METHODS OF PRODUCTION OF AN AMIDATED PEPTIDE THROUGH THE USE OF A FUSION PROTEIN

METHODS OF PRODUCTION OF AN AMIDATED PEPTIDE THROUGH THE USE OF A FUSION PROTEIN

The present invention is directed to the production of peptides, especially but not exclusively with carboxy-terminal modifications such as amidation, by recombinant means. 5

"Peptide" is a term loosely applied to a chain of amino acids, arbitrarily applied to sequences of three to over one hundred components, but possibly more, joined via their amino- and carboxy teraiini. There are many examples of naturally occurring peptides which function as hormones, messengers, growth factors, antimicrobials, 0 surfactants etc and a wide variety of medicinal and other applications can be envisaged.

Currently, there are at least three major sources of peptides, extraction from natural sources, chemical synthesis and from organisms transformed with recombinant 5 DNA constructs. The advantage of the route using transformed organisms is the biological fidelity of the synthetic process, the ability to synthesise chemically unfavourable sequences, the avoidance of chemical processes, using solvents, etc and, especially with longer peptides, cost-effectiveness.

0 The disadvantage of making peptides by recombinant technology is that the organisms used tend to be poor at synthesising, and if necessary secreting, short sequences of amino acids. Therefore, many of the methods considered for industrial use take advantage of fusion proteins where the short peptide sequence is made as either an amino- or a carboxy-terminal extension on another protein. 5 Although these fusion proteins can be produced in greater quantities, and often purified by exploiting special characteristics of the fusion partner to simplify purification, difficulties can be experienced in recovering the peptide. Proteins are chemically stable molecules and therefore require specific cleavage strategies in order to recover integral peptides with defined amino- and carboxy termini.

A wide panel of protein cleavage technologies can be envisaged. These range from chemical cleavage at specific amino acids to enzymatic cleavage using sequence-specific enzymes. Examples of chemical cleavage include cyanogen bromide cleavage after methionine residues and hydroxylamine cleavage between the amino acid pair asparagine - glycine. Examples of enzymes suitable for cutting at specific protein sequences include enterokinase, which cuts after the sequence (aspartic acid)₄ -lysine, and thrombin, which cuts after the basic amino acids lysine or arginine.

A common problem with both of these cleavage strategies is that sequence constraints operate on both the presence of internal sites within the peptide and the necessity to generate authentic ammo-termini. For example, cyanogen bromide is only useful when there are no internal methionines in the peptide and thrombin can cut at a number of different sites after basic amino acids. Enzymatic cleavage has additional problems in terms of process economics. The enzyme must come from an acceptable and validated source (a common source of enterokinase is calf gut endothelium) and be available in economically acceptable quantity.

Carboxy-terminal amidation is a common post-translational modification found on many biologically active peptides of potential commercial interest. Examples include calcitonin, magainin and etc. . In many instances, for example, calcitonin, the natural amidated peptide is nearly two thousand times as active as the non-amidated version.

There are many different chemical and biological methods designed to produce carboxy-terminal amidated peptides. However, as with any extra process step, each has disadvantages in terms of adding to the overall cost of the finished product.

This invention describes a method for the production of peptides as ammo-terminal extensions of fusion proteins in recombinant systems. We provide novel methods whereby cleavage of the peptide from the fusion protein and modifications of the peptide such as carboxy-amidation can occur as a series of linked reactions in a single process. Such an approach benefits from the low cost and fidelity of synthesis in a biological expression system without the disadvantages posed by the necessity of a separate cleavage step.

Thus, in a first aspect, the present invention provides a method for the production of a peptide which comprises the step of expressing the peptide as part of a fusion protein followed by release of the peptide from the fusion protein by an acyl- acceptor such as a sulphur containing reductant. Suitably, at least part of the fusion protein is a molecule capable of catalysing transfer of the peptide, as an acyl moiety, to a suitable acceptor such as a proximal sulphur atom to form the thio-ester.

