US005681733A United States Patent 19 (11 Patent Number: 5,681,733 Su et al. 45 Date of Patent: Oct. 28, 1997

54 NUCLEICACID SEQUENCES AND OTHER PUBLICATIONS EXPRESSION SYSTEMS FOR HEPARNASE Lohse, et al., The Journal of Biological Chemistry, vol. AND HEPARNASE DERVED FROM 267:34, pp. 24347-24335 (1992). FLAVOBACTERIUM HEPARINUM Sasisekharan, et al., Proc. Natl. Acad. Sci., USA, vol. 90, pp. 75) Inventors: Hongsheng Su, Longueuil; Francoise 3660-3664 (1993). Blain, Mtl.; Clark Bennett, Zimmermann, et al., Applied and Environmental Microbi Piérrefonds; Kangfu Gu, D.D.O., all of ology, vol. 56:11, pp. 2593-3594 (1990). Canada; Joseph Zimmermann, Elm Primary Examiner-Rebecca E. Prouty Grove, Wis.; Roy Musil, Carlsbad, Attorney, Agent, or Firm-Hale and Dorr LLP Calif. 57 ABSTRACT 73) Assignee: Ibex Technologies, Montreal, Canada The present invention describes the isolation and sequence of genes from Flavobacterium heparinum encoding heparin 21 Appl. No.: 258,639 and heparan sulfate degrading enzymes, heparinase II and 22 Fied: Jun. 10, 1994 heparinase III(EC 4.2.2.8). It further describes a method of expressing and an expression for heparinases I, II and III (51 Int. Cl. C12N 9788; C12N 15/60

using a modified ribosome binding region derived from a 52 U.S. C...... 435/232; 536/23.2 promoter from glycosaminoglycan lyase genes of F. hepari 58) Field of Search ...... 435/232; 536/23.2 num. Also, a multi-step protein purification method incor porating cell disruption, cation exchange chromatography, 56) References Cited affinity chromatography and hydroxylapatite chromatogra phy is outlined. Antibodies against a post-translational U.S. PATENT DOCUMENTS modification moiety common to Flavobacterium heparinum 5,389,539 2/1995 Sasisekharan et al...... 435/220 proteins and a method to obtain antibodies specific to these moieties and to the amino acid sequences of heparinases I, FOREIGN PATENT DOCUMENTS II and III are described. WO 9308289 4/1993 WIPO. WO 94.2618, 6/1994 WIPO. 6 Claims, 11 Drawing Sheets U.S. Patent Oct. 28, 1997 Sheet 1 of 11 5,681,733

plBhep AGGAAACAGAATTCATG S-D 10nt

pGhep AGGAGACAGAATTCATG

S-D 9nt p A 4hep AGGAGAATTCATG S-D 5 nt pGB AGGAGACAGGATCC S-D Ban II

FIG. 1 U.S. Patent Oct. 28, 1997 Sheet 2 of 11 5,681,733

SYNTHETIC OLIGONUCLEOTDES F. HEPARINUM Bom DIGESTED 5-CCCCCAGCCAGATTGTACATTCAGC-5 5-AAACCCCCCGACATTCCCGAACTAAAGA-3 CHROMOSOMAL DNA ADASH PHAGE ARMS

N

mm 350 bp PCR PRODUCT LIGATION

NOT LIGATION MDASH N N GENE LIBRARY

PLAOUE HYBRIDIZATION

SBPEHCK

pBLUESCRIPT (2961 bp)

LIGATION

BPEHCK

BPEHCK BPEHCK

pBSIB6-7 pBSIB6-21 (8.4 kbp) (8.4 kbp) SB SB

FIG.2 U.S. Patent Oct. 28, 1997 Sheet 3 of 11 5,681,733

P/ B YaN, MDASH (4.9 kbp) GENE LIBRARY

Bom H1

SYNTHETIC OLCONUCLEOTDES 5-TCAGGATTCATGCAAACCAAGGCCGG LIGATION ATGTGGTTTCCAA-5 B 5-GGACGATAACCACATTCGAGCATT-3

PCR

PARTIAL S.Pst Z DIGESTIONAND -mm-mm P/ B RE-LIGATION hep2 (5') U.S. Patent Oct. 28, 1997 Sheet 4 of 11 5,681,733 ATGAAAACACAATTATACCT GTATGTGATTTTTGTTCTAG TTGAACTTAT GGTTTTTACA 60 ACAAAGCCCT ATTCCCAAAC CAAGGCCGAT GTGGTTTGGA AAGACGTCGA TGGCCTATCT 120 ATCCCCATAC CCCCTAAGAC CCACCCGCGT TTCTATCTAC GTGACCAGCA AGTTCCTGAC 80 CTGAAAAACA CGATGAACGA CCCTAAACTG AAAAAAGTTT CGGCCGATAT GATCAAGATC 240 CAGGAAGACT CGAAGCCACC TGAATTCCT GAACTTAAAG ACTTTCGTTT TTATTTTAAC 500 CAGAAACCCC TACTGTAAG GGTTGAACTA ATCCCCCIGA ACTATCTGAT GACCAACCAT 560 CCAAAGGTAG GACGCGAAGC CATCACTTCA ATTATTGAA CCCTTGAAAC TCCAACTTTT 420 AAACCACCAC GTGATATTTC GAGAGGGATA CGCCTGTTTA TCG TACAGGGGCCATTGTG 480 TATGACTGGT CCTACCACA GCTGAAACCA GAAGAGAAAA CACGTTTTGT GAAGCCATTT 540 GTGACGCCG (CCAAAAIGCT CGAATGTGGT TATCCTCCGG TAAAAGACAA CTCTATTGTT 600 CCCCATCCT CCGAATCGAT GATCATGCCG GACCTCCTTT CTGTAGGGAT TCCCATTTAC 660 GATGAATTCC CTGAGATGTA TAACCTGGCT GCGGGTCGTTTTTTCAAAGA ACACCTGGTT 720 CCCCGCAACT CGTTTTATCC CTCCCATAAC TACCATCACG GTATGTCAA CCTGAACGTA 780 AGATTTACCA ACGACCTTTT TCCCCTCTCC ATATTAGACC CGATCGGCGC TCGTAATGTG 840 TTTAATCCAG GGCACCACTT TATCCTTTAT GACGCGATCT ATAAACGCCC CCCCGATGGA 900 CAGATTTAG CAGGTGGAGA TGTAGATTAT TCCAGGAAAA AACCAAAATA TATACGATG 960 CCTCCATTCCTTGCAGGTAG CTATATAAA GATGAAACC TTAATTACGA ATCCTGAAA 020 CATCCCAATC TTGAGCCACA TTCCAAATTC TTCCAATTTT TATCCCCCCA TACCCACTTC 1080 CGAAGTCGTA AGCCTGATGA TTTCCCACTT TCCAGGTACT CACCATCCCC TTTCCACC 1140 ATGATTGCCC GTACCGGATG GGGTCCGGAA AGTGTGATTG CAGAGATGAA AGTCAACGAA 1200 FIG.4A U.S. Patent Oct. 28, 1997 Sheet 5 of 11 5,681,733

TATTCCTTC TTAACCATCA GCATCAGGAT GCAGGAGCCT CCAGATCTA TTACAAAGGC 260 CCGCTGGCCA TAGATGCAGGCTCGTATACA GGTTCTTCAG GACGTTATAA CAGTCCCCAC 1520 AACAAGAACT TTTTTAAGCG CCATIC CACAATAGCT TGCTGATTTA CGATCCAAA 1380 GAAACTTTCA GTTCGTCGGG AATCGTGGA AGTGACCATA CCGATTTTCC TCCCAACGAT 440 CGTCGTCAGCCCCTCCCCCG AAAAGGTTGG ATTGCACCCC GCGACCTTAA ACAAATCCTG 500 GCAGGCGATT ICAGGACCGG CAAAATTCTT GCCCAGGGCT TTGGTCCCGA TAACCAAACC 560 CCTGAAA CTTATCTGAA AGGAGACATT ACACCACCTT ATICCCM AGTGAAGGAA 1620 GTAAAACGT CATTTCTATT CCTGAACCTT AAGGATGCCA AAGTCCCGC ACCGAGATC 1680 GTTTTTGACA ACGTAGTTCCTTCCAATCC GATTTTAAGA AGTTCTGGTT GTTGCACAGT 1740 ATTGACCACC CTGAAATAAA GGGGAATCAG ATTACCATAA AACGTACAAA AAACCGTGAT 800 AGTGGGATGT TGGTGAATAC GGCTTTCCTG CCGGATGCGG CCAATTCAAA CATTACCTCC 1860 ATTCGCGGCA AGGGCAAAGA CTCTGGGTG TTTGGTACCA AFTATACCAA TGATCCTAAA 1920 CCGGGCACGG ATGAACCATT GGAACGTCGA GAATGGCGTG TGGAAATCACTCCAAAAAAG 1980 GCAGCAGCCG AAGATTACTA CCTGAATGG ATACAGATTG CCGACAATAC ACAGCAAAAA 2040 TTACACGAGG TGAAGCGTATTGACGGGAC AAGGTTGTTG GTGGCAGCT TCCTGACAGG 200 ATACTTACTT TACCAAAAC TTCAGAAACT GTTGACGTCCCTTGGCTT TTCCGTTGTT 260 GGTAAAGGAA CATTCAAATT TGTGATGACC GATCTTTTAG CGGGTACCTG GCACGTCCG 2220 AAAGACCCAA AAATACTTTA TCCTGCGCT TCTCCAAAAG GTGATGATGG ACCCCTTTAT 2280 TTTGAAGGAA CTGAAGGAAC CTACCGTTTT TGAGATAA 2519

U.S. Patent Oct. 28, 1997 Sheet 7 of 11 5,681,733

SYNTHETIC OLGONUCLEODES F. HEPARINUM 5-GTNCCATTRTCNGCCTGGGCATGAAATCNCC-3 5-GNCATCAGTYCAGCCNCATAAAGGNTATGG-5 CHROMOSOMAL DNA

SBPEHCK

pBLUESCRIPT B C B (2961 bp) H 999 bp PCR PRODUCT

Bom H1/Clo 1

LIGATION

SB ADASH GENE LIBRARY CK pFB1 (3548 bp) PAQUE HYBRDZATION

Ps pBLUESCRIPT (296 bp)

pHindIIIBD pFB4 (5.63 kbp) (3.63 kbp)

FIG.6 U.S. Patent Oct. 28, 1997 Sheet 8 of 11 5,681,733

SYNTHETIC OIGONUCLEOTIDES 5-CGCGGATCCATCCAAAGCTCTCCATT-3 5-OGCCGATCCTCAAAGCTTGCCTTTCTC-3

pGB B BS HB 1333 BASE PAIR (4918 bp) 4-H PCR PRODUCT limitialispanoasagaa hep3A 3'

Born H

pGB-H3A3' (6251 bp) U.S. Patent Oct. 28, 1997 Sheet 9 of 11 5,681.733 ATGACTACGA AAATTTTTAA AAGGATCATT GTATTTCCTG TAATTGCCCT 50 ACGTCGGGA AATAACTTG CACAAAGCTC TTCCATTACC AGGAAAGATT 100 TTCACCACAT CAACCTTGAG TATTCCCCAC TGGAAAAGGT TAATAAAGCA 50 GTTCCTGCCG GCAACTATGA CGATGCGGCC AAAGCATTAC TCGCATACTA 200 CAGGGAAAAA AGTAACGCCA GGGAACCTGA TTTCAGTAAT GCAGAAAAGC 250 CTGCCGATAT ACGCCAGCCC ATAGATAAGGTTACGCGTGA AATCGCCGAC 300 AACGCTTTCG TCCACCAGTT TCAACCCCAC AAAGGCACG GCTATTTTGA 350 TTATGGAAA GACATCAACT GGCAGATGTG CCCCGAAAA GACAATGAAG 400 TACGCTCGCA GTTGCACCGT GTAAAATCGT CGCACGCTAT GGCCCTGGTT 450 TATCACGCTA CGGGOGATGAAAAATATGCA AGAGAATCGG TATATCAGTA 500 CACCCATTCG GCCAGAAAAA ACCCATTGGG CCTCTCCCAG GATAATGATA 550 AATTTGTGTG GCGGCCCCTTGAACTGTCGG ACAGGGTACA AAGTCTTCCC 600 CCAACCTTCA CCTTATTTGT AAACTCCCCA CCCTTACCC CAGCCTTTTT 650 AATCGAATTT TTAAACAGTT ACCACCAACA CGCCGATTAT TTATCTACGC 700 ATAGCCGA ACAGGGAAAC CACCGTTTA TTGAACCCCA ACGCAACTG 750 TTGCAGGGG ACTTCCC TGAATTTAAA GATTCACCAA GACGAGCCA 800 AACCCGCATA TCGGGCTCA ACACCGAGAT CAAAAAACAC GTTTATGCCG 850 ATCGGATGCA GTTTGAACTT TCACCAATTT ACCATGTAGC TGCCATCGAT 900 ATCTTCTTAA AGGCCTATGG TTCTCCAAAA CGAGTTAACC TTGAAAAAGA 950 ATTTCCCCAA TCTTATGTAC AAACTGTAGA AAATATGATT ATCCCCCTCA 000 FIG.8A U.S. Patent Oct. 28, 1997 Sheet 10 of 11 5,681,733 TCAGTATTTC ACTCCCAGAT TATAACACCC CTATGTTTCG AGATTCATCG 1050 ATTACAGATA AAAATTTCAC GATGGCACAGTTTCCCACCT CCCCCCCGC 100 TTTCCCCCCA AACCACCCCA TAAAATATT TGCTACAGAT GCCAAACAAG 50 GTAACGCGCC TAACTTTTTA TCCAAAGCAT GACCAATCC. GCITA 200 ACGTTTAGAA GCGGATGGGA AAAAATGCA ACCGTTATCG TATTAAAAGC 1250 CACTCCTCCC CGAGAATTTC ATCCCCACCC GGATAACGCC ACTTTTGAAC 300 TTTTATAAA CGCCAGAAAC TTTACCCCAG ACCCCCGGGT ATTTGTGTAT 1350 ACCCCCCACG AAGCCATCAT GAAACTGCCG AACTGGTACC GTCAAACCCG 1400 CATACACAGC ACGCTTACAC TCGACAATCA AAATATGGTC ATTACCAAAG 1450 CCCGGCAAAA CAAATCGGAA ACACGAAATA ACCTTGATGT CCTACCTAT 1500 ACCAACCCAA GCTATCCGAA TCTCGACCAT CAGCGCAGTG TACTTTTCAT 1550 CAACAAAAAA TACTTTCTCG TCATCGATAG CCCAATAGGC GAAGCTACCG 1600 GAAACCTGGG CGTACACTGG CAGCTTAAAG AAGACAGCAA CCCTGTTTTC 650 GATAACACAA AGAACCGGGT TTACACCACT TACAGAGATG GTAACAACC, 1700 GATGACCAA TCGTTGAATG CCGACAGGACCACCCTCAAT GAAGAAGAAG 1750 GAAAGGTATC TTATCTTTACAATAACGAGCTGAAAAGACC TCCTTTCCTA 1800 TTTGAAAAGC CTAAAAAGAA Iccococci CuTITC TCAGTATAGT 1850 TTATCCATAC GACCCCCAGA ACGCTCCAGA GATCAGCATA CGGGAAAACA 1900 AGGGCAATGA TTTTGAGAAA GGCAAGCTTA ATCTAACCCT TACCATTAAC 950 CGAAAACAAC ACCTTGTGTT GGTTCCTTAG 1980

