AU687751B2 - Compositions for the inhibition of protein hormone formation and uses therof - Google Patents

Description

~aaa 141~-4C-~ C 991_71

PCT

ANNOUNCEMENT OF THE LATER PUBLICATION OF INTERNATIONAL SEARCH REPORT INTERNATIONAL APPLICATION PUBLISHED UNDER THE PATENT COOPERATION TREATY (PCT) (51) International Patent Classification 5 (11) International Publication Number: WO 94/00555 CO7K 5/08, 13/00, C12P 21/08, C12Q 1/37 A3 A61K 37/02, 37/64, GOIN 33!573, 33/68 (43) International Publication Date: 6 January 1994 (06.01.94) A61K 39/395, GOIN 33/577 II C12N 15/28, 15/57, 9/64 (21) International Application Number: PCT/US93/06120 (74) Agents: McGARRIGLE, Philip. Jr. et al.: Chiron Corporation, 4560 Horton Street R440, Emervville. CA (22) International Filing Date: 25 June 1993 (25.06.93) 94608-2916 (US).

Priority data: (81) Designated States: AU, CA. JP. NO, European patent (AT.

07/905,546 25 June 1992(25.06.92) US BE, CH, DE. DK. ES, FR, GB, GR. IE, IT, LU, MC, r c NL, PT, SE).

Chirsor Corporohor\ (71) Applicant: CETUS ONCOLOGY CO Rp-A-Tfhsf [US/ US]; 4560 Horton Street, Emeryville, CA 94608-2916 Published With international search report.

Before the expiration of the time limi for amending the (72) Inventors: KRIEGLER, Michael P.O. Box 9894, Rancho claims and to be republished in the event of the receipt of Santa Fe, CA 92067 PEREZ, Carl 7275 Char- amendments.

mant Drive, #326, San Diego, CA 92122 HALEN- BECK, Robert, F. 136 Spring Grove Avenue, San Ra- (88) Date of publication of the international search report: fael, CA 94901 JEWELL, David, A. 427 Sher- 17 February 1994 (17.02.94) wood Drive, #307, Sausalito, CA 94065 KOTHS, Kirston, E. 2646 Mir Vista Drive, El Cerrito, CA 94530 R f SSEC fi ;i: 113 t J (54)Title: COMPOSITIONS FOR THE INHIBITION OF PROTEIN HORMONE FORMATION AND USES THEREOF (57) Abstract Compounds having proteolytic inhibitory activity useful for treating diseases, particularly as applied to the treatment of sepsis, AIDS or autoimmune diseases, resulting from a decrease, in the circulating level of mature protein hormones derived from the proteolytic cleavage of prohormone precursors such as 26 kD TNF, and method, for identifying compounds having the desired inhibitory activity are provided.

IU~aaB~V~P~-~ ~p~Wr~PrP -r m- ~rsasp~-- WO 94/00555 PCT/US93/06120 -1- COMPOSITIONS FOR THE INHIBITION OF PROTEIN HORMONE FORMATION AND USES THEREOF This is a continuation-in-part of U.S. Serial No. 07/395,253, filed August 16, 1989, which is pending.

Field of the Invention This invention is in the area of immunology/biochemistry, and particularly concerns the development of compositions and methods for identifying inhibitors of protein hormone release, and prophylactic and therapeutic uses of the inhibitors for treating diseases associated with elevated levels of the hormones. More specifically, the invention facilitates the identification of compositions and methods for identifying inhibitors of a TNF convertase. These inhibitors may be used to treat a variety of diseases, particularly sepsis, rheumatoid arthritis, cachexia, AIDS and autoimmune diseases, and thus affords the physician alternate treatment regimes.

Background of the Invention In the United States alone nosocomial bacteremia develops in about 194,000 patients per year, and of these about 75,000 die. Maki, 1981, Nosocomial Infect., (Dikson, page 183, Yrke Medical Books, Most of these deaths are attributable to six major gram-negative bacilli Pseudomonas aeruginosa, Escherichia coli, Proteus, Klebsiella, Enterobacter and Serratia. The current treatment for bacteremia is the administration of antibiotics which, have limited effectiveness in treatment of septic shock.

The precise pathology of bacter.mia is not completely elucidated ertheless, it is known th- "oaJ n xin o ysadllccharides (LPS), are the primary causative agent. LPS consists of at least three significant antigenic regions: lipid A; core polysaccharide; and O-specific polysaccharide. The latter is also referred to as O-specific chain or simply O-antigen. The O-specific chain region is a long-chain polysaccharide built up from repeating polysaccharide units.

The number of polysaccharide units differs among different bacterial species and -L L ~i a e~ -I se WO 94/00555 PCT/US93/06120 -2may vary from one to as many as six or seven monosaccharide units. While the O-specific chain varies among different gram-negative bacteria, the lipid A and core polysaccharides are similar if not identical.

Since LPS plays a key role in sepsis, many approaches have been pursued to neutralize its acti' ity. Presently, there is considerable work which suggest that antiLPS antibody will soon be a valuable clinical adjunct to the standard antibiotic therapy.

LPS initiates a cascade of biochemical events that eventually causes the death of the patient. It is widely believed that an early result of LPS introduction is the stimulation of macrophage cells and the production of tumor necrosis factor (TNF) as a result of LPS. Thus, considerable effort has been expended to produce neutralizing antibody to TNF, or other molecules that could inhibit its effects. It is likely that antibody to TNF will have valuable clinical applications. Tracey, et 1987, Nature, 330:662.

TNF has been shown to exist in both membrane-bound and soluble secreted forms.

Decker, t al., 1987, J. of Immunol., 138:957; Kriegler, e al., 1988, Cell, 53:45.

Human TNF has been cloned and shown to consist of a 17 kD polypeptide, plus an unusually long 76 amino acid putative signal leader sequence. The 17 kD molecule is a key agent involved in initiating the biochemical cascade responsible for sepsis. It has been proposed by Kriegler, et al., 1988, Cell, 53:45, that TNF may exist as both a membrane bound 26 kD form, and a soluble form corresponding to the 17 kD species. The 26 kD form is the precursor, or prohormone, of the mature 17 kD molecule. It has further been proposed by Kriegler, e al. above, that the two forms of TNF may have different biological effects, primarily as a result of differences in tissue distribution.

It will be appreciated that because TNF plays a key role in causing sepsis and other diseases that there is a need to identify and develop anti-TNF prophylactics/therapeutics. As mentioned above, anti-TNF antibody appears to be promising, and has been shown to be effective in baboons. However, these studies have involved the use of non-human TNF and non-human TNF antibody.

C~OI~ C~ CI-I-IC-~ r C 'WO 94/00555 PCT/US93/06120 -3- From a practical standpoint non-human anti-TNF antibody will have limited therapeutic application because of immunologic rejection of the antibody by a patient's immune system. Consequently, a human antibody, or a genetically engineered antibody consisting of the human constant region and the mouse variable region ("humanized antibody") is preferred.

TNF, in addition to playing a criucal role in sepsis, has recently been shown to be involved in initiating the expression of human immunodeficiency virus in human cells that carry latent virus. Folks et al., 1989, PNAS (USA), 86:2365. Thus, preventing or inhibiting the formation of the 17 kD, or lower molecular weight forms of TNF might be a valuable prophylactic for the treatment of AIDS patients by preventing the expression of virus that is latent in the patient.

TNF also plays a role in various autoimmune diseases, particularly rheumatoid arthritis. Duff, et al. 1987, International Conference on Tumor Necrosis Factor and Related Cytotoxins, 175:10. Thus, compounds or methods for inhibiting TNF action will have considerable application for the treatment of a variety of diseases of immunologic origin.

In addition to antibody, other molecules with TNF inhibitory activity are being sought Non-antibody TNF inhibitors are described by Seckinger, et al., 1988, J.

Exp. Med., 167:151, and Seckinger, et al, 1989, J. Biol. Chem,. 264:11966, and in EP 0 3083 1S inventors Wallach, et al. The inhibitors are present in the urine of febrile patients, and have been purified and shown to have molecular weights of about 27,000-33,000. These inhibitors are now known to be soluble forms of the TNF receptor. Although these molecules exhibit TNF-inhibitory activity, neither of the inhibitors has yet been shown to be effective in the treatment of sepsis in humans.

From the foregoing discussion it is apparent that there is a need to identify and develop additional anti-TNF inhibitors, both antibody based or otherwise, that are efficacious in the treatment of sepsis.

Summary of the Invention L~L~ ~LL~ IL I I PC-e~- WO 94/00555 PCT/US93/06120 -4- In its most general form, the invention described herein presents methods and compositions for inhibiting the production of a mature form of TNF, from its prohormone precursor, proTNF. These compositions are useful for preventing or treating diseases in patients associated with elevated circulating levels of mature TNF. The invention also relates to a method for identifying molecules that inhibit the production of a mature form of TNF. Such inhibitors are distinguishable from antiTNF antibody, which neutralizes TNF. This method can be used to identify medicaments such as prophylactics and/or therapeutics for the treatment of sepsis and other diseases caused by the production of mature TNF. These medicaments are able to interfere with the cleavage of the 26 kD proTNF prohormone by enzymes termed convertases. Thus, these medicaments inhibit the production of lower molecular weight sepsis-inducing molecules 17 kD TNF).

Specifically, the preferred inhibitors as described herein interfere with the activity of a TNF convertase to prevent removal of the N-terminal portion of the 26 kD molecule including at least the 76-aminoacid signal sequence to produce a maturJ form of TNF such as the 17 kD TNF. The invention also includes a class of compounds that are both inhibitors of a TNF convertase and effective in the prevention and/or treatment of sepsis. Compounds in this class include anti-convertase antibody, muteins of the prohormone form, and proteins or peptides that compete with the 26 kD form of TNF for binding to the convertase.

Also claimed are small molecular weight compounds that specifically inhibit the class of proteases that includes TNF convertases, or preferably, show selective specificity for inhibition of TNF convertase.

Additionally, the present inventors have purified a TNF convertase to nearhomogeneity, discovered its amino acid sequence, and compared it to known serine proteases. The purified TNF convertase contains an N-terminal amino acid sequence essentially identical to PR-3, a known neutrophil protease having the same molecular weight. They have also identified various inhibitors of TNF convertase and have tested them in in vitro and in vivo assays.

~YM YIC--QC ~P C I C_ C WO 94/00555 PCI'/US93/06120 Specifically, an object of the present invention is to provide small molecules that specifically inhibit TNF convertases.

Another object of the invention is a method for treating diseases such as septicemia, septic shock, cerebral malaria, rheumatoid arthritis, AIDS, cachexia, and graft-versus-host disease by administering a PR-3 inhibitor.

In one aspect of this invention, a method for identifying a prophylactic or therapeutic of a disease caused by a mature tumor necrosis factor (TNF) produced from a proTNF by cleavage of said proTNF by a TNF convertase is provided, the method comprising the steps of:(a) contacting the proTNF with an amount of the TNF convertase effective for cleaving the proTNF; measuring the conversion of the proTNF to the mature TNF in step repeating steps and further including a molecule sought to be identified as a prophylactic or therapeutic of diseases caused by the mature TNF; measuring the conversion of the proTNF to the mature TNF in step and comparing the conversion measured in step with the conversion measured in step to determine whether the molecule is a suitable prophylactic or therapeutic of diseases caused by mature TNF.

In another aspect of the invention, a therapeutic or prophylactic compound for treating a disease caused by a mature TNF produced from a proTNF by cleavage of said proTNF by a TNF convertase is provided, the therapeutic or prophylactic identified by a method comprising the steps of: contacting the proTNF with an amount of the TNF convertase effective for cleaving the proTNF; measuring the conversion of the proTNF to the mature TNF in step repeating steps and further including a molecule sought to be identified as a prophylactic or therapeutic of diseases caused by the mature TNF; measuring the conversion of the proTNF to the mature TNF in step and comparing the conversion measured in step with the conversion measured in step to identify whether the molecule is a suitable prophylactic or therapeutic of diseases caused by mature

TNF.

In yet another aspect of the invention, a method for treating a patient having a disease or susceptible to a disease caused by a mature TNF produced from a I I~ -LU ~01111 ~d I C- I WO 94/00555 PCT/IUS93/06120 -6proTNF by cleavage of said proTNF by a TNF convertase is provided, the method comprising administering to a patient in need of such treatment an effective amount of an inhibitor of a TNF convertase. In a preferred embodiment, the disease is selected from the group consisting of sepsis, rheumatoid arthritis, cachexia, cerebral malaria, AIDS, and graft-versus-host disease.

In a further aspect of this invention, a pharmaceutical composition for the treatment of a disease caused by a mature TNF produced from a proTNF by cleavage of said proTNF by a TNF convertase is provided, the composition comprising an effective amount of an inhibitor of a TNF-convertase and a pharmaceutically acceptable excipient.

Brief Description of the Drawings Figure 1, panel A, shows the restriction map of the DNA sequence that encodes 26 kD TNF. Panel B shows a hydrophobicity plot of 26 kD TNF, and panel C shows the DNA and amino acid sequences of the molecule.

Figure 2 shows the predicted amino acid sequence of the unprocessed precursor of human PR-3, derived from the DNA sequence of the cDNA clone.

Figure 3 shows the conversion of 26 kD TNF by TNF convertase. Lanes A, B, and C show various controls: TNF 6.8 cell lysate 26 kD transcription/translation and incubation controls. Lanes D, E, and F show the conversion of transcription/translation generated 26 kD TNF to predominately 17 kD TNF by convertase present in either HL60 S-1 cytosol uninduced and induced fractions, or a P-1 pellet fraction prepared from induced cells. G is a blank lane.

Figure 4 shows the effect of convertase inhibitors on the conversion of 26 kD TNF to its lower molecular weight forms as determined by gel electrophoresis.