In preferred embodiments the peptide is chemically modified, eg amidated at its carboxy terminus after release from the fusion protein. Suitably, the amidation step is carried out in the presence of a source of ammonium ions at a suitable pH and the amidation step occurs simultaneously with release of the peptide. Examples of amidated peptides which could be prepared using these methods include Salmon Calcitonin, Human Calcitonin, Lutenising hormone releasing hormone, Oxytocin, Gastrin neuropeptide Y, Vasopressin, Corticotrophin releasing hormone, Growth hormone releasing hormone, Human Calcitonin gene related peptide, Gastrin, D- tyr-trp-gly, phe-gly-phe-gly, gly-phe-gly, Melanocyte stimulation hormone precursor, Sectetin, Thyrotrophin releasing hormone, Amylin, Substance P, Pancreatic polypeptide, Cholecystokinin, Gastrin secretion factor, phe-his-ile, phe- tyr-tyr, Savagin, Mastoparin M, Caerulein and FMRF amide.

However, it is also possible to perform simple hydrolysis of the peptide in the absence of ammonia which would result in the formation of a free carboxylic acid terminal group. This makes the methods of the present invention suitable for the commercial production of peptides with free carboxy-termini for medicinal or other applications. Examples of such peptides include Hirulog, Magainin, thymosin alpha- 1, brain naturetic peptide, atrial naturetic peptide or bactericidal/permeability-increasing protein.

The methods of the present invention can for example utilise a commercially available expression vector designed for making proteins as fusion proteins. This vector incorporates a modified self-splicing protein, an intein, making it possible to liberate the protein from its fusion partner by a simple chemical reaction. The invention utilises modified chemical conditions/steps to result in cleavage of the fusion protein thereby liberating a desired peptide, which can be modified e.g. by ambition at the carboxy-terminus.

Inteins are proteins which are expressed with flanking protein sequences at both amino- and carboy-teπriini. The amino- and carboxy-terminal sequences have been named exteins in keeping with the DNA nomenclature of exons and introns. A seemingly typical member of the emerging family of inteins is the VMA1 gene product from yeast. This is approximately 50kDa in molecular mass and contains essential amino acids at the amino terminal (Cysteine) and at the carboxy-terminal (histamine and asparagine). In addition, the carboxy-terminal extein must start with a cysteine. At some point after translation is completed, the ammo-terminal peptide bond is broken and the extein transferred to the sulphur atom of the adjacent cysteine to form a thio-ester. This bond is then exchanged with the cysteine at the start of the carboxy-terminal extein and then, with participation of the adjacent asparagine, exchanged with the peptide bond at this end of the intein. The overall effect of these conceited reactions is that the two exteins are seamlessly joined and the intein is released.

A detailed understanding of these reactions has emerged following the analysis of a series of mutants where essential groups at either end of the intein and the proximal ends of the exteins have been systematically replaced. This knowledge has enabled the design of mutant inteins where the ammo-terminal extein can be replaced by any other protein and the self-splicing function has been disabled. However, cleavage of the resulting fusion protein is still possible by the addition of extraneous chemical agents such as the reductant dithiothreitol. The fusion protein is liberated as a thio-ester with the added reductant which gradually hydrolyses to the free acid in solution.

Calcitonin is an example of a medically and commercially important peptide suitable for manufacture using the methods described in this invention. It contains thirty-two amino acids and is amidated at the carboxy-terminus. The functional activity and amino acid sequence is highly conserved between species. Thus salmon Calcitonin , which was originally obtained mostly from natural sources but is now made by direct synthesis, is in widespread clinical use. In the past, therapies have focused on Paget's disease and hypocalcaemic shock. However, recently there has been a demand for larger amounts of material to treat osteoporosis in post-menopausal women. This application requires substantive quantities of material which makes the cost of production an increasingly important factor.

In order to make Calcitonin using the intein vector it is necessary to prepare complimentary oligonucleotides which encode the Calcitonin sequence flanked by restriction sites designed for insertion at the appropriate site 5' to the modified intein. These sites must be chosen so that the coding sequence of the peptide is in the same coding frame as the rest of the expressed protein. Suitable oligonucleotides can be made by any number of methods, known to those skilled in the art, including most obviously direct synthesis and polymerase chain reaction amplification from a natural sequence using primers designed to contain convenient restriction sites. This DNA construct is then transformed into a suitable expression system and the resulting fusion protein harvested.