FIG.8B U.S. Patent Oct. 28, 1997 Sheet 11 of 11 5,681,733

MTTKFKRI VFAVIALSSC NLAOSSSIT RKDFDHNLEYSCLEKWNKA WAACNYDDAA KALWHF PHKGYGYFDYK KALLAYYREK SKAREPDFSNAEKPADROP DKVTREMAD KALWHOFOPH KGYGYFDYCK PEPTIDE 3C

DIW LIK -NEWRHR WK DNWCMPVK DNEVRWOLHR WKWOAMALW YHATGDEKYA REWYOYSDWARKNPLGLSQ PEPTIDEA DNDKFWRPL EVSDRVOSLPPTFSLFVNSPAFTPAFLNEF LNSYHOOADY LSTHYAECGN HRLFEAORNL FACVSFPEFK DSPRWROTG SVLNTEKKO WYADGMCFEL SPYHWAAD IFLKAYGSAK RVNLEKEFPOSYWOTVENMI MALISISLPDYNTPMFGDSW TDKNFRMAO WKASPP FASWARVFPA NOAKYFATD CKOGKAPNFL SKALSNAGFY FRSCWOKNA TVMVLKASPP GEFH4OPWG IFELF GEFHAQPDNG TFELFKGRN FTPDACVFWY SCDEAMKLR NWYRQTRHS TTLDNQNW PEPTIDE 33 TKARQNKWE TGNNLDVLTY TNPSYPNLDH QRSVLFINKK YFLVIDRAIG EATGNLGVHW OLKEDSNPVF DKTKNRWYTT YRDGNLMIG SLNADRSLNEEECKVSYWY NKELKRPAFW FEKPKKNAGT (NFWSWYPY DGQKAPEIS RENKGNOFEK CKLNLTLT IN GKQQVLVP

