Human Pancreatic Elastase

curopaiscnes Katentamt

European Patent Office (jj) Publication number: O 1 98 645 B1 Office europeen des brevets

cUHOPeAN PATENT SPECIFICATION

jy nuinan pancreauc eiasiase.

rropnetor: sankyo company LIMITED, 27.04.85 JP 91986/85 No. 1-6, 3-chome Nihonbashi Honcho Chuo-ku, 26.07.85 JP 163964/85 Tokyo(JP) 23.10.85 JP 236686/85

inventor: Takiguchi, Yo C/o Bio-Science Research Laboratory, Sankyo Company Limited 22.10.86 Bulletin 86/43 Mo. 2-58, 1-chome, Hiromachi Shinagawa-ku rokyo140(JP) Inventor: Tani, Tokyo C/o Bio-Science Research ui^nv/aLiui i ui ii ic yi ai u li ic Ljctlcl II. Laboratory, Sankyo Company Limited 22.11.90 Bulletin 90/47 Mo. 2-58, 1-chome, Hiromachi Shinagawa-ku rokyo 140(JP) nventor: Kawashima, Ichiro C/o Bio-Science Research -ab., Sankyo Company Limited No. 2-58, 1-chome, \T BE CH DE FR GB IT LI LU NL SE Hiromachi Shinagawa-ku Tokyo 140(JP) nventor: Eurukawa, Hidehiko C/o Bio-Science Research Lab., Sankyo Company Limited Ao. 2-58, 1-chome, Hiromachi IWIVI VI VSI IbU . Shinagawa-ku =P-A- 0157 604 rokyo 140(JP) nventor: Ohmine, Toshinori C/o Bio-Science Research -ab., Sankyo Company Limited No. 3I0CHIMICA ET BIOPHYSICA ACTA, vol. 623, no. 1, 2-58, 1-chome, u 1980 -iiromachi Shinagawa-ku Tokyo 140(JP) i/lay 29, (Amsterdam); C. LARGMAN ET nventor: Ohsumi, Jun C/o Bio-Science U.::"Human Pancreatic Proelastase 2 Sequence of the Research Lab., Vctivation Peptide", 208-212 Sankyo Company Limited No. 2-58, 1-chome, Hiromachi pp. Shinagawa-ku Tokyo 140(JP) 0 representative: Gibson, Christian John Robert et al, A ARKS & CLERK 57/60 Lincoln's Inn Fields, London VC2A3LS(GB)

c iiiuiiuio iiuiii ii le puuuuaiion or me mention ot tne grant ot tne turopean patent, any person may give notice to le European Patent Office of opposition to the European patent granted. Notice of opposition shall be filed in a written reasoned tatement. It shall not be deemed to have been filed until the opposition fee has been paid (Art 99(1) European patent * c anvention). II- ACTORUM AG EP 0198 645 B1

uescription Background of the Invention

5 This invention relates to human pancreatic elastase. Elastase is a serine protease, capable of hydrolyzing the fibrous insoluble protein known as elastin. Elastin is a scleroprotein forming connective tissues, tendons, aortic integuments and cervical bundles of higher animals. Elastin is only slightly degraded by pepsin and trypsin. In the course of their study on arteriosclerosis, Balo' et ai observed degradation of the elastin fibers 10 of blood vessel walls, and postulated the presence of a degrading enzyme [Balo', J and Banga, I: Schweiz Z Pathol Bacteriol, 12, 350 (1949)]. Subsequently, Banga discovered an enzyme in the pancreas which specifically degrades elastin. The enzyme was isolated in the form of crystals and named "elastase" [Banga, I: Acta Physiol Acad Sci Hung, 3, 317 (1952)]. Elastase has been confirmed to exist in the pancreas of most animals, including humans, monkeys, 15 cats, rabbits, etc. The level is about 3.1 mg/g-pancreas in human beings, about 2.2 mg/g in bovine animals and about 10.2 mg/g in rats. A correlation is recognized between elastase activity and the age of a human being: a marked lowering in elastase activity is observed in the pancreas and plasma of males over 40 and females over 60 years [Loeven, W A and Baldwin, Maureen M: Gerontologia, 17, 170 (1971)]. In the case of patients with arteriosclerosis, the elastase activity in the pancreas was found to be 20 markedly lower than that of healthy people, and in some cases it had completely disappeared [Balo', J, and Banga, I: Nature, 178, 310 (1956)]. Subsequent studies have also demonstrated that elastase not only degrades elastin but also promotes elastin biosynthesis. Studies on the pharmacological action of elastase have been carried out in rats and rabbits, and have revealed the following effects: 25 1) inhibition of the deposition of lipids and calcium on arterial walls; 2) elimination of cholesterol and calcium from arterial walls; 3) selective degradation of denatured elastin; 4) promotion of synthesis of elastin fibers in the arterial walls; 30 5) lowering of serum lipids; 6) improvement of lipoprotein metabolism.

In clinical studies conducted on the basis of the studies mentioned above, the following effects have become apparent after oral administration of elastase: 35 1) restoration of elasticity and expandability of arterial walls; 2) improvement of serum lipid abnormality; 3) improvement of serum lipoprotein metabolism.

40 Elastase extracted and purified from porcine pancreas was used for the above studies. However, with administration of porcine elastase to human beings, there is the risk of antibody formation due to the foreign protein, and the danger of anaphylaxis with repeated administration (Japanese Patent Applica- tion Laid-open 11180/1984). Accordingly, human elastase is preferable for human use. However it has been extremely difficult to procure human elastase in sufficient quantities from traditional sources. 45 At present, two kinds of human pancreatic elastase are known, elastase 1 and elastase 2. They have not been fully characterized. For example, for pancreatic elastase 2, a partial sequence of 16 amino acid residues representing 12 amino acid residues in the activation peptide and 4 amino acid residues at the N-terminal end of the elastase have been sequenced [Largman, C et al; Biochim Biophys Acta, 623, 208 (1 980)]. This partial sequence is Cys Gly Asp Pro Thr Tyr Pro Pro Tyr Val Thr Arg Val Val Gly Gly. Objects of the Present Invention

It is an object of this invention to make human pancreatic elastase more readily available, with the pos- sibility of obtaining the elastase in substantially pure form. It is a further object of this invention to elimi- 55 nate the continuing dependency on human pancreas for adequate supplies of human pancreatic elastase. It is a yet further object to produce new elastase compounds and elastase-like compounds. Summary of the Present Invention

30 This invention embraces the use of genetic engineering for the production of human pancreatic elastase. Recombinant DNA technology is employed, in order to provide a base sequence coding for a human pancreatic elastase or a functionally similar compound. Correspondingly, human pancreatic elastase and functionally similar compounds now become readily available for the first time. Indeed, novel elastases have been discovered and are part of this invention. 35 More specifically, this invention provides a compound which includes an amino acid sequence for a EP0 198 645 B1

human pancreatic elastase selected from the group of human pancreatic elastase IIB, human pancreatic elastase IMA, human pancreatic elastase NIB, and any compounds corresponding to human pancreatic elastase IIB, IIIA or NIB with one or more deleted, inserted, replaced or altered amino acids. The sequences for human pancreatic elastases IIB, IIIA, and IIB are given below. The present invention also 5 embraces DNA which codes for the compounds provided by this invention. Thus, through the use of genetic engineering, it is now possible in accordance with the present invention to provide a DNA which codes for a compound capable of functioning as a human pancreatic elastase. Novel human pancreatic elastases are provided, as well as derivative compounds which func- tion as human pancreatic elastases. Such compounds, which include compounds arising from silent muta- 10 tions, fusion proteins and other compounds functionally effective as a human pancreatic elastase, are included. Thus, substantially similar protein effective as a human pancreatic elastase, that is, compounds corresponding to human pancreatic elastases with one or more deleted, replaced or altered amino acids, and compounds corresponding to human pancreatic elastases with one or more extra amino acids, are also included. More especially, precursor compounds such as proelastases and preproelastas- 15 es, especially when expressed by a recombinant DNA sequence, as well as elastases obtained by activation of such precursor compounds, are part of this invention. Further examples of precursor compounds within this invention include fusion proteins comprising an elastase (optionally in the form of a proelastase or preproelastase) and an amino acid sequence derived from another protein. In particular, such fusion proteins can be obtained when the DNA coding for an elastase is inserted into an expres- 20 sion gene downstream of a promoter. Elastase IIB

This invention provides human pancreatic elastase IIB having the following amino acid sequence of 25 formula (III): (N)-Cys Gly Val Ser Thr Tyr Ala Pro Asp Met Ser Arg Met Leu Gly Gly Glu Glu Ala Arg Pro Asn Ser Trp Pro Trp Gin Val Ser Leu Gin Tyr Ser Ser Asn Gly Gin Trp Tyr His 30 Thr Cys Gly Gly Ser Leu lie Ala Asn Ser Trp Val Leu Thr Ala Ala His Cys He Ser Ser Ser Arg lie Tyr Arg Val Met Leu Gly Gin His Asn Leu Tyr Val Ala Glu Ser Gly Ser Leu Ala Val Ser Val Ser Lys lie Val 35 Val His Lys Asp Trp Asn Ser Asn Gin Val Ser Lys Gly Asn Asp He Ala Leu Leu Lys Leu Ala Asn Pro Val Ser Leu Thr Asp Lys He Gin Leu Ala Cys Leu Pro Pro Ala Gly Thr He Leu Pro Asn Asn Tyr Pro Cys Tyr 40 Val Thr Gly Trp Gly Arg Leu Gin Thr Asn Gly Ala Leu Pro Asp Asp Leu Lys Gin Gly Arg Leu Leu Val Val Asp Tyr Ala Thr Cys Ser Ser Ser Gly Trp Trp Gly Ser Thr Val Lys Thr Asn Met He Cys Ala Gly Gly Asp 45 Gly Val lie Cys Thr Cys Asn Gly Asp Ser Gly Gly Pro Leu Asn Cys Gin Ala Ser Asp Gly Arg Trp Glu Val His Gly lie Gly Ser Leu Thr Ser Val Leu Gly Cys Asn Tyr Tyr Tyr Lys Pro Ser He Phe Thr Arg Val Ser 50 Asn Tyr Asn Asp Trp He Asn Ser Val lie Ala Asn Asn - (C) For the human pancreatic elastase IIB itself, the N-terminal end simply terminates with a hydrogen at- om. For derivatives of human pancreatic elastase lib, the protein can be extended with an amino acid sequence such as Met or all or part of (N) Met lie Arg Thr Leu Leu Leu Ser Thr Leu Val Ala Gly Ala Leu 55 Ser (C). The invention also provides a DNA coding for human pancreatic elastase IIB. Such a DNA can have a base sequence represented by the following formula (IV): (5')-TGT GGG GTC TCC ACT TAC GCG CCT GAT ATG TCT AGG ATG CTT GGA GGT GAA GAA GCG AGG 80 CCC AAC AGC TGG CCC TGG CAG GTC TCC CTG CAG TAC AGC TCC AAT GGC CAG TGG TAC CAC ACC TGC GGA GGG TCC CTG ATA GCC AAC AGC TGG GTC CTG ACG GCT GCC CAC TGC ATC AGC TCC TCC AGG ATC TAC CGC GTG ATG CTG GGC 35 CAG CAT AAC CTC TAC GTT GCA GAG TCC GGC

5 EP 0198 645 B1

I Uu U I G GCC GTC AGT GTC TCT AAG ATT GTG GTG CAC AAG GAC TGG AAC TCC AAC CAG GTC TCC AAA GGG AAC GAC ATT GCC CTG CTC AAA CTG GCT AAC CCC GTC TCC CTC ACC GAC AAG 5 ATC CAG CTG GCC TGC CTC CCT CCT GCC GGC ACC ATT CTA CCC AAC AAC TAC CCC TGC TAC GTC ACA GGC TGG GGA AGG CTG CAG ACC AAC GGG GCT CTC CCT GAT GAC CTG AAG CAG GGC CGG TTG CTG GTT GTG GAC TAT GCC ACC TGC 1 0 TCC AGC TCT GGC TGG TGG GGC AGC ACC GTG AAG ACG AAT ATG ATC TGT GCT GGG GGT GAT GGC GTG ATA TGC ACC TGC AAC GGA GAC TCC GGT GGG CCG CTG AAC TGT CAG GCA TCT GAC GGC CGG TGG GAG GTG CAT GGC ATC GGC AGC 1 5 CTC ACG TCG GTC CTT GGT TGC AAC TAC TAC TAC AAG CCC TCC ATC TTC ACG CGG GTC TCC AAC TAC AAC GAC TGG ATC AAT TCG GTG ATT GCA AAT AAC - X (3') wherein X represents a stop codon, that is TAA, TGA or TAG. It is to be recognized that the amino acid 20 sequence of formula (III) can arise from a different DNA base sequence to that of formula (IV), and such a modified DNA sequence is also part of this invention. The DNA may optionally have ATG at its 5'-end, which will then code for an additional Met at the N-terminal end of the amino acid sequence. As another option, the DNA may have at its 5'-end a part or all of the sequence (5') - ATG ATT AGG ACC CTG CTG CTG TCC ACT TTG GTG GCT GGA GCC CTC AGT - 25 (3'), which will then code for an extra sequence at the N-terminal end of the amino acid sequence comprising a part or all of (N) Met He Arg Thr Leu Leu Leu Ser Thr Leu Val Ala Gly Ala Leu Ser (C). Type IIB human pancreatic elastase of this invention may exist in a two-chain configuration formed by a disulfide-bond between Cys at position 1 and an inner Cys in the molecule when the bond is broken between amino acid 12 and amino acid 13. The present invention includes this two-chain elastase. 30 Elastase IIIA

This invention provides human pancreatic elastase IIIA having the following amino acid sequence of formula (V): 35 (N)-Val Val His Gly Glu Asp Ala Val Pro Tyr Ser Trp Pro Trp Gin Val Ser Leu Gin Tyr Glu Lys Ser Gly Ser Phe Try His Thr Cys Gly Gly Ser Leu lie Ala Pro Asp Trp Val Val Thr Ala Gly His Cys lie Ser Arg Asp 40 Leu Thr Tyr Gin Val Val Leu Gly Glu Tyr Asn Leu Ala Val Lys Glu Gly Pro Glu Gin Val He Pro lie Asn Ser Glu Glu Leu Phe Val His Pro Leu Trp Asn Arg Ser Cys Val Ala Cys Gly Asn Asp lie Ala Leu He Lys 45 Leu Ser Arg Ser Ala Gin Leu Gly Asp Ala Val Gin Leu Ala Ser Leu Pro Pro Ala Gly Asp He Leu Pro Asn Lys Thr Pro Cys Tyr He Thr Gly Trp Gly Arg Leu Tyr Thr Asn Gly Pro Leu Pro Asp Lys Leu Gin Gin Ala 50 Arg Leu Pro Val Val Asp Tyr Lys His Cys Ser Arg Trp Asn Trp Trp Gly Ser Thr Val Lys Lys Thr Met Val Cys Ala Gly Gly Tyr lie Arg Ser Gly Cys Asn Gly Asp Ser Gly Gly Pro Leu Asn Cys Pro Thr Glu Asp Gly 55 Gly Trp Gin Val His Gly Val Thr Ser Phe Val Ser Ala Phe Gly Cys Asn Phe lie Trp Lys Pro Thr Val Phe Thr Arg Val Ser Ala Phe lie Asp Trp He Glu Glu Thr He Ala Ser His - (C) 30 For the human pancreatic elastase IIIA itself, the N-terminal end simply terminates with a hydrogen at- om. For derivatives of human pancreatic elastase IIIA, the protein can be preceded with an amino acid sequence such as Met or a part or all of (N) Met Met Leu Arg Leu Leu Ser Ser Leu Leu Leu Val Ala Val Ala Ser Gly Tyr Gly Pro Pro Ser Ser His Ser Ser Ser Arg (C). The invention also provides a DNA coding for human pancreatic elastase IIIA. Such a DNA can have 35 a base sequence represented by the following formula (VI): EP 0 198 645 B1

(5')-GTT GTC CAT GGT GAG GAT GCG GTC CCC TAC AGC TGG CCC TGG CAG GTT TCC CTG CAG TAT GAG AAA AGT GGA AGC TTC TAC CAC ACG TGT GGC GGT AGC CTC ATC GCC CCC GAT TGG GTT 5 GTG ACT GCC GGC CAC TGC ATC TCG AGG GAT CTG ACC TAC CAG GTG GTG TTG GGT GAG TAC AAC CTT GCT GTG AAG GAG GGC CCC GAG CAG GTG ATC CCC ATC AAC TCT GAG GAG CTG TTT GTG CAT CCA CTC TGG AAC CGC TCG TGT GTG 10 GCC TGT GGC AAT GAC ATC GCC CTC ATC AAG CTC TCA CGC AGC GCC CAG CTG GGA GAT GCC GTC CAG CTC GCC TCA CTC CCT CCC GCT GGT GAC ATC CTT CCC AAC AAG ACA CCC TGC TAC ATC ACC GGC TGG GGC CGT CTC TAT ACC AAT 15 GGG CCA CTC CCA GAC AAG CTG CAG CAG GCC CGG CTG CCC GTG GTG GAC TAT AAG CAC TGC TCC AGG TGG AAC TGG TGG GGT TCC ACC GTG AAG AAA ACC ATG GTG TGT GCT GGA GGG TAC ATC CGC TCC GGC TGC AAC GGT GAC TCT GGA 20 GGA CCC CTC AAC TGC CCC ACA GAG GAT GGT GGC TGG CAG GTC CAC GGT GTG ACC AGC TTT GTT TCT GCC TTT GGC TGC AAC TTC ATC TGG AAG CCC ACG GTG TTC ACT CGA GTC TCC GCC TTC ATC GAC TGG ATT GAG GAG ACC ATA GCA 25 AGC CAC - X (3') wherein X represents a stop codon, that is TAA, TGA or TAG. It is to be recognized that the amino acid sequence of formula (V) can arise from a different DNA base sequence to the one of formula (VI), and such a modified DNA sequence is also part of this invention. The DNA may optionally have ATG at its 5'-end, which will then code for an additional Met at the N-ter- 30 minal end of the amino acid sequence. As another option, the DNA may optionally have at its 5'-end a part or all of the sequence (5') - ATG ATG CTC CGG CTG CTC AGT TCC CTC CTC CTT GTG GCC GTT GCC TCA GGC TAT GGC CCA CCT TCC TCT CAC TCT TCC AGC CGC - (3'), which will then code for an extra sequence at the N-terminal end of the amino acid sequence comprising a part or all of (N) Met Met Leu Arg Leu Leu Ser Ser Leu Leu Leu Val Ala Val Ser Gly Tyr Gly Pro Pro Ser Ser His Ser Ser Ser 35 Arg (C). Elastase 1MB

This invention provides human pancreatic elastase 1MB having the following amino acid sequence of 40 formula (VII): (N)- Val Val Asn Gly Glu Asp Ala Val Pro Tyr Ser Trp Pro Trp Gin Val Ser Leu Gin Try Glu Lys Ser Gly Ser Phe Tyr His Thr Cys Gly Gly Ser Leu He Ala Pro Asp Trp Val 45 Val Thr Ala Gly His Cys He Ser Ser Ser Arg Thr Tyr Gin Val Val Leu Gly Glu Tyr Asp Arg Ala Val Lys Glu Gly Pro Glu Gin Val He Pro lie Asn Ser Gly Asp Leu Phe Val His Pro Leu Trp Asn Arg Ser Cys Val 50 Ala Cys Gly Asn Asp He Ala Leu He Lys Leu Ser Arg Ser Ala Gin Leu Gly Asp Ala Val Gin Leu Ala Ser Leu Pro Pro Ala Gly Asp lie Leu Pro Asn Glu Thr Pro Cys Tyr He Thr Gly Trp Gly Arg Leu Tyr Thr Asn 55 Gly Pro Leu Pro Asp Lys Leu Gin Glu Ala Leu Leu Pro Val Val Asp Tyr Glu His Cys Ser Arg Trp Asn Trp Trp Gly Ser Ser Val Lys Lys Thr Met Val Cys Ala Gly Gly Asp lie Arg Ser Gly Cys Asn Gly Asp Ser Gly 60 Gly Pro Leu Asn Cys Pro Thr Glu Asp Gly Gly Trp Gin Val His Gly Val Thr Ser Phe Val Ser Ala Phe Gly Cys Asn Thr Arg Arg Lys Pro Thr Val Phe Thr Arg Val Ser Ala Phe He Asp Trp lie Glu Glu Thr He Ala 65 Ser His - (C)

5 EP 0198 645 B1

l-or the human pancreatic elastase IIIB itself, the N-termina! end simply terminates with a hydrogen at- om. For derivatives of human pancreatic elastase IIIB, the protein can be preceded with an amino acid sequence such as Met or a part or all of (N) Met Met Leu Arg Leu Leu Ser Ser Leu Leu Leu Val Ala Val Ala Ser Gly Tyr Gly Pro Pro Ser Ser Arg Pro Ser Ser Arg (C). 5 The invention also provides a DNA coding for human pancreatic elastase IIIB. Such DNA can have a base sequence represented by the following formula (VIII): (5")-GTT GTC AAT GGT GAG GAT GCG GTC CCC TAC AGC TGG CCC TGG CAG GTT TCC CTG CAG TAT GAG AAA AGC GGA AGC TTC TAC CAC ACC TGT 10 GGC GGT AGC CTC ATC GCC CCC GAC TGG GTT GTG ACT GCC GGC CAC TGC ATC TCG AGC TCC CGG ACC TAC CAG GTG GTG TTG GGC GAG TAC GAC CGT GCT GTG AAG GAG GGC CCC GAG CAG GTG ATC CCC ATC AAC TCT GGG GAC CTC TTT 1 5 GTG CAT CCA CTC TGG AAC CGC TCG TGT GTG GCC TGT GGC AAT GAC ATC GCC CTC ATC AAG CTC TCA CGC AGC GCC CAG CTG GGA GAC GCC GTC CAG CTC GCC TCA CTC CCT CCG GCT GGT GAC ATC CTT CCC AAC GAG ACA CCC TGC TAC 20 ATC ACC GGC TGG GGC CGT CTC TAT ACC AAC GGG CCA CTC CCA GAC AAG CTG CAG GAG GCC CTG CTG CCG GTG GTG GAC TAT GAA CAC TGC TCC AGG TGG AAC TGG TGG GGT TCC TCC GTG AAG AAG ACC ATG GTG TGT GCT GGA GGG GAC 25 ATC CGC TCC GGC TGC AAT GGT GAC TCT GGA GGA CCC CTC AAC TGC CCC ACA GAG GAT GGT GGC TGG CAG GTC CAT GGC GTG ACC AGC TTT GTT TCT GCC TTT GGC TGC AAC ACC CGC AGG AAG CCC ACG GTG TTC ACT CGA GTC TCC GCC 30 TTC ATT GAC TGG ATT GAG GAG ACC ATA GCA AGC CAC - X (31) wherein X represents a stop codon, that is TAA, TGA or TAG. It is to be recognized that the amino acid sequence of formula (VII) can arise from a different DNA base sequence to the one of formula (VIII), and such a modified DNA sequence is also part of this invention. 35 The DNA may optionally have ATG at its 5'-end, which will then code for an additional Met at the N-terminal end of the amino acid sequence. As another option, the DNA may have at its 5'-end a part or all of the sequence (5') - ATG ATG CTC CGG CTG CTC AGT TCC CTC CTC CTT GTG GCC GTT GCC TCA GGC TAT GGC CCA CCT TCC TCT CGC CCT TCC AGC CGC - (31), which will then code for an extra sequence at the N-terminal end of the amino acid sequence comprising a part or all of (N) Met Met Leu 40 Leu Leu Ser Ser Leu Leu Leu Val Ala Val Ala Ser Pro (cf Gly Tyr Giy Pro Ser Ser Arg Pro Ser Ser Arg

Processes of this Invention

45 The DNA of this invention, including the base sequences represented by the formulae (II), (IV), (VI) and (VIII) with optional extensions, can be produced for example in accordance with the steps (a) to (d): (a) separation of mRNA from human pancreas; (b) formation from the mRNA of a bank of cDNA in a suitable host; 50 (c) isolation from the cDNA plasmid bank of plasmids including cDNA coding for human elastase, for example by using cDNA of porcine elastase I or of rat elastase II as a probe; (d) isolation of the desired cloned DNA from the plasmid.

