CA1282020C - Proteins for the production of bioadhesives - Google Patents

Abstract

ABSTRACT Microbial production of a bioadhesive precursor protein is disclosed. The bioadhesive precursor protein can be hydroxylated and used as an adhesive in wet environments.

Description

~;~8~2~2~
1182-345A PROTEINS FO~ THE PRODUCTION OF
pp:570 BIOADHESIVES

BACKGROUND OF T~E INVENTION
This invention relates to proteins which are use~ul in the production of adhesives. More particularly! it relates to the microbial production, by techniques of recombinant DNA technology of proteins which can be used to produce bioadhesives which mainthin their adhesive properties in wet environments. Consequently, applications of these adhesives include marine and - general underwater adhesives, medical adhesives or - prosthetic joints~ joining or repairing arteries or intestines, or the like and dental cement.
The properties of adhesives generally must be tailored to meet the requirements of the particular environments in which they are to be used. Ideall~, an adhesive should be cured and it should maintain both its adhesivity and cohesivity under the conditions of useO
Curing is the altering of the physi~al properties of an adhesive by chemical means. In the case of the bioadhesives produced by the procedures described herein, curing i~ ely to be due to the cross-linking of adjacent uncured adhesive molecules by catalytic and/or chemical agents. Curing may involve polymerization, vulcanization and/or crosslinking.
Adhesivity, which is sometimes referred to as peel strength, is the ability of the substance to adhere to foreign surfaces. Cohesivity, which is sometimes referred to as shear strength, refers to the material's internal strength and resistance to plastic deformation and i5 particularly important in situations where the applied adhesive is subject to shear loading. Many adhesives which exhibit excellent peel strength and shear strength under dry conditions su~er a substantial or ~:8~

total loss of these properties in wet environments and, more importantly, cannot be cured in wet environments.
Thus, it has been particularly difficult to develop adhesives for use in wet environments such as marine environments or medical and dental applications.
Marine mussels and other sessile invertebrates have ~he ability to secrete adhesive substances by which they affix themselves to underwater structures. For example, the mussel Mytilus edulis deposits an adhesive substance ~ .. .
from the mussel foot which then becomes cured, forming a permanent attachment to the substrate. A major component of the adhesive plaque deposited by M. edulis has been identiied as a hydroxylated protein of about 130,00l) daltons (Waite, J.H., Biol. Chem.~ 25~:2911-2915 [March 10, 1983]). While this substance might make an excellent adhesive for use in wet environments, i~olation of the uncured adhesive from mussels or commercial use is not practical, since the extraction of 1 kg of the adhesive substance would require the slau~hter of about 5 to 10 million mussels.

SUMMA~Y OF THE INVENTIO~
This invention involves the microbial production, using techniques of recombinant DNA technology, of a protein useful in the production of a bioadhesive. The bioadhesive precursor protein which is produced by the process of the invention comprises a sequence of from about 100 to about 1500 amino acid residues which comprise from about 20% to 40~ proline residues; from about 10% to 40~ lysine residues; from about 10~ to 40%
tyrosine residues; and from 0 to about 40~ amino acid residues other than proline, lysine and tyrosine.
Preferably, the bioadhesive precursor protein of the invention is comprised of repeating decapeptides. Each decapeptide contains about 30~ proline residues, 20 lysine residues and 20~ tyrosine residues.

~ ~ ~2 ~

The protein produced by the process of the invention can be employed as a bioadhesive precursor. The adhesive properties of the protein are enhanced by hydroxylating approximately 50% of the tyrosine residues to L-3,4-dihydroxyphenylalanine (DOPA) and 60-100% oE the proline residues to 3- or 4-hydroxyproline by chemical or enzymatic means. The hydroxylated protein is cured to produce the desired physical properties in the bioadhesive. The hydroxylated bioadhesive precursor protein is analogous to the adhesive protein i~olated from the phenol gland of the mussel M. dulis (Waite, J.H., ~ iol . Chem., 258: 2911-2915 [March 10, 1983] ) .
The bioadhesive precursor protein is produced by the insertion in~o a microbial host, such as E. coli, S.
cerevisiae or Bo subtilis, of a replicable expression vehicle containing a chemically synthesized double-stranded DNA (dsDNA) sequence coding for the desired protein and expression of the synthetic dsDNA sequence in the host to yield the protein. -The synthetic dsDNA
sequence encoding the bioadhesive precursor protein of the invention is constructed of codons which are selected to optimize expression and provide for stable reproduction of the genetic information in the particular host microorganism employed.
The dsDNA sequence encoding the bioadhesive precursor protein can be linked, at its 5' end, to a sequence which encodes an N-terminal portion of a microbial protein in order to facil~tate transcription initiated at a microbial promoter. In such cases, the expressed protein will constitute a fusion of the bioadhesive precursor protein and the microbial protein fragment. The microbial protein fragment may include a signal peptide which, in the case of a host like B.
subtilis, facilitates secretion of the expression product across the cell membrane and into the surrounding medium, ,. ..

with attendant cleavage of the signal peptide.
Insertion, between t~e sequences encoding the bioadhesive precursor protein and the microbial protein fragment, of a dsDNA sequence encoding an amino acid sequence which is specifically cleavable by chemical or enzymatic methods provides a means of cleavage to separate the bioadhesive precursor protein from the microbial protein fragment.
BRIEF DESCRIPTION OF THE DRAWINGS
FIG. 1 is a diagram of E. coli plasmid pGX2213, containing the trpB and trpA regions of the tryptophan operon.
FIG. 2 is a diagram of B. subtilis plasmid pGX250 containing the -amylase gene.
FIG. 3 is a dia~ram of S. cerevisiae plasmid YpGXl.
FIG. 4 is a diagram of S cerevisiae plasmid YpGX60 containing a portion of the phosphoglycerate kinase gene.
DETAILED DESCRIPTION OF T~IE INVENTION

