15605, which describes specific sequences for 34 derivatives of aprotinin, some of which were reportedly active as elastase and cathepsin inhibitors. The amino acid substitutions in the derivatives were distributed throughout almost all positions of the aprotinin molecule.

Phage display methods have also been used to generate

KPI variants that inhibit factor Vila and kallikrein.

See Dennis et al . , J. Biol . Chem. 269:22129 and 269:22137

(1994) . The residues that could be varied in the phage display selection process were limited to positions 9-11, 13-17, 32, 36 and 37, and several of those residues were also held constant for each selection experiment. One of those variants was said to have a K, of 1.2nM for kallikrein, and had substitutions at positions 9 (Thr→Pro) , 13 (Arg→Lys) , 15 (Met→Leu) , and 37 (Gly→Tyr) . None of the inhibitors was tested for the ability to inhibit factor Xlla.

It is apparent, therefore, that new protease inhibitors that can bind to and inhibit the activity of serine proteases are greatly to be desired. In particular it is highly desirable to prepare peptides, based on human peptide sequences, that can inhibit selected serine proteases such as kallikrein; chymotrypsins A and B; trypsin; elastase; subtilisin; coagulants and procoagulants, particularly those in active form, including coagulation factors such as factors Vila, IXa, Xa, Xla, and Xlla; plasmin; thrombin; proteinase-3; enterokinase; acrosin; cathepsin; urokinase; and tissue plasminogen activator. It is also highly desirable to prepare novel protease inhibitors that can ameliorate one or more of the undesirable clinical manifestations associated with enhanced serine protease activity, for example by reducing pulmonary damage or blood loss during CPB.

Summary of the Invention The present invention relates to peptides that can bind to and preferably exhibit inhibition of the activity of serine proteases. Those peptides can also provide a means of ameliorating, treating or preventing clinical conditions associated with increased activity of serine proteases. Particularly, the novel peptides of the present invention preferably exhibit a more potent and specific (i.e., greater) inhibitory effect toward serine proteases of interest in comparison to known serine protease inhibitors. Examples of such proteases include: kallikrein; chymotrypsins A and B; trypsin; elastase; subtilisin; coagulants and procoagulants, particularly those in active form, including coagulation factors such as factors Vila, IXa, Xa, Xla, and Xlla; plasmin; thrombin; proteinase-3; enterokinase; acrosin; cathepsin; urokinase; and tissue plasminogen activator. In achieving the inhibition of serine protease activity, the invention provides protease inhibitors that can ameliorate one or more of the undesirable clinical manifestations associated with enhanced serine protease activity, for example, by reducing pulmonary damage or blood loss during CPB.

The present invention relates to protease inhibitors comprising the following amino acid sequences:

X'-Val-Cys-Ser-Glu-Gln-Ala-Glu-X'-Gly-X'-Cys-Arg- Ala-X⁴-X⁵-X⁶-X⁷-Trp-Tyr-Phe-Asp-Val-Thr-Glu-Gly- Lys-Cys-Ala-Pro-Phe-X⁸-Tyr-Gly-Gly-Cys-X⁹-X^I0-Xⁿ- X¹²-Asn-Asn-Phe-Asp-Thr-Glu-Glu-Tyr-Cys-Met-Ala- Val-Cys-Gly-Ser-Ala-He, wherein: X¹ is selected from Glu-Val-Val-Arg-Glu-, Asp, or Glu; X² is selected from Thr, Val, He and Ser; X³ is selected from Pro and Ala; X⁴ is selected from Arg, Ala, Leu, Gly, or Met; X⁵ is selected from He, His, Leu, Lys, Ala, or Phe; X⁶ is selected from Ser, He, Pro, Phe, Tyr, Trp, Asn, Leu, His, Lys, or Glu; X⁷ is selected from Arg, His, or Ala; X⁸ is selected from Phe, Val, Leu, or Gly; X⁹ is selected from Gly, Ala, Lys, Pro, Arg, Leu, Met, or Tyr; X¹⁰ is selected from Ala, Arg, or Gly; X¹¹ is selected from Lys, Ala, or Asn; and X¹² is selected from Ser, Ala, or Arg.

The invention relates more specifically to protease inhibitors comprising the following amino acid sequences: X¹-Val-Cys-Ser-Glu-Gln-Ala-Glu-X-Gly-X³-Cys-Arg- Ala-X⁴-X⁵-X⁶-X⁷-Trp-Tyr-Phe-Asp-Val-Thr-Glu-Gly- Lys-Cys-Ala-Pro-Phe-X⁸-Tyr-Gly-Gly-Cys-X⁹-X^,0-X^u- X^l2-Asn-Asn-Phe-Asp-Thr-Glu-Glu-Tyr-Cys-Met-Ala- Val-Cys-Gly-Ser-Ala-He, wherein X¹ is selected from Glu-Val-Val-Arg-Glu- , Asp, or Glu; X² is selected from Thr, Val, He and Ser; X³ is selected from Pro and Ala; X⁴ is selected from Arg, Ala, Leu, Gly, or Met; X^s is selected from He, His, Leu, Lys, Ala, or Phe; X⁶ is selected from Ser, He, Pro, Phe, Tyr, Trp, Asn, Leu, His, Lys, or Glu; X⁷ is selected from Arg, His, or Ala; X⁸ is selected from Phe, Val, Leu, or Gly; X⁹ is selected from Gly, Ala, Lys, Pro, Arg, Leu, Met, or Tyr; X¹⁰ is selected from Ala, Arg, or Gly; X¹¹ is selected from Lys, Ala, or Asn; X¹² is selected from Ser, Ala, or Arg; provided that when X⁴ is Arg, X⁶ is He; when X⁹ is Arg, X⁴ is Ala or Leu; when X⁹ is Tyr, X⁴ is Ala or X^s is His; and either X⁵ is not He; or X⁶ is not Ser; or X⁹ is not Leu, Phe, Met, Tyr, or Asn; or X¹⁰ is not Gly; or X¹¹ is not Asn; or X¹² is not Arg.

Another aspect of this invention provides protease inhibitors wherein at least two amino acid residues selected from the group consisting of X⁴, X⁵, X⁶, and X⁷ defined above differ from the residues found in the naturally occurring sequence of KPI. Another aspect of this invention provides protease inhibitors wherein X¹ is Asp or Glu, X² is Thr, X³ is Pro, and X¹² is Ser. Yet another aspect of this invention provides protease inhibitors wherein X¹ is Glu, X² is Thr, X³ is Pro, X⁴ is Met, X^s is He, X⁶ is Ser, X⁷ is Arg, x⁸ is Phe, X⁹ is Gly, X¹⁰ is Gly, and X¹¹ is Asn. Another aspect of this invention provides protease inhibitors wherein X¹ is Asp, X² is Thr, X³ is Pro, X⁴ is Arg, X⁵ is He, X⁶ is He, X⁷ is Arg, x⁸ is Val, X⁹ is Arg, X¹⁰ is Ala, and X" is Lys. Another aspect of this invention provides protease inhibitors wherein X¹ is Glu-Val-Val-Arg-Glu- , X² is Thr, X³ is Pro, X⁴ is Met, X⁵ is He, X⁶ is Ser, X⁷ is Arg, x⁸ is Phe, X⁹ is Gly, X¹⁰ is Gly, X¹¹ is Asn, and X¹² is Ala. Another aspect of this invention provides protease inhibitors wherein X¹ is Glu-Val-Val-Arg-Glu-, X² is Thr, X³ is Pro, X⁴ is Met, X⁵ is He, X⁶ is Ser, X⁷ is Arg, x⁸ is Phe, X⁹ is Gly, X¹⁰ is Gly, X¹¹ is Ala, and X¹² is Arg. Another aspect of this invention provides protease inhibitors wherein X¹ is Glu, X² is Thr, X³ is Pro, X⁴ is Met, X⁵ is He, X⁶ is Ser, X⁷ is Arg, x⁸ is Phe, X⁹ is Gly, X¹⁰ is Ala, X¹¹ is Asn, and X¹² is Arg. Another aspect of this invention provides protease inhibitors wherein X¹ is Glu-Val-Val-Arg-Glu-, X² is Thr, X³ is Pro, X⁴ is Met, X⁵ is He, X⁶ is Ser, X⁷ is Arg, x⁸ is Phe, X⁹ is Gly, X¹⁰ is Arg, X¹¹ is Asn, and X¹² is Arg. Another aspect of this invention provides protease inhibitors wherein X¹ is Glu- Val-Val-Arg-Glu- , X² is Thr, X³ is Pro, X⁴ is Met, X⁵ is He, X⁶ is Ser, X⁷ is Arg, x⁸ is Val, Leu, or Gly, X⁹ is Gly, X¹⁰ is Gly, X¹¹ is Asn, and X¹² is Arg. Another aspect of this invention provides protease inhibitors wherein X¹ is Glu-Val-Val-Arg-Glu- , X² is Thr, X³ is Pro, X⁴ is Met, X^s is He, X⁶ is Ser, X⁷ is Ala, x⁸ is Phe, X⁹ is Gly, X¹⁰ is Gly, X¹¹ is Asn, and X¹² is Arg. Another aspect of this invention provides protease inhibitors wherein X¹ is Glu- Val-Val-Arg-Glu-, X² is Thr, Val, or Ser, X³ is Pro, X⁴ is Ala or Leu, X⁵ is He, X⁶ is Tyr, X⁷ His, X⁸ is Phe, X⁹ is Gly, X¹⁰ is Gly, X¹¹ is Ala, and X¹² is Arg.

Yet another aspect of this invention provides protease inhibitors wherein X² is Thr, and X⁴ is Ala. Another aspect of this invention provides protease inhibitors wherein X² is Thr, and X⁴ is Leu. Another aspect of this invention provides protease inhibitors wherein X² is Val, and X⁴ is Ala. Another aspect of this invention provides protease inhibitors wherein X² is Ser, and X⁴ is Ala. Another aspect of this invention provides protease inhibitors wherein X² is Val, and X⁴ is Leu. Another aspect of this invention provides protease inhibitors wherein X² is Ser, and X⁴ is Leu.

Yet another aspect of this invention provides protease inhibitors wherein X¹ is Glu-Val-Val-Arg-Glu-, X² is Thr, X³ is Pro, X⁴ is Leu, X^s is Phe, X⁶ is Lys, X⁷ is Arg, X⁸ is Phe, X⁹ is Gly, X¹⁰ is Gly, X¹¹ is Ala, and X¹² is Arg. Another aspect of this invention provides protease inhibitors wherein X¹ is Glu-Val-Val-Arg-Glu-, X² is Thr, X³ is Pro, X⁴ is Leu, X⁵ is Phe, X⁶ is Lys, X⁷ is Arg, X* is Phe, X⁹ is Tyr, X¹⁰ is Gly, X" is Ala, and X¹² is Arg. Another aspect of this invention provides protease inhibitors wherein X¹ is Glu-Val-Val-Arg-Glu-, X² is Thr, X³ is Pro, X⁴ is Leu, X^s is Phe, X⁶ is Lys, X⁷ is Arg, X⁸ is Phe, X⁹ is Leu, X¹⁰ is Gly, Xⁿ is Ala, and X¹² is Arg. The present invention also relates to protease inhibitors comprising the following amino acid sequences:

X¹-Val-Cys-Ser-Glu-Gin-Ala-Glu-Thr-Gly-

Pro-Cys-X²-Ala-X³-X-X⁵-X⁶-Trp-Tyr-Phe-Asp- Val-Thr-Glu-Gly-Lys-Cys-Ala-Pro-Phe-Phe-

Tyr-Gly-Gly-Cys-Gly-Gly-Asn-Arg-Asn-Asn- Phe-Asp-Thr-Glu-Glu-Tyr-Cys-Met-Ala-Val- Cys-Gly-Ser-Ala-He, wherein: X¹ is selected from Glu-Val-Val-Arg-Glu- and Asp-Val-Val-Arg-Glu- ; X² is selected from Arg and Lys; X³ is selected from Met, Arg, Ala, Leu, Ser, Val; X⁴ is selected from He and Ala; X⁵ is selected from Ser, He, Ala, Pro, Phe, Tyr, and Trp; and X⁶ is selected from Arg, Ala, His, Gin, and Thr; provided that: when X² is Arg, X³ is Leu, and X⁴ is He, X⁵ cannot be Ser; and also provided that either X³ is not Met; or X⁴ is not He; or X⁵ is not Ser; or X⁶ is not Arg. Another aspect of this invention provides protease inhibitors wherein X³ is Arg or Met, and X⁵ is Ser or He. Yet another aspect of this invention provides protease inhibitors wherein X⁵ is selected from Phe, Tyr and Trp. Another aspect of this invention provides protease inhibitors wherein X³ is Ala or Leu.

A further aspect of this invention provides an isolated DNA molecule comprising a DNA sequence encoding a protease inhibitor of the invention. Another aspect of this invention provides an isolated DNA molecule comprising a DNA sequence encoding the protease inhibitor that further comprises an isolated DNA molecule operably linked to a regulatory sequence that controls expression of the coding sequence of the protease inhibitor in a host cell. Another aspect of this invention provides an isolated DNA molecule comprising a DNA sequence encoding the protease inhibitor operably linked to a regulatory sequence that controls expression of the coding sequence of the protease inhibitor in a host cell that further comprises a DNA sequence encoding a secretory signal peptide. That secretory signal peptide may preferably comprise the signal sequence of yeast alpha-mating factor. Another aspect of this invention provides a host cell transformed with any of the DNA molecules defined above. Such a host cell may preferably comprise E. coli or a yeast cell. When such a host cell is a yeast cell, the yeast cell may be selected from Saccharomyces cerevisiae and Pichia pastoris .

Another aspect of this invention provides a method for producing a protease inhibitor of the present invention, comprising the steps of culturing a host cell as defined above and isolating and purifying said protease inhibitor.

A further aspect of this invention provides a pharmaceutical composition, comprising a protease inhibitor of the present invention together with a pharmaceutically acceptable sterile vehicle.

An additional aspect of this invention provides a method of treatment of a clinical condition associated with increased activity of one or more serine proteases, comprising administering to a patient suffering from said clinical condition an effective amount of a pharmaceuti¬ cal composition comprising a protease inhibitor of the present invention together with a pharmaceutically acceptable sterile vehicle. That method of treatment may preferably be used to treat the clinical condition of blood loss during surgery.

Yet another aspect of this invention provides a method for inhibiting the activity of serine proteases of interest in a mammal comprising administering a therapeutically effective dose of a pharmaceutical composition comprising a protease inhibitor of the present invention together with a pharmaceutically acceptable sterile vehicle.

Another aspect of this invention provides a method for inhibiting the activity of serine proteases of interest in a mammal comprising administering a therapeutically effective dose of a pharmaceutical composition comprising a protease inhibitor of the present invention together with a pharmaceutically acceptable sterile vehicle, wherein said serine proteases are selected from the group consisting of: kallikrein; chymotrypsins A and B; trypsin; elastase; subtilisin; coagulants and procoagulants, particularly those in active form, including coagulation factors such as factors Vila, IXa, Xa, Xla, and Xlla; plasmin; thrombin; proteinase-3; enterokinase; acrosin; cathepsin; urokinase; and tissue plasminogen activator.

A further aspect of this invention relates to protease inhibitors comprising the following amino acid sequences:

X¹-Val-Cys-Ser-Glu-Gin-Ala-Glu-Thr-Gly-Pro-Cys- Arg-Ala-X²-X³-X⁴-Arg-Trp-Tyr-Phe-Asp-Val-Thr-Glu- Gly-Lys-Cys-Ala-Pro-Phe-Phe-Tyr-Gly-Gly-Cys-X^s- Gly-Asn-Arg-Asn-Asn-Phe-Asp-Thr-Glu-Glu-Tyr-Cys- Met-Al -Val-Cys-Gly-Ser-Ala-He, wherein X¹ is selected from Glu-Val-Val-Arg-Glu-, Asp, or Glu; X² is selected from Ala, Leu, Gly, or Met; X³ is selected from He, His, Leu, Lys, Ala, or Phe; X⁴ is selected from Ser, He, Pro, Phe, Tyr, Trp, Asn, Leu, His, Lys, or Glu; X^s is selected from Gly, Ala, Lys, Pro, Arg, Leu, Met, or Tyr; provided that when X⁵ is Arg, X² is Ala or Leu; when X^s is Tyr, X² is Ala or X³ is His; and either X³ is not He; or X⁴ is not Ser; or X⁵ is not Leu, Phe, Met, Tyr, or Asn. Another aspect of this invention provides a protease inhibitor as defined above wherein X¹ is Glu, X² is Met, X³ is He, X⁴ is He, and X⁵ is Gly.