From the DNA, the corresponding protein can then be prepared, for example by the steps of: 55 (1) inserting the DNA into an expression vector; (2) transforming a host organism; (3) culturing the transformed host under conditions resulting in expression of the DNA sequence; and (4) isolating a compound including the protein. so Considering the process in some more detail, several method are available which may be used in the extraction of mRNA from the human pancreas, including for instance the guanidine-thiocyanate-hot-phenol method, the guanidine thiocyanate-caesium chloride method, or the guanidine-thiocyanate-guanidine hydrochloride method. The last-mentioned method is preferred since it is superior to the other two meth- S5 ods in the following respects: the operation is simpler; recovery of extraction is high; extraction from a EP 0 198 645 B1

relatively large number of organs is possible; no contaminating DNA is entrained; and degradation of mRNA is low. Most of the mRNA existing in the cytoplasm of eukaryotic cells is known to have a poly(A) sequence at the 3'-end. Utilizing this specific feature, mRNA can be purified by adsorbing it onto a column of an oli- 5 go(dT) cellulose, and then eluting it. Using the mRNA obtained as a template, complementary double-stranded DNA can be synthesized with the use of a reverse transcriptase. Several methods can be employed for synthesising the double stranded DNA, including for example the S1 nuclease method [Efstratiadis, A eta[:Cell, 7, 279 (1976), the Land method [Land, H et a]: Nucleic Acids Res, 9, 2251 (1981)], the Okayama-Berg method [Okayama, H 10 and Berg, P: Mol Cell Biol, 2, 161 (1982)], the 0 Joon Yoo method [O Joon Yoo et a[: Proc Natl Acad Sci USA, 79 1 049 (1 982)], etc. For the present invention, the Okayama-Berg method is preferred. The double stranded cDNA is thus inserted into a suitable vector. The vector typically takes the form of a plasmid, especially a plasmid with a unique restriction endonuclease site, more especially a restriction site within a phenotypic sequence. Various techniques can be employed for inserting the cDNA into 15 the vector, including homopolymer tailing and the direct ligation of "sticky ends" produced following digestion of both the cDNA and the vector with a restriction endonuclease. The recombinant plasmids containing the cDNA can then be introduced into a suitable host, suitably a strain of E coH, for example strain RR1 , for transformation and thereby give the cDNA bank. The transformed hosts may then be selected on the basis of a phenotype modification used as a marker, for exam- 20 pie antibiotic resistance such as tetracycline resistance or ampicillin resistance. It is then necessary to select from amongst the transformants the clones which contain DNA coding for a human pancreatic elastase. This selection can be achieved by several methods, including the direct sequencing method and the colony hybridization method. In the present invention the colony hybridization method is preferred. For example, cDNA of porcine elastase I (Japanese Provisional Patent Pub- 25 lication No 207583/1985 corresponding to EP157604A published 9 October 1985 and further corresponding to USSN 716,189 filed March 26, 1985) labelled with or cDNA of rat elastase II labelled with 32p can be used as the probe for human elastase. The base sequence contained in the selected clone may, if desired, be determined. Determination can be effected, for example, by dideoxynucleotide termination sequencing using phage M1 3 [Messing, J et 30 ak Nucleic Acid Res, 9, 309 (1981) or by the Maxam-Gilbert method [Maxam, A M and Gilbert, W: Proc Natl Acad Sci USA, 74, 560 (1977)]. In the present invention, the dideoxynucleotide termination method is preferred, in order that clones having a complete DNA coding for a human pancreatic elastase can be identified. The elastase cDNA may then be isolated from the plasmid DNA of the selected clone, inserted into an 35 appropriate expression vector and introduced into an appropriate host in order to allow expression therein. The expression vector originally created may be cut with appropriate restriction endonucleases to construct further vectors containing the elastase cDNA. Bacteria such as E. coli or B. subtilis. yeasts, and mammalian cells can be employed as the host. 40 The DNA of the present invention may be introduced into the host by transformation using any of several methods, including the Hanahan method [Hanahan, D: J Mol Biol, 166. 557 (1983)], the calcium chloride method [Dagert, M and Ehrlick, S D: Gene, 6, 23 (1979)], the low pH method [page 49 of the Manual of Genetic Manipulation, edited by Yasuyuki Takagi, Kodansha Scientific (1982)], etc. The Hanahan method is currently preferred. 45 The host thus obtained (the transformed host) can be cultured in a suitable medium to produce and ac- cumulate the elastase or a substance having the same effect, followed by recovery thereof. Typical media for cuituring the transformed host include those comprising glucose, casamino acids, and other known components, for example, M9 medium [Miller, J: Experiments in Molecular Genetics, 431 to 433, Cold Spring Harbor Lab, New York (1972)] 50 The transformed host is generally cultured at 15 to 43°C for 3 to 24 hours, with aeration or stirring, if necessary. However, when mammalian cells are used as the host, it is usually necessary to carry out the culture for 3 to 1 0 days. After cuituring, the transformed host can be harvested in a conventional manner, for example by centrifugation or other known techniques. When B. subtilis. yeast or mammalians cells are employed as the 55 host, the elastase produced is generally released from the cell into the supernatant. However, when E coli is employed as the host, the elastase is mainly present as an undissolved protein in inclusion bodies within the cells. In cases in which E coli is used as the host, the elastase may be obtained by disruption of the cells. For example, the elastase can be obtained as a precipitate after suspending the cells in a buffer, rupturing the cells and centrifugation. The cells may be ruptured by conventional procedures, in- 60 eluding sonication treatment, lysozyme treatment or freeze-thaw treatment. Isolation of elastase from the supernatant or the precipitate may be practised according to the conventional methods known in the art for the purification of proteins. The typical elastases produced by the present invention include proelastases and other elastase derivatives. After any appropriate activation, the elastases generally exhibit comparable biological activi- 65 ty to those purified from human pancreatic fluid. They can be used for the same purposes and in the

7 EP 01 98 645 B1

same ways as ine extracted elastases. Ihus, the present invention further provides pharmaceutical compositions which comprise an elastase of this invention, together with a pharmaceutically acceptable carrier or diluent. 5 Examples of the Present Invention

The present invention is further illustrated by the following non-limiting Examples 2 to 4. Example 1 illustrates work on a closely human pancreatic elastase, Elastase 1 1 A. 10 Summary of the Drawings

In the Examples, reference is made to the accompanying drawings, in which:

Figure 1 is a restriction endonuclease cleavage map of part of plasmid pH2E2 obtained in Example 1, 15 wherein the stippled portion indicates the section coding a peptide considered to be the signal peptide, and the striped portion indicates the section coding the mature elastase protein. Figure 2 illustrates the construction of plasmid pSV2-E2A in Example 1. illustrates Figure 3 the construction of plasmid pBSEX-E2A in Example 1 . Figure 4 sets out a synthetic DNA linker containing part of the signal peptide region in the a-amylase 20 gene of Bacillus subtilis and the activation peptide N-terminal of elastase IIA cDNA, as employed in Ex- ample 1. Figure 5 illustrates the construction of plasmid pH2EX-2 in Example 1. The striped area represents the p-galactosidase gene. Broad arrows indicate the lactose promoter and operator, and narrow arrows show the direction of transcription. 25 Figure 6 is a restriction endonuclease cleavage map of part of plasmid pH2E8 obtained in Example 2, wherein the stippled portion indicates the section coding a peptide considered to be the signal peptide! and the striped portion indicates the section coding the mature elastase protein. Figure 7 illustrates the construction of plasmid pSV2-E2B in Example 2. Figure 8 illustrates the construction of plasmid pBSEX-E2B in Example 2. 30 Figure 9 sets out a synthetic DNA linker containing part of the signal peptide coding region in the a- amylase gene of Bacillus subtilis and the activation peptide N-terminal coding region of elastase IIB cDNA, as employed in Example 2. Figure 10 illustrates the construction of plasmid pH2EX-8 in Example 2. The striped area the represents p-galactosidase gene. Broad arrows indicate the lactose promoter and operator, and narrow arrows 35 show the direction of transcription. Figure 1 1 illustrates the construction of plasmid pSV2-CL2 in Example 3. Figure 1 2 illustrates the construction of plasmid pHEX010 in Example 3. Figure 13 sets out a synthetic DNA linker containing part of the signal peptide region in the a-amylase gene of Bacillus subtilis and the activation peptide N-terminal of elastase IIIA cDNA, as employed in Ex- 10 ample 3. Figure 14 illustrates the construction of plasmid PHEX002 in Example 3. The striped area represents the p-galactosidase gene. Broad arrows indicate the lactose promoter and operator, and narrow arrows show the direction of transcription. Figure 15 reproduces the results of polyacrylamide gel electrophoresis carried out in example 3, 15 where lane 1 corresponds to molecular weight markers, and lanes 2 to 5 correspond to E. coH YA21 , HB1 01 , MC41 00 and ^984, respectively, when transformed with plasmid pHEX002. Figure 16 illustrates the construction of plasmid pSV2-CL1 in Example 4. Figure 17 illustrates the construction of plasmid pHAM001 in Example 4. Figure 18 sets out a synthetic DNA linker containing part of the signal peptide region in the a-amylase 50 gene of Bacillus subtilis and the activation peptide N-terminal of elastase IIIB, cDNA, as employed in Ex- ample 4. Figure 19 illustrates the construction of plasmid pHEX102 in Example 4. The striped area represents the p-galactosidase gene. Broad arrows indicate the lactose promoter and operator, and narrow arrows show the direction of transcription. Example 1: Elastase IIIA

1) Separation of mRNA from human pancreas

SO Human pancreas (autopsy sample) weighing 5.5 g was homogenised and denatured in a homogeniser (a Polytron homogeniser from Kinematica GmbH, Germany, Polytron being a Trade Mark) in an adequate amount of guanidine thiocyanate solution (4M guanidine thiocyanate; 1% detergent, Sarcosyl from Sigma Chemical Co, USA; 20mM ethylenediaminetetraacetic acid (EDTA); 25mM sodium citrate, pH 7.0; 100mM 2-mercaptoethanol and 0.1% antifoaming agent, Antifoam A from Sigma Chemical Co USA), followed by >5 centrifugation, to obtain a supernatant. EP 0 198 645 B1

One volume of supernatant was added to 0.025 volumes of 1 M acetic acid and 0.75 volumes of ethanol, and the mixture cooled to -20°C for several hours, followed by centrifugation, to obtain a precipitate. The precipitate was suspended in a guanidine hydrochloride solution (7.5M guanidine hydrochloride; 25mM sodium citrate, pH 7.0; and 5mM dithiothreitol, DTT). To 1 volume of the suspension was add- 5 ed 0.025 volumes of 1 M acetic acid and 0.5 volumes of ethanol, and the resultant mix cooled to -20°C for several hours, followed by centrifugation. The resultant precipitate was suspended again in more of the guanidine hydrochloride solution, mixed as before with acetic acid and ethanol, cooled to -20°C and centrifuged, followed by collection of the precipitate. Next, the precipitate was washed several times with ethanol to remove guanidine hydrochloride, and then dissolved, in distilled water, followed by precipita- 10 tion of the RNA with ethanol. The precipitate was collected by centrifugation to give 53.9 mg of RNA. The RNA thus obtained was adsorbed onto an oligo(dT) cellulose column in a high concentation salt solution (0.5M NaCI; 20mM Tris-HCI, pH 7.5; 1mM EDTA; and 0.1% sodium dodecyl sulfate, SDS), and the mRNA-containing poly(A) was eluted with an elution buffer (10mM Tris-HCI, pH7.5; 1mM EDTA; and 0.05% SDS) to give 255 ug of mRNA. 15 2) Preparation of cDNA bank

The preparation of a cDNA bank was carried out according to the Okayama-Berg method. 5 ug of the mRNA and 24 units of a reverse transcriptase were incubated at 42°C for 60 minutes in 20 ul of a reac- 20 tion mixture (50mM of Tris-HCI, pH 8.3; 8mM of MgCI2; 30mM of KCI; 0.3 mM of DTT; 2mM each of dATP, dGTP, dCTP and dTTP; 10 uCi of oc-32P-dCTP; and 1.4 ug of vector primer DNA, purchased from PL-Pharmacia). The reaction was stopped by the addition of 2 ul of 0.25M EDTA and 1 ul of 10% SDS. Deproteiniza- tion was carried out with 20 ul of phenol-chloroform. After centrifugation of the reaction mixture, the 25 aqueous layer was collected. 20 ul of 4M ammonium acetate and 80 ul of ethanol were added to the aqueous layer and the mixture cooled to -70°C for 15 minutes.The precipitate was then collected by centrifugation and washed with 75% aqueous ethanol and dried under reduced pressure. The resultant precipitate was dissolved in 15 ul of terminal transferase reaction mixture (140mfv1 potas- sium cacodylate, of formula C2H6ASKO2; 30 mM Tris-HCI, pH 6.8; 1 mM cobalt chloride; 0.5mM DTT; 0.2 30 ug of poly(A); and 100mM of dCTP). After incubating the solution at 37°C for 3 minutes, 18 units of terminal deoxynucleotidyl transferase were added and allowed to react for 5 minutes. The above reaction was stopped by addition of 1 ul of 0.25M EDTA and 0.5 ul of 1 0% SDS. Deproteini- zation was then carried out with phenol-chloroform. After centrifugation of the reaction mixture, the aqueous layer was collected. 15 ul of 4M ammonium acetate and 60 ul of ethanol were added to the aque- 35 ous layer and mixed well. After cooling the mixture to -70°C for 15 minutes, the precipitate was collected by centrifugation. The precipitate was dissolved in 10 ul of a buffer for restriction endonuclease (50mM NaC1; 10mM Tris-HCI, pH 7.5; 10mM MgCl2,- and 1mM DTT). 2.5 units of the restriction enzyme Hindlll were added to the solution and the solution digested at 37°C for about 1 hour. After deproteinization with phenol-chloro- 40 form, precipitation was carried out with ethanol. After cooling to -70°C for 1 5 minutes, the precipitate was collected by centrifugation and dissolved in 10 ul of TE (10mM Tris-HCI, pH 7.5; 1mM EDTA). 1 ul of the above sample solution was then added to a reaction solution (10mM Tris-HCI, pH 7.5; 1mM EDTA; and 100mM NaCI), followed by addition of 10 ng of linker (purchased from PL-Pharmacia) bearing 45 oligo dG. The resulting mixture was heated at 65°C for 5 minutes and maintained at 42°C for 30 minutes. After cooling the reaction solution in ice water, 10 ul of 10-fold ligase buffer solution (10mM ATP; 660mM Tris-HCI, pH 7.5; 66mM MgCI2; and 100mM DTT), 78 ul of distilled water and 8 units of T4 DNA ligase were added to the solution and the reaction mixture maintained at 12°C overnight. Then, to the thus prepared mixture, 10 ul of 1M KCI, 1 unit of ribonuclease H, 33 units of DNA polymer- 50 ase I, 4 units of T4 DNA ligase, 0.5 |xl of nucleotide solution (20mM dATP, 20 mM dGTP, 20mM dCTP and 20mM dTTP) and 0.1 ul of 50 ug/ul bovine serum albumin (BSA) were added and the mixture maintained at 12°C for 1 hour and then at 25°C for 1 hour. After diluting the reaction solution 5-fold, E.coli strain RR1 was transformed according to the method of Hanahan [Hanahan, D; J Mol Biol 166, 557 (1983)] to prepare a human pancreatic cDNA bank. 55 3) Selection of transformed bacteria containing human pancreatic elastase HA cDNA

9 EPO 198 645 B1

sequence irom amino acia isi to amino acid 156 of the rat pancreatic elastase II. The 5'-end of the oligodeoxynucleotide was labelled with 32p by the method of Maxam, A M and Gilbert, W [Proc Natl Acad Sci, USA 74, 560 (1977)] to give a probe. Using the probe, a rat pancreatic elastase II cDNA clone was selected from the rat pancreatic cDNA bank. Then a fragment containing 200 bases was excised from the 5 rat pancreatic elastase II cDNA with restriction endnucleases Aatll and Hindlll and labelled with 32p according to the nick translation method [Rigby, P W et al: J Mol Biol 1J3, 237 (1977)]. The DNA fragment was then used as a probe for hybridization with the DNA fixed on the nitrocellulose filter, according to the method of Alwine, S C et a] [Proc Natl Acad Sci, USA 74, 5350 (1 977)]. 8 clones reactive with the probe were identified by autoradiography. A plasmid from one of the strains 10 was named "pH2E2". The restriction endonuclease cleavage map of the DNA inserted in pH2E2 was determined and is shown in Figure 1 . The base sequence was determined and is of the formula (5')-TT TTA CAG AAC TCC CAC GGA CAC ACC ATG ATA AGG ACG CTG CTG 1 5 CTG TCC ACT TTG GTG GCT GGA GCC CTC AGT TGT GGG GAC CCC ACT TAC CCA CCT TAT GTG ACT AGG GTG GTT GGC GGT GAA GAA GCG AGG CCC AAC AGC TGG CCC TGG CAG GTC TCC CTG CAG TAC AGC TCC AAT GGC AAG TGG TAC CAC 20 ACC TGC GGA GGG TCC CTG ATA GCC AAC AGC TGG GTC CTG ACG GCT GCC CAC TGC ATC AGC TCC TCC AGG ACC TAC CGC GTG GGG CTG GGC CGG CAC AAC CTC TAC GTT GCG GAG TCC GGC TCG CTG GCA GTC AGT GTC TCT AAG ATT GTG 25 GTG CAC AAG GAC TGG AAC TCC AAC CAA ATC TCC AAA GGG AAC GAC ATT GCC CTG CTC AAA CTG GCT AAC CCC GTC TCC CTC ACC GAC AAG ATC CAG CTG GCC TGC CTC CCT CCT GCC GGC ACC ATT CTA CCC AAC AAC TAC CCC TGC TAC 30 GTC ACG GGC TGG GGA AGG CTG CAG ACC AAC GGG GCT GTT CCT GAT GTC CTG CAG CAG GGC CGG TTG CTG GTT GTG GAC TAT GCC ACC TGC TCC AGC TCT GCC TGG TGG GGC AGC AGC GTG AAA ACC AGT ATG ATC TGT GCT GGG GGT GAT 35 GGC GTG ATC TCC AGC TGC AAC GGA GAC TCT GGC GGG CCA CTG AAC TGT CAG GCG TCT GAC GGC CGG TGG CAG GTG CAC GGC ATC GTC AGC TTC GGG TCT CGC CTC GGC TGC AAC TAC TAC CAC AAG CCC TCC GTC TTC ACG CGG GTC TCC W AAT TAC ATC GAC TGG ATC AAT TCG GTG ATT GCA AAT AAC TAA CCA AAA GAA GTC CCT GGG ACT GTT TCA GAC TTG GAA AGG TCA CAG AAG GAA AAT AAT ATA ATA AAG TGA CAA CTA TGC -(3') 15 Beginning at the underlined ATG triplet, the encoded amino acid sequence is of the formula: (N)- Met He Arg Thr Leu Leu Leu Ser Thr Leu Val Ala Gly Ala Leu Ser Cys Gly Asp Pro Thr Tyr Pro Pro Tyr Val Thr Arg Val Val Gly Gly Glu Glu Ala Arg 50 Pro Asn Ser Trp Pro Trp Gin Val Ser Leu Gin Tyr Ser Ser Asn Gly Lys Trp Tyr His Thr Cys Gly Gly Ser Leu lie Ala Asn Ser Trp Val Leu Thr Ala Ala His Cys He Ser Ser Ser Arg Thr Tyr Arg Val Gly Leu Gly 15 Arg His Asn Leu Tyr Val Ala Glu Ser Gly Ser Leu Ala Val Ser Val Ser Lys He Val Val His Lys Asp Trp Asn Ser Asn Gin He Ser Lys Gly Asn Asp He Ala Leu Leu Lys Leu Ala Asn Pro Val Ser Leu Thr Asp Lys iO lie Gin Leu Ala Cys Leu Pro Pro Ala Gly Thr He Leu Pro Asn Asn Tyr Pro Cys Tyr Val Thr Gly Trp Gly Arg Leu Gin Thr Asn Gly Ala Val Pro Asp Val Leu Gin Gin Gly Arg Leu Leu Val Val Asp Tyr Ala Thr Cys >5 Ser Ser Ser Ala Trp Trp Gly Ser Ser Val

0 EP 0 198 645 B1

Lys Thr Ser Met lie Cys Ala Gly Gly Asp Gly Val lie Ser Ser Cys Asn Gly Asp Ser Gly Gly Pro Leu Asn Cys Gin Ala Ser Asp Gly Arg Trp Gin Val His Gly He Val Ser 5 Phe Gly Ser Arg Leu Gly Cys Asn Tyr Tyr His Lys Pro Ser Val Phe Thr Arg Val Ser Asn Tyr He Asp Trp lie Asn Ser Val lie Als Asn Asn -(C) It can be seen that plasmid pH2E2 has a region coding for a signal peptide comprising 16 amino acids 10 and a region coding for the mature elastase protein termed elastase IIA. The elastase HA comprises 253 amino acids. Plasmid pH2E2 also contains 5' and 3' non-translated regions. There is as much as 84% homology between the amino acid sequence coded by the DNA insertion of plasmid pH2E2, and that of rat pancreatic elastase II reported by MacDonald, Ft J et al [Biochem, 21, 1453 (1982]. Accordingly, plasmid pH2E2 is believed to contain cDNA coding a human pancreatic 15 elastase II. With regard to the known sequence for human pancreatic elastase II, the 16 amino acid residue at the N-terminal of pH2E2 coding protein is identical to that of the human pancreatic elastase II reported by Largman, C, et al [Biochim Biophys Acta 623, 208 (1 980)]. Since pH2E2 contains an entire DNA coding for human pancreatic elastase HA, it is possible to produce large amounts of the elastase IIA by transferring the cDNA to an appropriate expression vector 20 and using, for example, E. coji, B.subtilis, yeast or mammalian cells as the host. 4) Production of human pancreatic elastase IIA using mammalian cells Construction of expression plasmid PSV2-E2A 25 In order to produce human elastase IIA using mammalian cells as host, the cDNA was ligated to an expression vector according to the procedure indicated in figure 2. The known pSV2 plasmid cortaining the SV40 promoter, enhancer, poly(A) signal and introns of the Small T antigen SV40 gene was used for construction of the expression vector. pSV2-p-globin is a plasmid based on the pSV2 vector with insert- 30 ed p-globin cDNA. Plasmid pH2E2 is the plasmid with inserted human elastase IIA cDNA. The reactions for with S1 nuclease, etc. were carried out according to the methods described in "Molecular Cloning" [Maniatis, T et al (ed) "Molecular Cloning" Cold Spring Harbor Lab (1982)]. The broad arrows in figure 2 represent the promoter derived from SV40, and the narrow arrows show the direction of transcription. From analysis of the restriction endonuclease digestion pattern, an expression plasmid (pSV2-E2A) 35 was selected in which vector and cDNA were ligated in the correct orientation for transcription. Transfection of COS 1 cells with expression plasmid pSV2-E2A

The constructed expression plasmid pSV2-E2A was transfected into COS 1 mammalian cells by the cal- 40 cium phosphate method described in the literature [Graham and van der Eb: Virology 52 456 (1973)]. Thus, 1 x 106 of COS 1 cells were seeded on Petri dishes (10 cm in diameter), and cultured overnight on Dulbecco-modified Eagle medium containing 10% fetal calf serum. Plasmid pSV2-E2A (300 ug) was suspended in sterile distilled water (12.55 ml), then 2.5 M CaCk (0.75 ml) was added and the suspension mixed thoroughly. A pipette was used to bubble air through the 45 system in order to maintain agitation, and 1.5 ml of 10 x HeBS solution (HEPES, 210 mM; NaCI, 1.37 M; KCI 4.7 mM; Na2HP04, 10.6 mM; glucose, 55.5 mM; pH 7.05) was dropped into the resulting solution to precipitate the DNA and calcium phosphate. The precipitate was allowed to stand at room temperature for 30 minutes to mature the precipitate, and 1 ml of the solution for each Petri dish was added to the COS 1 cells cultured in fresh medium containing 50 1 0% fetal calf serum. After cultivation of these cells for 12 hours at 37°C in the presence of 5% CO2, the culture medium was discarded in favour of a fresh, Dulbecco-modified Eagle medium containing no fetal calf serum. The cells were cultured for a further 48 hours at 37°C in the presence of 5% CO2. The transfected COS 1 cells obtained by this procedure were tested to detect the presence of human elastase IIA mRNA, while 55 the supernatant of the culture medium was tested for elastase activity. Extraction of mRNA from COS 1 cells