The bioadhesive precursor protein produced by the process of the invention comprises a sequence of from about lQ0 to about 1500 amino acid residues, preferably \ from about 600 to about 900 amino acid residues arranged in repeating decapeptides. The protein and preferably each decapeptide i5 comprised of about 20~ to about 40%
proline residues. The proline residues impart flexibility to the bioadhesivs and render the ~olecule non-globular, 50 that the bioadhesive is capable of conforming to the surface of a substrate and interacting with other adhesive molecules. The protein and preferably each decapeptide is also comprised of about 10 to about 40~ lysine residues. The lysine residues render the bioadhesive basic, which assists in bonding to underwater sur'aces, which are generally coated with a 1 r .
' ~:82~

thin film of acidic biological material It also provides reactive groups through which the prot~in can be cro~s-linked during the curing process. The protein and preferably each decapeptide also is comprised of about 10% to about 40% tyrosine residues. The phenolic tyrosine residues provide hydrogen bonding capability to the bioadhesive. Moreover, both the proline and tyrosine residues provide sites for hydroxylation. The addition of a hydroxyl group on the tyrosine to form DOPA ena~les the bioadhesive to strongly displace water molecules from the surface of a substrate. In addition to the proline, lysine and tyrosine residues, the protein and preferably each decapeptidé comprises from 0 to about 40% other amino acid residues. Preferably, these residues, if present, are selected from amino acids containing relatively small, non-reactive aliphatic side chain~, e.g., alanine, and hydroxyl containing amino acidsJ e.g., serine and threonine. These residues are preferably distributed throughout the protein chain so that no more than about 4 occur together in any given decapeptide sequence. Non-preferred amino acids are acidic amino acids, i . e., aspartic acid and glutamic acid, and cystine .
The bioadhesive precursor protein is produced by inserting a synthetic dsDNA sequence encoding the protein into a repl icable expression vector in which it is operably linked to a regulatory sequence that is capable of directing expression of the encoded protein in a host microorganism. Any suitable host microorganism including, for example, E. coli, S. cerevisiae or B.
subtilis can then be transformed with the expression vector, grown up and subjected to conditions under which the protein is expressed.
The dsDNA sequence encoding the protein is prepared by known methods of DNA synthesis. A suitable method fQr .

~2~

synthesizing the dsDNA sequence i5 the phosphite solid-phase method (Tetrahedron Letters, 21:719-722 [1980]).
Several criteria are considered in selecting the deoxyribonucleotide codons to incorporate into the synthetic dsDNA sequence coding for the protein. For example, if highly repeated DNA sequénces are employed to code for the protein, the repeated sequences may be susceptible to recombination and deletion. Therefore, ~o the degree possible, repeated DNA sequences are to be avoided. Due to the well-known degeneracy of the genetic code and thus the potential variability of the DNA
sequence that encodes a decapeptide, repetition in the DNA se~uence encoding a bioadhesive precursor protein that is comprised of repeated decapeptides can be minimized. Repetition in the DNA sequence can be further minimized by varyins the amino acid sequences of the individual decapeptides such that the bioadhesive precursor protein would be comprised of similar repeated decapeptides rather than identical repeated decapeptides.
The variations in the amino acid seguence of the decapeptide would be within the guidelines previously discussed. In addition to the avoidance of highly repeated DNA sequences, a particular microbial host ~ay have a bias in favor of using certain codons to code Eor specific amino acids, and it will be preEerred to employ these codons whenever possible. Generally, E. coli and _ subtilis have relatively low tolerances for repeated DNA sequences, whereas S. cerevislae tolerates them quite well. Accordingly, when designing a dsDNA coding sequence for expression of the protein in E coli or B.
subtil_s, avoidance of codon sequence repeat is a primary consideration, whereas preferred codon usage is a secondary consideration. When designing a dsDNA coding sequence for expression in S~ cerevisiae, preferred codon usage is a primary consideration and avoidance of codon sequence repeat is a secondary consideration.

.. .

A translation termination codon, which i5 recognized by proteins that release the translated polypeptide chain from the ribosome, must be inserted at the 3l end of the synthetic dsDNA s~quence coding for the protein. Either one termination codon (TGA or TAA) or a series of two termination codons may be utilized.
Shor~ synthetic linker sequences can be added to the 5' and 3' ends of the synthetic dsDNA sequence coding for the bioadhesive protein precursor to render it compatible with convenient restriction enzyme cleavage sites in the expression vector. The dsDNA encoding the bioadhesive precurssr protein is inserted into a replicable expression vector in such a way that it is under the control oE a regulatory sequence capable of directing expression in the host microorganism. Preferred host microorganisms are E. coli, B. subtilie and S.
cerevi~iae~ In general, those skilled in the art have achieved the greatest level of experience with expression of heterologous proteins in E. coli. ~owever, because of the repetitive nature of the dsDNA sequence involved, and because stable maintenance o~ repetitive DNA sequences is more characteristic of eukaryotic cells, efficient expression may be most easily achieved using S.
cerevisiae as a host. Moreover, S. cerevisiae DNA
__ _ . _ sequences have been identified which permit controlled expression oE foreign genes under large-scale fermentation conditions (Nucleic Acids Res., 10:2625-26~7 ~1982]). B. subtilis is a particularly useful host because of its ability to secrete proteins into the surrounding medium, thus facilitating large-scale production and recovery of a microbially-produced protein.
The synthetic dsDNA for the bioadhesive precursor protein is inserted into an expression vector under the control of a regulatory sequence containing a promoter, o~

