WO2000068363A2 - Elastase variants and substrates - Google Patents

Abstract

The present invention provides elastase variants having one or more amino acid substitutions at selected positions relative to a precursor elastase. In particular, an active-site histidine residue corresponding to histidine number 43 in human neutrophil elastase can be substituted with an amino acid having a volume smaller than that of histidine. In preferred embodiments, elastase variants of the invention have distinctive substrate specificity for substrates containing histidine residues at the position two amino acid residues to the N-terminal side of the site of cleavage. Preferred substrates for the elastase variants are also provided. The elastase variants and substrates of the invention are useful for a variety of therapeutic and non therapeutic applications.

Description

ELASTASE VARIANTS AND SUBSTRATES

FIELD OF THE INVENTION This invention relates to elastase variants and, in particular embodiments, to elastase variants having altered substrate specificity compared with wild-type elastases. Preferred elastase variants cleave substrates comprising histidine residues, preferably substrates comprising a histidine residue two amino acids to the amino-terminal (N-terminal) side of the cleavage site. The invention further relates to compositions, articles of manufacture, and kits comprising the elastase variants, and optionally including substrates selectively cleaved by the elastase variants. The invention also provides nucleic molecules encoding the novel variants and substrates, as well as recombinant materials and methods for producing them. In a particular aspect, the present invention provides methods for cleaving polypeptides at substrate sites, particularly those containing histidine residues.

BACKGROUND OF THE INVENTION Site-specific proteolysis is one of the most common forms of post-translational modifications of proteins (for review see Neurath, H. (1989) Trends Biochem. Sci., 14:268). In addition, proteolysis of fusion proteins in vitro is an important research and commercial tool (for reviews see Uhlen, M. and Moks, T. (1990) Methods

Enzymol., 185:129-143; Carter, P. (1990) in Protein Purification: From Molecular Mechanisms to Large-Scale Processes, M.R. Landisch, R.C. Wilson, CD. Painton, S.E. Builder, Eds. (ACS Symposium Series 427, American Chemical Society, Washington, D.C), Chap. 13, p.l 81-193; and Nilsson, B. et al. (1992) Current Opin. Struct. Biol., 2:569). Site-specific proteolysis is widely used in protein expression and recovery, where a protein of interest is expressed as a fusion protein containing one or more domains that must be removed to obtain the desired protein product. For instance, expressing a protein of interest as a fusion protein facilitates purification when the fusion protein contains an affinity domain such as glutathione-S-transferase, Protein A, or a poly-histidine tail. Liberating the protein product from the fusion protein requires selective and efficient cleavage of the added domain. Both chemical and enzymatic methods have been proposed (see references above). Enzymatic methods are generally preferred as they tend to be more specific and can be performed under mild conditions that avoid denaturation or unwanted chemical side-reactions. A number of natural and designed enzymes have been used for site-specific proteolysis. However, given the sequence requirements of the fusion protein junction and the possible existence of protease sequences within the desired protein product (Forsberg, G., Baastrup, B., Rondahl, H., Holmgren, E., Pohl, G., Hartmanis, M. and Lake, M. (1992) J. Prot. Chem., 11 :201-211), no enzyme is applicable to every situation. Site-specific proteolysis also has therapeutic applications. For example, sequence-specific proteases can be employed for antibody-directed enzyme activated prodrug therapy (ADEPT) (Wolfe, L.A. et al. (1999) Bioconj. Chem. 10:38-48). According to that application, an antibody directs a tethered catalyst to a cell or tissue- type of interest. A prodrug is also administered and remains inactive until acted on by the catalyst. The localization of the antibody-catalyst conjugate results in increased effective concentrations of active drug proximal to the therapeutic target, potentially reducing side-effects and increasing therapeutic benefit. In one application of this technique, sequence-specific proteases linked to antibodies are used with prodrugs incoφorating a selectively cleavable substrate site, such as the peptidyl prodrugs that have been produced from anti-cancer compounds. See, e.g., Chakravarty, P. K.. Carl, P. L., Weber, M. J. and Katzenellenbogen, J. A. (1983) J. Med. Chem. 26, 638-644. An expanded array of sequence-specific proteases, analogous to restriction endonucleases, would make site-specific proteolysis more widely applicable to processing fusion proteins or generating protein/peptide fragments either in vitro or in vivo.

Proteases are classified by a number of criteria including evolutionary relationships, mechanisms of action, and substrate specificities. The serine proteases include at least three structurally defined families (of which subtilisin, chymotrypsin, and carboxypeptidase II are representative members), which share a catalytic mechanism involving nucleophihc attack by a serine residue on a peptidyl substrate. A wide diversity of substrate specificities is attributed primarily to structural differences in the substrate-binding cleft of these proteins. Crystal structures have been determined for a number of chymotrypsin-like proteases, including porcine pancreatic elastase and human neutrophil elastase. Kinetic measurements of substrate preferences for these two mammalian elastases have shed light on structure-function relationships (Perona, J. and Craik, C. (1995) Protein Science, 4:337-360).

Serine proteases play a role in a host of physiological processes, ranging from facilitating digestion to functioning as biological regulators through proteolytic activation of precursor proteins. Elastase is one of several digestive enzymes produced in the pancreas. However, elastase is also found in the azurophilic granules of neutrophils. Human neutrophil elastase has been shown to be released upon neutrophil activation (Dewald, B., Rindler-Ludwig, R., Bretz, U. and Baggiolini, M. (1975) J. Exp. Med., 141 :709-732). Human neutrophil elastase is believed to play a role in normal tissue turnover, clearance of extracellular debris during wound healing, regulation of coagulation and immune responses, and host defenses against bacteria (Janoff, A. and Scherer, J. (1986) J. Exp. Med. 128:1137-1155; Bieth, J.G. (1986) in Regulation of Matrix Accumulation, Mecham, R., ed., 217-320, Academic Press, New York). Elastases have been implicated in several prevalent diseases including emphysema, cystic fibrosis and adult respiratory distress syndrome (Janoff, A. (1985) Ann. Rev. Med. 36:207-216). Because of the link between inappropriate release of elastase and disease, the development and study of elastase inhibitors has been a major focus of research related to this protease family.

Serine proteases have also been the subject of efforts to develop tools for site- specific proteolysis. For example, substrate-assisted catalysis has been demonstrated using mutants of the serine proteases subtilisin BPN' (Carter, P. and Wells, J. A.

(1987) Science 237:394-399; and U.S. Pat. No. 5,652,136) and trypsin (Corey, D. R., Willett, W. S., Coombs, G. S. and Craik, C. S. (1995) Biochemistry 34:11521- 11527). In these studies, an amino acid residue critical to the catalytic machinery was replaced with an amino acid lacking the critical functionality. The mutant proteases selectively cleaved substrates that contained amino acid residues with the functional group that was replaced in the mutants.

For example, the activity of subtilisin BPN' against peptidyl /?-nitroanilide substrates is reduced ~10⁶-fold by replacing the histidine at position 64 (as numbered from the amino-terminus of the protein) with alanine (H64A) (Carter, P. and Wells, J. A. (1987) Science 237:394-399). The function of this missing catalytic group can be partially restored by providing a substrate containing a histidine residue at either the position two amino acids to the amino-terminal (N-terminal) side of the substrate cleavage site (Carter, P. and Wells, J. A. (1987) Science 237:394-399) or of a position immediately to the carboxy-terminal (C-terminal) side of the cleavage site (Carter, P., Nilsson, B., Burnier, J. P., Burdick, D. & Wells, J. A. (1989) Proteins: Struct. Funct. Genet. 6:240-248; Matthews, D. J. & Wells, J. A. (1993) Science 260: 1113-1117). According to the standard nomenclature, Pn...P2-Pl-Pl'-P2'...Pn', where the scissile peptide bond is between the PI and PI' residues (Schechter, I. and Berger, A. (1967) Biochem. Biophys. Res. Commun. 27:157-162), substrates with histidines in either the P2 or PI' positions can be cleaved by H64A subtilisin. Furthermore, the catalytic efficiency (k_cat/K_M) of substrate-assisted catalysis by H64A subtilisin has been increased 150-fold by rational enzyme design and substrate optimization (Carter, P., Nilsson, B., Burnier, J. P., Burdick, D. & Wells, J. A. (1989) Proteins: Struct. Funct. Genet. 6:240-248; Carter, P., Abrahmsen, L. & Wells, J. A. (1991) Biochemistry 30:6142-6148).

Favorable substrates for such subtilisin variants have also been identified using substrate phage (Matthews, D. J. & Wells, J. A. (1993) Science 260:1113- 1117). For example, substrate phage have been engineered to express human growth hormone (HGH) with a randomized linker or substrate region. To identify favorable substrates, the substrate phage were bound to an immobilized hGH receptor. Selection of substrates was carried out by treating the bound phage with the protease of interest. Phage expressing suitable substrates were cleaved from the immobilized support. Sequential rounds of expression, binding, and cleavage, a process termed "panning," allowed for the identification of phage expressing favorable substrates. DNA sequencing of the linker region of the selected phage revealed the amino acid sequences of the favorable substrates.

To be useful in many applications, however, cleavage must be suitably efficient, as well as specific. Engineered protease-substrate pairs often have catalytic efficiencies well below that necessary for biotechnological and pharmaceutical applications. Thus, it would be of substantial interest to the fields of biochemistry, biotechnology and pharmaceuticals to develop proteases that selectively and efficiently cleave specific protein substrates.

SUMMARY OF THE INVENTION

The present invention provides elastase variants with amino acid sequences that differ from those of precursor elastases. According to the present invention, one or more amino acids of the precursor elastase, including at least one residue for conferring substrate specificity as described herein, are replaced with different amino acids. In one embodiment, the amino acid substitution produces an elastase variant having a substrate specificity that is substantially different from the substrate specificity of the precursor elastase. In variations of this embodiment, elastase variants are provided that are highly specific for efficient cleavage of substrates containing histidine residues. Preferred elastase variants are specific for the cleavage of substrates containing a histidine residue at position P2 of the substrate. In one embodiment, an elastase variant has an amino acid substitution at the position corresponding to amino acid residue 43 of elastase produced in human neutrophils (human neutrophil elastase). Substitution with an amino acid residue having a side chain volume smaller than histidine 43 of human neutrophil elastase is generally preferred. In a particularly preferred embodiment of the present invention, the amino acid corresponding to histidine 43 of human neutrophil elastase is substituted with an alanine residue (H43A elastase).

The invention further provides modified elastase variants in which an elastase variant is attached, directly or indirectly, to another molecule, such as a "targeting moiety" or "targeting domain" that directs the elastase variant to one or more specific cell or tissue types. Exemplary modified elastase variants include conjugates containing an elastase variant attached to an antibody or antibody fragment or other targeting moiety. Elastase variant conjugates comprising an elastase variant attached to a targeting moiety can be used, for example, to direct the localization of the elastase variant for therapeutic applications.

The invention also provides a substrate for the elastase variants of the invention. Preferred substrates have a substrate site comprising at least two amino acids and have a histidine residue at substrate position P2. Preferred substrates can also contain either arginine or methionine at position P4, glutamine or glutamic acid at position P3, and valine or threonine at position PI . At substrate position PI', preferred substrates contain either isoleucine, valine, threonine, or methionine. At position P2', a variety of hydrophobic amino acid residues are suitable, including isoleucine, phenylalanine, leucine, tyrosine, and tryptophan.