Lanes A,B,C,and D of panel 1 show, respectively; immunprecipitation of a cell lysate of the pFVXM-TNF6 transfected cell line TNF 6.8 (Kriegler, et al., 1988, in Cell, immunprecipitation of in vitro transcribed/translated 26 kD TNF, the effect of (1-((3((acetyloxyl)-7-methoxy-8-oxy-8-oxo-5-thio- -azabicyclo [4.2.0] oct-2-en- 2-yl) carbonyl) morpholine, S,S-dioxide, (6R-cis) on the conversion of ~s r~---pkr4pql p~C~C1 -I-s 1~ 1.

WO 94/00555 PCI7/US93/06120 -7- 26 kD TNF, and the conversion of 26 kD TNF in the absence of (1 -((3-(acetyloxyl)-7-methoxy-8-oxy-8oxo-5-thio- -azabicyclo oct-2-en- 2-yl) carbonyl) morpholine, S,S-dioxide, (6R-cis). Lanes A and B of panel 2 show, respectively; immunprecipitation of a cell lysate of the pFVXM-TNF6 transfected cell line TNF 6.8 (Kriegler, e al., 1988, in Cell, 53:45), and immunprecipitation of in vitro transcribed/translated 26 kD TNF. Lanes C and D show the conversion of 26 kD TNF in the presence and absence of 3,4dichloro-isocoumarin, respectively. Lanes E and F, show the conversion of 26 kD TNF in the presence and absence of elastinal, respectively.

Figure 5 shows additional gel electrophoretic assays on 26 kD TNF, demonstrating inhibition of purified human PR-3 from HL-60 cells by various serine protease inhibitors.

Figure 6A shows gel electrophoretic analysis of purified human neutrophil PR3 activity on 26 kD TNF, showing differential inhibitory activity of potential serine protease inhibitors.

Figure 6B shows similar results obtained using a colorimetric assay testing the same compounds.

Figure 7 shows the effect of prophylactic treatment of mice with a TNF convertase inhibitor prior to lethal injection with LPS: circulating serum TNF levels are decreased.

Figure 8 shows the effect of prophylactic treatment of mice with a TNF convertase inhibitor prior to lethal injection with LPS: survival is prolonged.

Detailed Description of the Invention Definitions To facilitate understanding the nature and scope of applicant's invention, several definitions regarding various aspects of the invention are presented below. It will be understood, however, that these definitions are general in nature, and encompassed within the definitions are meanings well known to those skilled in the art.

LLs B WO 94/00555 PCT/US93/06120 -8- "Sepsis" is herein defined to mean a disease resulting from gram positive or gram negative bacterial infection, the latter primarily due to the bacterial endotoxin, lipopolysaccharide (LPS). It can be induced by at least the six major gram-negative bacilli and these are Pseudonwnas aeruginosa, Escherichia coli, Proteus, Klebsiella, Enterobacter and Serratia.

The terms "prohormone" and "mature" hormone have the following meanings.

"Prohormone" is intended to cover proteins that contain a peptide segment which is removed during the in vivo production of the "mature" form of the hormone.

Preferably, these are proteins produced at least in part by cells of the immune system, such as T-cells or macrophages. The preferred embodiment of the invention is the 26 kD TNF prohormone, or "proTNF" as discussed in detail below. ProTNF is cleaved primarily to a 17 kD mature form, preferably having the N-terminal sequence of "mature TNF", Val-Arg-Ser-Ser. However, "mature TNF" is intended to include other cleavage products also formed from the prohormone. These cleavage products will retain the biological characteristics of the 17 kD form of mature TNF, and are truncated cleaved) forms of proTNF wherein at least the N-terminal 76-amino acid leader sequence is removed.

As used herein, "proTNF" refers to TNF having a molecular weight of about 26,000, which is the prohormone form of TNF-a (reference cloned sequence of Figure It is known that the propeptide segment of a prohormone varies in length depending on the species from which it is derived, but the amino acid sequence of this segment is highly conserved. Indeed, in the mouse, approximately 86% of the 79amino acids that make up the putative leader sequence of the prohormone are identical to the 76 known amino acids that compnse the putative leader of human TNF. Thus, it will be appreciated by tho skilled in the art that when reference is made to proTNF it is intended that the molecule can be derived from any particular species so that it may have a slightly altered leader sequence compared to the human sequence as is known in the art.

The term "convertase" or "TNF convertase", as used herein, refers to one or more enzymes normally present in an animal that are capable of cleaving 26 kD TNF to I Ib~e d y~C r-IP*~i~es C Ils lCw WO 94/00555 PCI/US93/06120 -9a mature TNF having TNF biological activity in trimeric form in a TNF bioassay.

In unstimulated cells, a convertase is recovered largely in fractions consisting substantially of membranes, although some activity is located in the cytosol. A TNF convertase is normally associated with cells that produce TNF. One TNF convertase is now known to be the serine protease "proteinase-3", also called "P-29B" or "myeloblastin".

The phrase "membrane-associated" as applied to TNF convertase indicate= a form of the convertase that is initially isolated in substantially insoluble form, as indicated by the presence of much of the convertase activity in a 30,000 x g pellet fraction. However, some TNF convertase is soluble when isolated from neutrophil granules.

"Recombinant antibody" refers to antibody wherein one porion of each of the amino acid sequences of heavy and light chain is homologous to corresponding sequences in antibody derived from a particular species or belonging to a particular class, while the remaining segment of the chains is homologous to corresponding sequences in another. Most commonly, in a recombinant antibody the variable region of both light and heavy chain copies the variable regions of antibody derived from one species of mammal, while the constant regions are homologous to the sequences in antibody derived from another. One example is "humanized" mouse antibody where the constant regions of the mouse antibody are replaced with a human constant region.

In its most general fo-m, the instant invention concerns methods and compositions for identifying inhibitors of diseases associated with the production of mature hormones from their prohormone forms. The preferred embodiment of a prohormone is 26 kD TNF, which is then cleaved to a lower molecular weight "mature" form, preferably 17 kD, which, in its multimeric (usually trimeric) form, is substantially involved in producing life-threatening physiological changes associated with sepsis. Thus, molecules which are capable of interfering with the conversion of the 26 kD TNF to the mature form are useful for preventing or treating sepsis.

I I IIIIII P~I 1911 .a II I WO 94/00555 PCi'/US93/06120 The assays described herein detect the conversion of a prohormone to its mature hormone form, with the preferred embodiment being the enzymatic conversion of the 26 kD molecular weight form of TNF to, preferably, a 17,000 molecular weight form. The enzyme responsible for the conversion is termed "TNF convertase". Thus, the invention is most readily presented in four parts. Part one shows the materials and methods for realizing proTNF, the 26 kD form of TNF.

Part two identifies sources of TNF-convertase, and methods for purifying the enzyme. Part three describes the identification of various convertase inhibitors.

Finally, part four of the invention presents a description of ways of using the inhibitors to treat patients suffering from sepsis or other diseases. Each of these sections will now be addressed separately.

Several patents/patent applications and scientific references are referred to below. The instant invention draws on some of the material and methods shown in these references, and thus it is intended that all of the references, in their entirety, be incorporated by reference.

1.26 kD TNF The TNF and proTNF of the current invention may be obtained in native, synthetic or recombinant forms by methods known in the art. While the recombinant systems described below render the 26 kD proTNF obtainable in considerable amounts and facilitate the assay procedures for TNF inhibitors, it will be appreciated that nonrecombinant systems may also be used. For instance, it has been shown that the 26 kD molecule can be identified in stimulated monocytes. Kriegler, et al., 1988, Cell, 5:45. Thus, a suitable assay procedare is to stimulate monocytes to produce the 26 kD proTNF molecule, and then to measure the disappearance of the 26 kD molecule as a result of action by the convertase. Preferably the 17,000 molecular weight mature TNF is generated.

The 26 kD proTNF is cleaved by convertase at one or more internal sites to generate "mature TNF". The major site is at the junction which separates the secreted form of TNF (the 17 kD species) from the leader sequence. The ~sPLar~ L L- WO 9(4/00555 PCI'/ US93/06120 11 sequence at this junction is Gln-Ala-Val-Arg-Ser-Ser-. The major cleavage site lies between alanine and valine, since valine is known to be the amino-terminal amino acid of the 17 kD molecule (the primary mature form). Several other species of TNF may be produced by the convertase, and these are the products of minor or secondary cleavage sites: for example, between the Val and the Arg in the sequence above, or between Pro and the Val located at +12 and +13 in the amino acid sequence. The assays described herein can monitor the inhibition of the conversion of the 26 kD proTNF species, or the appearance of a mature TNF form irrespective of its cleavage site, The proTNF form and mature TNF form have been cloned and expressed in a number of systems. For instance, the cloning of rabbit TNF is disclosed in EP 146,026, published June 26, 1985 (Dainippon Pharmaceutical Co., Ltd.) and EP 148,311, published July 17, 1985 (Asahi Kasei Kogyo Kabushiki). The cloning of human TNF having 151 and 155 amino acids (2 and 6 less than the native mature form) is disclosed in EP 155,549, published September 25, 1985 (Dainippon Pharmaceutical Co., Ltd.), and human TNF having 155 amino acids is disclosed in EP 158,286, published October 16, 1985 (Asahi Kasei Kogyo Kabushiki Kaisha) and corresponding GB 1,158,829A, published November 20, 1985. The cloning of mature TNF (157 amino acids) and various modified forms (muteins) thereof is disclosed in EP 168,214, published January 15, 1986 (Genentech) and PCT US 85/01921, filed October 3, 1985, (Cetus Corporation).

In addition, U.S. Patent Nos. 4,677,063 and 4,677,064 show cDNA sequences that encode the 26,000 and 17,000 forms of TNF, as well as muteins of these molecules.

The cDNA sequence that encodes the 26 kD TNF species is preferably obtained from the plasmid, pB11, described in commonly owned co-pending application, U.S. Serial No. 670,360, filed November 9,1984; and U.S. Patent Nos. 4,677,063 and 4,677,064. The plasmid pB11 contains the SV40 promoter in operable linkage to the TNF coding sequence, and thus is useful for expressing the 26 kD TNF species in eukaryotic host cells. Additionally, a second plasmid which pr ql c~rc~ers8 I 1 WO 94/00555 PCr/US93/06120 12contains the entire sequence which encodes the 26 kD TNF species is described in the forgoing U.S. patent application and patents. It is designated pE4. The plasmid pE4 is on deposit with the American Type Culture Collection, Accession No. 39894.

The cDNA sequence that encodes the 26 kD TNF species is present in the plasmid pB 11 as a PstI fragment. Thus, it is readily removed and inserted into any one of a number of suitable expression systems. The preferred expression system is the plasmid pFVXM, which is described in co-pending U.S. Serial No. 855,865, entitled Infective Drug Delivery System, inventor Kriegler, et l. (abandoned in favor of U.S. Serial No. 571,017, filed August 22, 1990). pFVXM is on deposit with the American Type Culture Collection and has Accession No. 67,103.

pFVXM is a retroviral vector that was derived from the plasmid pEVX described by Kriegler, et al., 1984, in Cell, 38:483. pEVX has a Moloney murine leukemia virus derived splice donor site 3' to the 5' long terminal repeat. It was previously shown that this splice donor sequence decreases the yield of correctly spliced translational templates of retroviral constructions. Thus, pEVX was engineered to remove the splice donor site, and replaced with an analogous Smal fragment of the Harvey murine sarcoma virus genome, which lacks the Moloney murine leukemia virus splice donor sequence. The resulting vector, pFVXM, lacks the Moloney murine leukemia virus spliced donor sequence and carries a viral packaging sequence. pFVXM has a convenient PstI site in which the DNA sequences that encodes the 26 kD TNF species can be inserted.

II.TNF Convertase TNF convertase activity arises from the proteolytic action of one or more enzymes. A variety of biological materials are available as sources of TNF convertase activity. These include tissues, cells, or extracts, or fluids associated therewith that are preferably, but not necessarily, of immunologic origin.

Moreover, established cell lines may also be utilized. Suitable sources would include human peripheral blood mononuclear cells, such as ieukocytes or cell lines I I sPL I I- WO 94/00555 IICT/US93/061 I2() 13 of leukocyte origin, preferably macrophages and monocytes. Neutrophils are a particularly useful sour-e- of TNFconvertase. Because of the ease of manipulating established cell lines, one preferred cell source of TNF convertase is Thus, the conversion of the 26 kD proTNF species to mature TNF can be affected by combining the 26 kD species with either intact HL60 cells, extracts derived therefrom, or media in which the HL60 cells were grown and thus contains TNF convertase activity. In some cell types, TNF convertase activity is present in the culture medium after the appropriate stimulation, which is discussed more below.

Further, because the TNF convertase activity is partially membrane-associated under certain conditions, it is possible to obtain a membrane fraction that may be utilized.

The procedures for isolating monocytes are well known in the art, as are other methods for culturing cell lines such as HL60. Briefly, monocytes may be prepared from peripheral blood by centrifugation first through Ficoll-hypaque and Percoll usirg standard procedures. This yields an enriched population of monocytes and lymphocytes, and the monocytes can be further enriched by plating the mixture of cells onto tissue culture dishes and incubating the cells for a time sufficient to permit the monocytes to adhere to the surface of the dishes. The lymphocytes are then washed off of the plates leaving primarily adherent monocytes. These cells may then be used as is, or can be stimulated to produce enhanced levels of TNF convertase using known monocyte activators, preferably lipopolysaccharide and phorbol myristate acetate. The cells may be fractionated, and either an extract or a membrane fraction prepared therefrom and employed in the assays described below.