ribosome binding site and translation initiation signal capable of effecting expression in the selected host~
The expression vector can be selected from plasmids and phages, wi~h plasmids generally being preferred. In order to facilitate expression, the synthetic dsDNA
encoding the protein may be linXed, at its S' end, to a sequence encoding an N-terminal portion of ~he microbial protein which is normally under the control of the particular regulatory sequence employed. For expression in B. subti1is, the 5' end preceding the se~uence enccding the bioadhesive precursor protein may also encode a signal peptide for a normally secreted protein which should allow the bioadhesive protein precursor to be secreted.
Expression of the protein in E. coli may be achi.eved by fusing the synthe~ic ds~NA sequence in-fra~e to a portion of an E. coli gene (including its regulatory sequence) that has already been cloned in an expression vector and is effi~iently expressed. For example, the dsDNA coding for the proteln can be inserted into cloned E. coli genes that code for one or more subunits o tryptophan synthetase or serine hydroxymethyl transferase. Expression vectors containing the E. coli -~
tryptophan synthetase and serine hydroxymethyl transferase genes have been deposited at the American Type Culture Collection, Rockville, Maryland, with accession numbers ATCC 39388 (in E. coli strain GX1668~ :
and ATCC 39214 (in K. aerogenes strain GX1704), respectively. Recombinant plasmids containing the synthetic dsDNA sequence for the bioadhesive precursor protein fused to a portion of one o these genes are used to transform E. coli and the transformants are assayed for production of the fused protein using known methods.
The synthetic dsDNA sequence encoding the5 bioadhesive precursor protein can be inserted into an S.

cerevisiae expression vector so that an in-frame fusion is created with an efficiently expressed yeast gene pres~nt on the vector. For example~ the synthetic dsDNA
sequence can be inserted into YpGXl, followed by insertion of a portion of the S. cerevisiae PGK
structural gene plus its regulatory sequences. The PGK
gene encodes phosphoglycerate kinase. Phosphoglycerate kinase is involved in glucose metabolism and can constitute up to 5% of the total yea~t cell soluble tO protein. A suitable yeast vector, identified as YpGX60 eneoding phosphoglycerate kinase and YpGXl have been deposited at the American Type Culture Collection, Rockville, Maryland with the accession numbers ATCC 3g693 and ATCC 396~2, respectively. ~he recombinan~ plasmids resulting from insertion of the syntlletic dsDNA sequence into the yeast expression vectors are used to transform S. cerevisiae and the transformants are assayed for the production of the fused protein by known methods.
Expression of th~ bioadhes~ve precursor protein in a bacillus host, such as B. subtilisr can be achieved by fusing the synthetic dsDNA sequence in-frame to an efficiently expressed bacillus gene. Advantageously this bacillus gene contains a sequence encoding a signal peptide which is required for the transport of the microbial protein across the cell membrane with attendant cleavage of the signal peptide. In this manner, the bioadhesive precursor protein, together with any portion of the microbial protein fragment which follows the signal peptide, may be secreted from the baclllus host into the surrounding medium, thereby simplifying recovery and purification of the protein.
A suitable vector for expression of the bioadhesive precursor protein in B. subtilis is the plasmid pGX2509.
This plasmid has been deposited at the American Type 3S Culture Collection, Rockville, Maryland, with the :
. ~ .

$~o~

accession number ATCC 39854. It contains the regulatory region and structural gene for the ~-a~ylase from B.
~ ]~lir~ n- strain ATCC 23844, including the signal peptide. The synthetic dsDNA encoding the bioadhesive S precursor protein can be inserted withi.n the structural gene for the mature ~amylase so that the B subtilis transformant will secrete a protein comprising a fusion of an N-terminal fragment of -amylase with the bioadhesive precursor protein. Alternatively, the -amylase gene can be al~ered by known techniques ofoligonucleotide-directed in vitro mutagenesis (Nucleic Acids Res.t 10:6487-6500 E1982~) so that the dsDNA
. _ encoding the bioadhesive precursor protein can be inserted directly a~ter the -amylase signal peptide coding sequence. In such a case, the B. subtilis transformant will secrqte the bioadhesive precursor protein directly, thus eliminating the need for cleavage to remove a microbial protein fragment from the N
terminus.
In order to separate the bioadhesive precursor protein from a fusion protein, a short DNA sequence encoding a cleavable linker sequence of one or more amino acids is advantageously interposed between the 3' end of the sequence encoding the microbial protein fragment and the 5' end of the synthetic dsDNA sequence encoding the bioadhesive precursor protein. The interposed DNA
sequence encodes a sequence of one or more amino acids which can be selectively cleave~ by ~hemical or enzymatic means. When the bioadhesive precursor protein is devoid of methionine residues~ a preferred cleavable linker sequence is methionine, which is encoded by the sequence ATG. When a methionine residue is interposed between the microbial protein fragment and the bioadhesive precursor protein, the expressed fusion protein can be treated with cyanogen bromide, which selectively cleaves on the ~L~82~X~

carboxyl side of the methionine residue to release the bioadhesive precursor protein rom the microbial protein fragment. Other known cleavable linker sequences can be employed to join the bioadhesive precursor protein to the microbial protein ~ragment, provided only that the cleavable sequence is one which does not occur within the bioadhesive precursor protein itself. For example, a DNA
sequence encoding the amino acid sequence Asp-Asp-Asp~
Asp-Lys can be inserted between the microbial protein fragment and the bioadhesive precursor protein. This sequence is recognized and cleaved on the carboxyl side of the Lys residue by the enzyme bovine enterokinase.
Other suitable amino acid sequence~ which are specifically recognized and cleaved by selected elzymes are known to those skilled in the art (see published European Patent Application No~ 035384).
The expression vector containing the inserted dsDNA
coding for the bioadhesive precursor protein is used to transform the host microorganism using known techniques of transformatlon. The trans~ormant is then grown up in a fermentor and subjected to conditions under which the polypeptide is expressed. The expression product is recovered by conventional techniques. In ~he case oE
both yeast te.g., S. cerevisi~e) and E. coli hosts, the ~ . .
expression product will generally be sequestered within the cell, necessit~ting cell lysis prior to recovery. If the protein is produced in a bacillus host (e.g., ~.
subtilis) using a DNA sequence encoding an appropriate signal peptide, the protein attached to the signal peptide will pass through the cell membrane, with attendant cleavage of the si~nal peptide, and may result in secretion of the protein into the surrounding medium.
Following recovery~ if necessary, the expression product can be treated chemically or enzymatically to ~eparate the bioadhesive precursor protein from any fused ~x~o~