In one embodiment, the substrate is a polypeptide comprising a substrate site as described above. In a variation of this embodiment, the polypeptide is a fusion protein, and the substrate site separates two domains of the fusion protein. In a preferred fusion protein, the substrate site links a polypeptide of interest to an affinity domain. Such a fusion protein can be purified via the affinity domain, and a suitable elastase variant can be used to cleave the polypeptide of interest from the affinity domain.

In another variation, the preferred substrate is a peptidyl prodrug, and cleavage at the substrate site produces an active drug.

The invention includes compositions, such as pharmaceutical compositions, comprising an elastase variant and optionally a substrate according to the invention. The invention also provides articles of manufacture comprising such compositions and related kits.

Other aspects of the invention include nucleic acid molecules encoding the elastase variants and substrates of the invention. The invention includes vectors, preferably expression vectors, comprising the nucleic acid molecules; host cells containing such vectors; and methods for producing a variant elastase or a substrate according to the invention.

The invention further includes methods for cleaving a substrate in which an elastase variant of the invention is contacted with a substrate under conditions such that the elastase variant cleaves the substrate.

Another aspect of the invention is a method of producing an elastase or elastase variant that does not require zymogen activation, along with a nucleic acid molecule, vector, and host cell suitable for use in this production method. The nucleic acid molecule encodes an elastase lacking a prosequence.

BRIEF DESCRIPTION OF THE FIGURES FIGURE 1 shows a molecular model of the predicted interactions between elastase and a substrate containing histi dines at P2 (HisP2) (model 1 in Table 1) and PI' (HisPT) (model 2 in Table 1) positions. The substrate is shown together with the main chain of elastase and the catalytic triad, H43, D90, and SI 75. Molecular modeling was based upon the X-ray crystallographic structure of elastase complexed with the turkey ovomucoid inhibitor (Bode, W., Wei, A.Z., Huber, E.M., Meyer, E., Travis, J., and Neumann, S. (1986) EMBO J., 5:2453-2458):

FIGURE 2 (A) Schematic representation and (B) nucleotide sequence of a synthetic gene for the expression of wild-type elastase in P. pastoris highlighting the catalytic triad residues and the amino terminal residue, 13. The elastase gene is preceded and followed by sequences encoding the Matα peptide and hexahistidine, respectively. The 5 ' and 3 ^' ends of synthetic DNA fragments used in assembling the synthetic gene are indicated by lowercase, as are amino acid residues in the Matα peptide.

FIGURE 3 shows linker sequences of substrate phage clones identified after 7 rounds of panning that are (A) sensitive or (B) resistant to cleavage by H43A elastase. Also shown are summaries of (Q H43A sensitive and (D) H43A resistant linkers. FIGURE 4 shows cleavage of fusion proteins composed of the synthetic Z domain of S. aureus protein A joined by a linker sequence to E. coli alkaline phosphatase (Z-AP fusion proteins) by H43A and wild-type elastase. Z-AP fusion proteins (4 μM) with histidine-containing (LI) or non-histidine containing (LI 4) linkers were digested with 50 nM H43A or 0.1 nM wild-type elastase in the absence (-) or presence (+) of 1 mM phenymethylsulfonyl fluoride (PMSF) (24 h, 37 °C). The bands with apparent molecular weight of 54 kDa, 47 kDa, 31 kDa and 28 kDa represent Z-AP fusion protein, AP, H43A elastase and a minor contaminant in the fusion protein, respectively.

DETAILED DESCRIPTION OF THE INVENTION Definitions

Terms used in the claims and specification are defined as set forth below unless otherwise specified.

The term "amino acid" or "amino acid residue," as used herein, includes naturally occurring L-amino acids or residues, unless otherwise specifically indicated. The commonly used one- and three-letter abbreviations for amino acids are used herein (Lehninger, A. L. (1975) Biochemistry, 2d ed., pp. 71-92, Worth Publishers, N. Y.). The term also includes D-amino acids as well as chemically modified amino acids, such as amino acid analogs, naturally occurring amino acids that are not usually incorporated into proteins, and chemically synthesized compounds having the characteristic properties of amino acids (collectively, "atypical" amino acids). For example, analogs or mimetics of phenylalanine or proline, which allow the same conformational restriction of the peptide compounds as natural Phe or Pro are included within the definition of "amino acid."

Exemplary atypical amino acids, include, for example, those described in International Publication No. WO 90/01940 as well as 2-amino adipic acid (Aad) which can be substituted for Glu and Asp; 2-aminopimelic acid (Apm), for Glu and Asp; 2-aminobutyric acid (Abu), for Met, Leu, and other aliphatic amino acids; 2- aminoheptanoic acid (Ahe), for Met, Leu, and other aliphatic amino acids; 2- aminoisobutyric acid (Aib), for Gly; cyclohexylalanine (Cha), for Val, Leu, and He; homoarginine (Har), for Arg and Lys; 2, 3-diaminopropionic acid (Dpr), for Lys, Arg, and His; N-ethylglycine (EtGly) for Gly, Pro, and Ala; N-ethylasparagine (EtAsn), for Asn and Gin; hydroxyllysine (Hyl), for Lys; allohydroxyllysine (Ahyl), for Lys; 3- (and 4-) hydoxyproline (3Hyp, 4Hyp), for Pro, Ser, and Thr; allo-isoleucine (Aile), for He, Leu, and Val; P-amidinophenylalanine, for Ala; N-methylglycine (MeGly, sarcosine), for Gly, Pro, and Ala; N-methylisoleucine (Melle), for He; norvaline (Nva), for Met and other aliphatic amino acids; norleucine (Nle), for Met and other aliphatic amino acids; ornithine (Orn), for Lys, Arg, and His; citrulline (Cit) and methionine sulfoxide (MSO) for Thr, Asn, and Gin; N-methylphenylalanine (MePhe), trimethylphenylalanine, halo (F, CI, Br, and I) phenylalanine, and trifluorylphenylalanine, for Phe.

"Elastase," "precursor elastase" and the like refer to a carbonyl hydrolase that generally acts to cleave peptide bonds adjacent to small hydrophobic amino acids. Elastases generally have molecular weights of about 22 to about 33 kD. As defined herein, elastases are serine endoproteases characterized by a common catalytic triad that distinguishes them from, for example, the subtilisin-related proteases. Although both types of enzymes have a catalytic triad including aspartate, histidine, and serine, the order of these residues (reading from the N-terminus) is His-Asp-Ser for elastases and Asp-His-Ser for subtilisin-related proteases. Thus, as used herein, the terms "elastase" and "precursor elastase" refer to proteases that have the elastase-type catalytic triad. "Precursor elastase" refers to an elastase, the sequence of which, when mutated according to the invention, gives rise to the elastase variant of the invention. "Elastase variant" and the like refer to an elastase-type serine protease having an amino acid sequence that differs from that of a precursor elastase by at least an amino acid substitution as described herein for conferring substrate specificity. Preferably, at least one such amino acid substitution significantly alters substrate specificity.

"Prosequence" refers to a sequence of amino acids covalently linked to the N- terminal portion of the mature form of an elastase. The prosequence is proteolytically removed to produce the "mature" form of the elastase. Many proteolytic enzymes are found in nature as translational proenzyme products and, in the absence of post- translational processing, are expressed in this fashion.

A "signal sequence" or "presequence" refers to any sequence of amino acids covalently linked to the N-terminal portion of the mature form of an elastase, or to the N-terminal portion of a prosequence, that participates in the secretion of the mature or pro forms of the elastase. This definition of signal sequence is a functional one, meant to include all amino acid sequences, encoded by the N-terminal portion of the elastase gene or other secretable proteins, that facilitate the secretion of elastase or other carbonyl hydrolases under native conditions. A "pre" form of an elastase or elastase variant consists of an elastase or elastase variant having a "pre" or "signal" sequence operably linked its N-terminus. A "pro" form of an elastase or elastase variant consists of an elastase or elastase variant having a prosequence operably linked to the N-terminus of the elastase or the elastase variant. A "prepro" form of an elastase or elastase variant consists of an elastase or elastase variant having a prosequence operably linked to the N-terminus of the elastase or elastase variant and a "pre" or "signal" sequence operably linked to the

N-terminus of the prosequence.

An elastase or elastase variant that does not require zymogen activation is one that possesses carbonyl hydrolase activity, i.e., one that does not require cleavage of a prosequence for activity. An exemplary elastase that does not require zymogen activation is the mature form.

"Operably linked," when describing the relationship between two protein or

DNA regions or domains, simply means that the regions or domains are functionally related to each other. For example, a presequence is operably linked to a protein if it functions as a signal sequence, participating in the secretion of the protein, generally involving the cleavage of the presequence. Operably linked protein domains are often contiguous, but this is not a requirement. Examples of operably linked DNA regions include a promotor or ribosome binding site and a coding sequence. The promoter is operably linked to the coding sequence if it controls the transcription of the sequence; a ribosome binding site is operably linked to the coding sequence if it is positioned so as to permit translation. As with protein regions or domains, operably linked DNA regions need not be contiguous. For example, enhancers need not be contiguous with a coding sequence to enhance transcription of the coding sequence. Residue positions in elastase and elastase variants are designated herein by the three-letter or one-letter code for the amino acid, followed by the position number, as numbered from the N-terminus of the pro form of native human neutrophil elastase, residue Ser 1 (this corresponds to position 28 from the initiating methionine in the prepro form). When referring to elastase variants, the precursor amino acid residue is followed by the residue number and the new or substituted amino acid residue. Thus, a substitution of alanine at His43 of elastase is expressed as "His43Ala" or "H43A."

Residues in two or more polypeptides are said to "correspond" if they are either homologous (i.e., occupying similar positions in either primary, secondary, or tertiary structure) or analogous (i.e., having the same or similar functional capacities). As is well known in the art, homologous residues can be determined by aligning the polypeptide sequences based on amino acid sequence and/or structure. Those skilled in the art understand that it may be necessary to introduce gaps in either sequence to produce a satisfactory alignment.

Alignments based on amino acid sequence are carried out by aligning those residues known to be invariant in all elastases for which sequences are known, including the catalytic triad, which is His43, Asp90, and Serl75 in human neutrophil elastase. Alignment of invariant residues is followed by alignment of residues known to be conserved in elastases for which sequences are known. Residues are said to be "conserved" if the same residue appears in a majority of wild-type elastases and/or if the majority of the differences between the residues in different elastases represent conservative amino acid substitutions, as defined below. A satisfactory alignment generally aligns at least about 50%, preferably about 75%, and more preferably about 100%), of conserved residues.

The term "conservative amino acid substitution" is used herein to refer to the replacement of an amino acid with a functionally equivalent amino acid. Functionally equivalent amino acids are generally similar in size and/or character (e.g., charge or hydrophobicity) to the amino acids they replace. Amino acids of similar character can be grouped as follows:

(1) hydrophobic: His, Tip, Tyr, Phe, Met, Leu, He, Val, Ala;

(2) neutral hydrophobic: Cys, Ser, Thr;

(3) polar: Ser, Thr, Asn, Gin;

(4) acidic/negatively charged: Asp, Glu;

(5) charged: Asp, Glu, Arg, Lys, His;

(6) basic/positively charged: Arg, Lys, His;

(7) basic: Asn, Gin, His, Lys, Arg;

(8) residues that influence chain orientation: Gly, Pro; and

(9) aromatic: Trp, Tyr, Phe, His.

The following table shows exemplary and preferred conservative amino acid substitutions.