In order to confirm the existence of human elastase IIA mRNA transcribed from the expression plas- 60 mid in the transfected COS 1 cells, mRNA was extracted from the COS 1 cells and assayed by Northern blot hybridization according to the following procedure. After cultivation of the COS 1 cells for 48 hours, 1 ml of guanidine thiocyanate solution (4M guanidine thiocyanate; 1% Sarcosyl; 20 mM EDTA; 25 mM sodium citrate, pH 7.0; 100 mM 2-mercaptoethanol; 0.1% Antifoam A) was added to each Petri dish in order to lyse the cells. 65 High molecular DNA molecules were degraded to low molecular ones by passing the solution several

11 EP 0 198 645 B1

times through a 21 guage injection needle. The resultant solution was layered on a solution containing 5.7 M cesium chloride and 0.1 M EDTA, and the solution centrifuged at 20°C, 30,000 rpm for 17 hours using a Hitachi RPS 40 swing rotor. The resultant RNA precipitate was washed with a small amount of ethanol and dissolved in distilled water (300 ul). 5 According to the method of Aviv and Leder [Proc Natl Acad Sci, USA 69, 1408 (1972)], the extracted

20 Since the human elastase IIA cDNA in pSV2-E2A has a signal peptide region, secretion of expressed elastase into the medium as a proelastase was expected. Elastase activity in the medium after 48 hours cultivation was accordingly determined. Tris-HCI buffer solution (pH 8.0) was added to the supernatant of the culture medium (1 ml) to a final concentration of 0.2 M, 10 mg/ml trypsin (50 ul) was added, and an activation treatment of the proelastase was carried out for 15 minutes at 25°C. Next, 50 ul of soybean 25 trypsin inhibitor solution (10 mg/ml) and a synthetic substrate, glutaryl-Ala-Ala-Pro-Leu-p-nitroanilide were added, and the enzyme reaction was carried out by incubation at 25°C. After completion of the reaction, liberated p-nitroanilide was determined by measuring absorbance at 410 nm. The elastase activity in the culture medium of COS 1 cells transfected with pSV2-E2A was distinctly detected, and the production of active human elastase IIA in COS 1 cells was found. 30 The activation treatment with trypsin or other activator was indispensable. The production of human elastase IIA in COS 1 cells transfected with plasmid pSV2-E2A was tran- sient. However, if plasmid pSV2-E2A is linked with a suitable selection marker (for example, the neo gene, dihydrofolate reductase gene, or other resistance marker) and is transfected in to CHO or other cells, a cell line suitable for sustained production of human elastase IIA can be prepared. 35 5) Production of human pancreatic elastase IIA using Bacillus subtilis Construction of expression vector PBSEX-E2A

40 The construction method for the expression vector is illustrated in figure 3. Plasmid pH2E2 containing human elastase IIA cDNA was digested with restriction endonuclease Hindlll to a linear DNA, and then partially digested with restriction enzyme Pvull. A DNA fragment of about 1.3 Kb containing the elastase IIA cDNA was isolated. This fragment was partially digested with restriction endonuclease Avail to give a fragment containing cDNA coding for the amino acid sequence beginning from the third amino acid 45 counted from the N-terminal end of the elastase IIA activation peptide. The specific oligonucleotide sequence shown in Figure 4 was synthesised. This sequence codes for part of the signal peptide of B. sub- tHjs a-amylase, and two amino acids (Cys, Gly) at the N-terminal end of the elastase IIA activation peptide. The fragments were ligated with T4 DNA ligase. On the other hand, plasmid pTUB228 [Ohmura, K etal: J Biochem 95 87 (1 984)] containing the a-amyla- 50 se gene of Bacillus subtilis was digested with restriction endonuclease Hindlll to isolate a DNA fragment of 428 base pairs containing the a-amylase promoter and a part of the signal peptide, and separately a DNA fragment of 5,100 base pairs containing the replication origin. The DNA fragment of 428 base pairs and the DNA fragment of 5,100 base pairs were further digested with restriction endonuclease Hpall and Smal, respectively, to give DNA fragments of 385 base pairs and of 4566 base pairs. >5 The three DNA fragments were ligated with T4 DNA ligase, and incorporated into protoplasts of Bacil- ]us subtilis 207 - 25 strain (m768 hsrM recE4 amyE07 arol906 leuA8 Iys21; derived from Marburg strain) according to conventional procedures. After regeneration, cultivation on a medium containing 10 ug/ml of kanamycin allowed selection of the transformed strains which could grow on this medium. Selec- tion of the desired plasmid was achieved through colony 3 hybridization using as probe the synthetic oligonucleotide of figure 4. A positive clone was identified and the plasmid shown to be the intended one by determination of the base sequence. The expression plasmid obtained in this manner was designated plasmid pBSEX-E2A.

2 EP 0 198 645 B1

Elastase activity in supernatant of the culture broth

Bacillus subtilis 207-25 strain transformed with the human elastase IIA expression plasmid pB- SEXE2A was cultured on a reciprocal shaker in 1 I of LG medium (1% Bacto trypton, Difco; 0.5% yeast 5 extract, Difco; 0.5% NaCI; 0.2% glucose; pH 7.0) containing 50 ug/ml of kanamycin. The cuituring was performed at 35°C for 48 hours. After completion of the culture, the culture medium was cooled to 4°C and centrifuged at 3000 x G for 5 minutes, and the cells discarded. Ammonium sulfate was added to the supernatant to 55% saturation, and the solution was stirred at 4°C for 12 hours. Insoluble material formed in this treatment was precipi- 10 tated by centrifugation at 8000 x G for 20 minutes, the supernatant was discarded, and the precipitate was dissolved in 20 ml of 0.2 M Tris-HCI buffer solution (pH 8.0). This solution was dialyzed against 1 1 of 0.2 M Tris-HCI buffer solution (pH 8.0) for 1 6 hours, and insoluble material was removed by centrifugation at 8000 x G for 20 minutes. The dialyzed solution was taken as a crude elastase IIA sample solution, and the activity was determined by the following procedure. 15 Determination of Elastase Activity in the Sample Solution

Proelastase was activated by trypsin treatment of the sample solution, and allowed to react at 25°C with a synthetic substrate, glutaryl-Ala-Ala-Pro-Leu-p-nitroanilide. Liberated p-nitroanilide was deter- 20 mined by measuring absorbance at 410 nm. Since elastase activity was distinctly detected in the sample solution from the 207-25 strain transformed with pBSEX-E2A, the production of active human elastase IIA in Bacillus subtilis was demonstrated. 6) Production of human pancreatic elastase IIA using Escherichia coli 25 Various different promoters may be used, including for instance the tryptophan (trp) promoter, lactose (lac) promoter, tryptophan-lactase (tac) promoter, lipoprotein (Ipp) promoter, lambda {X) Pl promoter derived from bacteriophage, or polypeptide chain elongation factor (tuf B) promoter. An example with the lactose promoter is as follows. 30 As shown in figure 5, plasmid pH2E2 containing human elastase IIA cDNA was digested with restriction endonuclease Hindlll to form a linear DNA, and then partially digested with restriction enzyme Pvull to isolate a DNA fragment of about 1 .3 Kb containing elastase IIA cDNA. This fragment was partially digested with restriction endonuclease Ddel to give a fragment containing cDNA coding for an amino acid sequence down stream from the activation peptide of elastase IIA. By 35 treatment of this DNA fragment DNA polymerase Klenow fragment and four kinds of deoxyribonucleo- side 5'-triphosphates (dATP, dCTP, dGTP, dTTP), both ends were converted to blunt ends. The resultant DNA fragment was inserted into the Smal site of plasmid pUC8. Escherichia coli strain JM1 03 was transformed with the plasmid. Plasmid was extracted from several transformed strains, and through the use of restriction endonuclease mapping, plasmid with the same ori- 40 entation of cDNA transcription and of the promoter for the lactose operon was selected. This elastase IIA expression plasmid was designated as pH2EX-2. The base sequence and corresponding amino acid sequence for around the 5'-end of the elastase cD- NA in pH2EX-2 are as follows. 45 3-galactosidase 3^ — elastase Ser^Leu Thr Met lie Thr Asn Ser Cys Gly Asp ... 50 ATG ACC ATG ATT ACG AAT TCC CTC AGT TGT GGG GAC ...

55 Therefore, the human elastase-HA expressed by pH2EX-2 is shown to be a fused protein in which two amino acids (Leu, Ser) derived from the signal peptide and six amino acids derived from p-galactosidase combine at the N-terminal. Several Escherichia coli strains were transformed with pH2EX-2 to give strains capable of producing human elastase IIA. The strains containing expression plasmid were inoculated in 2 x TY-ampicillin medi- 60 urn (1.6% Bacto trypton; 1% yeast extract; 0.5% NaCI; 50 ug/ml ampicillin) and cultured at 37°C for 15 hours. After completion of the culture, cells were harvested by centrifugation of the culture medium, 2.4 x 108 cells were suspended in 15 ul of SDS solution (2% SDS, 5% 2-mercaptoethanol; 10% glycerin; 60mM Tris-HCI, pH 6.8). The cell suspension was heated at 1 00°C for 3 minutes, and protein was analyzed by SDS polyacrylamide gel electrophoresis, according to the method of Laemmli et a] [Nature, 227 65 680(1970)].

13 tF Ul 98 645 B1

as a resun, a large quantity ot elastase production was found in strain YA21 . The production yield of elastase fusion protein was 20% of the total protein produced by the YA21 strain. Though a considerable production of elastase fusion protein in %984 strain was observed, the production yield was less than half that in the YA21 strain. The production yields in two other Escherichia coli strains (HB101 and 5 MC41 00 strains) were low. Since the elastase fusion protein forms inclusion bodies in cells, purification was coparatively easy. Thus, from 1 I of a culture medium of YA21 strain transformed with plasmid pH2EX-2, 6.5 g of cells forming inclusion bodies could be obtained. The obtained cells (6.5 g) were lysed in 50 mM Tris-HCI buffer solution (pH 8.0) containing 0.2 mg/ml lysozyme and 1 mg/ml deoxycholic acid. Intact cells were removed by 10 low speed centrifugation (1,500 x G, 10 minutes) and inclusion bodies were recovered as a precipitate by high speed centrifugation (11,000 x G, 20 minutes). Since the precipitate still contained much ceil debris, the precipitate was suspended in 50 mM Tris-HCI buffer solution containing 5 mg/ml octylphenoxy poly- ethoxyethanol surfactant (Triton X-100, Trade Mark) and washed by high speed centrifugation (11,000 x G, 20 minutes). A washed pellet of inclusion bodies was suspended in a small amount of Tris-HCI buffer 15 solution and stored at 4°C. The purified inclusion body (300 mg) containing about 50% elastase IIA fusion protein could be obtained by this procedure. The inclusion body produced by YA21 strain transformed with pH2EX-2 was shown to contain the elastase fusion protein by immuno-blotting. Though most of the elastase produced by Escherichia coli is in the insoluble fraction of the cells as in- 20 elusion bodies, a smaller proportion exists in a soluble state and retains enzyme activity. Determination of the enzyme activity was carried out as follows. The strain x984 strain transformed with plasmid pH2EX-2 was cultured with shaking in 1 I of 2 x TY-ampicillin medium at 37°C for 15 hours. After completion of the cuituring, the cells were harvested by centrifugation at 3000 x G for 5 minutes and suspended in 20 ml of buffer solution A (50 mM Tris-HCI, pH 8.0; 1mM EDTA; 50 mM NaCI). 10 mg of lysozyme 25 was added, and the suspension incubated at 5°C for 20 minutes. Deoxycholic acid to a final concentration of 1 mg/ml was added to the resultant suspension which was then incubated at 20°C. Deoxyribonucle- ase to a final concentration of 0.1 mg/ml was added, and the cells were disrupted in a Polytron homogenizer. After removal of the cell debris by centrifugation at 80,000 x G, 40 minutes, the lysate was subjected to column chromatography on high molecular weight, cross-linked dextran (Sephadex G-75, a gel 30 filtration material able to separate substances of molecular weight 1 ,000 to 5,000). The fractions having elastase activity were purified by antibody affinity chromatography, and used for the following determination of elastase activity. Proelastase was activated by trypsin treatment of the sample solution. Next, p-nitroanilide liberated by the reaction with a synthetic substrate, glutaryl-Ala-Ala-Pro-Leu-p-nitroanilide, was determined by 35 measuring the absorbance at 410 nm. The reaction was carried out by incubation at 25°C in 0.2 M Tris- HCI buffer solution (pH 8.0) containing the synthetic substrate. As a result, elastase activity was distinctly detected in the sample solution of strain %984 transformed with plasmid pH2EX-2, and therefore the production of active human elastase IIA was demonstrated. 10 7) Production of human pancreatic elastase IIA using veast

In the case of yeast as a host (and as with a Escherichia coli. Bacillus subtilis or mammalian cells), human pancreatic elastase IIA cDNA could be linked with a suitable expression vector using known techniques, introduced into the host cells and expressed. Elastase activity in the culture medium was con- 15 firmed. Saccharomyces cerevisiae described in "Japanese Guidelines for Recombinant DNA Research" may be used as host, but S288C strain and other strains are practically suitable. On the other hand, YEp13 and other vectors were appropriate as a vector. As promoter, the ADHI gene coding for alcohol dehy- drogenase gene, and other promoters, are suitable. 50 Example 2: Elastase IIB

1) Separation of mRNA from human pancreas

I5 Human pancreas (autopsy sample) weighing 5.5 g was homogenized and denatured in a homogeniser (a Polytron homogeniser from Kinematica GmbH, Germany, Polytron being a Trade Mark) in an adequate amount of guanidine thiocyanate solution (4M guanidine thiocyanate; 1% detergent, Sarcosyl from Sigma Chemical Co, USA; 20mM ethylenediaminetetraacetic acid (EDTA); 25mM sodium citrate, pH 7.0; 100mM 2-mercaptoethanoi and 0.1% antifoaming agent, Antifoam A from Sigma Chemical Co USA), followed by >0 centrifugation, to obtain a supernatant. One volume of supernatant was added to 0.025 volumes of 1 M acetic acid and 0.75 volumes of ethanol, and the mixture cooled to -20°C for several hours, followed by centrifugation, to obtain a precipitate. The precipitate was suspended in a guanidine hydrochloride solution (7.5M guanidine hydrochloride; 25mM sodium citrate, pH 7.0; and 5mM dithiothreitol, DTT). To 1 volume of the suspension was add- i5 ed 0.025 volumes of IM acetic acid and 0.5 volumes of ethanol, and the resultant mix cooled to -20°C for

4 EP 0 198 645 B1

several hours, followed by centrifugation. The resultant precipitate was suspended again in more of the guanidine hydrochloride solution, mixed as before with acetic acid and ethanol, cooled to -20°C and centrifuged, followed by collection of the precipitate. Next, the precipitate was washed several times with ethanol to remove guanidine hydrochloride, and then dissolved in distilled water, followed by precipita- 5 tion of the RNA with ethanol. The precipitate was collected by centrifugation to give 53.9 mg of RNA. The RNA thus obtained was adsorbed onto an oligo(dT) cellulose column in a high concentration salt solution (0.5M NaCI; 20mM Tris-HCI, pH 7.5; 1mM EDTA; and 0.1% sodium dodecyl sulfate; SDS), and the mRNA-containing poly(A) was eluted with an elution buffer (10mM Tris-HCI, pH7.5; ImM EDTA; and 0.05% SDS) to give 255 ug of mRNA. 10 2) Preparation of cDNA bank

The DNA from 4,700 clones of the human pancreatic cDNA bank was fixed onto nitrocellulose filter according to the method of Grunstein, M and Hogness, D S [Proc Natl Acad Sci, USA, 72, 3961 (1975)]. 55 In order to select a clone including a plasmid containing human pancreatic elastase IIB cDNA, a rat pancreatic elastase II cDNA fragment was employed as a probe. The rat pancreatic elastase II cDNA was obtained by the following method. Firstly a cDNA bank was prepared using mRNA extracted from rat pancreas. Then, an oligodeoxynucleotide 5'd(GGCATAGTCCACAACCA)3' comprising 17 bases was synthesized complementary to a part of the known base sequence of rat pancreatic elastase II mRNA 60 [MacDonald, R et al: Biochem, 21, 1453 (1982)]. The oligodeoxynucleotide corresponds to the amino acid sequence from amino acid 151 to amino acid 156 of the rat pancreatic elastase II. The 5'-end of the oligodeoxynucleotide was labelled with 32p by the method of Maxam, A M and Gilbert, W [Proc Natl Acad Sci, USA 74> 560 (1977)] to give a probe. Using the probe, a rat pancreatic elastase II cDNA clone was selected from the rat pancreatic cDNA bank. Then a fragment containing 200 bases was excised from the 65 rat pancreatic elastase II cDNA with restriction endonucleases Aatll and Hindlll and labelled with 32P

15 EP 0198 645 B1

according to tne nicK translation method [Rigby, P W et al: J Mol Biol 113, 237 (1977)]. The DNA fragment was then used as a probe for hybridization with the DNA fixed on the nitrocellulose filter, according to the method of Alwine, S C et a[ [Proc Natl Acad Sci, USA 74, 5350 (1 977)]. 8 clones reactive with the probe were identified by autoradiography. A plasmid from one of the strains 5 was named "pH2E8". The restriction endonuclease cleavage map of the DNA inserted in pH2E8 was determined and is shown in Figure 6. The base sequence was determined and is of the formula (5')-C TTA CAG AAC TCC CAC GGA CAC ACC ATG ATT AGG ACC CTG CTG 1 0 CTG TCC ACT TTG GTG GCT GGA GCC CTC AGT TGT GGG GTC TCC ACT TAC GCG CCT GAT ATG TCT AGG ATG CTT GGA GGT GAA GAA GCG AGG CCC AAC AGC TGG CCC TGG CAG GTC TCC CTG CAG TAC AGC TCC AAT GGC CAG TGG TAC CAC 15 ACC TGC GGA GGG TCC CTG ATA GCC AAC AGC TGG GTC CTG ACG GCT GCC CAC TGC ATC AGC TCC TCC AGG ATC TAC CGC GTG ATG CTG GGC CAG CAT AAC CTC TAC GTT GCA GAG TCC GGC TCG CTG GCC GTC AGT GTC TCT AAG ATT GTG 20 GTG CAC AAG GAC TGG AAC TCC AAC CAG GTC TCC AAA GGG AAC GAC ATT GCC CTG CTC AAA CTG GCT AAC CCC GTC TCC CTC ACC GAC AAG ATC CAG CTG GCC TGC CTC CCT CCT GCC GGC ACC ATT CTA CCC AAC AAC TAC CCC TGC TAC 25 GTC ACA GGC TGG GGA AGG CTG CAG ACC AAC GGG GCT CTC CCT GAT GAC CTG AAG CAG GGC CGG TTG CTG GTT GTG GAC TAT GCC ACC TGC TCC AGC TCT GGC TGG TGG GGC AGC ACC GTG AAG ACG AAT ATG ATC TGT GCT GGG GGT GAT 30 GGC GTG ATA TGC ACC TGC AAC GGA GAC TCC GGT GGG CCG CTG AAC TGT CAG GCA TCT GAC GGC CGG TGG GAG GTG CAT GGC ATC GGC AGC CTC ACG TCG GTC CTT GGT TGC AAC TAC TAC TAC AAG CCC TCC ATC TTC ACG CGG GTC TCC 35 AAC TAC AAC GAC TGG ATC AAT TCG GTG ATT GCA AAT AAC TAA CCA AAA GAA GTC CCT GGG ACT GTT TCA GAC TTG GAA AGG TCA CAG AAG GAA AAT AAT ATT ATA TAA AGT GAC AAC TAT GCA AAT CAC -(3') iO Beginning at the underlined ATG triplet, the encoded amino acid sequence is of the formula: (N)- Met He Arg Thr Leu Leu Leu Ser Thr Leu Val Ala Gly Ala Leu Ser Cys Gly Val Ser Thr Tyr Ala Pro Asp Met Ser Arg Met Leu Gly Gly Glu Glu Ala Arg *5 Pro Asn Ser Trp Pro Trp Gin Val Ser Leu Gin Tyr Ser Ser Asn Gly Gin Trp Tyr His Thr Cys Gly Gly Ser Leu He Ala Asn Ser Trp Val Leu Thr Ala Ala His Cys He Ser Ser Ser Arg lie Tyr Arg Val Met Leu Gly 50 Gin His Asn Leu Tyr Val Ala Glu Ser Gly Ser Leu Ala Val Ser Val Ser Lys lie Val Val His Lys Asp Trp Asn Ser Asn Gin Val Ser Lys Gly Asn Asp lie Ala Leu Leu Lys Leu Ala Asn Pro Val Ser Leu Thr Asp Lys >5 lie Gin Leu Ala Cys Leu Pro Pro Ala Gly Thr He Leu Pro Asn Asn Tyr Pro Cys Tyr Val Thr Gly Trp Gly Arg Leu Gin Thr Asn Gly Ala Leu Pro Asp Asp Leu Lys Gin Gly Arg Leu Leu Val Val Asp Tyr Ala Thr Cys 30 Ser Ser Ser Gly Trp Trp Gly Ser Thr Val Lys Thr Asn Met lie Cys Ala Gly Gly Asp Gly Val lie Cys Thr Cys Asn Gly Asp Ser Gly Gly Pro Leu Asn Cys Gin Ala Ser Asp Gly Arg Trp Glu Val His Gly lie Gly Ser 35 Leu Thr Ser Val Leu Gly Cys Asn Tyr Tyr

6 EP0198 645 B1

Tyr Lys Pro Ser He Phe Thr Arg Val Ser Asn Tyr Asn Asp Trp lie Asn Ser Val lie Ala Asn Asn - (C) It can be seen that plasmid pH2E8 has a region coding for a signal peptide comprising 16 amino acids 5 and a region coding for the mature elastase protein termed elastase IIB. The elastase IIB comprises 253 amino acids. Plasmid pH2E8 also contains 5' and 3' non-translated regions. There is as much as 77% homology between the amino acid sequence coded by the DNA insertion of plasmid pH2E8, and that of rat pancreatic elastase II reported by MacDonald, R J et a[ [Biochem, 21, 1453 (1982)]. Accordingly, plasmid pH2E8 is believed to contain cDNA coding a human pancreatic 10 elastase II. With regard to the known sequence for human pancreatic elastase II, the 16 amino acid residue at the N-terminal of pH2E2 coding protein is clearly different to that of the human pancreatic elastase II reported by Largman, C, etal [Biochim Biophys Acta 623, 208 (1980)]. Thus the human pancreatic elastase IIB appears to be a novel protein. Since pH2E8 contains an entire DNA coding for human pancreatic elastase IIB, it is possible to pro- 15 duce large amounts of the elastase IIB by transferring the cDNA to an appropriate expression vector and using, for example, E. coli. B. subtilis. yeast or mammalian cells as the host. 4) Production of human pancreatic elastase IIB using mammalian cells 20 Construction of expression plasmid PSV2-E2B

In order to produce human elastase IIB using mammalian cells as host, the cDNA was ligated to an expression vector according to the procedure indicated in figure 7. The known pSV2 plasmid containing the SV40 promoter, enhancer, poly(A) signal and introns of the Small T antigen gene was used for con- 25 struction of the expression vector. pSV2-p-globin is a plasmid based on the pSV2 vector with inserted p-globin cDNA. Plasmid pH2E8 is the plasmid with inserted human elastase IIB cDNA. The reactions with S1 nuclease, etc. were carried out according to the methods described in "Molecular Cloning" [Maniatis, T et al (ed) "Molecular Cloning" Cold Spring Harbor Lab (1982)]. The broad arrows in figure 7 represent the promoter derived from SV40, and the narrow arrows show the direction of transcription. 30 From analysis of the restriction endonuclease digestion pattern, an expression plasmid (pSV2-E2B) was selected in which vector and cDNA were ligated in the correct orientation for transcription.