FIG.9 5,681,733 1. 2 NUCLEICACID SEQUENCES AND 1849-1857 (1985), heparinase II as described by Zimmer EXPRESSION SYSTEMS FOR HEPARNASE mann and Cooney, U.S. Pat. No. 5,169,772, and heparinase DAND HEPARNASE DERVED FROM III (E.C. 4.2.2.8) as described by Lohse and Linhardt, J. FLAWOBACTERIUM HEPARINUM Biol. Chem. 267: 24347- 24355 (1992). These enzymes catalyze an eliminative cleavage of the (d.1-4) carbohy BACKGROUND OF THE INVENTION drate bond between glucosamine and hexuronic acid resi dues in the heparin/heparan sulfate backbone. The three This invention is directed to cloning, sequencing and enzyme variants differ in their action on specific carbohy expressing heparinase II and heparinase III from Flavobac drate residues. Heparinase I cleaves at O-D-GlcNp2S6S terium heparinum. (1->4)o-L-IdoAp2S, heparinase III at ot-D-GlcNp2Ac The heparin and heparan sulfate family of molecules is 10 (or2S)6OHC1-)4)8-D-GlcAp and heparinase II at either comprised of glycosaminoglycans of repeating glucosamine linkage as described by Desai et al., Arch. Biochem. Biophys. and hexuronic acid residues, eitheriduronic or glucuronic, in 306(2): 461-468 (1993). Secondary cleavage sites for each which the 2, 3 or 6 position of glucosamine or the 2 position enzyme also have been described by Desai et al. of the hexuronic acid may be sulfated. Variations in the Heparinase I has been used clinically to neutralize the extent and location of sulfation as well as conformation of 15 anticoagulant properties of heparinas summarized by Baugh the alternating hexuronic acid residue leads to a high degree and Zimmermann, Perfusion Rev. 1(2): 8-13, 1993. Hepa of heterogeneity of the molecules within this family. rinase I and III have been shown to modulate cell-growth Conventionally, heparin refers to molecules which possess a factor interactions as demonstrated by Bashkin et al., J. Cell high sulfate content, 2.6 sulfates per disaccharide, and a Physiol. 151:126-137 (1992) and cell-lipoprotein interac higher amount of iduronic acid. Conversely, heparan sulfate tions as demonstrated by Chappell et al., J. Biol. Chem. contains lower amounts of sulfate, 0.7 to 1.3 sulfates per 268(19):14168-14175 (1993). The availability of heparin disaccharide, and less iduronic acid. However, variants of degrading enzymes of sufficient purity and quantity could intermediate composition exist and heparins from all bio lead to the development of important diagnostic and thera logical sources have not yet been characterized. peutic formulations. Specific sulfation/glycosylation patterns of heparin have 25 SUMMARY OF THE INVENTON been associated with biological function, such as the anti Prior to the present invention, partially purified hepari thrombin binding site described by Choay et al., Thrombosis nases II and III were available, but their amino acid Res. 18: 573-578 (1980), and the fibroblast growth factor sequences were unknown. Cloning these enzymes was dif binding site described by Turnbullet al., J. Biol. Chem. 267: 30 ficult because of toxicity to the host cells. The present 10337-10341 (1992). It is apparent from these examples inventors were able to clone the genes for heparinases II and that heparin's interaction with certainmolecules results from III, and herein provide their nucleotide and amino acid the conformation imparted by specific sequences and not sequences. solely due to electrostatic interactions imparted by its high Amethod is described for the isolation of highly purified sulfate composition. Heparin interacts with a variety of 35 heparin and heparan sulfate degrading enzymes from F. mammalian molecules, thereby modulating several biologi heparinum. Characterization of each protein demonstrated cal events such as hemostasis, cell proliferation, migration that heparinases I, II and III are glycoproteins. All three and adhesion as summarized by Kjellen and Lindahl, Ann proteins are modified at their N-terminal amino acid residue. Rey Biochem 60:443-475 (1991) and Burgess and Macaig, Antibodies generated by injecting purified heparinases into Ann. Rey. Biochen. 58: 575-606 (1989). Heparin, extracted rabbits yielded anti-sera which demonstrated a high degree from bovine lungs and porcine intestines, has been used as of cross reactivity to proteins from F. heparinum. Polyclonal an anticoagulant since its antithrombotic properties were antibodies were separated by affinity chromatography into discovered by McLean, Am. J. Physiol. 41:250–257 (1916). fractions which bind the amino acid portion of the proteins Heparin and chemically modified heparins are continually and a fraction which binds the post-translational modifica under review formedical applications in the areas of wound 45 tion allowing for the use of these antibodies to specifically healing and treating vascular disease. distinguish the native and recombinant forms of each hepa Heparin degrading enzymes, referred to as heparinases or rinase protein. heparin lyases, have been identified in several microorgan Amino acid sequence information was used to synthesize isms including: Flavobacterium heparinum, Bacteriodes sp. oligonucleotides that were subsequently used in a poly and Aspergillus nidulans as summarized by Linhardt et al., 50 merase chain reaction (PCR) to amplify a portion of the Appl. Biochem. Biotechnol. 12: 135-177 (1986). Heparan heparinase II and heparinase III genes. Amplified regions sulfate degrading enzymes, referred to as heparitinases or were used in an attempt to identify clones from a DASH-II heparan sulfate lyases, have been detected in platelets gene library which contained F. heparinum genomic DNA. (Oldberg et al., Biochemistry 19:5755-5762 (1980)), tumor Natural selection against clones containing the entire hepa (Nakajima et al., J. Biol. Chem. 259:2283-2290 (1984)) and 55 rinase II and III genes was observed. This was circumvented endothelial cells (Gaal et al., Biochen. Biophys. Res. Comm. by cloning sections of the heparinase I gene separately, and 161: 604-614 (1989)). Mammalian heparanases catalyze the by screening host strains for stable maintenance of complete hydrolysis of the carbohydrate backbone of heparan sulfate heparinase III clones. Expression of heparinase II and III at the hexuronic acid (1->4) glucosamine linkage (Nakajima was achieved by use of a vector containing a modified et al., J. Cell. Biochem. 36: 157-167 (1988)) and are ribosome binding site which was shown to increase the inhibited by the highly sulfated heparin. However, accurate expression of heparinase I to significant levels. biochemical characterizations of these enzymes has thus far This patent describes the gene and amino acid sequences been prevented by the lack of a method to obtain homoge for heparinase II and III from F. heparinum, which may be neous preparations of the molecules. used in conjunction with suitable expression systems to Flavobacterium heparinum produces heparin and heparan 65 produce the enzymes. Also described, is a modified ribo sulfate degrading enzymes termed heparinase I (E.C. some binding sequence used to express heparinase I, II, and 4.2.2.7) as described by Yang et al., J. Biol. Chem. 260(3): . 5,681,733 3 4 BRIEF DESCRIPTION OF THE DRAWINGS nations thereof. Such gene sequences may comprise FIG. 1 shows the modifications to the tac promoter genomic DNA which may or may not include naturally ribosome binding region, which were evaluated for the level occurring introns. moreover, such genomic DNA may be of expression of heparinase I. The original sequence, as obtained in association with promoter regions or poly A found in pBhep, and the modified sequences, as found in sequences. The gene sequences, genomic DNA or cDNA pGhep and pa4hep, are shown with the Shine-Dalgarno may be obtained in any of several ways. Genomic DNA can sequences (S-D) and the heparinase I gene start codon, be extracted and purified from suitable cells, such as brain underlined. The gap (in nucleotides, nt) between these cells, by means well known in the art. Alternatively, mRNA regions is indicated below each sequence. The ribosome can be isolated from a cell and used to produce cDNA by binding region for pGB contains no start codon, and has a 10 reverse transcription or other means. BamHI site (underlined) in place of the EcoRI site Recombinant DNA (GAATTC) found in pChep. By the term "recombinant DNA” is meant amolecule that FIG. 2 shows the construction of plasmids used to has been recombined by in vitro splicing cDNA or a sequence the heparinase I gene from Flavobacterium hep genomic DNA sequence. arinum. Restriction sites are: N- Not, Nc-Nicol, Sa-Sal, 15 Cloning Vehicle B=BamHI, P=PstI, E=EcoRI, H=HindIII, C=Clai and A plasmid or phage DNA or other DNA sequence which K=Kpn. is able to replicate in a host cell. The cloning vehicle is FIG. 3 shows the construction of pCBH2, a plasmid characterized by one or more endonuclease recognition sites capable of directing the expression of active heparinase II in at which is DNA sequences may be cut in a determinable fashion without loss of an essential biological function of the E. coli from tandem tac promoters (double arrow heads). DNA, which may contain a marker suitable for use in the Restriction sites are: B-BamHI, P-Pst I. identification of transformed cells. Markers include for FIG. 4 shows the nucleic acid sequence for the heparinase example, tetracycline resistance or amplicillin resistance. II gene from Flavobacterium heparinum (SEQUID NO:1). The word vector can be used to connote a cloning vehicle. FIG. 5 shows the amino acid sequence for heparinase II 25 Expression Control Sequence from Flavobacterium heparinum (SEQUID NO:2). The A sequence of nucleotides that controls or regulates leader peptide sequence is underlined. The mature protein expression of structural genes when operably linked to those starts at Q-26. Peptides 2A, 2B and 2C are indicated at their genes. They include the lac systems, the trp system major corresponding positions within the protein. operator and promoter regions of the phage lambda, the FIG. 6 shows the construction of plasmids used to 30 control region offd coat protein and other sequences known sequence the heparinase III gene from Flavobacterium hep to control the expression of genes in prokaryotic or eukary arinum. Restriction sites are: S-Sal, B=BamHI, P-Pst, otic cells. E=EcoRI, H=HindIII, C=Clal and K=Kpn. Expression vehicle FIG. 7 shows the construction of pCBH3, a plasmid A vehicle or vector similar to a cloning vehicle but which capable of directing the expression of active heparinase III 35 is capable of expressing a gene which has been cloned into in E. coli from a tandem taq promoter (double arrow heads). it, after transformation into a host. The cloned gene is Restriction sites are: S=Sal, B=BamHI, PsPst, E=EcoRI, usually placed under the control of (i.e., operable linked to) H=HindIII, Bs=BspEI, C=Clai and K=Kpn. certain control sequences such as promoter sequences. FIG. 8 shows the nucleic acid sequence for the heparinase Expression control sequences will vary depending on III gene from Flavobacterium heparinum (SEQUID NO:3). whether the vectoris designed to express the operably linked FIG. 9 shows the amino acid sequence for heparinase III gene in a prokaryotic or eukaryotic host and may addition from Flavobacterium heparinum (SEQUID NO:4). The ally contain transcriptional elements such as enhancer leader peptide sequence is underlined. The mature protein elements, termination sequences, tissue-specificity starts at Q-25. Peptides 3A, 3B and 3C are indicated at their elements, and/or translational initiation and termination corresponding positions within the protein. 45 sites. DEALED DESCRIPTION OF THE Promoter NVENTION The term "promoter" is intended to refer to a DNA To aid in the understanding of the specification and sequence which can be recognized by an RNA polymerase. claims, including the scope to be given such terms, the SO The presence of such a sequence permits the RNA poly following definitions are provided. merase to bind and initiate transcription of operably linked Gene gene sequences. By the term "gene” is intended a DNA sequence which Promoter region encodes through its template or messenger RNA a sequence The term "promoter region” is intended to broadly include of amino acids characteristic of a specific peptide. Further, 55 both the promoter sequence as well as gene sequences which the termincludes intervening, non-coding regions, as well as may be necessary for the initiation of transcription. The regulatory regions, and can include 5' and 3' ends. presence of a promoter region is, therefore, sufficient to Gene sequence cause the expression of an operably linked gene sequence. The term "gene sequence” is intended to refer generally to Operably Linked a DNA molecule which contains one or more genes, or gene As used herein, the term "operably linked” means that the fragments, as well as a DNA molecule which contains a promoter controls the initiation of expression of the gene. A non-transcribed or non-translated sequence. The term is promoter is operably linked to a sequence of proximal DNA further intended to include any combination of gene(s), gene if upon introduction into a host cell the promoter determines fragments(s), non-transcribed sequence(s) or non-translated the transcription of the proximal DNA sequence or sequence(s) which are present on the same DNA molecule. 65 sequences into one or more species of RNA. A promoter is The present sequences may be derived from a variety of operably linked to a DNA sequence if the promoter is sources including DNA, synthetic DNA, RNA, or combi capable if initiating transcription of that DNA sequence. 5,681,733 5 6 Prokaryote ATG start codon, and a recombinant nucleotide sequence The term "prokaryote” is meant to include all organisms encoding. Also provided are modifications to this vector without a true nucleus, including bacteria. comprising changing the length and sequence of the Shine Host Dalgarno sequence, and also by reducing the spacing The term "host" is meant to include not only prokaryotes, between the Shine-Dalgarno sequence and the start codonto but also such eukaryotes as yeast and filamentous fungi, as 8,7,6, 5, 4 or fewer nucleotides. Methods of expressing the well as plant and animal cells. The terms includes organisms heparinases using these novel expression vectors comprise a or cell that is the recipient of a replicable expression vehicle. preferred embodiment of the invention. The present invention is based on the cloning and expres Expression of the heparinases in differenthosts may result sion of two previously uncloned enzymes. Although hepa 10 in different post-translational modifications which may alter rinases II and II had been partially purified previously, no the properties of the heparinases, or a functional derivative amino acid sequences were available. Specifically, the thereof, in eukaryotic cells, and especially mammalian, invention discloses the cloning, sequencing and expression insect and yeast cells, Especially preferred eukaryotic hosts of heparinases II and III from Flavobacterium heparinum are mammalian cells either in vivo, in animals or in tissue and the use of a modified ribosome binding region for 5 culture. Mammalian cells provide post-translational modi expression of these genes. In addition to the nucleotide fications to recombinant heparinases which include folding sequences, the amino acid sequences of heparinases II and and/or glycosylation at sites similar or identical to that found II are also provided. The invention further provides for the native heparinases. Most preferably, mammalian host expressed heparinases I, II and III, as well as methods of cells include brain and neuroblastoma cells. expressing those enzymes. A nucleic acid molecule, such as DNA, is said to be Cloning was accomplished using degenerate and "guess "capable of expressing a polypeptide if it contains expres mer” nucleotide primers derived from amino acid sequences sion control sequences which contain transcriptional regu of fragments of the heparinases, purified as described below latory information and such sequences are "operably linked' in detail. The amino acid sequences were previously to the nucleotide sequence which encodes the polypeptide. unavailable. Cloning was exceptionally difficult because of 25 An operable linkage is a linkage in which a sequence is the unexpected problem of F. heparinum DNA toxicity in E. connected to a regulatory sequence (or sequences) in such a coli. The inventors discovered techniques for solving this way as to place expression of the sequence under the problem, as described below in detail. Based on this influence or control of the regulatory sequence. Two DNA disclosure, one skilled in the art can readily clone additional sequences (such as a heparinases encoding sequence and a heparinases and other proteins from F. heparinum or from promoter region sequence linked to the 5' end of the encod additional sources using the novel methods described ing sequence) are said to be operably linked if induction of promoter function results in the transcription of the hepari Expression of the heparinases is a further disclosure of the nases encoding sequence mRNA and if the nature of the present invention. To express heparinases I, II and III, linkage between the two DNA sequences does not (1) result transcriptional and translational signals recognizable by an 35 in the introduction of a frame-shift mutation, (2) interfere appropriate host are necessary. The cloned heparinases with the ability of the expression regulatory sequences to encoding sequences, obtained through the methods direct the expression of the heparinases, or (3) interfere with described above, and preferably in a double-stranded form, the ability of the heparinases template to be transcribed by may be operably linked to sequences controlling transcrip the promoter region sequence. Thus, a promoter region tional expression in an expression vector, and introduced would be operably linked to a DNA sequence if the promoter into a host cell, either prokaryote or eukaryote, to produce were capable of effecting transcription of that DNA recombinant heparinases or a functional derivative thereof. sequence. Depending upon which strand of the heparinases encoding The precise nature of the regulatory regions needed for sequence is operably linked to the sequences controlling gene expression may vary between species or cell types, but transcriptional expression, it is also possible to express 45 in general includes, as necessary, 5' non-transcribing and 5' heparinases antisense RNA or a functional derivative non-translating (non-coding) sequences involved with ini thereof. tiation of transcription and translation respectively, such as For the expression of heparinases I, II and III in E. coli, the TATA box, capping sequence, CAAT sequence, and the vectors were constructed wherein expression was driven by like. Especially, such 5' non-transcribing control sequences two repeats of the tac promoter. Modifications of the ribo will include a region which contains a promoter for tran some binding region of this promoter were made by intro scriptional control of the operably linked gene. ducing mutations with the polymerase chain reaction. In a If desired, a fusion product of the heparinases may be