We have isolated TNF convertase from 12 liters of HL60 culture by isolating the cell membrane fraction, solubilizing it in a 0.5% Nonidet P-40 detergent, subjecting the solution to anion exchange chromatography, cation exchange-HPLC, anion exchange-HPLC, and reverse-phase HPLC to yield 20 t of 1,000-fold purified TNF convertase, which is equivalent to -320 Units at an 18% yield.

The convertase was found to have a molecular weight of approximately 29-30 kD r~ ~18 I

I_

WO 94/00555 PCT/US93/06120 14by SDS-PAGE analysis (silver-stained). The convertase was sequenced, and the first amino acids were found to be identical, within experimental error, to the mature N-terminal sequence of a, known neutrophil proteinase, PR-3 (Campanelli et al, 1990, J. Exp. Med., 172:17091715). The purified convertase was shown S to cleave the 26 kD proTNF to the 17 kD mature form.

As described more fully below, the amino acid sequence for PR-3 has been elucidated, as predicted from the sequence of the cDNA clone is shown in Figure 2. PR-3 is known in the art as a protease having activities unrelated to TNF processing. It is classified as a human polymorphonuclear leukocyte serine proteinase that degrades elastin, fibronectin, laminin, vitronectin, and collagen type IV; see Rao et al., 1991, J. Biol. Chem.. 266:9540-9548. By SDS-PAGh.

analysis purified PR-3 has been reported to have a major band at 26.8 kD with two smaller bands having slightly larger molecular masses, possibly representing different glycosylated species, see Rao et al., supra. PR-3 is structurally similar to other serine proteases, such as elastase, cathepsin G, mouse granzyme B, rat mast cell protease II, human lymphocyte protease, and chymotrypsin; see Campanelli et al., 1990, J. Exp. Med., 172:1709-1715. PR-3 is inhibited by o< 2 -macroglobulin, phenylmethyl-sulfonyl fluoride (PMSF), and Oqantitrypsin.

Sequencing of the PR-3 digestion products of radiolabelled 26 kD TNF show that PR-3 prefers to cleave the proTNF to produce an N-terminal Val-Arg-Ser sequence (amino acids 1-3 of the 17 kD mature form) although cleavage may occur to produce an N-terminal Arg-Ser-Ser (amino acids 2-4) or Val-Ala-His (amino acids 1315) sequences. Rao et al., supra, reports that PR-3 prefers small aliphatic amino acids in the substrate cleavage site. Serine proteases such as elastase, cathepsin G, and plasmin do not efficiently convert the 26 kD proTNF to the 17 kD mature form.

PR-3 may also be isolated from neutrophils. Neutrophils are separated from human blood, then granules and membranes are isolated, and the mixture is fractionated on RP-HPLC, as described below.

III ,ys IJ__M I I I WO 94/00555 PCTI/US93/06120 15 As shown below, PR-3 is inhibited by peptide diphenyl phosphonate inhibitors, elastinal, and dichloro-isocumarin (DCI). The peptide diphenyl phosphonate inhibitors include Boc-Val-Pro-Val-p(OPh) 2 and Boc-Ala-Pro-Val-p(OPh) 2 Boc-Ala-Gln-Alap(OPh) 2 and Boc-Leu-Ala-Gln-Ala-p(OPh) 2 have also been tested and have much less inhibitory activity. "Boc" means tert-butyloxycarbonyl and "p(OPh) 2 represents the diphenyl phosphonate moiety, wherein the formula -COOH group is replaced with -P(=O)(O-phenyl)2. See Oleksyszyn t al., 1991, Biochem.. 30:485. It will be appreciated that other peptide diphenyl phosphonate molecules may inhibit PR-3.

Potential inhibitors may be constructed using the procedures shown in Oleksyszyn et al, supra, using small aliphatic peptides, for an example. Once the potential inhibitors are made, they may be tested in the assays shown below. Modeling studies predict that Boc-Val-Pro-His-p(OPh) 2 will be a potent PR-3 inhibitor.

III.Inhibitors of TNF convertase Activity-Prophylactics or Therapeutics of Sepsis Inhibitors of convertase activity will also be prophylactics or therapeutics that may be used in the treatment of sepsis and certain other diseases in which circulating TNF has been implicated, including rheumatoid arthritis and cachexia. Inhibitors of TNF convertase can be identified by procedures that enable one to measure the conve,'sion of proTNF to mature TNF. Several such assay procedures are descibed herein, and in Example 4 below. A suitable assay would consist of combining 26 kD proTNF, a TNF convertase, and a putative inhibitor. It will be understood by those skilled in the art that the inhibitory material may be added to the convertase before the convertase is added to TNF, or it can be added to TNF prior to, or immediately after adding the convertase. The order of addition may facilitate identification of inhibitors, but it is not determinative. If a substance has inhibitory activity, this can be revealed by electrophoretic analysis of the solution which will reveal, relative to control reaction, an increase in the amount of the 26 kD species, and concomitantly a decrease in mature TNF species. Applicants have also identified a colorimetric assay to detect convertase inhibitors. The assay l1I I i3 WO 94/00555 PC'/US93/06120 16is convenient and correlates with the autoradiographic assay for cleavage of 26 kD TNF. The colorimetic assay is described in detail in Example 4. Also see Kam et 1992, FEBS, 297(.12):119-123.

Other compoundr vith anti-convertase activity include anti-convertase antibody, either polyclonal or mosioclonal, or recombinant antibody. Preferably these antibodies will be humanized antibodies. Monoclonal antibody to the convertase may be produced using the general procedures described by Kohler, G. and Milstein, 1975, Nature, 256:495, which have been modified over the years as is known in the art. These initial studies involved fusing murine lymphocytes and drug selectable plasmacytomas to produce hybridomas. Subsequently, the technique has been applied to produce hybrid cell lines that secrete human monoclonal antibodies. The latter procedures are generally described in Abrams, 1986, Methods in Enzymology, 121:107, but other modifications are known to those skilled in the art. Regardless of whether murine or human antibody is produced, the antibody-secreting cells are combined with the fusion partner and the cells fused with a suitable fusing agent, preferably polyethylene glycol, and more preferably polyethylene glycol 1000. The latter is added to a cell pellet containing the antibody-secreting cells and the fusion partner in small amounts over a short period of time accompanied with gentle agitation. After the addition of the fusing agent, the cell mixture is washed to remove the fusing agent and any cellular debris, and the cell mixture consisting of fused and unfused cells seeded into appropriate cell culture chambers containing selective growth media. After a period of several weeks, hybrid cells are apparent, and may be identified as to antibody production and subcloned to ensure the availability of a stable hybrid cell line.

A preferred antibody is human monoclonal antibody which can be produced from lymphocytes sensitized with convertase either in vivo or in vitro and immortalized as antibody-producing hybrid cell lines, thereby making available a renewable source of the desired antibody. In vitro immunization techniques are well known in the art, while in vitro techniques are generally described by Luben, R. and L bl s _I I 1_ WO 94/00555 PCT/US93/06120 17- Mohler, 1980, Molecular Immunology, 17:635, Reading, C. Method in Enzymology, 121 (Part One):18, or Voss, 1986, Methods in Enzymology, 121:27. A number of in vitro immunization systems have been shown to be effective for sensitizing human B-cells. Reading, 1982, J. of Immun, Methods, 53:261.

It will be apparent to those skilled in the art, that in lieu of immunizing individuals directly with TNF convertase, lymphocytes may be isolated from individuals that are experiencing, or have experienced a bacteremic attack. For example, human patients having Wegener's granulomatosis are natural source of antiPR-3 antibodies and also contain human cells suitable for deriving human monoclonals.

A fraction of these lymphocytes will be sensitized to the convertase and may be used to produce permanent antibody-secreting hybrid cell lines. For example, immunocompromised human patients are generally susceptible to bacterial infections, particularly those suffering from various malignancies, extensive burs, etc., and lymphocytes isolated therefrom may be a source of antibody-secreting cells.

Sensitized lymphocytes can be immortalized by viral transformation. The preferred viral transformation technique for human lymphocytes involves the use of Epstein-Barr virus. The virus is capable of transforming human B-cells, and has been used to generate human monoclonal antibodies. Crawford, D. et al., 1983, J. of General Virology, 64:697; Kozbor, V. and Roder, 1983, L Immun. Today, 4:72.

Another procedure whereby sensitized lymphocytes may be immortalized consists of a combination of the above two techniques, dtat is viral transformation and cell fusion. The preferred combination consists of transforming antibody-secreting cel.'s with Epstein-Barr virus, and subsequently fusing the transformed cells to a suitable fusion partner. The fusion partner may be a mouse myeloma cell line, a heteromyeloma line, or a human myeloma line, or other immortalized cell line.

PCT No. 81/00957; Schlom e al., 1980, PNAS (USA), 77:6841; Croce i al., 1980, Nature, 288:488. The preferred fusion partner is a mouse-human ~Y WO 94/00555 PCT/US93/06120 18hetero-hybrid, and more preferred is the cell line designated F3B6. This cell line is on deposit with the American Type Culture Collection, Accession No. HB8785.

It was deposited April 18, 1985. The procedures for generating F3B6 are described in EPA No. 174,204.

Techniques applicable to the use of Epstein-Barr virus transformation and the production of immortal antibody-secreting cell lines are presented by Roder, J. et al., 1986, Methods in Enzymology, 121:140. Basically, the procedure consists of isolating Epstein-Barr virus from a suitable source, generally an infected cell line, and exposing the target antibody-secreting cells to supernatants containing the virus. The cells are washed and cultured in an appropriate cell culture n.dium.

Subsequently, virally transformed cells present in the cell culture can be identified by the presence of the Epstein-Barr viral nuclear antiger., and transformed antibody-secreting cells can be identified using standard methods known in the art.

It will be apparent to those skilled in the art that while a preferred embodiment of the instant invention is a neutralizing anti-TNF convertase monoclonal antibody, singly or in combination, that the antibody(s) may be altered and still maintain biological activity. Thus, encompassed within the scope of the invention is antibody modified by reduction to various size fragments, such as F(ab'h, Fab, Fv, or the like. Also, the hybrid cell lines that produce the antibody may be considered to be a source of the DNA that encodes the desired antibody, which may be isolated and transferred to cells by known genetic techniques to produce genetically engineered antibody. An example of the latter would be the production of single-chain antibody having the antibody combining site of the hybridomas described herein. Single-chain antibodies are described in U.S. Patent No.

4,704,692. A second example of genetically engineered antibody is recombinant, or chimeric antibody. Methods for producing recombinant antibody are shown in U.S. Patent No. 4,816,567, to Cabilly, t al.; Japanese Patent Application No.

84169370, filed August 15, 1984; U.S. Serial No. 644,473, filed August 27, 1984; British Patent Application No. 8422238, filed on September 3, 1984; Japanese Patent Application, No. 85239543, filed October 28, 1985; U.S. Serial a -sC--b WO 94/00555 PCI/US93/06120 19 No. 793,980 on November 1, 1985; U.S. Serial No. 77,528, filed July 24, 1987.

Also, British Patent Application No. 867679, filed March 27, 1986 describes methods for producing an altered antibody in which at least parts of the complementary determining regions (CDRs) in the light or heavy chain variable domains have been replaced by analogous parts of CDRs from an antibody of different specificity. Using the procedures described therein, it is feasible to construct recombinant antibody having the CDR region of one species grafted onto antibody from a second species that has its CDR region replaced. The preferred embodiment in this instance is a murine anti-convertase antibody CDR region that replaces the CDR region of human antibody.

In addition to antibodies, compounds that compete with 26 kD proTNF for binding to the convertase will inhibit or reduce the conversion of 26 kD proTNF to the mature form, and may thus be useful medicaments for treating sepsis and other diseases. One such class of reagents consists of peptides, polypeptides, or proteins, or other compounds synthetic, or naturally occurring, that have TNF convertase-binding activity similar to or better than the 26 kD proTNF. Preferred peptides or proteins are those that contain amino acid sequences similar to that found at the junction between the 76 amino acid leader sequence of proTNF and the 17 kD mature form. On such sequence is Leu-Ala-Gln-Ala-Val-Arg-Ser-Ser-Ser, where the second Ala is the residue present on the leader remaining in the membrane after the cleavage event, and Val is the N-terminal amino acid of the mature TNF. It is important to note that while the sequence is shown to consist of nine amino acids, that what is minimally intended is a peptide containing at least the dipeptide sequence Ala-Val that is recognized by the convertase.

An alternate embodiment of a peptide/protein convertase inhibitor is one that has the amino acid sequence, a sequence that is functionally similar to (SEQ ID NO: This peptide spans tv- TNF convertase cleavage sites, and thus would prevent the formation of the 17 kD mature TNF, among others. The first and dominant cleavage site is between alanine and valine at positions -1 and 1; and secondary 1- -b WO 94/00555 PCIT/US93/06120 sites are between valine and arginine at positions 1 and and proline and valine at positions +12 and +13. These positions correspond to the amino acid sequence shown in Figure 1.

A second class of competitive inhibitors consists of compounds including the sequence shown above, that is (SEQ ID NO: but wherein certain amino acids have been altered or deleted to yield a non-cleavable substrate. A preferred embodiment of this peptide is a 26 kD proTNF mutein produced by standard site-specific mutagenesis techniques. Most preferred is a mutein wherein the alanine or valine or both are substituted or deleted.

The peptides described above can be made by techniques well known in the art, such as, for example, the Merrifield solid-phase method described in Science, 232:341347 (1985). The procedure may use commercially available synthesizers such as a Biosearch 9500 automated peptide machine, with cleavage of the blocked amino acids being achieved with hydrogen fluoride, and the peptides purified by preparative HPLC using a Waters Delta Prep 3000 instrument, on a 15-20 _m Vydac C4 PrepPAK column.