microbial- or linker-derived protein fragment. The microbially produced bioadhesive precursor protein can be converted to a bioadhesive having excellent underwater adhesive properties by hydroxylating the tyrosine residues and curing the hydro~ylated protein.
Preferably, at least about 50% of the tyrosine residues are hydroxylated following the recovery of the prote-n.
The hydroxylation can be carried out by the use of commercially available enzyme~ such as mushroom polyphenoloxidase.
The bioadhesives produced by the methods of this inven~ion can be used in a conventional manner and, if desired, may be admixed with conventional synthetic polymer adhesives, fillers, coacervates and/or adjuvants generally employed in the adhesive industry. ~hey are particularly useful where performance in wet environments is de~ired, such as marine adhesives or adhesives for medical or dental use, protective coatings or solder for electrical circuit~.
~he ~olowing examples are intended to Eurther illustrate the prac~ice of the invention descr;bed herein and are not intended to limit the scope of the ~nvention.
In the examples, the method employed for the \ synthesis of oligodeoxyribonucleotides i8 the phosphite solid-phase method (Matteucci, M D. and Caruthers, ~.H., Tetrahedron Letters, 21:719-722 [19~01 using an automated solid-phase DNA synthesizer manufactured ~)y Applied Bio~y~tems, Inc. The starting materials such as the four appropriately protected 5'-dimethoxytrityl-2'-deoxyribonucleoside-3'-phosphoramidites as well as the solid support such as silica and controlled pore glass (CPG) ~Adams, S.P., Kavka, K.S., Wykes, E.J., Holder, S.B. and Gallappi, G.R., J _Amer. Chem. Soc., 105:661-663 [1983]) derivatized with appropriately protected 5'-~ % 8 dimethyltrityl-2'-deoxyribonucleosides~ are commercially available~
rrhe DNA synthesis proceeds from the 3'-end to the 5'-end. Following is a specific example for the synthesis of the single strand 5' GCG AAA CCA AGT TAC CCA CCG ACC TAC AAA 3' The derivatized solid support containing approximately 1 ~m of protected 5'-dimethoxytrityl-2'-deoxyadenosine is loaded in a synthesis column and placed into the automated DNA synthesizer. The coupling cyc~e consists of detritylation of the solid support with 2%
trichloroacetic acid in dichloromethane: washing with methanol, anhydrous nitromethane and then with anhydrous acetonitrile; simultaneous addition of appropriately protected 5'-dimethoxytrityl~ deoxyribonucleoside-3'-phosphoramidate ~10 ~Im) in acetonitrile and tetrazole (30 ~m) in acetonitrile, incubation for 3-5 minutes, o~idation with iodine in a mixture of tetrahydrofuran, lutidine and water 12:1:2]; capping of unreacted 5'- ;
hydroxyl groups with acetic anhydride and dimethyl-aminopyridine in tetrahydrofuran; and final washing with nitromethane followed by anhydrous acetonitrile. The coupling cycle is repeated until the desired length of DN~ is obtained. The DNA is then partially deprotected with the treatment with thiophenoxide in dioxane/triethylamine and it is released from the solid support by several (2-4~ brief treatments t5-10 minutles) with concentrated ammonium hydroxide. Completely deprotected DNA is obtained by heating the concentrated ammonium hydroxide solution at 60-65C for 8-14 hours.
The DNA is then purified either by ion exchange HPLC
usin~ weak ion-exchange resin and a linear gradient of 0.1 M to 0.4 M phosphate buffer containing 20%
acetonitrile, or by acrylamide gel electrophoresis~ The purified DNA is enzymatically phosphorylated at the 5'-end and characterized prior to subsequent ligation.

,, .

z~

EXAMPLE I
Production in E. coli of NH2-(Ala-Lys-prQ
Ser-Tyr-Pro-Pro-Thr-Tvr-Lvs) -COOH

2 Q , The synthetio dsDNA coding for each of the twenty repeating decapeptide units can be selected ~rom se~uences in which the coding strand has the formula GCN AAR CCN (AGY or TCN ) TAY CCN CCN ACN TAY AAR
wherein G, A, T and C represent deoxyribonucleotides containing the bases guanine, adenine, thymine and cytosine, respectively; R represents a deoxyribo-nucleotide containing guanine or adenine; Y represents a deoxyribonucleotide ~ontaining cytosine or thymine; and N
represents G, A, T or C~
For expression of the 20-repeat protein described above in E. coli or B. subtilis, the following ive double-stranded oligodeoxyribonucleotide sequences are used in preparing the dsDNA insert encoding the bioadhesive precursor protein.
S' *GCGAAACCAAGTTACCCACCGACCTACAAA
GGTTCAATGGGTGGCTGGATGTTTCGCTTT* 5' S' *GCGAAACCCAGCTATCCTCCGACATATAAA
GGGTCGATAGGAGGCTGTATATTTCGCTTT* 5' 5' *GCGAAACCTTCTTATCCGCCTACCTATAAG
: GGAAGAATAGGCGGATGGATATTCCGCTTT* 5' : 30 5' *GCGAAACCGAGTTACCCACCAACGTACAAG
GGCTCAATGGGTGGTTGCATGTTCCGCT~T* 5' S' *GCGAAACCGTCGTACCCGCCCACCTACAAA
GGCAGCATGGGCGGGTGGATGTTTCGCTTT* 5' These oligodeoxyribonucleotides were selected with a view toward minimizing repeated DNA sequences which might lead to deletion or recombination by the host. The five oligodeoxyribonucleotides are ligated to each other randomly in order to produce a sequence coding for 20 repeats of the decapeptide.