In a further refinement of the system the fusion protein also comprises a label, which allows for identification and/or purification of the fusion protein, and thus the peptide, by affinity or other chromatographic methods. Examples of a suitable label include a specific chitin-binding domain, or part thereof, a repeat of acidic or basic amino acids, a poly-histidine sequence, glutathione S transferase and lysozyme. For example, the carboxy-teπninus of an intein can be fused with a specific chitin-binding domain. This binds tightly to a packed column of chitin beads and can be used for the affinity-purification of the intact fusion protein. After extensive washing, the column can then be treated with an appropriate cleavage reagent and the liberated target peptide eluted.

Any expression system which can operate on a commercial scale is suitable although the intein based vector described above is designed for use in E. coli. Other vectors can be designed for optimal use in a particular expression system. For example, if a mammalian expression system was chosen, then protein-encoding regions should have optimised codon usage for that particular system. Expression could also be improved by use of a smaller affinity tag for identification and/or purification such as a repeat of acidic or basic amino acids as described above, to permit resolution from contaminating proteins by ion-exchange chromatography or by the inclusion of a poly-histidine sequence for purification on a metal chelate matrix. A further modification which could improve secretion from a mammalian system (the current E. coli vector is designed for intracellular protein production) would be to add a secretory leader sequence to the calcitonin to promote secretion into the media or into the milk of transgenic animals. Suitably, such a leader sequence should be removed during the secretory process by natural processing enzymes.

Examples of expression systems which could be used to express peptide fusion proteins include bacteria (E.coli, B.subtilis etc.), yeast (S. cerevisiae, P.pastoralis etc.), insect cells (S. frugiperda), mammalian expression systems (Chinese hamster ovary, baby hamster kidney etc.), transgenic mammalian expression in milk or other body fluids (preferably pig, cow, sheep, goat, rabbit etc) and plants (potato, corn, etc). In the case of an E.coli expression system, the initiator methionine will be retained in the expression product. Thus, where the peptide of interest is one which does not include an additional methionine in its sequence, this initiator methionine can be removed using cyanogen bromide. One example of such a peptide is Calcitonin.

Expression could be optimised for any of these systems, and for intracellular or extracellular production, by the appropriate selection of leader sequence, codon usage, intein or mutant thereof, and purification strategy. The skilled person will appreciate that this invention is not tied to any particular manifestation of intein or any species as a source. For instance, it may not be necessary to use a whole intein molecule, much of the sequence may be irrelevant to the desired process and perhaps most of the molecule is functionally unnecessary. Indeed, other proteins outside the definition of "intein" may be capable of transferring the peptide bond at the carboxy-terminus of the target peptide to an appropriate thiol group thus creating the thio-ester group which is necessary for cleavage with concomitant amidation.

Thio-esters are relatively reactive, chemical groups-, compared to either peptide bonds or oxygen-esters, and are therefore readily converted to amides under mild reactive conditions. There are two points in the normal cleavage and release pathway during which the fused peptide can be converted to a carboxy-terminal amide. The first and probably most suitable point is after the peptide has been released from the fusion partner by the addition of a thiol reagent. The preferred reagent is dithiothreitol but any number of sulphur-containing reductants could also function effectively. This reaction is essentially a thiol-interchange reaction where the thiol-ester formed between the carboxy-terminus of the peptide and the sulphur of the intein cysteine is transferred to one of the dithiothreitol sulphur atoms. As with any chemical reaction, the acyl shift reaction, between the amine of the cysteine at the an o-terminus of the intein and the sulphur of the same amino acid residue, is an equilibrium. With the yeast intein described above this equilibrium is shifted in favour of the amine group and the thio-ester is a minor component.

The added thiol reagent removes this thio-ester species and therefore drives the reaction in the direction of making more thio-ester until effectively all of the peptide is released as free thio-ester. The released thio-ester is relatively stable to hydrolysis by water (which would generate the unwanted free acid) and is thus suitable for cleavage by any chemical conditions which will promote amide formation.