preferred modification of the expression vector, the minimal constructed. For example, the sequence coding for hepari consensus Shine-Delgarno sequence was improved by intro nases may be linked to a signal sequence which will allow ducing a single mutation (AGGAA-AGGAG), which had 55 secretion of the protein from, or the compartmentalization of the further advantage of decreasing the number of nucle the protein in, a particular host. Such signal sequences otides between the Shine-Delgarno sequence and the ATG maybe designed with or without specific protease sites such start codon. Furthermodifications were produced using PCR that the signal peptide sequence is amenable to subsequent in which the gap between the Shine-Delgarno sequence and removal. Alternatively, the native signal sequence for this the start codon were further reduced. Using the same protein may be used. techniques, additional modifications in this region, including Transcriptional initiation regulatory signals can be insertions and deletions, can be produced to create addi selected which allow for repression or activation, so that tional heparinase expression vectors. As a result, an expres expression of the operably linked genes can be modulated. sion vector for the expression of heparinases is provided Based on this disclosure, one skilled in the art can readily which comprises a modified ribosome binding region con 65 place the sequences of the present invention in additional taining a 5 base pair Shine-Dalgarno sequence, a 9 base pair expression vectors and transforminto a variety of bacteria to spacer region between the Shine-Dalgarno sequence and the obtain recombinant heparinase II or heparinase III. 5,681,733 7 8 Once the vector or DNA sequence containing the the protein's antigenicity. Methods of increasing the antige construct(s) is prepared for expression, the DNA construct nicity of a protein are well-known in the art and include, but (s) is introduced into an appropriate host cell by any if a are not limited to coupling the antigen with a heterologous variety of suitable means, including transfection. After the protein (such as globulin or B-galactosidase) or through the introduction of the vector, recipient cells are grown in a inclusion of an adjuvant during immunization. selective medium, which selects for the growth of vector For monoclonal antibodies, spleen cells from the immu containing cells. Expression of the cloned gene sequence(s) nized animals are removed, fused with myeloma cells, such results in the production of heparinase I, II or III, or in the as SP2/0-Ag14 myeloma cells, and allowed to become production of a fragment of one of these proteins. This monoclonal antibody producing hybridoma cells. expression can take place in a continuous manner in the 10 Any one of a number of methods well known in the art can transformed cells, or in a controlled manner, for example, be used to identify the hybridoma cell which produces an expression which follows induction of differentiation of the antibody with the desired characteristics. These include transformed cells (for example, by administration of bro screening the hybridomas with an ELISA assay, western blot modeoxyuracil to neuroblastoma cells or the like). analysis, or radioimmunoassay (Lutz et al., Exp. Cell Res. The expressed protein is isolated and purified in accor 15 175:109-124 (1988)). dance with conventional conditions, such as extraction, Hybridomas secreting the desired antibodies are cloned precipitation, chromatography, electrophoresis, or the like. and the class and subclass is determined using procedures Detailed procedures for the isolation of the heparinases is known in the art (Campbell, A. M., Monoclonal Antibody discussed in detail in the examples below. Technology: Laboratory Techniques in Biochemistry and The invention further provides functional derivatives of 20 Molecular Biology, Elsevier Science Publishers, the sequences of heparinase II, heparinase III, and the Amsterdam, The Netherlands (1984)). modified ribosome binding site. As used herein, the term For polyclonal antibodies, antibody containing antisera is "functional derivative" is used to define any DNA sequence isolated from the immunized animal and is screened for the which is derived by the original DNA sequence and which presence of antibodies with the desired specificity using one still possesses the biological activities of the native parent 25 of the above-described procedures. molecule. A functional derivative can be an insertion, a The present invention further provides the above deletion, or a substitution of one or more bases in the described antibodies in detectably labelled form. Antibodies original DNA sequence. The substitutions can be such that can be detectably labelled through the use of radioisotopes, they replace a native amino acid with another amino acid affinity labels (such as biotin, avidin, etc.), enzymatic labels that does not substantially effect the functioning of the 30 (such as horseradish peroxidase, alkaline phosphatase, etc.), protein. Those skilled in the art will recognize that likely fluorescent labels (such as FITC or rhodamine, etc.), para substitutions include positively the functioning of the magnetic atoms, chemiluminescent labels, and the like. protein, such as a small, neutrally charged amino acid Procedures for accomplishing such labelling are well-known replacing another small, neutrally charged amino acid. in the art; for example, see Sternberger, L. A. et al., J. Those of skill in the art will recognize that functional 35 Histochem. Cytochem. 18:315 (1970); Byer, E. A. et al., derivatives of the heparinases can be prepared by mutagen Meth. Enzym, 62:308 (1979); Engval, E. et al., Immunol. esis of the DNA using one of the procedures known in the 109:129 (1972); Goding, J. W., J. Immunol. Meth. 13:215 art, such as site-directed mutagenesis. In addition, random (1976). mutagenesis can be conducted and mutants retaining func The present invention further provides the above tion can be obtained through appropriate screening. described antibodies immobilized on a solid support. The antibodies of the present invention include mono Examples of such solid supports include plastics, such as clonal and polyclonal antibodies, as well fragments of these polycarbonate, complex carbohydrates such as agarose and antibodies. Fragments of the antibodies of the present inven sepharose, acrylic resins such as polyacrylamide and latex tion include, but are not limited to, the Fab, the Fab2, and the beads. Techniques for coupling antibodies to such solid Fc fragment. 45 supports are well known in the art (Weir et al., Handbook of The invention also provides hybridomas which are Experimental Immunology, 4th Ed., Blackwell Scientific capable of producing the above-described antibodies. A Publications, Oxford, England (1986)). The immobilized hybridoma is an immortalized cell line which is capable of antibodies of the present invention can be used for immu secreting a specific monoclonal antibody. noaffinity purification of heparinases. In general, techniques for preparing polyclonal and mono 50 Having now generally described the invention, the same clonal antibodies as well as hybridomas capable of produc will be understood by a series of specific examples, which ing the desired antibody are well-known in the art are not intended to be limiting. (Campbell, A. M., "Monoclonal Antibody Technology: Laboratory Techniques in Biochemistry and Molecular EXAMPLE 1. Biology.” Elsevier Science Publishers, Amsterdam, The 55 Netherlands (1984); St. Groth et al., J. immunol. Methods Purification of Heparinases 35:1-21 (1980)). Heparin lyase enzymes were purified from cultures of Any mammal which is known to produce antibodies can Flavobacterium heparinum, F. heparinum was cultured in a be immunized with the pseudogene polypeptide. Methods 15 L computer-controlled fermenter using a variation of the for immunization are well-known in the art. Such methods defined nutrient medium described by Galliher et al., Appl include subcutaneous or interperitoneal injection of the Environ. Microbiol. 41(2):360-365 (1981). Those fermen polypeptide. One skilled in the art will recognize that the tations designed to produce heparin lyases incorporate semi amount of heparinase used for immunization will vary based purified heparin (Celsus Laboratories) in the media at a on the animal which is immunized, the antigenicity of the concentration of 1.0 g/L as the inducer of heparinase syn peptide and the site of injection. 65 thesis. Cells were harvested by centrifugation and the The protein which is used as an immunogen may be desired enzymes released from the periplasmic space by a modified or administered in an adjuvant in order to increase variation of the osmotic shock procedure described by 5,681,733 10 Zimmermann and Cooney, U.S. Pat. No. 5,262,325, herein hydrochloride, were examined as solubilizing agents. Of incorporated by reference. these, only guanidine HCl, at 6M, was able to solubilize the A semi-purified preparation of the heparinase enzymes heparinase 1 inclusion bodies. However, the highest degree was achieved by a modification of the procedure described of purification was obtained by sequentially washing the by Zimmermann et al., U.S. Pat. No. 5,262,325. Proteins inclusion bodies in 3M urea and 6M guanidine HC1. The from the crude osmolate were adsorbed onto cation urea wash step served to removed contaminating E. coli exchange resin (CBX, J.T. Baker) at a conductivity of 1-7 proteins and cell debris prior to solubilizing of the aggre Imho. Unbound proteins from the extract were discarded gated heparinase I by guanidine HCI. and the resin packed into a chromatography column (5.0 cm Recombinant heparinase I was prepared by growing E. i.d.x100 cm). The bound proteins eluted at a linear flow rate coli Y1090CpGHep1), a strain harboring a plasmid contain of 3.75 cmmin' with step gradients of 0.01M phosphate, ing the heparinase I gene expressed from tandem tac 0.01Mphosphate/0.1M sodium chloride, 0.01M phosphated promoters, in Luria broth with 0.1M IPTG. The cells were 0.25M sodium chloride and 0.01M phosphate/1.0M sodium concentrated by centrifugation and resuspended in 1/10th chloride, all at pH 7.0.0.1. Heparinase II elutes in the 0.1M volume buffer containing 0.01M sodium phosphate and 15 0.2M sodium chloride at pH 7.0. The cells were disrupted by NaCl fraction, while heparinases 1 and 3 elute in the 0.25M sonication, 5 minutes with intermittent 30 second cycles, fraction. power setting #3 and the inclusion bodies concentrated by Alternately, the 0.1M sodium chloride step was elimi centrifugation, 7,000xg,5 minutes. The pellets were washed nated and the three heparinases co-eluted with 0.25M two times with cold 3M urea for 2 hours at pH, 7.0 and the sodium chloride. The heparinase fractions were loaded insoluble material recovered by centrifugation. Heparinase I directly onto a column containing cellufine sulfate (5.0 cm was unfolded in 6M guanidine HCl containing 50 mMDTT i.d.X30cm, Amicon) and eluted at a linear flow rate of 2.50 and refolded by dialysis into 0.1M ammonium sulfate. cm-min' with step gradients of 0.01M phosphate, 0.01M Additional contaminating proteins precipitated in the 0.1M phosphate/0,2M sodium chloride, 0.01M phosphate/0.4M ammonium sulfate and could be removed by centrifugation. sodium chloride and 0.01M phosphate/1.0M sodium Heparinase Ipurified by this method had a specific activity chloride, all at pH 7.0.0.1. Heparinase II and 3 elute in the 25 of 42.21 IU/mg and was 90% pure by SDS-PAGE/scanning 0.2M sodium chloride fraction while heparinase I elutes in densitometry analysis. The enzyme can be further purified the 0.4M fraction. by cation exchange chromatography, as described above, The 0.2M sodium chloride fraction from the cellufine yielding a heparinase I preparation that is more than 99% sulfate column was diluted with 0.01M sodium phosphate to pure by SDS-PAGE/scanning densitometry analysis. give a conductance of less than 5 pmhos. The solution was further purified by loading the material onto a hydroxyla EXAMPLE 2. patite column (2.6 cm i.d.X20 cm) and eluting the bound Characterization of Heparinases protein at a linear flow rate of 1.0 cm-min-1 with step gradients of 0.01M phosphate, 0.01M phosphate/0.35M The molecular weight and kinetic properties of the three 35 heparinase enzymes have been accurately reported by Lohse sodium chloride, 0.01M phosphate/0.45M sodium chloride, and Linhardt, J. Biol. Chem. 267:24347–24355 (1992). 0.01M phosphate/0.65M sodium chloride and 0.01M However, an accurate characterization of the proteins' post phosphate/1.0M sodium chloride, all at pH 7.0.0.1. Hepa translational modifications had not been carried out. Hepa rinase IIelutes in a single protein peakin the 0.45M sodium rinases I, II and III, purified as described herein, were chloride fraction while heparinase III elutes in a single analyzed for the presence of carbohydrate moieties. Solu protein peak in the 0.65M sodium chloride fraction. tions containing 2 ug of heparinases I, II and III and Heparinase I was further purified by loading material recombinant heparinase I were brought to pH 5.7 by adding from the cellufine sulfate column, diluted to a conductivity 0.2M sodium acetate. These protein samples underwent less than 5 Imhos, onto a hydroxylapatite column (2.6 cm carbohydrate biotinylation following protocol 2a, described i.d.X20 cm) and eluting the bound protein at a linear flow 45 in the GlycoTrack kit (Oxford Glycosystems). 30 ul of each rate of 1.0 cmmin' with a linear gradient of phosphate biotinylated protein solution was subjected to SDS-PAGE (0.01 to 0.25M) and sodium chloride (0.0 to 0.5M). Hepa (10% gel) and transferred by electroblotting at 170 mA rinase I elutes in a single protein peak approximately mid constant current to a nitrocellulose membrane. Detection of way through the gradient. the biotinylated carbohydrate was accomplished by an alka The heparinase enzymes obtained by this method were 50 line phosphatase-specific color reaction after attachment of analyzed by SDS-PAGE using the technique of Laemmli, a streptavadin-alkaline phosphatase conjugate to the biotin Nature 227; 680–685 (1970), and the gels quantified by a groups. These analyses revealed that heparinases I and Iare scanning densitometer (Bio-Rad, Model GS-670). Hepari glycosylated and heparinase III and recombinant heparinase nases I, II and III displayed molecular weights of 42,500+2, I are not. 000, 84,000-4,200 and 73,000+3,500 Daltons, respectively. 55 Polyclonal antibodies generated in rabbits injected with All proteins displayed purities of greater than 99%. Purifi wild type heparinase I could be fractionated into two popu cation results for the heparinase enzymes are shown in Table lations as described below. It appears that one of these 1. fractions recognizes a post-translational moiety common to Heparinase activities were determined by the spectropho proteins made in F. heparinum, while the other fraction tometric assay described by Yang et al. Amodification of this specifically recognizes amino acid sequences contained in assay incorporating a reaction buffer comprised of 0.018M heparinase I. AI heparinase enzymes made in F. heparinum Tris, 0.044M sodium chloride and 1.5 g/L heparan sulfate at were recognized by the “non-specific' antibodies but not pH 7.5 was used to measure heparan sulfate degrading heparinase made in E. coli. The most likely candidate for the activity. non-protein antigenic determinant from heparinase I is the Recombinant heparinase I forms intracellular inclusion 65 carbohydrate component; thus, the Western blot experiment bodies which require denaturation and protein refolding to indicates that all lyases made in F. heparinum are glycosy obtain active heparinase. Two solvents, urea and guanidine lated. 5,681,733 11 12 Purified heparinases II and III were analyzed by the clonal antibodies in rabbits. Each of heparinase I, II and ITI technique of Edman to determine the N-terminal amino acid was carried through the following standard immunization residue of the mature protein. However, the Edman chem procedure: The primary injection consisted of 0.5-1.0 mg of istry was unable to liberate an amino acid, indicating that a purified protein dissolved in 1 ml of sterile phosphate post-translational modification had occurred at the buffered Saline, which was homogenized with 1 ml of N-terminal amino acid of both heparinases. One nmol Freund's adjuvant (Cedarlane Laboratories Ltd.). This samples of heparinases II and III were used for deblocking protein-adjuvant emulsion was used to inject New Zealand with pyroglutamate aminopeptidase. Control samples were White female rabbits; 1 ml per rabbit, 0.5 ml per rear leg, produced by mock deblocking 1 nmol protein samples i.m., in the thigh muscle near the hip. After 2 to 3 weeks, the without adding pyroglutamate aminopeptidase. All samples 10 rabbits were given an injection boost consisting of 0.5-1.0 were placed in 10 mMNHCO, pH 7.5, and 10 mM DTT mg of purified protein dissolved in sterile phosphate buff (100 ul final volume). To non-control samples, 1 m J of ered Saline homogenized with 1 ml of incomplete Freund's pyroglutamate aminopeptidase was added and all samples adjuvant (Cedarlane Laboratories, Ltd.). Again after 2 to 3 were incubated for 8 hr at 37° C. After incubation, an weeks, the rabbits were given a third identical injection additional 0.5 mJ of pyroglutamate aminopeptidase was 15 boost. added to non-control samples and all samples were incu A blood sample was collected from each animal from the bated for an additional 16 h at 37° C. central artery of the ear approximately 10 days following the Deblocking buffers were exchanged for 35% formic acid final injection boost. Serum was prepared by allowing the using a 10,000 Dalton cut-off Centricon unit and the sample sample to clot for 2 hours at 22°C. followed by overnight was dried under vacuum. The samples were subjected to incubation at 4°C., and clearing by centrifugation at 5,000 amino acid sequence analysis according to the method of rpm for 10 min. The antisera were diluted 1:100,000 in Edman. Tris-buffered Saline (pH 7.5) and carried through Western The properties of the three heparinase proteins from blot analysis to identify those sera containing anti Flavobacterium heparinum are listed in Table 2. heparinase I, II or II antibodies. Heparinases II and III were digested with cyanogen 25 Antibodies generated against wild type heparinase I, but bromide in order to produce peptide fragments for isolation. not recombinant heparinase I, displayed a high degree of The protein solutions (1-10 mg/ml protein concentration) cross reactivity against other F. heparinum proteins. This were brought to a DTT concentration of 0.1M, and incubated was likely due to the presence of an antigenic post at 40° C. for 2 hr. The samples were frozen and lyophilized translational modification common to F. heparinum proteins under vacuum. The pellet was resuspended in 70% formic 30 but not found on proteins synthesized in E. coli. To explore acid, and nitrogen gas was bubbled through the solution to this further, recombinantheparinase I was immobilized onto exclude oxygen. Astock solution of CNBr was made in 70% Sepharose beads and packed into a chromatography column. formic acid and the stock solution was bubbled with nitro Purified anti-heparinase I (wildtype) antibodies were loaded gen gas and stored in the dark for short time periods. For onto the column and the unbound fraction collected. Bound addition of CNBr, a 500 to 1000 times molar excess of CNBr 35 antibodies were eluted in 0.1M glycine, pH 2.0. IgG was to methionine residues in the protein was used. The CNBr found in both the unbound and bound fractions and subse stock was added to the protein solutions, bubbled with quently used in Western blot experiments. Antibody isolated nitrogen gas and the tube was sealed. The reaction tube was from the unbound fraction non-specifically recognized F. incubated at 24° C. for 20 hr, in the dark. heparinum proteins but no longer detected recombinant The samples were dried down partially under vacuum, heparinase I (E. coli), while the antibody isolated from the water was added to the sample, and partial lyophilization bound fraction only recognized heparinase I, whether syn was repeated. This washing procedure was repeated until the thesized in F. heparinum or E. coli.This resultindicated that, as hypothesized, two populations of antibodies are formed sample pellets were white. The peptide mixtures were solu by exposure to the wild-type heparinase I antigen: one bilized in formic acid and applied to a Vydac C. reverse 45 phase HPLC column (4.6 mm i.d.X30 cm) and individual specific for the protein backbone and the other recognizing peptide fragments eluted at a linear flow rate of 6.0 a post-translationally modified moiety common to F. hep cm-min' with a linear gradient of 10 to 90% acetonitrile in arinum proteins. 