The invention relates to a method wherein a peptide fragment coupling is carried out enzymatically in the presence of a ligase to synthesise a peptide comprising the sequence His-X-Glu-Gly-Thr-Phe-Thr-Ser-Asp-Val-Ser-

Ser-Tyr-Leu-Glu-Gly-Gln-Ala-Ala-Y-Glu-Phe-lle-Ala-Trp-Leu -Val-Z-Gly-Arg-Gly.

Several peptides comprising the amino acid sequence H-His-X-Glu-Gly-Thr-Phe-Thr-Ser-Asp-Val-Ser- Ser-Tyr-Leu-Glu-Gly-GIn-Ala-Ala-Y-Glu-Phe-lle-Ala-Trp-Leu-Va l-Z-Gly-Arg-Gly-OH are well known in the art as insulinotropic peptides. These peptides include GLP-1 , Liraglutide and Semaglutide.

Human GLP-1 (Glucagon-like peptide-1) has the formula H-His-Ala-Glu-Gly-Thr-Phe-Thr-Ser-Asp-Val- Ser-Ser-Tyr-Leu-Glu-Gly-GIn-Ala-Ala-Lys-Glu-Phe-lle-Ala-Trp- Leu-Val-Lys-Gly-Arg-Gly-OH.

Liraglutide is an Arg ²⁰ -GLP-1 analogue substituted on the e-amino group of the lysine in position 20 of the above sequence with a Glu-spaced palmitic acid. Thus, Liraglutide has the formula H-His-Ala-Glu-Gly-Thr-Phe- Thr-Ser-Asp-Val-Ser-Ser-Tyr-Leu-Glu-Gly-Gln-Ala-Ala-Lys(Pal- y-Glu)-Glu-Phe-lle-Ala-Trp-Leu-Val-Arg-Gly-Arg-Gly- OH (see also Figure 1 , all chiral amino acid residues are L-amino acid residues). In Lys(Pal-y-Glu) the e-amino-group of the Lys residue is linked with the g-Glu carboxylic side-chain and the Glu is N-pal m itoylated .

Semaglutide has the formula H-His-Aib-Glu-Gly-Thr-Phe-Thr-Ser-Asp-Val-Ser-Ser-Tyr-Leu-Gl u-Gly- Gln-Ala-Ala-Lys(AEEA-AEEA-y-Glu-17-carboxyheptadecanoyl)-Glu -Phe-lle-Ala-Trp-Leu-Val-Arg-Gly-Arg-Gly-OH. Herein AEEA-AEEA-y-Glu-17-carboxyheptadecanoyl is N-(17-carboxy-l-oxoheptadecyl)-L-y-glutamyl-2-[2-(2- aminoethoxy)ethoxy]acetyl-2-[2-(2-aminoethoxy)ethoxy]acetyl (see also Figure 2, all chiral amino acid residues are L- amino acid residues).

These peptides can, e.g., be used in the treatment of diabetes II. Further, e.g., Liraglutide can be used in the treatment of obesity, as injectable adjunct to a reduced-calorie diet and increased physical activity for chronic weight management in adult patients.

Processes for synthesizing peptides, including oligopeptides like GLP-1 , Liraglutide and Semaglutide, are known in the art. Methods to synthesise insulinotropic peptides, such as GLP-1 and analogues thereof are described in WO2007147816 and in WO2016/046753. In the ‘BACKGROUND OF THE INVENTION’ of WO2016/046753 a detailed description is given of suitable preparation methods, notably recombinant methodology, sequential synthesis on a solid support , solid phase synthesis of Liraglutide involving coupling a peptide sequence containing amino acid residues (1-10) to a sequence containing amino acid residues (11-31), or solid phase synthesis of Liraglutide involving the preparation of peptide sequences containing amino acid residues (1-4), (15-16) and (17- SI), coupling the peptides containing amino acid residues (15-16) with (17-31) and sequential addition of amino acids before coupling with the peptide containing amino acid sequence (1-4).

In accordance with WO2016/046753 GLP-1 peptides are prepared in a process comprising liquid or solid phase peptide synthesis or a combination thereof, wherein the process comprises a final coupling step in which at least two fragments are coupled at a terminal Gly residue, and wherein at least one of the fragments is prepared by coupling of at least two sub-fragments. Liraglutide is in particular obtained by coupling His-Ala-Glu-Gly and Thr-Phe-Thr-Ser- Asp-Val-Ser-Ser-Tyr-Leu-Glu-Gly-Gln-Ala-Ala-Lys(Pal-Glu-OX)- Glu-Phe-lle-Ala-Trp-Leu-Val-Arg-Gly-Arg-Gly-OH. In this sequence X represents H or a protecting group for the Glu a-carboxylic acid group.

As follows from‘BACKGROUND OF THE INVENTION’ of WO2016/046753 there remains a need for discovering new methods for the synthesis of GLP-1 proteins such as Liraglutide or Semaglutide to provide a better, more efficient and/or cheaper process or to provide a product which can be more readily purified in order to achieve a product with improved yield and purity. In particular, it expresses the need to provide a method for preparing GLP-1 and analogues, such as Liraglutide or Semaglutide, especially on an industrial scale, which should not require the use of toxic or otherwise undesirable reagents in good yields and which can be readily purified to obtain a product having high purity.

An enzymatic synthesis of GLP-1 or an analogue thereof, like Liraglutide or Semaglutide is not suggested in WO2007147816 and in WO2016/046753, which both focus on chemical synthesis.