Transfection of COS 1 cells with expression plasmid pSV2-E2B

35 The constructed expression plasmid pSV2-E2B was transfected into COS 1 (mammalian cells) by the calcium phosphate method described in the literature [Graham and Van der Eb: Virology 52 456 (1 973)]. Thus, 1 x 106 of COS 1 cells were sprayed on Petri dishes (10 cm in diameter), and cultured overnight on Dulbecco-modified Eagle medium containing 10% fetal calf serum. Plasmid pSV2-E2B (300 ug) was suspended in sterile distilled water (12.55 ml), then 2.5 M CaCl2 (0.75 40 ml) was added and the suspension mixed thoroughly. A pipette was used to bubble air through the system in order to maintain agitation, and 1.5 ml of 10 x HeBS solution (HEPES, 210 mM; NaCI, 1.37 M; KCI 4.7 mM; Na2HP04, 10.6 mM; glucose, 55.5 mM; pH 7.05) was dropped into the resulting solution to precipitate the DNA and calcium phosphate. The precipitate was allowed to stand at room temperature for 30 minutes to mature the precipitate, and 45 1 ml of the solution for each Petri dish was added to the COS 1 cells cultured in fresh medium containing 10% fetal calf serum. After cultivation of these cells for 12 hours at 37°C in the presence of 5% CO2, the culture medium was discarded in favour of a fresh, Dulbecco-modified Eagle medium containing no fetal calf serum. The cells were cultured for a further 48 hours at 37°C in the presence of 5% C02. The transfected COS 1 50 cells obtained by this procedure were tested to detect the presence of human elastase IIB mRNA, while the supernatant of the culture medium was tested for elastase activity. Extraction of mRNA from COS 1 cells

55 In order to confirm the existence of human elastase IIB mRNA transcribed from the expression plasmid in the transfected COS 1 cells, mRNA was extracted from the COS 1 cells and assayed by the North- ern blot hybridization according to the following procedure. After cultivation of the COS 1 cells for 48 hours, 1 ml of guanidine thiocyanate solution (4M guanidine thiocyanate; 1% Sarcosyl; 20 mM EDTA; 25 mM sodium citrate, pH 7.0; 100 mM 2-mercaptoethanol; 0.1% 60 Antifoam A) was added to each Petri dish in order to lyse the cells. High molecular DNA molecules were degraded to low molecular ones by passing the solution several times through a 21 guage injection needle. The resultant solution was layered on a solution containing 5.7 M cesium chloride and 0.1 M EDTA, and the layered solution centrifuged at 20°C, 30,000 rpm for 17 hours using a Hitachi RPS 40 swing rotor. The resultant RNA precipitate was washed with a small 65 amount of ethanol and dissolved in distilled water (300 ul).

17 EP 0198 645 B1

Hccoraing ro me metnoa or aviv ana Leder [Proc Natl Acad Sci, USA 69, 1408 (1972)], the extracted total RNA was purified on an oligo(dT) cellulose column to give several ug of purified mRNA. Half of the purified mRNA was used for Northern blot hydridization by the Thomas method [Proc Natl Acad Sci, USA 77 5201 (1980]. 32p-Labelled human elastase IIB cDNA was used as a probe by the nick translation 5 method [Rigby, P W etal: J Mol Biol 113, 237 (1977)]. A major band at about 1 .8 Kb and a minor band at about 1.0 Kb which hydridized with the probe were detected only in the mRNA of COS 1 cells transfected with plasmid pSV2-E2B. It was assumed that in the transcription of the plasmid pSV2-E2B, the 1.8 Kb mR- NA is formed on termination of transcription at the poly(A) signal contained in the vector, while the 1.0 Kb mRNA is formed on termination of transcription at the poly(A) signal in the cDNA. The results from the 1 o Northern blot hybridization coincided with these assumptions. The results thus show that using SV40 promoter in COS 1 cells, the plasmid pSV2-E2B can synthesize large quantities of human elastase IIB mRNA. Elastase activity in supernatant of the culture broth 15 Since the human elastase IIB cDNA in pSV2-E2B has a signal peptide region, secretion of expressed elastase into the medium as a proelastase was expected. Elastase activity in the medium after 48 hours cultivation was accordingly determined. Tris-HCI buffer solution (pH 8.0) was added to the supernatant of the culture medium (1 ml) to a final concentration of 0.2 M, 10 mg/ml trypsin (50 ul) was added, and an 20 activation treatment of the proelastase was carried out for 15 minutes at 25°C. Next, 50 ul of soybean trypsin inhibitor solution (10 mg/ml) and a synthetic substrate, glutaryl-Ala-Ala-Pro-Leu-p-nitroanilide were added, and the enzyme reaction was carried out by incubation at 25°C. After completion of the reaction, liberated p-nitroanilide was determined by measuring absorbance at 410 nm. The elastase activity in the culture medium of COS 1 cells transfected with pSV2-E2B was distinctly 25 detected, and the production of active human elastase IIB in COS 1 cells was found. The activation treatment with trypsin or other activator was indispensable. The production of human elastase IIB in COS 1 cells transfected with plasmid pSV2-E2B was tran- sient. However, if plasmid pSV2-E2B is linked with a suitable selection marker (for example, the neo gene, dihydrofolate reductase gene, or other resistance marker) and is transfected in to CHO or other 30 cells, a cell line suitable for sustained production of human elastase IIB can be prepared. 5) Production of human pancreatic elastase IIB using Bacillus subtilis Construction of expression vector 0BSEX-E2B 35 The construction method for the expression vector is illustrated in figure 8. Plasmid pH2E8 containing human elastase IIB cDNA was digested with restriction endonuclease Hindlll to a linear DNA, and then partially digested with restriction enzyme Pvull. A DNA fragment of about 1.3 Kb containing the elastase IIB cDNA was isolated. This fragment was partially digested with restriction endonuclease Hhal to give a 10 fragment containing cDNA coding for the amino acid sequence beginning from the eighth amino acid counted from the N-terminal end of the elastase IIB activation peptide. The specific oligonucleotide sequence shown in Figure 8 was synthesised. This sequence codes for part of the signal peptide of B. suf> tiNjs a-amylase, and seven amino acids at the N-terminal end of the elastase IIB activation peptide The fragments were ligated with T4 DNA ligase. 15 On the other hand, plasmid pTUB228 [Ohmura, K eta]: J Biochem 95 87 (1 984)] containing the a-amylase gene of Bacillus subtilis was digested with restriction endonuclease Hindlll to isolate a DNA fragment of 428 base pairs containing the a-amylase promoter and a part of the signal peptide, and separately a DNA fragment of 5,100 base pairs containing the replication origin. The DNA fragment of 428 base pairs and the DNA fragment of 5,100 base pairs were further digested with restriction endonuclease Hpall >o and Smal, respectively, to give DNA fragments of 385 base pairs and of 4566 base pairs. The three DNA fragments were ligated with T4 DNA ligase, and incorporated into protoplasts of Bacil- Jus subtilis 207 - 25 strain (rrrf68 hsrM recE4 amyE07 aro1906 leuA8 Iys21 ; derived from Marburg strain) according to conventional procedures. After regeneration, cultivation on a medium containing 10 ug/ml of kanamycin allowed selection 5 of the transformed strains which could grow on this medium. Selection of the desired plasmid was achieved through colony hybridization using as probe the synthetic oligonucleotide of figure 9. A positive clone was identified and the plasmid shown to be the intended one by determination of the base sequence. The expression plasmid obtained in this manner was designated plasmid pBSEX-E2B. l0 Elastase activity in supernatant of culture broth

Bacillus subtilis 207 - 25 strain transformed with the human elastase IIB expression plasmid pBSEX- E2B was cultured on a reciprocal shaker in 1 1 of LG medium (1% Bacto trypton, Difco; 0.5% yeast extract, Difco; 0.5% NaCI; 0.2% glucose; pH 7.0) containing 50 ug/ml of kanamycin. The cuituring was per-

a EP 0 198 645 B1

formed at 35°C for 48 hours. After completion of the culture, the culture medium was cooled to 4°C and centrifuged at 3000 x G for 5 minutes, and the cells discarded. Ammonium sulfate was added to the supernatant to 55% saturation, and the solution was stirred at 4°C for 12 hours. Insoluble material formed in this treatment was precipi- 5 tated by centrifugation at 8000 x G for 20 minutes, the supernatant was discarded, and the precipitate was dissolved in 20 ml of 0.2 M Tris-HCI buffer solution (pH 8.0). This solution was dialyzed against 1 1 of 0.2 M Tris-HCI buffer solution (pH 8.0) for 16 hours, and insoluble material was removed by centrifugation at 8000 x G for 20 minutes. The dialyzed inner solution was taken as a crude elastase IIB sample solution, and the activity was determined by the following procedure. 10 Determination of Elastase Activity in the Sample Solution

Proelastase was activated by trypsin treatment of the sample solution, and allowed to react at 25°C with a synthetic substrate, glutaryl-Ala-Ala-Pro-Leu-p-nitroanilide. Liberated p-nitroanilide was deter- 15 mined by measuring absorbance at 410 nm. Since elastase activity was distinctly detected in the sample solution from the 207-25 strain transformed with pBSEX-E2B, the production of active human elastase IIB in Bacillus subtilis was demonstrated. 6) Production of human pancreatic elastase IIB using Escherichia coli 20 Various different promoters may be used, including for instance the tryptophan (trp) promoter, lactose (lac) promoter, tryptophan-lactose (tac) promoter, lipoprotein (ipp) promoter, lambda (X) Pl promoter derived from bacteriophage, or polypeptide chain elongation factor (tuf B) promoter. An example with the lactose promoter is as follows. 25 As shown in figure 10, plasmid pH2E8 containing human elastase IIB cDNA was digested with restriction endonuclease Hindlll to form a linear DNA, and then partially digested with restriction enzyme Pvull to isolate a DNA fragment of about 1 .3 Kb containing elastase IIB cDNA. This fragment was partially digested with restriction endonuclease Ddel to give a fragment containing cDNA coding for an amino acid sequence down stream from the activation peptide of elastase IIB. By 30 treatment of this DNA fragment with DNA polymerase Klenow fragment and four kinds of deoxyribonuci- eoside 5'-triphosphates (dATP), dCTP, dGTP, dTTP), both ends were converted to blunt ends. The resultant DNA fragment was inserted into the Smal site of plasmid pUC8. Escherichia coli strain JM103 was transformed with the plasmid. Plasmid was extractd from several transformed strains, and through the use of restriction endonuclease mapping, plasmid with the same ori- 35 entation of cDNA transcription and of the promoter for the lactose operon was selected. This elastase IIB expression plasmid was designated as pH2EX-8. The base sequence and corresponding amino acid sequence for around the 5'-end of the elastase cD- NA in pH2EX-2 are as follows. 40 P-galactosidase ^elastase

Thr Met lie Thr Asn Ser Leu Ser Cys Gly Asp ... 45 ATG ACC ATG ATT ACG AAT TCC CTC AGT TGT GGG GTC ...

Therefore, the human elastase IIB expressed by pH2EX-8 is shown to be a fusion protein in which two 50 amino acids (Leu, Ser) derived from the signal peptide and six amino acids derived from p-galactosidase combine at the N-terminal. Several Escherichia coli strains were transformed with pH2EX-8 to give strains capable of producing human elastase IIB. The strains containing expression plasmid were inoculated in 2 x TY-ampicillin medium (1.6% Bacto trypton; 1% yeast extract; 0.5% NaCI; 50 ug/ml ampicillin) and cultured at 37°C for 15 55 hours. After completion of the culture, the cell suspension was harvested by centrifugation of the culture medium, 2.4 x 108 cells were suspended in 15 ul of SDS solution (2% SDS, 5% 2-mercaptoethanol; 10% glycerin; 60 mM Tris-HCI, pH 6.8). The cell suspension was heated at 100°C for 3 minutes, and protein was analyzed by SDS polyacrylamide gel electrophoresis, according to the method of Laemmli et al [Nature, 227 680 (1970)]. 60 As a result, a large quantity of elastase production was found in strain YA21 . The production yield of elastase fusion protein was 20% of the total protein produced by the YA21 strain. Though a considerable production of elastase fusion protein in x984 strain was observed, the production yield was less than half that in the YA21 strain. The production yields in two other Escherichia coli strains (HB101 and MC4100 strains) were low. 65 Since the eiastase fusion protein forms inclusion bodies in cells, purification was comparatively easy.

19 EP 0 198 645 B1

Thus, from 1 I of a culture medium of YA21 strain transformed with plasmid pH2EX-8, 6.5 g of cells forming inclusion bodies could be obtained. The obtained cells (6.5 g) were lysed in 50 mM Tris-HCI buffer solution (pH 8.0) containing 0.2 mg/ml lysozyme and 1 mg/ml deoxycholic acid. Intact cells were removed by low speed centrifugation (1,500 x G, 10 minutes) and inclusion bodies were recovered as a precipitate by 5 high speed centrifugation (11,000 x G, 20 minutes). Since the precipitate still contained much cell debris, the precipitate was suspended in 50 mM Tris-HCI buffer solution containing 5 mg/ml octylphenoxy poly- ethoxyethanol surfactant (Triton X-100, Trade Mark) and washed by high speed centrifugation (11,000 x G, 20 minutes). A washed pellet of inclusion bodies was suspended in a small amount of Tris-HCI buffer solution and stored at 4°C. 10 The purified inclusion body (300 mg) containing about 50% elastase IIB fusion protein could be obtained by this procedure. The inclusion body produced by YA21 strain transformed with pH2EX-8 was shown to contain the elastase fusion protein by immuno-blotting. Though most of the elastase produced by Escherichia coli is in the insoluble fraction of the cells as inclusion bodies, a smaller proportion exists in a soluble state and retains enzyme activity. Determination 15 of the enzyme activity was carried out as follows. The strain x984 strain transformed with plasmid pH2EX-8 was cultured with shaking in 1 I of 2 x TY-ampicillin medium at 37°C for 15 hours. After completion of the cuituring, the cells were harvested by centrifugation at 3000 x G for 5 minutes and suspended in 20 ml of buffer solution A (50mM Tris-HCI, pH 8.0; 1mM EDTA; 50 mM NaCI). 10 mg of lysozyme was added, and the suspension incubated at 5°C for 20 minutes. Deoxycholic acid to a final concentra- 20 tion of 1 mg/ml was added to the resultant suspension which was then incubated at 20°C. Deoxyribonucle- ase to a final concentration of 0.1 mg/ml was added, and the cells were disrupted in a Polytron homogenizer. After removal of the ceil debris by centrifugation at 80,000 x G, 40 minutes, the lysate was subjected to column chromatography on high molecular weight, cross-linked dextran (Sephadex G-75 ). The fractions having elastase activity were purified by antibody affinity chromatography, and used for the 25 following determination of elastase activity. Proelastase was activated by trypsin treatment of the sample solution. Next, p-nitroanilide liberated by the reaction with a synthetic substrate, glutaryl-Ala-AIa-Pro-Leu-p-nitroanilide, was determined by measuring the absorbance at 410 nm. The reaction was carried out by incubation at 25°C in 0.2 M Tris- HCI buffer solution (pH 8.0) containing the synthetic substrate. As a result, elastase activity was dis- 30 tinctly detected in the sample solution of x984 strain transformed with plasmid pH2EX-8, and therefore the production of active human elastase IIB in B coli was demonstrated. (7) Production of human pancreatic elastase IIB using veast

35 In the case of yeast as a host (and as with a Escherichia coli. Bacillis subtilis or mammalian cells), human pancreatic elastase IIB cDNA could be ligated to a suitable expression vector using known techniques, introduced into the host cells and expressed. Elastase activity in the culture medium was confirmed. Saccharomvces cerevisiae described in "Japanese Guidelines for Recombinant DNA Research" may 40 be used as host, but S288C strain and other strains are practically suitable. On the other hand, YEp13 and other vectors were appropriate as a vector. As promoter, the ADH1 gene coding for alcohol dehy- drogenase gene, and other promoters, are suitable. Example 3: Elastase IIIA 45 1) Separation of mRNA for human pancreas

Human pancreas (autopsy sample) weighing 5.5 g was homogenised and denatured in a homogeniser (a Polytron homogeniser from Kinematica GmbH, Germany, Polytron being a Trade Mark) in an adequate 50 amount of guanidine thiocyanate solution (4M guanidine thiocyanate; 1% detergent, Sarcosyl from Sigma Chemical Co, USA; 20mM ethylenediaminetetraacetic acid (EDTA); 25mM sodium citrate, pH 7.0, 100mM 2-mercaptoethanol and 0.1% antifoaming agent, Antifoam A from Sigma Chemical Co USA), followed by centrifugation, to obtain a supernatant. One volume of supernatant was added to 0.025 volumes of 1 M acetic acid and 0.75 volumes of etha- 55 nol, and the mixture cooled to -20°C for several hours, followed by centrifugation, to obtain a precipitate. The precipitate was suspended in a guanidine hydrochloride solution (7.5M guanidine hydrochloride; 25mM sodium citrate, pH 7.0; and 5mM dithiothreitol, DTT). To 1 volume of the suspension was added 0.025 volumes of 1 M acetic acid and 0.5 volumes of ethanol, and the resultant mix cooled to -20°C for several hours, followed by centrifugation. The resultant precipitate was suspended again in more of the so guanidine hydrochloride solution, mixed as before with acetic acid and ethanol, cooled to -20°C and centrifuged, followed by collection of the precipitate. Next, the precipitate was washed several times with ethanol to remove guanidine hydrochloride, and then dissolved in distilled water, followed by precipitation of the RNA with ethanol. The precipitate was collected by centrifugation to give 53.9 mg of RNA. The RNA thus obtained was adsorbed onto an oligo(dT) cellulose column in a high concentration salt 35 solution (0.5M NaCI; 20mM Tris-HCI, pH 7.5; 1mM EDTA; and 0.1% sodium dodecyl sulfate, SDS), and

>0 EP 0 198 645 B1

the mRNA-containing poly(A) was eluted with an elution buffer (10mM Tris-HCI, pH7.5; 1mM EDTA; and 0.05% SDS) to give 255 ug of mRNA. 2) Preparation of cDNA bank 5 The preparation of a cDNA bank was carried out according to the Okayama-Berg method. 5 ug of the mRNA and 24 units of a reverse transcriptase were incubated at 42°C for 60 minutes in 20 ul of a reaction solution (50mM of Tris-HCI, pH 8.3; 8mM of MgCI2; 30mM of KCI; 0.3mM of DTT; 2mM each of dATP, dGTP, dCTP and dTTP; 10 uCi of a-szp-dCTP; and 1.4 ug of vector primer DNA, purchased from 10 PL-Pharmacia). The reaction was stopped by the addition of 2 ul of 0.25M EDTA and 1 ul of 10% SDS. Deproteiniza- tion was carried out with 20 ul of phenol-chloroform. After centrifugation of the reaction mixture, the aqueous layer was collected. 20 ul of 4M ammonium acetate and 80 ul of ethanol were added to the aqueous layer and the mixture cooled to -70°C for 15 minutes. The precipitate was then collected by centrifu- 15 gation and washed with 75% aqueous ethanol and dried under reduced pressure. The resultant precipitate was dissolved in 15 ul of terminal transferase reaction solution (140mM po- tassium cacodylate, of formula C2H6ASKO2; 30 mM Tris-HCI, pH 6.8; 1mM cobalt chloride; 0.5mM DTT; 0.2 ug of poly(A); and 100mM of dCTP). After incubating the solution at 37°C for 3 minutes, 18 units of terminal deoxynucleotidyl transferase were added and allowed to react for 5 minutes. 20 The above reaction was stopped by addition of 1 ul of 0.25M EDTA and 0.5 ul of 1 0% SDS. Deproteini- zation was then carried out with phenol-chloroform. After centrifugation of the reaction mixture, the aqueous layer was collected. 15 ul of 4M ammonium acetate and 60 uJ of ethanol were added to the aqueous layer and mixed well. After cooling the mixture to -70°C for 15 minutes, the precipitate was collected by centrifugation. 25 The precipitate was dissolved in a buffer for restriction endonuclease (50mM NaCI; 10mM Tris-HCI, pH 7.5; 10mM MgCl2; and 1mM DTT). 2.5 units of the restriction enzyme Hindlll were added to the solution and the solution incubated at 37°C for about 1 hour. After deproteinization with phenol-chloroform, precipitation was carried out with ethanol. After cooling to -70°C for 15 minutes, the precipitate was collected by centrifugation and dissolved in 30 10 ul of TE (10mM Tris-HCI, pH 7.5; 1mM EDTA). 1 ul of the above sample solution was then added to a reaction solution (10mM Tris-HCI, pH 7.5; 1mM EDTA; and 100mM NaCI), followed by addition of 10 ng of linker (purchased from PL-Pharmacia) bearing oligo dG. The resulting mixture was heated at 65°C for 5 minutes and maintained at 42°C for 30 minutes. After cooling the reaction solution in ice water, 10 ul of 10-fold ligase buffer solution (10mM ATP; 35 660mM Tris-HCI, pH 7.5; 66mM MgCI2; and 100mM DTT), 78 u.1 of distilled water and 8 units of T4 DNA concentrated ligase were added to the solution and the reaction mixture maintained at 1 2°C overnight. Then, to the thus prepared mixture, 1 0 ul of 1 M KCI, 1 unit of ribonuclease H, 33 units of DNA polymerase I, 4 units of T4 DNA ligase, 0.5 ul of nucleotide solution (20mM dATP, 20 mM dGTP, 20mM dCTP and 20mM dTTP) and 0.1 ul of 50 ug/ul bovine serum albumin (BSA) were added and the mixture main- 40 tained at 1 2°C for 1 hour and then at 25°C for 1 hour. After diluting the reaction solution 5-fold, E coji strain RR1 was transformed according to the method of Hanahan [Hanahan, D; J Mol Biol 166, 557 (1983)] to prepare a human pancreatic cDNA bank. 3) Selection of transformed bacteria containing human pancreatic elastase IIIA cDNA 45 The DNA from 4,700 clones of the human pancreatic cDNA bank was fixed onto nitrocellulose filter according to the method of Grunstein, M and Hogness, D S [Proc Natl Acad Sci, USA, 72, 3961 (1975)]. A porcine pancreatic elastase I cDNA fragment was excised with restriction endonuclease Pstl and labelled with 32p according to the nick translation method [Rigby, P W et §]:J. Mol. Biol 113, 237 (1977)]. 50 Using the DNA fragment as probe, it was hybridized with the DNA fixed on the filter at 35°C in a solution of 40% formamide, 0.75 M sodium chloride and 0.075 M sodium citrate, along with a solution of 0.1% BSA, 0.1% polysucrose (Ficoll, Trade Mark), 0.1% polyvinyl pyrrolidone, 0.5% SDS, 10 mM sodium phosphate and 1 00 ug/ml of salmon spermatozoon DNA. 7 microorganism strains which hybridized with the probe were identified by autoradiography. A plasmid 55 from one of the strains was named pCL2, and the inserted cDNA was named CL2.

4) The amino acid sequence and elastase IIIA coded by plasmid pCL2

In CL2, there exists only one open reading frame. This sequence codes for 270 amino acids. The en- 60 coded amino acid sequence showed some homology with trypsin and chymotrypsin, to a degree of only about 30%. However, there was a higher homology, 57%, with porcine pancreatic elastase I. There also existed an amino acid sequence (Gly-Asp-Ser-Gly-Gly-Pro) around the active center which is be com- monly found in the same region in serine proteases such as trypsin, chymotrypsin, etc. Further, there is a 45th histidine residue, 95th aspartic acid residue and 189th serine residue which can participate in the 65 characteristic charge relay of serine proteases.