The peptide diphenyl phosphonates described above are also used as inhibitors.

Useful peptides may be attached to Boc and the diphenyl phosphonate moiety (see Oleksyszyn et al., 1991, Biochem.. 30:485) and tested in a convertase inhibition assay. Preferred peptides are Boc-Val-Pro-Val-p(OPh) 2 Boc-Ala-Pro-Val-p(OPh) 2 and BocVal-Pro-His-p(OPh) 2 However, it will be seen that other peptide diphenyl phosphonates may be used in the inhibition assays described below to identify further TNF convertase inhibitors. Examples are disclosed below and are shown in Oleksyszyn et al, 1991, Biochem., 30:485.

The specificity of the identified TNF convertase PR-3 is similar to enzymes such as elastase, which typically cleave immediately following certain neutrally charged amino acids, such as between valine, proline, and alanine residues. Thus, in addition to the peptide inhibitors mentioned above, a variety of other inhibitors known to inhibit elastase may also generally inhibit an enzyme that cleaves the 26 III~ I~ ~R WO 94/00555 I'CI/ US93/06120 -21 kD proTNF. Those compounds that inhibit TNF convertase can be identified using the assays described below. A variety of elastase inhibitors are commercially available from suppliers such as Boehringer Mannheim Biochemicals, or are known in the art. Doherty, e al., 1986, Nature, 322:192; U.S. Patent Nos. 4,711,886; 4,797,396; 4,717,722; and 4,699,904. The preferred elastase inhibitors are modified cephalosporin antibiotics, such as those shown by Doherty, et al., above. More preferred is (1-((3-((acetyloxyl)-7-methoxy-8oxy-8-oxo-5-thio-l-azabicyclo oct-2-en- 2-yl) carbonyl) morpholine, S,Sdioxide, (6R-cis). Also, Stetler, t al., 1986, Nucleic Acids Research, 14:7883, describe a cDNA clone that codes for an inhibitor of neutrophil elastase. However, preferred inhibitors are those which inhibit PR-3 more effectively than elastase, since elastase activity may help ameliorate septic shock by, for example, degrading circulating TNF or releasing soluble TNF receptors which, in turn, inhibit circulating TNF. (See Scuderi, 1991, Cellular Immunology, 135:299-313).

Additionally, inhibitors may be found by modeling the crystal structure for PR3 by adapting the known structure for the closely homologous elastase molecule.

Computer models known in the art may be constructed to establish important contact points in the substrate-binding site of PR-3. Potential inhibitors may be designed based on this information and then tested in the present assay systems, as well as in relevant animal models for septic shock.

Recombinant techniques may be used to obtain the inhibitors, the proTNFs, mature TNFs or TNF convertases described herein. Most of the recombinant techniques that are described herein that may be used to transform cells, fabric"te vectors, extract messenger RNA, and the like are widely practiced in biotechnology and most practitioners are familiar with the standard materials and methods employed. However, for convenience, the following paragraphs are offered as a guideline.

A.General Cloning Techniques I I I-,,-sLLI~I L~ ~JI-e ~p s I WO 94/00555 PCT/US93/06120 -22- Construction of suitable vectors containing the desired TNF coding sequence employs standard ligation and restriction techniques which are well understood in the art. Isolated vectors, DNA sequences, or synthesized oligonucleotides are cleaved, tailored, and religated in the form desired.

Site-specific DNA cleavage is performed by treating with suitable restriction enzyme(s) under conditions which are generally understood in the art, and the particulars of which are specified by the manufacturer of these commercially available restriction enzymes. See, New England Biolabs, Product Catalog.

In general, about 1 1 /g of plasmid or DNA sequence is cleaved by one unit of enzyme in about 20p.1 of buffer solution. In the examples herein, typically, an excess of restriction enzyme is used to ensure complete digestion of the DNA substrate. Incubation times of about 1-2 hours at about 37 C are workable, although variations can be tolerated. After each incubation, protein is removed by extraction with phenol/chloroform, and may be tollowed by ether extraction, and the nucleic acid recovered from aqueous fractions by precipitation with ethanol followed by chromatography using a Sephadex G-50 spin column. If desired, size separation of the cleaved fragments may be performed by polyacrylamide gel or agarose gel electrophoresis using standard techniques. A general description of size separations is found in Methods in Enzvmology, 1980, 65:499560.

Restriction cleaved fragments may be blunt-ended by treating with the large fragment of E. coli DNA polymerase I, that is, the Klenow fragment, in the presence of the four deoxynucleotide triphosphates (dNTPs) using incubation times of about 15-25 minutes at 20-25 °C in 50 mM Tris pH 7.6, 50 mM NaCl, 6 mM MgCI 2 6 mM DTT and 10 mM dNTPs. After treatment with Klenow, the mixture is extracted with phenol/chloroform and ethanol precipitated. Treatment under appropriate conditions with S nuclease results in hydrolysis of single-stranded portions.

Ligations are performed in 15-30.1 volumes under the following standard conditions and temperatures: 20 mM Tris-Cl pH 7.5, 10 mM MgCl, 10 mM @9 DTT, 33Ag/ml BSA, 10 mM-50 mM NaCI, and 1 mM ATP, 0.3-0.6 (Weiss) I B" l WO 94/00555 PCT/US93/06120 -23units T4 DNA ligase at 4°C for "sticky end" ligation, or for "blunt-end" ligations.

Intermolecular "sticky end" ligations are usually performed at 33-100p-g/ml total DNA concentration. In blunt-end ligations, the total DNA concentration of the ends is about lp-M.

In vector construction employing "vector fragments," the vector fragment is commonly treated with bacterial alkaline phosphatase (BAP) in order to remove the 5' phosphate and prevent religation of the vector. BAP digestions are conducted at pH in approximately 150 mM Tris, in the presence of Na and Mg 2 using about 1 Unit of BAP per tg of vector at 60 0 C for about 1 hour.

Nucleic acid fragments are recovered by extracting the preparation with phenol/chloroform, followed by ethanol precipitation. Alternatively, religation can be prevented in vectors which have been double-digested by additional restriction enzyme digestion of the unwanted fragments.

In the constructions set forth below, correct ligations are confirmed by first transforming the appropriate E. coli strain with the ligation mixture. Successful transformants are selected by resistance to ampicillin, tetracycline or other antibiotics, or using other markers depending on the mode of plasmid construction, as is understood in the art. Miniprep DNA can be prepared from the transformants by the method of Ish-Howowicz et al., 1981, Nucleic Acids Res., 2:2989, and analyzed by restriction and/or sequenced by the dideoxy method of Sanger et al., 1977, PNAS (USA, 74:5463 as further described by Messing et al., 1981, Nucleic Acids Res., 9:309, or by the method of Maxam et al., 1980 Methods in Enzymology, 65:499.

Host strains used in cloning in M13 consist of E. coli strains susceptible to phage infection, such as E. coli K12 strain DG98. The DG98 strain has been deposited with ATCC July 13, 1984 and has Accession No. 1965.

Depending on the host cell used, transformation is done using standard techniques appropriate to such cells. Calcium treatment employing calcium chloride, as described by Cohen, 1972, PNAS (USA) 69:2110, or the RbC12 method described ,0 by Maniatis eg al., 1984, Molecular Cloning: A Laboratory Manual, Cold Spring p I 31-LCI WO 94/00555 PCT/US93/06120 -24- Harbor Press, p. 254, may be used for procaryotes. Transfection may also achieved using a modification of the calcium phosphate precipitation technique of Graham t al., 1973, Virology, 52:456 or Wigler t al., 1978, Cel, 14:725.

B.Oligonucleotide Probes Synthetic oligonucleotides are prepared by the triester method of Matteucci et al., 1981, J. Am Chem. Soc., 103:3185 or using commercially available automated oligonucleotide synthesizers. Kinasing of single strands prior to annealing or for labeling is achieved using an excess, approximately 10 Units of polynucleot .ie kinase to 0.1 nmole substrate in the presence of 50 mM Tris, pH 7.6, 10 mM MgCI 2 5 mM dithiothreitol, 1-2 mM ATP, 1.7 pmoles P-ATP (2.9mQ /mmole), 0.1 mM spermidine, 0.1 mM EDTA.

C.Mutagenesis Mutagenesis can be carried out using any number of procedures known in the art.

These techniques are described by Smith, 19S5, Annual Review of Genetics, 19:423, and modifications of some of the techniques are described in Methods in Enzvmology, 154. part E, (eds.) Wu and Grossman (1987), chapters 17, 18, 19, and 20. The preferred procedure is a modification of the gapped-duplex site-directed mutagenesis method. The general procedure is described by Kramer et al., in chapter 17 of the Methods in Enzymology, above.

Conventional M13 mutagenesis methods involve annealing a short synthetic oligonucleotide to single stranded M13 DNA having a cloned target coding sequence that is sought to be mutagenized. The oligonucleotide is almost, but not entirely complementary to the target sequence and has at least one mispaired nucleotide. After the annealing reaction, the remaining portion of the single stranded DNA must be filled in to give heteroduplex DNA that can be transfected into a suitable host cell which allows for the expression of the mutation. In the gapped-duplex method, a partial DNA duplex is constructed that has only the 4 target region exposed, unlike the conventional methods which have the target /VT 0 _~LII -II I CI 3 WO 94/00555 PCT/US93/06120 region and the rest of the single-stranded M13 DNA exposed. Like the conventional methods, a short oligonucleotide is annealed to the target region, and extended and ligated to produce a heteroduplex. However, because only a small portion of single-stranded DNA is available for hybridization in the gappedduplex method, the oligonucleotide does not anneal to undesired sites within the M13 genome. Further, this method has the additional advantage of introducing fewer errors during the formation of the heteroduplex since only a very small region of DNA on either side of the target region has to be filled in.

More specifically, the gapped-duplex method involves cloning the target DNA sequence into an appropriate M13 phage that carries selectable markers, such as for example the stop codon amber mutation. The latter allows for negative selection in a host cell that cannot suppress the effects of the mutation. Preferably the phage is M13mp9 which contains two amber codons in critical phage genes.

Thus, the sequence that encodes 26 kD TNF is cloned into M13mp9 amber+, and single-stranded DNA is prepared therefrom using standard techniques. Next, double-stranded replicative form DNA from M13 GAP, a genetically engineered M13 derivative that lacks the amber codons is cleaved with HincII restriction enzyme. The base sequen, f M13 GAP is similar to M13mpl8, which lacks both the amber codons and the sequence between base pairs 6172 and 6323. This deletion flanks the multiple cloning sites of the M13mp series and generates a unique HinclI site. Gapped-duplex DNA is formed, using standard DNA/DNA hybridization techniques, consisting of singlestranded DIA having the amber codons, and a second strand of DNA from HinclI digested M13 GAP lacking both the amber codons and the TNF coding sequences. Thus, the only portion of the gapped-duplex that is exposed is the 26 kD TNF target sequence. The desired oligonucleotide is annealed to the gapped-duplex DNA and any remaining gaps filled in with DNA polymerase and the nicks sealed with DNA ligase to produce a heteroduplex. The latter is transfected, preferably into a mismatch repair deficient host, and mixed phage produced. From the mixed phage population, phage carrying unmutated 26 kD TNF DNA, which also have the amber mutations, can

I

"C I C II WO 94/00555 PCT/US93/06120 -26be selected against by infecting the mixed phage population into a host cell that cannot suppress the amber mutation. Clones can then be screened for phage that carry the desired TNF mutation.

L, ~Idt II__ I WO 94/00555 PCT/US93/06120 -27- IV.Methods of Use of TNF Convertase Inhibitors Compounds identified as having TNF convertase-inhibitory activity will also have prophylactic or therapeutic applications in the treatment of sepsis. Because o \e onset of sepsis is associated with an increase in circulating mature TNF, these inhibitors may be used prophylacticall3 in those instances where there is a risk of bacterial infection, partVfularly in a pre-operative setting. Similarly, when there is an early diagnosis of sepsis, the inhibitors will have beneficial therapeutic effects in substantially reducing the amount of the soluble, 17 kD form of TNF that is produced.

Increases in circulating mature TNF are also associated with the diseases rheumatoid arthritis, cachexia, cerebral malaria and graft-versus-host disease.

Thus, the inhibitors of this invention will also have useful prophylactic or therapeutic applications in the treatment of these diseases.

Another medical application for inhibitors of convertase is for the treatment of AIDS. It has been shown that TNF causes the activation of latent human immunodeficiency virus. Folks et al., 1989, PNAS (USA). 86:2365. Thus, preventing or inhibiting the formation of mature T'hF, by inhibition of TNF convertase would be a valuable treatment for AIDS, and would preferably be used to treat patients that are infected with the virus that is in a latent phase.

The inhibitors of this invention may be administered at concentrations that are therapeutically effective for prevention of sepsis, AIDS, etc. To accomplish these goals, the peptides and chemical compounds are administered parenterally via intravascular [intraarterial or intravenous], intramuscular, or subcataneous routes). Methods to accomplish this administration are known to those of ordinary skill in the art.

Before administration to patients, formulants or pharmaceutically acceptable excipients may be added to the peptides ano chemical compounds. A liquid formulation is preferred. For example, these formulants may include oils, polymers, vitamins, carbohydrates, amino acids, buffers, albumin, surfactants, or bulking agents. Preferably carbohydrates include sugar or sugar alcohols such as ~~ll1-~9 q ~Bd E WO 94/00555 PCT/US93/06120 28 mono-, di-, or polysaccharides, or water soluble glucans. The saccharides or glucans can include fructose, dextrose, lactose, glucose, mannose, sorbose, xylose, maltose, sucrose, dextran, pullulan, dextrin, alpha and beta cyclodextrin, soluble starch, hydroxethyl starch and carboxymethylcelloluose, or mixtures thereof.