Preferred Conservative

Original Residue Exemplary Conservative Substitution Substitution

Ala Val, Leu, He Val

Arg Lys, Gin, Asn Lys

Asn Gin, His, Lys, Arg Gin

Asp Glu Glu

Cys Ser Ser

Gin Asn Asn

Glu Asp Asp

Gly Pro Pro

His Asn, Gin, Lys, Arg Asn

He Leu, Val, Met, Ala, Phe Leu

Leu He, Val, Met, Ala, Phe He

Lys Arg, Gin, Asn Arg

Met Leu, Phe, He Leu Preferred Conservative

Original Residue Exemplary Conservative Substitution Substitution

Phe Leu, Val, He, Ala Leu

Pro Gly Gly

Ser Thr Thr

Thr Ser Ser

Trp Tyr Tyr

Tyr Trp, Phe, Thr, Ser Phe

Val He, Leu, Met, Phe, Ala Leu

Alignments of polypeptides based on structure can be carried out for elastases whose structure has been determined by x-ray crystallography. Alignment is achieved after the best model has been oriented and positioned to give the maximum overlap of atomic coordinates of non-hydrogen atoms of the two polypeptides. The best model is the crystallographic model giving the lowest R factor for experimental diffraction data at the highest resolution available.

Σ \ Fo (h) \ - \ Fc (h) R factor = h

∑ \ Fo (h) \ h

Corresponding residues in polypeptides aligned in this manner are those for which the atomic coordinates of two or more of the main chain atoms of a particular amino acid residue of one polypeptide are within 0.13 nm and preferably 0.1 nm of the main chain atoms of the other amino acid residue (N on N, CA on CA, C on C and O on O). Alignments can be carried out on the basis of structure and sequence, as described, for example in Greer, J. (1990) Proteins: Struct. Funct. Genet. 7:317-334, which shows the alignment of porcine elastase with 34 other serine proteases.

Corresponding residues that are functionally analogous are those that influence protein structure or function (e.g., substrate binding or catalysis) in a similar manner.

"Percent sequence identity" is used herein to describe the relatedness of two nucleic acid or amino acid sequences. This value is calculated by determining the number of corresponding residues that are identical between two aligned sequences and dividing by the total number of residues in the longest sequence. Substrate sites are numbered as Pn...P2-Pl-Pl'-P2'...Pn'. The "PI" residue refers to the position preceding (N-terminal to) the scissile peptide bond of the substrate as defined by Schechter and Berger (Schechter, I. and Berger, A., (1967) Biochem. Biophys. Res. Commun. 27:157-162). Similarly, the term "P"' is used to refer to the position following (C-terminal to) the scissile peptide bond of the substrate. Increasing numbers refer to the next consecutive position preceding (e.g., P2 and P3) and following (e.g., P2' and P3') the scissile bond. As used with reference to the substrates of the invention, the scissile peptide bond is the bond that is cleaved by the elastase variants of the invention. Substrate sites can also be numbered without reference to the scissile bond as X_rX₂-X₃-X₄-X₅ . . . . X_n9 where X, is the N-terminal residue in the substrate site.

A change in substrate specificity is defined as a difference between the k_cat/K_M ratio of the precursor elastase and the elastase variant. The k_cat/K_M ratio is a measure of catalytic efficiency. Elastase variants with increased or decreased k_cat/K_M ratios compared to the precursor elastase from which they were derived are described herein. A greater (i.e. numerically larger) k_cat/K_M ratio for a particular substrate indicates that the variant cleaves the target substrate more efficiently. Therefore, a variant having a greater k_cat/K_M ratio for a given substrate is preferred. A change in substrate specificity is said to be significant if the precursor and variant k_cat/K_M ratios for a particular substrate differ by at least two-fold. An enzyme's specificity or discrimination between two or more competing substrates is determined by the ratios of k_cat/K_M for the two substrates (Fersht, A.R. (1985) in Enzyme Structure and Mechanism, W.F. Freeman and Co., N.Y. p. 112). A high degree of specificity is useful for the hydrolysis of a particular substrate in a mixture of substrates, limiting undesired hydrolysis of the non-specific substrates. Mutating an enzyme can produce a variant with distinct substrate specificity. In producing an enzyme variant, an increase in k_cat/K_M ratio for one substrate can be accompanied by a reduction in k_cat/K_M ratio for another substrate. This shift in substrate specificity indicates that the variant with the increased k_cat/K ratio for the particular substrate has utility in cleaving the particular substrate over the precursor enzyme.

The term "purified," as used herein, refers to a composition that has been separated from at least one component normally found with the composition in its natural source. This term applies to such a composition, regardless of whether the composition is purified to homogeneity or present in a heterogeneous mixture and regardless of whether the composition is subsequently joined to or mixed with other components.

The term "nucleic acid molecule" encompasses single-stranded and double-stranded DNA molecules, including genomic DNA, cDNA, DNA produced by an amplification reaction (such as polymerase chain reaction ["PCR"]), and DNA produced by oligonucleotide synthesis and or ligation of smaller fragments, as well as RNA molecules, such as mRNA. Genomic DNA can include non-transcribed and transcribed regions (such as 5 ' and 3 ' non-coding regions, introns, and coding regions). cDNA and mRNA molecules contain sequences corresponding to transcribed regions.

"Vector" refers to a DNA construct containing a nucleic acid molecule of interest. Such a vector can be propagated stably or transiently in a host cell. The vector may be a plasmid, a phage particle, or simply a potential genomic insert. Once transformed into a suitable host, the vector may replicate and function independently of the host genome, or may, in some instances, integrate into the host genome. The terms "plasmid" and "vector" are sometimes used interchangeably herein as the plasmid is the most commonly used form of vector at present. However, the invention is intended to include other forms of vectors that serve equivalent functions and that are, or become, known in the art. "Expression vector" refers to a DNA construct containing a nucleic acid molecule that is operably linked to a suitable control sequence capable of effecting the expression of the nucleic acid in a suitable host. Exemplary control sequences include a promoter to effect transcription, an optional operator sequence to control transcription, a sequence encoding suitable mRNA ribosome binding sites, and sequences that control termination of transcription and translation.

"Host cell" refers to a cell capable of maintaining a vector either transiently or stably. Host cells of the invention include, but are not limited to, bacterial cells, yeast cells, insect cells, plant cells and mammalian cells. Other host cells known in the art, or which become known, are also suitable for the invention. The terms "targeting moiety" or "targeting domain" are used herein to describe a chemical moiety or protein domain, respectively, that is capable of directing an attached molecule, such as an "elastase" variant of the invention, to a specific cell or tissue type. The targeting moiety or domain is a member of a pair of binding partners. The targeting moiety or domain binds its cognate binding partner with sufficient affinity and specificity to allow preferential binding to the binding partner in a complex mixture of substances. Preferably, the association constant (Ka) for the binding of the targeting moiety or domain to its cognate binding partner is at least about 10⁵ M^"1. Exemplary targeting moieties that provide sufficiently high affinity binding include antibodies and fragments thereof (e.g., Fab fragments). "Prodrug" refers to a precursor or derivative form of a drug that is less active than the drug itself and is capable of being enzymatically activated or converted to the active drug. Examples of cytotoxic drugs that can be derivatized to produce a prodrug for use in this invention include, but are not limited to, chemotherapeutic agents and cytotoxic polypeptides. See, e.g., Wilman (1986) Biochemical Society Transactions 14:375-382, 615 Meeting Belfast; Stella, et al (1985) Directed Drug Delivery, Borchardt (ed.) pp. 247-267, Humana Press. Exemplary prodrugs include phosphate-containing prodrugs, thiophosphate-containing prodrugs, glycolsylated prodrugs or optionally substituted phenylacetamide-containing prodrugs.

As used to describe a prodrug, the term "active" refers to biological and/or immunological activity.

The term "chemotherapeutic agent" is defined herein as any chemical compound useful in the treatment of cancer. The term "cancer" refers to the physiological condition in mammals that is characterized by unregulated cell growth. Examples of cancer include, but are not limited to, carcinoma, lymphoma, blastoma, sarcoma, and leukemia. More particular examples of such cancers include squamous cell cancer, small-cell lung cancer, non-small-cell lung cancer, gastric cancer, pancreatic cancer, glial cell tumors such as glioblastoma and neurofibromatosis, cervical cancer, ovarian cancer, liver cancer, bladder cancer, hepatoma, breast cancer, colon cancer, colorectal cancer, endometrial carcinoma, salivary gland carcinoma, kidney cancer, renal cancer, prostate cancer, vulval cancer, thyroid cancer, hepatic carcinoma, and various types of head and neck cancer. Examples of chemotherapeutic agents include Adriamycin, Doxorubicin, 5-Fluorouracil (5-FU), Cytosine arabinoside (Ara-C), Cyclophosphamide, Thiotepa, Busulfan, Cytoxin, Taxol, Methotrexate, Cisplatin, Melphalan and any related nitrogen mustard, Ninblastine, Bleomycin, Etoposide, Ifosfamide, Mitomycin C, Mitoxantrone, Vincristine, VP-16,

Vinorelbine, Carboplatin, Teniposide, Daunomycin, Carminomycin, Aminopterin, Dactonomycin, a Mitomycin, Νicotinamide, an Esperamicin, and an endocrine therapeutic (such as diethylstilbestrol [DES], Tamoxifen, a luteinizing hormone releasing hormone-antagonizing drug, an anti-progestin, etc.). The term "cytotoxic polypeptide" refers to a polypeptide that inhibits a cellular function or kills cells. Cytotoxic polypeptides suitable for use as cytotoxic drugs include, but are not limited to, toxins of bacterial, fungal, plant, or animal origin and fragments thereof and an oncogene product/tyrosine kinase inhibitor, such as a peptide that inhibits binding of a tyrosine kinase to a SH2-containing substrate protein (see WO 94/07913, for example). Elastase Variants The present invention provides elastase variants that have at least one amino acid substitution as described herein for conferring substrate specificity. In preferred embodiments, the elastase variants have significantly altered substrate specificities, as compared to their precursor elastases. The elastase variants of the invention may include other amino acid substitutions so long as they retain the substrate specificity of the elastase variants described herein. Preferred elastase variants are highly specific for efficient cleavage of substrates containing histidine residues. In one embodiment, the elastase variants are specific for the cleavage of substrates containing a histidine residue at position P2 of the substrate, and in particular, the substrates exemplified herein.

In one embodiment, an elastase variant has an amino acid substitution at the position corresponding to amino acid residue 43 of human neutrophil elastase. Identification of this position in a given precursor elastase is carried out by determining a suitable alignment of the precursor elastase sequence with that of human neutrophil elastase. Substitution with an amino acid residue having a side chain volume smaller than that of histidine is generally preferred. In a particularly preferred embodiment of the present invention, the elastase variant is substituted with an alanine residue. Where the precursor elastase is human neutrophil elastase, this variant is termed "H43A elastase."

Although the present invention is exemplified with variants of human neutrophil elastase (see Examples 1-6), the invention can be carried out using any elastase as a precursor elastase. The precursor elastase can be from any species, but is preferably a mammalian elastase, and is most preferably a human elastase.