1% trifluoroacetic acid. Fragments recovered from these This finding provides both a means to purify specific reactions were subjected to amino acid sequence determi 50 anti-heparinase antibodies and a tool for characterizing the nation using an Applied Biosystems 745A Protein wild-type heparinase I protein. Sequencer. Three peptides isolated from heparinase II gave sequences: EFPEMYNLAAGR (SEQUID NO:5), KPAD EXAMPLE 4 IPEVKDGR (SEQU ID NO:6), and LAGDFVT GKILAQGFG PDNQTPDYTYL (SEQU ID NO:7) and 55 Construction of a F heparinum Gene Library were named peptides 2A, 2B and 2C respectively. Three A Flavobacterium heparinum chromosomal DNA library peptides from heparinase III gave sequences: LIK was constructed in lambda phage DASHIL 0.4 ug of F. NEVRWQLHRVK (SEQU ID NO:8), VLKASPPGEF heparinum chromosomal DNA was partially digested with HAQPDNGTFELFI (SEQUID NO:9) and KALVHWFW restriction enzyme Sau3A to produce a majority of frag PHKGYGYFDYGKDIN (SEQU ID NO:10) and were ments around 20 kb in size, as described in Maniatis, et al., named peptides 3A, 3B and 3C, respectively. Molecular Cloning Manual, Cold Spring Harbor (1982). This DNA was phenol/chloroform extracted, ethanol EXAMPLE 3 precipitated, ligated with DASHII arms and packaged with Antibodies to the Heparinase Proteins packaging extracts from a DASHI/BamHI Cloning Kit 65 (Stratagene, La Jolla, Calif.). The library was titered at Heparinases I, II and III and recombinant heparinase I, approximately 10 pfu/ml after packaging, amplified to purified as described herein, were used to generate poly 10 pfu/ml by the plate lysis method, and stored at -70° C. 5,681,733 13 14 as described by Silhavy, T. J., et al. in Experiments in DNA was then digested with BamHI and the single-stranded Molecular Genetics, Cold Spring Harbor Laboratory, 1992. ends were made double-stranded by filling-in with Klenow The F. heparinum chromosomal library was titered to fragment. The blunt-end DNA was ligated and transformed about 300 pfu/plate, overlaid on a lawn of E. coli, and into E. coli strain FTB1. A plasmid which contained a unique allowed to transfect the cells overnight at 37° C., forming 5 BamHIsite and noheparinase Igene DNA was purified from plaques. The phage plaques were transferred to nitrocellu a kanamycin resistant colony and was designated plasmid, lose paper, and the phage DNA bound to the filters, as pGB. DNA sequence analysis revealed that plasmid pGB described in Maniatis, et al., ibid. contained the modifiedribosome binding site, shown in FIG. EXAMPLE 5 10 1. A Modified Ribosome Binding Region for the EXAMPLE 6 Expression of Flavobacterium heparinum Glycosaminoglycan Lyases Nucleic Acid Encoding Heparinase II Four "guessmer” oligonucleotides were designed using The gene for the mature heparinase I protein was cloned 15 into the EcoRI site of the vector, p39, where its expression information from two peptide sequences 2A and 2B and use was driven by two repeats of the tac promoter (from expres of the consensus codons for Flavobacterium, shown in Table sion vector, pKK223-3, Brosius, and Holy, Proc. Natl. Acad. 3. These were: Sci. USA 81: 6929-6933 (1984)). In this vector, p3hep, the O 5'-GAATTCCCTGAGATGTACAATCTGGCCGC- (SEQUID No: 11) first codon, ATG, for heparinase 1 is separated by 10 3', nucleotides from a minimal Shine-Dalgarno sequence AGGA (Shine and Dalgarno, Proc. Natl. Acad. Sci. USA 5'-CCGGCAGCCAGATTGTACATTTCAGG-3', (SEQUID NO: 12) 71:1342-1346 (1974)), FIG. 1. This construct was trans 5'-AAACCCGCCGACATTCCCGAAGTAAA (SEQUID NO: 13) AGA-3', formed into the E. coli strain, JM109, grown at 37° C. and 25 induced with 1 mM IPTG, 2 hours before harvesting. Cells and were lysed by sonication, the cell membrane fraction was SCGAAAGTCTTTTACTTCGGGAATGTC (SEQUID NO: 14) pelleted and the supernatant was saved. The membrane GGC-3', fraction was resuspended in 6M guanidine-HCl in order to named 2-1, 2-2, 2-3 and 2-4, respectively. The oligonucle solubilize inclusion bodies containing the recombinant otides were synthesized with a Bio/CAN (Mississauga, heparinase I enzyme. The soluble heparinase I was refolded Ontario) peptide synthesizer. Pairs of these oligonucleotides by diluting in 20 mM phosphate buffer. The enzyme activity were used as primers in PCR reactions. F. heparinum was determined in the refolded pellet fraction, and in the chromosomal DNA was digested with restriction endonu supernatant fraction. Low levels of activity were detected in 35 cleases Sall, Xbal or Not, and the fragmented DNA com the supernatant and the pellet fractions. Analysis of the bined for use as the template DNA. Polymerase chain fractions by SDS-PAGE indicated that both fractions may reaction mixtures were produced using the DNA Amplifi contain minor bands corresponding to the recombinanthepa cation Reagent Kit (Perkin Elmer Cetus, Norwalk, Conn.). rinase I. The PCR amplifications were carried out in 100 ulreaction volume containing 50 mM KC1, 10 mM Tris HCI, pH 9, In an attempt to increase expression levels from pBhep, 0.1% Triton X-100, 1.5 mM MgCl, 0.2 mM of each of the two mutations were introduced as indicated in FIG. 1. The four deoxyribose nucleotide triphosphates (dNTPs), 100 mutations were produced to improve the level of translation pmol of each primer, 10 ng of fragmented F. heparinum of the heparinase I mRNA by increasing the length of the genomic DNA and 2.5 units of Taq polymerase (Bio/CAN Shine-Dalgarno sequence and by decreasing the distance 45 Scientific Inc., Mississauga, Ontario). The samples were between the Shine-Dalgarno sequence and the ATG-start placed on an automated heating block (DNA thermal cycler, site. Using PCR, a single base mutation converting an A to Barnstead/Thermolyne Corporation, Dubuque, Iowa) pro a Gimproved the Shine-Dalgarno sequence from a minimal grammed for step cycles of: denaturation temperature 92° C. AGGA sequence to AGGAG while decreasing the distance (1 minute), annealing temperatures of 37°C., 42 C. or 45° C. (1 minute) and extension temperature 72 C. (2 minutes). between the Shine-Dalgarno sequence and the translation These cycles were repeated 35 times. The resulting PCR start site from 10 to 9 base pairs. This construct was named products were analyzed on a 1.0% agarose gel containing pGhep. In the second construct, pa4hep, 4 nucleotides 0.6ug/ml ethidium bromide, as described by Maniatis, et al., (AACA) were deleted using PCR, in order to lengthen the ibid. DNA fragments were produced by oligonucleotides Shine-Dalgarno sequence to AGGAG as well as moving it to 55 2-2 and 2-3. The fragments, 250 bp and 350 bp in size, were within 5 base pairs of the ATG-start site. first separated on 1% agarose gel electrophoresis, and the The different constructs were analyzed as described DNA extracted from using the GENECLEAN I kit (Biof above. Refolded pellets from E. coli transformed with CAN Scientific, Mississauga, Ontario). Purified fragments pGhep displayed approximately a 7x increase in heparinase were ligated into pTZ/PC (Tessier and Thomas, unpublished) previously digested with Not, FIG. 2, and the I activity, as compared to refolded pellets from E. coli ligation mixture used to transform E. coli FTB1, as containing pBhep. On the other hand, E. coli containing described in Maniatis et al., ibid. Allrestriction enzymes and pA4hep displayed 2-3 times less activity than the pBhep T4DNA ligase were purchased from New England Biolabs containing E. coli. The levels of heparinase 1 activity in the (Mississauga, Ontario). supernatants were similar. 55 Strain FTB1 was constructed in our laboratory. The F Plasmid, pBhep, was digested with EcoRI and treated episome from the XL-1 Blue E. coli strain (Stratagene, La with S1 nuclease to form blunt-ended DNA. The plasmid Jolla, Calif.), which carries the lac Iq repressor gene and 5,681,733 15 16 produces 10 times more lac repressor than wild type E. coli, a negative-selective effect on any E. coli cells that harbor it. was moved, as described by J. Miller, Experiments in This toxic affect had not been observed previously with Molecular Genetics, Cold Spring Harbor Laboratory (1972), other F. heparinum chromosomal DNA fragments. into the TB1 E. coli strain, described by Baker, T.A., et al., A second strategy employed to circumvent the unexpected Proc. Natl. Acad. Sci. 81:6779-6783 (1984). The FTB1 5 problem of F. heparinum DNA toxicity in E. coli was to background permits a more stringentrepression of transcrip digest the chromosomal DNA fragment with a restriction tion from plasmids carrying promoters with a lac operator endonuclease which would divide the fragment, and if (i.e., lac and Taq promoters). Colonies resulting from the possible the heparinase II, gene into two pieces, FIG. 2. transformation of FTB1 were selected on LB agar contain O These fragments could be cloned individually. DNA ing amplicillin and screened using the blue/white screen sequence analysis of the PCR insert in plasmid, pCE14, provided by X-gal and IPTG included in the agar medium, demonstrated that BamHI and EcoRI sites were present in as described by Maniatis, et al., ibid. Transformants were the insert. Hybridization experiments also demonstrated that analyzed by colony cracking and mini-preparations of DNA the BamHI digested F. heparinum DNA in phage HIS were made for enzyme restriction analysis using the RPM 15 produced two bands 1.8 and 5.5 kb in size. Analysis of kit (Bio/CAN Scientific Inc., Mississauga, Ontario). Ten hybridization data indicated that the 1.8 kb band contains the plasmids contained inserts of the correct size, which were 5' end and the 5.5 kb band contains the 3' end of the gene. released upon digestion with EcoRI and HindIII. Furthermore, a 5kbB.coRIF. heparinum chromosomal DNA DNA sequencing revealed that one of the plasmids, O fragment hybridized with the PCR probe. The 1.8, 5, and 5.5 pCE14, contained a 350 bp PCR fragment had the expected kb fragments containing heparinase II gene sequences were DNA sequence as derived from peptide 2C. DNA sequences inserted into pBluescript, as described above. Two clones, were determined by the dideoxy-chain termination method pBSIB6-7 and p3SIB6-21, containing the 5.5 kb BamHI of Sanger et al., Proc. Natl. Acad. Sci. 74:5463-5467 (1978). insert in different orientations were isolated and one Sequencing reactions were carried out with the Sequenase 25 plasmid, p8SB213, was isolated which contained the 1.8 kb Kit (U.S. Biochemical Corp., Cleveland, Ohio) and S BamHI fragment. No clones containing the 5 kb EcoRI. dATP (Amersham Canada Ltd., Oakville, Ontario, Canada), fragment were isolated, even though extensive screening of as specified by the supplier. possible clones was done. The heparinase II gene was cloned from a F. heparinum 30 The molecular weight of heparinase II protein is approxi chromosomal DNA library, FIG. 2, constructed as described mately 84 kD, so the size of the corresponding gene would above. Ten plaque-containing filters were hybridized with be approximately 2.4 kb. The 1.8 and 5.5 kb BamHI the DNA probe, produced from the gel purified insert of chromosomal DNA fragments could include the entire hepa pCE14, which was labeled using a Random Labeling Kit rinase II gene. The plasmids plbSIB6-7, p3SIB6-21 and (Boehringer Mannheim Canada, Laval, Quebec). Plaque 35 pBSIB2-13, FIG. 2, were used to produce nested deletions hybridization was carried out, as described in Maniatis et al., with the Erase-a-Base system (Promega Biotec, Madison ibid, at 65° C. for 16 hours in a Tek Star hybridization oven Wis.). These plasmids were used as templates for DNA (Bio/CAN Scientific, Mississauga, Ontario). Subsequent sequence analysis using universal and reverse primers and washes were performed at 65° C.: twice for 15 min. in 2x oligonucleotide primers derived from known heparinase II SSC, once in 2XSSC/0.1% SDS for 30 min, and once in 0.5x sequence. Because parts of the gene were relatively G-Crich SSC/0.1% SDS for 15 min. Positive plaques were harvested and contained numerous strong, secondary structures, the using plastic micropipette tips and confirmed by dot blot sequence analysis was, at times, performed using reactions analysis, as described by Maniatis et al., ibid. Six of the in which the dGTP was replaced by dIP. Analysis of the phages, which gave strong hybridization signals, were used 45 DNA sequence, FIG. 4, indicated that there was a single, for Southern hybridization analysis, as described by continuous open reading frame containing codons for 772 Southern, E.M., J. Mol. Biol. 98:503–517 (1975). This amino acid residues, FIG.S. Searching for a possible signal analysis showed that one phage, HIS, contained a 5.5 kb peptide sequence using Geneworks (Intelligenetics, Moun Xbal DNA fragment which hybridized with the probe. tain View, Calif) suggested that there are two possible sites Cloning the 5.5 kb Xbal fragment into the Xbal site of any for processing of the protein into a mature form: Q-26 of following vectors: pIZIPC, pBluescript (Stratagene, La (glutamine) and D-30 (aspartate). N-terminal amino acid Jolla Calif.), p C18 (described in Yanisch-Perron et al., sequencing of deblocked, processed heparinase II indicated Gene 33:103-119 (1985)), and p0K12 (described in Vierra that the mature protein begins with Q-26, and contains 747 andMessing, Gene 100:189-194 (1991)), was unsuccessful, 55 amino acids with a calculated molecular weight of 84.545 even though the FTB1 background was used to repress Daltons, FIG. 5. plasmid promoter-derived transcription. Vector, pCK12, a low copy number plasmid derived from pACYC184 EXAMPLE 7 (approximately 10 copies/cell, Chang, A. C. Y. and Cohen, S. N., J. Bact. 134:1141-1156 (1978) was used in an Expression of Heparinase II in E. coli attempt to circumvent the toxic effects of a foreign DNA The vector, pGB, was used for heparinase IIexpressionin fragment in E. coli by minimizing the number of copies of E. coli, FIG. 3. pGB contains the modifiedribosome binding the toxic foreign fragment. In addition, insertion of the entire region from p(hep, FIG. 1, and a unique BamHI site, Not chromosomal DNA insert of the HIS phage into 65 whereby expression of a DNA fragment inserted into this plasmid pCK12 plasmid, was unsuccessful. It was con site is driven by a double tac promoter. The vector also cluded that this region of F. heparinum chromosome imparts includes a kanamycin resistance gene, and the lacIq gene to 5,681,733 17 18 allow induction of transcription with IPTG. Initially, a gel the heparinase III gene, using oligonucleotides synthesized purified 5.5 kb BamHI fragment from psSIB6-21 was on the basis of amino acid sequence information, required ligated with BamHI digested pCB and transformed into eliminating some of the DNA sequence possibilities. An FTB1, which was selected on LB agar with kanamycin. Six assumed codon usage was calculated based on known DNA of the resulting colonies contained plasmids with inserts in sequences for genes from other Flavobacterium species. the correct orientation for expression of the open reading Sequences for 17 genes were analyzed and a codon usage frame. Pst digestion and religation of one of the plasmids, table was compiled, Table 3. forming pCBIID, deleted 3.5 kb of the 5.5 kb BamHI fragment and removed a BamHI site leaving only one O Four oligonucleotides were designed by choosing each BamHI site directly after the Shine-Dalgarno sequence. codon according to the codon usage table. These were: Finally, two synthetic oligonucleotides were designed: 5'-GAATTCCATCAGTTTCAG CCGCATAAA-3'(SEQU 5'-TGAGGATTCATG CAAACCAAGGCCGATGT ID NO:17).5'-GAATTCTTTATGCGGCTGAAACTGATG GGTTTGGAA-3' (SEQU ID NO: 15), and 3' (SEQU ID NO:18),5'- 5'-GGAGGATAACCACATTCGAGCATT-3' (SEQU ID 15 GAATTCCCGCCGGGCGAATTTCATGC-3'(SEQUID NO:16) for use in a PCR to produce a fragment containing NO:19) and 5'-GAATTCGCATGAAATTCGCCCGGCGG aBamHIsite and an ATG start codon upstream of the mature 3'(SEQUID NO:20), and were named oligonucleotides 3-1, protein encoding sequence and a downstream BamHIsite, 3-2, 3-3 and 3-4, respectively. These oligonucleotides were FIG. 3. Lambda clone HII-I, isolated at the same time as 20 used in all possible combinations, in an attempt to amplify lambda clone HIS, was used as template DNA. a portion of the heparinase III gene using the polymerase Cloning the blunt-end PCR product into pIZ/PC was chain reaction. The PCR amplifications were carried out as unsuccessful, using FTB1 as the host. Cloning the BamHI described above. Cycles of: denaturation temperature 92°C. digested PCR product into the BamHI site of pBluescript, (1 minute), annealing temperatures ranging from 37° to 55 again using FTB1 as the host, resulted in the isolation of 2 25 C., (1 minute) and extension temperature 72°C. (2 minutes) plasmids containing the PCR fragment, after screening of were repeated 35 times. Analysis of the PCR reactions as 150 possible clones. One of these, pFBSQTK-9, which was described above demonstrated that no DNA fragments were sequenced with reverse and universal primers, contained an produced by these experiments. accurate reproduction of the DNA sequence from the hepa 30 A second set of oligonucleotides was synthesized and was rinase II gene. The BamHI digested PCR fragment from comprised of 32 base sequences, in which the codon usage pBSQTK-9 was inserted into the BamHI site of pCBIID in table was used to guess the third position of only half of the such orientation that the ATG site was downstream of the codons. The nucleotides within the parentheses indicate Shine-Dalgarno sequence. This construct, pCBH2, placed degeneracies of two or four bases at a single site. These the mature heparinase II gene under control of the tac Wee:

5-GG(ACGT)GAATTTCCATGCCCAGCC(ACGT)GA(CTAATGG(ACGT)AC-3', (SEQUID NO: 21) 5-GTCACGTCCATT(AG)TC(ACGTGGCTGGGCATGAAATTC(ACGT)CC-3', (SEQUID NO: 22) 5-GTCACGTCATCAGTTCCT)CAGCC(ACGTCATAAAGG(ACGT)TATGG-3, (SEQUID NO: 23) and 5'-CCCATA(ACGTCCTTTATG(ACGTyGGCTG(AGAACTGATG(ACGT)AC-3', (SEQUID NO: 24) promoters in pCB, FIG. 3. Strain E. coli FTB1(pGBH2) was and were named oligonucleotides 3-5, 3-6, 3-7 and 3-8. grown in LB medium containing 50 ug/mlkanamycin at 37 respectively. These oligonucleotides were used in an attempt C. for 3 h. Induction of the tac promoter was achieved by to amplify a portion of the heparinase III gene using the polymerase chain reaction, and the combination of 3-6 and adding 1 mmol PTG and the culture placed at either room 50 3-7 gave rise to a specific 983 bp PCR product. An attempt temperature or 30° C. Heparin and heparan sulfate degrad was made to clone this fragment by blunt end ligation into ing activity was measured in the cultures after growth for 4 E. coli vector, pBluescript, as well as two specifically hours using the method described by Yang et al., ibid. designed vectors for the cloning of PCR products, pIZIPC Heparin degrading activities of 0.36 and 0.24 IU/mg protein and pCRII from the TA cloning ' kit (InVitrogen and heparan sulfate degrading activities of 0.49 and 0.44 55 Corporation, San Diego, Calif.). All of these constructs were IU/mg protein were observed at room temperature and 30° transformed into the FTB1 E. Coli strain. Transformants C., respectively. were first analyzed by colony cracking, and subsequently mini-preparations of DNA were made for enzyme restriction EXAMPLE 8 analysis. No clones containing this PCR fragment were isolated. Nucleic Acid Encoding Heparinase III A third set of oligonucleotides was synthesized incorpo The amino acid sequence information obtained from rating BamHendonuclease sequences on the ends of the 3-6 peptides derived from heparinase III, FIG. 9, purified as and 3-7 oligonucleotide sequences. A 999 base pair DNA described herein, reverse translated into highly degenerate 65 sequence was obtained using the polymerase chain reaction oligonucleotides. Therefore, a cloning strategy relying on with F. heparinum chromosomal DNA as the target. the polymerase chain reaction amplification of a section of Attempts were made to clone the amplified DNA into the 5,681,733 19 20 BamH1 site of the high copy number plasmid pbluescript tem (Promega Corporation, Madison, Wis.) or sequenced and the low copy number plasmids pBR322 and using synthetic oligonucleotide primers. pACYC184. All of these constructs were again transformed Sequence analysis revealed a single continuous open into the FTB1 E. coli strain. More than 500 candidates were reading frame, without a translational termination codon, of screened, yet no transformants containing a plasmid harbor 1929 base pairs, corresponding to 643 amino acids. Further ing the F. heparinum DNA were obtained. Once again, it was screening of the lambda library led to the identification of a concluded that this region of F. heparinum chromosome 673 bp Kpn fragment which was similarly cloned into the imparts a negative-selective effect on E. coli cells that harbor KpnI site of peluescript, creating plasmid prB4. The ter it. 10 mination codon was found within the KpnIfragment adding As in the case for isolation of the heparinase II gene, the an extra 51 base pairs to the heparinase III gene and an PCR fragment was split in order to avoid the problem of additional 16 amino acid to the heparinase III protein. The foreign DNA toxicity. Digestion of the 981 bp BamHI complete heparinase III gene was later found to be included digested heparinase IIIPCR fragment with restriction endo within a 3.2 kilobase Pst fragment from lambda plaque nuclease ClaI produced two fragments of 394 and 587 bp. 15 #118. The complete heparinase III gene from Flavobacte The amplified F. heparinum region was treated with Claland rium is thus 1980 base pairs in length, FIG. 8, and encodes the two fragments separated by agarose gel electrophoresis. a 659 amino acid protein, FIG. 9. N-terminal amino acid The 587 and 394 base pairfragments were ligated separately sequencing of deblocked, processed heparinase III indicated into plasmid p3luescript that had been treated with restric that the mature protein begins with Q-25, and contains 635 tion endonucleases BamHI and Cla. In addition, the entire amino acids with a calculated molecular weight of 73,135 981 bp PCR fragment was purified and ligated into BamHI Daltons, FIG. 9. cut pBluescript. The ligated plasmids were inserted into the XL-1 Blue E. coli. Transformants containing plasmids with EXAMPLE 9 inserts were selected on the basis of their ability to form 25 white colonies on LB-agar plates containing X-gal, IPTG Expression of Heparinase III in E. coli and 50 ug?ml ampicillin, as described by Maniatis. Plasmid PCR was used to generate a mature, truncated heparinase pFB1 containing the 587 bp F heparinum DNA fragment III gene, which had 16 amino acids deleted from the and plasmid pFB2 containing the entire 981 base pair 30 carboxy-terminus of the protein. An oligonucleotide com fragment were isolated by this method. The XL-1 Blue prised of 5'-CGCGGATCCATGCAAAGCT CTTCCATT-3' strain, which, like strain FTB1, contains the lacq repressor (SEQUID NO:25) was designed to insert an ATG start site gene on an F episome, allowed for stable maintenance of the immediately preceding the codon for the first amino acid complete BamHIPCR fragment, unlike FTB1. The reason (Q-25) of mature heparinase III, while an oligonucleotide for this discrepancy is not apparent from the genotypes of 35 comprised of 5'- CGCGG ATCCT CA the two strains (i.e., both are rec A, etc.). AAGCTTGCCTTTCTC-3' (SEQU ID NO:26), was DNA sequence analysis of the F heparinum DNA in designed to insert a termination codon after the last amino plasmid plb1 showed that it contained a sequence encoding acid of the heparinase III gene on the 2.7 kb HindIII peptide Hep3-B while the F heparinum insert in plasmid fragment. Both oligonucleotides also contained a BamHI pFB2 contained a DNA sequence encoding peptides Hep3-D site. Plasmid pHindIIIBD was used as the template in a PCR and Hep3-B, FIG. 9. This analysis confirmed that these reaction with an annealing temperature of 50° C. A specific inserts were part of the gene encoding heparinase III. fragment of the expected size, 1857 base pairs, was The PCR fragment insert in plasmid pFB1 was labeled obtained. This fragment encodes a protein of 620 amino with P-ATP using a Random Primed DNA Labeling kit 45 acids with a calculated MW of 71.535 Da. It was isolated (Boehringer Mannheim, Laval, Quebec), and was used to and inserted in the BamHI site of the expression vector pGB. screen the F. heparinum ADASHII library, FIG. 6, con This construct was named pCB-H3A3', FIG. 7. structed as described herein. The lambda library was plated To add the missing 3' region of heparinase III, the out to obtain approximately 1500 plaques, which were BspEI/Sal I restriction fragment from pCB-H3A3' was transferred to nitrocellulose filters (Schleicher & Schuel, 50 removed and replaced with the BspEI/SalI fragment from Keene, N.H.). The PCR probe was purified by ethanol pFB5. The construct containing the complete heparinase III precipitation. Plaque hybridization was carried out using the gene was named pCBH3, FIG. 7. Recombinant heparinase conditions described above. Eight positive lambda plaques III is a protein of 637 amino acids with a calculated were identified. Lambda DNA was isolated from lysed 55 molecular weight of 73,266 Daltons. E. coli strain XL-1 bacterial cultures as described in Maniatis and further ana Blue(pGBH3) was grown at 37°C. in LB medium contain lyzed by restriction analysis and by Southern blotting using ing 75 lug/ml kanamycin to an ODoo of 0.5, at which point a Hybond-N nylon membrane (Amersham Corporation, the tac promoter from pCB was induced by the addition of Arlington Heights, Ill.) following the protocol described in 1 mM PTG. Cultures were grown an additional 2-5 hours Maniatis. A 2.7 kilobase HindIII fragment from lambda at either 23° C., 30° C. or 37° C. The cells were cooled on plaque #3, which strongly hybridized to the PCR probe, was ice, concentrated by centrifugation and resuspended in cold isolated and cloned in pbluescript, in the XL-1 Blue E. coli PBS at Vioth the original culture volume. Cells were lysed by background, to yield plasmid pHindIIIBD, FIG. 6. This sonication and cell debris removed by centrifugation at clone was further analyzed by DNA sequencing. The 65 10,000Xg for 5 minutes. The pellet and supernatantfractions sequence data was obtained using successive nested dele were analyzed for heparan sulfate degrading (heparinase III) tions of pHindIIIBD generated with the Erase-a-Base Sys activity. Heparan sulfate degrading activities of 1.29, 5.27 5,681,733 21 22 and 3.29 IU/ml were observed from cultures grown at 23°, 30° and 37° C., respectively. TABLE 2-continued The present invention describes a methodology for obtaining highly purified heparin and heparan sulfate Propertied of heparinases from degrading proteins by expressing the genes for these pro Flavobacterium heparinum teins in a suitable expression system and applying the steps of cell disruption, cation exchange chromatography, affinity sample heparinase I heparinase II heparinase chromatography and hydroxylapatite chromatography. 10 Variations of these methods will be obvious to those skilled N-terminal peptide QQKKSG QTKAIDV QSSSIT in the art from the foregoing detailed description of the glycosylation yes yes maybe invention. Such modifications are intended to come within the scope of the appended claims. H-heparin, HS-heparan sulfate 15 TABLE 1. TABLE 3 Purification of heparinase enzymes from Flavobacterium hepatinum fermentations Codon usage table for Flavobacterium and Escherichia coli activity specific activity yield 20 sample (IU) (IU/mg) (%) consensus codon fermentatio crimentation amino acid codion(s) E. coli Flavobacterium heparin degrading 39,700 106 OO heparan sulfate degrading 75,400 ND 100 25 A. GCT, GCC, GCG, GCA GCT GCC osmolate C TGI, TGC ETHER ETHER heparin degrading 15,749 ND 40 D GAT, GAC ETHER ETHER heparan sulfate degrading 42,000 ND 56 E GAG, GAA GAA GAA cation exchange F TTC, TTT ETHER IT heparin degrading 12,757 ND 32 30 G GGC, GGA, GGG, GGT GGC or GGT GGC heparan sulfate degrading 27,540 ND 37 H CAC, CAT CAT CAT cellufine sulfate ATC, ATA, ATT ATA AC heparin degrading 8,190 ND 21. K AAAAAG AAA AAA heparan sulfate degrading 9,328 30.8 12 35 L CTT, CTA, CIG, TIG, CG CTG hydroxylapatite TTA, CTC heparinase 7,150 15.3 18 M ATG ATG ATG heparinase II 2,049 28.41 3 N AAC, AAT AAC AAT heparinase II 5,150 44.46 7 P CCC, CCT, CCA, CCG CCG CCG 40 Q CAG, CAA CAG CAG R CGT, AGA, CGC, CGA, CGT CGC TABLE 2 AGG, CGG Propertied of heparinases from S TCA, TCC, TCG, TCT, TCT Flavobacterium heparinar 45 AGC, AGT T ACG, ACC, ACT, ACA ACC or ACT ACC or ACA sample heparinase I heparinase if heparinase III is a w-sus 4 as p W GTC, GTA, GTT, GTG GTT Km (M) 17.8 57.7 29.4 W GG TGG TGG Keat (s) 157 23.3 164 substrate H H and HS HS Y TAC, TAT ETHER TAT specificity