Sugar alcohol is defined as a C 4 to Cg hydrocarbon having an -OH group and includes galactitol, inositol, mannitol, xylitol, sorbitol, glycerol, and arabitol.

Mannitol is most preferred. These sugars or sugar alcohols mentioned above may be used individually or in combination. There is no fixed limit to amount used as long as the sugar or sugar alcohol is soluble in the aqueous preparation.

Preferably, the sugar or sugar alcohol concentration is between 1.0 w/v% and more preferable between 2.0 and 6.0 Preferably amino acids include levorotary forms of carnitine, arginine, and betaine; however, other amino acids may be added. Preferred polymers include polyvinylpyrrolidone (PVP) with an average molecular weight between 2,000 and 3,000, or polyethylene glycol (PEG) with an average molecular weight between 3,000 and 5,000. It is also preferred to use a buffer in the composition to minimize pH changes in the solution before lyophilization or after reconstitution. Most any physiological buffer may be used, but citrate, phosphate, succinate, and glutamate buffers or mixtures thereof are preferred. Most preferred is a citrate buffer.

Preferably, the concentration is from 0.01 to 0.3 molar. Surfactants that can be added to the formulation are shown in EP Nos. 270,799 and 268,110.

Additionally, the present peptides and chemical compounds can be chemically modified by covalent conjugation to a polymer to increase their circulating half-life, for example. Preferred polymers, and methods to attach them to peptides, are shown in U.S. Patent Nos. 4,766,106, 4,179,337, 4,495,285, and 4,609,546 which are all hereby incorporated by reference in their entireties.

Preferred polymers are polyoxyethylated polyols and polyethylene glycol (PEG).

PEG is soluble in wter at room temperature and has the general formula:

R(O-CH

2

-CH

2 where R can be hydrogen, or a protective group such as an alkyl or alkanol group. Preferably, the protective group has between 1 and 8 i I L-P~ 9 ~1 rat 1 WO 94/00555 PC/US93/06120 -29carbons, more preferably it is methyl. The symbol n is a positive integer, preferably between 1 and 1,000, more preferably between 2 and 500. The PEG has a preferred average molecular weight between 1000 and 40,000, more preferably between 2000 and 20,000, most preferably between 3,000 and 12,000.

Preferably, PEG has at least one hydroxy group, more preferably it is a terminal hydroxy group. It is this hydroxy group which is preferably activated to react with a free amino group on the inhibitor.

Water-soluble polyoxyethylated polyols are also useful in the present invention.

They include polyoxyethylated sorbitol, polyoxyethylated glucose, polyoxyethylated glycerol (POG), etc. POG is preferred. One reason is because the glycerol backbone of polyoxyethylated glycerol is the same backbone occurring naturally in, for example, animals and humans in mono-, di-, triglycerides.

Therefore, this branching would not necessarily be seen as a foreign agent in the body. The POG has a preferred molecular weight in the same range as PEG. The structure for POG is shown in Knauf et al., 1988, 1. Bio. Chem.

263:15064-15070, and a discussion of POG conjugates is found in U.S. Patent No.

4,766,106, both of which are hereby incorporated by reference in their entireties.

After the liquid pharmaceutical composition is prepared, it is preferably lyophilized to prevent degradation and to preserve sterility. Methods for lyophilizing liquid compositions are known to those of ordinary skill in the art.

Just prior to use, the composition may be reconstituted with a sterile diluent (Ringer's solution or sterile saline, for example) which may include additional ingredients. Upon reconstitution, the composition is preferably administered to subjects using those methods that are known to those skilled in the art.

Insoluble inhibitors can be formulated by combination with one or more solubilizers. Preferred solubilizers include: ethanol; oils, such as corn oil; PEG; propylene glycol; and non-ionic surfactants. Preferred co-solvents have a molecular weight between 50 and 1,000, more preferably between 100 and 600.

Preferably their concentration is between 1 and 75% w/w, more preferably between 10 and 50%. The concentration of ethanol is preferably between 0.1% ~a I ~1 br WO 94/00555 PCT/US93/06120 and 20%, more preferably between 1 and Preferred non-ionic surfactants have a hydrophile-lipophile balance between 14 and 40, more preferably between and 20, most preferably between 17 and 19. Preferably, the non-ionic surfactants have a molecular weight in the range between 100 and 250,000, more preferably between 4,000 and 200,000, most preferably between 6,000 and 150,000. Preferably, the non-ionic surfactants are effective in the concentration range of 0.005% to 10% w/v, more preferably in the range of 0.01 to 5% w/v, most preferably in the range of 5 to 2.5% w/v. Preferably, the non-ionic surfactants include those commonly used in the pharmaceutical, food, and cosmetic industries. Preferred non-ionic surfactants include: polyoxyethylene sorbitan fatty acid esters Tweens), polyethylene glycol esters, polyethylene fatty acid esters, block copolymers of ethylene oxide and propylene oxide Pluronics), ethylated fatty alcohol ethers laureth-12), octylphenoxy polyethyoxy ethanol compounds Tritons), and polyoxyethylated castor oil Cremophor).

These non-ionic surfactants can be produced by means known in the art or purchased from commercial suppliers.

Other non-ionic surfactants can be determined by using the present screening method. In this method a non-ionic surfactant is added to an insoluble inhibitor.

The resulting solution is mixed or homogenized and allowed to stand for 24 hours at room temperature. If the inhibitor remains in solution, as measured by RP-HPLC, GC, or visual or spectrophotometric clarity, then the surfactant is useful to solubilize the inhibitor.

Having generally described what the applicants believe their invention to be, presented below are examples that are illustrative of the scope of the invention. It will be appreciated by those skilled in the art that the examples are not intended to be construed as limiting the invention to the materials and methods shown as there are numerous substitutions that can be made therein without departing from the scope of the invention.

Example 1 Isolation and Identification of a TNF Convertase IB I c,, U J I I WO 94/00555 PCT/US93/06120 31 cells were obtained from the American Type Culture Collection (Rockville, MD) and grown in T-175 flasks containing RPMI 1640 medium supplemented with 20% fetal bovine serum (GIBCO) and L-glutamine. Batches totalling 3 liters of HL60 cells were grown to confluency and harvested. The cells were resuspended in approximately 120 ml of a hytonic buffer and lysed by nitrogen cavitation (400 psi, 30 minutes at 4 0 The homogenate was centrifuged at 10,000 x g for 10 minutes, and both the supernatant and the cell debris pellet were stored at -20 C.

cell debris from 3 batches of HL60 cell cultures were thawed in 250 ml of 10 mM Tris pH 8.5 containing 0.5% NP-40, 5 mM EDTA, and 2 g/ml leupeptin (DEAE buffer) and dialyzed for 4 hours in the same buffer. The protease inhibitors used during purification were shown to have no effect on the convertase activity detected in HL60 lysates. Particulates were removed by centrifugation (10,000 x g, 10 minutes) and the sample fractionated by anion exchange chromatography on a DEAESepharose column (2.6 x 21 cm, Pharmacia) eluted with a 680 ml, NaCI gradient from 0-0.8 M. Fractions containing TNF convertase activity were identified throughout the purification using the 3S-proTNF convertase assay. Pooled DEAE fractions were dialyzed into 20 mM sodium phosphate buffer, pH 6.5, containing 01% NP-40, 1 mM EDTA, and 1 p.g/ml leupeptin, divided into three equal portions and each subjected to cation exchange HPLC on a TSK-SP-5PW column (7.5 x 75 mm, BioRad) eluted with a sodium chloride gradient from 0-0.6 M. Fractions enriched in convertase activity were pooled and dialyzed into DEAE buffer containing 0.1% The pooled material from the SP column was divided into three portions, and each was subjected to anion exchange HPLC on a TSK-DEAE-5PW column x 75 mm, BioRad) eluted with a 45-minute, sodium chloride gradient from 0-0.6 M. The pool of convertase activity was further purified by RP-HPLC on a Vydac C4 column using an acetonitrile/0.1 TFA mobile phase.

This treatment provided a 1,000-fold purification, resulting in 20 p. of convertase (approximately 320 Units) at an 18% yield. Fractions from the RP-HPLC were I p ~PR ~pM II~ WO 94/00555 PCI'/US93/06120 32 tested for convertase activity and sized on SDS-PAGE. The fraction that contained convertase activity contained a protein having a molecular mass of approximately 28-31 kD. The convertase was sequenced, and an 18-amino-acid sequence at the N-terminus proved to be identical to that of the serine protease PR-3. PR-3 was subsequently isolated from human neutrophils, essentially using published procedures, and it was found to have the same activity as TNF convertase in the 35 SproTNF assay.

The identification of PR-3 as a TNF convertase was further strengthened by Nterminal sequencing of cyanogen bromide cleavage fragments, as well as amino acid composition of PR-3 both of which agreed (within experimental error) with the published amino acid sequence of mature, active PR-3 (Campanelli et a, 1990, J. Exp. Med.,, 172:1709-1715).

Example 2 Cloning and Recombinant Expression of Human PR-3 RNA was purified from HL60 cells and a cDNA library was constructed in the plasmid pGEM. Construction of the cDNA used C tailing of cDNA and G tailing of the vector, then ligation into the plasmid (Gene Transfer and Expression, 1990, pgs 114135). Clones were screened using a unique oligonucleotide probe derived from the known sequence of myeloblastin (Bories et 1989, Cell, 59:959-968).

Sequencing of one clone MY17 was performed using plasmid double-strand sequencing and the Sequenase kit and an automated ABI sequencer. Sequence for MY17 is shown in Figure 2. Novel features for the sequence include 5 nucleotide differences from the original publication by Bories et aL., 1989, Cell, 5:959-968, and three nucleotide differences from the Campanelli .t aL, 1990, J. Exp. Med., 172:17091715. Additional 5' sequence and an additional 5' methionine coding sequence was found. The two carboxyl terminal amino acids in PR-3, arginine and proline, are similar to that of Bories eg aL, supr, but differ from the glycine and proline sequence from Campanelli eA aL, supra.

rrrm~--- IIII C1IR IL WO 94/00555 PCI'/ S93/06120 33 Transient mammalian expression of PR-3 was performed by cloning the 1.0 Kb HinIII-EcoRI PR-3 fragment from MY17 into the PstI site of SR-ocvector. COS cells were transiently transfected using the DEAE/Dextran method. Kriegler, 1990, Gene Transfer and Expression, pp. 99-100, Stockton Press. Transient expression revealed low levels of PR-3 expression by Western blot analysis. PR-3 was mutagenized to optimize its expression in mammalian, bacterial and insect expression systems. The PR3 gene in the pGEM vector was mutagenized using oligonucleotide directed mutagenesis. Two constructs were made; A) delta zymogen PR-3 and B) delta signal peptide PR-3.

A)Delta zymogen PR-3 was made using an oligonucleotide that deletes the codons for amino acids at position -1 and -2 (glutamic acid and alanine, respectively).

This gene can be removed from pGEM by EcoRI digestion, and the gene transferred to SR-oLfor transient mammalian expression and pcDNA I for production of stable transfectants.

B)Delta signal PR-3 was made using an oligonucleotide that deleted the leader and added an ATG prior to the position 1 isoleucine of the mature protein. This gene can be removed from pGEM by EcoRI digestion, and transferred to SR-a and pcDNA I for transient and stable mammalian expression. In addition, this construct was placed in DG160 a\,Pl based bacterial expression vector at 8-12 nucleotides from the Shine-Dalgarno ribosomal binding site.

Another construct, the cecropin B PR-3 construct, was made so that the insect leader for cecropin B was placed before the position 1 isoleucine of the mature PR-3 protein. This was placed in the insect vector, pAcC13.

D)For optimization of bacterial expression, mutagenesis of the third nucleotide from a purine to pyrimidine in the codons for the first 2-8 amino acids of delta signal PR-3 was performed using overlapping synthetic oligonucleotides and polymerase chain reaction amplification of the synthetic fragment. This fragment will be cloned into the 5' smal site of PR-3, to decrease the GC content of the RNA and facilitate expression.

i .Example 3 m 0 1 ~1Le-~l pi I C-I ~s WO 94/00555 PCP/US93/06120 -34- Conversion of 26 kD proTNF to Mature TNI The vector pFVXM, on deposit with the American Type Culture Collection, Accession No. 67,103, was used to produce a vector pFVXM-TNF6, which contains the DNA sequence that encodes the 26 kD TNF species. To produce the latter vector, the plasmid B11 which contains the cDNA sequence that encodes the 26 kD TNF species was treated with PstI, which excises the coding sequence.

The fragment was purified using standard electrophoretic techniques. Next, the vector pFVXM was treated with PstI, and the PstI fragment from pBl1 containing the 26 kD coding sequence was inserted into the polylinker region of the vector using standard techniques, as described above, to produce pFVX-TNF6.

pFVX-TNF6 was used to produce the cell line TNF 6.8, as described by Kriegler et aL., 1988, above, or as described in U.S. Serial No. 395,254, entitled "Cleavage Site Blocking Antibody to Prohormone Proteins and Uses Thereof," filed August 16, 1989.

TNF 6.8 expresses both 26 kD and 17 kD TNF. Figure 3 shows the conversion of 26 kD TNF by convertase activity present in HL60 cells. The production of labelled 26 kD TNF by in vitro transcription/translation, and analysis by gel electrophoresis is described below in Example 4. Note that the S-1 cytosol or pellet fractions cause the near complete conversion of 26 kD TNF to a 17 kD species. Figure 3 also shows, for comparative purposes, 26 kD and 17 kD TNF in a lysate of TNF 6.8 cells.

pFVXM and the plasmid pBl 1 were both amplified in E. coli strain HB101.