Furthermore, the precursor elastase can be one of a number of naturally occurring isoforms or allelic variants. The precursor elastase can be obtained from any tissue or cellular source, including, for example, pancreas and monocytes, in addition to, neutrophils. In a preferred embodiment, the precursor elastase is human neutrophil elastase. In addition to naturally occurring elastases, recombinant elastases or synthetic elastases can serve as the precursor elastase of the invention. For example, a gene encoding human neutrophil elastase can be assembled as described in Dennis, M.S. Carter, P. and Lazarus, R.A. (1993) Proteins: Struct. Funct. Genet. 15:312-321, and the assembled gene can be expressed to produce a recombinant elastase (see also Fig. 2). The amino acid sequence of recombinant or synthetic precursor elastases can differ from wild-type elastase sequences (i.e., can have amino acid substitutions, insertions, and deletions). However, such precursor elastases typically have the wild- type catalytic triad together, with at least about 30%, preferably about 50%>, more preferably about 70%, and most preferably about 90%>, sequence identity with human neutrophil elastase (as shown in Fig. 2) when these polypeptides are aligned on the basis of amino acid sequence and/or protein structure.

As discussed above, the invention encompasses elastase variants with as little as one amino acid substitution conferring substrate specificity as described herein, compared to the precursor elastase. However, those skilled in the art can readily design elastase variants having additional amino acid sequence variations. More specifically, the elastase variants of the invention include those having amino acid substitutions, insertions, or deletions (including N- and/or C-terminal truncations, i.e., elastase variant fragments), especially outside the critical domains that significantly influence overall protein structure, substrate specificity, or catalytic activity. Such domains are known for a number of elastases and can be predicted for others based on comparative modeling. See, e.g., Greer, J. (1990) Proteins: Struct. Funct. Genet. 7:317-334. Functional amino acid residues within such domains (or, in fact, anywhere in the proteins) can be determined by standard alanine scanning mutagenesis, as described by Cunningham, B., and J.A. Wells (1989) Science

244:1081-1085. In this technique, a series of single amino acid-substituted variants can be constructed in which each residue in a putative functional domain is replaced with a neutral amino acid, such as alanine. Testing of the series of variants in a functional assay, such as the cleavage assays described in the examples, allows identification of important functional residues. Such residues are generally not deleted in elastase variants of the invention, unless the function of such residues is deemed unnecessary for the desired variant. However, important functional residues may be replaced with other amino acids. When it is desirable to preserve the function of the original amino acid, conservative amino acid substitutions are generally preferred. Amino acid substitutions that preserve function can also be determined empirically using the substrate phage technique (Matthews, D. J. & Wells, J. A. (1993) Science 260:1113-1117), which allows rapid screening of variants having any amino acid residue at a given position. This technique can also be used to produce elastase variants having amino acid sequence changes (beyond those exemplified herein) that alter substrate specificity and/or cleavage efficiency. The substrate phage technique is described in detail in Examples 3-4.

Elastase variants according to the present invention typically have at least about 70%, preferably about 80%, more about preferably 90%, and most preferably about 95%, amino acid sequence identity with the precursor elastase. The invention includes mature forms of elastase variants, as well as pre-, pro- and prepro-forms of such variants. These variants can be substantially full-length, which, as used herein, means that the elastase variant is at least about 95% as long as a corresponding wild- type elastase. As discussed above, however, the invention also includes elastase variants having internal deletions and/or that are truncated at the N- and or C- terminus. Such elastase variants are preferably at least about 30%, 50%, 70%>, or 90% as long as a corresponding wild-type elastase. Pre-, pro-, and prepro- sequences can, if desired, be attached to elastase variants having internal deletions or terminal truncations. In one embodiment, the inclusion of pre- and prepro-sequences in elastase variants is preferred since this facilitates the expression and secretion of the elastase variants.

An elastase variant according to the invention can be attached, directly or indirectly, to one or more other molecules or chemical groups to form a modified elastase variant. In a preferred embodiment, an elastase variant is attached to a targeting moiety or domain that directs the elastase variant to one or more specific cell or tissue types. Such targeted variants can be used, e.g., to direct the localization of the variant for therapeutic applications. Exemplary modified elastase variants according to this embodiment include a conjugate containing an elastase variant attached to an antibody or antibody fragment specific for a cell surface receptor or antigen or a fusion protein including an elastase variant attached to a targeting domain, such as a receptor binding domain.

Elastase variants can be attached to other by molecules any of a variety of means familiar to those of skill in the art. Covalent attachment is typically the most convenient, but other forms of attachment can be employed, depending on the application. Examples of suitable forms of covalent attachment include the bonds resulting from the reaction of molecules bearing activated chemical groups with amino acid side-chains as well as the peptide bonds formed during translation of mRNA. Peptide bonds are conveniently employed, e.g., when the molecule to be attached to the elastase variant is a polypeptide. In this case, the elastase variant and attached polypeptide can be expressed as a fusion protein.

The attachment can be direct or the elastase variant can be separated from the attached molecule by a linker. The length of the linker can vary, but linkers that provide spatial separation between the functional domains of the elastase variant and the attached molecule are preferred. Typically, the linker is long enough to ensure that the elastase variant can bind to and cleave the intended substrate. If the attached molecule is a targeting moiety or domain, the linker is generally long enough to ensure that the targeting moiety or domain can bind its cognate binding partner. Preferably, the linker spans a distance of at least about 30 angstroms and more preferably at least about 60 angstroms. Those skilled in the art recognize that various combinations of atoms provide for variable length molecules based upon known bond lengths (see, e.g., Morrison and Boyd (1977) Organic Chemistry, 3rd ed., Allyn and Bacon, Inc.). Polypeptide linkers of about 5 to about 35 amino acids are preferred; those of between about 5 to about 20 amino acids are more preferred; and those of about 5 to about 10 are most preferred. Generally, shorter polypeptide linkers are preferable to longer polypeptide linkers, provided the elastase variant is active. In one embodiment, the linker molecule includes a flexible peptide chain. Suitable flexible peptide linkers are known or can be readily determined by those familiar with protein structure. Examples include Gly-Pro-Gly-Gly, Gly-Gly, or Pro- Gly. Elastase variants and modified elastase variants retain the ability to bind and cleave a substrate of the invention, preferably with an affinity that allows the variants to compete for binding with wild-type elastase at physiological concentrations. In addition, such variants retain the ability to cleave a suitable substrate. Thus, amino acid sequence variations or other modifications that significantly impair these functions are avoided. In light of the teachings herein, those skilled in the art can readily design a large number of variants that preserve the ability to bind and cleave a suitable substrate. The activity of such variants can be confirmed by a simple cleavage assay, such as those described in Examples 4-6. Generally, the k_cat/K_M ratio (catalytic efficiency) of an elastase variant or modified

2 1 1 elastase variant according to the invention is between about 1 x 10 M^" s^" to about 1 x 10⁸ M^'V¹. More typically the k_cat/K_M ratio is between about 1 x 10 M^" s^" and about

7 1 1

1 x 10 M^" sA Preferred elastase variants and modified elastase variants have improved catalytic efficiency, as compared to their respective precursor elastases. For ease of description, most of the remaining aspects of the invention are described in terms of elastase variants. However, those of skill in the art will appreciate that the invention encompasses the same aspects related to modified elastase variants.

Elastase variants according to the invention can be synthesized using standard techniques, but are generally more conveniently produced using recombinant techniques, as described in the examples below. Precursor elastase genes or gene sequences can cloned, for instance, based on homology to known elastases, such as human neutrophil elastase. With a precursor elastase gene in hand, a nucleic acid molecule encoding the elastase variant can be generated by any of a variety of mutagenesis techniques. See, e.g., Sambrook, J., Fritsch, E. F., and Maniatis, T. (1989) in Molecular Cloning: A Laboratory Manual, Second Edition, Cold Spring Harbor, N.Y. Examples include site-specific mutagenesis (Kunkel et al., (1991) Methods Enzymol., 204:125-139; Carter, P., et al, (1986) Nucl. Acids Res. 10:6487), cassette mutagenesis (Wells, J.A., et al., (1985) Gene 34:315), and restriction selection mutagenesis (Wells, J.A., et al., (1986) Philos. Trans. R. Soc, London Ser. A, 317:415). In a preferred embodiment of the invention, the sequence of an elastase coding region is used as a guide to design a synthetic nucleic acid molecule encoding the elastase variant that can be incoφorated into a vector of the present invention. Methods for constructing synthetic genes are well-known to those of skill in the art. See, e.g., Dennis, M. S., Carter, P. and Lazarus, R. A. (1993) Proteins: Struct. Funct. Genet., 15:312-321.

Substrates The invention also provides a substrate for an elastase variant according to the invention. The substrate includes a substrate site having at least two amino acids and a histidine residue at position P2. Preferred substrates also contain either arginine or methionine at position P4, glutamine or glutamic acid at position P3, and/or valine or threonine at position PI . Preferences for substrate amino acid residues at the PI' and P2' positions are less stringent. At position PI', preferred substrates contain either isoleucine, valine, threonine, or methionine. At position P2', a variety of hydrophobic amino acid residues are suitable, including isoleucine, phenylalanine, leucine, tyrosine, and tryptophan. Thus, an exemplary substrate contains Met-Glu-His-Val- Val-Tyr (SEQ ID NO. _) at positions P4-P3-P2-P1-P1'-P2' respectively. Exemplary two-amino acid substrate sites include His- Val and His-Thr.

The substrate can be naturally occurring, recombinant, or synthetic. Naturally occurring substrates can be purified using suitable conventional purification methods. Recombinant or synthetic substrates can be produced as described above for elastase variants.

The substrate can be a relatively short peptide or a longer polypeptide. Generally, the substrate site is positioned between two molecules, domains, or chemical groups to be separated by cleavage (hereafter "molecules to be separated"). The substrate site can be formed by attaching the two molecules to be separated directly or installing a histidine at the junction. Alternatively, a peptide including the substrate site can be attached to the molecules to be separated. The two molecules to be separated can be attached to one another, to a histidine residue, or to a peptide including the substrate site by any suitable means, such as those discussed above in connection with modified elastase variants. Covalent attachment is generally preferred. The attachment can be direct or indirect. When the molecules to be separated are large relative to the substrate site (or adopt a conformation that reduces access to substrate site), a linker can be employed to ensure that the substrate site is available for cleavage. The linker can be positioned on either side, or both sides, of the substrate site, in between the substrate site and the molecule(s) to be separated. The linker preferably includes a flexible peptide chain. The considerations for selecting a suitable linker for use in a substrate are as described above for modified elastase variants. In a preferred embodiment, the substrate site is incoφorated into a fusion protein. An exemplary fusion protein includes an affinity domain that aids in protein purification, a substrate site, a polypeptide of interest, and an optional linker sequence (or sequences) adjacent to the substrate site. The affinity domain is typically positioned at the N-terminus with the polypeptide of interest at the C-terminus. In a variation of this embodiment, a flexible peptide linker is positioned between the substrate site and the affinity domain to provide better separation of the two regions.

Examples of affinity domains suitable for use in fusion proteins include glutathione-S-transferase, which binds glutathione; protein A (or derivative or fragments thereof), which binds IgG molecules; polyhistidine sequences, particularly hexahistidine sequences that bind metal affinity columns; maltose binding protein, which binds maltose, human growth hormone, which binds the human growth hormone receptor or any of a variety of other proteins or protein domains that can bind to an affinity support with an association constant (Ka) of > 10⁵ M^"1.

The N-terminal residues of the polypeptide of interest may alter the efficiency of substrate linker cleavage (Bauer, C. A., Brayer, G. D., Sielecki, A. R. and James, M. N. (1981) Eur. J. Biochem., 120:289-294; Bizzozero, S. A. and Dutler, H. (1987) Arch. Biochem. Biophys., 256:662-676). To increase cleavage efficiency, position PI' preferably contains a non-polar amino acid, and P2' preferably contains a hydrophobic amino acid. In some cases, it is preferable to add residues to the N-terminus of the polypeptide of interest or to mutate the N-terminal amino acids of this polypeptide to produce an optimal substrate for an elastase variant according to the invention.