SEQUENCE LISTING

( 1) GENERAL INFORMATION: ( i i i ) NUMBER OF SEQUENCES: 26

( 2) INFORMATION FOR SEQID No:1: ( i) SEQUENCE CHARACTERISTICs: ( A LENGTH: 2339 base pairs (B) TYPE: nucleic acid (C) STRANDEDNESS: double (D) TOPOLOGY: linear

5,681,733 25 26 -continued

G GAT G.A.T G G A C C C CTA TTI GAA G GAA CT GAA G GAA C CACC GT TT TTGAGATAA 23 39

( 2) INFORMATION FOR SEQIDNO2: ( i) SEQUENCE CHARACTERISTICS: (A) LENGTH: 772 amino acids (B) TYPE: amino acid (D) TOPOLOGY: linear ( i i ) MOLECULETYPE: protein ( xi ) SEQUENCE DESCRIPTION: SEQID NO:2:

Me t L. ys A rig G in L. eu Ty r Le u Ty r Wa. e. Phe Wa Wa Wa G 1 u eu 1. 5 15

Me t W P e Thr Ly is G y Ty r S e r G 1 in T. r ... y s A a A sp Wa 1 Wa 1 O 25 3 0

Lys A sp Wa As p G 1 y Wa S e M t 1 e Pro Pro His 35 a 0 45

A. I g L. eu eu A rig G in G in a A sp I. c. lu As in Arg 50 55 6 O Me t As in A sp P y s L. eu Wa 1 A 1 a As p Me t e I y s Me t 65 O 5 8 O

G in G u Trip ... ys A 1 a As p e P. c. G Wa As p Phe Arg 85 9 O 95

P. e. As in G in Gly Le u Thr Wa A r g Wa 1 1 I eu Me t A 1 a 1 O 5 1 O

As in L. eu Me t Pro a G 1 y G A la I e

S e e L. eu T P he A a G 1 y 13 35 4.0

1 e I e G 1 y Le Phe Me t Wa 1 T. r A 1 a 1 e Wa 1 s 155 6 O

As p Tr p A sp G 1 a P G 1 u. G r A rig Phe O

W A a P he Wa A rig I e A 1 a Met L. eu u 18 O

Ly is A sp I e Wa Gly H is A 1 a G Trip Me t I 1 e 9 is 2 0 2 O 5

Me t A sp Let u L. eu w a Gly I e A 1 a e Ty r A sp Phi e Pro 2 5 22 O

G Me t As in A 1 a A a G 1 y Arg P he P e Ly is is 225 23 O 35 2 - O

A 1 a As in P. e. Ty r His As in Ty r Hi is G in G 1 y Me t Se 2 4 5 25 O 255

Let u As in Wa A rig P he Thr As in I tu P. e. A 1 a Le e L. eu 2, 6-0

As p Me t gly A 1 a Gly As in Wa P e As in Gy G in G in P a e e 275 28 0. 285

L. eu Tyr A sp A 1 a 1 e A rig A rig G 1 e I. e. A 1 a 29 O

G y As p Wa As p A rig Thr Me t 3 20

A 1 a L. eu. A 1 a As p I e As in Ty r 3 25 3 35

G P. e. e Lys A sp c. As a W G P H is L et P he u 34 0. 3 4 5 3 5 O

5,681,733 29 -continued

7 O

( 2 INFORMATION FOR SEQID NO:3: (i) SEQUENCE CHARACTERISTICS: (A) LENGTH: 1980 base pairs (B) TYPE: nucleic acid (C) STRANDEDNESS: double (D) TOPOLOGY: linear ( i i ) MOLECULETYPE: DNA (genomic) ( xi ) SEQUENCE DESCRIPTION: SEQID NO:3:

A GA CAC GA AAA TTIAA AAG GAT CAT GTATTG CTG AAT GCC CT is 0.

A CGT CGG GA AATAACT G. CA CAAA G CTC T CCATTACC AGGAAAGAT 0 0

GACCA CAT CAAC CTAG A T C C G GAC G GAAAAG. T. TAATAAA G CA 50

GT T G C GCCG G CAA CAT GA CGA G CG GCC AAA G CATTAC G G CATACT 2 O O

CA. GGGAAAAA A GTAAG GC CA GG GAA C C T GA. TT CA GTA.A.T G CAGAAAAG C 250

CT GCC GATAT A C GCCA C C C ATAGA TAAG G TAC GC GT CA AA T G GCC GAC 3 O 0.

AAG G CTTGG T C CAC CAGT CAA CCG CAC AAAG G CAC G G CTATTGA 3 is

A T G G T AAA GACA CAA CT GCACA GT G CCG GAAAA GACAATGAAG 4 O

A CGC G CA TG CAC CT T.A.A.A.A. G. GT G G CAG GCTA C C C is 4 50

TAT CAC CTA C. G. G. G. C. GAT GA AAAATA G CA. A GA GAA. GGG ATACA GTA 5 O CA. G. C. GATT G GCCA GAAAAA A C C CATT G. G. G. cc T G T cc cag GATAATGATA 55 AA T G T G T G GCG G C C C CTT GAA G T G T C G A CAG G GACA A.A. G. CTT CCC 6 OO

C CAA CCTT CA GCTTATTTGT AAA CT C C CA GCCTTTACCC A cc TTTTT 55

A. A. G. GAATTT TTAAACA GTT A. CCACCAA CA. GGCC GATA TAT CTAC GC 00

ATA T G C CGA A CAG G GAA. A. C. CA C C GTTTAT TTGAA C C CA A C GCAA CTG 5

T. G. CAG G G TAT C C C C GAATAAA GATT CAC CAA GAT G. AGG CA 8 00

AA C C G G CATA TCG. G. G. C. A A.C.A C C GA. GAT CAAAAAACAG GT ATCCG 850

A T G G GAT G CA G GAA CT T CAC CAATT A CCA. G. A. C. T C CAT CGA 9 O

ATCTT CTTAA AG GCC TAG G TCT CAAAA. CGAGTAACC T GAAAAAGA 95

A T C C G CAA TCTTAT GAC AAA CGT AGA AAAA. GAT A. GGC GCT GA O O 0

CA GTATTTC ACT G CCAGAT AAA CACCC CATT G is AGAT CA. G. G. O SO

ATACA GATA AAAA. T. CAG GA GCACA G TTT G. C. CAG CT G. G. G. C. C. G. G. G. 1 O

T C C C G CA AAC CAGGC CA AAAA. TAT TT GCTA CAGA GG CAAACAAG SO

AAG C GCC TAACTTA TCCAAA G CAT GAG CAAT G. C. AG CTTTA 20 0

A CGT TAGAA GCG GA. GGGA TAAAAA T G CA A C C GT TAT G. G. TATAAAAGC 25 O

CA C C C C C GGG GAATTT c A G C C CAGCC G GATAACGG G ACTTGAAC 3 O O.

TAAAA GGG CAGAAAC TAC C C CAG A C GCC GGG GT A A 35

A C C C C C CA. C. c. AAG C T CA GAA C C C GC AACT G GTAC C GT CAAA C C CG 4 O O

(CATACA CA. Gi C A CGCTA CAC TCGA CAA CA AAAT AT G. G. C. ATAC CAAA. G. 4 S O C C C C C CAAAA CA.A.A. GGGAA. A CAG GAA ATA A C C GAT GT G. cAC CAT 5 O O. A CCAA C C CAA C CAT C C GAA TCT G GACCA CA G C G CA GT TACTTT CA 550

(CAA CAAAAAA. TACTT T. CTGG T CAT C GAAG GGCAA AG. G. C. GAA G CTAC CG 6 O O.

GAA. A CCT G. G. G. C GT ACA CT GG CA CAAA. G. A.A.G.A. CAG CAA CCCT GTTTC 55 0. 5,681,733

-continued

GATAAA CAA A GAA C C G G G T ACAC CA CT TACA GA, GA. G. GAA CAA C CT 7 00

GAT GAT CCAA CG T GAA T G CG GA CA. G. G.A. C. CAGCCT CAAT GAA GAA. GAA G 1 50

GAA. A. GG AT C T TA GTT AC AATAAG GAG C T GAAAAGA C C T G CTC GA 1800

T GAAAAGC (CTAAAAA GAA T G C C GCACA CAAAA TT G CAG TATA GT 1 85 0.

TAC CATAC GA CG GCCA GA. A. G. G. C C CAGA, GAT CAG CAA CGG GAAAACA 19 O O.

A. GGG CAATGA TT GA. G.A.A.A. GGCAAG CTA AT CTA ACC CT TAC CAT AAC 1950

G GAAAACAA C A G CT G T G TT G G T C CTTAG 1980

( 2) INFORMATION FOR SEQID NO:4: ( ; ) SEQUENCE CHARACTERISTICS: ( A LENGTH: 659 amino acids (B) TYPE: amino acid (D) TOPOLOGY: linear i i MOLECULETYPE: protein ( x i ) SEQUENCE DESCRIPTION: SEQID No.4: Met Thr Thr Lys I e Phe Lys Arg 1 e e Wa Phi e A a Wa 1 e A 1 a 1. 5 O L. eu Se r S e r G 1 y As in I e E. eu A 1 a G in Se r S e r S e r I 1 e Thr Arg Lys 2 0 25 3 O A s p Ph e A s p H is I 1 et As in L. eu G 1 u Tyr Ser G 1 y Le u Gi i u Lys V a 1. As in 35 4 O 45 Lys A a Wa 1. A 1 a. A 1 a G y As in Ty r A s p A s p A 1 a. A a Lys A 1 a Le u L. eu SO 55 60 A la Tyr Ty r A rig G1 u y s Ser Lys A a Arg G 1 u Pro A. s p Ph e Ser A. s. n. 55 7 0 T 5 8 D A 1 a G 1 u Lys Pro A la A s p e Arg G in Pro I e A s p Lys Wa Th T A rig 85 9 O 95 G 1 u Me t . A 1 a. A s p Lys A a Le u Wa 1 H is G 1 in P he G 1 in Pro H is Lys G 1 y O 1 O 5 1 1 0 Tyr G 1 y Tyr Phe A s p Tyr G i y Lys A s p I 1 e A is in Trp G 1 in Met Trp Pro 15 1 ... O 25 Wa 1 Lys A s p As in Glu Wa Arg Tr p G 1 in Le u H is Arg Wa 1 ly s Trip Trip 30 1 35 1 4 0 G n. A a Me t A 1 a L. eu Wa Tyr H is A. 1 a. Thr G 1 y A s p G 1 u Lys Ty r A 1 a 45 5 O 55 1 6 O A rig Gi i u Trip W a Tyr G 1 in Tyr Ser A. sp r p A 1 a. A rig Lys As in Pro Le u 1 65 1 O T 5 G l y L. eu Ser G in A s p As in A s p Lys Phe Wa 1 Trip Arg Pro E eu G 1 u Wa 1 8 0. 85 19 O Se r A s p A rig Wa G 1 n Ser Le u Pro Pro Thr Phe Ser Le u Ph e V a As in 195 20 2 O 5 Ser Pro A 1 a Phe Thr Pro A 1 a Ph e Le u Me t G 1 u Ph e Le u As in Ser Ty r 2 1 0 2 5 22 O His Gl in G in A 1 a. A s p Ty I L. eu Ser Thr H is Ty r A 1 a G 1 u G 1 in G 1 y As in 225 23 O. 235 2 A.. O His Arg Le u Ph e G 1 u A 1 a G 1 in A rig As n L. eu Ph e A1 a G y W a 1 Ser P he 2 4 5 25 O 25 s Pro G i u Ph e Ly is A. s p Ser Pro Arg Tr p Arg G 1 in Thr G 1 y e Sier V a 1 25 O 2 65 27 O Le u As in Thr G 1 u I e Lys Lys G 1 in Val Ty r A 1 a. A s p G 1 y Me t G 1 in Phe 275 28 0. 2. 85 5,681,733 33 34 -continued

G u L. eu Ser Pro I 1 e Ty r H is Wa A 1 a A la I 1 e As p I 1 e Phe ... eu Ly is 9 O 295 3 O

A 1 a Tyr G 1 y Ser A 1 a Lys A rig Wa As in Le u G 1 u Lys G P e Pro G in 3 O 5 3 O 3 15 3 2 O

Ser Tyr V a 1 G 1 in Th r V a 1 G As a Met I e Me t A 1 a I. et u I 1 e S e e 3 25 3 3 O 335

Ser L. eu Pro A s p Ty r A sm. Me t P he G 1 y A sp S e I 1 e. 3 0. 3 4 5

A s p Lys A. s. n. Phe Arg Me t A 1 a G 1 in Phe A l a Se r A 1 a Wa 1 P e 355 3 50 3, 55

Pro A a As in G 1 a. A 1 a. I 1 e I. ys P he A 1 a Thr G 1 y G 1 in G 1 y 3 O 3 5

... y s Al a Pro A. s. n. Phe L. eu See A 1 a eu Ser A 1 a G 1 y P he Tyr 385 39 0 395 4 O

Thr Ph e Arg Ser G 1 y Tr p A sp As in A a Thr Wa Me t W 1 L. eu 1 O 4 5

A 1 a Se T Pro Pro G 1 y G u P he is A 1 a G 1 in P o A sp As in P he a 20 4 is

G 1 u. L. eu Ph e I le Ly is G 1 y Arg As in P. e. Tr Pro A sp A 1 a Wa 1 P e 4 35 4 4 O 4 45

Wa I Ty r S e r G 1 y A. s p G 1 u A 1 a e Me t As n. 50 4. 55

G 1 in Thr Arg I le H is Ser T. r thr G 1 a Wa 1 4 6 5 4 4. T 5 4 8

I e Thr I y s A 1 a. A rig Gil in As in G 1 Thr As in As a I eu A sp 49 O 4.95

Wall Le u Thr Tyr Thr As in Ty r Pir o As in A sp is in Arg 50 O 5 : 5 5 1 O

Sc r Wall I e u Ph e 1 e As in Lys P he e u Wa 1 1 e A sp Arg A 1 a 5 5 is 20 52 is

I 1 e G 1 y G 1 u A a Thr G y As in eu G 1 y Wa is Tr p 1 in L. eu G 1 535 5 4 O

A s p Ser A a n Pro Wa Phe A sp A rig Wa Thr 5 : 5 550 555 5 60

Ty r A rig A s p Gly. As in As a L. eu Me it I e G in S As in A rig 5 65 57 0.