21 EP 0 198 645 B1

moreover, tne encoded amino acid sequence adjacent the carboxyl terminal, which in elastase forms a pocket for binding with the substrate is very similar between human elastase IIIA and the porcine elastase I. There is also an encoded 211th valine residue and 223rd threonine residue, which are characteristic of a substrate-binding pocket of the porcine elastase I. 5 The base sequence was of the formula (5")- ATC ATC ACA AAA CTC ATG ATG CTC CGG CTG CTC AGT TCC CTC CTC CTT GTG GCC GTT GCC TCA GGC TAT GGC CCA CCT TCC TCT CAC TCT TCC AGC CGC 1 0 GTT GTC CAT GGT GAG GAT GCG GTC CCC TAC AGC TGG CCC TGG CAG GTT TCC CTG CAG TAT GAG AAA AGT GGA AGC TTC TAC CAC ACG TGT GGC GGT AGC CTC ATC GCC CCC GAT TGG GTT GTG ACT GCC GGC CAC TGC ATC TCG AGG GAT 1 5 CTG ACC TAC CAG GTG GTG TTG GGT GAG TAC AAC CTT GCT GTG AAG GAG GGC CCC GAG CAG GTG ATC CCC ATC AAC TCT GAG GAG CTG TTT GTG CAT CCA CTC TGG AAC CGC TCG TGT GTG GCC TGT GGC AAT GAC ATC GCC CTC ATC AAG 20 CTC TCA CGC AGC GCC CAG CTG GGA GAT GCC GTC CAG CTC GCC TCA CTC CCT CCC GCT GGT GAC ATC CTT CCC AAC AAG ACA CCC TGC TAC ATC ACC GGC TGG GGC CGT CTC TAT ACC AAT GGG CCA CTC CCA GAC AAG CTG CAG CAG GCC 25 CGG CTG CCC GTG GTG GAC TAT AAG CAC TGC TCC AGG TGG AAC TGG TGG GGT TCC ACC GTG AAG AAA ACC ATG GTG TGT GCT GGA GGG TAC ATC CGC TCC GGC TGC AAC GGT GAC TCT GGA GGA CCC CTC AAC TGC CCC ACA GAG GAT GGT 30 GGC TGG CAG GTC CAC GGT GTG ACC AGC TTT GTT TCT GCC TTT GGC TGC AAC TTC ATC TGG AAG CCC ACG GTG TTC ACT CGA GTC TCC GCC TTC ATC GAC TGG ATT GAG GAG ACC ATA GCA AGC CAC TAG AAC CAA GGC CCA GCT GGC AGT 35 GCT GAT CGA TCC CAC ATC CTG AAT AAA GAA TAA AGA TCT CTC AGA AAA -(3') Beginning with the underlined ATG triplet, the encoded amino acid sequence is of the formula: (N)-Met Met Leu Arg Leu Leu Ser Ser Leu Leu Leu Val Ala Val Ala Ser GLy Tyr 40 Gly Pro Pro Ser Ser His Ser Ser Ser Arg Val Val His Gly Glu Asp Ala Val Pro Tyr Ser Trp Pro Trp Gin Val Ser Leu Gin Tyr Glu Lys Ser Gly Ser Phe Tyr His Thr Cys Gly Gly Ser Leu lie Ala Pro Asp Trp Val 45 Val Thr Ala Gly His Cys lie Ser Arg Asp Leu Thr Tyr Gin Val Val Leu Gly Glu Tyr Asn Leu Ala Val Lys Glu Gly Pro Glu Gin Val He Pro He Asn Ser Glu Glu Leu Phe Val His Pro Leu Trp Asn Arg Ser Cys Val 50 Ala Cys Gly Asn Asp He Ala Leu He Lys Leu Ser Arg Ser Ala Gin Leu Gly Asp Ala Val Gin Leu Ala Ser Leu Pro Pro Ala Gly Asp lie Leu Pro Asn Lys Thr Pro Cys Tyr lie Thr Gly Trp Gly Arg Leu Tyr Thr Asn 55 Gly Pro Leu Pro Asp Lys Leu Gin Gin Ala Arg Leu Pro Val Val Asp Tyr Lys His Cys Ser Arg Trp Asn Trp Trp Gly Ser Thr Val Lys Lys Thr Met Val Cys Ala Gly Gly Tyr lie Arg Ser Gly Cys Asn Gly Asp Ser Gly 30 Gly Pro Leu Asn Cys Pro Thr Glu Asp Gly Gly Trp Gin Val His Gly Val Thr Ser Phe Val Ser Ala Phe Gly Cys Asn Phe lie Trp Lys Pro Thr Val Phe Thr Arg Val Ser Ala Phe lie Asp Trp lie Glu Glu Thr He Ala 35 Ser His -(C)

!2 EP0 198 645 B1

The amino acid sequence of the coded protein appears at a glance to be analogous with the amino acid composition of the human pancreatic elastase I reported by Largman et a] [Largman, C etal: Biochemistry 15, 2491 (1976)3, However, the present elastase was clearly different in detail from the known one, and the molecular weights were also different. The isoelectric point calculated according to the amino acid se- 5 quence was in the range of 7.5 to 7.8. This value was similar to the isoelectric point of 7.6 of the human pancreatic proelastase I reported by Scheele, G et al [Scheele et a]: Gastroenterology 80, 461 (1 981 )]. The active form of the human elastase may be produced by cleaving at the carboxy group of the 28th arginine residue (counted from the N-terminal) and retaining the longer, C-terminal portion. Cleaving can be effected for example with trypsin. 10 Since pCL2 contains an entire cDNA coding for human pancreatic elastase IIIA, it is possible to produce large amounts of the elastase IIIA by transferring the cDNA to an appropriate expression vector and using, for example, E coN, B.subtilis. yeast or mammalian cells as the host. 5) Production of human pancreatic elastase IIIA using animal cells 15 Construction of expression plasmid pSV2-CL2

In order to produce human elastase IIIA using mammalian cells as host, the cDNA was ligated to an expression vector according to the procedure indicated in figure II. The known pSV2 plasmid containing 20 the SV40 promoter, enhancer, poly(A) signal and introns of the Small T antigen gene was used for construction of the expression vector. pSV2-p-globin is a plasmid based on the pSV2 vector with inserted p-globin cDNA. Plasmid pCL2 is the plasmid with inserted human elastase IIIA cDNA. The reactions with S1 nuclease, etc. were carried out according to the methods described in "Molecular Cloning" [Maniatis, Tet al (ed) "Molecular Cloning" Cold Spring Harbor Lab (1982)]. The broad arrows in figure 11 represent 25 the promoter derived from SV40, and the narrow arrows show the direction of transcription. From analysis of the restriction endonuclease cleavage pattern, an expression plasmid (pSV2-CL2) was selected which incorporated the desired fragments oriented in the correct sense for the transcription of the vector and cDNA.

30 Transfection of COS 1 cells with expression plasmid pSV2-CL2

The constructed expression plasmid pSV2-CL2 was transfected into COS 1 (mammalian cells) by the calcium phosphate method described in the literature [Graham and Van der Eb: Virology 52 456 (1973)]. Thus, 1 x 106 of COS 1 cells were seeded on Petri dishes (10 cm in diameter), and cultured overnight on 35 Dulbecco-modified Eagle medium containing 1 0% fetal calf serum. Plasmid pSV2-CL2 (300 ug) was suspended in sterile distilled water (12.55 ml), then 2.5 M CaCl2 (0.75 ml) was added and the suspension mixed thoroughly. A pipette was used to bubble air through the system in order to maintain agitation, and 1.5 ml of 10 x HeBS solution (HEPES, 210 mM; NaCI, 1.37 M; KCI 4.7 mM; Na2HP04, 10.6 mM; glucose, 55.5 mM; pH 7.05) was dropped into the resulting solution to precipi- 40 tate the DNA and calcium phosphate. The precipitate was allowed to stand at room temperature for 30 minutes to mature the precipitate, and 1 ml of the solution for each Petri dish was added to the COS 1 cells cultured in fresh medium containing 10% fetal calf serum. After cultivation of these cells for 12 hours at 37°C in the presence of 5% C02, the culture medium 45 was discarded in favour of a fresh, Dulbecco-modified Eagle medium containing no fetal calf serum. The cells were cultured for a further 48 hours at 37°C in the presence of 5% C02. The transfected COS 1 cells obtained by this procedure were tested to detect the presence of human elastase IIIA mRNA, while the supernatent of the culture medium was tested for elastase activity. 50 Extraction of mRNA from COS 1 cells

In order to confirm the existence of human elastase IIIA mRNA transcribed from the expression plasmid in the transfected COS 1 cells, mRNA was extracted from the COS 1 cells and assayed by the North- ern blot hybridization according to the following procedure. 55 After cultivation of the COS 1 cells for 48 hours, 1 ml of guanidine thiocyanate solution (4M guanidine thiocyanate; 1% Sarcosyl; 20 mM EDTA; 25 mM sodium citrate, pH 7.0; 100 mM 2-mercaptoethanol; 0.1% Antifoam A) was added to each Petri dish in order to lyse the cells. High molecular DNA molecules were degraded to low molecular ones by passing the solution several times through a 21 guage injection needle. The resultant solution was layered on a solution containing 5.7 60 M cesium chloride and 0.1 M EDTA, and the solution centrifuged at 20°C, 30,000 rpm for 17 hours using a Hitachi RPS 40 swing rotor. The resultant RNA precipitate was washed with a small amount of ethanol and dissolved in distilled water (300 ul). 81 0 ug of total mRNA was isolated from about 1 08 COS 1 cells. According to the method of Aviv and Leder [Proc Natl Acad Sci, USA 69, 1408 (1972)], the extracted total RNA was purified on an oligo(dT) cellulose column to give several ug of purified mRNA. Half of the 65 purified mRNA was used for Northern blot hydridization by the Thomas method [Proc Natl Acad Sci,

23 EP 0198 645 B1

USA 77 5201 (1980]. 32p-Labelled human elastase IIIA cDNA was used as a probe by the nick translation method [Rigby, PW etal: J Mol Biol m, 237 (1 977)]. A major band at about 1 .8 Kb and a minor band at about 1.0 Kb which hydridized with the probe were detected only in the mRNA of COS 1 cells transfected with plasmid pSV2-CL2. It was assumed that in the transcription of the plasmid pSV2-CL2, the 1.8 Kb mR- 5 NA is formed on termination of transcription at the poly(A) signal contained in the vector, while the 1 .0 Kb mRNA is formed on termination of transcription at the poly(A) signal in the cDNA. The results from the Northern blot hybridization coincided with these assumptions. The results thus show that using SV40 promoter in COS 1 cells, the plasmid pSV2-CL2 can synthesize large quantities of human elastase IIIA mRNA, and most of the transcribed mRNA is not terminated at the 10 poly(A) signal in the cDNA but instead at the poly(A) signal in the vector. Elastase activity in supernatant of the culture medium

Since the human elastase IIIA cDNA in pSV2-CL2 has a signal peptide region, secretion of expressed 15 elastase into the medium as a proelastase was expected. Elastase activity in the medium after 48 hours cultivation was accordingly determined. The asssay was carried out using the method of Bieth et a[ using a synthetic substrate [Front Matrix Biol 6 1 (1978)]. 200 ul of Tris-HCI buffer solution (pH 8.5) was added to the supernatant of the culture medium (1 ml). 10 mg/ml trypsin (50 uJ) was added, and an activation treatment of the proelastase was carried out for 15 20 minutes at 25°C. Next, 50 ul of soybean trypsin inhibitor solution (10 mg/ml) and 10.4 ul of a synthetic substrate, succinyl-Ala-Ala-Ala-p-nitroanilide (125mM) dissolved in N-methylpyrrolidone were added. The enzyme reaction was carried out by incubation at 37°C for one hour. After completion of the reaction, liberated p-nitroanilide was determined by measuring absorbance at 410 nm. In this assay of the elastase activity using the culture broth of COS1 cells transfected with pSV2- 25 CL2, an OD410 value of 0.21 was found. The presence of elastase was thus recognized in the culture broth. However, the hydrolytic activity for the synthetic substrate was 5.2%, when compared with the hydrolytic activity of a culture supernatant from COS1 cells transfected with an expression plasmid pSV2-PEL1 containing the porcine elastase I cDNA linked to the pSV2 vector. It is also reported that the hydrolytic activity of the human elastase purified by Largman et a] from hu- 30 man pancreas for the synthetic substrate Suc-Ala-Ala-Ala-p-nitroanilide is about 3.5%, when compared with porcine elastase [Largman, C et al Biochemistry 15, 2491 (1976)]. These results are summarized in the following table.

35 Sample hydrolytic activity (%)

Culture supernatant of cells 100

transfected with pSV2-PELl

Culture supernatant of cells 5.2

transfected with pSV2-CL2

50 Purified porcine elastase IOO

Purified human elastase 3.5 55

Thus, Suc-Ala-Ala-Ala-p-nitroanilide which is a very sensitive substrate for porcine elastase, is not a sensitive substrate for assay of human elastase activity. The production of human elastase IIIA in COS 1 cells transfected with plasmid pSV2-CL2 was tran- 60 sient. However, if plasmid pSV2-CL2 is ligated to a suitable selection marker (for example, the neo gene, dihydrofolate reductase gene, or other resistance marker) and is introduced in to CHO or other cells, a cell line suitable for sustained production of human elastase IIIA can be prepared.

24 EP 0 198 645 B1

6) Production of human pancreatic elastase IIIA using Bacillus subtilis Construction of expression vector pHEXOIQ 5 The construction method for the expression vector is illustrated in figure 12. Plasmid containing human elastase IIIA cDNA was digested with restriction endonucleases Hindlll and Bglll, and a DNA fragment of 712 base pairs containing a part of the elastase IIIA cDNA was isolated by agarose gel electrophoresis. This fragment was ligated using T4 DNA with a synthetic oligonucleotide shown in figure 12 comprising 93 base pairs coding for a part of the signal peptide of a-amylase of Bacillus subtilis and 24 amino 10 acids on the amino terminal side of elastase IIIA. A DNA fragment of 805 base pairs was isolated by agarose gel electrophoresis. On the other hand, plasmid pTUB228 [Ohmura, K et al: J Biochem 95 87 (1984)] containing the a-amylase gene of Bacillus subtilis was digested with restriction endonuclease Hindlll to isolate a DNA fragment of 428 base pairs containing the a-amylase promoter and a part of the signal peptide, and separately a 15 DNA fragment of 5,100 base pairs containing the replication origin. The DNA fragment of 428 base pairs and the DNA fragment of 5,100 base pairs were further digested with restriction endonuclease Hpall and Bell, respectively, to give DNA fragments of 385 base pairs and of 4173 base pairs. The three DNA fragments were ligated with T4 DNA ligase, and incorporated into protoplasts of Bacil- lus subtilis 207 - 25 strain (nxr68 hsrM recE4 amyE07 aro1906 leuA8 Iys21; derived from Marburg strain) 20 according to conventional procedures. After regeneration, cultivation on a medium containing 10 ug/ml of kanamycin allowed selection of the transformed strains which could grow on this medium. Selection of the desired plasmid was achieved through colony hybridization using as probe the synthetic oligonucleotide of figure 13. A positive clone was identified and the plasmid shown to be the intended one by determination of the base 25 sequence. The expression plasmid obtained in this manner was designated plasmid pHEX010. Confirmation of the Product

Bacillus subtilis 207-25 strain transformed with the human elastase IIIA 30 expression plasmid pHEX010 was cultured on a reciprocal shaker in 1 I of LG medium (1% Bacto trypton, Difco; 0.5% yeast extract, Dif- co; 0.5% NaCI; 0.2% glucose; pH 7.0) containing 50 ug/ml of kanamycin. The cuituring was performed at 35°C for 48 hours. After completion of the culture, the culture medium was cooled to 4°C and centrifuged at 3000 x G for 5 and the ceils discarded. Ammonium sulfate added to the to 55% 35 minutes, was supernatant saturation, and the solution was stirred at 4°C for 12 hours. Insoluble material formed in this treatment was precipi- tated by centrifugation at 8000 x G for 20 minutes, the supernatant was discarded, and the precipitate was dissolved in 50mM of Tris-HCI buffer solution (pH 8.0). This solution was dialyzed against 500 ml of 50mM Tris-HCI buffer solution (pH 8.0) for 16 hours, and insoluble material was removed by centrifugation at 8000 G for 20 minutes. The solution taken crude elastase IIIA 40 x dialyzed was as a sample solution, and the activity was determined by the following procedure. Elastase activity in the sample solution was determined by the method described below. The synthetic substrate, N-carbobenzoxy-L-alanine-p-nitrophenyl ester (Sigma), was dissolved in 50 mM Tris-HCI solution (pH 8.0) to give a 0.1 mM solution. To 0.25 ml of this substrate solution, 0.25 ml of the elastase solution added and for 37°C. the absorb- 45 sample was allowed to react 30 minutes at Then, ance at 410 nm was determined. At the same time, to the sample solution was added elastatinal or a-1 -antitrypsin, an elastase inhibitor, to a final concentration of 0.1 mg/ml, in order to examine the inhibition to the activity. The absorbance of the 207-25 strain into which the plasmid rightly oriented with the promoter was introduced increased by 0.32 in comparison with the control solution. The elastase activity could be inhibited either of the above inhibitors. 50 by two In this procedure, the reaction of the restriction endonuclease and the other enzymes was carried out using the buffer solutions and the reaction conditions mentioned in the literature accompanying the com- mercial enzyme products. Production of human pancreatic elastase IIIA using Escherichia coli 55 7) Various different promoters may be used, including for instance the tryptophan (trp) promoter, lactose (lac) promoter, tryptophan-lactose (tac) promoter, lipoprotein (Ipp) promoter, lambda (X) P|_ promoter derived from bacteriophage, or polypeptide chain elongation factor (tuf B) promoter. An example with 60 the lactose promoter is as follows. As shown in Figure 14, the plasmid pCL2 containing the human elastase IIIA cDNA was digested with restriction endonucleases EcoRI and Bglll to give a DNA fragment of 1215 bp containing human elastase IIIA cDNA. This fragment was partially digested with restriction enzyme Fnu4HI and then treated with S1 nuclease to give a DNA fragment of 788 bp with blunt ends. Separately, plasmid pUC8 was digested with 65 restriction endonuclease Smal and treated with phosphatase to obtain a DNA fragment containing the

25 EP0198 645 B1

promoter ana operator region ot the lactose operon and a part of the p-galactosidase coding region. The two DNA fragments were added to 30 ul of a solution containing T4 DNA ligase (66 mM of Tris- HCI, pH 7.6; 6.6 mM of MgCI2; 10 mM of DTT; 1mM of ATP; 2.5 units of T4 DNA ligase) and incubated at 6°C for 72 hours to ligate the fragmetns. 5 The thus constructed human elastase expression plasmid was named pHEX002. Various strains of Escherichia coji were transformed with pHEX002, and strains capable of producing human elastase could be obtained. The human elastase expressed in this manner is a fused protein in which 8 amino acids originating from p-galactosidase combine with the N-terminal of mature human elastase, as shown below. 10 The DNA base sequence adjacent the 5'-terminai end of the elastase gene in plasmid pHEX002 and the encoded amino acid sequence are as follows:

e-Gaiactosidase ^-Elastase 15 — trypsin or

cloptripain 20 . Met Thr Met lie Thr Arg%al Asn Ser Val ...

ATC ACC ATG ATT ACG AAT TCC CGC GTT GTC 25

The strain containing the elastase expression plasmid was inoculated on 2 x TY-ampicillin medium (1.6% Bactotrypsin; 1% Yeast Extract; 0.5% NaCI; 50 ug/ml ampicillin) and cultured for 15 hours at 37°C. After completion of the culture, cells were harvested by centrifugation of the culture medium. 2.4 x 108 ce||s 30 were suspended in 15 ul of SDS solution [2% SDS; 5% 2-mercaptoethanol; 10% glycerin; 60 mM Tris- HCI, pH 6.8]. After heating this solution at 100°C for 3 minutes, product protein was analyzed by SDS poiyacrylamide gel electrophoresis, according to the method of Laemmli et al [Nature 227, 680 (1 970)]. As a result, a large amount of elastase production was found in the YA21 strain. In this respect, reference is made to Figure 15, where lane 1 corresponds to molecular weight markers, and lanes 2 to 5 corre- 35 spond to E coN YA21, HB101, MC4100 and x984, respectively, when transformed with plasmid pHEX002. The yield of elastase fused protein was 33% of the total protein produced by the YA21 strain. Though a considerable production of elastase fused protein was observed in x984 strain, the yield was less than half that in the YA21 strain (14% of the total microorgansim cell protein). The yields of elastase fused protein in two other Escherichia coli strains (HB101 and MC4100 strains) were low. io Since the elastase fused protein forms inclusion bodies in cells, purification was comparatively easy. Thus, from 1 1 of a culture medium of YA21 strain transformed with plasmid pHEX002, 6.5 g of cells forming inclusion bodies could be obtained. The cells (6.5 g) were lysed in 50 mM Tris-HCI buffer solution (pH 8.0) containing 0.2 mg/ml lysozyme and 1 mg/ml deoxycholic acid. Intact cells were removed by low speed centrifugation (1,500 x G, 10 minutes) and inclusion bodies were recovered as a precipitate by high i5 speed centrifugation (11,000 x G, 20 minutes). Since the precipitate still contained much cell debris, the precipitate was suspended in 50 mM Tris-HCI buffer solution containing 5 mg/ml octylphenoxy polyethox- yethanol surfactant (Triton X-100, Trade Mark) and washed by high speed centrifugation (11,000 x G, 20 minutes). A washed pellet of inclusion bodies was suspended in a small amount of Tris-HCI buffer solution and stored at 4°C. 50 The purified inclusion body (370 mg) containing about 58% elastase IIIA fused protein could be obtained by this procedure. The inclusion body produced by YA21 strain transformed with pH2EX-2 was shown to contain the elastase fused protein by immuno-blotting. Most of the elastase produced by Escherichia coli is in insoluble parts of the cells as inclusion bodies, but a smaller proportion exists in a soluble state and retains enzyme activity. Determination of the en- !5 zyme activity was carried out as follows. The strain x984 transformed with plasmid pHEX002 was cultured with shaking in 1 I of 2 x TY-ampicillin medium at 37°C for 15 hours. After completion of the cuituring, the cells were harvested by centrifugation at 3000 x G for 5 minutes and suspended in 20 mi of buffer solution A (50 mM Tris-HCI, pH 8.0; 1mM EDTA; 50 mM NaCI). 10 mg of lysozyme was added, and the suspension incubated at 5°C for 20 minuts. Deoxycholic acid to a final concentration of 1 mg/ml was add- 30 ed to the resultant suspension which was then incubated at 20°C. Deoxyribonuclease to a final concentration of 0.1 mg/ml was added, and the cells were disrupted in a Polytron homogenizer. Aftrer removal of the cell debris by centrifugation at 80,000 x G, 40 minutes, the lysate was subjected to column chromatography on high molecular weight, cross-linked dextran (Sephadex G-75 ). The fractions having elastase activity were purified by antibody affinity chromatography, and used for the following determi- 35 nation of elastase activity.

I6 EPO 198 645 B1

Elastase activity in the sample solution was determined by the following procedure. N-carbobenzoxy- L-alanine-p-nitrophenyl ester (Sigma), a sythetic substrate, was dissolved in 50 mM Tris-HCI (pH 8.0) to give a 0.1 mM solution. To 0.25 ml of this substrate solution, 0.25 ml of the elastase sample solution was added, and, after reaction for 30 minutes at 37°C, the absorbance at 410 nm was determined. The 5 elastase activity was detected by the increase in the absorbance of 0.41 . At the same time, when elastatinal or a-1 -antitrypsin (an elastase inhibitor), was added to this sample solution to a final concentration of 0.1 mg/ml, it was found that the inhibitor inhibited the activity.