The invention also relates more specifically to protease inhibitors comprising the following amino acid sequences: Glu-Val-Val-Arg-Glu-Val-Cys-Ser-Glu-Gln-Ala-Glu- Thr-Gly-Pro-Cys-Arg-Ala-X¹-X²-X³-Arg-Trp-Tyr-Phe- Asp-Val-Thr-Glu-Gly-Lys-Cys-Ala-Pro-Phe-Phe-Tyr- Gly-Gly-Cys-X⁴-Gly-Asn-Arg-Asn-Asn-Phe-Asp-Thr- Glu-Glu-Tyr-Cys-Met-Ala-Val-Cys-Gly-Ser-Ala-Ile, wherein X¹ is selected from Ala, Leu, Gly, or Met; X² is selected from He, His, Leu, Lys, Ala, or Phe; X³ is selected from Ser, He, Pro, Phe, Tyr, Trp, Asn, Leu, His, Lys, or Glu; X⁴ is selected from Gly, Arg, Leu, Met, or Tyr; provided that when X¹ is Ala, X² is He, His, or Leu; when X¹ is Leu, X² is He or His; when X¹ is Leu and X² is He, X³ is not Ser; when X¹ is Gly, X² is He; when X⁴ is Arg, X¹ is Ala or Leu; when X⁴ is Tyr, X¹ is Ala or X² is His; and either X¹ is not Met, or X² is not He, or X³ is not Ser, or X⁴ is not Gly.

A further aspect of this invention provides a protease inhibitor as defined above wherein X¹ is Met, X³ is Ser, and X⁴ is Gly. Another aspect of this invention provides a protease inhibitor wherein X² is selected from His, Ala, Phe, Lys, and Leu. Another aspect of this invention provides a protease inhibitor wherein X² is His. Another aspect of this invention provides a protease inhibitor wherein X² is Ala. Another aspect of this invention provides a protease inhibitor wherein X² is Phe. Another aspect of this invention provides a protease inhibitor wherein X² is Lys. Another aspect of this invention provides a protease inhibitor wherein X² is Leu. Another aspect of this invention provides a protease inhibitor wherein X¹ is Met, X² is He, and X⁴ is Gly.

Yet another aspect of this invention provides a protease inhibitor wherein X³ is He. Another aspect of this invention provides a protease inhibitor wherein X³ is Pro. Another aspect of this invention provides a protease inhibitor wherein X³ is Phe. Another aspect of this invention provides a protease inhibitor wherein X³ is Tyr. Another aspect of this invention provides a protease inhibitor wherei'n X³ is Trp. Another aspect of this invention provides a protease inhibitor wherein X³ is Asn. Another aspect of this invention provides a protease inhibitor wherein X³ is Leu.

An additional aspect of this invention provides a protease inhibitor wherein X³ is Lys. Another aspect of this invention provides a protease inhibitor wherein X³ is His. Another aspect of this invention provides a protease inhibitor wherein X³ is Glu. Another aspect of this invention provides a protease inhibitor wherein X¹ is Ala. Another aspect of this invention provides a protease inhibitor wherein X² is He. Another aspect of this invention provides a protease inhibitor wherein X³ is Phe, and X⁴ is Gly. Another aspect of this invention provides a protease inhibitor wherein X³ is Tyr, and X⁴ is Gly. Another aspect of this invention provides a protease inhibitor wherein X³ is Trp, and X⁴ is Gly.

Yet another other aspect of this invention provides a protease inhibitor wherein X³ is Ser or Phe, and X⁴ is

Arg or Tyr. Another aspect of this invention provides a protease inhibitor wherein X² is His or Leu, X³ is Phe, and X⁴ is Gly. Another aspect of this invention provides a protease inhibitor wherein X¹ is Leu. Another aspect of this invention provides a protease inhibitor wherein

X² is His, X³ is Asn or Phe, and X⁴ is Gly. Another aspect of this invention provides a protease inhibitor wherein X² is He, X³ is Pro, and X⁴ is Gly. Another aspect of this invention provides a protease inhibitor wherein X¹ is Gly, X² is He, X³ is Tyr, and X⁴ is Gly.

Another aspect of this invention provides a protease inhibitor wherein X¹ is Met, X² is His, X³ is Ser, and X⁴ is Tyr.

Additionally, another aspect of this invention relates to protease inhibitors comprising the following amino acid sequences: X^l-Val-Cys-Ser-Glu-Gln-Ala-Glu-X²-Gly-Pro-Cys-

Arg-Ala-X³-X⁴-X⁵-X⁶-Trp-Tyr-Phe-Asp-Val-Thr-Glu-

Gly-Lys-Cys-Ala-Pro-Phe-Phe-Tyr-Gly-Gly-Cys-X⁷-

Gly-Asn-Arg-Asn-Asn-Phe-Asp-Thr-Glu-Glu-Tyr-Cys-

Met-Ala-Val-Cys-Gly-Ser-Ala-He, wherein X¹ is selected from Glu-Val-Val-Arg-Glu-, Asp, or Glu; X² is selected from Thr, Val, He and Ser; X³ is selected from Arg, Ala, Leu, Gly, or Met; X⁴ is selected from He, His, Leu, Lys, Ala, or Phe; X⁵ is selected from Ser, He, Pro, Phe, Tyr, Trp, Asn, Leu, His, Lys, or Glu; X⁶ is selected from Arg, His, or Ala; and X⁷ is selected from Gly, Ala, Lys, Pro, Arg, Leu, Met, or Tyr.

Another aspect of this invention provides a protease inhibitor as defined above wherein at least two amino acid residues selected from the group consisting of X³, X⁴, X⁵, and X⁶ differ from the residues found in the naturally occurring sequence of KPI. Another aspect of this invention provides a protease inhibitor wherein X¹ is Glu-Val-Val-Arg-Glu-, X² is Thr, Val, or Ser, X³ is Ala or Leu, X⁴ is He, X^s is Tyr, X⁶ is His and X⁷is Gly. Another aspect of this invention provides a protease inhibitor wherein X² is Thr, and X³ is Ala. Another aspect of this invention provides a protease inhibitor wherein X² is Thr, and X³ is Leu. Another aspect of this invention provides a protease inhibitor wherein X² is Val, and X³ is Ala. Another aspect of this invention provides a protease inhibitor wherein X² is Ser, and X³ is Ala. Another aspect of this invention provides a protease inhibitor wherein X² is Val, and X³ is Leu. Another aspect of this invention provides a protease inhibitor wherein X² is Ser, and X³ is Leu. Another aspect of this invention provides a protease inhibitor wherein X¹ is Glu-Val-Val-Arg-Glu-, X² is Thr, X³ is Leu, X⁴ is Phe, X^s is Lys, X⁶ is Arg and X⁷ is Gly. Another aspect of this invention provides a protease inhibitor wherein X¹ is Glu-Val-Val-Arg-Glu-, X² is Thr, X³ is Leu, X⁴ is Phe, X⁵ is Lys, X⁶ is Arg and X⁷ is Tyr. Another aspect of this invention provides a protease inhibitor wherein X¹ is Glu-Val-Val-Arg-Glu-, X² is Thr, X³ is Leu, X⁴ is Phe, X⁵ is Lys, X⁶ is Arg and X⁷ is Leu.

Other objects, features and advantages of the present invention will become apparent from the following detailed description. It should be understood, however, that the detailed description and the specific examples, while indicating preferred embodiments of the invention, are given by way of illustration only, since various changes and modifications within the spirit and scope of the invention will become apparent to those skilled in the art from this detailed description.

Brief Description of the Drawings

Figure 1 shows the strategy for the construction of plasmid pTW10:KPI. Figure 2 shows the sequence of the synthetic gene for KPI (1→57) fused to the bacterial phoA secretory signal sequence.

Figure 3 shows the strategy for construction of plasmid pKPI-61.

Figure 4 shows the 192 bp -Xbal-Hindlll synthetic gene fragment encoding KPI (l-»57) and four amino acids from yeast alpha-mating factor.

Figure 5 shows the synthetic 201 bp -Ωbal-Hindlll fragment encoding KPI(-4→57) in PKPI-61.

Figure 6 shows the strategy for the construction of plasmid pTW113.

Figure 7 shows plasmid PTW113, encoding the 445 bp synthetic gene for yeast alpha-factor-KPI(-4-*57) fusion. Figure 8 shows the amino acid sequence for KPI (-4→57) .

Figure 9 shows the strategy for constructing plasmid pTW6165.

Figure 10 shows plasmid, PTW6165, encoding the 445 bp synthetic gene for alpha-factor-KPI(-4→57; M15A, S17W) fusion.

Figure 11 shows the sequences of the annealed oligonucleotide pairs used to construct plasmids PTW6165, pTW6166, pTW6175, pBG028, pTW6183, pTW6184, pTW6185, pTW6173, and pTW6174.

Figure 12 shows the sequence of plasmid PTW6166 encoding the fusion of yeast alpha-factor and KPI(-4→57; M15A, S17Y) .

Figure 13 shows the sequence of plasmid PTW6175 encoding the fusion of yeast alpha-factor and KPI(-4→57; M15L, S17F) .

Figure 14 shows the sequence of plasmid PBG028 encoding the fusion of yeast alpha-factor and KPI(-4-»57; M15L, S17Y) . Figure 15 shows the sequence of plasmid PTW6183 encoding the fusion of yeast alpha-factor and KPI(-4→57; I16H, S17F) . Figure 16 shows the sequence of plasmid PTW6184 encoding the fusion of yeast alpha-factor and KPI(-4→57; I16H, S17Y) .

Figure 17 shows the sequence of plasmid PTW6185 encoding the fusion of yeast alpha-factor and KPI(-4-»57; I16H, S17W) .

Figure 18 shows the sequence of plasmid PTW6173 encoding the fusion of yeast alpha-factor and KPI(-4→57; M15A, I16H) . Figure 19 shows the sequence of plasmid PTW6174 encoding the fusion of yeast alpha-factor and KPI(-4-*57; M15L, I16H) .

Figure 20 shows the amino acid sequence of KPI 4-57; M15A, S17W) . Figure 21 shows the amino acid sequence of KPI -4→57; M15A, S17Y) .

Figure 22 shows the amino acid sequence of KPI -4→57; M15L, S17F) .

Figure 23 shows the amino acid sequence of KPI (-4-*57; M15L, S17Y) .

Figure 24 shows the amino acid sequence of KPI -4→57; I16H, S17F) .

Figure 25 shows the amino acid sequence of KPI -4→57; I16H, S17Y) . Figure 26 shows the amino acid sequence of KPI -4→57; I16H, S17W) .

Figure 27 shows the amino acid sequence of KPI -4→57; M15A, S17F) .

Figure 28 shows the amino acid sequence of KPI (-4→57; M15A, I16H) .

Figure 29 shows the amino acid sequence of KPI 4→57; M15L, I16H) .

Figure 30 shows the construction of plasmid pSP26:Aιηp:Fl. Figure 31 shows the construction of plasmid pglll. Figure 32 shows the construction of plasmid pP oA:KPI:gIII.

Figure 33 shows the construction of plasmid pLGl. Figure 34 shows the construction of plasmid pAL51. Figure 35 shows the construction of plasmid pAL53. Figure 36 shows the construction of plasmid PSP26:Amp:FI:PhoA:KPI:gill.

Figure 37 shows the construction of plasmid pDWl #14. Figure 38 shows the coding region for the fusion of phoA-KPI (1→55) -genelll.

Figure 39 shows the construction of plasmid PDWl 14- 2.

Figure 40 shows the construction of KPI Library 16- 19.

Figure 41 shows the expression unit encoded by the members of KPI Library 16-19.

Figure 42 shows the phoA-KPI (1→55) -genelll region encoded by the most frequently occurring randomized KPI region.

Figure 43 shows the construction of pDD185 KPI (-4→57; M15A, S17F) .

Figure 44 shows the sequence of alpha-factor fused to KPI (-4→57; M15A, S17F) . Figure 45 shows the inhibition constants (KjS) determined for purified KPI variants against the selected serine proteases kallikrein, factor Xa, and factor Xlla. Figure 46 shows the inhibition constants (KjS) determined for KPI variants against kallikrein, plasmin, and factors Xa, Xla, and Xlla.

Figure 47 shows the post-surgical blood loss in pigs in the presence (KPI) and absence (NS) of KPI 185-1 (M15A, S17F) .

Figure 48 shows the post-surgical hemoglobin loss in pigs in the presence (KPI) and absence (NS) of KPI 185-1 (M15A, S17F) .

Figure 49 shows the oxygen tension in the presence and absence of KPI, before CPB, immediately after CPB, and at 60 and 180 minutes after the end of CPB. Figure 50 summarizes the results shown in Figures 47- 49.

Figure 51 shows the sequence of plasmid PTW6166 encoding the fusion of yeast alpha-factor and KPI(-4→57; M15A, S17Y) . Figure 52 shows the sequence of plasmid PTW6175 encoding the fusion of yeast alpha-factor and KPI(-4→57; M15L, S17F) .

Figure 53 shows the sequence of plasmid PBG028 encoding the fusion of yeast alpha-factor and KPI(-4-»57; M15L, S17Y) .

Figure 54 shows the inhibition constants (KjS) determined for KPI variants against kallikrein, plasmin, and factor Xlla.

Detailed Description

The present invention provides peptides that can bind to and preferably inhibit the activity of serine proteases. These inhibitory peptides can also provide a means of ameliorating, treating or preventing clinical conditions associated with increased activity of serine proteases. The novel peptides of the present invention preferably exhibit a more potent and specific (i.e., greater) inhibitory effect toward serine proteases of interest than known serine protease inhibitors. Examples of such proteases include: kallikrein; chymotrypsins A and B; trypsin; elastase; subtilisin; coagulants and procoagulants, particularly those in active form, including coagulation factors such as factors Vila, IXa, Xa, Xla, and Xlla; plasmin; thrombin; proteinase-3; enterokinase; acrosin; cathepsin; urokinase; and tissue plasminogen activator.

Peptides of the present invention may be used to reduce the tissue damage caused by activation of the proteases of the contact pathway of the blood during surgical procedures such as cardiopulmonary bypass (CPB) . Inhibition of contact pathway proteases reduces the "whole body inflammatory response" that can accompany contact pathway activation, and that can lead to tissue damage, and possibly death. The peptides of the present invention may also be used in conjunction with surgical procedures to reduce activated serine protease-associated perioperative and postoperative blood loss. For instance, perioperative blood loss of this type may be particularly severe during CPB surgery. Pharmaceutical compositions comprising the peptides of the present invention may be used in conjunction with surgery such as CPB; administration of such compositions may occur preoperatively, perioperatively or postoperatively. Examples of other clinical conditions associated with increased serine protease activity for which the peptides of the present invention may be used include: CPB- induced inflammatory response; post-CPB pulmonary injury; pancreatitis; allergy-induced protease release; deep vein thrombosis; thrombocytopenia; rheumatoidarthritis; adult respiratory distress syndrome; chronic inflammatory bowel disease; psoriasis; hyperfibrinolytic hemorrhage; organ preservation; wound healing; and myocardial infarction. Other examples of preferable uses of the peptides of the present invention are described in U.S. Patent No. 5,187,153.

The invention is based upon the novel substitution of amino acid residues in the peptide corresponding to the naturally occurring KPI protease inhibitor domain of human amyloid β-amyloid precursor protein (APPI) . These substitutions produce peptides that can bind to serine proteases and preferably exhibit an inhibition of the activity of serine proteases. The peptides also preferably exhibit a more potent and specific serine protease inhibition than known serine protease inhibitors. In accordance with the invention, peptides are provided that may exhibit a more potent and specific inhibition of one or more serine proteases of interest, e.g., kallikrein, plasmin and factors Xa, Xla, Xlla, and Xlla.