The second point where the peptide exhibits a thio-ester is to the intein itself but as described above, this species is a minor component. However, even here it would be possible to design chemical conditions to allow simultaneous release of the peptide as an amidated species.

The conditions expected to favour thio-ester cleavage with concomitant amidation are many and the following is only meant to illustrate a representitive selection of the possible reagents and reactions. Chemically, amides can be formed by the cleavage of thio-esters with ammonia and related compounds. This requires conditions where the positive charge of the carbonyl is enhanced (which is an effect of the adjacent sulphur atom) and the lone pair of electrons on the nitrogen of ammonia are available. In aqueous solution the positively-charged ammonium ion, provided by a salt such as ammonium phosphate or sulphate, is in equilibrium with uncharged ammonia, the reactive species, and the concentration of free ammonia is thus increased with a lowering of the hydrogen ion concentration. It is therefore expected that the reaction promoting the formation of the amide product, although likely to proceed at relatively low pH values, for example pH 4.0 to 6.0, will occur more rapidly as the pH is increased in the range 6.0 to 9.0 or even 10.0, where the equilibrium is shifted significantly in favour of ammonia formation.

The optimal range will be a compromise between the highest pH which will be tolerated by the peptide substrate itself and the lowest pH whereby the reaction still proceeds at an acceptable rate. This optimum range will be deterrnined by the sequence of the peptide itself and other factors relating to the properties of the fusion partner and to process-related, especially purification, issues. Similar conditions and constraints are likely to apply whether the cleavage/amidation reactions occur simultaneously or sequentially. There are many other chemical conditions which can be envisaged by those skilled in the art, both in aqueous and non-aqueous systems, which could achieve the desired reactions. The above is meant simply as an illustration of a suitable method and is not intended to exclude these other possible approaches.

The invention will now be described by way of the following examples, which should not be construed as limiting the invention in any way.

Example 1 : Glycine Extended Salmon Calcitonin

1.1. Cloning Strategy

Vector pCYBl, obtainable from New England Biolabs, containing a Ndel site for translation initiation and a Sapl site directly adjacent to the intein, was used to clone and express glycine extended salmon calcitonin (sCT-G). The sCT-G coding sequence was synthesised as two complementary single stranded oligonucleotides of 103bases and 104bases. The codon usage was optimised for expression in E.coli. Annealing of the two strands produced 5' overhangs complementary to the Ndel (5' end) and the Sapl site (3' end). The double stranded oligonucleotide was inserted into pCYBl digested with Ndel and Sapl. The expression of the fusion gene is under the control of the P^ promoter and is regulated by IPTG due to the presence of a lacl^q gene on the vector.

1.2. Fusion Protein Expression and Analysis

The pCYBl vector containing sCT-G was transfected into DH5-α, cells grown, induced with IPTG, harvested and lysed by sonication. Expressed fusion was captured on chitin agarose which was washed and then boiled in SDS-PAGE sample buffer. The supernatant was run on 16% SDS-PAGE gels and the protein visualised with coomassie stain or electroblotted to PVDF membrane for N- terminal sequencing. The sequence analysis indicated that the sCT-G was N- terminally truncated at two positions; Ser2 and Thr6. 1.3. Fusion Protein Cleavage and Peptide Amidation

Chitin agarose bound fusion was washed with 20mM Hepes pH 8.0, 40mM DTT (cleavage buffer A) or with cleavage buffer A supplemented with 3.0M ammonium bicarbonate (cleavage buffer B) and incubated at 4°C overnight. Released sCT-G was washed from the column and captured on a cation exchange resin then eluted with a salt step. C18 RP-HPLC analysis after digestion with trypsin showed that the product formed with cleavage buffer B contained greater than 90 % amidated C-terminus while the product with cleavage buffer A had a mixture of carboxylic C-terminus and an adduct extended by a single Cys residue, presumably from the intein N-terminus (Figure 1).

Example 2 : Leutenizing Hormone Releasing Hormone (LHRH)

2.1 Cloning Strategy

Cloning was preformed exactly as described for sCT-G except that the oligonucleotides contained the LHRH coding sequence (Tan, L. and Rousseau, P. Biochem. Biophys. Res. Com. 109: 1061-1071 (1982)).