Ligation of the fragments was carried out using standard conditions. Plasmid DNA was isolated after the ligation procedure and the correct orientation of the TNF encoding sequences was established by restriction analysis.

Plasmid DNA was prepared according to the procedure of Birnboim and Doly, as described in Nucleic Acid Research, 2:1513 (1979). The plasmid DNA was banded twice in cesium chloride, and exhaustively dialyzed against TE buffer consisting of 10 mM Tris, pH 8.0, and 1 mM EDTA.

s1~1_1 I_ L~~ P- r I C I WO 94/00555 PC'/US93/06120 35 Example 4 TNF Convertase Assays A.Jn Vitro Transcription/Translation Assay A preferred assay procedure consists of in vitro transcription/translation to produce the 26 kD molecule, followed by treatment with convertase in the presence or absence of compounds being tested for convertase inhibitory activity. The procedure entails in vitro transcription/translation of the TNF cDNA present in the plasmid B11. Thus, the sequence was removed from pBll by P.tI digestion and was inserted into the PstI site of pGEM-3 (obtainable from Promega Biotec). The resulting plasmid, termed pGEM-TNF14, was amplified in E. coli using established techniques, and plasmid DNA was prepared according to the procedure of Birboim and Doly, described above. Plasmid DNA was in vitro transcribed by linearizing it with HindIII, and the linearized plasmid templates used to prepare capped transcripts with T7 RNA polymerase and an in vitro transcription kit supplied by Promega Biotec. Transcription was performed using standard techniques as suggested by the manufacturer's instructions.

The mRNA was translated in v':o in the presence of "S-cysteine to produce 26 kD TNF. A rabbit reticulocyte lysate translation kit was used, also supplied by Promega Biotec, and the conditions recommended by the manufacturer were followed.

35 S-cysteine-labelled 26 kD TNF was used to assay for convertase inhibitors as follows. 25 .1 of in vitro translated material was combined with 250 ~l of convertase activity partially purified from uninduced HL60 cells, plus compounds to be assayed for inhibitory activity. The convertase was produced by harvesting 2 x 109 HL60 cells, and isolating S-1 and P-30 fractions totalling 18 and 6 ml, respectively. 2501 of the P-30 fraction was used, although the S-1 fraction may also be used. The assay was carried out at 300C for 1 hour, essentially as described above. Next, the reaction mixture was immunoprecipitated with anti-TNF polyclonal antisera and protein A Sepharose, pelleted and washed. The A0 bound protein was eluted and electrophoresed using SDS-PAGE. The gel was 111 L~_s I I r sr e I WO 94/00555 PCI'/US93/06120 -36fixed in 40% methanol, 10% acetic acid, soaked in Enlightening (Dupont), dried, and exposed to x-ray film which was subsequently developed. The gel electrophoretic profiles of 26 kD TNF treated with HL60 convertase and varying dilutions of the potential inhibitory compound, revealed those compounds with inhibitory activity.

Using the above assay, it was determined that 3,4-dichloro-isocoumarin and elastinal at concentrations of 100 g/ml and 5 mg/ml, respectively, inhibit the convertase. It was also shown that (1.-((3-((acetyloxyl)-7-methoxy-8-oxy-8-oxo-5-thio 1-azabicyclo oct-2-en- 2-yl) carbonyl) morpholine, S,S-dioxide, (6R-cis) at a concentration of 1 mM inhibits convertase activity. These results are shown in Figure 4.

The above assay was also used with pure HL60 cell PR-3 to test a variety of proteinase inhibitors for TNF convertase inhibitory activity, as shown in Figure Pure PR-3 (0.3 p.g/ml) was preincubated for 30 minutes with the following inhibitors prior to addition of "S-labelled 26 kD-TNF and assaying as described above: DCI (45 t a-2-macroglobulin (1 mg/ml), PMSF (2C PM), leupeptin (2 fg/ml), EDTA (10 mM), or pepstatin 2 g/ml). The first three of these inhibitors showed significant inhibitory activity.

I~ eP %A~e~YIJ~Pg- ~11111~- P~pl I s~llO gl WO 94/00555 W US93/06 120 37 B.Monocyte Assay The 26 kD form of TNF can also be produced by stimulated monocytes which produce 26 kD TNF, as described by Kriegler, et al., 1988, £CIJ Briefly, human monocytes are purified from human blood by, ;ation, and subsequently enriched based on the adherence of monocytes to Julture dishes.

Centrifugation consists of purifying the monocytes through Ficoll-hypaque and percoll obtainable from Pharmacia. The manufacturers recommended procedures followed. Next, the mixture of cells resulting from the centrifugation step, consisting of monocytes and lymphocytes, are plated onto tissue culture dishes containing RPMI media supplemented with 20% fetal calf serum. The dishes are incubated for 30 minutes at 37 C after which they are extensively rinsed with the same media. This treatment removes non-adherent lymphocytes and leaves only adherent monocytes.

Monocyte 26 kD TNF is radiolabelled as follows. The monocytes are incubated for 3 hours at 370C in RPMI media supplemented with 20% fetal calf serum, 100 ng/ml lipopolysaccharide, and 10 Ag/ml phorbol myristate acetate for 30 minutes at 37OC. The latter two compounds induce the expression of TNF. The RPMI media is cysteineminus, and the fetal calf serum present at a final concentration of The serum is dialyzed prior to use to remove any cysteine present. After the 30-minute incubation period, 100 uCi 3 sS-cysteine is added, and the cells are radiolabelled for 3 hours at 370C, after which they are lysed and used to assay for convertase activity. The steps for carrying out the assay, as well as identifying inhibitors of the convertase, are similar to those described above.

C.Colorimetric Assay for Convertase Inhibition TNF convertase inhibition can also be measured by a colorimetric assay. In this type of assay, the actual conversion of proTNF to mature TNF is measured indirectly using a colorimetric TNF convertase substrate. By "colorimetric TNF convertase substrate" is meant a compound that is cleaved by a TNF convertase to J0 release a compound that absorbs light of a certain frequency. One such substrate ICI-l I= P- I1~Y l~yl~yyl _I I -O s- WO 94/00555 W 4CIOI 40iS93/06120 38is Boc -AlaONp (Bachem Bioscience, Inc., Philadelphia, PA). Other such substrates can be discerned from the structure of TNF convertase and other serine proteases. Although the example herein uses purified native PR-3 as the TNF convertase, it is contemplated that recombinant PR-3 or other TNF convertases can be used in this assay as well.

Peptide diphenyl phosphonate inhibitors were synthesized and stored as lyophilized solids as described in Oleksyszyn and Powers, 1991, Biochem., 30:485493.

Inhibitor solutions (10 mg/ml) were prepared in 100% dimethyl sulfoxide (DMSO) and diluted into aqueous buffers upon initiation of the experiments. 3,4, dichloro-isocumarin was purchased from CalBiochem. Purified PR-3 (10 A 1, 0.1 p.g/ml) was mixed with varying concentrations of protease inhibitor (400 ml final volume) in 20 mM sodium phosphate buffer, pH 7.0, containing 0.1 M sodium chloride. Aliquots (40 I-1) were removed at selected times and diluted 1/10 into a colorimetric assay for convertase, containing 0.5-1 mM BOC-Ala-ONp (prepared fresh from a 50 mM stock in 100% methanol) in 0.02 M sodium phosphate buffer, pH 7.0, 0.1 M sodium chloride. The increase in absorbance was monitored at 347 nm on a Hewlett Packard 8450A spectrophotometer, and using an extinction co-efficient of 5.5 x 10' M-'cm'.

Example Peptide Diphenyl phosphonate Inhibitors of TNF Convertase Several peptide diphenyl phosphonates were tested for inhibitory activity: BocVal-Pro-Val-p(OPh) 2 (VPV), Boc-Ala-Pro-Val-p(OPh) 2

(APV),

Boc-Ala-Gln-Alap(OPh) 2 (AQA), and Boc-Leu-Ala-Gln-Ala-p(OPh) 2

(LAQA).

The peptides were prepared by chemical synthesis using the Merrifield method and the diphenyl phosphonates were prepared according to the method similar to the one shown in Oleksyszyn et al,, supra.

The peptide diphenyl phosphonates were tested in the colorimetric assay described in Example 3, for inhibition of TNF convertase/PR-3 activity. The results are shown in Figure 6. VPV and APV demonstrated inhibitory activity. AQA and I C a I III L I I WO 94/00555 I'Cl'/ US93/6 6120 39 LAQA showed marginal, if any inhibition at the concentrations tested. Dichloro isocoumarin (DCI) showed 100% inhibition in the assay.

Example 6 TNF Mutein/Antibody/Peptide Inhibitors of Convertase Activity The following compounds will have convertase inhibitory activity and can be prepared as follows. These compounds may be tested for inhibitory activity as described in Exampz 4 above.

A.Anti-Convertase Antibody Monoclonal or polyclonal antibody is prepared that binds to the convertase and thereby sterically prevents the convertase from binding to 26 kD TNF or otherwise neutralizes the enzymatic activity of the convertase. The procedure consists of immunizing an appropriate host animal with a membranous fraction of HL60 cells producing TNF convertase. Alternatively, purified TNF convertase may be used from native or recombinant sources. For example, PR-3 from human neutrophils may elicit anti-TNF convertase antibodies. A sufficient amount of material should be used to elicit an immune response, and usually this will consist of between pg to 10 mg per kilogram of body weight. Immunization may be conducted with adjuvant in a biologically acceptable buffer, as is known in the art. The best immunization route can be determined experimentally, and the primary immunization may be followed by one or more secondary immunizations depending on the strength of the immune response to the initial immunization.

The presence of neutralizing anti-convertase antibody in the sera may be detected using the convertase assay described above wherein antisera is present in the assay mixture. Inhibition of the conversion of the 26 kD TNF species to species having the molecular weight of mature TNF indicates the presence of a neutralizing antibody. It is, of course, assumed that the proper controls are conducted to insure that anti-sera from non-immunized animals is not inhibitory. Polyclonal 4 30 antibody may be purified as described below.

VT o7i/ WO 94/00555 PCT/US93/06120 Monoclonal antibody to the convertase may be produced using either in vivo or in vitro immunization techniques, and sensitized lymphocytes resulting therefrom can be used to prepare hybrid cell lines that secrete the appropriate monoclonal antibody. Rodent, preferably of murine origin, or human antibody is most preferred. The in vitro immunization procedure involves sensitizing lymphocytes to the convertase by immunizing either mice or humans, and isolating therefrom the antibody-secreting cell fraction and immortalizing the cells therein by one of several procedures. An alternate embodiment is to isolate lymphocytes that have already been sensitized to the convertase from septic patients or Wegener's granulomatosis patients as described above.

(i)Murine Antibody For in vivo immunization of mice, the procedure of Kohler and Milstein described in Nature, 256:495 (1975) may be followed, or modified procedures such as those shown by Fendly et 1987, Hybridoma, 6:359; Buck,e al., 1988, In Vitro, 18:377. In vitro techniques are generally described by Luben, R. and Mohler, 1980, Molecular Immunology, 17:635, Reading, Methods in Enzymology, 121 (Part One):18, or Voss, 1986, Methods in Enzymology, 121:27.

Mice are ii..,nunized with 1 mg/ml of a membranous fraction of HL60 cells previously shown to be positive for convertase activity. Alternatively, a smaller amount of purified TNF convertase may be employed. The immunization is carried out in complete Freund's adjuvant. Two additional immunizations, or boosts, are performed at monthly intervals without adjuvant, and one month after the last boost the mice are given an I.V. boost of 10 vg of membranous material.

Three days after the I.V. boost, mice are sacrificed, their spleens removed, and the spleenocytes isolated and fused to an immortalized drug selectable myeloma partner cell line. Numerous such myeloma lines are known in the art, most of which are incapable of growth in HAT supplemented cell culture media. A typical myeloma cell line is SP-2/OAg 14. Thus, the hybridomas are formed by 30 combining splenocytes and myeloma cells in a 5:1 ratio, which generally consists -I rC lp L~ IC- YIPI I WO 94/00555 W)/07UtS93/O20 -41 of 2 x 106 myeloma cells to 1 x 1( 7 splenocytes. The cell mixture is pelleted, media removed and fusion affected by the addition of 1.0 ml of 40% solution of polyethylene glycol 1500 by dropwise addition over 60 seconds at room temperature, followed by a 60 second incubation at 37 C. To the cell suspension with gentle agitation is added 9 ml of Dulbecco's Modified Eagles medium over minutes. Cell clumps in the mixture are gently resuspended, the cells washed to remove any residual PEG and plated in microtiter plates at about 2 x 105 cells/well in DMEM supplemented with 20% fetal calf serum. After 24 hours, the cells are fed a 2 x solution of hypoxanthine and azaserine selection medium.

Media from wells that exhibit positive cell growth may be screened for neutralizing monoclonal antibody to the convertase. The preferred assay is the convertase assay described in Example 2, above, wherein media sought to be tested for antibody activity is present in the assay. More preferred is to combine culture supernatants from 3-8 microtiter wells, and assay the mixture. If the mixture is positive, then media from each well may be assayed independently to identify the secreting hybridoma(s). Many assays are known in the art and can detect soluble, or non-soluble antigens, and are shown by Langone, J. and Van Vinakis, Methods of Enzvmology, 92, Part E (1983).

Regardless of whether the antibody is polyclonal or monoclonal it is desirable to purify the antibody by standard techniques as is known in the art, or described by Springer, 1980, Monoclonal Antibodies,:194, (Eds. Kennett, T. McKearn and K.

Bechtol, Plenum Press, New York. Generally this consists of at least one ammonium sulfate precipitation of the antibody using a 50% ammonium sulfate solution. Antibody affinity columns may also be used.