The present invention also includes pharmaceutical compositions comprising an effective amount of at least one of the peptides of the invention, in combination with a pharmaceutically acceptable sterile vehicle, as described in REMINGTON'S PHARMACEUTICAL SCIENCES: DRUG RECEPTORS AND RECEPTOR THEORY, (18th ed.), Mack Publishing Co., Easton, PA (1990). A. Selection of sequences of KPI variants

The sequence of KPI is shown in Table l. Table 2 shows a comparison of this sequence with that of aprotinin, with which it shares about 45% sequence identity. The numbering convention for KPI shown in Table l and used hereinafter designates the first glutamic acid residue of KPI as residue 1. This corresponds to residue number 3 using the standard numbering convention for aprotinin. The crystal structure for KPI complexed with trypsin has been determined. See Perona et al., J. Mol . Biol . 230:919 (1993) . The three-dimensional structure reveals two binding loops within KPI that contact the protease. The first loop extends from residue Thr⁹ to He¹⁶, and the second loop extends from residue Phe³² to Gly³⁷. The two protease binding loops are joined through the disulfide bridge extending from Cys¹² to Cys³⁶. KPI contains two other disulfide bridges, between Cys³ and Cys⁵³, and between Cys²⁸ to Cys⁴⁹. This structure was used as a guide to inform our strategy for making the amino acid residue substitutions that will be most likely to affect the protease inhibitory properties of KPI. Our examination of the structure indicated that certain amino acid residues, including residues 9, 11, 13-18, 32, and 37-40, appear to be of particular significance in determining the protease binding properties of the KPI peptide. In a preferred embodiment of the invention two or more of those KPI peptide residues are substituted; such substitutions preferably occurring among residues 9, 11, 13-18, 32, and 37-40. In particular, we found that those substituted peptides, including peptides comprising substitutions of at least two of the four residues at positions 15-18, may exhibit more potent and specific serine protease inhibition toward selected serine proteases of interest than exhibited by the natural KPI peptide domain. Such substituted peptides may further comprise one or more additional substitutions at residues 9, 11, 13, 14, 32 and 37-40; in particular, such peptides may further comprise a substitution at positions 9 or 37, or an additional substitution at residue 13. In particular, the peptides of the present invention preferably exhibit a greater potency and specificity for inhibiting one or more serine proteases of interest (e.g., kallikrein, plasmin and factors Vila, IXa, Xa, Xla, and Xlla) than the potency and specificity exhibited by native KPI or other known serine protease inhibitors. That greater potency and specificity may be manifested by the peptides of the present invention by exhibiting binding constants for serine proteases of interest that are less than the binding constants exhibited by native KPI, or other known serine protease inhibitors, for such proteases.

As an initial guide to informing the choices of amino acid substitution for preparation of KPI variants, the sequences and protease inhibitory activities of aprotinin and KPI are compared. Aprotinin is twice as potent as wild-type KPI with respect to inhibition of human plasma kallikrein, and is 100-fold more potent as an inhibitor of human plasmin. There are three amino acid differences between aprotinin and wild-type KPI in the first protease binding loop extending from residues 9 to 17. A series of KPI variants may then be created, using the methods detailed below, where the residues present in aprotinin at positions 13, 15 and 17 are substituted with the residues found in KPI. The effect of such substitutions upon KPI inhibition of plasma kallikrein and plasmin is then determined.

These results show that substitution of arginine at position 13 by lysine significantly reduces the activity of the resulting protein as an inhibitor of plasma kallikrein. Similarly, substituting positions 15 and 17 of KPI with the corresponding residues found in aprotinin also decreases potency of the KPI variants against kallikrein. Substitutions of aprotinin residues at positions 13 and 15, however, increase the potency of KPI toward plasmin. The single change of methionine to arginine at position 15 (designated M15R) decreases the K, against plasmin more than 10-fold. The change of serine to isoleucine at position 17 (S17I) decreases the potency of KPI toward plasmin.

It is observed that single-amino acid substitutions in the first protease binding loop are generally additive, that is, combinations of single amino-acid substitutions, each of which individually enhance the potency toward plasmin, result in variants with even higher potency. The substitution R13K results in a plasmin K_j of 12.3, and the further exchange of M15R results in a K_; that is reduced to 1.45.

It appears, therefore, from these results that combinations of favorable single amino acid substitutions can result in enhanced potency of KPI variants. It is further apparent that substitution in KPI with the residues found in the aprotinin first protease binding loop is not always useful. Although aprotinin is a more potent kallikrein inhibitor than KPI, none of the combinations of aprotinin residues in KPI improve kallikrein inhibition. To further investigate substitutions that might usefully enhance protease inhibition, a series of single substitutions in KPI is prepared where charged residues in the first protease binding loop are systematically replaced with alanine. This is intended to determine whether substitutions at these sites affect potency toward plasma kallikrein, factor Xlla or plasmin.

It is found that replacement of arginine at position 13 (R13A) drastically reduces KPI inhibition of kallikrein, Xlla or plasmin. The replacement I16A, however, significantly increases the K_j towards both kallikrein and plasmin, suggesting that this amino acid position is critical to inhibition of these proteases. The S17A substitution has little effect. The substitution R18A has little effect upon plasmin inhibition, but significantly impacts inhibition of kallikrein and factor Xlla. These results suggest that substitutions at positions 116 and R18 have the potential to significantly alter the potency of KPI toward kallikrein or plasmin. These results also suggest that substitutions at residues M15 and S17 could have major effects upon inhibition of kallikrein, Xlla or plasmin. To investigate this further, two sets of yeast expression plasmids are prepared, using the methods described in detail below, in which either M15 or S17 are replaced with all possible amino acids.

Yeast are transformed with these two sets of plasmids, and 100 individual colonies are picked at random from each transformation. Small cultures are grown from each of these colonies, and their conditioned broth is harvested and tested for kallikrein inhibiting activity. The plasmids from colonies yielding cultures expressing KPI variants more potent than wild-type KPI are isolated, and the KPI domain are sequenced. It is found that only four 4 substitutions at position 15: M15A,M15L,M15S,M15V; and 4 substitutions at position 17: S17P,S17F,S17Y and S17W, result in KPI variants with improved potency toward kallikrein. Combinations of these position 15 and 17 mutants are then prepared to test if their effects on potency of protease inhibition are additive. Four of these double mutants ( [M15A,S17Y] , [M15A.S17W] , [M15L,S17Y] and

[M15L,S17F]) are substantially more potent toward kallikrein and factor Xlla than the single amino acid substitutions on which they are based.

The results of changing arginine at positions 18 for alanine also suggest that substitutions at position 18 could affect inhibition of kallikrein and factor Xlla. The KPI double variant M13A,S17W (named TW6165 below) is used to construct a series of variants where all possible amino acid substitutions other than Cys and Arg are placed at position 18. Of these variants, three

([M13A,S17W, R18H] , [M13A,S17W, R18Q] , and [M13A,S17W, R18T] ) are found to exhibit enhanced inhibition of kallikrein and Factor Xlla.

The results described above relate to proteins having the N-terminal sequence EWREVCS- et seg. , as found in KPI (-4-»57) . The present invention also relates, however to proteins wherein the N-terminal sequence may be varied, preferably by substituting aspartic acid at the N-terminus in place of the glutamic acid (i.e. the N- ter inal sequence is DWREVCS-). Other N-terminal sequences that may be used will be apparent to the skilled artisan, including a sequence lacking the first four amino acids of KPI(-4-»57), i.e. having the sequence EVCS-.

By way of example, and as set forth in greater detail below, the serine protease inhibitory properties of peptides of the present invention were measured for the serine proteases of interest — kallikrein, plasmin and factors Xa, Xla, and Xlla. Methodologies for measuring the inhibitory properties of the KPI variants of the present invention are known to those skilled in the art, e.g., by determining the inhibition constants of the variants toward serine proteases of interest, as described in Example 4, infra. Such studies measure the ability of the novel peptides of the present invention to bind to one or more serine proteases of interest and to preferably exhibit a greater potency and specificity for inhibiting one or more serine protease of interest than known serine protease inhibitors such as native KPI.

The ability of the peptides of the present invention to bind one or more serine proteases of interest, particularly the ability of the peptides to exhibit such greater potency and specificity toward serine proteases of interest, manifest the clinical and therapeutic applications of such peptides. The clinical and therapeutic efficacy of the peptides of the present invention can be assayed by in vitro and in vivo methodologies known to those skilled in the art, e.g., as described in Example 5, infra. Table 1: SEQUENCE OP KPI:

1 10 20 30

V R E V C S E Q A E T G P C R A M I S R W Y F D V T E G K C A P

40 50

F F Y G G C G G N R N N F D T E E Y C M A V C G S A I

Table 2: COMPARISON OF KPI AND APROTININ SEQUENCES^;

1 10 20 30 40 50

KPI : VREVCSEOAETGPCRAMISRWYFDVTEGKCAPFFYGGCGGNRNNFDTEEYCMAVCGSAI

I I MM I I I I I I MM MM I II II I

BPTI : RPDFCLEPPYTGPCKJUtllRYFYNAKAGLCOTFVYGGCRAKRNNFKSAEDCMRTCGGA 1 10 20 30 40 50

B. Methods of producing KPI variants

The peptides of the present invention can be created by synthetic techniques or recombinant techniques which employ genomic or cDNA cloning methods.

1. Production by chemical synthesis

Peptides of the present invention can be routinely synthesized using solid phase or solution phase peptide synthesis. Methods of preparing relatively short peptides such as KPI by chemical synthesis are well known in the art. KPI variants could, for example be produced by solid-phase peptide synthesis techniques using commercially available equipment and reagents such as those available from Milligen (Bedford, MA) or Applied Biosystems-Perkin Elmer (Foster City, CA) . Alternatively, segments of KPI variants could be prepared by solid-phase synthesis and linked together using segment condensation methods such as those described by Dawson et al., Science 266:776 (1994). During chemical synthesis of the KPI variants, substitution of any amino acid is achieved simply by replacement of the residue that is to be substituted with a different amino acid monomer.

2. Production by recombinant DNA technology

(a) Preparation of genes encoding KPI variants

In a preferred embodiment of the invention, KPI variants are produced by recombinant DNA technology. This requires the preparation of genes encoding each KPI variant that is to be made. Suitable genes can be constructed by oligonucleotide synthesis using commercially available equipment, such as that provided by Milligen and Applied Biosystems, supra . The genes can be prepared by synthesizing the entire coding and non- coding strands, followed by annealing the two strands. Alternatively, the genes can be prepared by ligation of smaller synthetic oligonucleotides by methods well known in the art. Genes encoding KPI variants are produced by varying the nucleotides introduced at any step of the synthesis to change the amino acid sequence encoded by the gene.

Preferably, however, KPI variants are made by site- directed mutagenesis of a gene encoding KPI. Methods of site-directed mutagenesis are well known in the art. See, for example, Ausubel et al., (eds.) CURRENT PROTOCOLS IN MOLECULAR BIOLOGY (Wiley Interscience, 1987); PROTEIN ENGINEERING (Oxender & Fox eds., A. Liss, Inc. 1987) . These methods require the availability of a gene encoding KPI or a variant thereof, which can then be mutagenized by known methods to produce the desired KPI variants. In addition, linker-scanning and polymerase chain reaction ("PCR") mediated techniques can be used for purposes of mutagenesis. See PCR TECHNOLOGY (Erlich ed. , Stockton Press 1989) ; CURRENT PROTOCOLS IN MOLECULAR BIOLOGY, vols. 1 & 2, loc. cit .

A gene encoding KPI can be obtained by cloning the naturally occurring gene, as described for example in U.S. Patents Nos. 5,223,482 and 5,187,153, which are hereby incorporated by reference in their entireties. In particular, see columns 6-9 of U.S. Patent No. 5,187,153. See also PCT Application No. 93/09233. In a preferred embodiment of the invention a synthetic gene encoding KPI is produced by chemical synthesis, as described above. The gene may encode the 57-amino acid KPI domain shown in Table 1, or it may also encode additional N-terminal amino acids from the APPI protein sequence, such as the four amino acid sequence (Glu-Val-Val-Arg, designated residues -4 to -1) immediately preceding the KPI domain in APPI.

Production of the gene by synthesis allows the codon usage of the KPI gene to be altered to introduce convenient restriction endonuclease recognition sites, without altering the sequence of the encoded peptide. In a preferred embodiment of the invention, the synthetic KPI gene contains restriction endonuclease recognition sites that facilitate excision of DNA cassettes from the KPI gene. These cassettes can be replaced with small synthetic oligonucleotides encoding the desired changes in the KPI peptide sequence. See Ausubel, supra .

This method also allows the production of genes encoding KPI as a fusion peptide with one or more additional peptide or protein sequences. The DNA encoding these additional sequences is arranged in-frame with the sequence encoding KPI such that, upon translation of the gene, a fusion protein of KPI and the additional peptide or protein sequence is produced. Methods of making such fusion proteins are well known in the art. Examples of additional peptide sequences that can be encoded in the genes are secretory signal peptide sequences, such as bacterial leader sequences, for example ompA and phoA, that direct secretion of proteins to the bacterial periplasmic space. In a preferred embodiment of the invention, the additional peptide sequence is a yeast secretory signal sequence, such as c-- mating factor, that directs secretion of the peptide when produced in yeast. Additional genetic regulatory sequences can also be introduced into the synthetic gene that are operably linked to the coding sequence of the gene, thereby allowing synthesis of the protein encoded by the gene when the gene is introduced into a host cell. Examples of regulatory genetic sequences that dan be introduced are: promoter and enhancer sequences and transcriptional and translational control sequences. Other regulatory sequences are well known in the art. See Ausubel et al., supra, and Sambrook et al., supra. Sequences encoding other fusion proteins and genetic elements are well known to those of skill in the art. In a preferred embodiment of the invention, the KPI sequence is prepared by ligating together synthetic oligonucleotides to produce a gene encoding an in-frame fusion protein of yeast α-mating factor with either KPI (1→57) or KPI (-4→57) .

The gene constructs prepared as described above are conveniently manipulated in host cells using methods of manipulating recombinant DNA techniques that are well known in the art. See, for example Sambrook et al., MOLECULAR CLONING: A LABORATORY MANUAL, Second Edition, (Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY 1989) , and Ausubel, supra. In a preferred embodiment of the invention the host cell used for manipulating the KPI constructs is E. coli . For example, the construct can be ligated into a cloning vector and propagated in E. coli by methods that are well known in the art. Suitable cloning vectors are described in Sambrook, supra, or are commercially available from suppliers such as Promega (Madison, WI) , Stratagene (San Diego, CA) and Life Technologies (Gaithersburg, MD) .

Once a gene construct encoding KPI has been obtained, genes encoding KPI variants are obtained by manipulating the coding sequence of the construct by standard methods of site-directed mutagenesis, such as excision and replacement of small DNA cassettes, as described supra. See Ausubel, supra, and Sinha et al., supra. See also U.S. Patent 5,373,090, which is herein incorporated by reference in its entirety. See particularly, columns

4-12 of U.S. Patent 5,272,090. These genes are then used to produce the KPI variant peptides as described below.

Alternatively, KPI variants can be produced using phage display methods. See, for example, Dennis et al. supra, which is hereby incorporated by reference in its entirety. See also U.S. Patent Nos. 5,223,409 and 5,403,484, which are hereby also incorporated by reference in their entireties. In these methods, libraries of genes encoding variants of KPI are fused in- frame to genes encoding surface proteins of filamentous phage, and the resulting peptides are expressed

(displayed) on the surface of the phage. The phage are then screened for the ability to bind, under appropriate conditions, to serine proteases of interest immobilized on a solid support. Large libraries of phage can be used, allowing simultaneous screening of the binding properties of a large number of KPI variants. Phage that have desirable binding properties are isolated and the sequences of the genes encoding the corresponding KPI variants is determined. These genes are then used to produce the KPI variant peptides as described below.