In another embodiment, the substrate is a prodrug including a substrate site selectively cleavable by an elastase variant of the invention, wherein cleavage produces an active drug. A prodrug of the invention includes an inhibitory molecule, chemical group, or domain attached, directly or indirectly, to a peptide including the substrate site, which is itself attached, directly or indirectly, to the drug. The drug is preferably a chemotherapeutic agent or cytotoxic polypeptide as defined above. For polypeptide drugs, the prodrug is typically a fusion protein. The design of prodrugs useful in therapeutic applications, such as, for example, the treatment of cancer, is within the level of skill in the art in light of the teachings herein.

Compositions, Articles of Manufacture, and Kits Including Elastase Variants or

Substrates The invention provides compositions, including pharmaceutical compositions, comprising elastase variants and/or substrates of the invention. The compositions optionally include other components, as for example, a storage solution, such as a suitable buffer, e.g., a physiological buffer. In a preferred embodiment the composition is a pharmaceutical composition and the other component is a physiologically acceptable carrier, excipient, or stabilizer, such as are described in Remington's Pharmaceutical Sciences (1980) 16th editions, Osol, ed., 1980.

A physiologically acceptable carrier, excipient, or stabilizer suitable for use in the invention is non-toxic to recipients at the dosages employed, and can include a buffer (such as a phosphate buffer, citrate buffer, and buffers made from other organic acids), an antioxidant (e.g., ascorbic acid), low-molecular weight (less than about 10 residues) polypeptide, a protein (such as serum albumin, gelatin, and an immunoglobulin), a hydrophilic polymer (such as polyvinylpyrrolidone), an amino acid (such as glycine, glutamine, asparagine, arginine, and lysine), a monosaccharide, a disaccharide, and other carbohydrates (including glucose, mannose, and dextrins), a chelating agent (e.g., ethylenediaminetetratacetic acid [EDTA]),a sugar alcohol (such as mannitol and sorbitol), a salt-forming counterion (e.g., sodium), and/or an anionic surfactant (such as Tween , Pluronics , and PEG). In one embodiment, the physiologically acceptable carrier is an aqueous pH-buffered solution.

Preferred embodiments include sustained-release pharmaceutical compositions. An exemplary sustained-release composition has a semipermeable matrix of a solid hydrophobic polymer to which the elastase variant or substrate is attached or in which the elastase variant or substrate is encapsulated. Examples of suitable polymers include a polyester, a hydrogel, a polylactide, a copolymer of L- glutamic acid and T-ethyl-L-glutamase, non-degradable ethylene-vinylacetate, a degradable lactic acid-glycolic acid copolymer, and poly-D-(-)-3-hydroxybutyric acid. Such matrices are in the form of shaped articles, such as films, or microcapsules.

Exemplary sustained release compositions include elastase variants attached, typically via ε-amino groups, to a polyalkylene glycol (e.g., polyethylene glycol [PEG]). Attachment of PEG to proteins is a well-known means of reducing immunogenicity and extending in vivo half-life (see, e.g., Abuchowski, J., et al.

(1977) J. Biol. Chem. 252:3582-86) and is thus particularly useful for elastase variants intended for therapeutic uses, such as prodrug therapy. Any conventional "pegylation" method can be employed, provided the "pegylated" variant is capable of binding and cleaving substrate. In another embodiment, a sustained-release composition includes a liposomally entrapped elastase variant or substrate. Liposomes are small vesicles composed of various types of lipids, phospholipids, and/or surfactants. These components are typically arranged in a bilayer formation, similar to the lipid arrangement of biological membranes. Liposomes containing elastase variants or substrates are prepared by known methods, such as, for example, those described in Epstein, et al. (1985) PNAS USA 82:3688-92, and Hwang, et al., (1980) PNAS USA, 77:4030-34. Ordinarily the liposomes in such preparations are of the small (about 200-800 Angstroms) unilamellar type in which the lipid content is greater than about 30 mol. percent cholesterol, the specific percentage being adjusted to provide the optimal therapy. Useful liposomes can be generated by the reverse-phase evaporation method, using a lipid composition including, for example, phosphatidylcholine, cholesterol, and PEG-derivatized phosphatidylethanolamine (PEG-PE). If desired, liposomes are extruded through filters of defined pore size to yield liposomes of a particular diameter. Pharmaceutical compositions can also include an elastase variant or substrate adsorbed onto a membrane, such as a silastic membrane, which can be implanted, as described in International Publication No. WO 91/04014.

Pharmaceutical compositions of the invention can be stored in any standard form, including, e.g., an aqueous solution or a lyophilized cake. Such compositions are typically sterile when administered to recipients. Sterilization of an aqueous solution is readily accomplished by filtration through a sterile filtration membrane. If the composition is stored in lyophilized form, the composition can be filtered before or after lyophilization and reconstitution.

The invention also provides articles of manufacture including such compositions and related kits. The invention encompasses any type of article including a composition of the invention, but the article of manufacture is typically a container, preferably bearing a label identifying the composition contained therein. The container can be any formed from any material that does not react with the contained composition and can have any shape or other feature that facilitates use of the composition for the intended application. A container for a pharmaceutical composition of the invention generally has a sterile access port, such as, for example, an intravenous solution bag or a vial having a stopper pierceable by a hypodermic injection needle.

Kits of the invention generally include one or more such articles of manufacture and preferably include instructions for use. Exemplary kits include (1) multiple containers, each containing an elastase variant with a different substrate specificity, and (2) a container including an elastase variant of the invention along with a container including a substrate for that elastase variant. A preferred example of the latter is a container including a pharmaceutical composition comprising an elastase variant that cleaves a prodrug together with a container including a pharmaceutical composition comprising the prodrug and, optionally, instructions for use of the compositions in the treatment of a disease or disorder, such as cancer.

Nucleic Acid Molecules. Vectors, and Host Cells The present invention also includes nucleic acid molecules encoding the elastase variants and nucleic acid molecules encoding the substrates according to the invention.

A nucleic acid molecule of the present invention can be incoφorated into a vector for propagation and/or expression in a host cell. Such vectors typically contain a replication sequence capable of effecting replication of the vector in a suitable host cell (i.e., an origin of replication) as well as sequences encoding a selectable marker, such as an antibiotic resistance gene. Upon transformation of a suitable host, the vector can replicate and function independently of the host genome or integrate into the host genome. Vector design depends, among other things, on the intended use and host cell for the vector, and the design of a vector of the invention for a particular use and host cell is within the level of skill in the art.

If the vector is intended for expression of an elastase variant, the vector includes one or more control sequences capable of effecting and/or enhancing the expression of an operably linked elastase variant coding sequence. Control sequences that are suitable for expression in prokaryotes, for example, include a promoter sequence, an operator sequence, and a ribosome binding site. Control sequences for expression in eukaryotic cells include a promoter, an enhancer, and a transcription termination sequence (i.e., a polyadenylation signal).

An elastase expression vector can also include other sequences, such as, for example, nucleic acid sequences encoding a signal sequence or an amplifiable gene. As discussed above, a signal sequence directs the secretion of a polypeptide fused thereto from a cell expressing the protein. In the expression vector, nucleic acid encoding a signal sequence is linked to an elastase variant coding sequence so as to preserve the reading frame of the elastase variant coding sequence. The inclusion in a vector of a gene complementing an auxotrophic deficiency in the chosen host cell allows for the selection of host cells transformed with the vector.

A vector of the present invention is produced by linking desired elements by ligation at convenient restriction sites. If such sites do not exist, suitable sites can be introduced by standard mutagenesis (e.g., site-directed or cassette mutagenesis) or synthetic oligonucleotide adaptors or linkers can be used in accordance with conventional practice.

The present invention also provides a host cell containing a vector of this invention. A wide variety of host cells are available for propagation and/or expression of vectors. Examples include prokaryotic cells (such as E. coli and strains of Bacillus, Pseudomonas, and other bacteria), yeast or other fungal cells (including S. cerevesiae and R. pastoris), insect cells, plant cells, and phage, as well as higher eukaryotic cells (such as human embryonic kidney cells and other mammalian cells). Host cells according to the invention include cells in culture and cells present in live organisms, such as transgenic plants or animals. A vector of the present invention is introduced into a host cell by any convenient method, which will vary depending on the vector-host system employed. Generally, a vector is introduced into a host cell by transformation (also known as "transfection") or infection with a virus (e.g., phage) bearing the vector. If the host cell is a prokaryotic cell (or other cell having a cell wall), convenient transformation methods include the calcium treatment method described by Cohen, et al. (1972) Proc. Natl. Acad. Sci., USA, 69:2110-14. If a prokaryotic cell is used as the host and the vector is a phagemid vector, the vector can be introduced into the host cell by infection. Yeast cells can be transformed using polyethylene glycol, for example, as taught by Hinnen (1978) Proc. Natl. Acad. Sci, USA, 75: 1929-33. Mammalian cells are conveniently transformed using the calcium phosphate precipitation method described by Graham, et al. (1978) Virology, 52:546 and by Gorman, et al. (1990) DNA and Prot. Eng. Tech., 2:3-10. However, other known methods for introducing DNA into host cells, such as nuclear injection, electroporation, and protoplast fusion also are acceptable for use in the invention.

Production Methods To produce elastase variants recombinantly, host cells containing an elastase variant expression vector are prepared and cultured under conditions suitable for cell growth and for expression of the elastase variant. In particular, the culture medium contains appropriate nutrients and growth factors for the host cell employed. The nutrients and growth factors are, in many cases, well known or can be readily determined empirically by those skilled in the art. Suitable culture conditions for mammalian host cells, for instance, are described in Mammalian Cell Culture (Mather ed., Plenum Press 1984) and in Barnes and Sato (1980) Cell 22:649. In addition, the culture conditions should allow transcription, translation, and protein transport between cellular compartments. Factors that affect these processes are well-known and include, for example, DNA/RNA copy number; factors that stabilize DNA; nutrients, supplements, and transcriptional inducers or repressors present in the culture medium; temperature, pH and osmolality of the culture; and cell density. The adjustment of these factors to promote expression in a particular vector-host cell system is within the level of skill in the art. Principles and practical techniques for maximizing the productivity of in vitro mammalian cell cultures, for example, can be found in Mammalian Cell Biotechnology: a Practical Approach (Butler ed., IRL Press (1991). The cell culture procedure employed in the production of an elastase variant of the present invention can be any of a number of well-known procedures for large- or small-scale production of proteins. These include, but are not limited to, the use of a shaken flask, a fluidized bed bioreactor, a roller bottle culture system, and a stirred tank bioreactor system. An elastase variant can be produced, for instance, in a batch, fed-batch, or continuous mode. Methods for recovery of recombinant proteins produced as described above are well-known and vary depending on the expression system employed. For example, if, as is typical, the elastase variant includes a signal sequence, the elastase variant is recovered from the culture medium or the periplasm. Conveniently, the variant is secreted into the periplasmic space as a mature protein. The elastase variants can also be expressed intracellularly and recovered from cell lysates.