' . ' :: ' ' ~28~0X~

The autoinated DNA synthesi~er is employed to synthesize the ten single-strandeà oligodeoxyribo-nucleotides having the sequences shown above. There are also synthesized two single-stranded oligodeoxyribo-nucleotides which can be annealed to form the followingblunt-ended linker fragment for the 5~ end of ~he dsDNA
insert.
NcoI
~
Xba I BamEI I I I
5' CTA GAG GG~ TCC ATG
GAT CTC CCT AGG TAC CGC TTT* 5' Finally, there are synthesized two single--stranded oligodeoxyribonucleotides which can be annealed to form the followlng blunt-ended linker fragment for the 3~ end of the dsDNA insert.

StoD
~jEcoRI
5' *GCG AAA TGA ATT C
ACT TAA G 5' The 5' ends of the single-stranded oligodeoxyribo-nucleotides labelled with asterisks are phosphorylated by treatment with adenosine triphosphate in the presence of polynucleotide Xinase.
The ten oligodeoxyribonucleotides (1.6 ~9 each) representing the decapeptide and the four linker fragment oligodeoxyribonucleotides ~0.4 ~g each) are annealed and ligated in the presence of T4 DNA ligase. The ligation reaction produces a family of blunt-ended dsDNA's of varying length having an average of about 20 of the decapeptide-encoding fragments, ligated in random order, preceded and followed by the 5' and 3' linker fragments, respectively. The synthetic dsDNA thus produced encodes ~320X~

a series of repeated decapeptides directly preceded by the sequence Leu-Gly-Ser-Met~ encoded by the 5' linker and followed by the dipeptide Ala-Lys, encoded by the GCGAAA sequence preceding the stop codon (TGA) in the

3' linker. The synthetic dsDNA from the liqation reaction is run on a 6~ polyacrylamide gel according to the procedure of Maniatis, et al~ (Biochemistry, 14:3787-3794 [1975]). The band corresponding in size to 20 decapeptide-encoding sequences i5 cut from the gel and the dsDNA is ele~troeluted from the gel. If it is desired to produce a protein havin~ fewer or greater than 20 decapeptide repeats, then the ratio of oligodeoxyribo-nucleotide fragments representing the decapeptlde to linker fragments in the ligation mixture is adjusted proportionately and the dsDNA fragment of desired size is isolated from the gel.
The synthetic dsDNA sequence is then inserted into the single EIpaI ~ite of plasmid pGX2213. Referring to Fig. 1? this pla~mid contains the trpB and tr~A regions of the tryptophan operon under the control of a tac (hybrid tr~/lac) promoter. ~he plasmid has been deposited, in an E. coli host ~train GX166B), at the American Type Culture Collection, Rockvilla, Maryland, with accession number \ ATCC 39388. The plasmid (l ~ g) is cleaved by treatment with ~e~I (1 unit). Cleavage occur~ within the trpB
region, leaving 1122 base pair at the 5' end of the trpB
gene linked to the tac promoter. The linearised plasmid has blunt ends. The dsDNA insert containing the decapeptide-encoding sequences (0.2 ~g) i9 blunt-encl liga';ed to the linearized pGx2213 using T4 DNA liga~e.
The recircularized plasmids are used to transform E.
coli strain JM109. ~his host strain is commercially available, e.g., from P-L Biochemicals, Inc., Bethesda Research Lboratorie j Inc. and New England BioLabs, Inc.

~7 It contains the lacIq gene which overproduces the lac repressor protein which regulates expression from the tac promoter. Expression from the tac promoter can be induced by the addition of isopropyl~ -D-thiogalactoside (IPTG)o The host is also recA , which reduces the likelihood of recombination of repeated decapeptide-encoding sequences in the dsDNA insert. The transformed E. coli JM109 cells are inoculated onto LB-agar plates containing ampicillin and grown for 24 hours. Plasmid DNA
prepared from the re~ultant colonies is screened by restriction analysis to isolate clones containing a single dsDNA insert in the proper orientation. The isolated clones contain a fused gene which codes for a protein containing the first 374 amino acids of the _~eB
gene product fused to the five amino acids coded for by the linker at the 5' end of the synthetic dsD~A insert ~Leu-Glu Gly-Ser-Met), followed by twenty repeats of the decapeptide sequence of the insert and terminating with Ala-Lys, which i5 coded for by the portion of the linker at the 3' end of the synthetic insert preceding the stop codon. The coding region i~ un~er the control of the tac promoter, including the untranslated region following the TGA stop codonO
Transformants containing ~he single dsDNA insert in the proper orientation are inoculated into 2-liter culture flasks containing Luria broth and ampicillin and are grown to mid-lo~ phase (OD600~ 0.5). IPTG (0.25 mM, final concentration) i8 added to induce expression.
After 8-16 hours, the fusion protein is expressed at high level~ in the host cells. The cell~ are harvested by centrifugation, lysed by sonication, and the fusion protein i5 recovered by conventional protein recovery techniques. ~he fusion protein is treated with cyanogen bromide, which cleaves the protein on the carboxyl side `~.

2~) ~8 of the methionine residue immediately preceding the first decapeptide sequence, thereby separating the bioadhesive precursor protein from the N-terminal fragment of the ~B gene product and the linker-derived peptide fragment. The bioadhesive precursor protein is then isolated by conventional procedures.