(b) Expression of KPI variant peptides Once genes encoding KPI variants have been prepared, they are inserted into an expression vector and used to produce the recombinant peptide. Suitable expression vectors and corresponding methods of expressing recombinant proteins and peptides are well known in the art. Methods of expressing KPI peptides are described in U.S. Patent 5,187,153, columns 9-11, U.S. Patent 5,223,482, columns 9-11, and PCT application 93/09233, pp. 49-67. See also Ausubel et al., supra, and Sambrook et al., supra . The gene can be expressed in any number of different recombinant DNA expression systems to generate large amounts of the KPI variant, which can then be purified and tested for its ability to bind to and inhibit serine proteases of interest.

Examples of expression systems known to the skilled practitioner in the art include bacteria such as E. coli , yeast such as Saccharomyces cerevisiae and Pichia pastoris, baculovirus, and mammalian expression systems such as in Cos or CHO cells. In a preferred embodiment, KPI variants are expressed in Pichia pastoris. In another preferred embodiment the KPI variants are cloned into expression vectors to produce a chimeric gene encoding a fusion protein of the KPI variant with yeast Qf-mating factor. The mating factor acts as a signal sequence to direct secretion of the fusion protein from the yeast cell, and is then cleaved from the fusion protein by a membrane-bound protease during the secretion process. The expression vector is transformed into S. cerevisiae, the transformed yeast cells are cultured by standard methods, and the KPI variant is purified from the yeast growth medium.

Recombinant bacterial cells expressing the peptides of the present invention, for example, E. coli , are grown in any of a number of suitable media, for example LB, and the expression of the recombinant antigen induced by adding IPTG to the media or switching incubation to a higher temperature. After culturing the bacteria for a further period of between 2 and 24 hours, the cells are collected by centrifugation and washed to remove residual media. The bacterial cells are then lysed, for example, by disruption in a cell homogenizer and centrifuged to separate dense inclusion bodies and cell membranes from the soluble cell components. This centrifugation can be performed under conditions whereby dense inclusion bodies are selectively enriched by incorporation of sugars such as sucrose into the buffer and centrifugation at a selective speed. If the recombinant peptide is expressed in inclusion bodies, as is the case in many instances, these can be washed in any of several solutions to assist in the removal of any contaminating host proteins, then solubilized in solutions containing high concentrations of urea (e.g., 8M) or chaotropic agents such as guanidine hydrochloride in the presence of reducing agents such as S-mercaptoethanol or DTT (dithiothreitol) . At this stage it may be advantageous to incubate the peptides of the present invention for several hours under conditions suitable for the peptides to undergo a refolding process into a conformation which more closely resembles that of native KPI. Such conditions generally include low protein concentrations less than 500 μg/ml, low levels of reducing agent, concentrations of urea less than 2M and often the presence of reagents such as a mixture of reduced and oxidized glutathione which facilitate the interchange of disulphide bonds within the protein molecule. The refolding process can be monitored, for example, by SDS-PAGE or with antibodies which are specific for the native molecule (which can be obtained from animals vaccinated with the native molecule isolated from parasites) . Following refolding, the peptide can then be purified further and separated from the refolding mixture by chromatography on any of several supports including ion exchange resins, gel permeation resins or on a variety of affinity columns. Purification of KPI variants can be achieved by standard methods of protein purification, e.g., using various chromatographic methods including high performance liquid chromatography and adsorption chromatography. The purity and the quality of the peptides can be confirmed by amino acid analyses, molecular weight determination, sequence determination and mass spectrometry. See, for example, PROTEIN PURIFICATION METHODS — A PRACTICAL APPROACH, Harris et al., eds. (IRL Press, Oxford, 1989). In a preferred embodiment, the yeast cells are removed from the growth medium by filtration or centrifugation, and the KPI variant is purified by affinity chromatography on a column of trypsin-agarose, followed by reversed-phase HPLC.

C. Measurement of protease inhibitory properties of KPI variants

Once KPI variants have been purified, they are tested for their ability to bind to and inhibit serine proteases of interest in vitro . The peptides of the present invention preferably exhibit a more potent and specific inhibition of serine proteases of interest than known serine protease inhibitors, such as the natural KPI peptide domain. Such binding and inhibition can be assayed for by determining the inhibition constants for the peptides of the present invention toward serine proteases of interest and comparing those constants with constants determined for known serine protease inhibitors, e.g., the native KPI domain, toward those proteases. Methods for determining inhibition constants of protease inhibitors are well known in the art. See Fersht, ENZYME STRUCTURE AND MECHANISM, 2nd ed. , W.H. Freeman and Co., New York, (1985).

In a preferred embodiment the inhibition experiments are carried out using a chromogenic synthetic protease substrate, as described, for example, in Bender et al., J. Amer. Chem. Soc. 88:5890 (1966). Measurements taken by this method can be used to calculate inhibition O 96/35788 PCΪ7US96/06384

- 34 - constants (K, values) of the peptides of the present invention toward serine proteases of interest. See Bieth in BAYER-SYMPOSIUM V "PROTEINASE INHIBITORS", Fritz et al., eds., pp. 463-69, Springer-Verlag, Berlin, Heidelberg, New York, (1974) . KPI variants that exhibit potent and specific inhibition of one or more serine proteases of interest may subsequently be tested in vivo. In vitro testing, however, is not a prerequisite for in vivo studies of the peptides of the present invention.

D. Testing of KPI variants in vivo

The peptides of the present invention may be tested, alone or in combination, for their therapeutic efficacy by various in vivo methodologies known to those skilled in the art, e.g., the ability of KPI variants to reduce postoperative bleeding can be tested in standard animal models. For example, cardiopulmonary bypass surgery can be carried out on animals such as pigs in the presence of KPI variants, or in control animals where the KPI variant is not used. The use of pigs as a model for studying the clinical effects associated with CPB has previously been described. See Redmond et al., Ann. Thorac. Surg. 56:474 (1993) .

The KPI variant is supplied to the animals in a pharmaceutical sterile vehicle by methods known in the art, for example by continuous intravenous infusion. Chest tubes can be used to collect shed blood for a defined period of time. The shed blood, together with the residual intrathoracic blood found after sacrifice of the animal can be used to calculate hemoglobin (Hgb) loss. The postoperative blood and Hgb loss is then compared between the test and control animals to determine the effect of the KPI variants.

E. Therapeutic use of KPI variants

KPI variants of the present invention found to exhibit therapeutic efficacy (e.g., reduction of blood loss following surgery in animal models) may preferably be used and administered, alone or in combination or as a fusion protein, in a manner analogous to that currently used for aprotinin or other known serine protease inhibitors. See Butler et al., supra . Peptides of the present invention generally may be administered in the manner that natural peptides are administered. A therapeutically effective dose of the peptides of the present invention preferably affects the activity of the serine proteases of interest such that the clinical condition may be treated, ameliorated or prevented. Therapeutically effective dosages of the peptides of the present invention can be determined by those skilled in the art, e.g., through in vivo or in vitro models. Generally, the peptides of the present invention may be administered in total amounts of approximately 0.01 to approximately 500, specifically 0.1 to 100 mg/kg body weight, if desired in the form of one or more administrations, to achieve therapeutic effect. It may, however, be necessary to deviate from such administration amounts, in particular depending on the nature and body weight of the individual to be treated, the nature of the medical condition to be treated, the type of preparation and the administration of the peptide, and the time interval over which such administration occurs. Thus, it may in some cases be sufficient to use less than the above amount of the peptides of the present invention, while in other cases the above amount is preferably exceeded. The optimal dose required in each case and the type of administration of the peptides of the present invention can be determined by one skilled in the art in view of the circumstances surrounding such administration. Such peptides can be administered by intravenous injections, in situ injections, local applications, inhalation, oral administration using coated polymers, dermal patches or other appropriate means. Compositions comprising peptides of the present invention are advantageously administered in the form of injectable compositions. Such peptides may be preferably administered to patients via continuous intravenous infusion, but can also be administered by single or multiple injections. A typical composition for such purpose comprises a pharmaceutically acceptable carrier. Pharmaceutically acceptable carriers include aqueous solutions, non-toxic excipients, including salts, preservatives, buffers and the like, as described in REMINGTON'S PHARMACEUTICAL SCIENCES, pp. 1405-12 and 1461-87 (1975) and THE NATIONAL FORMULARY XIV., 14th Ed. Washington: American Pharmaceutical Association (1975) . Aqueous carriers include water, alcoholic/aqueous solutions, saline solutions, parenteral vehicles such as sodium chloride, Ringer's dextrose, etc. Intravenous vehicles include fluid and nutrient replenishers. Preservatives include antimicrobials, anti-oxidants, chelating agents and inert gases. The pH and exact concentration of the various components of the composition are adjusted according to routine skills in the art. See GOODMAN AND GILMAN'S THE PHARMACOLOGICAL BASIS FOR THERAPEUTICS (7th ed.). The peptides of the present invention may be present in such pharmaceutical preparations in a concentration of approximately 0.1 to 99.5% by weight, specifically 0.5 to 95% by weight, relative to the total mixture. Such pharmaceutical preparations may also comprise other pharmaceutically active substances in addition to the peptides of the present invention. Other methods of delivering the peptides to patients will be readily apparent to the skilled artisan.

Examples of mammalian serine proteases that may exhibit inhibition by the peptides of the present invention include: kallikrein; chymotrypsins A and B; trypsin; elastase; subtilisin; coagulants and procoagulants, particularly those in active form, including coagulation factors such as thrombin and factors Vila, IXa, Xa, Xla, and Xlla; plasmin; proteinase-3; enterokinase; acrosin; cathepsin; urokinase; and tissue plasminogen activator. Examples of conditions associated with increased serine protease activity include: CPB-induced inflammatory response; post-CPB pulmonary injury; pancreatitis; allergy-induced protease release; deep vein thrombosis; thrombocytopenia; rheumatoid arthritis; adult respiratory distress syndrome; chronic inflammatory bowel disease; psoriasis; hyperfibrinolytic hemorrhage; organ preservation; wound healing; and myocardial infarction. Other examples of the use of the peptides of the present invention are described in U.S. Patent No. 5,187,153.

The inhibitors of the present invention may also be used for inhibition of serine protease activity in vi tro, for example during the preparation of cellular extracts to prevent degradation of cellular proteins. For this purpose the inhibitors of the present invention may preferably be used in a manner analogous to the way that aprotinin, or other known serine protease inhibitors, are used. The use of aprotinin as a protease inhibitor for preparation of cellular extracts is well known in the art, and aprotinin is sold commercially for this purpose.

The present invention, thus generally described, will be understood more readily by reference to the following examples, which are provided by way of illustration and are not intended to be limiting of the present invention.

EXAMPLES Example 1. Expression of wild-type KPI (-4-*57)

A. Construction of PTW10 :KPI Plasmid PTW10:KPI is a bacterial expression vector encoding the 57 amino acid form of KPI fused to the bacterial phoA signal sequence. The strategy for the construction of PTW10:KPI is shown in Figure 1.

Plasmid pcDNAII (Invitrogen, San Diego, CA) was digested with PvuII and the larger of the two resulting PvuII fragments (3013 bp) was isolated. Bacterial expression plasmid pSP26 was digested with AEluI and RsrII , and the 409 bp Mlul-Rsrll fragment containing the pTrp promoter element and transcription termination signals was isolated by electrophoresis in a 3% NuSieve Agarose gel (FMC Corp., Rockland, ME). Plasmid pSP26, containing a heparin-binding EGF-like growth factor (HB- EGF) insert between the Ndel and Hindlll sites, is described as pNA28 in Thompson et al . , J. Biol . Chem. 269:2541 (1994) . Plasmid pSP26 was deposited in host E. coli W3110, pSP26 with the American Type Culture Collection (ATCC) , 12301 Parklawn Drive, Rockville, Maryland, 20852, USA under the conditions specified by the Budapest Treaty on the International Recognition of the Deposit of Microorganisms (Budapest Treaty) . Host E. coli W3110, pSP26 was deposited on 3 May 1995 and given Accession No. 69800. Availability of the deposited plasmid is not to be construed as a license to practice the invention in contravention of the rights granted under the authority of any government in accordance with its patent laws.

The ends of the Mlul-Rsrll fragment were blunted using DNA polymerase Klenow fragment by standard techniques. The blunted fragment of pSP26 was then ligated into the large PvuII fragment of plasmid pCDNAII, and the ligation mixture was used to transform E. coli strain MC1061. Ampicillin-resistant colonies were selected and used to isolate plasmid pTWIO by standard techniques.

A synthetic gene was constructed encoding the bacterial phoA secretory signal sequence fused to the amino terminus of KPI(l-*57) . The synthetic gene contains cohesive ends for Ndel and Hindlll, and also incorporates restriction endonuclease recognition sites for Agrel, RsrII, Aatll and BamHI, as shown in Figure 2. The synthetic phoA-KPI gene was constructed from 6 oligonucleotides of the following sequences (shown 5'→3'):

6167:

TATGAAACAAAGCACTATTGCACTGGCACTCTTACCGTTACTGTTTACCC CTGTGACAAAAGCCGAGGTGTGCTCTGAA

6169: CTCGGCTTTTGTCACAGGGGTAAACAGTAACGGTAAGAGTGCCAGTGCAA TAGTGCTTTGTTTCATA

6165:

CAAGCTGAGACCGGTCCGTGCCGTGCAATGATCTCCCGCTGGTACTTTGA CGTCACTGAAGGTAAGTGCGCTCCATTCTTT 6166 :

GCACTTACCTTCAGTGACGTCAAAGTACCAGCGGGAGATCATTGCACGGC ACGGACCGGTCTCAGCTTGTTCAGAGCACAC

6168:

TACGGCGGTTGCGGCGGCAACCGTAACAACTTTGACACTGAAGAGTACTG CATGGCAGTGTGCGGATCCGCTATTTAAGCT

6164:

AGCTTAAATAGCGGATCCGCACACTGCCATGCAGTACTCTTCAGTGTCAA AGTTGTTACGGTTGCCGCCGCAACCGCCGTAAAAGAATGGAGC

The oligonucleotides were phosphorylated and annealed in pairs: 6167 + 6169, 6165 + 6166, 6168 + 6164. In 20 μl T4 DNA Ligase Buffer (New England Biolabs, Beverley, MA) , 1 μg of each oligonucleotide pair was incubated with 10 U T4 Polynucleotide Kinase (New England Biolabs) for 1 h at 37°C, then heated to 95°C for l minute, and slow-cooled to room temperature to allow annealing. All three annealed oligo pairs were then mixed for ligation to one another in a total volume of 100 μl T4 DNA Ligase Buffer, and incubated with 400 U T4 DNA Ligase (New England Biolabs) overnight at 15°C. The ligation mixture was extracted with an equal volume of phenol:CHC1₃ (1:1), ethanol-precipitated, resuspended in 50 μl Restriction Endonuclease Buffer #4 (New England Biolabs) and digested with Ndel and Hindlll. The annealed, ligated and digested oligos were then subjected to electrophoresis in a 3% ΝuSieve Agarose gel, and the 240 bp Ndel-Hindlll fragment was excised. This gel- purified synthetic gene was ligated into plasmid pTWIO which had previously been digested with Ndel and Hindlll, and the ligation mixture was used to transform E. coli strain MC1061. Ampicillin-resistant colonies were selected and used to prepare plasmid pTW10:KPI. This plasmid contains the phoA-KPI(l-»57) fusion protein inserted between the pTrp promoter element and the transcription termination signals.

B. Construction of pKPI-61

The strategy for constructing pKPI-61 is shown in Figure 3. Plasmid pTW10:KPI was digested with Agel and Hindlll; the resulting 152 bp Ag-el-Hindlll fragment containing a portion of the KPI synthetic gene was isolated by preparative gel electrophoresis. An oligonucleotide pair (129 + 130) encoding the 9 amino- terminal residues of KPI(1→57) and 4 amino acids of yeast c.-mating factor was phosphorylated and annealed as described above.