The elastase variant can be purified from culture medium or a cell lysate by any method capable of separating the variant from components of the host cell or culture medium. Typically, the elastase variant is separated from host cell and/or culture medium components that would interfere with the intended use of the elastase variant. As a first step, the culture medium or cell lysate is usually centrifuged or filtered to remove cellular debris. The supernatant is then typically concentrated or diluted to a desired volume or diafiltered into a suitable buffer to condition the preparation for further purification. The elastase variant is typically further purified using well-known techniques.

The technique chosen will vary depending on the properties of the elastase variant. If, for example, the elastase variant is expressed as a fusion protein containing an affinity domain, purification typically includes the use of an affinity column containing the cognate binding partner. For instance, elastase variants fused with hexahistidine or similar metal affinity tags can be purified by fractionation on an immobilized metal affinity column.

The following exemplary procedures can be used or adapted for purifying an elastase variant of the invention: fractionation on an immunoaffinity column, fractionation on an ion-exchange column, ammonium sulfate or ethanol precipitation, reverse phase HPLC chromatography on silica, isoelectric focusing, SDS-PAGE, or gel filtration.

In another embodiment, the invention provides an elastase production method that produces an elastase or elastase variant that does not require zymogen activation. The method employs a nucleic acid molecule encoding an elastase or elastase variant according to the invention, in which the prosequence is absent. An exemplary nucleic acid molecule encoding human neutrophil elastase lacking two N-terminal amino acids is described in Example 1. In the encoded elastase, the two-residue prosequence of human neutrophil elastase, Serl-Glu2, is deleted, so the protein begins with Ile3 of human neutrophil elastase. As described in the example, the elastase or elastase variant according to this embodiment can be linked to a signal sequence.

The nucleic acid molecule encoding an elastase or elastase variant that does not require zymogen activation can be inserted into a vector, which can be introduced into a host cell, as discussed above. To produce the elastase or elastase variant, the nucleic acid molecule is inserted into an expression vector, which is introduced into a suitable expression host. The active elastase or elastase variant is produced by culturing the expression host and recovering the active elastase or elastase variant. The selection of vectors, host cells, and cell culture methods is as described above in connection with elastase variants and substrates.

Cleavage and Therapeutic Methods

The invention includes a method for cleaving a substrate in which an elastase variant is contacted with a substrate, which includes a substrate site according to the invention. This step is carried out, in vivo or in vitro, under conditions that allow the elastase variant to cleave the substrate. Exemplary reaction conditions are provided in the examples.

In one embodiment, the substrate site is incoφorated into a fusion protein, as described above, and cleavage separates two protein domains, such as, for example, an affinity domain and a polypeptide of interest. In this embodiment, the cleavage method can be used as a step in affinity purification of the polypeptide. In another embodiment, the cleavage method is used in therapy, preferably prodrug therapy, such as, for example, antibody-directed prodrug therapy (ADEPT). This embodiment employs a modified elastase variant including an elastase variant attached to a targeting moiety or domain that directs the elastase variant to a particular cell or tissue type, such as a tumor. In prodrug therapy, a pharmaceutical composition including the modified elastase variant is administered to a patient along with a pharmaceutical composition including a prodrug substrate that is specifically cleaved by the modified elastase variant. Cleavage of the substrate converts the prodrug to an active drug.

In one embodiment, the pharmaceutical compositions of the invention are administered to an animal, typically a mammal. In a variation of this embodiment, the compositions are used to treat humans.

Methods for administering pharmaceutical compositions according to the invention do not differ from known methods for administering therapeutic proteins. Suitable routes of administration include, for example, intravenous, intraperitoneal, intracerebral, intramuscular, intraocular, intraarterial, or intralesional routes.

Pharmaceutical compositions of the invention can be administered continuously by infusion, by bolus injection, or, where the compositions are sustained-release preparations, by methods appropriate for the particular preparation.

Dosages for pharmaceutical compositions according to the invention depend, for example, upon the therapeutic objectives, the route of administration, and the condition of the recipient. Accordingly, it is necessary for the clinician to titer the dosage and modify the route of administration as required to obtain the optimal therapeutic effect. A typical daily dosage can range from about 1 μg/kg up to about 100 mg/kg of body weight or more per day, but is typically between about 10 μg/kg/day to 10 mg/kg/day. Generally, the clinician begins with a low dosage of a pharmaceutical composition and increases the dosage until the desired therapeutic effect is achieved.

In prodrug therapy, the modified elastase and substrate need not be administered simultaneously, or by the same route of administration, as long as both reach the intended target cell or tissue. The delivery method(s) employed are preferably selected to increase substrate cleavage in the vicinity of the target, while reducing the amount of cleavage occurring elsewhere in the body. This consideration is particularly important when the prodrug is a cytotoxic agent, such as a cytotoxic polypeptide used for cancer chemotherapy. The administration of modified elastase variants and substrates of the invention can be combined with other therapeutic regimens. For the treatment of cancer, radiation and/or a chemotherapeutic agent can be administered concomitantly with an elastase variant and prodrug substrate. Suitable preparation and dosing schedules for chemotherapeutic agents are as recommended by the manufacturer or as determined empirically by the clinician. Preparation and dosing schedules for standard chemotherapeutic agents are found in Chemotherapy Service Perry ed., (Williams & Wilkins (1992)). Administration of the chemotherapeutic agent can precede, or follow, administration of the elastase variant and/or substrate, or the chemotherapeutic agent can be given simultaneously with either or both. Antibodies against tumor-associated antigens, such as antibodies that bind EGFR, ErbB-2, ErbB- 3, or ErbB-4 receptor, or vascularendothelial factor (VEGF) can also be co- administered with a pharmaceutical composition(s) of the invention, as can one or more cytokines.

EXAMPLES Example 1 Design of Human Neutrophil Elastase Variant This example describes the design of an elastase variant according to the invention using molecular modeling. The 3-dimensional structure of elastase complexed with turkey ovomucoid inhibitor third domain (Bode, W., Wei, A. Z., Huber, E. M., Meyer, E., Travis, J. and Neumann, S. (1986) EMBO J., 5:2453-2458) was analyzed using Insight II 95.0 (Molecular Simulations, San Diego, CA) running on an Indigo work station (Silicon Graphics, Mountain View, CA). The P2 and PI' residues of the inhibitor were replaced with histidine. These modeled histidines were adjusted manually to mimic the interaction between the catalytic H43 and the other catalytic triad residues, SI 75 and D90. Hydrogen bond angles and distances as well as angles between the catalytic or substrate histidines and the other catalytic groups were estimated (Carter, P. and Wells, J. A. (1987) Science 237:394-399). A priori the non-ionized imidazoyl ring of a substrate histidine might exist in either NδlH or Nε2H tautomers, in which Nδl and Nε2 atoms are protonated, respectively. The catalytic histidine of elastase, H43, adopts the NδlH tautomer with Nε2 poised to accept a proton from the catalytic serine, SI 75 (Bode, W., Wei, A. Z., Huber, E. M., Meyer, E., Travis, J. and Neumann, S. (1986) EMBO J., 5:2453-2458). The histidine in model peptides also commonly adopt the NδlH tautomer although this depends upon the local environment (Creighton, T. E. (1984) in Proteins, Structure and Molecular Properties, (W. H. Freeman, New York) pp. 15). Therefore, the structural mimicry of the catalytic histidine by P2 and PI' substrate histidines in the more likely NδlH tautomer was investigated before considering the Nε2H tautomer.

The substrate histidines were initially modeled such that their Nδl and Nε2 atoms best approximate the position of the corresponding nitrogens from the catalytic histidine (Fig. 1). Molecular details were investigated by comparing substrate and catalytic histidines in their torsion angles and possible hydrogen-bond interactions with other catalytic residues (Table 1). The modeled P2 histidine and the catalytic histidme (model 1) have very similar hydrogen bond angles with and distances to SI 75, albeit at the expense of a χ₂ torsion angle that falls outside the range of observed histidine rotamers (Ponder, J. W. and Richards, F. M. (1987) J. Mol. Biol., 193:775-791). A histidine at PI' is less favorable than one at P2 in mimicking the interaction between H43 and other members of the catalytic triad (Table 1). In particular, a PI' histidine, unlike a P2 histidine, is too distant from D90 to form a direct hydrogen bond (models 3 and 1, respectively). When the dihedral angles of the P2 (model 2) and PI' (model 4) histidines are constrained to ideality, the only plausible modeled hydrogen bond is between the P2 histidine and SI 75.

Analogous models were constructed in which the substrate histidine was in the Nε2H tautomer with the unprotonated NδlH poised to accept a proton from SI 75. Modeling results with the Nε2H tautomer (Table 1 , models 1 -4) were broadly similar to those with the NδlH tautomer (Table 1, models 5-8). This significant structural mimicry observed between substrate and catalytic histidines supported replacement of the elastase catalytic histidine with alanine (H43A) and investigation of the ability of P2 and PI' substrate histidines to substitute functionally for the missing catalytic group.

Table 1. Bond angles and distances modeled for substrate-assisted catalysis by P2 and PI' substrate histidines in elastase

Angles Distances (A)

Histidine Dihedral H bonds Nε2 His -> Nδl His -> Catalytic His - -> His P2 or Pl' residue Xi X2 Ser-> His His-> Asp Oγ Ser Oδl Asp Oδ2 Asp Nε2 / Nε2 Nδl / Nδl

His43 71° -103° 166° 145° 2.48 3.26 2.56

His P2 model 1 72° 130° 172° 155° 2.34 4.34 3.36 0.84 1.42

His P2 model 2 60° 90° 161° 126° 2.45 4.68 3.42 0.28 1.62

His PI' model 3 -169° -166° 167° 140° 3.00 7.44 7.38 1.22 4.99

His PI' model 4 -180° -90° 167° 139° 3.74 7.62 7.54 1.82 5.30

Nδl His -> Nε2 His -> Catalytic His -> His P2 or Pl' Oγ (Ser) Oδl Asp Oδ2 Asp Nε2/ Nδl Nδl/ Nε2

His P2 model 5 72° -50° 122° 147° 2.47 3.69 2.75 1.21 0.52

His P2 model 6 60° -90° 141° 125° 2.48 4.27 3.21 1.29 1.23

His PI' model 7 -169° 14° 166° 145° 2.56 6.26 6.09 1.90 3.87

His PI' model 8 -180° 90° 175° 157° 3.66 6.37 6.43 2.31 4.29

Histidine dihedral angles are defined by χ, (N-Cα-Cβ-Cγ) and χ₂ (Cα-Cβ-Cγ-Cδ) whereas the Ser-> His and His -> Asp angles repres angles Oγ Ser-Hδ Ser-Nε2 His and Nδl His-Hε His-Oδ2 Asp (models 1 to 4) or Oγ Ser-Hδ Ser-Nδl His and Nε2 His-Hε His-Oδ2 Asp (mo respectively. Distances and angles were determined as described in Materials and Methods.

Example 2 Construction and Purification of Elastase Variants. Active elastase was expressed by juxtaposing a leader sequence directly against the start of a synthetic gene for mature elastase. In vivo cleavage of the leader sequence in R. pastoris released active elastase with the authentic amino terminus. This strategy obviates the needed for zymogen activation by exogenous protease addition as previously used for recombinant elastase (Okano, K., Aoki, Y., Shimizu, H. and Naruto, M. (1993) Biochem. Biophys. Res. Commun. 167:1326-1332). This avoids a source of protease contamination which might otherwise potentially confound activity measurements of catalytic triad mutants.