EXAMPLE II
Production in B. su~tilis of NH~ la-Lys-Pro-Ser-T~,rr-Pro-Pro-Thr-TYr-Lvs) -COOH
n The automated DNA synthesizer is employed to synthesize the same ten oligodeoxyribonucleotides representing the decapeptides that are produced in Example I, and the 5' ends o~ the oligodeoxyribo~
nucleotides are phosphorylated by treatment with adenosine triphosphate in the presence of polynucleotide kinase.
The following single-stranded ~ligodeoxyribo-nucleotides are synthesized which can be annealed to provide a linker for the 5' end of the dsDNA insert.
5' AGCTTTATGATG
AATACTACCGCTTT* 5' The following oligodeoxyribonucleotides are synthesized and annealed to provide a llnXer for the 3' end of the dsDNA insert.
5' *GCGA M TGATAAGGATCCA
ACTATTCCTAGGTTCCA 5' The S' ends o~ the linker fragments labelled with the asterisks are phosphorylated by treatment with adenosine triphosphate in the presence of polynucleotide kinase.

82~0 The ten oligodeoxyribonucleotides (1.6 ~g each) representing the decapeptide and the four linker fragment oligodeoxyribonucleotides (0.4 ~g each) are annealed and ligated in the presence of T4 DNA ligase. The ligation reaction produces a family of dsDNA's of varying length ha~ing an average of about 20 repeats of the decapeptide-encoding fragments, ligated in random order, preceded and followed by the 5' and 3' linker fragments, respectively.
The synthetic dsDNA thus produced encodes a series of repeated decapeptides directly preceded by the sequence Ser-Phe-Met-Met, encoded by the 5' linker, and follo~ed by the dipeptide Ala-Lys, encoded by the GCGAAA sequence preceding the two stop codons (TGA, TAA) in the 3' linker. The synthetic dsDNA from the ligation reaction 15 i5 run on a 6% polyacrylamide gel according to the procedure of Maniatis et al., ~ . The band corresponding in size to 20 decapeptide-en~oding sequences is cut from the gel and the dsDNA is electroeluted.
The synthetic dsDNA sequence is then inserted into the single Hind~II site of plasmid pGX2509. Referring to Fig. 2, this plasmid contains the regulatory region and structural gene for ~-amylase derived from B.
amyl-oliquefaci-ens. A B. subtiIis strain containing the plasmid (GX7027~ has been deposited at the American Type Culture Collection, Rockville, Maryland, wîth the accession number ATCC 39854. The plasmid ~1 ~g) is cleaved by treatment with HindlII (1 unit). The HindIII-digested plasmid is ligated with the dsDNA insert at a total DNA conoentration of about 50 ~g/ml.
The ligation mixture is used to transform B.
subtilis strain lS53 from the Bacillus ~enetic Stock Center, Columbus, Ohio, using the method of S. Chang and S.N. Cohen, Molec._Gen. Genet., 168~ 115 ~1979).
Transformants are screened for the failure to produce ~-~2~2C~O

amylase and for the presence of a HindIII insert of the desired size. The oriçntation of the insert is determined by digesting with ~amHI and determining the size of the smaller fragment by gel electrophoresis. For a 600 base pair fragment encoding a 2Q-repeats of the decapeptide, the smaller BamHI fragment will be about 1500 base pairs if the fragment has been inserted in the proper orientation and abou~ 900 base pairs if the fragment has been inserted in the opposite orientation.
The transformants containing the single insert encoding the 20-repeat decapeptide in the proper orientation are grown, in the presence of kanamycin ~10 ~ g/ml.)~ in the following fermentation broth: tryptone (33 9/l.), yeast ex~ract (20 g/l.), ~aCl (7.4 g/lo)~ 3M
NaOH (12 ml/l.) Na2~PO4 (8 g/l.~, KH2PO4 (4 g/1.), casamino a~ids (20 9/1.), glucose (10 9/l.) and MnC12 (O.06 m~). The production and secretion of the 20-repeat decapeptide fusion protein is detected immunolgically.
The fu~ion protein is recovered by conventional protein recovery techniques and treated with cyanogen bromide to separate the bioadhesive precursor protein from the - -amylase fragment and the linker-derived peptide. The bioadhesive precursor protein i5 then isolated by \ conventional procedures.

EX~MPLE _ Production in B~ subtilis of N~2-~Ala-Lys-Pro-Ser-'ryr Pro-Pro-Thr-Tyr-Lys)20-COO~ fused directly _ toa -amyla~e_~ignal pe~tide ~he plasmid containing the 20-decapeptide coding sequence produced according to the procedure of Example II is dige~ted with XbaI and partially digested with BamHI. The XbaI-~am~I fragment contaîning the dsDNA
insert is cloned into the phage vector U13mpl8 from P-~

,,$,~ .
.~

nzo Biochemicals Div., Pharmacia, Inc. Using the procedure for oligonucleotide-directed in vitro mutagenesis of Zoller &
Smith ~Nucleic Acids Res., 10:6487-6500 [1982]), the entire DN~ sequence between the end of the ~-amylase signal sequence and the beginning of the decapeptide-encoding sequence is deleted. The exact oligodeoxyribonucleotide which i5 used in this procedure depends on which of the five double-stranded oligedeoxyribonucleotides occurs first in the sequence of the dsDNA insert. For example, if the first decapeptide-encoding oligodeoxyribonucleotide listed in Example I
occurs first, the following oligodeoxyribonucleotide can be used.

3' End of ~ Oliyodeoxyribo-amylase signal nucleotide #l After isolating the deletion mutant in which the 20-repeat decapeptide sequence is fused directly to the amylase signal sequence, the RFI DNA i8 isolated as described by Zoller and Smith, su~ra. The PvuII-BamHI
fragment i9 ligated to pGx2509 which has been digested with PvuII and BamHI to produce a circularized plasmid \ which codes for the 20-repeat decapeptide fused directly to the ~-amylase signal sequence. The plasmid is used to transform B. subtilis strain lS53 by the procedure of S.
Chang and S.N. Cohen, supra. If the transformants secrete the encoded polypeptide into the growth medium with concomitant cleavage of the signal peptide, the desired bioadhesive precursor protein can be recovered by conventional techniques from the growth medium without the need for chemical or enzymatic cleavage to remove bacterial protein sequences from the N-terminus.