2.2. Fusion Protein Expression and Analysis

As described for sCT-G. N-terminal sequencing demonstratred the LHRH was extended at the N-terminus by a single Met residue, retained from the E.coli initiation signal.

2.3. Fusion Protein Cleavage and Peptide Amidation

The LHRH fusion was treated in the same manner as the sCT-G fusion until the final cation capture step. The column wash was applied directly to an elecrospray mass spectrometer and the data reconstructed to give the mass of the parent ion (Figure 2). LHRH from cleavage buffer B (as described in example 1) resulted in a parent ion with a mass of 133 IDa consitent with the Met extended, amidated molecule. LHRH from cleavage buffer A (as described in example 1) gave a parent ion mass of 1332 Da consistent with the Met extended free acid. The difference of IDa is the expected mass difference between an amide and carboxylic acid.

EXAMPLE 3 Cloning of Human Amylin

The IMPACT I (Intein Mediated Purification with an Affinity Chitin-binding Tag) protein purification system from New England Biolabs (NEB) offers 4 E. coli expression vectors, which differ in their available cloning sites. Human Amylin is cloned using the NEB vector pCYBl, which contains a Ndel site for translation initiation and a Sapl site directly adjacent to the intein.

The Human Amylin sequence is synthesised as two complementary single stranded oligo nucleotides of 115 and 116 bases respectively. The codon usage is optimised for expression in E. coli. Annealing of the two strands produces 5' overhangs complementary to the Ndel (5' end) and the Sapl site (3' end). The double stranded oligo nucleotide can be inserted directly into pCYBl which has previously been digested with both Ndel and Sapl.

PCYBl:

5' CATATGGCTAGC GGC TCT TCC TGC TTT 3' 3' GTATACCGATCG CCG AGA AGG AGC AAA 5'

Ndel Sapl

Human Amylin:

Ndel BspMI

CATATGAAATGCAACACCGCGACCTGCGCGACCCAGCGCCTGGCG GTATACTTTACGTTGTGGCGCTGGACGCGCTGGGTCGCGGACCGC

Mbol Ddel AACTTCCTGGTGCATAGCAGCAACAACTTCGGCGCGATCCTGAGC TTGAAGGACCACGTATCGTCGTTGTTGAAGCCGCGCTAGGACTCG

AGCACCAACTGGGCAGCAACACCTATTGCTTT TCGTGGTTGACCCGTCGTTGTGGATAACGAAA

EXAMPLE 4 General Protcol for expressing peptide fusion proteins in E. coli and purification of fusion protein and released peptide

Expression in bacteria requires transformation of cells with an expression construct using any one of a range of standard methods (Maniatis et al, supra). After cell growth, it is usual to induce expression of the target fusion protein using a combination of an inducible promoter, for example the β-galactosidase promoter, and a small molecule inducer such as IPTG. The fusion protein is then recovered after cell harvesting and breakage and then purified by affinity chromatography. Most usually, this involves passing the clarified cell lysate through a column of an appropriate affinity matrix displaying a ligand to which the fusion protein binds. Contaminants are then washed from the matrix before either specific elution of the fusion protein or cleavage of the bound fusion protein in situ. For instance, with the Impact vector described in Examples 2 and 3, the fusion protein containing lysozyme would be purified by cation exchange chromatography. In this case, cleavage in situ is probably not an option, unless cleavage conditions can be found which do not promote elution of the fusion protein. Under these circumstances, cleavage in solution phase would be required. Cleavage of the fusion protein whilst bound to a matrix simplifies the subsequent purification of the peptide.

Cleavage of the fusion protein can be done by the direct addition of a thiol acyl- acceptor, such as lOmM DTT, to yield a thioester intermediate, which can subsequently be converted to the amide by treatment with ammonia salts at a pH above 6.0. Simultaneous cleavage and conversion t an amide may also be possible with the addition of a suitable mixture of acceptor thiol and ammonia salt.

Released peptide is then further purified, if necessary, using conventional techniques such as solvent partitioning and HPLC.