129: CTAGATAAAAGAGAGGTGTGCTCTGAACAAGCTGAGA 130: CCGGTCTCAGCTTGTTCΑGAGCACACCTCTCTTTTAT

The annealed oligonucleotides were then ligated to the Agrel-Hindlll fragment of the KPI (1→57) synthetic gene. The resulting 192 bp -Xbal-Hindlll synthetic gene (shown in Figure 4) was purified by preparative gel electrophoresis, and ligated into plasmid pUC19 which had previously been digested with -Xbal and Hindlll. The ligation products were used to transform E. coli strain MC1061. Ampicillin-resistant colonies were picked and used to prepare plasmid PKPI-57 by standard methods. To create a synthetic gene encoding KPI(-4→57) , PKPI-57 was digested with .Xbal and Agel and the smaller fragment replaced with annealed oligos 234 + 235, which encode 4 amino acid residues of yeast o.-mating factor fused a 4 amino acid residue amino-terminal extension of KPI(1→57) .

234: CTAGATAAAAGAGAGGTTGTTAGAGAGGTGTGCTCTGAACAAGCTGAGA 235: CCGGTCTC^GCTTGTTCAGAGCACACCTCTCTAACAACCTCTL ITTAT

The 4 extra amino acids are encoded in the amyloid

0-protein precursor/protease nexin-2 (APPI) which contains the KPI domain. The synthetic 201 bp Xbal-

Hindlll fragment encoding KPI(-4-»57) in pKPI-61 is shown in Figure 5.

C. Assembly of pTW113

The strategy for the construction of PTW113 is shown in Figure 6. Plasmid pSP35 was constructed from yeast expression plasmid pYES2 (Invitrogen, San Diego, CA) as follows. A 267 bp PvuII -Xbal fragment was generated by PCR from yeast αf-mating factor DNA using oligos 6274 and 6273:

6274: GGGGGCAGCTGTATAAACGATTAAAA 6273: GGGGGTCTAGAGATACCCCTTCTTCTTTAG

This PCR fragment, encoding an 82 amino acid portion of yeast α-mating factor, including the secretory signal peptide and pro-region, was inserted into pYES2 that had been previously digested with PvuII and .Xbal. The resulting plasmid is denoted pSP34.

Two oligonucleotide pairs, 6294 + 6292 were then ligated to 6290 + 6291, and the resulting 135 bp fragment was purified by gel electrophoresis.

6294: CTAGATAAAAGAGAGGCΠGAGGCΓCACGCTGAAGGTACΓTTCACTTC 6290: TGACGTCTCTTCTTACTTGGAAGGTCAAGCTGCTAAGGAATTCAT

CGCTTGGTTGGTCAAAGGTAGAGGTTAAGCTTA

6291: CTAGTAAGCTTAACCTCTACCTTTGACCAACCAAGCGATGAATTC CTTAGCA

6292: GCTTGACCTTCCAAGTAAGAAGAGACGTCAGAAGTGAAAGTACCT TCAGCGTGAGCCTCAGCCTCTCTTTTAT

The resulting synthetic fragment was ligated into the Xbal site of pSP34, resulting in plasmid pSP35. pSP35 was digested with .Xbal and Hindlll to remove the insert, and ligated with the 201 bp .Xbal-Hindlll fragment of pKPI-61, encoding KPI(-4→57). The resulting plasmid pTWH3, encodes the 445 bp synthetic gene for the cv- factor-KPI(-4-»57) fusion. See Figure 7.

D. Transformation of yeast with pTW113 Saccharomyces cerevisiae strain ABL115 was transformed with plasmid pTW113 by electroporation by the method of Becker et al., Methods Enzymol . 194:182 (1991) . An overnight culture of yeast strain ABL115 was used to inoculate 200 ml YPD medium. The inoculated culture was grown with vigorous shaking at 30°C to an ODβooOf 1.3-1.5, at which time the cells were harvested by centrifugation at 5000 rpm for 5 minutes. The cell pellet was resuspended in 200 ml ice-cold water, respun, resuspended in 100 ml ice-cold water, then pelleted again. The washed cell pellet was resuspended in 10 ml ice-cold IM sorbitol, recentrifuged, then resuspended in a final volume of 0.2 ml ice-cold IM sorbitol. A 40 μl aliquot of cells was placed into the chamber of a cold 0.2 cm electroporation cuvette (Invitrogen) , along with 100 ng plasmid DNA for pTW113. The cuvette was placed into an Invitrogen Electroporator II and pulsed at 1500 V, 25 μF, 100 Ω . Electroporated cells were diluted with 0.5 ml IM sorbitol, and 0.25 ml was spread on an SD agar plate containing IM sorbitol. After 3 days' growth at 30°C, individual colonies were streaked on SD + CAA agar plates.

E. Induction of pTW113 /AB 115, purification of KPI (-4→57)

Yeast cultures were grown in a rich broth and the galactose promoter of the KPI expression vector induced with the addition of galactose as described by Sherman, Methods Ehzymol . 194:3 (1991). A single well-isolated colony of pTW113/ABL115 was used to inoculate a 10 ml overnight culture in Yeast Batch Medium. The next day, IL Yeast Batch Medium which had been made 0.2% glucose was inoculated to an OD^ of 0.1 with the overnight culture. Following 24 hours at 30°C with vigorous shaking, the IL culture was induced by the addition of 20 ml Yeast Galactose Feed Medium. Following induction, the culture was fed every 12 hours with the addition of 20 ml Yeast Galactose Feed Medium. At 48 hours after induction, the yeast broth was harvested by centrifugation, then adjusted to pH 7.0 with 2M Tris, pH 10. The broth was subjected to trypsin-Sepharose affinity chromatography, and bound KPI(-4-_*57) was eluted with 20mM Tris pH 2.5. See Schilling et al . , Gene 98:225 (1991). Final purification of KPI(-4→57) was accomplished by HPLC chromatography on a semi-prep Vydac C4 column in a gradient of 20% to 35% acetonitrile. The sample was dried and resuspended in PBS at 1-2 mg/ml. The amino acid sequence of KPI(-4→57) is shown in Figure 8.

Example 2. Recombinant Expression of site-directed KPI(-4-_*57) variants

Expression vectors for the production of specific variants of KPI(-4-*57) were all constructed using the pTWH3 backbone as a starting point. For each KPI variant, an expression construct was created by replacing the 40 bp RsrII-Aatll fragment of the synthetic KPI gene contained in pTW113 with a pair of annealed oligonucleotides which encode specific codons mutated from the wild-type KPI(-4→57) sequence. In the following Examples the convention used for designating the amino substituents in the KPI variants indicates first the single letter code for the amino acid found in wild-type KPI, followed by the position of the residue using the numbering convention described supra, followed by the code for the replacement amino acid. Thus, for example, M15R indicates that the methionine residue at position 15 is replaced by an arginine.

A. Construction of pTW6165

The strategy for constructing pTW6165 is shown in Figure 9. Plasmid pTW113 was digested with RsrII and Aatll, and the larger of the two resulting fragments was isolated. An oligonucleotide pair (812 + 813) was phosphorylated, annealed and gel-purified as described above.

812: GTCCGTGCCGTGCΑGCTATCTGGCGCTGGTACTTTGACGT 813: CAAAGTACCAGCGCCAGATAGCTGCACGGCACG

The annealed oligonucleotides were ligated into the RsrII and Aatll-digested pTWH3, and the ligation product was used to transform E. coli strain MC1061. Transformed colonies were selected by ampicillin resistance. The resulting plasmid, pTW6165, encodes the 445 bp synthetic gene for the cv-factor-KPI(-4-»57; M15A, S17W) fusion. See Figure 10.

B . Construction of pTW6166, pTW6175, pBG028 , pTW6183 , pTW6184 , pTW6185, pTW6173 , pTW6174 .

Construction of the following KPI(-4→57) variants was accomplished exactly as outlined for pTW6165. The oligonucleotides utilized for each construct are denoted below, and the sequences of annealed oligonucleotide pairs are shown in Figure 11. Figures 12-19 show the synthetic genes for the c.-factor fusions with each

KPI(-4→57) .variant.

pTW6166: KPI(-4→57; M15A, S17Y) — See Figure 12 814: GTCCGTGCCGTGCAGCTATCTACCGCTGGTACTTTGACGT 815: CAAAGTACCAGCGGTAGATAGCTGCACGGCACG

pTW6175: KPI(-4→57; M15L, S17F) —See Figure 13

867: GTCCGTGCCGTGCATTGATCTTCCGCTGGTACTTTGACGT

868: CAAAGTACCAGCGGAAGATCAATGCACGGCACG

pBG028: KPI(-4-»57; M15L, S17Y) —See Figure 14 1493: GTCCGTGCCGTGCTTTGATCTACCGCTGGTACTTTGACGT

1494: CAAAGTACCAGCGGTAGATCAAAGCACGGCACG

pTW6183: KPI(-4→57; I16H, S17F) —See Figure 15

925: GTCCGTGCCGTGCAATGCACTTCCGCTGGTACTTTGACGT

926: CAAAGTACCAGCGGAAGTGCATTGCACGGCACG

pTW6184: KPI(-4→57; I16H, S17Y) —See Figure 16

927: GTCCGTGCCGTGCAATGCACTACCGCTGGTACTTTGACGT 928: CAAAGTACCAGCGGTAGTGCATTGCACGGCACG

pTW6185: KPI(-4→57; I16H, S17W) —See Figure 17

929: GTCCGTGCCGTGCAATGCACTGGCGCTGGTACTTTGACGT

930: CAAAGTACCAGCGCCAGTGCATTGCACGGCACG

pTW6173: KPI(-4→57; M15A, I16H) —See Figure 18

863: GTCCGTGCCGTGCAGCTCACTCCCGCTGGTACTTTGACGT

864: CAAAGTACCAGCGGGAGTGAGCTGCACGGCACG

PTW6174: KPI(-4-»57; M15L, I16H) — See Figure 19

865: GTCCGTGCCGTGCATTGCACTCCCGCTGGTACTTTGACGT 866: CAAAGTACCAGCGGGAGTGCAATGCACGGCACG

C. Transformation of yeast with expression vectors Yeast strain ABL115 was transformed by electroporation exactly according to the protocol described for transformation by pTW113.

D. Induction of transformed yeast strains, purification of KPI (-4→57) variants.

Cultures of yeast strains were grown and induced, and recombinant secreted KPI(-4-*57) variants were purified according to the procedure described for KPI(-4-»57) . The amino acid sequences of KPI(-4→57) variants are shown in

Figures 20-29.

Example 3. Identification of KPI (-4-»57; M15A, S17F) DD185 by phage display.

A. Construction of vector pSP26:Amp:Fl The construction of pSP26:Amp:Fl is outlined in

Figure 30. Vector pSP26:Amp:Fl contributes the basic plasmid backbone for the construction of the phage display vector for the phoA:KPI fusion, PDWl #14. pSP26:Amp:Fl contains a low-copy number origin of replication, the ampicillin-resistance gene (Amp) and the Fl origin for production of single-stranded phagemid DNA. The ampicillin-resistance gene (Amp) was generated through polymerase chain reaction (PCR) amplification from the plasmid genome of PUC19 using oligonucleotides 176 and 177.

176: GCCATCGATGGTTTCTTAAGCGTCAGGTGGC-ACTTTTC

177: GCGCCAATTCTTGGTCTACGGGGTCTGACGCTCAGTGGAACGAA

The PCR amplification of Amp was done according to standard techniques, using Taq polymerase (Perkin-Elmer Cetus, Norwalk, CT) . Amplification from plasmid pUC19 with these oligonucleotides yielded a fragment of 1159 bp, containing PflMI and Clal restriction sites. The PCR product was digested with PflMI and Clal and purified by agarose gel electrophoresis in 3% NuSieve

Agarose (FMC Corp.). Bacterial expression vector pSP26

(supra; was digested with PflMI and Clal and the larger vector fragment was purified. The PflMI-Clal PCR fragment was ligated into the previously digested pSP26 containing the Amp gene. The ligation product was used to transform E. coli strain MC1061 and colonies were selected by ampicillin resistance. The resulting plasmid is denoted pSP26:Amp.

The Fl origin of replication from the mammalian expression vector pcDNAII (Invitrogen) was isolated in a 692 bp Earl fragment. Plasmid pcDNAII was digested with Earl and the resulting 692 bp fragment purified by agarose gel electrophoresis. Earl-Notl adapters were added to the 692 bp Earl fragment by ligation of two annealed oligonucleotide pairs, 179 + 180 and 181 + 182. The oligo pairs were annealed as described above.

179 GGCCGCTCTTCC 180 AAAGGAAGAGC 181 CTAGAATTGC 182 GGCCGCAATTC The oligonucleotide-ligated fragment was then ligated into the single Notl site of PSP26 :Amp to yield the vector pSP26: Amp:F1.

B . Construction of vector pglll The construction of pglll is outlined in Figure 31. The portion of the phage genelll protein gene contained by the PDWl #14 phagemid vector was originally obtained as a PCR amplification product from vector ml3mp8. A portion of ml3mp8 genelll encoding the carboxyl-terminal 158 amino acid residues of the genelll product was isolated by PCR amplification of ml3mp8 nucleotide residues 2307-2781 using PCR oligos 6162 and 6160.

6162: GCCGGATCCGCTATTTCCGGTGGTGGCTCTGGTTCC 6160: GCCAAGCTTATTAAGACTCCTTATTACGCAG

The PCR oligos contain BamHI and Hindlll restriction recognition sites such that PCR from ml3mp8 plasmid DΝA with the oligo pair yielded a 490 bp BamHI-Hindlll fragment encoding the appropriate portion of genelll. The PCR product was ligated between the BamHI and Hindlll sites within the polylinker of PUC19 to yield plasmid pglll.

C. Construction of pPhoA : KPI : gill Construction of pPhoA:KPI:gIII is outlined in

Figure 32. A portion of the phoA signal sequence and KPI fusion encoded by the phage display vector PDWl #14 originates with pPhoA:KPI:gIII. The 237 bp Ndel-Hindlll fragment of pTW10:KPI encoding the entire phoA:KPI (l-*57) fusion was isolated by preparative agarose gel electrophoresis, and inserted between the Ndel and Hindlll sites of pUC19 to yield plasmid pPhoA:KPI. The

490 bp BamHI-Hindlll fragment of pglll encoding the C- terminal portion of the genelll product was then isolated and ligated between the BamHI and Hindlll sites of pPhoA:KPI to yield vector pPhoa:KPI:gIII. The pP2ιoA:KPI:glll vector encodes a 236 amino acid residue fusion of the phoA signal peptide, KPI (1→57) and the carboxyl-terminal portion of the genelll product.

D. Construction of pLGl

Construction of pLGl is illustrated in Figure 33. The exact genelll sequences contained in vector PDWl #14 originate with phage display vector pLGl.- A modified genelll segment was generated by PCR amplification of the genelll region from pglll using PCR oligonucleotides 6308 and 6305.

6308: AGCTCCGATCTAGGATCCGGTGGTGGCTCTGGTTCCGGT 6305: GCAGCGGCCGTTAAGCTTATTAAGACTCCT

PCR amplification from pglll with these oligonucleotides yielded a 481 bp BamHI-Hindlll fragment encoding a genelll product shortened by 3 amino acid residues at the amino-terminal portion of the segment of the genelll fragment encoded by pglll. A 161 bp Ndel- BamHI fragment was generated by PCR amplification from bacterial expression plasmid pTHW05 using oligonucleotides 6306 and 6307.

6306: GATCCTTGTGTCCATATGAAACAAAGC

6307: CACGTCGGTCGAGGATCCCTAACCACGGCCTTTAACCAG

The 161 bp Ndel-BamHI fragment and the 481 bp BamHI- Hindlll fragment were gel-purified, and then ligated in a three-way ligation into PTW10 which had previously been digested with Ndel and Hindlll. The resulting plasmid pLGl encodes a phoA signal peptide-insert-genelll fusion for phage display purposes. E. Construction of pALΞl

Construction of pALSl is illustrated in Figure 34. Vector pAL51 contains the genelll sequences of pLGl which are to be incorporated in vector pDWl #14. A 1693 bp fragment of plasmid pBR322 was isolated, extending from the BamHI site at nucleotide 375 to the PvuII site at position 2064. Plasmid pLGl was digested with Asp718I and BamHI, removing an 87 bp fragment. The overhanging Asp718I end was blunted by treatment with Klenow fragment, and the PvuII-BamHI fragment isolated from pBR322 was ligated into this vector, resulting in the insertion of a 1693 bp "stuffer" region between the Asp718I and BamHI sites. The 78 bp NdeI -Asp718I region of the resulting plasmid was removed and replaced with the annealed oligo pair 6512 + 6513.