,..~

EXAMPLE IV
Production in S. cerevisiae of NH2-~Ala-Lys-Pro-~ ~3~
For expression in S. erevisiae, the following double~stranded oligodeoxyribonucleotide ~equence is used in preparing the dsDNA insert eneoding the bioadbesive precursor protein.
GCTAAGCCATCTTACCCACCAACCTACAAG
GGTA~AATGGGTGGTTGGATGTTCCGATTC
The automated DNA ~ynthesizer is used to synthesize the following oligodeoxyribonucleotides:
A 5' GCT AAG CCA TCT TAC CCA CCA ACC TAC AAG
B 5' CTT AGC CTT G~A GGT TGG TGG GTA AGA TGG
: C 5' GAA ~TC ~TC GAC AT~
D 5' CTT AGC CAT GTC GAC GAA TTC
E 5' GCT AAG ~AA ~CT TGG ATC C

Oligodeoxyribonucleotides A, B, D and E are phosphorylated at the 51 ends by treatment with adenosine triphosphate in the presence of polynucleotide kinase.
The oligodeoxyribonucleotides are annealed and ligated in the presence of T4 DNA ligase at a ratio of 4A: 4B: 1C:ID:1E:1F to produce a dsDNA insert comprisi~g a 5' iinker (C and D), a 3' linker (E and F~ and a repeated decapeptide coding sequence (A and B) which contains the preferred codons for expression in S. cerevisiae. Th1e 5' lin~er contains EcoRI and SalI cleavage sites. The 3' linker contain~ HindIII and BamHI cleavage sites. The ligation product is run on a 6~ polyacrylamide gel according to the procedure of Maniatis et al., su~
The band corresponding to a dsDNA having approximately 20 repeats of the decapeptide coding sequence is cut from the gel and dsDNA is electroeluted from the gel.

. .
. ;~
. ~ ' :' -.

~ ~8 The isolated ds~NA fragment is digested with SalI
and HindIII to generate staggered ends. The dsDNA
fragment thus produced is inserted into YpGX1, which is represented in Fig. 3. YpGX1 has unique SalI and HindIII
restriction sites within the tetracycline resistance gene. The plasmid (2 ~g) is digested with SalI and HindIII and the resulting linearized plasmid is ligated wi~h the dsDNA insert in the presence of T4 DNA ligase~
The recircularized plasmid is used to transform ~. coli strain JM109 and the transformants are grown separately on LB-agar plates containing ampicillin and then transferred onto LB~ ampicillin and L~+ tetracycline plates. Plasmids from the ApRTcS colonies are analyzed by restriction endonuclease digestion to identify those plasmids containing the dsDNA insert in the proper orientation.
An S cerevisiae phosphoylycera~e kinase (PGR ) promoter and partial structural gene f~agment i5 isolated from plasmid YpGX60, which is represented in Fig. 4.
YpGX60 is diyested with SalI and the ~2000-base pair fragment corresponding to the PGK promoter and the flrst 229 codons of the s'cructural gene is isolated by gel electrophoresis. The plasmid YpGX1 into which the synthetic dsDNA fragment encoding 20 repeats of the 25 decapeptide has been inserted is digested with SalI. The PGK fragment isolated from plasmid YpGX60 is ligated to the linearized YpGX1 containing the 20 decapeptide coding insert in the presence of T4 DNA ligase. The resulting plasmid is us~d to transform S. cerevisiae YGXC18. The transformants are grown on Y~BD ~ tryptophan plates and screened im~unologically to identify colonies that produce the PGK-adhesive fusion protein. The isolated transformants are inoculated into 2-liter flasks containing YNBD + tryptophan and grown overnight at 30C.
The cells are harvested by centrifugation and lysed in a 2~

French press. The fusion protein is recovered by conventional protein recovery techniques and treated with cyanogen bromide, which cleaves the protein on the carboxyl side of the methionine residue immediately preceding the first decapeptide sequence, to separate ~he bioadhesive precursor protein from the N-terminal fragment of PGK and the linker-derived peptide. The bioadhesive precursor protein is then isolated by conventional procedures.

.

Claims (34)

The embodiments of the invention in which an exclusive property or privilege is claimed are defined as follows:

1. A protein which is capable of being converted to a bioadhesive, said protein comprising a sequence of from about 100 to about 1500 amino acids which comprise from about 20% to about 40% proline residues; from about 10% to about 40% lysine residues; from about 10% to about 40%
tyrosine residues; and from 0% to about 40% amino acid residues other than proline, lysine and tyrosine.

2. A protein as claimed in claim 1, wherein the protein has from 600 to 900 amino acid residues.

3. A protein as claimed in claim 1 or 2 which comprises repeating decapeptide units each of which has an amino acid composition as described in claim 1.

4. A protein as claimed in claim 1 or 2 which comprises repeating decapeptide units of the formula:
-Ala-Lys-Pro-Ser-Tyr-Pro-Pro-Thr-Tur-Lys-.

5. A dsDNA sequence which codes for a protein comprising a sequence of from about 100 to about 1500 amino acids which comprise from about 20% to about 40%
proline residues from about 10% to about 40% lysine residues; from about 10% to about 40% tyrosine residues;
and from 0% to about 40% amino acid residues other than proline, lysine and tyrosine.

6. A dsDNA sequence as claimed in claim 5, wherein the encoded protein has from 600 to 900 amino acid residues.

7. A dsDNA sequence as claimed in claim 5 or 6 which encodes repeating decapeptide units.

8. A dsDNA sequence as claimed in claim 5 or 6 wherein the coding strand of the dsDNA comprises repeated units which are each individually selected from units in which the coding strand has the formula GCN AAR CCN (AGY orTCN) TAY CCN CCN ACN TAY AAR
wherein G, A, T and C represent deoxyribo-nucleotides containing the bases guanine, adenine, thymine and cytosine, respectively; R
represents a deoxyribonucleotide containing guanine or adenine; Y represents a deoxyribonucleotide containing cytosine or thymine; and N represents G, A, T or C.