8) Production of human pancreatic elastase IIIA using yeast 10 In the case of yeast as a host (and as with a Escherichia coli. Bacillus subtilis or mammalian cells), human pancreatic elastase IIIA cDNA could be ligated to a suitable expression vector using known techniques, inserted into the host ceils and expressed. Elastase activity in the culture medium was confirmed. Saccharomvces cerevisiae described in "Japanese Guidelines for Recombinant DNA Research" may 15 be used as host, but S288C strain and other strains are practically suitable. On the other hand, YEp13 and other vectors were appropriate as a vector. As promoter, the ADH1 gene coding for alcohol dehy- drogenase gene, and other promoters, are suitable. Example 4: Elastase IIIB 20 1) Separation of mRNA from human pancreas

Human pancreas (autopsy sample) weighing 5.5 g was homogenised and denatured in a homogeniser (a Polytron homogeniser from Kinematica GmbH, Germany, Polytron being a Trade Mark) in an adequate 25 amount of guanidine thiocyanate solution (4M guanidine thiocyanate; 1% detergent, Sarcosyl from Sigma Chemical Co, USA; 20 mM ethylenediaminetetraacetic acid (EDTA); 25mM sodium citrate, pH 7.0; 100mM 2-mercaptoethanol and 0.1% antifoaming agent, Antifoam A from Sigma Chemical Co USA), followed by centrifugation, to obtain a supernatant. One volume of supernatant was added to 0.025 volumes of 1 M acetic acid and 0.75 volumes of etha- 30 nol, and the mixture cooled to -20°C for several hours, followed by centrifugation, to obtain a precipitate. The precipitate was suspended in a guanidine hydrochloride solution (7.5M guanidine hydrochloride; 25mM sodium citrate, pH 7.0; and 5mM dithiothreitol, DTT). To 1 volume of the suspension was added 0.025 volumes of 1 M acetic acid and 0.5 volumes of ethanol, and the resultant mix cooled to -20°C for several hours, followed by centrifugation. The resultant precipitate was suspended again in more of 35 the guanidine hydrochloride solution, mixed as before with acetic acid and ethanol, cooled to -20°C and centrifuged, followed by collection of the precipitate. Next, the precipitate was washed several times with ethanol to remove guanidine hydrochloride, and then dissolved in distilled water, followed by precipitation of the RNA with ethanol. The precipitate was collected by centrifugation to give 53.9 mg of RNA. 40 The RNA thus obtained was adsorbed onto an oligo(dT) cellulose column in a high concentration salt solution (0.5M NaCI; 20mM Tris-HCI, pH 7.5; 1mM EDTA; and 0.1% sodium dodecyl sulfate, SDS), and the mRNA-containing poly(A) was eluted with an elution buffer (10mM Tris-HCI, pH7.5; 1mM EDTA; and 0.05% SDS) to give 255 ug of mRNA. 45 2) Preparation of cDNA bank

27 EP 0198 645 B1

The precipitate was dissolved in a buffer for restriction endonuclease (50mM NaCI; 10mM Tris-HCI, pH 7.5; 10mM MgCI2; and 1mM DTT). 2.5 units of the restriction enzyme Hindlll were added to the solution and the solution incubated at 37°C for about 1 hour. After deproteinization with phenol-chloroform, precipitation was carried out with ethanol. 5 After cooling to -70°C for 15 minutes, the precipitate was collected by centrifugation and dissolved in 10 ul of TE (10mM Tris-HCI, pH 7.5; 1mM EDTA). 1 ul of the above sample solution was then added to a reaction solution (10mM Tris-HCI, pH 7.5; 1mM EDTA; and 100mM NaCI), followed by addition of 10 ug of linker (purchased from PL-Pharmacia) bearing oligo dG. The resulting mixture was heated at 65°C for 5 minutes and maintained at 42°C for 30 minutes. 10 After cooling the reaction solution in ice water, 10 ul of 10-fold ligase buffer solution (10mM ATP; 660mM Tris-HCI, pH 7.5; 66mM MgCI2; and 100mM DTT), 78 ul of distilled water and 8 units of T4 DNA concentrated ligase were added to the solution and the reaction mixture maintained at 12°C overnight. Thus, to the thus prepared mixture, 10 ul of 1M KCI, 1 unit of ribonuclease H, 33 units of DNA polymerase I, 4 units of T4 DNA ligase, 0.5 ul of nucleotide solution (20mM dATP, 20 mM dCTP, 20 mM cDTP 15 and 20mM dTTP) and 0.1 ul of 50 ug/ul bovine serum albumin (BSA) were added and the mixture maintained at 12°C for 1 hour and then at 25°C for 1 hour. After diluting the reaction solution 5-fold, E. coji strain RR1 was transformed according to the method of Hanahan [Hanahan, D; J Mol Biol 166. 557 (1983)] to prepare a human pancreatic cDNA bank. 20 3) Selection of transformed bacteria containing human pancreactic elastase IIIB cDNA

The DNA from 4,700 clones of the human pancreactic cDNA bank was fixed onto nitrocellulose filter according to the method of Grunstein, M and Hogness, D S [Proc Natl Acad Sci, USA, 72, 3961 (1975)]. A porcine pancreatic elastase I cDNA fragment was exised with restriction endonuclease Pstl and la- 25 belled with 32p according to the nick translation method [Rigby, p W et a]:J. Mol. Biol 113, 237 (1 977)]. Us- ing the DNA fragment as probe, it was hybridized with the DNA fixed on the filter at 35°C in a solution of 40% formamide, 0.75 M sodium chloride and 0.075 M sodium citrate (SSC), along with a solution of 0.1% BSA, 0.1% polysucrose (Ficoll, Trade Mark), 0.1% polyvinyl pyrrolidone, 0.5% SDS, 10 mM sodium phosphate and 1 00 ug/ml of salmon spermatozoon DNA. 30 10 microorganism strains which hybridized with the probe were identified by autoradiography. A plasmid from one of the strains was named pCL1 , and the inserted cDNA was named CL1 . 4) The amino acid sequence and elastase IIIB coded bv plasmid pCL1

35 In CL1, there exists only one open reading frame. This sequence codes for 270 amino acids. The encoded amino acid sequence showed some homology with trypsin and chymotrypsin, to a degree of only about 30%. However, there was a higher homology, 50%, with porcine pancreatic elastase I. There also existed an amino acid sequence (Gly-Asp-Ser-Gly-Gly-Pro) around the active center which is be com- monly found in the same region in serine proteases such as trypsin, chymotrypsin, etc. Further, there is 40 a 45th histidine residue, 95th aspartic acid residue and 189th serine residue which can participate in the characteristic charge relay of serine proteases. Moreover, the encoded amino acid sequence adjacent the carboxyl terminal, which in elastase forms a pocket for binding with the substrate is very similar between human pancreatic elastase IIIB and the porcine elastase I. There is also an encoded 211th valine residue and 223rd threonine residue, which are 45 characteristic of a substrate-binding pocket of the porcine elastase I. The base sequence was of the formula (5> CCT ATC ATC GCA AAA CTC ATG ATG CTC CGG CTG CTC AGT TCC CTC CTC CTT GTG GCC GTT GCC TCA GGC TAT 50 GGC CCA CCT TCC TCT CGC CCT TCC AGC CGC GTT GTC AAT GGT GAG GAT GCG GTC CCC TAC AGC TGG CCC TGG CAG GTT TCC CTG CAG TAT GAG AAA AGC GGA AGC TTC TAC CAC ACC TGT GGC GGT AGC CTC ATC GCC CCC GAC TGG GTT 55 GTG ACT GCC GGC CAC TGC ATC TCG AGC TCC CGG ACC TAC CAG GTG GTG TTG GGC GAG TAC GAC CGT GCT GTG AAG GAG GGC CCC GAG CAG GTG ATC CCC ATC AAC TCT GGG GAC CTC TTT GTG CAT CCA CTC TGG AAC CGC TCG TGT GTG 60 GCC TGT GGC AAT GAC ATC GCC CTC ATC AAG CTC TCA CGC AGC GCC CAG CTG GGA GAC GCC GTC CAG CTC GCC TCA CTC CCT CCG GCT GGT GAC ATC CTT CCC AAC GAG ACA CCC TGC TAC ATC ACC GGC TGG GGC CGT CTC TAT ACC AAC 65 GGG CCA CTC CCA GAC AAG CTG CAG GAG GCC

28 EP 0 198 645 B1

CTG CTG CCG GTG GTG GAC TAT GAA CAC TGC TCC AGG TGG AAC TGG TGG GGT TCC TCC GTG AAG AAG ACC ATG GTG TGT GCT GGA GGG GAC ATC CGC TCC GGC TGC AAT GGT GAC TCT GGA 5 GGA CCC CTC AAC TGC CCC ACA GAG GAT GGT GGC TGG CAG GTC CAT GGC GTG ACC AGC TTT GTT TCT GCC TTT GGC TGC AAC ACC CGC AGG AAG CCC ACG GTG TTC ACT CGA GTC TCC GCC TTC ATT GAC TGG ATT GAG GAG ACC ATA GCA 10 AGC CAC TAG AAC CAA GGC CCA GCT GGC AGT GCT GAT CGA TCC CAC ATC CTG AAT AAA GAA TAA AGA TCT CTC AGA AAA A-(3') Beginning with the underlined ATG triplet, the encoded amino acid sequence is of the formula: (N)- Met Met Leu Arg Leu Leu Ser Ser 15 Leu Leu Leu Val Ala Val Ala Ser Gly Tyr Gly Pro Pro Ser Ser Arg Pro Ser Ser Arg Val Val Asn Gly Glu Asp Ala Val Pro Tyr Ser Trp Pro Trp Gin Val Ser Leu Gin Tyr Glu Lys Ser Gly Ser Phe Tyr His Thr Cys 20 Gly Gly Ser Leu He Ala Pro Asp Trp Val Val Thr Ala Gly His Cys lie Ser Ser Ser Arg Thr Tyr Gin Val Val Leu Gly Glu Tyr Asp Arg Ala Val Lys Glu Gly Pro Glu Gin Val He Pro lie Asn Ser Gly Asp Leu Phe 25 Val His Pro Leu Trp Asn Arg Ser Cys Val Ala Cys Gly Asn Asp lie Ala Leu lie Lys Leu Ser Arg Ser Ala Gin Leu Gly Asp Ala Val Gin Leu Ala Ser Leu Pro Pro Ala Gly Asp lie Leu Pro Asn Glu Thr Pro Cys Tyr 30 He Thr Gly Trp Gly Arg Leu Tyr Thr Asn Gly Pro Leu Pro Asp Lys Leu Gin Glu Ala Leu Leu Pro Val Val Asp Tyr Glu His Cys Ser Arg Trp Asn Trp Trp Gly Ser Ser Val Lys Lys Thr Met Val Cys Ala Gly Gly Asp 35 He Arg Ser Gly Cys Asn Gly Asp Ser Gly Gly Pro Leu Asn Cys Pro Thr Glu Asp Gly Gly Trp Gin Val His Gly Val Thr Ser Phe Val Ser Ala Phe Gly Cys Asn Thr Arg Arg Lys Pro Thr Val Phe Thr Arg Val Ser Ala 40 Phe He Asp Trp lie Glu Glu Thr lie Ala Ser His -(C) The amino acid sequence of the coded protein appears at a glance to be analogous with the amino acid composition of the human pancreatic elastase I reported by Largman et a] [Largman, C eta]: Biochemistry 15, 2491 (1976)]. However, the present elastase was clearly different in detail from the known one, and 45 the molecular weights were also different. The active form of the human elastase may be produced by cleaving at the carboxy group of the 28th arginine residue (counted from the N-terminal) of the preproelastase, and retaining the longer, C-terminal portion. Cleaving can be effected for example with trypsin. The degree of homology in encoded amino acid sequence between preproelastase IIIB of CL1 and preproelastase IIIA of CL2 is as high as 93.3%. The degree of homology in base sequence between pre- 50 proelastase CL2 and CL1 is as high as 95.3%. Hence the elastases are closely related proteins. In the human pancreas, the amount of mRNA corresponding to CL1 is about 1 0% of that for CL2. Since pCL1 contains an entire DNA coding for human pancreatic elastase IIIB, it is possible to produce large amounts of the elastase IIIB by transferring the cDNA to an appropriate expression vector and using, for example. E. coli. B. subtilis. yeast or mammalian cells as the host. 55 5) Production of human pancreatic elastase IIIB using mammalian cells Construction of expression plasmid PSV2-CL1

60 In order to produce human elastase IIIB using mammalian cells as host, the cDNA was ligated to an expression vector according to the procedure indicated in figure 16. The known pSV2 plasmid containing the SV40 promoter, enhancer, poly(A) signal and introns of the Small T antigen gene was used for construction of the expression vector. pSV2-p-globin is a plasmid based on the pSV2 vector with inserted p-globin cDNA. Plasmid pCL1 is the plasmid with inserted human elastase IIIB cDNA. The reactions with 65 S1 nuclease, etc, were carried out according to the methods described in "Molecular Cloning" [Maniatis,

29 EP 0 198 645 B1

i et ai (ed) Molecular uoning" Gold Spring Harbor Lab (1982)]. The broad arrows in figure 16 represent the promoter derived from SV40, and the narrow arrows show the direction of transcription. From analysis of the restriction endonuclease excision pattern, an expression plasmid (pSV2-CL1) was selected which incorporated the desired fragments oriented in the correct sense for the transcrip- 5 tion of the vector and cDNA.

Transfection of COS 1 cells with expression plasmid PSV2-CL1

The constructed expression plasmid pSV2-CL1 was transfected into COS 1 (mammalian cells) by the 10 calcium phosphate method described in the literature [Graham and van der Eb: Virology 52 456 (1 973)]. Thus, 1 x 106 of COS 1 cells were seeded on Petri dishes (10 cm in diameter), and cultured overnight on Dulbecco-modified Eagle medium containing 10% fetal calf serum. Plasmid pSV2-CL1 (300 ug) was suspended in sterile distilled water (12.55 ml), then 2.5 M CaCI2 (0.75 ml) was added and the suspension mixed thoroughly. A pipette was used to bubble air through the system 15 in order to maintain agitation, and 1.5 ml of 10 x HeBS solution (HEPES, 210 mM; NaCI, 1.37 M; KCI 4.7 mM; Na2HP04, 10.6 mM; glucose, 55.5 mM; pH 7.05) was dropped into the resulting solution to precipitate the DNA and calcium phosphate. The precipitate was allowed to stand at room temperature for 30 minutes to mature the precipitate, and 1 ml of the solution for each Petri dish was added to the COS 1 cells cultured in fresh medium containing 20 1 0% fetal calf serum. After cultivation of these cells for 12 hours at 37°C in the presence of 5% C02, the culture medium was discarded in favour of a fresh, Dulbecco-modified Eagle medium containing no fetal calf serum. The cells were cultured for a further 48 hours at 37°C in the presence of 5% C02. The transfected COS 1 cells obtained by this procedure were tested to detect the presence of human elastase IIIB mRNA, while 25 the supernatant of the culture medium was tested for elastase activity. Extraction of mRNA from COS 1 cells

In order to confirm the existence of human elastase IIIB mRNA transcribed from the expression plas- 30 mid in the transfected COS 1 cells, mRNA was extracted from the COS 1 cells and assayed by the North- ern blot hybridization according to the following procedure. After cultivation of the COS 1 cells for 48 hours, 1 ml of guanidine thiocyanate solution (4M guanidine thiocyanate; 1% Sarcosyl; 20 mM EDTA; 25 mM sodium citrate, pH 7.0; 100 mM 2-mercaptoethanoi; 0.1% Antifoam A) was added to each Petri dish in order to lyse the cells. 35 High molecular DNA molecules were degraded to low molecular ones by passing the solution several times through a 21 guage injection needle. The resultant solution was layered on a solution containing 5.7 M cesium chloride and 0.1 M EDTA, and the solution centrifuged at 20°C, 30,000 rpm for 17 hours using a Hitachi RPS 40 swing rotor. The resultant RNA precipitate was washed with a small amount of ethanol and dissolved in distilled water (300 ul). 40 According to the method of Aviv and Leder [Proc Natl Acad Sci, USA 69, 1408 (1972)], the extracted total RNA was purified on an oligo(dT) cellulose column to give several ug of purified mRNA. Half of the purified mRNA was used for Northern blot hydridization by the Thomas method [Proc Natl Acad Sci, USA 77 5201 (1908)]. 32p-Labelled human elastase IIIA cDNA was used as a probe by the nick translation method [Rigby, P W etal: J Mol Biol 113, 237 (1977)]. A major band at about 1 .8 Kb and a minor band at 45 about 1 .0 Kb which hydridized with the probe were detected only in the mRNA of COS 1 cells transfected with plasmid It assumed pSV2-CL1 . was that in the transcription of the plasmid pSV2-CL1 , the 1 .8 Kb mR- NA is formed on termination of transcription at the poly(A) signal contained in the vector, while the 1.0 Kb mRNA is formed on termination of transcription at the poly(A) signal in the cDNA. The results from the Northern blot hybridization coincided with these assumptions. 50 The results thus show that using SV40 promoter in COS 1 cells, the plasmid pSV2-CL1 can synthesize large quantities of human elastase IIIB mRNA, and most of the transcribed mRNA is not terminated at the poly(A) signal in the cDNA but instead at the poly(A) signal in the vector. Elastase activity in supernatant of the culture medium >5 Since the human elastase IIIB cDNA in pSV2-CL1 has a signal peptide region, secretion of expressed elastase into the medium as a proelastase was expected. Elastase activity in the medium after 48 hours cultivation was accordingly determined. The asssay was carried out using the method of Bieth et a] using a synthetic substrate [Front Matrix Biol 6 1 (1978)]. 30 200 ul of Tris-HC1 buffer solution (pH 8.5) was added to the supernatant of the culture medium (1 ml). 10 mg/ml trypsin (50 |xl) was added, and an activation treatment of the proelastase was carried out for 15 minutes at 25°C. Next, 50 ul of soybeam trypsin inhibitor solution (10 mg/ml) and 10.4 ul of a synthetic substrate, succinyl-Ala-Ala-Ala-p-nitroanilide (125mM) dissolved in 125mM of N-methylpyrrolidone were added. The enzyme reaction was carried out by incubation at 37°C for one hour. After completion 35 of the reaction, liberated p-nitroanilide was determined by measuring absorbance at 410 nm.

io EP 0 198 645 B1

Elastase activity in the culture medium of COS 1 cells transfected with pSV2-CL1 was found, after activation with trypsin. The production of human elastase IIIB in COS 1 cells transfected with plasmid pSV2-CL1 was tran- sient. However, if plasmid pSV2-CL1 is ligated to a suitable selection marker (for example, the neo gene, 5 dihydrofolate reductase gene, or other resistance marker) and is introduced in to CHO or other cells, a cell line suitable for sustained production of human elastase IIIB can be prepared.

6) Production of human pancreatic elastase IIIB using Bacillus subtilis 10 Construction of expression vector pHAMOOl

The construction method for the expression vector is illustrated in figure 17. Plasmid containing human elastase IIIB cDNA was digested with restriction endonucleases Hindlll and Bglll, and a DNA fragment of 712 base pairs containing a part of the elastase IIIB cDNA was isolated by agarose gel electrophore- 15 sis. This fragment was ligated using T4 DNA with a synthetic oligonucleotide of 85 base pairs shown in figure 1 8 which codes for a part of the signal peptide of a-amylase of Bacillus subtilis and for 24 amino acid residues on the amino terminal side of elastase IIIB. A DNA fragment of 797 base pairs was isolated by agarose gel electrophoresis. On the other hand, plasmid pTUB228 [Ohmura, K et aJ: J Biochem 95 87 (1984)] containing the a-amyla- 20 se gene of Bacillus subtilis was digested with restriction endonuclease Hindlll to isolate a DNA fragment of 428 base pairs containing the a-amylase promoter and a part of the signal peptide, and separately a DNA fragment of 5,100 base pairs containing the replication origin. The DNA fragment of 428 base pairs and the DNA fragement of 5,100 base pairs were further digested with restriction endonuclease Hpall and Bell, respectively, to give DNA fragments of 385 base pairs and of 4173 base pairs. 25 The three DNA fragments were ligated with T4 DNA ligase, and incorporated into protoplasts of Bacil- lus subtilis 207 - 25 strain (nrr68 hsrM recE4 amyE07 arol906 leuA8 Iys21 ; derived from Marburg strain) according to conventional procedures. After regeneration, cultivation on a medium containing 10 ug/ml of kanamycin allowed selection of the transformed strains which could grow on this medium. Selection of the desired plasmid was achieved through colony hybridization using the 30 as probe synthetic oligonucleotide of figure 1 8. A positive clone was identified and the plasmid shown to be the intended one by determination of the base sequence. The expression plasmid obtained in this manner was designated plasmid pHAMOOl . Confirmation of the Product 35 Bacillus subtilis 207 - 25 strain transformed with the human elastase IIIB expression plasmid pHAMOOl was cultured on a reciprocal shaker in 1 I of LG medium (1% Bacto trypton, Difco; 0.5% yeast extract, Dif- co; 0.5% NaCI; 0.2% glucose; pH 7.0) containing 50 ug/ml of kanamycin. The cuituring was performed at 35°C for 48 hours. 40 After completion of the culture, the culture medium was cooled to 4°C and centrifuged at 3000 x G for 5 minutes, and the cells discarded. Ammonium sulfate was added to the supernatant to 55% saturation, and the solution was stirred at 4°C for 12 hours. Insoluble material formed in this treatment was precipi- tated by centrifugation at 8000 x G for 20 minutes, the supernatant was discarded, and the precipitate was dissolved in 50mM Tris-HCI buffer solution (pH This solution was 500 ml of 45 8.0) dialyzed against 50mM Tris-HC1 buffer solution (pH 8.0) for 16 hours, and insoluble material was removed by centrifugation at 8000 x G for 20 minutes. The dialyzed solution was taken as a crude elastase IIIB sample solution, and the activity was determined by the following procedure. Elastase activity in the sample solution was determined by the method described below. The synthetic substrate, N-carbobenzoxy-L-alanine-p-nitrophenyl ester (Sigma), was dissolved in 50 50 mM Tris-HCI solution (pH 8.0) to give a 0.1 mM solution. To 0.25 ml of this substrate solution, 0.25 ml of the elastase sample solution was added and allowed to react for 30 minutes at 37°C. Then, the absorbance at 41 0 nm was determined. At the same time, to the sample solution was added elastatinal or a-1 -antitrypsin, an elastase inhibitor, to a final concentration of 0.1 mg/ml, in order to examine the inhibition to the activity. The absorbance of the 207-25 strain into which the plasmid rightly oriented with the promoter was introduced increased by 0.24 in comparison with the control solution. The elastase activity could be inhibited by either of the above two inhibitors. In this procedure, the reaction of the restriction endonuclease and the other enzymes was carried out using the buffer solutions and the reaction conditions mentioned in the literature accompanying the com- mercial products. 60 enzyme 7) Production of human pancreatic elastase IIIB using Escherichia coli

Various different promoters may be used, including for instance the tryptophan (trp) promoter, lactose (lac) promoter, tryptophan-lactose (tac) promoter, lipoprotein (Ipp) promoter, lambda (X) Pl 65 promot-

31 EP 0 198 645 B1

er derived tram bacteriophage, or polypeptide chain elongation factor (tuf B) promoter. An example with the lactose promoter is as follows. As shown in Figure 19, the plasmid pCL1 containing the human elastase IIIB cDNA was digested with restriction endonucleases EcoRI and Bglll to give a DNA fragment of 1640 bp containing human elastase 5 IIIB cDNA. This fragment was partially digested with restriction enzyme Fnu4HI and then treated with S1 nuclease to give a DNA fragment of 788 bp with blunt ends. Separately, plasmid pUC8 was digested with restriction endonuclease Smal and treated with phosphatase to obtain the DNA fragment containing the promoter and operation region of the lactose operon and a part of the p-galactosidase coding region. The two DNA fragments were added to 30 ul of a solution containing T4 DNA ligase (66 mM of Tris- 10 HCI, pH 7.6; 6.6 mM of MgCI2; 10 mM of DTT; 1 mM of ATP; 2.5 units of T4 DNA ligase) and incubated at 6°C for 72 hours to ligate the fragments. The thus constructed human elastase expression plasmid was named pHEX102. Various strains of Escherichia coJi were transformed with pHEX102, and strains capable of producing human elastase could be obtained. 15 The human elastase expressed in this manner is a fused protein in which 8 amino acids originating from p-galactosidase combine with the N-terminal of mature human elastase, as shown below. The DNA base sequence adjacent the 5'-terminal end of the elastase gene in plasmid pHEX102 and the encoded amino acid sequence are as follows:

p-Galactosidase ^ Elastase

trypsin or

clostripain L Met Thr Met He Thr Asn Ser Arg Val val ... ATC ACC ATG ATT ACG AAT TCC CGC GTT GTC ...

I ne strain containing the elastase expression plasmid was inoculated on 2x TY-ampicillin medium (1.6% Bactotryptone; 1% Yeast Extract; 0.5% NaCI; 50 ul/ml ampicillin) and cultured for 15 hours at 37°C. Af- 35 ter completion of the culture, cells were harvested by centrifugation of the culture medium. 2.4x 108 cells were suspended in 15 ul of SDS solution [2% SDS; 5% 2-mercaptoethanol; 10% glycerin; 60 mM Tris-HCI, pH 6.8]. After heating this solution at 100°C for 3 minutes, product protein was analyzed by SDS polyacrylamide gel electrophoresis, according to the method of Laemmli et al 1[Nature 227, 680 (1970)]. - - 40 Since the elastase fused protein forms inclusion bodies in cells, purification was comparatively easy. Thus, from 1 I of a culture medium of YA21 strain transformed with plasmid pHEX102, 6.6 g of cells forming inclusion bodies could be obtained. The cells (6.6 g) were lysed in 50 mM Tris-HCI buffer solution (pH 8.0) containing 0.2 mg/ml lysozyme and 1 mg/ml deoxycholic acid, intact cells were removed by low speed centrifugation (1,500 x G, 10 minutes) and inclusion bodies were recovered as a precipitate by high *5 speed centrifugation (11,000 x G, 20 minutes). Since the precipitate still contained much cell debris, the precipitate was suspended in 50 mM Tris-HCI buffer solution containing 5 mg/ml octylphenoxy polyethox- yethanol surfactant (Triton X-100, Trade Mark) and washed by high speed centrifugation (11,000 x G, 20 minutes). A washed pellet of inclusion bodies was suspended in a small amount of Tris-HCI buffer solution and stored at 4°C. 50 The purified inclusion body (350 mg) containing nearly 50% elastase IIIB fused protein could be obtained by this procedure. The inclusion body produced by YA21 strain transformed with pHEX102 was shown to contain the elastase fused protein by immuno-blotting. Most of the elastase produced by Escherichia coli is in insoluble parts of the cells as inclusion bodies, but a smaller proportion exists in a soluble state and retains enzyme activity. Determination of the en- 55 zyme activity was carried out as follows. The strain %984 transformed with plasmid pHEX102 was cultured with shaking in 1 I of 2 x TY-amplicillin medium at 37°C for 15 hours. After completion of the cuituring, the cells were harvested by centrifugation at 3000 x G for 5 minutes and suspended in 20 ml of buffer solution A (50 mM Tris-HCI, pH 8.0; 1mM EDTA; 50 mM NaCI). 10 mg of lysozyme was added, and the suspension incubated at 5°C for 20 minutes. so Deoxycholic acid to a final concentration of 1 mg/ml was added to the resultant suspension which was then incubated at 20°C. Deoxyribonuclease to a final concentration of 0.1 mg/ml was added, and the cells were disrupted in a Polytron homogenizer. After removal of the cell debris by centrifugation at 80,000 x G, 40 minutes, the lysate was subjected to column chromatography on high molecular weight, cross- linked dextran (Sephadex G-75 ). The fractions having elastase activity were purified by antibody affini- 35 ty chromatography, and used for the following determination of elastase activity.