6512: TATGAAACAAAGCACTATTGCACTGGCACTCTTACCGTTACTGTT TACCCCGGTGACCAAAGCCCACGCTGAAG

6513: GTACCTTCAGCGTGGGCTTTGGTCACCGGGGTAAACAGTAACGGT AAGAGTGCCAGTGCAATAGTGCTTTGTTTCA

The newly created 74 bp NdeI -Asp718I fragment encodes the phoA signal peptide, and contains a BstEII cloning site. The resulting plasmid is denoted pAL51.

F. Construction of pAL53

Construction of pAL53 is outlined in Figure 35. Plasmid pAL53 contributes most of the vector sequence of pDWl #14, including the basic vector backbone with Amp gene, Fl origin, low copy number origin of replication, genelll segment, phoA promotor and phoA signal sequence.

Plasmid pAL51 was digested with Ndel and Hindlll and the resulting 2248 bp Ndel-Hindlll fragment encoding the phoA signal peptide, stuffer region and genelll region was isolated by preparative agarose gel electrophoresis. The Ndel-Hindlll fragment was ligated into plasmid pSP26: Amp:Fl between the Ndel and Hindlll sites, resulting in plasmid pAL52. The phoA promoter region and signal peptide was generated by amplification of a portion of the E. coli genome by PCR, using oligonucleotide primers 405 and 406.

405: CCGGACGCGTGGAGATTATCGTCACTG 406: GCTTTGGTCACCGGGGTAAACAGTAACGG

The resulting PCR product is a 332 bp ΛTluI -Bs EII fragment which contains the phoA promoter region and signal peptide sequence. This fragment was used to replace the 148 bp MluI-BstEII segment of PAL52, resulting in vector pAL53.

G. Construction of pSP26:Amp:Fl :PhoA:KPI:gIII Construction of pSP26: Amp:F1: hoA:KPI:gIII is illustrated in Figure 36. This particular vector is the source of the KPI coding sequence found in vector pDWl #14. Plasmid pPhoa:KPI:gIII was digested with Ndel and Hindlll, and the resulting 714 bp Ndel-Hindlll fragment was purified, and then inserted into vector pSP26: Amp:Fl between the Ndel and Hindlll sites. The resulting plasmid is denoted pSP26:Amp:Fl:PhoA:KPI:gIII.

H. Construction of pDWl #14

Construction of pDWl #14 is illustrated in Figure 37. The sequences encoding KPI were amplified from plasmid pSP26:Amp:Fl: hoA:KPI:gIII by PCR, using oligonucleotide primers 424 and 425.

424: CTGTTTACCCCGGTGACCAAAGCCGAGGTGTGCTCTGAACAA 425: AATAGCGGATCCGCACACTGCCATGCAGTACTCTTC

The resulting 172 bp BstEII-BamHI fragment encodes most of KPI (ι-»55) . This fragment was used to replace the stuffer region in pAL53 between the BstEII and BamHI sites. The resulting plasmid, PDWl #14, is the parent KPI phage display vector for preparation of randomized KPI phage libraries. The coding region for the phoA-KPI (l-*55) -genelll fusion is shown in Figure 38. I. Construction of pDWl 14 -2

Construction of pDWl 14-2 is illustrated in Figure 39. The first step in the construction of the KPI phage libraries in pDWl #14 was the replacement of the Agrel-BamHI fragment within the KPI coding sequence with a stuffer fragment. This greatly aids in preparation of randomized KPI libraries which are substantially free of contamination of phagemid genomes encoding wild-type KPI sequence. Plasmid pDWl #14 was digested with Agrel and BamHI, and the 135 bp Agrel-BamHI fragment encoding KPI was discarded. A stuffer fragment was created by PCR amplification of a portion of the PBR322 Tet gene, extending from the BamHI site at nucleotide 375 to nucleotide 1284, using oligo primers 266 and 252.

266: GCTTTAAACCGGTAGGTGGCCCGGCTCCATGCACC 252: CGAATTCACCGGTGTCATCCTCGGCACCGTCACCCT

The resulting 894 bp Agrel-BamHI stuffer fragment was then inserted into the Agel/BamEI-digested pDWl #14 to yield the phagemid vector pDWl 14-2. This vector was the starting point for construction of the randomized KPI libraries.

J. Construction of KPI Library 16-19

Construction of KPI Library 16-19 is outlined in Figure 40. Library 16-19 was constructed to display KPI-genelll fusions in which amino acid positions Ala¹⁴, Met", He¹⁶ and Ser¹⁷ are randomized. For preparation of the library, plasmid pDWl 14-2 was digested with Ag-el and BamHI to remove the stuffer region, and the resulting vector was purified by preparative agarose gel electrophoresis. Plasmid pDWl #14 was used as template in a PCR amplification of the KPI region extending from the Agrel site to the BamHI site. The oligonucleotide primers used were 544 and 551.

544: GGGCTGAGACCGGTCCGTGCCGT(NNS)₄CGCTGGTACTTTGACGTC 551: GGAATAGCGGATCCGCACACTGCCATGCAG

Oligonucleotide primer 544 contains four randomized codons of the sequence NNS, where N represents equal mixtures of A/G/C/T and S an equal mixture of G or C. Each NNS codon thus encodes all 20 amino acids plus a single possible stop codon, in 32 different DNA sequences. PCR amplification from the wild-type KPI gene resulted in the production of a mixture of 135 bp Agrel- BamHI fragments all containing different sequences in the randomized region. The PCR product was purified by preparative agarose gel electrophoresis and ligated into the Agel/BamHI digested PDWl 14-2 vector. The ligation mixture was used to transform E. coli ToplOF¹ cells

(Invitrogen) by electroporation according to the manufacturer's directions. The resulting Library 16-19 contained approximately 400,000 independent clones. The potential size of the library, based upon the degeneracy of the priming PCR oligo #544 was 1,048,576 members. The expression unit encoded by the members of Library 16-19 is shown in Figure 41.

K. Selection of Library 16-19 with human plasma kallikrein

KPI phage were prepared and amplified by infecting transformed cells with M13K07 helper phage as described by Matthews et al . , Science 260:1113 (1993). Human plasma kallikrein (Enzyme Research Laboratories, South Bend, IN) , was coupled to Sepharose 6B resin. Prior to phage binding, the immobilized kallikrein resin was washed three times with 0.5 ml assay buffer (AB = lOOmM Tris-HCl, pH 7.5, 0.5M NaCl, 5mM each of KC1, CaCl₂, MgCl₂, 0.1% gelatin, and 0.05% Triton X-100) . Approximately 5xl0⁹ phage particles of the amplified Library 16-19 in PBS, pH 7.5, containing 300mM NaCl and 0.1% gelatin, were bound to 50 μl kallikrein resin containing 15 pmoles of active human plasma kallikrein in a total volume of 250 μl. Phage were allowed to bind for 4 h at room temperature, with rocking. Unbound phage were removed by washing the kallikrein resin three times in 0.5 ml AB. Bound phage were eluted sequentially by successive 5 minute washes: 0.5 ml 50mM sodium citrate, pH 6.0, 150mM NaCl; 0.5 ml 50mM sodium citrate, pH 4.0, 150mM NaCl; and 0.5 ml 50mM glycine, pH 2.0, 150mM NaCl. Eluted phage were neutralized immediately and phagemids from the pH 2.0 elution were titered and amplified for reselection. After three rounds of selection on kallikrein-Sepharose, phagemid DNA was isolated from 22 individual colonies and subjected to DNA sequence analysis.

The most frequently occurring randomized KPI region encoded: Ala¹⁴-Ala¹⁵-Ile¹⁶-Phe¹⁷. The phoA-KPI-genelll region encoded by this class of selected KPI phage is shown in Figure 42. The KPI variant encoded by these phagemids is denoted KPI (1→55; M15A, S17F) .

L. Construction of pDD185 KPI ( -4→57; M15A, S17F) Figure 43 outlines the construction of pDD185 KPI

(-4-*57; M15A, S17F) . The sequences encoding KPI (l-*55;

M15A, S17F) were moved from one phagemid vector, pDWl (16-19) 185, to the yeast expression vector so that the

KPI variant could be purified and tested.

Plasmid pTW113 encoding wild-type KPI (-4-*57) was digested with Agrel and BamHI and the 135 bp Agrel-BamHI fragment was discarded. The 135 bp Agrel-BamHI fragment of pDWl (16-19) 185 was isolated and ligated into the yeast vector to yield plasmid pDD185, encoding α-factor fused to KPI (-4→57; M15A, S17F) . See Figure 44.

Λ. Purification of KPI ( -4→57; M15A, S17F) pDD185 Transformation of yeast strain ABL115 with pDD185, induction of yeast cultures, and purification of KPI (-4→57; M15A, S17F) pDD185 was accomplished as described for the other KPI variants.

N. Construction of KPI Library 6 — M15A, with residues 14, 16-18 random.

Library 6 was constructed to display KPI-genelll fusions in which amino acid positions Ala¹⁴, He¹⁶, Ser¹⁷ and Arg¹⁸ are randomized, but position 15 was held constant as Ala. For preparation of the library, plasmid pDWl #14 was used as template in a PCR amplification of the KPI region extending from the Agrel site to the BamHI site. The oligonucleotide primers used were 551 and 1003.

1003: GCTGAGACCGGTCCGTGCCGTNNSGCA(NNS)₃TGGTACTTTGACGTC

551: GGAATAGCGGATCCGCACACTGCCATGCAG

Oligonucleotide primer 1003 contained four randomized codons of the sequence NNS, where N represents equal mixtures of A/G/C/T and S an equal mixture of G or C. Each NNS codon thus encodes all 20 amino acids plus a single possible stop, in 32 different DNA sequences. PCR amplification from the wild-type KPI gene resulted in the production of a mixture of 135 bp Agrel-BamHI fragments all containing different sequences in the randomized region. The PCR product was phenol extracted, ethanol precipitated, digested with BamHI and purified by preparative agarose gel electrophoresis. Plasmid pDWl 14-2 was digested with BamHI, phenol extracted and ethanol precipitated. The insert was ligated at high molar ratio to the vector which was then digested with Agrel to remove the stuffer region. The vector containing the insert was purified by agarose gel electrophoresis and recircularized. The resulting library contains approximately 5xl0⁶ independent clones.

0. Construction of KPI Library 7 — residues 14 -18 random.

Library 7 was constructed to display KPI-genelll fusions in which amino acid positions Ala¹⁴, Met¹⁵, He¹⁶, Ser¹⁷ and Arg¹⁸ are randomized. For preparation of the library, plasmid pDWl #14 was used as template in a PCR amplification of the KPI region extending from the Agrel site to the BamHI site. The oligonucleotide primers used were 551 and 1179. 1179: GCTGAGACCGGTCCGTGCCGT(NNS)jTGGTACTTTGACGTC

551: GGAATAGCGGATCCGCACACTGCCATGCAG

Oligonucleotide primer 1179 contains five randomized codons of the sequence NNS, where N represents equal mixtures of A/G/C/T and S an equal mixture of G or C. Each NNS codon thus encoded all 20 amino acids plus a single possible stop, in 32 different DNA sequences. PCR amplification from the wild-type KPI gene resulted in the production of a mixture of 135 bp Agrel-BamHI fragments all containing different sequences in the randomized region. The PCR product was phenol extracted, ethanol precipitated, digested with BamHI and purified by preparative agarose gel electrophoresis. Plasmid pDWl 14-2 was digested with BamHI, phenol extracted and ethanol precipitated. The insert was ligated at high molar ratio to the vector which was then digested with Agrel to remove the stuffer region. The vector containing the insert was purified by agarose gel electrophoresis and recircularized. The resulting library contains approximately lxlO⁷ independent clones.

P. Selection of Libraries 6 & 7 with human factor Xlla

KPI phage were prepared and amplified by infecting transformed cells with M13K07 helper phage (Matthews and Wells, 1993) . Human factor Xlla (Enzyme Research Laboratories, South Bend, IN), was biotinylated as follows. Factor Xlla (0.5 mg) in 5mM sodium acetate pH 8.3 was incubated with Biotin Ester (Zymed) at room temperature for 1.5 h, then buffer-exchanged into assay buffer (AB) . Approximately lxlO¹⁰ phage particles of each amplified Library 6 or 7 in PBS, pH 7.5, containing 300mM NaCl and 0.1% gelatin, were incubated with 50 pmoles of active biotinylated human factor Xlla in a total volume of 200 μl. Phage were allowed to bind for 2 h at room temperature, with rocking. Following the binding period, 100 μl Strepavidin Magnetic Particles (Boehringer Mannheim) were added to the mixture and incubated at room temperature for 30 minutes. Separation of magnetic particles from the supernatant and wash/elution buffers was carried out using MPC-E-1 Neodymium-iron-boron permanent magnets (Dynal) . Unbound phage were removed by washing the magnetically bound biotinylated XHa-phage complexes three times with 0.5 ml AB. Bound phage were eluted sequentially by successive 5 minute washes: 0.5 ml 50mM sodium citrate, pH 6.0, 150mM NaCl; 0.5 ml 50mM sodium citrate, pH 4.0, 150mM NaCl; and 0.5 ml 50mM glycine, pH 2.0, 150mM NaCl. Eluted phage were neutralized immediately and phagemids from the pH 2.0 elution were titered and amplified for reselection. After 3 or 4 rounds of selection with factor Xlla, phagemid DNA was isolated from individual colonies and subjected to DNA sequence analysis.

Sequences in the randomized regions were compared with one another to identify consensus sequences appearing more than once. From Library 6 a phagemid was identified which encoded M15L, S17Y, R18H. From Library 7 a phagemid was identified which encoded M15A, S17Y, R18H.

0. Construction of pBGOlS KPI ( -4→57; M15L, S17Y, R18H) , pBG022 ( -4→57; M1SA, S17Y, R18H)

The sequences encoding KPI (l-»55; M15L, S17Y, R18H) and KPI (1→55; M17A, S17Y, R18H) were moved from the phagemid vectors to the yeast expression vector so that the KPI variant could be purified and tested.

Plasmid pTW113 encoding wild-type KPI (-4-_*57) was digested with Agel and BamHI and the 135 bp Ag l-BamHI fragment was discarded. The 135 bp Agel-BamHI fragment of the phagemid vectors were isolated and ligated into the yeast vector to yield plasmids pBG015 and pBG022, encoding alpha-factor fused to KPI (-4→57; M15L, S17Y, R18H) , and KPI (-4→57; M15A, S17Y, R18H) , respectively. R. Construction of pBG029 KPI ( -4→5 , T9V, M15L, S17Y, R18H)

Plasmid pBG015 was digested with .Xbal and RsrII, and the larger of the two resulting fragments was isolated. An oligonucleotide pair (1593 + 1642) was phosphorylated, annealed and gel-purified as described previously.

1593: CTAGATAAAAGAGAGGTTGTTAGAGAGGTGTGCTCTGAACAAGCT GAGGTTG

1642: GACCAACCTCAGCTTGTTCAGAGCACACCTCTCTAA CAACCTCTCTTTTAT

The annealed oligonucleotides were ligated into the Xbal and RsrII-digested pBG015, and the ligation product was used to transform E. coli strain MC1061 to ampicillin resistance. The resulting plasmid pBG029, encodes the 445 bp synthetic gene for the alpha-factor-KPI (-4-*57; T9V, M15L, S17F, R18H) fusion.

S. Construction of pBG033 KPI ( -4→57; T9V, M15A, S17Y, R18H)

Plasmid pBG022 was digested with Xbal and RsrII, and the larger of the two resulting fragments was isolated. An oligonucleotide pair (1593 + 1642) was phosphorylated, annealed and gel-purified as described previously. The annealed oligonucleotides were ligated into the Xbal and RsrII-digested pBG022, and the ligation product was used to transform E. coli strain MC1061 to ampicillin resistance. The resulting plasmid pBG033, encodes the 445 bp synthetic gene for the alpha-factor-KPI (-4→57; T9V, M15A, S17F, R18H) fusion.