Construction of Synthetic Elastase Gene. A synthetic gene encoding elastase was assembled from 24 synthetic oligonucleotides (54- to 68-mer) sharing 4 bp overlaps with adjacent oligonucleotides as described (Dennis, M. S., Carter, P. and Lazarus, R. A. (1993) Proteins: Struct. Funct. Genet., 15:312-321). The H43A mutation and a hexahistidine-encoding sequence were installed by site-directed mutagenesis to produce the H43A elastase gene (Kunkel, T. A., Roberts, J. D. and Zakour, R. A. (1987) Methods Enzymol., 154:367-382) and the nucleotide sequence verified.

Wild-type and H43 A elastase genes were cloned as Xhol-EcoRl fragments into the P. pastoris expression vector, pPIC9 (Invitrogen, Carlsbad, CA) to create pPIC9WThne and pPIC9H43Ahne, respectively. The fragments were designed to encode the sequence Glu-Lys-Arg N-terminal to Ile3, the first amino acid residue of the mature elastase protein. The translated product is a fusion protein of an 85 amino acid signal sequence of the mating typε-α-factor (Matα) with the 247 amino acid mature elastase gene. The Matα signal sequence contains a cleavage site for the P. pastoris Kex2 protease. Use of restriction fragments encoding the Glu-Lys-Arg amino-terminus resulted in the retention of the Re*2 cleavage site in the translated product. The 2 residue pro sequence of elastase, Serl-Glu2, was puφosely omitted in anticipation that cleavage by Kex 2 protease in vivo would generate the amino terminus found in the mature active enzyme rather than the inactive zymogen. The H43A elastase gene Xhol-EcoRl fragment was also cloned into the mammalian expression vector, pRK5 (Suva, L. J., Winslow, G. A., Wettenhall, R. E., Hammonds, R. G., Moseley, J. M., Diefenbach-Jagger, H., Rodda, C. P., Kemp, B. E., Rodriguez, H., Chen, E. Y., Hudson, P. J., Martin, T. J. and Wood, W. I. (1987) Science, 237:893-896), to create pRK5H43Ahne.

Elastase Production. Human embryonic kidney 293 cells (American Type Culture Collection, CRL-1573) were transfected with ρRK5H43Ahne (Gorman, C. M., Gies, D. R. and McCray, G. (1990) DNA Prot. Engin. Tech., 2:3-10). The conditioned media was ultrafiltrated, dialyzed against 50 mM Tris-HCl (pH 8.0), 1 M NaCl (buffer A) and applied to a 1 ml nickel nitrilotriacetic acid (Ni ⁺ -NT A)

Superflow column (Qiagen, Valencia, CA). The resin was washed with 20 ml buffer A containing 10 mM imidazole and the H43A elastase eluted with a gradient of 10- 200 mM imidazole in buffer A. Purified elastase from 293 cells and was dialyzed 4 times against 100 mM Tris-HCl (pH 8.0), flash frozen and stored at -70 C. The yield of elastase was estimated by amino acid hydrolysis and ELISA.

Larger quantities of H43A elastase together with the wild-type enzyme were obtained following expression in P. pastoris GS115 cells following the manufacturer's recommendations (Invitrogen).

Elastase was affinity-purified from culture supernatants using 5 mg each of the anti-elastase monoclonal antibodies, 4E4 and 5A1 (Genentech), immobilized on 10 ml CNBr-activated Sepharose 4B ™ (Amersham Pharmacia Biotech). The resin was washed with 500 ml phosphate-buffered saline (pH 7.4) and the elastase eluted with 50 mM triethylamine (pH 11.0). Eluted fractions were neutralized with 1.0 M Tris- HCl (pH 7.0), adjusted to 5 mM EDTA (0.5 h, 4 °C) then 50 mM MgCl₂ (0.5 h, 4 °C). The elastase was then adjusted to 25 mM imidazole and applied to a 3 ml Ni -NTA column. The resin was washed with 60 ml 25 mM imidazole in 50 mM Tris-HCl (pH 8.0) and the elastase eluted with 350 mM imidazole in 50 mM Tris-HCl (pH 7.5).

Elastase ELISA. Elastase purified from human neutrophils (Calbiochem, La Jolla, CA) was quantified by amino acid analysis to provide an ELISA standard. Individual wells of a 96-well Immuno plate (Nunc, Rochester, NY) were coated with 1 μg bovine pancreatic trypsin inhibitor (BPTI, Sigma, St Louis, MO) or 1 μg of the 4E4 monoclonal antibody. The plates were blocked with 1 % (w/v) bovine serum albumin (Intergen, Purchase, NY), incubated with samples or standards, then with a sheep anti-elastase polyclonal antibody (Biodesign International, Kennebunk, ME) followed by a horseradish peroxidase conjugate of a mouse anti-sheep polyclonal antibody (Sigma). Peroxidase activity was detected with o-phenylenediamine dihydrochloride (Sigma) and the reaction was quenched with 4 M HCl. The absorbance at 450 nm was measured with a Spectramax 340 plate reader and SoftMaxPro 1.2.0 software (Molecular Devices, Sunnyvale, CA). The signal response was found to be linear over the range 0-10 ng/ml and 0-50 ng/ml elastase for the 4E4 monoclonal antibody and BPTI-coated wells, respectively. Similar results were obtained with both BPTI and 4E4 coats, suggesting that the recovered elastase is fully active.

The levels of H43A and wild-type elastase in P. pastoris reached plateaus of 1.0 μg/ml and 0.1 μg/ml, respectively, following 72 h induction with methanol. The elastase variants were affinity-purified using 2 anti-elastase antibodies followed by immobilized-metal affinity chromatography (IMAC). H43A and wild-type elastase were recovered in up to 5% yield and > 90%> homogeneity as evidenced by ELISA and SDS-PAGE, respectively. The amino termini of H43A and wild-type elastase were found to be IVGGRRAR, consistent with Kex 2 cleavage immediately following the Matα domain of the fusion protein.

Expression levels of H43A elastase from 293 cells were > 4-fold lower than those obtained with P. pastoris. Nevertheless it proved possible to purify > 50 μg of H43A elastase by IMAC from a total of ~5 liters conditioned media. This was sufficient for phage panning work whereas subsequent studies were undertaken with H43A produced in P. pastoris. Example 3 Construction Substrate Phage Libraries A phage library was constructed starting from the template phGH-LIB-G3 (Matthews, D. J. & Wells, J. A. (1993) Science 260:1113-1117) by site-directed mutagenesis (Kunkel, T. A., Roberts, J. D. and Zakour, R. A. (1987) Methods Enzymol., 154:367-382) using the oligonucleotide 5'- AGCTGTGGCCCAGGTGGTNNSN

NSCACNNSNNSNNSGGTGGTCCAGGGTCGACTGGCGGTGGCTCT 3', where N = T, C, G or A and S = G or C. The template contains eight stop codons and introduces a frame shift between the sequences encoding human growth hormone (hGH) and Ml 3 gene III (gill) so that only mutagenized phagemids will give rise to hGH-displaying phage. Correctly mutagenized phage contain the linker sequence GPGGX₃HX₂GGPG, where X is any amino acid, juxtaposed between hGH and the carboxy terminal domain of the Ml 3 gene III protein.

Example 4 Substrate Phage Selection The library was propagated and panned on hGH receptor as previously described (Matthews, D. J. & Wells, J. A. (1993) Science 260:1113-1117) with the following modifications. Phage were released with 1 μM H43A elastase derived from 293 cells (0.5 h, 25 C). Protease resistant phage, i.e., those still bound to the plates after treatment with H43A elastase, were then eluted with 50 mM glycine (pH 2.0). The selection procedure was then repeated 6 times. Clones from the protease- sensitive and protease-resistant pools were sequenced after 4 and 7 rounds of panning. This library of 6 x 10 clones is large enough to represent a significant proportion of the 3.4 x 10 (32 ) possible codon permutations from the NNS randomization strategy. However, sampling of the library was limited by the number of phage that were captured on immobilized hGH receptor <10 . Phage released with H43A elastase were propagated and subjected to further rounds of panning, as were protease resistant phage. After 7, but not 4, rounds of panning, a strong consensus sequence had emerged for the H43A sensitive (Fig. 3A, C) but not H43A resistant clones (Fig. 3B, D). The H43A elastase sensitive clones have predominantly R or M at position X, (P4), E or Q at X₂ (P3), and V or T at X₄ (PL). The specificity is broader at the other positions with M, T, V and I being frequently found at X₄ (PL) and Y, W, L and F at X₅ (P2') (Fig. 3C).

Example 5 Cleavage of Fusion Proteins Histidine-Dependant Proteolysis by H43A Elastase. The ability of H43A elastase to cleave sequences identified from substrate phage was investigated using Z- AP fusion proteins constructed with 9 such sequences as linkers (Table 2, L1-L9). The phagemid, pZAP, encodes a fusion protein (Z-AP) in which the synthetic Z domain of Staphylococcus aureus protein A is joined by a linker sequence to Escherichia coli alkaline phosphatase (Carter, P., Nilsson, B., Burnier, J. P., Burdick, D. and Wells, J. A. (1989) Proteins: Struct. Funct. Genet., 6:240-248). Different linker sequences were installed by site-directed mutagenesis (Kunkel, T. A., Roberts, J. D. and Zakour, R. A. (1987) Methods Enzymol., 154:367-382). Z-AP fusion proteins were expressed in E. coli and purified as described previously (Carter, P., Nilsson, B., Burnier, J. P., Burdick, D. and Wells, J. A. (1989) Proteins: Struct. Funct. Genet., 6:240-248). Z-AP fusion proteins (6 μM) were digested (0.5-10 h, 37 °C) with either H43A (1 μg/ml) or wild-type (20 ng/ml) elastase in 100 mM Tris-HCl (pH 8.0), 5 mM EDTA in the presence (H43A) or absence (wild-type) of 1 mM phenylmethylsulfonyl fluoride (PMSF). The digests were terminated with Tris- glycine-SDS sample buffer and analyzed on 8% SDS-polyacrylamide gels (Novex, San Diego, CA). Gels were stained for 3 hours with Serva blue G (Serva, Heidelberg, Germany) and destained for 4-5 h in 10% (v/v) acetic acid 20%> (v/v) ethanol. Initial rates of Z-AP cleavage were estimated from 5-8 successive time points, under conditions where < 10 % of the fusion protein was digested.

Nine additional fusion proteins were designed, five by combining residues frequently observed in phage-derived sequences (L10-L14), three with a PI' histidine (L15-L17) and a control sequence lacking a histidine (LI 8). These 18 different Z-AP fusion proteins allowed assessment of proteolysis by H43A elastase, including histidine dependence and the subsite position of the histidine (P2 or PI') as well as specificity at other subsites (Table 2).

Z-AP fusion proteins were secreted from E. coli grown in shake flasks, and recovered in yields of 0.3-0.8 mg/L by IgG affinity chromatography. The initial rate of cleavage of Z-AP fusion proteins (54 kDa) by H43A and wild-type elastase was determined from the release of AP (47 kDa) as followed by SDS-PAGΕ and scanning laser densitometry (Table 2). Z domain release was not followed because of its small size (7 kDa) and weak staining with Coomassie blue (Carter, P., Nilsson, B., Burnier, J. P., Burdick, D. and Wells, J. A. (1989) Proteins: Struct. Funct. Genet., 6:240-248). Several histidine-containing Z-AP fusion proteins were cleaved by H43A elastase including six of the nine phage-derived and four of the eight designed sequences. In each case, the histidine residue was located at the P2 position as evidenced by amino terminal sequence analysis of the AP product (Table 2). Cleavage by H43A elastase with a PI' histidine was not detected, even for 3 puφosely designed linkers (L15-L17) that were cleaved by wild-type elastase (Table 2). The absence of cleavage by H43A elastase at other histidines in Z-AP fusion proteins likely reflects that only two out of twelve of these sites have favorable PI residues (LVAH VTS and SQΕH TGS) and these sites are at least partially buried within AP (Sowadski, J. M., Handschumacher, M. D., Krishna Murthy, H. M., Foster, B. A. and Wyckoff, H. W. (1985) J. Mol. Biol., 186:417-433).