12 EP 0 198 645 B1

Elastase activity in the sample solution was determined by the following procedure. N-carbobenzoxy- L-alanine-p-nitrophenyl ester (Sigma), a synthetic substrate, was dissolved in 50 mM Tris-HCI (pH 8.0) to give a 0.1 mM solution. To 0.25 ml of this substrate solution, 0.25 ml of the elastase sample solution was added, and, after reaction for 30 minutes at 37°C, the absorbance at 410 nm was determined. The 5 elastase activity was detected by the increase in the absorbance of 0.36. At the same time, when elastatinal or a-1 -antitrypsin (an elastase inhibitor), was added to this sample solution to a final concentration of 0.1 mg/ml, it was found that the inhibitor inhibited the activity. 8) Production of human pancreatic elastase IIIB using veast 10 In the case of yeast as a host (and as with a Escherichia coli. Bacillus subtilis or mammalian cells), human pancreatic elastase IIIB cDNA could be ligated to a suitable expression vector using known techniques, introduced into the host cells and expressed. Elastase activity in the culture medium was confirmed. 15 Saccharomvces cerevisiae described in "Japanese Guidelines for Recombinant DNA Research" may be used as host, but S288C strain and other strains are practically suitable. On the other hand, YEp13 and other vectors were appropriate as a vector. As promoter, the ADH1 gene coding for alcohol dehy- drogenase gene, and other promoters, are suitable. 20 Claims

1 . A compound which includes an amino acid sequence for a human pancreatic elastase selected from the group of human pancreatic elastase IIB, human pancreatic elastase IIIA, human pancreatic elastase 25 IIIB, and any compounds corresponding to human pancreatic elastase IIB, IIIA or IIIB with one or more deleted, inverted, replaced or altered amino acids, said amino acid sequences of human pancreatic elastase IIB, human pancreatic elastase IIIA, and human pancreatic elastase IIIB respectively being;- human pancreatic elastase IIB: (N)-Cys Gly Val Ser Thr Tyr Ala Pro Asp Met 30 Ser Arg Met Leu Gly Gly Glu Glu Ala Arg Pro Asn Ser Trp Pro Trp Gin Val Ser Leu Gin Tyr Ser Ser Asn Gly Gin Trp Tyr His Thr Cys Gly Gly Ser Leu He Ala Asn Ser Trp Val Leu Thr Ala Ala His Cys He Ser 35 Ser Ser Arg He Tyr Arg Val Met Leu Gly Gin His Asn Leu Tyr Val Ala Glu Ser Gly Ser Leu Ala Val Ser Val Ser Lys lie Val Val His Lys Asp Trp Asn Ser Asn Gin Val Ser Lys Gly Asn Asp He Ala Leu Leu Lys 40 Leu Ala Asn Pro Val Ser Leu Thr Asp Lys He Gin Leu Ala Cys Leu Pro Pro Ala Gly Thr lie Leu Pro Asn Asn Tyr Pro Cys Tyr Val Thr Gly Trp Gly Arg Leu Gin Thr Asn Gly Ala Leu Pro Asp Asp Leu Lys Gin Gly 45 Arg Leu Leu Val Val Asp Tyr Ala Thr Cys Ser Ser Ser Gly Trp Trp Gly Ser Thr Val Lys Thr Asn Met lie Cys Ala Gly Gly Asp Gly Val lie Cys Thr Cys Asn Gly Asp Ser Gly Gly Pro Leu Asn Cys Gin Ala Ser Asp 50 Gly Arg Trp Glu Val His Gly He Gly Ser Leu Thr Ser Val Leu Gly Cys Asn Tyr Tyr Tyr Lys Pro Ser He Phe Thr Arg Val Ser Asn Tyr Asn Asp Trp He Asn Ser Val He Ala Asn Asn - (C) 55 human pancreatic elastase IIIA: (N)-Val Val His Gly Glu Asp Ala Val Pro Tyr Ser Trp Pro Trp Gin Val Ser Leu Gin Tyr Glu Lys Ser Gly Ser Phe Tyr His Thr Cys Gly Gly Ser Leu He Ala Pro Asp Trp Val 60 Val Thr Ala Gly His Cys lie Ser Arg Asp Leu Thr Tyr Gin Val Val Leu Gly Glu Tyr Asn Leu Ala Val Lys Glu Gly Pro Glu Gin Val lie Pro lie Asn Ser Glu Glu Leu Phe Val His Pro Leu Trp Asn Arg Ser Cys Val 65 Ala Cys Gly Asn Asp He Ala Leu lie Lys

33 EP 0 198 645 B1

Leu ser Arg Ser Ala Gin Leu Gly Asp Ala Val Gin Leu Ala Ser Leu Pro Pro Ala Gly Asp lie Leu Pro Asn Lys Thr Pro Cys Tyr lie Thr Gly Trp Gly Arg Leu Try Thr Asn 5 Gly Pro Leu Pro Asp Lys Leu Gin Gin Ala Arg Leu Pro Val Val Asp Tyr Lys His Cys Ser Arg Trp Asn Trp Trp Gly Ser Thr Val Lys Lys Thr Met Val Cys Ala Gly Gly Tyr He Arg Ser Gly Cys Asn Gly Asp Ser Gly 1 0 Gly Pro Leu Asn Cys Pro Thr Glu Asp Gly Gly Trp Gin Val His Gly Val Thr Ser Phe Val Ser Ala Phe Gly Cys Asn Phe lie Trp Lys Pro Thr Val Phe Thr Arg Val Ser Ala Phe lie Asp Trp He Glu Glu Thr lie Ala 15 Ser His - (C) or human pancreatic elastase IIIB (N)-Val Val Asn Gly Glu Asp Ala Val Pro Tyr Ser Trp Pro Trp Gin Val Ser Leu Gin Tyr Glu Lys Ser Gly Ser Phe Tyr His Thr Cys 20 Gly Gly Ser Leu lie Ala Pro Asp Trp Val Val Thr Ala Gly His Cys He Ser Ser Ser Arg Thr Tyr Gin Vai Val Leu Gly Glu Tyr Asp Arg Ala Val Lys Glu Gly Pro Glu Gin Val lie Pro He Asn Ser Gly Asp Leu Phe 25 Val His Pro Leu Trp Asn Arg Ser Cys Val Ala Cys Gly Asn Asp He Ala Leu lie Lys Leu Ser Arg Ser Ala Gin Leu Gly Asp Ala Val Gin Leu Ala Ser Leu Pro Pro Ala Gly Asp lie Leu Pro Asn Glu Thr Pro Cys Tyr 30 lie Thr Gly Trp Gly Arg Leu Tyr Thr Asn Gly Pro Leu Pro Asp Lys Leu Gin Glu Ala Leu Leu Pro Val Val Asp Tyr Glu His Cys Ser Arg Trp Asn Trp Trp Gly Ser Ser Val Lys Lys Thr Met Val Cys Ala Gly Gly Asp 35 He Arg Ser Gly Cys Asn Gly Asp Ser Gly Gly Pro Leu Asn Cys Pro Thr Glu Asp Gly Gly Trp Gin Val His Gly Val Thr Ser Phe Val Ser Ala Phe Gly Cys Asn Thr Arg Arg Lys Pro Thr Val Phe Thr Arg Val Ser Ala 40 Phe lie Asp Trp lie Glu Glu Thr lie Ala Ser His - (C). 2. A compound as claimed in claim 1 which is a proelastase, preproelastase or fusion protein. 3. A DNA which codes for a compound according to claim 1 . 4. A DNA which codes for a compound according to claim 2. 45 5. A DNA as claimed in claim 3 which includes a base sequence selected from the group consisting of the following three sequences and functionally similar base sequences: (5")-TGT GGG GTC TCC ACT TAC GCG CCT GAT ATG TCT AGG ATG CTT GGA GGT GAA GAA GCG AGG CCC AAC AGC TGG CCC TGG CAG GTC TCC CTG 50 CAG TAC AGC TCC AAT GGC CAG TGG TAC CAC ACC TGC GGA GGG TCC CTG ATA GCC AAC AGC TGG GTC CTG ACG GCT GCC CAC TGC ATC AGC TCC TCC AGG ATC TAC CGC GTG ATG CTG GGC CAG CAT AAC CTC TAC GTT GCA GAG TCC GGC 55 TCG CTG GCC GTC AGT GTC TCT AAG ATT GTG GTG CAC AAG GAC TGG AAC TCC AAC CAG GTC TCC AAA GGG AAC GAC ATT GCC CTG CTC AAA CTG GCT AAC CCC GTC TCC CTC ACC GAC AAG ATC CAG CTG GCC TGC CTC CCT CCT GCC GGC SO ACC ATT CTA CCC AAC AAC TAC CCC TGC TAC GTC ACA GGC TGG GGA AGG CTG CAG ACC AAC GGG GCT CTC CCT GAT GAC CTG AAG CAG GGC CGG TTG CTG GTT GTG GAC TAT GCC ACC TGC TCC AGC TCT GGC TGG TGG GGC AGC ACC GTG 35 AAG ACG AAT ATG ATC TGT GCT GGG GGT GAT

S4 EP 0 198 645 B1

GGC GTG ATA TGC ACC TGC AAC GGA GAC TCC GGT GGG CCG CTG AAC TGT CAG GCA TCT GAC GGC CGG TGG GAG GTG CAT GGC ATC GGC AGC CTC ACG TCG GTC CTT GGT TGC AAC TAC TAC 5 TAC AAG CCC TCC ATC TTC ACG CGG GTC TCC AAC TAC AAC GAC TGG ATC AAT TCG GTG ATT GCA AAT AAC -X (3'); (5')-GTT GTC CAT GGT GAG GAT GCG GTC CCC TAC AGC TGG CCC TGG CAG GTT TCC CTG CAG TAT 1 0 GAG AAA AGT GGA AGC TTC TAC CAC ACG TGT GGC GGT AGC CTC ATC GCC CCC GAT TGG GTT GTG ACT GCC GGC CAC TGC ATC TCG AGG GAT CTG ACC TAC CAG GTG GTG TTG GGT GAG TAC AAC CTT GCT GTG AAG GAG GGC CCC GAG CAG 1 5 GTG ATC CCC ATC AAC TCT GAG GAG CTG TTT GTG CAT CCA CTC TGG AAC CGC TCG TGT GTG GCC TGT GGC AAT GAC ATC GCC CTC ATC AAG CTC TCA CGC AGC GCC CAG CTG GGA GAT GCC GTC CAG CTC GCC TCA CTC CCT CCC GCT GGT 20 GAC ATC CTT CCC AAC AAG ACA CCC TGC TAC ATC ACC GGC TGG GGC CGT CTC TAT ACC AAT GGG CCA CTC CCA GAC AAG CTG CAG CAG GCC CGG CTG CCC GTG GTG GAC TAT AAG CAC TGC TCC AGG TGG AAC TGG TGG GGT TCC ACC GTG 25 AAG AAA ACC ATG GTG TGT GCT GGA GGG TAC ATC CGC TCC GGC TGC AAC GGT GAC TCT GGA GGA CCC CTC AAC TGC CCC ACA GAG GAT GGT GGC TGG CAG GTC CAC GGT GTG ACC AGC TTT GTT TCT GCC TTT GGC TGC AAC TTC ATC TGG 30 AAG CCC ACG TGT TTC ACT CGA GTC TCC GCC TTC ATC GAC TGG ATT GAG GAG ACC ATA GCA AGC CAC - X (3'); (5')-GTT GTC AAT GGT GAG GAT GCG GTC CCC TAC AGC TGG CCC TGG CAG GTT TCC CTG CAG TAT 35 GAG AAA AGC GGA AGC TTC TAC CAC ACC TGT GGC GGT AGC CTC ATC GCC CCC GAC TGG GTT GTG ACT GCC GGC CAC TGC ATC TCG AGC TCC CGG ACC TAC CAG GTG GTG TTG GGC GAG TAC GAC CGT GCT GTG AAG GAG GGC CCC GAG CAG 40 GTG ATC CCC ATC AAC TCT GGG GAC CTC TTT GTG CAT CCA CTC TGG AAC CGC TCG TGT GTG GCC TGT GGC AAT GAC ATC GCC CTC ATC AAG CTC TCA CGC AGC GCC CAG CTG GGA GAC GCC GTC CAG CTC GCC TCA CTC CCT CCG GCT GGT 45 GAC ATC CTT CCC AAC GAG ACA CCC TGC TAC ATC ACC GGC TGG GGC CGT CTC TAT ACC AAC GGG CCA CTC CCA GAC AAG CTG CAG GAG GCC CTG CTG CCG GTG GTG GAC TAT GAA CAC TGC TCC AGG TGG AAC TGG TGG GGT TCC TCC GTG 50 AAG AAG ACC ATG GTG TGT GCT GGA GGG GAC ATC CGC TCC GGC TGC AAT GGT GAC TCT GGA GGA CCC CTC AAC TGC CCC ACA GAG GAT GGT GGC TGG CAG GTC CAT GGC GTG ACC AGC TTT GTT TCT GCC TTT GGC TGC AAC ACC CGC AGG 55 AAG CCC ACG GTG TTC ACT CGA GTC TCC GCC TTC ATT GAC TGG ATT GAG GAG ACC ATA GCA AGC CAC - X (3'); where X represents TAA, TGA or TAG. 6. A host transformed with a DNA in accordance with claim 3. 60 7. A host transformed with a DNA in accordance with claim 4. 8. A host transformed with a DNA in accordance with claim 5. 9. A transformed host according to claim 6 which is capable of expressing a compound functioning as a human pancreatic elastase. 10. A transformed host according to claim 7 which is capable of expressing a compound functioning as 65 a human pancreatic elastase.

35 tP 0 1 98 645 Bl

iransrormea nosi accoraing to claim o wnich is capable of expressing a compound functioning as a human pancreatic elastase. 12. A pharmaceutical composition which comprises a human pancreatic elastase according to claim 1 together with a pharmaceutically acceptable carrier. 5 Patentanspriiche

1 . Eine Verbindung, welche eine Aminosauresequenz fur menschliche Bauchspeichelelastase beinhaltet, die aus der von menschlicher Bauchspeichelelastase IIB, menschlicher Bauchspeichelelastase 10 IIIA, menschlicher Baucspeichelelastase IIIB und igendwelchen der menschlichen Bauchspeichelelasta- se IIB, IIIA und IIIB mit einer oder mehreren entfallenden, invertierten, ersetzten oder abgeanderten Aminosauren entsprechenden Verbindungen gebildeten Gruppe ausgewahlt ist, wobei die Aminosaure- sequenzen der menschlichen Bauchspeichelelastase IIB, der menschlichen Bauchspeichelelastase IIIA und der menschlichen Bauchspeichelelastase IIIB die folgenden sind: 15 menschliche Bauchspeichelelastase IIB: (N)-Cys Gly Val Ser Thr Tyr Ala Pro Asp Met Ser Arg Met Leu Gly Gly Glu Glu Ala Arg Pro Asn Ser Trp Pro Trp Gin Val Ser Leu Gin Tyr Ser Ser Asn Gly Gin Trp Tyr His 20 Thr Cys Gly Gly Ser Leu lie Ala Asn Ser Trp Val Leu Thr Ala Ala His Cys lie Ser Ser Ser Arg lie Tyr Arg Val Met Leu Gly Gin His Asn Leu Tyr Val Ala Glu Ser Gly Ser Leu Ala Val Ser Val Ser Lys He Val 25 Val His Lys Asp Trp Asn Ser Asn Gin Val Ser Lys Gly Asn Asp He Ala Leu Leu Lys Leu Ala Asn Pro Val Ser Leu Thr Asp Lys He Gin Leu Ala Cys Leu Pro Pro Ala Gly Thr He Leu Pro Asn Asn Tyr Pro Cys Tyr 30 Val Thr Gly Trp Gly Arg Leu Gin Thr Asn Gly Ala Leu Pro Asp Asp Leu Lys Gin Gly Arg Leu Leu Val Val Asp Tyr Ala Thr Cys Ser Ser Ser Gly Trp Trp Gly Ser Thr Val Lys Thr Asn Met He Cys Ala Gly Gly Asp 35 Gly Val He Cys Thr Cys Asn Gly Asp Ser Gly Gly Pro Leu Asn Cys Gin Ala Ser Asp Gly Arg Trp Glu Val His Gly lie Gly Ser Leu Thr Ser Val Leu Gly Cys Asn Tyr Tyr Tyr Lys Pro Ser lie Phe Thr Arg Val Ser 10 Asn Tyr Asn Asp Trp lie Asn Ser Val lie Ala Asn Asn - (C) menschliche Bauchspeichelelastase IIIA: (N)-Val Val His Gly Glu Asp Ala Val Pro Tyr Ser Trp Pro Trp Gin Val Ser Leu Gin Tyr t5 Glu Lys Ser Gly Ser Phe Tyr His Thr Cys Gly Gly Ser Leu He Ala Pro Asp Trp Val Val Thr Ala Gly His Cys lie Ser Arg Asp Leu Thr Tyr Gin Val Val Leu Gly Glu Tyr Asn Leu Ala Val Lys Glu Gly Pro Glu Gin i0 Val He Pro He Asn Ser Glu Glu Leu Phe Val His Pro Leu Trp Asn Arg Ser Cys Val Ala Cys Gly Asn Asp lie Ala Leu He Lys Leu Ser Arg Ser Ala Gin Leu Gly Asp Ala Val Gin Leu Ala Ser Leu Pro Pro Ala Gly 5 Asp He Leu Pro Asn Lys Thr Pro Cys Tyr He Thr Gly Trp Gly Arg Leu Try Thr Asn Gly Pro Leu Pro Asp Lys Leu Gin Gin Ala Arg Leu Pro Val Val Asp Tyr Lys His Cys Ser Arg Trp Asn Trp Trp Gly Ser Thr Val i0 Lys Lys Thr Met Val Cys Ala Gly Gly Tyr lie Arg Ser Gly Cys Asn Gly Asp Ser Gly Gly Pro Leu Asn Cys Pro Thr Glu Asp Gly Gly Trp Gin Val His Gly Val Thr Ser Phe Val Ser Ala Phe Gly Cys Asn Phe lie Trp 15 Lys Pro Thr Val Phe Thr Arg Val Ser Ala

3 EP 0 198 645 B1

Phe lie Asp Trp lie Glu Glu Thr lie Ala Ser His - (C) oder menschliche Bauchspeichelelastase IIIB: (N)-Val Val Asn Gly Glu Asp Ala Val Pro Tyr 5 Ser Trp Pro Trp Gin Val Ser Leu Gin Tyr Glu Lys Ser Gly Ser Phe Tyr His Thr Cys Gly Gly Ser Leu lie Ala Pro Asp Trp Val Val Thr Ala Gly His Cys He Ser Ser Ser Arg Thr Tyr Gin Val Val Leu Gly Glu Tyr 1 0 Asp Arg Ala Val Lys Glu Gly Pro Glu Gin Val He Pro He Asn Ser Gly Asp Leu Phe Vai His Pro Leu Trp Asn Arg Ser Cys Val Ala Cys Gly Asn Asp He Ala Leu He Lys Leu Ser Arg Ser Ala Gin Leu Gly Asp Ala 15 Val Gin Leu Ala Ser Leu Pro Pro Ala Gly Asp lie Leu Pro Asn Glu Thr Pro Cys Tyr He Thr Gly Trp Gly Arg Leu Tyr Thr Asn Gly Pro Leu Pro Asp Lys Leu Gin Glu Ala Leu Leu Pro Val Val Asp Tyr Glu His Cys 20 Ser Arg Trp Asn Trp Trp Gly Ser Ser Val Lys Lys Thr Met Val Cys Ala Gly Gly Asp He Arg Ser Gly Cys Asn Gly Asp Ser Gly Gly Pro Leu Asn Cys Pro Thr Glu Asp Gly Gly Trp Gin Val His Gly Val Thr Ser Phe 25 Val Ser Ala Phe Gly Cys Asn Thr Arg Arg Lys Pro Thr Val Phe Thr Arg Val Ser Ala Phe He Asp Trp He Glu Glu Thr lie Ala Ser His - (C). 2. Eine Verbindung nach Anspruch 1 , welche eine Proelastase, eine Praproelastase oder ein Fusions- 30 protein ist. 3. Eine DNS, welche eine Verbindung nach Anspruch 1 codiert. 4. Eine DNS, welche eine Verbindung nach Anspruch 2 codiert. 5. Eine DNS nach Anspruch 3, welche eine Basensequenz beinhaltet, die aus der von den folgenden drei Sequenzen und funktionell ahnlichen Basensequenzen gebildeten Gruppen ausgewahlt ist: 35 (5')-TGT GGG GTC TCC ACT TAC GCG CCT GAT ATG TCT AGG ATG CTT GGA GGT GAA GAA GCG AGG CCC AAC AGC TGG CCC TGG CAG GTC TCC CTG CAG TAC AGC TCC AAT GGC CAG TGG TAC CAC ACC TGC GGA GGG TCC CTG ATA GCC AAC AGC 40 TGG GTC CTG ACG GCT GCC CAC TGC ATC AGC TCC TCC AGG ATC TAC CGC GTG ATG CTG GGC CAG CAT AAC CTC TAC GTT GCA GAG TCC GGC TCG CTG GCC GTC AGT GTC TCT AAG ATT GTG GTG CAC AAG GAC TGG AAC TCC AAC CAG GTC 45 TCC AAA GGG AAC GAC ATT GCC CTG CTC AAA CTG GCT AAC CCC GTC TCC CTC ACC GAC AAG ATC CAG CTG GCC TGC CTC CCT CCT GCC GGC ACC ATT CTA CCC AAC AAC TAC CCC TGC TAC GTC ACA GGC TGG GGA AGG CTG CAG ACC AAC 50 GGG GCT CTC CCT GAT GAC CTG AAG CAG GGC CGG TTG CTG GTT GTG GAC TAT GCC ACC TGC TCC AGC TCT GGC TGG TGG GGC AGC ACC GTG AAG ACG AAT ATG ATC TGT GCT GGG GGT GAT GGC GTG ATA TGC ACC TGC AAC GGA GAC TCC 55 GGT GGG CCG CTG AAC TGT CAG GCA TCT GAC GGC CGG TGG GAG GTG CAT GGC ATC GGC AGC CTC ACG TCG GTC CTT GGT TGC AAC TAC TAC TAC AAG CCC TCC ATC TTC ACG CGG GTC TCC AAC TAC AAC GAC TGG ATC AAT TCG GTG ATT 60 GCA AAT AAC -X (3'); (5')-GTT GTC CAT GGT GAG GAT GCG GTC CCC TAC AGC TGG CCC TGG CAG GTT TCC CTG CAG TAT GAG AAA AGT GGA AGC TTC TAC CAC ACG TGT GGC GGT AGC CTC ATC GCC CCC GAT TGG GTT 65 GTG ACT GCC GGC CAC TGC ATC TCG AGG GAT