9. A fusion protein comprising an N-terminal portion of a microbial protein fused to the bioadhesive convertible protein of claim 1.

10. A fusion protein as claimed in claim 9, wherein said microbial protein is an E. coli protein.

11. A fusion protein as claimed in claim 10, wherein said microbial protein is E. coli tryptophan synthetase.

12. A fusion protein as claimed in claim 9, wherein said microbial protein is a bacillus protein.

13. A fusion protein as claimed in claim 12, wherein said microbial protein is B. amyloliquefaciens .alpha.- amylase.

14. A fusion protein as claimed in claim 9, wherein said microbial protein is a S. cerevisiae protein.

15. A fusion protein as claimed in claim 14, wherein the microbial protein is S. cerevisiae phosphoglycerate kinase.

16. A fusion protein as claimed in claim 9, wherein the microbial protein contains a signal peptide which is capable of effecting secretion of the fusion protein through the cell membrane of a bacillus host.

17. A fusion protein as claimed in claim 9, wherein the microbial protein fragment is fused to the bioadhesive convertible protein through an enzymatically or chemically cleavable linker sequence of amino acids which does not occur within the amino acid sequence of the bioadhesive convertible protein.

18. A fusion protein as claimed in claim 17, wherein said cleavable linker sequence is a methionine residue.

19. A replicable expression vector comprising a plasmid which contains a transcription promoter, ribosome binding site and translation initiation signal operably linked to a dsDNA sequence encoding a bioadhesive precursor protein comprising a sequence of from about 100 to about 1500 amino acids which comprise from about 20% to about 40% proline residues; from about 10% to about 40%
lysine residues; and from about 10% to about 40% tyrosine residues; and from 0% proline residues; to about 40% amino acid residues other than proline, lysine and tyrosine.

20. A replicable expression vector as claimed in claim 19, further comprising a dsDNA sequence, attached to the 5' end of the sequence encoding the bioadhesive precursor protein, said dsDNA sequence encoding an N-terminal portion of a microbial protein.

21. A replicable expression vector as claimed in claim 20, wherein the sequence encoding the N-terminal portion of the microbial protein is joined to the sequence encoding the bioadhesive precursor protein through a dsDNA
sequence encoding an enzymatically or chemically cleavable linker sequence of amino acids.

22. A replicable expression vector as claimed in claim 21, wherein the sequence encoding the cleavable linker is ATG.

23. A replicable expression vector as claimed in claim 19, 20 or 21, wherein the encoded bioadhesive precursor protein contains from 600 to 900 amino acid residues.

24. A replicable expression vector as claimed in claim 19, 20 or 21, wherein the dsDNA sequence encoding the bioadhesive precursor protein comprises repeating decapeptide units.

25. A replicable expression vector as claimed in claim 19, 20 or 21, wherein the dsDNA sequence encoding the bioadhesive precursor protein comprises repeating units each, individually, having a coding strand of the formula GCN AAR CCN (AGY or TCN ) TAY CCN CCN ACN TAY AAR
wherein G, A, T and C represent deoxyribonuc-leotide containing the bases guanine, adenine, thymine and cytosine, respectively; R repre-sents a deoxyribonucleotide containing guanine or adenine; Y represents a deoxyribonucleotide containing cytosine or thymine; and N represents G, A, T or C.

26. A method of producing a bioadhesive precursor protein which comprises:
(a) chemically synthesizing a dsDNA segment encoding a bioadhesive precursor protein of from about 100 to 1500 amino acid residues which comprise from about 20% to about 40% proline residues; from about 10%
to about 40% lysine residues; from about-10% to about 40% tyrosine residues; and from 0% to about 40% amino acid residues other than proline, lysine and tyrosine;

(b) inserting the chemically synthesized dsDNA segment into a replicable expression vector in which it is operably linked to a transcription promoter, ribosome binding site and translation initiation signal;
(c) transforming a host microorganism with the replicable expression vector containing the chemically synthesized dsDNA insert;
and (d) expressing the bioadhesive precursor protein in the transformant microorganism.

27. A method as claimed in claim 26, wherein the protein encoded by the chemically synthesized dsDNA
contains from 600 to 900 amino acid residues.

28. A method as claimed in claim 26, wherein the chemically synthesized dsDNA segment further comprises a dsDNA seguence, attached to the 5' end of the sequence encoding the bioadhesive precursor protein, said dsDNA
sequence encoding an N-terminal portion of a microbial protein.

29. A method as claimed in claim 28, wherein the sequence encoding the N-terminal portion of the microbial protein is joined to the sequence encoding the bioadhesive precursor protein through a dsDNA sequence encoding an enzymatically or chemically cleavable linker sequence of amino acids.

30. A method as claimed in claim 29, wherein the sequence encoding the cleavable linker is ATG.

31. A method as claimed in claim 28, wherein the dsDNA sequence encoding the N-terminal portion of the microbial protein includes a sequence encoding a signal peptide which is capable of effecting secretion of the expressed protein through the cell membrane of a bacillus host.

32. A method as claimed in claim 26, wherein the host microorganism is selected from E. coli, B. subtilis and S. cerevisiae.

33. A method as claimed in claim 26, wherein the dsDNA segment encoding the bioadhesive precursor protein is comprised of repeating decapeptides each of which has an amino acid composition as described in claim 1.

34. A method as claimed in claim 26, wherein the dsDNA segment encoding the bioadhesive precursor protein comprises repeating units each, individually, having a coding strand of the formula.
GCN AAR CCN (AGY or TCN ) TAY CCN CCN ACN TAY AAR
wherein G, A, T and C represent deoxyribo-nucleotide containing the bases guanine, adenine, thymine and cytosine, respectively;
R represents a deoxyribonucleotide containing guanine or adenine; Y represents a deoxyribo-nucleotide containing cytosine or thymine; and N represents G, A, T or C.