I

q _C 1 WO 94/00555 I'CI'/1JS93/06120 -42- (ii)Human Monoclonal Antibody Peripheral blood lymphocytes are isolated from septic patients, and then infected with Epstein-Barr virus and the infected lymphocytes immortalized by fusion to a selectable myeloma cell line, and the hybrid cell lines so generated isolated and characterized as to antibody production.

More specifically, mononuclear cells are separated on Ficoll-hypaque (Pharmacia), and monocytes depleted froni the mixture by adherence to plastic. Standard laboratory techniques were utilized to effect these procedures. Next, nonadherent cells are enriched for antibody producers by antigen-specific panning. Panning is a technique generally known in the art, and involves incubation of a population of antibody-secreting cells on a plastic surface coated with the appropriate antigen.

Those cells that express antibody on their surface bind antigen, and consequently adhere to the plastic surface, whereas cells that do not express cell surface antibody, do not adhere and can be removed by washing. Thus, specific antibodysecreting cells are enriched for by this technique.

More specifically, 6-well plates (Costar) are coated with purified TNFconvertase or a membrane fraction containing convertase prepared from either induced or uninduced HL60 cells, as described above, such that 150 p.g of membranous material is coated per well in phosphate buffered saline at 4 0 C overnight. The wells are blocked after the overnight incubation period with phosphate buffered saline containing 1% bovine serum albumin for at least 1 hour at 4 0 C, and subsequently washed with phosphate buffered saline/BSA. Next, 10' lymphocytes in 1 ml of PBS/BSA are added to each well of the six well plates. The lymphocytes are allowed to incubate on the plates for 70 minutes, after which any nonadherent cells are removed by aspiration. The adherent cells are incubated with cell culture medium (IMDM, Sigma Chemical Co., St. Louis, Missouri) containing 10% fetal calf serum.

The adherent cells are subjected to Epstein-Barr virus transformation by adding an equal amount of culture media obtained from growing the Epstein-Barr virus -0 infected marmoset cell line, B95-8, and thus containing the virus, to media bathing i I LYCl I LL -s~I bs L a WO 94/00555 I'C1'/1JS93/06 120 43 the adherent cells. The cells were cultured in this environment at 3 IoC for 3 hours, and in this way the lymphocytes in the adherent cell population are subjected to Epstein-Barr infection. Following the infection period, the cells are washed and plated onto 96 well microti're plates at a density of about 10' 105 cells/well in IMDM medium, plus 10% fetal calf serum, and 30% conditioned medium. The latter is derived from a lymphoblastoid cell line, preferably The medium also contains 5 x 10' M 2mercaptoethanol, 50 rig/ml gentamycin sulfate (Sigma), and 600 ng/ml cyclosporine A (Sandimmun, Sandoz, Basel, Switzerland).

After about 14 to 21 days of incubation, cell culture supernatants are combined and screened for TNF convertase-neutralizing activity as described above.

Positive hybridomas are subcultured at low density, retested for neutralizing antibody, and grown up and fused to the cell line F3B6 using polyethylene glycol and the plate fusion technique known in the art. The latter technique is described by Larrick, 1985, in Human Hybridomas and Monoclonal Antibodies, E.G.

Engleman, S.K.H. Foung, Larrick, and A.A. Raubitschek, Editors, Plenum Press, New York, page 446. F3B6 is a heteromyeloma cell line that is sensitive to growth in media containing 100rn hypoxanthine, 5 ytg/ml azaserine and 5 iAM ouabain. Finally, the resulting hybrids are again screened to insure that they produce neutrnlizing anti-convertase antibody.

B.26 kD Muteins 26 kD TNF muteins are described that compete for binding to the convertase, thereby inhibiting or reducing its activity. The preferred mutein embodiments are those having valine at positions 1 and/or 13; or alanine at position -1 and/or proline at position 12, replaced or deleted. The muteins are constructed using a modification of the site-directed mutagenesis gapped-duplex method.

The following solutions/buffers are used to perform the desired procedures: 5 x gapped-duplex buffer (GDB) consisting of 0.938 M KC1, 0.063 M Tris, pH 10 x PEL consisting of 1.0 M KCL, 0.30 M Tris, 0.15 M MgClI, 0.02 M DTT, I e -I C -I a WO 94/00555 IC'I'/ US93/06 120 -44pH 7.5; 10 x KB consisting of 0.50 M Tris, 0.10 M MgCl 2 0.05 M DTT, 0.001 M EDTA, pH 8.0; a solution containing 0.25 mM dCTP, dATP, dGTP, dTTP, made fresh from 10 mM stocks; an ATP solution consisting of 0.1 M ATP made by dissolving 60 mg of ATP in 0.8C ri of H 2 0 and adjusting the pH to 7.0 with 0.1 M NaOH in a final volume of 1.0 ml with H 2 0; 20% PEG/2.5 M N:aCl; M NaOAc; and TE-saturated phenol.

Various bacterial strains and phage are employed to yield the desired muteins and these are BMH 71-18, JM103 for growing phage strains; HB2154: MutL, Su', made competent for DNA transformation; and HB2151: Su', used as lawn cells during transformation; M13 GAP, the RF is used for the formation of the gapped-duplex; and M13mpl9amber, the 26 kD TNF target DNA is cloned in this vector, and ssDNA isolated for the formation of gapped-duplex.

Phage are infected into an appropriate bacterial strain, grown up, and titered as follows. In making a large-scale preparation of either phage for ssDNA or cells for dsDNA, or RF DNA, the same infection protocol is used.

Plaque-purified phage is produced using standard techniques. Briefly, this consists of streaking phage supernatants on agar plates, followed by careful overlay with ml of soft agar and 100 (l of fresh overnight culture of BMH 71-18. Next, isolated plaques are picked and incubated with a 1:50 dilution of fresh overnight culture of BMH 71-18 in R26 or R17 10 mM MgCL2 with shaking at 37 C for 4.5-6 hours. R17 (N-Z amine broth) consist of 10 g N-Z amine, type A, 5 g NaCI with H 2 0 to 1 lier, while R26 consist of 8 g tryptone, 5 g yeast extract, 5 g NaC1, with water to 1 liter (YT broth). The phage stock is titered, and phage infected into bacteria at a multiplicity of infection (MOI) of 10. After incubating the culture with shaking at 37 0 C for 5 hours the cell suspension is pelleted, and the supernatant saved for ssDNA isolation, and the cells for RF isolation. RF DNA is isolated using established plasmid DNA isolation techniques, while ssDNA is isolated as follows.

250 ml of phage supernatant is spun down hard, after which 200 ml of the supernatant is decanted, followed by adding 50 ml of 20% PEG/2.5 M NaCI, and 73 0 ^o^

__I

WO 94r/00555 WC4AJ'e S93/06 120 45 incubation overnight at 4 C, or on ice for 30 minutes. This mixture is also spun down hard, and the supernatant decanted. The bottle is spun again to pellet the phage precipitate along the sides of the bottle, and the remaining fluid aspirated with a Pasteur pipette. The pellet is resuspended in 5.0 ml of 1 x TE, and stored at 4 C, after which 0.5 ml of is extracted twice with 0.5 ml of TE saturated phenol. To the aqueous layer is added 0.050 ml of 3.0 M NaOAc and 1.0 ml ethanol. The mixture is placed in a dry ice bath for 10 minutes, and centrifuged for 10 minutes in a microfuge at 4 0 C. The pellet is dried, and resuspended in 200 y-l of 1 x TE. This material may be stored in 0.050 ml aliquots at -200C until used in the mutagenesis of 26 kD TNF.

The following deletions and substitutions in Table 1 are preferred proTNF muteins. These muteins can be prepared using appropriate oligonucleotides by methods known in the art.

01~- Ip c CP ~9~P T~QIIIICI~II~IPSII~ 91 ~a ICL~I Table 1 Deletions AVAL 1 AVAL 13 AVAL 1 APRO 12 AVAL 1 AVAL 13 Substitutions (VAL 1 ALA 1) (VAL 13 ALA 13) (VAL 1 GLY 1) (VAL 13 GLY 13) (VAL 1 LEU 1) (VAL 13 LEU 13) (VAL 1 MET 1) (VAL 13 MET 13) (VAL 1 PHE 1) (VAL 13 PI-IE 13) (VAL 1 HIS 1) (VAL 13 HIS 13) (VAL 1 THR 1) (VAL 13 TI-R 13) (ALA 1, VAL 1 GLN 1, HIS 1) (PRO 12, VAL 13 GLN 12, HIS 13) (ALA 1, VAL 1 GLN 1, HIS 1) (PRO 12, VAL 13 SER 12, THR 13) 20 The oligonucleotides are kinased using the following reaction solution and conditions: 31l 10 x KB buffer, 3X 10 nmM rATP (1:10 dilution of 0.1 M rATP stock, 2X mutagenic oligonucleotide (100 pmole/X) 21 X H20, and 1 X polynucleotide kinase (10 Units/X). The reaction is run at 37"C for minutes, and then at 65-68C for 5 minutes. Next 24X of the kinased oligonucleotide is diluted with 56 X of HO to give 2pmole/X.

The gapped-duplex is formed as described below, followed by annealing the oligonucleotides. The following reagents are combined in a total volume of 40 X:8 X5 x GDB buffer, 0.50 pmole ssDNA, and 0.10 pmole IlincII linearized a M13 GAP RF DNA. 10 X is removed for future use, and the remaining 30 X is treated sequentially as follows: 100"C for 3 minutes, 65"C for 5 minutes, followed by cooling to room temperature to 30 minutes, and then placing the reaction mixture on ice. Next, 10 k of gapped-duplex and 10X of control ungapped material is subject to electrophoresis on agarose gel to check gapped-duplex formation. If the gel shows the presence of a third band, the gapped-duplex has formed, and the kinased oligonucleotides can be annealed to the duplex by combining 16 X of gapped-duplex reaction mixture and 4 X I Y I~ ls LI II I~---eP-l I 1A -9 of diluted kinased oligonucleolide and heating the mixture to 05"C for 3 minutes, followed by cooling to room temperature for 20 minules.

The heteroduplex is completed by the appropriate extension and ligation reactions consisting of combining the following reagents in a total volume of 40 X: 10 X gapped-duplex and primer, 4 X 10 x PEL buffer, 4 X dNTP's (0.25 mM solution made from 10 mM stocks, 3 X ATP (10 X of 0.1 M ATP stock 1490 X HIO 0,662 mM), 17 X H20, 1 X Klenow (5 u/X) and 1 X T4 DNA ligase (0.6 Weiss u/X, diluted stock with 1 x PEL). The r'eaction is conducted at 16"C for 2 hours, followed by transformation of 10 of the extension/ligation mixture into 200 X of thawed competent I-B2154 cells. The cells are kept at 0"C for 30 minutes, and then 42"C for 1.5 minutes, following by plating various volumes of the transformation mix 50 1, 10X etc.) with 100 X of fresh overnight culture of HB2151 cells 3.OX of soft agar.

The resulting plaques are screened using the plaque hybridization procedure.

While a variety of such procedures are known, a description of the preferred procedure follows. Plates are replicated onto duplicate nitrocellulose filter papers (S S type BA85) and the DNA fixed to the filter by sequential treatment for 5 minutes with 0.5N NaOH plus 1.5 M NaCI; 1.0 M NaCI plus M Tris-HC1 pH 7,4; and 2 x SSC (standard saline citrate). Filters are air- 20 dried and baked at 80"C for 2 hours, in vacuo.

The duplicate filters are prehybridized at 55"C for 2 hours with 10 ml per filter of DNA hybridization buffer, 5 x SSC, pH 7.0, 5 x Donhardt's solution (polyvinylpyrrolidone, plus Ficoll and bovine serum albumin; 1 x 0.02% of each), 50 mM sodium phosphate buffer at pH 7.0, 5 mM EDTA. 0.1% SDS, and 100 pg/ml yeast RNA. The prehybridization buffer is removed and the samples hybridized with the appropriate kinased probe, specifically, kinased oligonucleotides as shown above, under conditions which depend on the stringency desired. About 2 x 106 cpm/ml total is used. Typical moderately stringent conditions employ a temperature of 42"C plus 50% formamide for 24-36 hours with 1-5 ml/filter of DNA hybriization buffer containing probe.

SFor higher I-1 PLI s -L i- II C I WO 94/00555 UlI'(/ US93/06 120 -48stringencies high temperatures and shorter times are employed. The preferred hybridization conditions consists of hybridizing the probes to the filters in 5 x SSC, Denhardt's solution, 50 mM NaPO 4 pH 7.0, 5 mM EDTA, 0.1% SDS, and 100 mg/ml yeast RNA at 10 0 C below the TM of the oligonucleotide used to do the screening. Next, the filters are washed twice, 30 minutes each wash, at room temperature with 2 x SSC, 0.1% SDS, then washed once with 2 x SSC and 0.1% SDS at 5 0 C below the T m of the oligonucleotide used to screen, and air-dried.

Finally, the filters are autoradiographed at -70-C for 36 hours. Autoradiography reveals those plaques containing the virus that carries the muteins of interest.

In addition to constructing muteins wherein valine at position 2 and/or 13 have been deleted or substituted, large deletion muteins may be produced that encompass the two predominate cleavage sites of 26 kD TNF. A preferred embodiment mutein lacks the amino acids spanning the region -9 to 14, as shown in Figure 1. This mutein was constructed using the materials and methods described above and the oligonucleotide, CP375 which has the following sequence (SEQ ID NO: 2).