Table 2. Proteolysis of Z-AP fusion proteins by H43A and wild-type elastase

Initial cleavage rates

Z-AP υ H 3A ^υWιld-type ^υWιld-type

linker ^υH43A

Phage :-derived

LI REHV VY 5.0 ±0.1 510± 50 1.0 xlO²

L2 MEHTΨVY 1.4±0.1 750 ±50 5.3 x 10²

L3 REHVHIF 0.7 ±0.1 330 ±20 5.0 xlO²

L4 REHvHlW 0.60 ±0.01 510 ±50 8.5 x 10²

L5 REHvUlY 0.4 ±0.1 300 ± 40 7.5 x 10²

L6 RQHvUlY 0.20 ± 0.02 140 ±10 7.0 x 10²

L7 REHvllI < 0.002 80 ±10 >4.0xl0⁴

L8 REHT4Λ/Y < 0.002 210±10 >1.0x 10⁵

L9 REHT^IY < 0.002 150±10 >7.5xl0⁴

Designed

L10 MEHVUVY 7.2 ± 0.2 2700 ± 300 3.8 x 10²

Lll REHv vW 3.6 ±0.1 380 ±10 1.1 x 10²

L12 RQHvUvY 1.10 ±0.04 240 ± 30 2.2 xlO²

L13 REHV TY 0.5 ±0.1 1090 ±10 2.2 x 10³

L14 RQHTIVY < 0.002 100 ±10 >5.0x 10⁴

L15 REAV^HY < 0.004 1800 ±20 >4.5xl0⁵

L16 MEAV^HY < 0.004 2200 ± 200 >5.5xl0⁵

L17 MEAT^HY < 0.004 800 ± 40 >2.0xl0⁵

L18 REAV VY < 0.002 3500 ±100 >1.8xl0⁶

Initial cleavage rates of Z-AP substrates presented as means (± 2 SD) of measurements performed in duplicate or triplicate. All substrates were cleaved by wild-type elastase (-1), whereas some substrates were cleaved at the same position by H43A elastase (li). A Z-AP fusion protein containing the linker, REHVVY (LI), was readily cleaved to completion following extensive digestion with H43A or wild-type elastase (Fig. 4). Replacement of the histidine in the linker with alanine, REAVVY (LI 8), abolished detectable cleavage by H43A but not wild-type elastase (Fig. 4). Thus a P2 histidine is apparently a necessary but not sufficient condition for proteolysis by H43A but not wild-type elastase. H43A, but not wild-type elastase, is resistant to PMSF inhibition (Fig. 4). Thus, the catalytic histidine is apparently required for stable sulfonylation of the active-site serine as previously observed for H64A subtilisin BPN' (Carter, P. and Wells, J. A. (1987) Science 237:394-399). Subsite Specificity ofH43A and Wild-Type Elastase. Efficient cleavage of Z-

AP fusion proteins by H43A elastase apparently requires a P2 histidine as well as favorable residues at other subsites (P4 to P2') (Table 2). E. g., valine is strongly preferred over threonine at PI (L10 vs- L2, LI vs L8, L12 vs L14), glutamate is favored over glutamine at P3 (LI vs L12, L5 vs L6) and methionine is preferable to arginine at P4 position (L10 vs LI, L2 vs L8). Valine is favored over isoleucine (LI vs L5, L12 vs L6, ) and threonine (LI vs L13) at the PI' position, whereas tyrosine (LI, L5), tryptophan (LI 1) and phenylalanine (L3), but not isoleucine (L7), are favored residues at P2'. A designed linker, MEHVVY (LIO) containing the preferred residues identified at each subsite position, proved to be a more favorable substrate than any of the phage-derived sequences (L1-L9) evaluated. Beyond the P2 position, similar subsite specificity trends were observed for wild-type and H43A elastase. One exception is that threonine is slightly preferred over valine at the PI' position for wild- type elastase, whereas valine is very strongly favored over threonine at this position by H43A elastase (LI 3 vs LI). The subsite specificity of H43A elastase is strongly dependent upon residues at neighboring subsites. E. g., with arginine at P4, two linkers that were uncleavable with H43A elastase were converted to good substrates by replacement of PI threonine with valine (L8 vs LI and L14 vs L12, respectively). In contrast, in the context of the more favorable P4 residue, methionine, replacement of threonine with valine at PI resulted in a much more modest 5-fold enhancement in cleavage rate (L2 vs L10). Thus installing a favorable residue at one subsite diminishes the enhancement in substrate cleavage upon subsequent installation of a highly favorable residue at a second subsite.

Example 6

Comparison of Proteolysis Rates by H43A and Wild-Type Elastase The most favorable Z-AP fusion protein substrates for H43A elastase (LI, L2 and LIO) are cleaved at initial rates within 100 to 380-fold of those observed with wild-type elastase (Table 2). A more detailed kinetic analysis was undertaken to distinguish between effects upon k_cat and K_M (Fersht, A. in Enzyme Structure and Mechanism, Second edition, Freeman and Co., N.Y.). For some fusion proteins, initial cleavage rates were estimated over a range of substrate concentration (0.02—4 μM) and k_cat and K_M values estimated by a non-linear least squares fit of the data to the Michaelis-Menten equation using Kaleidagraph 3.0.8 (Synergy Software, Reading, PA). Kinetic parameters for hydrolysis of Z-AP fusion proteins with histidine, REHVVY (LI), and non-histidine, REAVVY (LI 8), -containing linkers were obtained from initial cleavage rates determined over a range of substrate concentrations (Table 3). The specificity constant, k_cat/K_M, for H43A elastase was only 160-fold lower than for the wild-type enzyme against the histidine-containing substrate (LI). This reflects an 80-fold decrease in k_cat and a 2-fold increase in K_M. In contrast, replacing the histidine in the substrate by alanine increased the activity with wild-type elastase by 7-fold and abolished detectable cleavage by H43 A elastase.

Table 3. Kinetic analysis of Z-AP cleavage by H43A and wild-type elastase

Elastase Z-AP -at κ_M kca/Kr i variant substrate s -1 nM s - 1 M

H43A REHVVY (1.69 ± 0.08) x lO^"3 600 ± 70 2.8 x 10³

H43A REAVVY < 6 x l0^"7 ND ND

Wild-type REHVVY (1.40 ± 0.04) x lO^"1 320 ± 20 4.4 x lO⁵

Wild-type REAVVY 1.03 ± 0.04 340 ± 30 3.0 x l0⁶

ND, No detectable cleavage.

The present invention has of necessity been discussed herein by reference to certain specific methods and materials. It is to be understood that the discussion of these specific methods and materials in no way constitutes any limitation on the scope of the present invention, which extends to any and all alternative materials and methods suitable for accomplishing the ends of the present invention.

All references cited herein are expressly incoφorated by reference.

Claims

What is claimed is:

1. A purified elastase variant having an amino acid sequence different from that of a precursor elastase, said difference comprising substitution of an active-site histidine residue coπesponding to residue 43 in human neutrophil elastase with a different amino acid residue such that said elastase variant has substrate specificity substantially different from said precursor elastase.

2. The elastase variant of Claim 1 wherein said different amino acid residue has a side chain volume less than the side chain volume of histidine

3. The elastase variant of Claim 2 wherein said different amino acid residue is alanine.

4. The elastase variant of Claim 1 wherein said precursor elastase is human neutrophil elastase.

5. The elastase variant of Claim 4 wherein said different amino acid residue has a side chain volume less than the side chain volume of histidine.

6. The elastase variant of Claim 5 wherein said different amino acid residue is alanine.

7. A modified elastase variant comprising the elastase variant of Claim 1 attached, directly or indirectly, to another molecule.

8. The modified elastase variant of Claim 7 wherein the other molecule is a targeting moiety or domain.

9. The modified elastase variant of Claim 8 wherein the targeting moiety comprises an antibody or antibody fragment.

10. The modified elastase variant of Claim 8 wherein the elastase variant and the targeting domain are portions of a fusion protein.

1 1. A purified substrate for an elastase variant of Claim 1, said substrate comprising a substrate site including: at least two amino acids; and a histidine residue at the P2 position.

12. The substrate of Claim 1 1 wherein said substrate site further comprises either a valine or threonine residue at the PI position.

13. The substrate of Claim 12 wherein said substrate site further comprises either an arginine or methionine residue at position P4.

14. The substrate of Claim 12 wherein said substrate site further comprises either a glutamine or glutamic acid residue at position P3.

15. The substrate of Claim 11, wherein the substrate is a fusion protein comprising at least two domains separated by said substrate site.

16. The fusion protein of Claim 15 wherein one domain is an affinity domain.

17. The substrate of Claim 1 1 wherein said substrate is a prodrug.

18. A composition comprising the elastase variant of Claim 1 and a physiologically acceptable carrier.

19. An article of manufacture comprising: a container; and the composition of Claim 18 contained in said container.

20. A composition comprising the substrate of Claim 1 1 and a physiologically acceptable carrier.

21. An article of manufacture comprising: a container; and the composition of Claim 20 contained in said container.

22. A kit comprising: first and second containers; the composition of Claim 18 contained in said first container; and a purified substrate for said elastase variant, said substrate comprising a substrate site including: at least two amino acids; and a histidine residue at the P2 position

23. A nucleic acid molecule encoding the elastase variant of Claim 1, wherein, if said nucleic acid molecule is naturally occurring, said nucleic acid molecule is purified.

24. A vector comprising the nucleic acid molecule of Claim 23.

25. A host cell comprising the vector of Claim 24.

26. A method of producing an elastase variant comprising: culturing the host cell of Claim 25; and recovering the elastase variant.

27. A nucleic acid molecule encoding the substrate of Claim 11, wherein, if said nucleic acid molecule is naturally occurring, said nucleic acid molecule is purified.

28. A vector comprising the nucleic acid molecule of Claim 27.

29. A host cell transformed with the vector of Claim 28.

30. A method of producing a substrate for an elastase variant comprising: culturing the host cell of Claim 29; and recovering the substrate.

31. A method for cleaving a substrate comprising the step of contacting the elastase variant of Claim 1 with a substrate comprising a substrate site, said substrate site comprising: at least two amino acids; and a histidine residue at the P2 position.

32. The method of Claim 31 wherein said substrate site further comprises either a valine or threonine residue at the PI position.

33. The method of Claim 32 wherein said substrate site further comprises either an arginine or methionine residue at position P4.

34. The method of Claim 33 wherein said substrate site further comprises either a glutamine or glutamic acid residue at position P3.

35. A nucleic acid molecule encoding an elastase or elastase variant lacking a prosequence.

36. A vector comprising the nucleic acid molecule of Claim 35.

37. A host cell comprising the vector of Claim 36.

38. A method of producing an elastase or elastase variant that does not require zymogen activation comprising: culturing the host cell of Claim 37; and recovering the elastase or elastase variant.