Th r S e r L. eu As in G 1 u, G 1 u G: G1 y I y s Va. 1 Se: w a 1 58 0 585

G i u I. eu Lys Arg Pro A 1 a P he Wa P e G 1 u Lys Lys Ly is As a 595 6 O O 6 05

G 1 y Thr G 1 in As a Ph e V a 1 S I e Wa Ty r As p G 1 y G in 6 10 6 15 520

A a PI o G 1 u I e Sier I 1 e A rig G u As a Lys G 1 y As in As p Phe G Ly is 625 3 6 3.5 6 40

G l y Ly s eu. As a L. eu Thr I e u Thr e As n G 1 y Il y s G a G in ... et u Wa 6 4.5 6 is 0. is 55

e Wa 1 P r is

( 2 INFORMATION FOR SEQID NO5: ( i ) SEQUENCE CHARACTERISTICS: (A LENGTH: 12 amino acids (B) TYPE: amino acid (D TOPOLOGY: linear 5,681,733 35 36 -continued ( i i ) MOLECULETYPE: peptide ( xi ) SEQUENCE DESCRIPTION: SEQID No.5: G 1 u Ph e Pro G i u Me t Ty r A is in L. eu. A 1 a. A 1 a G 1 y Arg 5 1 O

( 2) INFORMATION FOR SEQID NO:6: ( i SEQUENCE CHARACTERISTICS: (A) LENGE: 12 amino acids (B) TYPE: amino acid (D) TOPOLOGY: linear ( i i MOLECULETYPE: peptide ( x i ) SEQUENCE DESCRIPTION: SEQDD NO:6: y s Pro A 1 a. A s p I 1 e Pro G 1 u Wall Lys A s p G 1 y Arg 1. 5 1 0

( 2 INFORMATION FOR SEQID NO:7; ( i) SEQUENCE CHARACTERISTICS: (A. LENGTH: 27 amino acids (B) TYPE: amino acid ( D TOPOLOGY: Finear ( i i MOLECULETYPE: peptide ( x i SEQUENCE DESCRIPTION: SEQID NO:7: Le u A 1 a G 1 y A s p P he V a 1 Thr G 1 y Ly is I 1 e Le u A 1 a G 1 in G 1 y P he G 1 y 5 O 5 Pro A. sp. As n G 1 in Thr Pro A. s p Tyr Thr Ty r Le u 2 0 25

( 2) INFORMATION FOR SEQID No:8: ( i SEQUENCE CHARACTERISTICS: (A) LENGH: 14 amino acids (B) TYPE:ainino acid (D) TOPOLOGY: linear ( i i MOLECULETYPE: peptide ( x i ) SEQUENCE DESCRIPTION: SEQID NO:8: Let u I e Lys As in G 1 u Wa 1 Arg Tr p G 1 in Le u EE is Arg Wa 1 Lys 5 O

( 2) INFORMATION FOR SEQID NO:9: ( i SEQUENCE CHARACTERISTICS: ( A LENGTH: 23 amino acids (B) TYPE:amino acid (D) TOPOLOGY: linear ( i i) MOLECULETYPE: peptide ( xi ) SEQUENCE DESCRIPTION: SEQED NO;9: Wa L. eu Lys Al a Ser Pro Pro G y G i u Ph e H is A 1 a G 1 in Pro A s p As in 1. 5 1 O 15 G 1 y Thr P he G 1 u Le u Ph e I 1 e 2 O

( 2 INFORMATION FOR SEQID No:10: ( i) SEQUENCE CHARACTERISTICS: ( A LENGTH: 23 amino acids ( B TYPE:amino acid ( D TOPOLOGY: linear 5,681,733

-continued

( i i MOLECULE TYPE: peptide ( xi ) SEQUENCE DESCRIPTION: SEQID No:10: Lys A 1 a L. eu Wa 1 H is Trp Phe Trp Pro H is Ly is G 1 y Tyr G 1 y Tyr Phe 1. 5 1 O 15 A s p Tyr G 1 y Lys Asp I 1 e A sn O

( 2 INFORMATION FOR SEQID NO:11: ( i) SEQUENCE CHARACTERISTICS: (A) LENGTH: 29 base pairs (B) TYPE: nucleic acid (C) STRANDEDNESS: double (D) TOPOLOGY: linear ( i i ) MOLECULE TYPE: DNA genomic) ( xi ) SEQUENCE DESCRIPTION: SEQID No:11:

GAA. T C C CT G. AGA GTA CAA. T C C C C C GC 29

( 2 INFORMATION FOR SEQID No:12: ( i) SEQUENCE CHARACTERISTICS: (A) LENGTH; 26 base pairs (B) TYPE: nucleic acid (C) STRANDEDNESS: double D TOPOLOGY: linear (ii) MOLECULE TYPE. DNA (genomic) ( xi ) SEQUENCE DESCRIPTION: SEQID NO:12:

CCG GCA C C A GA GTA CAT TTCACG 26

( 2) INFORMATION FOR SEQID NO:13: ( i ) SEQUENCE CHARACTERISTICS: (A) LENGTH: 29 base pairs (B) TYPE: nucleic acid ( Cy STRANDEDNESS: double ( D TOPOLOGY: linear ( i i ) MOLECULETYPE: DNA (genomic) ( xi ) SEQUENCE DESCRIPTION: SEQID NO:13:

AAA C C C C C C G A CAT C C CGA AGAAAACA 29

( 2) INFORMATION FOR SEQID No:14: ( i SEQUENCE CHARACTERISTICS: ( A LENGTH: 29 base pairs (B) TYPE: mucleic acid (C) STRANDEDNESS: double (D) TOPOLOGY: linear ( i i MOLECULETYPE: DNA (genomic) ( x i SEQUENCE DESCRIPTION: SEQID NO:14:

CGAAA GT C. A. C.C.G. G. A.A.T GT C GCSC 29

(2) INFORMATION FOR SEQED NO:15: ( i SEQUENCE CHARACTERISTICS: (A) LENGTH: 38 base pairs (B) TYPE: nucleic acid ( C ) STRANDEDNESS: double (D) TOPOLOGY: linear 5,681,733 39 40 -continued ( i i MOLECULETYPE: DNA (genomic) ( x i ) SEQUENCE DESCRIPTION: SEQED NO:15:

GAG GATT CA T G CAAA CCAA GCC GAT G. G. GTTT G GAA 3. 8

( 2) INFORMATION FOR SEQID No:16: ( i SEQUENCE CHARACTERISTICS: (A) LENGTH: 24 base pairs (B) TYPE: nucleic acid (C)SIRANDEDNESS: double (D) TOPOLOGY: linear ( i i MOLECULETYPE: DNA (genomic) ( x i ). SEQUENCE DESCRIPTION: SEQD NO:16:

G GAG GATA A C CA. CAT CGA G CATT 2 a.

( 2) INFORMATION FOR SEQID NO:17: ( i) SEQUENCE CHARACTERISTICS: (A) LENGTH: 27 base pairs (B) TYPE: nucleic acid (C) SERANDEDNESS: double (D) TOPOLOGY: linear ( i i ) MOLECULETYPE: DNA genomic) ( xi ) SEQUENCE DESCRIPTION: SEQID No:17: GAATTC CA. C. A. GTT CAGCC G CATAAA 27

( 2 INFORMATION FOR SEQID NO:18: ( i ) SEQUENCE CHARACTERISEICS: ( A LENGTH: 27 base pairs (B) TYPE: nucleic acid ( C ) STRANDEDNESS: double ( D TOPOLOGY: linear ( i i MOLECULETYPE: DNA (genomic) ( x i SEQUENCEDESCRIPTION: SEQID NO:18:

GAA. T CTA GCG GCT GAA. A. C. GATG 27

( 2) INFORMATION FOR SEQID NO:19: ( i SEQUENCE CHARACTERISTICs: (A) LENGTH: 26 base pairs (B) TYPE: nucleic acid (C) STRANDEDNESS: double (D) TOPOLOGY: linear ( i i MOLECULE TYPE: DNA (genomic) ( x i SEQUENCE DESCRIPTION: SEQD NO:19: GAATT C C C G C C C GC GAAT. T. CAT G. C. 2 6

( 2) INFORMATION FOR SEQID No:20: ( i) SEQUENCE CHARACTERISTICS: A LENGTH: 26 base pairs (B) TYPE: nucleic acid (C) STRANDEDNESS: double (D) TOPOLOGY: linear ( i i ) MOLECULETYPE: DNA (genomic) ( xi ) SEQUENCE DESCRIPTION: SEQID NO:20:

GAA C GCAT G.A.A.A.T. C. G. C C C G G C G 2 6 5,681,733 41 42 -continued

( 2) INFORMATION FOR SEQID NO2: (i) SEQUENCE CHARACTERISTICS: (A) LENGTH: 29 base pairs (B) TYPE: nucleic acid (C)SIRANDEDNESS: double ( D TOPOLOGY: linear ( i i) MOLECULETYPE: DNA (genomic) ( xi ) SEQUENCE DESCRIPTION: SEQED NO:21:

GG GAATCC AT GCC CAGCC GAA.A.T G SAC 29

( 2 INFORMATION FOR SEQID NO:22: ( i SEQUENCE CHARACTERISTICS: (A) LENGTH: 28 base pairs (B) TYPE: nucleic acid ( Cy STRANDEDNESS: double (D TOPOLOGY: linear ( i i MOLECULETYPE: DNA (genomic) ( xi ) SEQUENCE DESCRIPTION: SEQID NO:22:

GT C CATT C GCT. G. G. CAT G AAATCCC 2 8

( 2 INFORMATION FOR SEQID NO:23: ( i SEQUENCE CHARACTERISTICS: ( A LENGTH: 28 base pairs (B) TYPE: nucleic acid (CSTRANDEDNESS: double ( D TOPOLOGY: linear ( i i ) MOLECULETYPE: DNA (genomic) ( xi ) SEQUENCE DESCRIPTION: SEQED NO:23:

GT CAT CASTT CAGCC CAAA AG GTAT G. G. 8

( 2) INFORMATION FOR SEQID NO:24: ( i ). SEQUENCE CHARACTERISTICS: (A) LENGTH: 28 base pairs (B) TYPE: nucleic acid (C) STRANDEDNESS: double (D) TOPOLOGY: linear ( i i ) MOLECULE TYPE: DNA genomic) ( xi ) SEQUENCEDESCRIPTION: SEQID No.24:

CCCAA C C. A. G. G. GC GAA CT AGAC 8

( 2) INFORMATION FOR SEQID NO:25: (i) SEQUENCE CHARACTERISTICS: ( A LENGTH: 27 base pairs (B) TYPE: nucleic acid (C) STRANDEDNESS: double (D TOPOLOGY: linear ( i i ) MOLECULETYPE: DNA (genomic) ( xi ) SEQUENCE DESCRIPTION: SEQED NO:25: C. G. C. G. A. CCA G CAAA CT C T C CAT 27

( 2 INFORMATION FOR SEQID No:26: ( i y SEQUENCE CHARACTERISTICS: 5,681,733 43 44 -continued (A) LENGTH: 27 base pairs (B) TYPE: nucleic acid (c) STRANDEDNESS: double (D TOPOLOGY: linear ( i i MOLECULE TYPE: DNA (genomic ( x i SEQUENCE DESCRIPTION: SEQID NO:26:

CGC G GA. C. CT CAAA C CTTG. C. CTTT CT C 2

We claim: 4. Isolated and purified Flavobacterium heparinum hepa

1. Isolated, recombinant non-glycosylated Flavobacte- 15 rinase III whereinw the amino terminus is not a modified rium8 heparinum heparinase- II in substantially pure form. pyroglutamate5. Isolated andresidue. purified heparinase III comprising the 2. Isolated and purified recombinant non-glycosylated amino acid sequence of SEQID NO:4 wherein the amino heparinase II comprising the amino acid sequence of SEQ terminus is not a modified pyroglutamate residue. D NO:2. 6. Isolated and purified heparinase III comprising the 3. Isolated and purified heparinase II comprising the ' amino acid sequence of SEQID NO.4 as expressed in E. coli amino acid sequence of SEQID NO2 as expressed in E. coli cells. cells. s: : ; ; ;