C.Protein/Peptide Inhibitors Peptides having the following amino acid sequences are synthesized by the solid-phase method, described in detail by Merrifield, 1985), Science, 232:341-347: Gln-Ala-Val-Arg-Ser-Ser-Ser; (SEQ ID NO: (SEQ ID NO: 3 (SEQ ID NO: and (SEQ ID NO: A Biosearch 9500 automated peptide machine is used with hydrogen fluoride cleavage, and purification by preparative HPLC using a Waters Delta Prep 3000 instrument, on a 15-20p.m Vydac C4 PrepPAK column.

TNF convertase inhibitory activity of these peptides is shown by performing the assay described above in the presence of varying amounts of each peptide. Gel electrophoresis and Western blotting of the reaction m'ture shows an inhibition of conversion of the 26 kD proTNF to the 17 kD mature form.

A, rAQ.

I 5-n WO 94/00555r~ WCI'/ I S93/06 120 49- Example 7 TNF Convertase Inhibitory Activity of DCI in L929 Mice DCI specifically suppresses the release of TNF but not IL-6 from mouse macrophages as shown below.

Release of TNF by macrophages after stimulation by LPS is a major source of TNF. In these studies peritoneal macrophages were purified by adhesion, cultured in 24 well plates and LPS was added to induce secretion of TNF. Analysis of the kinetics of TNF release showed a maximal peak at 3 hours. DCI was then added in dimethyl sulfoxide excipient to cultures. The control cultures had P'MSO alone added in equivalent concentrations. Suprnatants were collected and assayed for TNF and IL-6. Results show that TNF secretion is markedly suppressed with DCI but not control excipient. In contrast the IL-6 response was not significantly altered, thus ruling out a nonspecific toxic effect (see Table 2).

Since DCI was able to specifically suppress LPS induced TNF secretion in murine macrophages the therapeutic effect of administration of DCI to mice injected with LPS was examined.

Stability and formulation studies showed that DCI when dissolved in corn oil was stable and retained serine protease inhibitor activity. Infection of DCI/oil into mice showed an LD 50% at a dose of 1 mg/ml. This represented a maximal tolerated dose of DCI that could be administered.

The kinetics of TNF and IL-6 in mice injected with a lethal dose .f LPS was studied. TNF showed a sharp peak at 2 hours with return to baseline. IL-6 showed a slower gradual increase. Injection of DCI 1 hour before the LPS dose resulted in a marked inhibition of serum TNF secretion (see Figure Also, there was no delayed ;ncrease in TNF measured up to the 6-hour time point. This was true for both i' .!nunoreactive mouse TNF measured by ELISA and bioactive TNF measured by lysis of L929 cells. IL-6 levels were not reduced by this therapy.

The effect of DCI on survival of mice injected with a dose of LPS that results in 100% death of animals by 24 hours was also investigated. Results show that ~11~1~ ~a pg I~IPIIPS I u II I I L prophylactic therapy with DCI could prolong survival of mice (see Figure 8).

There was a dose response relationship noted by 0.75 mg being more effective than 0.5 mg.

In summary, these studies show that DCI is able to spec.ically inhibit LPS induced TNF production by murine macrophages. This specificity of inhibition of TNC could also be seen in animals injected with a lethal dose of

LPS.

Furthermore, the survival of animals was prolonged with DCI therapy in a dose related manner. These studies show that DCI (a serine protease inhibitor) may be beneficial in a sepsis model in prolonging survival by suppression of the systemic release of TNF.

Table 2 S

S..

*a.

06a. 00 a a Oa S a Sample TNF (ng/ml) IL-6 (pg/ml) DMSO control 6.9 299 (tg/ml) 0.05 189 ~u 1~91 II L C C -I 1 r WO 94/00555 PC/US93/06120 -51 Example 8 Protective Effect of TNF Convertase Inhibitors in the Treatment of Sepsis Compounds that are effective inhibitors of convertase activity are shown to prevent sepsis in a baboon model system as follows. Anti-TNF convertase antibody, murine, human, or recombinant, at a concentration of 5 mg/kg is administered in a single I.V. bolus 60 minutes before the animals are challenged with a lethal dose of E. coli, and 2 mg/kg simultaneously with the E. coli challenge. The antibody is administered in a physiologically balanced salt solution, and about 4 x 1010 E.

coli organisms are used. The E. coli dose is infused over a 2 hour period.

Animals that receive the antibody are protected for at least 7 days, whereas control animals that are administered only the balanced salt solution expire within 16 to 32 hours.

Similar protection is attributable to the TNF mutein convertase inhibitors shown in Example 5. The muteins are administered at a concentration of 5 mg/kg in a single I.V. bolus 60 minutes before the animals are challenged with 4 x 10'0 E.

coli organisms. The baboons also receive 2 mg/kg of the muteins simultaneously with the E. coli challenge.

Finally, the peptides shown in Example 5, that is, Gln-Ala-Val-Arg-Ser-Ser-Ser and, (SEQ ID NO: 1) are tested as described above and yield similar protective effects.

Example 9 Modelling of Human PR-3 Onto Human Elastase Crystal Structure and Use of Inhibitor-Enzyme Complex Models to Predict Novel PR-3 Inhibitors A model for the TNF convertase PR-3 was constructed by determining structural similarities shared between PR-3 and other serine proteinases. A final 3-D model of the enzyme was generated by first determining that the PR-3 sequence shared a highest degree of sequence homology with human neutrophil elastase (HNE). The crystal structure HNE (Navia t al., 1989, PNAS (UTSA), was used as a scaffold to build a three dimensional representation of the PR-3 protein using the s L L I se r I a I c WO 94/00555 PI'C/ US93/06120 -52 computer program Homology (Biosym, San Diego). The model was further refined by two rounds of minimization using the computer program Discover (Biosym, San Diego). The design of potential inhibitors that differentiate between HNE and PR-3 is determined by the unique and similar amino acids found in the active sites of these enzymes. Most notably, the catalytic triad common to this class of proteinases is spatially conserved. Within the binding pocket of the P1 residue (S1 site) several significant differences in amino acid side chains are proposed by the model. The following described object compound of the present invention takes into account the unique aspartic acid and leucine amino acids found within the S pocket of the PR-3 model and can be represented by the following general formula.

R2 R0 RiY 0 Ra N R4

H

in which

R

2 are lower alkyl, optionally substituted ar(lower)alkyl, cyclo(lower)alkyl(lower)alkyl or optionally substituted heterocyclic(lower)alkyl, natural amino acids, -OH, -NH2, lower alkylimino or lower alkylene;

R

3 is pyroyl, imidazoyl, butylamine, or ethyl-epoxide; and

R

4 is aldehyde, diphosphonylate, ethoxycourmarinyl. chloromethyl and difluoromethyl ketonyl.

An example of a PR-3 inhibitor based on this model is Boc-Val-Pro-Hisp(OPh) 2 Inhibition of PR-3 activity by such a compound is unexpected in light of the generally accepted belief that elastase and PR-3 selectively bind and cut after L I y 9C-- 1 -C-T 9- h WO 94/00555 PC'/US93/06120 53 residues quite different from histidine, namely those with short aliphatic side chains such as alanine.

The present invention has been described with reference to specific embodiments.

However, this application is intended to cover those changes and substitutions which may be made by those skilled in the art without departing from the spirit and the scope of the appended claims.

I- ~dBdl WO 94/00555 WO 940555Pf US93/06 120 54 Sgquence Listing Gln-Ala-Val-Arg-Ser-Ser-Ser-Arg-Thr-Pro-Ser-Asp-Lys-Pro-Val-Ala (SEQ ID NO: 1) GMIIGCTACA ACATGGAGGT CCCTGGGGGA (SEQ ID NO: 2) Pro-Leu- Ala- G In- Ala-Val-Arg-S er-Ser- Ser- Arg-Thir- Pro- Ser- Asp-Lys-Pro- Val-Ala- HisVal-Val-Ala (SEQ ID NO: 3) Arg-Thr-Pro-S er-A sp-Lys-Pro-Val-A la-Hi s-Val -Val -Ala (SEQ 11D NO: 4) Pro-Leu-Ala-Gln-Ala-Val-Arg-Ser-Ser-Ser-Arg-Thr-Pro (SEQ ID NO:

Claims (28)

1. A method of identifying a prophylactic or therapeutic of a disease caused by a mature tumor necrosis factor (TNF) produced from a proTNF by cleavage of said proTNF by a TNF convertase, the method comprising the steps of: contacting the proTNF with an amount of TNF convertase effective for cleaving the proTNF; measuring the conversation of the proTNF to the mature TNF in step repeating steps and further including a molecule sought to be identified as a prophylactic or therapeutic of diseases caused by the mature TNF; comparing the conversion measured in step with the conversion measured in step and determining whether tne molecule is a suitable prophylactic or therapeutic of diseases caused by mature TNF.

2. A method according to claim 1, wherein the proTNF is 26 kD 20 TNF.

3. A method according to claim 1 or claim 2, wherein the disease is selected from the group consisting of sepsis, rheumatoid arthritis and AIDS.

4. A method according to claim 3, wherein the disease is sepsis. A method according to any one of claims 1 to 4, wherein the TNF convertase is proteinase-3 (PR-3).

6. A method according to claim 5, wherein the PR-3 is native PR-3. i

7. A method according to claim 5, wherein the PR-3 is recombinant PR-3. 3~ 1~ r~-A -urs r 11;

8. A method according to any one of claims 1 to 7, wherein the conversion of proTNF to mature TNF in steps and is approximated by an enzymatic assay measuring TNF convertase cleavage of a TNF converlase substrate.

9. A therapeutic or prophylactic compound for tratling a disease caused by a mature TNF produced from a proTNF by cleavage of said proTNF by a TNF convertase, the therapeutic or prophylactic identified by a method comprising the steps of: contacting the proTNF with an amount of the TNF convertase effective for cleaving the proTNF; measuring the conversion of the proTNF to Ihe mature TNF in step repeating steps and further including a molecule sought to be identified as a prophylactic or therapeutic of diseases caused by the mature TNF; comparing the conversion measured in step with the conversion measured in step and determining whether !he molecule is a suitable i 20 prophylactic or therapeutic of diseases caused by mature TNF.

10. A therapeutic or prophylactic according to claim 9, wherein the proTNF is 26 kD TNF. V. 11. A therapeutic or prophylactic according to claim 9 or claim wherein the disease is selected from the group consisting of sepsis, rheunatoid arthritis and AIDS.

12. A therapeutic or prophylactic according to claim 11, wherein i the disease is sepsis,

13. A therapeutic or prophylactic according to any one of claims 9 to 12, wherein the TNF convertase is proteinase-3 (PR-3). IUII~~ ,3sR~I~_~S ~CI c-'~l~pBIL~a

14. A thernpeutic orI prophylactic accorling to claiii 13, whereinl the PR-3 is nialive PR-3. A therapeutic compound for treating a disease caused by a mature TNF produced fronm a proTNF by cleavage of said proNF by a TNF convertase, the compound being a non-cleavable mutein of proTNF.

16. A therapeutic compound according to clainm 15. wherein the non-cleavable mutein includes a substitution or deletion at amino acid 1 (valine) or anino acid 13 (valine).

17. A method for treating a patient having a disease or susceptible to a disease caused by a mature TNF produced from a proTNF by cleavage of' said proTNF by a TNF convertase, the method comprising administering to a patient in need of such treatment an effective amount of an inhibitor of a TNF convertase.

18. A method according to claim 17, wherein the proTNF is 26 kD TNF.

19. A method according to claim 17 or claiiim 18, wheroiin the disease is selected from the group consisting of sopsis, rheoumatoid arthritis, cachexia cerebral malaria, AIDS and graftversus-host disease.

20. A method according to claim 19, wherein the disease is sepsis. V9 a *e~21. A niethod according to any one of claims 17 to 20, wherein the TNF convertase is proteinase-3 (PR-3).

22. A minethod according to claim 21, wherein the PR-3 is native PR- .3

23. A method of according to claim 21, wherein thlie PR-3 is recombinant PR-3. -r qlel a IC~ r

24. A method according to any one of claims 17 to 23, wherein the inhibitor is a peptide diphenyl phosphonate having the formula Boc- X- p(OPh)2, wherein X is an oligopeptide Val-Pro-His. 25, A method according to any one of claims 17 to 23, wherein the inhibitor is a neutralizing antibody to PR-3.

26. A pharmaceutical composition for the treatment of a disease caused by a mature TNF produced from a proTNF by cleavage of said proTNF by a TNF convertase, the composition comprising an effective amount of an inhibitor of a TNF convertase and a pharmaceutically acceptable excipient.

27. A pharmaceutical composition according to claim 26, wherein the disease is selected from the group consisting of sepsis, rheumatoid arthritis and AIDS.

28. A pharmaceutical composition according to claim 27, wherein the disease is sepsis. 20 29. A pharmaceutical composition according to any one of claims 26 to 28, wherein the TNF convertase is proteinase-3 (PR-3).

30. A pharmaceutical composition according to claim 29, wherein the PR-3 is native PR-3. 4oo:

31. A pharmaceutical composition according to claim 29, wherein the PR-3 is recombinant PR-3.

32. A pharmaceutical composition according to any one of claims 26 to 31, wherein the inhibitor is a peptide diphenyl phosphonate having the formula Boc- X- p(OPh) wherein X is an oligopeptide selected from the C: .i group consisting of Val-Pro-Val, Ala-Pro-Val and Val-Pro-His.

33. A pharmaceutical composition according to claim 32, wherein X is Val-Pro-His. c I- e~ 9 A4 popid dphallyI )1J~l phoato hanvinig the, Iuiiiuiihi,, Hioc- X- p(O P1I), vvhorcoi n X is till cligopep tide selec ted humll I he group consistIing of' Val-Pro-I-is.

35. A method for tremiting a patient for autoimimune diseases, the method comp.rising adiministering to a patient in need of'such treat nient an effective amount of an inhibitor of'a TNF convertai,. 6 4099 *.G 009