T. Selection of Library 16-19 wi th human factor Xa

KPI phage were prepared and amplified by infecting transformed cells with M13K07 helper phage (Matthews and Wells, 1993) . Human factor Xa (Haematologic Technologies, Inc., Essex Junction, VT) was coupled to Sepharose 6B resin. Prior to phage binding, the immobilized Xa resin was washed three times with 0.5 ml assay buffer (AB = lOOmM Tris-HCl, pH 7.5, 0.5M NaCl, 5mM each of KC1, CaCl₂, MgCl₂, 0.1% gelatin, and 0.05% Triton X-100) . Approximately 4xl0¹⁰ phage particles of the amplified Library 16-19 in PBS, pH 7.5, containing 300mM NaCl and 0.1% gelatin, were bound to 50 μl Xa resin in a total volume of 250 μl. Phage were allowed to bind for 4 h at room temperature, with rocking. Unbound phage were removed by washing the Xa resin three times in 0.5 ml AB. Bound phage were eluted sequentially by successive 5 minute washes: 0.5 ml 50mM sodium citrate, pH 6.0, 150mM NaCl; 0.5 ml 50mM sodium citrate, pH 4.0 150mM NaCl; and 0.5 ml 50mM glycine, pH 2.0, 150mM NaCl. Eluted phage were neutralized immediately and phagemids from the pH 2.0 elution were titered and amplified for reselection. After three rounds of selection on Xa- Sepharose, phagemid DNA was isolated and subjected to DNA sequence analysis.

Sequences in the randomized Ala¹⁴-Ser¹⁷ region were compared with one another to identify consensus sequences appearing more than once. A phagemid was identified which encoded KPI (1→55; M15L, I16F, S17K) .

ϋ. Construction of pDD131 KPI ( -4→57; M15L, I16F, S17K)

The sequences encoding KPI (l-»55; M15L, I16F, S17K) were moved from the phagemid vector to the yeast expression vector so that the KPI variant could be purified and tested.

Plasmid pTW113 encoding wild-type KPI (-4→57) was digested with Agrel and BamHI and the 135 bp Agrel-BamHI fragment was discarded. The 135 bp Agel-Ba HI fragment of the phagemid vector was isolated and ligated into the yeast vector to yield plasmid pDD131, encoding alpha- factor fused to KPI (-4→57; M15L, I16F, S17K) . V. Construction of pDD134 KPI ( -4→57; M15L, I16F, S17K, G37Y)

Plasmid pDD131 was digested with Aatl and BamHI, and the larger of the two resulting fragments was isolated. An oligonucleotide pair (738 + 739) was phosphorylated, annealed and gel-purified as described previously.

738: ClACTGAAGGTAAGTGCGCTCCΑTTC l lACGGCGGTTGCrrACGGCAACCGT AACAACTTTGACACTGAAGAGTACTGCATGGCAGTGTGCG

739: GATCCGCACΑCrraCCΑTGCAGTACTCriTC^ CGTAGCAACCGCCGTAAAAGAATGGAGCGCACTTACCTTCAGTGACGT

The annealed oligonucleotides were ligated into the Aatl and BamHI-digested pDD131, and the ligation product was used to transform E. coli strain MC1061 to ampicillin resistance. The resulting plasmid pDD134, encodes the 445 bp synthetic gene for the alpha-factor-KPI (-4→57; M15L, I16F, S17K, G37Y) fusion.

W. Construction of pDD135 KPI ( -4→57; M15L, I16F, S17K, G37L)

Plasmid pDD131 was digested with Aatll and BamHI, and the larger of the two resulting fragments was isolated. An oligonucleotide pair (724 + 725) was phosphorylated, annealed and gel-purified as described previously.

738: ClACTGAAGGTAAGTGCGCrCCArr ri ACGGCGGTTGCTACGGCAACCGT AACAACTTTGACACTGAAGAGTACTGCATGGCAGTGTGCG

739: GATCCGCΑCΑCTOCCΑTG<^GTACTCTTC^

CGTAGCAACCGCCGTAAAAGAATGGAGCGCACTTACCTTCAGTGACGT

The annealed oligonucleotides were ligated into the

Aatll and BamHI-digested pDD131, and the ligation product was used to transform E. coli strain MC1061 to ampicillin resistance. The resulting plasmid pDD135, encodes the 445 bp synthetic gene for the alpha-factor-KPI (-4-»57; M15L, I16F, S17K, G37L) fusion.

Example 4. Kinetic analysis of KPI(-4-»57) variants

The concentrations of active human plasma kallikrein, factor Xlla, and trypsin were determined by titration with p-nitrophenyl p' -guanidinobenzoate as described by Bender et al., supra, and Chase et al., Biochem. Biophys . Res. Commun. 29:508 (1967). Accurate concentrations of active KPI(-4→57) inhibitors were determined by titration of the activity of a known amount of active-site-titrated trypsin. For testing against kallikrein and trypsin, each KPI(-4-_*57) variant (0.5 to lOOnM) was incubated with protease in low-binding 96-well microtiter plates at 30°C for 15-25 min, in lOOmM Tris-HCl, pH 7.5, with 500mM NaCl, 5mM KC1, 5mM CaC12, 5mMMgC12, 0.1% Difco gelatin, and 0.05% Triton X-100. Chromogenic synthetic substrate was then be added, and initial rates at 30°C recorded by the SOFTmax kinetics program via a THERMOmax microplate reader (Molecular Devices Corp., Menlo Park, CA) . The substrates used were N-α-benzoyl-L-Arg p-nitroanilide

(lmM) for trypsin (20nM) , and N-benzoyl-Pro-Phe-Arg p- nitroanilide (0.3mM) for plasma kallikrein (InM) . The

Enzfitter (Elsevier) program was used both to plot fractional activity (i.e., activity with inhibitor, divided by activity without inhibitor) , a, versus total concentration of inhibitor, I,, and to calculate the dissociation constant of the inhibitor (Kj) by fitting the curve to the following equation:

,-ι ^[E] <^{+ [ I} -^+K-~J ^{{ [E]} *^{+ l I] t+}*i>^a-* ^lE] ttJl _e

2 [E] _e

The K;S determined for purified KPI variants are shown in Figure 45. The most potent variants, KPI

(-4→57; M15A, S17F) DD185 and KPI (-4→57; M15A, S17Y)

TW6166 are 115-fold and 100-fold more potent, respectively, as a human kallikrein inhibitor than wild- type KPI (-4-*57). The least potent variant, KPI (-4→57; I16H, S17W) TW6185 is still 35-fold more potent than wild-type KPI. For testing against factor Xlla, essentially the same reaction conditions were used, except that the substrate was N-benzoyl-He-Glu-Gly-Arg p-nitroaniline hydrochloride and its methyl ester (obtained from Pharmacia Hepar, Franklin, OH) , and corn trypsin inhibitor (Enzyme Research Laboratories, South Bend, IN) was used as a control inhibitor. Factor Xlla was also obtained from Enzyme Research Laboratories.

Various data for inhibition of the serine proteases of interest kallikrein, plasmin, and factors Xa, Xla, and Xlla by a series of KPI variants are given in Figure 46. The results indicate that KPI variants can be produced that can bind to and preferably inhibit the activity of serine proteases. The results also indicate that the peptides of the invention may exhibit the preferable more potent and specific inhibition of one or more serine proteases of interest.

Example 5. Effect of KPI variant KPI185-1 on postoperative bleeding

A randomized, double-blinded study using an acute porcine cardiopulmonary bypass (CPB) model was used to investigate the effect of KPI185-1 on postoperative bleeding. Sixteen pigs (55-65 kg) underwent 60 minutes of hypothermic (28°C) open-chest CPB with 30 minutes of cardioplegic cardiac arrest. Pigs were randomized against a control solution of physiological saline (NS; n=8) or KPI-185 (n=8) groups. During aortic cross- clamping, the tricuspid valve was inspected through an atriotomy which was subsequently repaired. Following reversal of heparin with protamine, dilateral thoracostomy tubes were placed and shed blood collected for 3 hours. Shed blood volume and hemoglobin (Hgb) loss were calculated from total chest tube output and residual intrathoracic blood at time of sacrifice. Total blood loss was significantly reduced in the

KPI185-1 group (245.75 ± 66.24 ml vs. 344.25 ± 63.97 ml, p=0.009). In addition, there was a marked reduction in total Hgb loss in the treatment group (13.59 ± 4.26 gm vs. 23.61 ± 4.69 gm, p=0.0005). Thoracostomy drainage

Hgb was significantly increased at 30 and 60 minutes in the control group [6.89 ± 1.44 vs. 4.41 ± 1.45 gm/dl

(p=0.004) and 7.6 ± 1.03 vs. 5.26 ± 1.04 gm/dl

(p=0.000.2), respectively]. Preoperative and post-CPB hematocrits were not statistically different between the groups. These results are shown in graphical form in Figures 47-50.

The invention has been disclosed broadly and illustrated in reference to representative embodiments described above. Those skilled in the art will recognize that various modifications can be made to the present invention without departing from the spirit and scope thereof.

Claims (133)

What Is Claimed Is:

1. A protease inhibitor comprising the sequence:

X¹-Val-Cys-Ser-Glu-Gln-Ala-Glu-X²-Gly-X³- Cys-Arg-Ala-X⁴-X⁵-X⁶-X⁷-Trp-Tyr- Phe-Asp- Val-Thr-Glu-Gly-Lys-Cys-Al -Pro-Phe-X⁸- Tyr-Gly-Gly-Cys-X⁹-X^I0-X^u-X^I -Asn-Asn-Phe- Asp-Thr-Glu-Glu-Tyr-Cys-Met-Ala-Val-Cys- Gly-Ser-Ala-He, wherein:

X¹ is selected from Glu-Val-Val-Arg-Glu-, Asp, or Glu;

X² is selected from Thr, Val, He and Ser; X³ is selected from Pro and Ala; X⁴ is selected from Arg, Ala, Leu, Gly, or Met; X^s is selected from He, His, Leu, Lys, Ala, or Phe; X⁶ is selected from Ser, He, Pro, Phe, Tyr, Trp, Asn, Leu, His, Lys, or Glu;

X⁷ is selected from Arg, His, or Ala; X⁸ is selected from Phe, Val, Leu, or Gly; X⁹ is selected from Gly, Ala, Lys, Pro, Arg, Leu, Met, or Tyr;

X¹⁰ is selected from Ala, Arg, or Gly; X^u is selected from Lys, Ala, or Asn; X¹² is selected from Ser, Ala, or Arg; provided that: when X⁴ is Arg, X⁶ is lie; when X⁹ is Arg, X⁴ is Ala or Leu; when X⁹ is Tyr, X⁴ is Ala or X⁵ is His; and either X^s is not He; or X⁶ is not Ser; or X⁹ is not Leu, Phe, Met, Tyr, or Asn; or X¹⁰ is not Gly; or X¹¹ is not Asn; or X¹² is not Arg.

2. A protease inhibitor comprising the sequence:

X'-Val-Cys-Ser-Glu-Gln-Ala-Glu-Thr-Gly- Pro-Cys-Arg-Ala-X²-X³-X⁴-Arg-Trp-Tyr-Phe- Asp-Val-Thr-Glu-Gly-Lys-Cys-Ala-Pro-Phe- Phe-Tyr-Gly-Gly-Cys-X⁵-Gly-Asn-Arg-Asn- Asn-Phe-Asp-Thr-Glu-Glu-Tyr-Cys-Met- la- Val-Cys-Gly-Ser-Ala-He, wherein:

X¹ is selected from Glu-Val-Val-Arg-Glu-, Asp, or Glu;

X² is selected from Ala, Leu, Gly, or Met;

X³ is selected from He, His, Leu, Lys, Ala, or Phe;

X⁴ is selected from Ser, He, Pro, Phe, Tyr, Trp, Asn, Leu, His, Lys, or Glu;

X^s is selected from Gly, Ala, Lys, Pro, Arg, Leu, Met, or Tyr; provided that: when X⁵ is Arg, X² is Ala or Leu; when X^s is Tyr, X² is Ala or X³ is His; and either X³ is not He; or X⁴ is not Ser; or X⁵ is not Leu, Phe, Met, Tyr, or Asn.

3. A protease inhibitor comprising the sequence:

Glu-Val-Val-Arg-Glu-Val-Cys-Ser-Glu-Gln- Ala-Glu-Thr-Gly-Pro-Cys-Arg-Ala-X¹-X²-X³- Arg-Trp-Tyr-Phe-Asp-Val-Thr-Glu-Gly-Lys- Cys-Ala-Pro-Phe-Phe-Tyr-Gly-Gly-Cys-X⁴- Gly-Asn-Arg-Asn-Asn-Phe-Asp-Thr-Glu-Glu- Tyr-Cys-Met-Ala-Val-Cys-Gly-Ser-Ala-He, wherein:

X¹ is selected from Ala, Leu, Gly, or Met; X² is selected from He, His, Leu, Lys, Ala, or Phe; X³ is selected from Ser, He, Pro, Phe, Tyr, Trp, Asn, Leu, His, Lys, or Glu;

X⁴ is selected from Gly, Arg, Leu, Met, or Tyr; provided that: when X¹ is Ala, X² is He, His, or Leu; when X¹ is Leu, X² is He or His; when X¹ is Leu and X² is He, X³ is not Ser; when X¹ is Gly, X² is lie; when X⁴ is Arg, X¹ is Ala or Leu; when X⁴ is Tyr, X¹ is Ala or X² is His; and either X¹ is not Met, or X² is not He, or X³ is not Ser, or X⁴ is not Gly.

4. A protease inhibitor according to claim 1, wherein at least two amino acid residues selected from the group consisting of X⁴, X⁵, X⁶, and X⁷ differ from the residues found in the naturally occurring sequence of KPI.

5. A protease inhibitor according to claim 1, wherein X¹ is Asp or Glu, X² is Thr, X³ is Pro, and X¹² is Ser.

6. A protease inhibitor according to claim 5, wherein X¹ is Glu, X² is Thr, X³ is Pro, X⁴ is Met, X⁵ is He, X⁶ is Ser, X⁷ is Arg, x⁸ is Phe, X⁹ is Gly, X¹⁰ is Gly, and X¹¹ is Asn.

7. A protease inhibitor according to claim 5, wherein X¹ is Asp, X² is Thr, X³ is Pro, X⁴ is Arg, X^s is He, X⁶ is He, X⁷ is Arg, x⁸ is Val, X⁹ is Arg, X¹⁰ is Ala, and X¹¹ is Lys.

8. A protease inhibitor according to claim 1, wherein X¹ is Glu-Val-Val-Arg-Glu-, X² is Thr, X³ is Pro, X⁴ is Met, X⁵ is He, X⁶ is Ser, X⁷ is Arg, x⁸ is Phe, X⁹ is Gly, X¹⁰ is Gly, X¹¹ is Asn, and X¹² is Ala.

9. A protease inhibitor according to claim 1, wherein X¹ is Glu-Val-Val-Arg-Glu-, X² is Thr, X³ is Pro, X⁴ is Met, X⁵ is He, X⁶ is Ser, X⁷ is Arg, x⁸ is Phe, X⁹ is Gly, X¹⁰ is Gly, X¹¹ is Ala, and X¹² is Arg.

10. A protease inhibitor according to claim 1, wherein X¹ is Glu, X² is Thr, X³ is Pro, X⁴ is Met, X^J is He, X⁶ is Ser, X⁷ is Arg, x⁸ is Phe, X⁹ is Gly, X¹⁰ is Ala, X¹¹ is Asn, and X¹² is Arg.

11. A protease inhibitor according to claim 1, wherein X¹ is Glu-Val-Val-Arg-Glu-, X² is Thr, X³ is Pro, X⁴ is Met, X⁵ is He, X⁶ is Ser, X⁷ is Arg, x⁸ is Phe, X⁹ is Gly, X¹⁰ is Arg, x" is Asn, and X¹² is Arg.

12. A protease inhibitor according to claim 1, wherein X¹ is Glu-Val-Val-Arg-Glu-, X² is Thr, X³ is Pro, X⁴ is Met, X^s is He, X⁶ is Ser, X⁷ is Arg, x⁸ is Val, Leu, or Gly, X⁹ is Gly, X¹⁰ is Gly, X¹¹ is Asn, and X¹² is Arg.