37 EP 0 198 645 B1

O I la AOO I AO UAla la I la la I la I I la lata 1 GAG TAC AAC CTT GCT GTG AAG GAG GGC CCC GAG CAG GTG ATC CCC ATC AAC TCT GAG GAG CTG TTT GTG CAT CCA CTC TGG AAC CGC TCG TGT GTG 5 GCC TGT GGC AAT GAC ATC GCC CTC ATC AAG CTC TCA CGC AGC GCC CAG CTG GGA GAT GCC GTC CAG CTC GCC TCA CTC CCT CCC GCT GGT GAC ATC CTT CCC AAC AAG ACA CCC TGC TAC ATC ACC GGC TGG GGC CGT CTC TAT ACC AAT 1 0 GGG CCA CTC CCA GAC AAG CTG CAG CAG GCC CGG CTG CCC GTG GTG GAC TAT AAG CAC TGC TCC AGG TGG AAC TGG TGG GGT TCC ACC GTG AAG AAA ACC ATG GTG TGT GCT GGA GGG TAC ATC CGC TCC GGC TGC AAC GGT GAC TCT GGA 15 GGA CCC CTC AAC TGC CCC ACA GAG GAT GGT GGC TGG CAG GTC CAC GGT GTG ACC AGC TTT GTT TCT GCC TTT GGC TGC AAC TTC ATC TGG AAG CCC ACG TGT TTC ACT CGA GTC TCC GCC TTC ATC GAC TGG ATT GAG GAG ACC ATA GCA 20 AGC CAC - X (31); (5')-GTT GTC AAT GGT GAG GAT GCG GTC CCC TAC AGC TGG CCC TGG CAG GTT TCC CTG CAG TAT GAG AAA AGC GGA AGC TTC TAC CAC ACC TGT GGC GGT AGC CTC ATC GCC CCC GAC TGG GTT 25 GTG ACT GCC GGC CAC TGC ATC TCG AGC TCC CGG ACC TAC CAG GTG GTG TTG GGC GAG TAC GAC CGT GCT GTG AAG GAG GGC CCC GAG CAG GTG ATC CCC ATC AAC TCT GGG GAC CTC TTT GTG CAT CCA CTC TGG AAC CGC TCG TGT GTG 30 GCC TGT GGC AAT GAC ATC GCC CTC ATC AAG CTC TCA CGC AGC GCC CAG CTG GGA GAC GCC GTC CAG CTC GCC TCA CTC CCT CCG GCT GGT GAC ATC CTT CCC AAC GAG ACA CCC TGC TAC ATC ACC GGC TGG GGC CGT CTC TAT ACC AAC 35 GGG CCA CTC CCA GAC AAG CTG CAG GAG GCC CTG CTG CCG GTG GTG GAC TAT GAA CAC TGC TCC AGG TGG AAC TGG TGG GGT TCC TCC GTG AAG AAG ACC ATG GTG TGT GCT GGA GGG GAC ATC CGC TCC GGC TGC AAT GGT GAC TCT GGA 10 GGA CCC CTC AAC TGC CCC ACA GAG GAT GGT GGC TGG CAG GTC CAT GGC GTG ACC AGC TTT GTT TCT GCC TTT GGC TGC AAC ACC CGC AGG AAG CCC ACG GTG TTC ACT CGA GTC TCC GCC TTC ATT GAC TGG ATT GAG GAG ACC ATA GCA J5 AGC CAC -X (3'); worin X fur TAA, TGA oder TAG steht. 6. Ein Wirt, welcher mit einer DNS gemaB Anspruch 3 transformiert ist. 7. Ein Wirt, welcher mit einer DNS gemaB Anspruch 4 transformiert ist. 8. Ein Wirt, welcher mit einer DNS gemaB Anspruch 5 transformiert ist. 50 9. Ein transformierter Wirt gemaB Anspruch 6, welcher befahigt ist, eine als menschliche Bauchspei- chelelastase fungierende Verbindung zu exprimieren. 10. Ein transformierter Wirt gemaB Anspruch 7, welcher befahigt ist, eine als menschliche Bauchspei- chelelastase fungierende Verbindung zu exprimieren. 11. Ein transformierter Wirt gemaB Anspruch 8, welcher befahigt ist, eine als menschliche Bauchspei- 15 chelelastase fungierende Verbindung zu exprimieren. 12. Pharmazeutische Mischung, welche eine menschliche Bauchspeichelelastase gemaB Anspruch 1 zusammen mit einem pharmazeutisch annehmbaren Trager enthalt. u nevenaicauons

1. Compose comprenant une sequence d'aminoacides d'une elastase pancreatique humaine choisie parmi I'elastase pancreatique humaine IIB, I'elastase pancreatique humaine IIIA, I'elastase pancreatique humaine IIIB et tous composes correspondant a I'elastase pancreatique humaine IIB, IIIA ou IIIB avec un 5 ou plusieurs aminoacides supprimes, inverses, remplaces ou modifies, lesdites sequences d'amino- EP 0 198 645 B1

acides de I'elastase pancreatique humaine IIB, de I'elastase pancreatique humaine IIIA et de I'elastase pancreatique humaine IIIB etant respectivement: elastase pancreatique humaine IIB: (N)-Cys Gly Val Ser Thr Tyr Ala Pro Asp Met 5 Ser Arg Met Leu Gly Gly Glu Glu Ala Arg Pro Asn Ser Trp Pro Trp Gin Val Ser Leu Gin Tyr Ser Ser Asn Gly Gin Trp Tyr His Thr Cys Gly Gly Ser Leu He Ala Asn Ser Trp Val Leu Thr Ala Ala His Cys lie Ser 1 o Ser Ser Arg lie Tyr Arg Val Met Leu Gly Gin His Asn Leu Tyr Val Ala Glu Ser Gly Ser Leu Ala Val Ser Val Ser Lys He Val Val His Lys Asp Trp Asn Ser Asn Gin Val Ser Lys Gly Asn Asp He Ala Leu Leu Lys 1 5 Leu Ala Asn Pro Val Ser Leu Thr Asp Lys He Gin Leu Ala Cys Leu Pro Pro Ala Gly Thr He Leu Pro Asn Asn Tyr Pro Cys Tyr Val Thr Gly Trp Gly Arg Leu Gin Thr Asn Gly Ala Leu Pro Asp Asp Leu Lys Gin Gly 20 Arg Leu Leu Val Val Asp Tyr Ala Thr Cys Ser Ser Ser Gly Trp Trp Gly Ser Thr Val Lys Thr Asn Met lie Cys Ala Gly Gly Asp Gly Val lie Cys Thr Cys Asn Gly Asp Ser Gly Gly Pro Leu Asn Cys Gin Ala Ser Asp 25 Gly Arg Trp Glu Val His Gly He Gly Ser Leu Thr Ser Val Leu Gly Cys Asn Tyr Tyr Tyr Lys Pro Ser He Phe Thr Arg Val Ser Asn Tyr Asn Asp Trp He Asn Ser Val He Ala Asn Asn - (C) 30 elastase pancreatique humaine IIIA: (N)-Val Val His Gly Glu Asp Ala Val Pro Tyr Ser Trp Pro Trp Gin Val Ser Leu Gin Tyr Glu Lys Ser Gly Ser Phe Tyr His Thr Cys Gly Gly Ser Leu He Ala Pro Asp Trp Val 35 Val Thr Ala Gly His Cys He Ser Arg Asp Leu Thr Tyr Gin Val Val Leu Gly Glu Tyr Asn Leu Ala Val Lys Glu Gly Pro Glu Gin Val lie Pro He Asn Ser Glu Glu Leu Phe Val His Pro Leu Trp Asn Arg Ser Cys Val 40 Ala Cys Gly Asn Asp lie Ala Leu He Lys Leu Ser Arg Ser Ala Gin Leu Gly Asp Ala Val Gin Leu Ala Ser Leu Pro Pro Ala Gly Asp lie Leu Pro Asn Lys Thr Pro Cys Tyr He Thr Gly Trp Gly Arg Leu Try Thr Asn 45 Gly Pro Leu Pro Asp Lys Leu Gin Gin Ala Arg Leu Pro Val Val Asp Tyr Lys His Cys Ser Arg Trp Asn Trp Trp Gly Ser Thr Val Lys Lys Thr Met Val Cys Ala Gly Gly Tyr He Arg Ser Gly Cys Asn Gly Asp Ser Gly 50 Gly Pro Leu Asn Cys Pro Thr Glu Asp Gly Gly Trp Gin Val His Gly Val Thr Ser Phe Val Ser Ala Phe Gly Cys Asn Phe He Trp Lys Pro Thr Val Phe Thr Arg Val Ser Ala Phe He Asp Trp lie Glu Glu Thr He Ala 55 Ser His - (C) ou elastase pancreatique humaine IIIB: (N)-Val Val Asn Gly Glu Asp Ala Val Pro Tyr Ser Trp Pro Trp Gin Val Ser Leu Gin Tyr Glu Lys Ser Gly Ser Phe Tyr His Thr Cys 60 Gly Gly Ser Leu lie Ala Pro Asp Trp Val Val Thr Ala Gly His Cys lie Ser Ser Ser Arg Thr Tyr Gin Val Val Leu Gly Glu Tyr Asp Arg Ala Val Lys Glu Gly Pro Glu Gin Val He Pro He Asn Ser Gly Asp Leu Phe 65 Val His Pro Leu Trp Asn Arg Ser Cys Val

39 EP 0198 645 B1

aia uys uiy Asn Asp lie Aia Leu lie Lys Leu Ser Arg Ser Ala Gin Leu Gly Asp Ala Val Gin Leu Ala Ser Leu Pro Pro Ala Gly Asp lie Leu Pro Asn Glu Thr Pro Cys Tyr 5 lie Thr Gly Trp Gly Arg Leu Tyr Thr Asn Gly Pro Leu Pro Asp Lys Leu Gin Glu Ala Leu Leu Pro Val Val Asp Tyr Glu His Cys Ser Arg Trp Asn Trp Trp Gly Ser Ser Val Lys Lys Thr Met Val Cys Ala Gly Gly Asp 1 0 He Arg Ser Gly Cys Asn Gly Asp Ser Gly Gly Pro Leu Asn Cys Pro Thr Glu Asp Gly Gly Trp Gin Val His Gly Val Thr Ser Phe Val Ser Ala Phe Gly Cys Asn Thr Arg Arg Lys Pro Thr Val Phe Thr Arg Val Ser Ala 15 Phe He Asp Trp He Glu Glu Thr lie Ala Ser His - (C). 2. Compose selon la revendication 1 , lequel est une proelastase, une preproelastase ou une proteine de fusion. 3. ADN codant pour un compose selon la revendication 1 . 20 4. ADN codant pour un compose selon la revendication 2. 5. ADN selon la revendication 3, lequel comprend une sequence de bases choisie parmi les trois sequences suivantes et les sequences de bases fonctionnellement similaires: (5')-TGT GGG GTC TCC ACT TAC GCG CCT GAT ATG TCT AGG ATG CTT GGA GGT GAA GAA GCG AGG 25 CCC AAC AGC TGG CCC TGG CAG GTC TCC CTG CAG TAC AGC TCC AAT GGC CAG TGG TAC CAC ACC TGC GGA GGG TCC CTG ATA GCC AAC AGC TGG GTC CTG ACG GCT GCC CAC TGC ATC AGC TCC TCC AGG ATC TAC CGC GTG ATG CTG GGC 30 CAG CAT AAC CTC TAC GTT GCA GAG TCC GGC TCG CTG GCC GTC AGT GTC TCT AAG ATT GTG GTG CAC AAG GAC TGG AAC TCC AAC CAG GTC TCC AAA GGG AAC GAC ATT GCC CTG CTC AAA CTG GCT AAC CCC GTC TCC CTC ACC GAC AAG 35 ATC CAG CTG GCC TGC CTC CCT CCT GCC GGC ACC ATT CTA CCC AAC AAC TAC CCC TGC TAC GTC ACA GGC TGG GGA AGG CTG CAG ACC AAC GGG GCT CTC CCT GAT GAC CTG AAG CAG GGC CGG TTG CTG GTT GTG GAC TAT GCC ACC TGC ♦0 TCC AGC TCT GGC TGG TGG GGC AGC ACC GTG AAG ACG AAT ATG ATC TGT GCT GGG GGT GAT GGC GTG ATA TGC ACC TGC AAC GGA GAC TCC GGT GGG CCG CTG AAC TGT CAG GCA TCT GAC GGC CGG TGG GAG GTG CAT GGC ATC GGC AGC 15 CTC ACG TCG GTC CTT GGT TGC AAC TAC TAC TAC AAG CCC TCC ATC TTC ACG CGG GTC TCC AAC TAC AAC GAC TGG ATC AAT TCG GTG ATT GCA AAT AAC -X (3')" (5')-GTT GTC CAT GGT GAG GAT GCG GTC CCC TAC 30 AGC TGG CCC TGG CAG GTT TCC CTG CAG TAT GAG AAA AGT GGA AGC TTC TAC CAC ACG TGT GGC GGT AGC CTC ATC GCC CCC GAT TGG GTT GTG ACT GCC GGC CAC TGC ATC TCG AGG GAT CTG ACC TAC CAG GTG GTG TTG GGT GAG TAC 15 AAC CTT GCT GTG AAG GAG GGC CCC GAG CAG GTG ATC CCC ATC AAC TCT GAG GAG CTG TTT GTG CAT CCA CTC TGG AAC CGC TCG TGT GTG GCC TGT GGC AAT GAC ATC GCC CTC ATC AAG CTC TCA CGC AGC GCC CAG CTG GGA GAT GCC iO GTC CAG CTC GCC TCA CTC CCT CCC GCT GGT GAC ATC CTT CCC AAC AAG ACA CCC TGC TAC ATC ACC GGC TGG GGC CGT CTC TAT ACC AAT GGG CCA CTC CCA GAC AAG CTG CAG CAG GCC CGG CTG CCC GTG GTG GAC TAT AAG CAC TGC i5 TCC AGG TGG AAC TGG TGG GGT TCC ACC GTG

0 EP 0 198 645 B1

AAG AAA ACC ATG GTG TGT GCT GGA GGG TAC ATC CGC TCC GGC TGC AAC GGT GAC TCT GGA GGA CCC CTC AAC TGC CCC ACA GAG GAT GGT GGC TGG CAG GTC CAC GGT GTG ACC AGC TTT 5 GTT TCT GCC TTT GGC TGC AAC TTC ATC TGG AAG CCC ACG TGT TTC ACT CGA GTC TCC GCC TTC ATC GAC TGG ATT GAG GAG ACC ATA GCA AGC CAC - X (3'); (5')-GTT GTC AAT GGT GAG GAT GCG GTC CCC TAC 1 0 AGC TGG CCC TGG CAG GTT TCC CTG CAG TAT GAG AAA AGC GGA AGC TTC TAC CAC ACC TGT GGC GGT AGC CTC ATC GCC CCC GAC TGG GTT GTG ACT GCC GGC CAC TGC ATC TCG AGC TCC CGG ACC TAC CAG GTG GTG TTG GGC GAG TAC 1 5 GAC CGT GCT GTG AAG GAG GGC CCC GAG CAG GTG ATC CCC ATC AAC TCT GGG GAC CTC TTT GTG CAT CCA CTC TGG AAC CGC TCG TGT GTG GCC TGT GGC AAT GAC ATC GCC CTC ATC AAG CTC TCA CGC AGC GCC CAG CTG GGA GAC GCC 20 GTC CAG CTC GCC TCA CTC CCT CCG GCT GGT GAC ATC CTT CCC AAC GAG ACA CCC TGC TAC ATC ACC GGC TGG GGC CGT CTC TAT ACC AAC GGG CCA CTC CCA GAC AAG CTG CAG GAG GCC CTG CTG CCG GTG GTG GAC TAT GAA CAC TGC 25 TCC AGG TGG AAC TGG TGG GGT TCC TCC GTG AAG AAG ACC ATG GTG TGT GCT GGA GGG GAC ATC CGC TCC GGC TGC AAT GGT GAC TCT GGA GGA CCC CTC AAC TGC CCC ACA GAG GAT GGT GGC TGG CAG GTC CAT GGC GTG ACC AGC TTT 30 GTT TCT GCC TTT GGC TGC AAC ACC CGC AGG AAG CCC ACG GTG TTC ACT CGA GTC TCC GCC TTC ATT GAC TGG ATT GAG GAG ACC ATA GCA AGC CAC - X (3'); X representant TAA, TGA ou TAG. 35 6. Hote transforme par un ADN selon la revendication 3. 7. Hote transforme par un ADN selon la revendication 4. 8. Hote transforme par un ADN selon la revendication 5. 9. Hote transforme selon la revendication 6, lequel est capable d'exprimer un compose fonctionnant comme une elastase pancreatique humaine. 40 10. Hote transforme selon la revendication 7, lequel est capable d'exprimer un compose fonctionnant comme une elastase pancreatique humaine. 1 1 . Hote transforme selon la revendication 8, lequel est capable d'exprimer un compose fonctionnant comme une elastase pancreatique humaine. 12. Composition pharmaceutique comprenant une elastase pancreatique humaine selon la revendica- 15 tion 1 , conjointement avec un vehicule pharmaceutiquement acceptable.

fcr U lyo 040 Bl

fvun Mg.z. rvun Pstiv\ PvuIIx^

HindIIl\.

mi Hindlll Hindlll digestion

rsu mnam i I findlll I —

partial digestion IPvuIIPstI lindUI/Bglll PvuII j/3 digestion 1 1 n a in J — — — I

linrJIII BgLII sti partial digestion

'Stl PvuII -J

i nuclease 1 nuclease PvuII

UD NA ligase EP 0 198 645 B1

Fig.3. Pvu" HindlH PvuII „ Hindlll PvulK^^PvuII Hindlll digestion ^XHindIII f^X^ll erlindlll HindI11 r.nrfin pairs)� c1,ndI11 (A28base Smal * I Kind III digestion HpallIpall digestion iindltt/Smal digestion Hindlll Pvul i Hindlll ■ ■

Hindlll Hindlll ^«PQU 'vull partial Avail digestion C^aSoiqI (385base pairs) Hindlll^ PvuII

(4566 base pairs) [Avail partial Avail \ PvuII d|9estion Hpall AvnTT

Synthetic DNA ^linker

DNA ligase PvuII Hpall 1 t

frT4 DNA ligase Hindlll Hpall

pBSEX-E2Af J

7SmaI/\ iPvuII / signal peptidase recognition site

1 I paH Avail i'lCGGCGGCT GCG TGT GGGUI InCGCCGACGCACACncqn) 5'

ALA ALA ALA CYS GLY EP 0 198 645 B1

pvuII PvuII Fig. 5 PvuIIv^i^PvuII /^^\V,PvuII

Hindlll Smal Iqc £

digestion jHindlll

Hindlll PvuII Hindlll 1 — i 1— //-«

| PvuII partial digestion Odel i Hindlll Odel Ddel PvuII Smal digestion

I Odel partial digestion * Smal Ddel PvuII Smal

lfK°&8l/dNTPs

.pvuii

T4 ONA ligase

EP 0 198 645 B1

Fig. 7. PvuII

Hindlll

PstI Hindlll ion jHindlll digest

Hindlll PstI Pvyll Hindlll 1 1 1 /A-i ^ I PvuII partial digestion PstI' \ Hind III/Bg III PvuII Hindlll [^r^l digestion

I PstI partial digestion Hindlll Bglll PstI * PvuII r> 1

i SI nuclease PvuII

S1 nuclease

U DNA ligase EP 0 198 645 B1

PvuII Hindlll PvuIIvJ — ^JvuII PvuII

(«8base Hindlll V < Hindlll pairs) Smal I Hindlll digestion Hpall digestion Hindlll/Smal Hindlll PjVuH Hindlll digestion ~ Hindlll J— - — it-* ^ Hpall I PvuII partial Hindm1" PvuII *'W*'m (385 base pairs) Hindlll

IHhal digestion Hpall Hhal Hhal PvuII

Synthetic DNA linker ~T \ ^Smal (4566 base pairs)

UONA ligase

Hpall PvuII

DNA JH ligase

HindUI ^*HpaII

/SmaI/\ IPvuII J EP 0 198 645 B1

Fig. 9.

Signal peptidase recognition site

Hpall Hhal

5' CG GCG GCT GCG TGT GGG GTC TCC ACT TAC GCG 3;

3' i7] C6C CGA CGC ACA CCC CAG AGG TGA ATG C| 5/

Ala Ala Ala Cys Gly Val Ser Thr Tyr Ala fcHU 198 645 Bl

Fig.10.

T4 DNA ligase f EP 0 198 645 B1

Fig.11 .

SaoIIIA-Mjf^ Hindlll

Saul IIA fragment^ „^n?r

Hindlll sauIIIA / Bgl *////////>* SauIIIA

Blunt ends Hindlll pSV2-CL2 Htndlll/Bgt

Hindlll Bgl

Hindlll digestion

Addition of Hindlll linker

Hindlll

Hindlll digestion

SEZZh

Hindlll

Hindlll EP 0 198 645 B1

Fig. 12. y Hindlll u. jitt Bgl II

f\ Htndlll Hindlll digestion 1— {pTiitrnfl j ^ HinHTTf Hindlll-] V_

Hindlll /Bell Hpall digestion Hind III /Bglll digestion digestion

Hindlll •-^Hpall (385 base pairs)

. Hindlll Hpall Hindm Hindlllindlll Bglll

Synthetic OHAv / (712 base pairs) "Bell (4173 base pairs) ^

(805 base pairs)

T^DNA ligase

Hindlll x-^Hpall

PEX010J

(Bell/ Bglll) Recognition site of the signal peptidase H n ^-1 1 r— 6C0 LCC GCT GCG AGT GCT GTT GCT CAT fifif f& 111 Si 3*1 CGC CGA CGC TCA CGA CAA Ala Ala Ala Ser Ala Val Val His Gly

rfr SE5 SB CCC TAC AGC TGG CCC TGG CTC CTA CGC CAG GGG ATG TCG ACC GGfi ACC wo Asp Ala Val Pro Tyr Ser Trp Pro Trp iindlll CAG GTT TCC CTG CAG TAT GAG AAA AGT GGAl3 ??tAU GTC CAA AGG GAC GTC ATA CTC TTT tpa liln Val Ser Leu Gin Tyr Glu Lys Ser Gly EP 0198 645 B1

Bgl II EP 0 198 645 B1

Fig .15.

I 2 3 A 5 ^aif t:"-(.«K*.BJ * i^miHwu-t— nw ■Jill Miwr I

1 : Molecular weight marker

2:E.coli YA21/pHEX002

3: E. coli HB101/pHEX002

4 :E.coli MCA100/pHEX002

5 E: coli X9B4/ pHEX002 EPO 198 645 B1

Fig.16

Pstl-W Hindlll - PstI 540bp Pstl-^ fragment / PstI Hindni/Bgl Bgl Hindlll^ Partial .fragment/ digestion

S1 nuclease Hindlll PSV2-CL1 Hindlll/Bgl

Hindlll V/W///A

Hindlll digestion Addition of Hindlll linker

Hindlll

Hindlll digestion

^JZZZZn

Hindlll Hindlll EP 0 198 645 B1

Fig.17.

Hindin Hindiii \ f ,-HindHI Hindlll . (pTUB228) Ji2f±£!L HindIH-| Hindin J rH,ndII! Bell U28 base pairs)

Hindlll/Bcll Hpall digestion Hindlll/Bglll digestion digestion

Hindlll "-♦Hpall (385 base pairs) Hindlll

Hpall Hindlll Hindlll Bglll C^BclI Synthetic DNA / (712 base pairs) (85 base pairs) Y (4173 base pairsL HpII 1T4DMA ligase Bglll

(797 base pairs) T4ONA ligase

Hindlll Hpall

(Bell /Bglll) Fig.18.

Recognition site of the signal peptidase *Val Pro Ala Ala Ala Val Asn Gly Glu Asp 1 CG GCG GCT GCG GTT GTC AAT GGT GAG GAT 3' CGC CGA CGC CAA CAG TTA CCA CTC CTA

Recognit ion site of Hpall

kla Val Pro Tyr ber irp fro irp uin vai CG GTC CCC TAC AGC TGG CCC TGG CAG GTT ;GC CAG GGG ATG TCG ACC GGG ACC GTC CAA

ler Leu Gin Tyr Blu Lys ber uiy oer fCC CTG CAG TAT GAG AAA AGC GGA -3' tGG GAC GTC ATA CTC TTT TCG CCT TCG A-5' T Recognition site of Hindlll EP 0 198 645 B1

Fig.19.

Elastase

EcoRI

Smal digestion EcoRI /Bglll digestion

EcoRI Bglll SmQj Smal

Fnu4HI partial digesti S1 nuclease

Blunt end

(789 base pairs)

T^DNA ligase