13. A protease inhibitor according to claim 1, wherein X¹ is Glu-Val-Val-Arg-Glu-, X² is Thr, X³ is Pro, X⁴ is Met, X^s is He, X⁶ is Ser, X⁷ is Ala, x⁸ is Phe, X⁹ is Gly, X¹⁰ is Gly, X¹¹ is Asn, and X¹² is Arg.

14. A protease inhibitor according to claim 1, wherein X¹ is Glu-Val-Val-Arg-Glu-, X² is Thr, Val, or Ser, X³ is Pro, X⁴ is Ala or Leu, X⁵ is He, X⁶ is Tyr, X⁷ His, X⁸ is Phe, X⁹ is Gly, X¹⁰ is Gly, X¹¹ is Ala, and X¹² is Arg.

15. A protease inhibitor according to claim 14, wherein X² is Thr, and X⁴ is Ala.

16. A protease inhibitor according to claim 14, wherein X² is Thr, and X⁴ is Leu.

17. A protease inhibitor according to claim 14, wherein X² is Val, and X⁴ is Ala.

18. A protease inhibitor according to claim 14, wherein X² is Ser, and X⁴ is Ala.

' 19. A protease inhibitor according to claim 14, wherein X² is Val, and X⁴ is Leu.

20. A protease inhibitor according to claim 14, wherein X² is Ser, and X⁴ is Leu.

21. A protease inhibitor according to claim 1, wherein X¹ is Glu-Val-Val-Arg-Glu-, X² is Thr, X³ is Pro, X⁴ is Leu, X⁵ is Phe, X⁶ is Lys, X⁷ is Arg, X⁸ is Phe, X⁹ is Gly, X¹⁰ is Gly, x" is Ala, and X¹² is Arg.

22. A protease inhibitor according to claim 1, wherein X¹ is Glu-Val-Val-Arg-Glu-, X² is Thr, X³ is Pro, X⁴ is Leu, X⁵ is Phe, X⁶ is Lys, X⁷ is Arg, X⁸ is Phe, X⁹ is Tyr, X¹⁰ is Gly, X¹¹ is Ala, and X¹² is Arg.

23. A protease inhibitor according to claim 1, wherein X¹ is Glu-Val-Val-Arg-Glu-, X² is Thr, X³ is Pro, X⁴ is Leu, X⁵ is Phe, X⁶ is Lys, X⁷ is Arg, X⁸ is Phe, X⁹ is Leu, X¹⁰ is Gly, X¹¹ is Ala, and X¹² is Arg.

24. A protease inhibitor according to claim 2, wherein X¹ is Glu, X² is Met, X³ is He, X⁴ is He, and X^s is Gly.

25. A protease inhibitor according to claim 3, wherein X¹ is Met, X³ is Ser, and X⁴ is Gly.

26. A protease inhibitor according to claim 25, wherein X² is selected from His, Ala, Phe, Lys, and Leu.

27. A protease inhibitor according to claim 26, wherein X² is His.

28. A protease inhibitor according to claim 27, wherein X² is Ala.

29. A protease inhibitor according to claim 27, wherein X² is Phe.

30. A protease inhibitor according to claim 27, wherein X² is Lys.

31. A protease inhibitor according to claim 27, wherein X² is Leu.

32. A protease inhibitor according to claim 3, wherein X¹ is Met, X² is He, and X⁴ is Gly.

33. A protease inhibitor according to claim 32, wherein X³ is He.

34. A protease inhibitor according to claim 32, wherein X³ is Pro.

35. A protease inhibitor according to claim 32, wherein X³ is Phe.

36. A protease inhibitor according to claim 32, wherein X³ is Tyr.

37. A protease inhibitor according to claim 32, wherein X³ is Trp.

38. A protease inhibitor according to claim 32, wherein X³ is Asn.

39. A protease inhibitor according to claim 32, wherein X³ is Leu.

40. A protease inhibitor according to claim 32, wherein X³ is Lys.

41. A protease inhibitor according to claim 32, wherein X³ is His.

42. A protease inhibitor according to claim 32, wherein X³ is Glu.

43. A protease inhibitor according to claim 3, wherein X¹ is Ala.

44. A protease inhibitor according to claim 43, wherein X² is He.

45. A protease inhibitor according to claim 44, wherein X³ is Phe, and X⁴ is Gly.

46. A protease inhibitor according to claim 44, wherein X³ is Tyr, and X⁴ is Gly.

47. A protease inhibitor according to claim 44, wherein X³ is Trp, and X⁴ is Gly.

48. A protease inhibitor according to claim 44, wherein X³ is Ser or Phe, and X⁴ is Arg or Tyr.

49. A protease inhibitor according to claim 43, wherein X² is His or Leu, X³ is Phe, and X⁴ is Gly.

50. A protease inhibitor according to claim 3, wherein X¹ is Leu.

51. A protease inhibitor according to claim 50, wherein X² is His, X³ is Asn or Phe, and X⁴ is Gly.

52. A protease inhibitor according to claim 50, wherein X² is He, X³ is Pro, and X⁴ is Gly.

53. A protease inhibitor according to claim 3, wherein X¹ is Gly, X² is He, X³ is Tyr, and X⁴ is Gly.

54. A protease inhibitor according to claim 3, wherein X¹ is Met, X² is His, X³ is Ser, and X⁴ is Tyr.

55. An isolated DNA molecule comprising a DNA sequence encoding a protease inhibitor according to claim l.

56. An isolated DNA molecule according to claim 55, operably linked to a regulatory sequence that controls expression of the coding sequence in a host cell.

57. An isolated DNA molecule according to claim 56, further comprising a DNA sequence encoding a secretory signal peptide.

58. An isolated DNA molecule according to claim 57, wherein said secretory signal peptide comprises the signal sequence of yeast alpha-mating factor.

59. A host cell transformed with a DNA molecule according to claim 55.

60. A host cell according to claim 59, wherein said host cell is E. coli or a yeast cell.

61. A host cell according to claim 60, wherein said yeast cell is selected from Pichia pastoris and Saccharomyces cerevisiae.

62. A method for producing a protease inhibitor, comprising the steps of culturing a host cell according to claim 59 and isolating and purifying said protease inhibitor.

63. A pharmaceutical composition, comprising a protease inhibitor according to claim 1, together with a pharmaceutically acceptable sterile vehicle.

64. A method of treatment of a clinical condition associated with increased activity of one or more serine proteases, comprising administering to a patient suffering from said clinical condition an effective amount of a pharmaceutical composition according to claim 63.

65. The method of treatment of claim 64, wherein said clinical condition is blood loss during surgery.

66. A method for inhibiting the activity of serine proteases of interest in a mammal comprising administering a therapeutically effective dose of a pharmaceutical composition according to claim 63.

67. The method of claim 66, wherein said serine proteases are selected from the group consisting of: kallikrein; chymotrypsins A and B; trypsin; elastase; subtilisin; coagulants and procoagulants, particularly those in active form, including coagulation factors such as factors Vila, IXa, Xa, Xla, and Xlla; plasmin; thrombin; proteinase-3; enterokinase; acrosin; cathepsin; urokinase; and tissue plasminogen activator.

68. A protease inhibitor comprising the sequence: x'-Val-Cys-Ser-Glu-Gln-Ala-Glu-X²-Gly-Pro- Cys-Arg-Ala-X³-X⁴-X⁵-X⁶-Trp-Tyr-Phe-Asp- Val-Thr-Glu-Gly-Lys-Cys-Ala-Pro-Phe-Phe- Tyr-Gly-Gly-Cys-X⁷-Gly-Asn-Arg-Asn-Asn- Phe-Asp-Thr-Glu-Glu-Tyr-Cys-Met-Ala-Val- Cys-Gly-Ser-Ala-He, wherein:

X¹ is selected from Glu-Val-Val-Arg-Glu-, Asp, or Glu;

X² is selected from Thr, Val, He and Ser; X³ is selected from Arg, Ala, Leu, Gly, or Met; X⁴ is selected from He, His, Leu, Lys, Ala, or Phe; X⁵ is selected from Ser, He, Pro, Phe, Tyr, Trp, Asn, Leu, His, Lys, or Glu;

X⁶ is selected from Arg, His, or Ala; and X⁷ is selected from Gly, Ala, Lys, Pro, Arg, Leu, Met, or Tyr.

69. A protease inhibitor according to claim 68, wherein at least two amino acid residues selected from the group consisting of X³, X⁴, X⁵, and X⁶ differ from the residues found in the naturally occurring sequence of KPI.

70. A protease inhibitor according to claim 68, wherein X¹ is Glu-Val-Val-Arg-Glu-, X² is Thr, Val, or Ser, X³ is Ala or Leu, X⁴ is He, X^s is Tyr, X⁶ is His and X⁷is Gly.

71. A protease inhibitor according to claim 70, wherein X² is Thr, and X³ is Ala.

72. A protease inhibitor according to claim 70, wherein X² is Thr, and X³ is Leu.

73. A protease inhibitor according to claim 70, wherein X² is Val, and X³ is Ala.

74. A protease inhibitor according to claim 70, wherein X² is Ser, and X³ is Ala.

75. A protease inhibitor according to claim 70, wherein X² is Val, and X³ is Leu.

76. A protease inhibitor according to claim 70, wherein X² is Ser, and X³ is Leu.

77. A protease inhibitor according to claim 68, wherein X¹ is Glu-Val-Val-Arg-Glu-, X² is Thr, X³ is Leu, X⁴ is Phe, X^s is Lys, X⁶ is Arg and X⁷ is Gly.

78. A protease inhibitor according to claim 68, wherein X¹ is Glu-Val-Val-Arg-Glu-, X² is Thr, X³ is Leu, X⁴ is Phe, X^s is Lys, X⁶ is Arg and X⁷ is Tyr.

79. A protease inhibitor according to claim 68, wherein X¹ is Glu-Val-Val-Arg-Glu-, X² is Thr, X³ is Leu, X⁴ is Phe, X⁵ is Lys, X⁶ is Arg and X⁷ is Leu.

80. A protease inhibitor comprising the sequence:

X'-Val-Cys-Ser-Glu-Gln-Ala-Glu-Thr-Gly- Pro-Cys-X²-Ala-X³-X⁴-X⁵-X⁶-Trp-Tyr-Phe-Asp- Val-Thr-Glu-Gly-Lys-Cys-Ala-Pro-Phe-Phe- Tyr-Gly-Gly-Cys-Gly-Gly-Asn-Arg-Asn-Asn- Phe-Asp-Thr-Glu-Glu-Tyr-Cys-Met-Ala-Val- Cys-Gly-Ser-Ala-He, wherein: X¹ is selected from Glu-Val-Val-Arg-Glu- and Asp- Val-Val-Arg-Glu-;

X² is selected from Arg and Lys;

X³ is selected from Met, Arg, Ala, Leu, Ser, Val;

X⁴ is selected from He and Ala;

X⁵ is selected from Ser, He, Ala, Pro, Phe, Tyr, and Trp; and

X⁶ is selected from Arg, Ala, His, Gin, and Thr; provided that: when X² is Arg, X³ is Leu, and X⁴ is He, X⁵ cannot be Ser; and also provided that either X³ is not Met; or X⁴ is not He; or X⁵ is not Ser; or X⁶ is not Arg.

81. A protease inhibitor according to claim 80, wherein X⁵ is selected from Phe, Tyr and Trp.

82. A protease inhibitor according to claim 80, wherein X⁴ is He.

83. A protease inhibitor according to claim 82, wherein X² is Lys.

84. A protease inhibitor according to claim 83, wherein X³ is Met.

85. A protease inhibitor according to claim 84, wherein X⁵ is Ser.

86. A protease inhibitor according to claim 84, wherein X⁵ is He.

87. A protease inhibitor according to claim 83, wherein X³ is Arg.

88. A protease inhibitor according to claim 87, wherein X^s is Ser.

89. A protease inhibitor according to claim 87, wherein X^s is He.

90. A protease inhibitor according to claim 82, wherein X² is Arg.

91. A protease inhibitor according to claim 90, wherein X³ is Arg or Met, and X^s is Ser or He.

92. A protease inhibitor according to claim 91, wherein X³ is Arg.

93. A protease inhibitor according to claim 92, wherein X⁵ is Ser.

94. A protease inhibitor according to claim 92, wherein X⁵ is He.

95. A protease inhibitor according to claim 91, wherein X³ is Met.

96. A protease inhibitor according to claim 95, wherein X⁵ is Ser.

97. A protease inhibitor according to claim 95, wherein X⁵ is He.

98. A protease inhibitor according to claim 82, wherein X³ is Ala.

99. A protease inhibitor according to claim 82, wherein X³ is Leu.

100. A protease inhibitor according to claim 82, wherein X³ is Ser.

101. A protease inhibitor according to claim 82, wherein X³ is Val.

102. A protease inhibitor according to claim 82, wherein X⁵ is Pro. O 96/35788 PC17US96/06384

- 75 -

103. A protease inhibitor according to claim 82, wherein X⁵ is Phe.

104. A protease inhibitor according to claim 82, wherein X⁵ is Tyr.

105. A protease inhibitor according to claim 82, wherein X⁵ is Trp.

106. A protease inhibitor according to claim 104, wherein X³ is Ala or Leu.

107. A protease inhibitor according to claim 106, wherein X³ is Ala.

108. A protease inhibitor according to claim 106, wherein X³ is Leu.

109. A protease inhibitor according to claim 105, wherein X³ is Ala.

110. A protease inhibitor according to claim 109, wherein X^s is His.

111. A protease inhibitor according to claim 109, wherein X⁵ is Gin.

112. A protease inhibitor according to claim 109, wherein X⁵ is Thr.

113. An isolated DNA molecule comprising a DNA sequence encoding a protease inhibitor according to claim 80.

114. An isolated DNA molecule according to claim 113, operably linked to a regulatory sequence that controls expression of the coding sequence in a host cell.

115. An isolated DNA molecule according to claim

114, further comprising a DNA sequence encoding a secretory signal peptide.

116. An isolated DNA molecule according to claim

115, wherein said secretory signal peptide comprises the signal sequence of yeast alpha-mating factor.

117. A host cell transformed with a DNA molecule according to claim 113.

118. A host cell according to claim 117, wherein said host cell is E. coli or a yeast cell.

119. A host cell according to claim 118, wherein said yeast cell is selected from Pichia pastoris and Saccharomyces cerevisiae.

120. A method for producing a protease inhibitor, comprising the steps of culturing a host cell according to claim 117 and isolating and purifying said protease inhibitor.

121. A pharmaceutical composition, comprising a protease inhibitor according to claim 80, together with a pharmaceutically acceptable sterile vehicle.

122. A method of treatment of a clinical condition associated with increased activity of one or more serine proteases, comprising administering to a patient suffering from said clinical condition an effective amount of a pharmaceutical composition according to claim 121.

123. The method of treatment of claim 122, wherein said clinical condition is blood loss during surgery.

124. A method for inhibiting the activity of serine proteases of interest in a mammal comprising administering a therapeutically effective dose of a pharmaceutical composition according to claim 121.

125. The method of claim 124, wherein said serine proteases are selected from the group consisting of: kallikrein; chymotrypsins A and B; trypsin; elastase; subtilisin; coagulants and procoagulants, particularly those in active form, including coagulation factors such as factors Vila, IXa, Xa, Xla, and Xlla; plasmin; thrombin; proteinase-3; enterokinase; acrosin; cathepsin; urokinase; and tissue plasminogen activator.

126. A protease inhibitor according to claim 81, wherein X⁴ is He.

127. A protease inhibitor according to claim 126, wherein X⁵ is Phe.

128. A protease inhibitor according to claim 126, wherein X⁵ is Tyr.

129. A protease inhibitor according to claim 126, wherein X^s is Trp.

130. A protease inhibitor according to claim 128, wherein X³ is Ala or Leu.

131. A protease inhibitor according to claim 130, wherein X³ is Ala.

132. A protease inhibitor according to claim 130, wherein X³ is Leu.

133. A protease inhibitor according to claim 129, wherein X³ is Ala.