However, fully chemical synthesis of peptides has disadvantages, as also discussed in the above cited prior art. Further, peptides longer than 15 amino acids are often very difficult to synthesize on the solid phase due to side reactions. As a consequence, purification is troublesome. Therefore, peptides longer than 10 amino acids are often synthesized by a combination of solid phase synthesis of side-chain protected oligopeptide fragments which are subsequently chemically condensed in solution, e.g. as in a 10 + 10 condensation to make a peptide of 20 amino acids. The major drawback of chemical side-chain protected oligopeptide fragment condensation is that, upon activation of the C-terminal amino acid residue of the acyl donor, racemisation occurs. In contrast, enzyme-catalysed peptide couplings are completely devoid of racemisation and have several other advantages over chemical peptide synthesis such as the absence of side reactions on the side-chain functionalities. For industrial application, an enzymatic peptide synthesis concept based on a kinetic approach, i.e. using an acyl donor C-terminal ester is most attractive (see for instance N. Sewald and H.-D. Jakubke, in:“Peptides: Chemistry and Biology”, 1st reprint, Ed. Wiley-VCH Verlag GmbH, Weinheim 2002).

A problem with enzymatic coupling in aqueous solutions is that the presence of water tends to promote hydrolysis rather than coupling. Some reports have been published on the enzymatic condensation of oligopeptide fragments in aqueous solution (Kumaran et al. Protein Science, 2000, 9, 734; Bjorup et al. Bioorg. Med. Chem. 1998, 6, 891 ; Homandberg et al. Biochemistry, 1981 , 21 , 3387; Komoriya et al. Int. J. Pep. Prot. Res. 1980, 16, 433).

It was found by Wells et al. (US 5,403,737) that the enzymatic condensation of oligopeptides in aqueous solution could be significantly improved by altering the active site of subtilisin BPN’, a subtilisin from B. amyloliquefaciens (SEQ ID NO: 2). When two mutations were introduced, i.e. S221C and P225A, a subtilisin BPN’ variant called subtiligase was obtained having a 500-fold increased synthesis over hydrolysis ratio (S/H ratio) as compared to wild-type subtilisin BPN’. In further experiments, Wells et al. added five additional mutations to subtiligase to make the enzyme more stable (Proc. Natl. Acad. Sci. USA, 1994, 91 , 12544). The new mutant called stabiligase appeared moderately more resistant to sodium dodecasulphate and guanidinium hydrochloride, but hydrolysis was still a major side reaction. In WO 2016/056913 a solution is provided for the undesirably high hydrolytic activity encountered with enzymes like subtiligase or stabiligase when used for (oligo)peptide synthesis in an aqueous environment, by providing subtilisin BPN’ variants or a homologues thereof, with specific mutations. These variants or homologues are in particular suitable to catalyse the synthesis of peptides by coupling a first peptide fragment and a second peptide fragment, wherein the first fragment is a peptide C-terminal ester or thioester and the second fragment is a peptide nucleophile having an N-terminally unprotected amine.

The inventors considered to apply enzymatic fragment condensation for the synthesis of GLP-1 , Liraglutide and Semaglutide starting from peptide fragments mentioned in WO2007147816 or WO2016/046753. Amongst others, the inventors considered to couple the 10-mer peptide having amino acid residues 1-10 to the 21-mer peptide containing amino acid residues 11-31 of Liraglutide, Semaglutide or GLP-1 by enzymatic fragment condensation, with the 10-mer as the (thio)ester and the 21-mer as the nucleophile. For the coupling of the peptide C-terminal (thio)ester having amino acid residues 1-10 to a peptide nucleophile containing amino acid residues 11-31 , the presence of a serine at both PT and P2’ was found to be a disadvantage for the peptide nucleophile. Further possible reasons for a lack of effective coupling could be the presence of a non-hydrophobic amino acid at P4 (threonine) of the peptide C-terminal (thio)ester. Further, the inventors attempted to couple a peptide C-terminal (thio)ester having amino acid residues 1-4 to a peptide nucleophile containing amino acid residues 5-31, without success. The inventors concluded that in particular the presence of histidine at P4 and/or the presence of glycine at P1 of the peptide C-terminal (thio)ester are detrimental to effective coupling. As illustrated in Examples 1 and 2 of present disclosure, it was found that it is possible to prepare these peptides by enzymatic coupling in the presence of a ligase, also in an aqueous reaction medium, but that the yield was unexpectedly low for several processes designed on the basis of scientific considerations, such as the consideration that a ligase like a subtilisin variant or homologue thereof favours coupling of C-terminal peptide (thio)esters that have a hydrophobic amino acid residue at the P4 position (the fourth amino acid from the C-terminal end) of the peptide C-terminal ester or thioester. Amongst others, a method was carried out wherein the peptide with the amino acid sequence of Liraglutide (‘H-Liraglutide-1 -31 -OH‘) was prepared by enzymatically coupling the 13-mer C-terminal ester His-Ala-Glu-Gly-Thr-Phe-Thr-Ser-Asp-Val-Ser-Ser-Tyr-OCam-Leu -OH and the 18-mer peptide nucleophile H-Leu-Glu-Gly-Gln-Ala-Ala-Lys-Glu-Phe-lle-Ala-Trp-Leu-Val-Ar g-Gly-Arg-Gly, in the presence of a Subtilisin BPN’ variant. Although it was expected that the presence of the hydrophobic amino acid residue, Val, at the P4 position of the C-terminal ester would make these fragments particularly good fragments to make the Liraglutide amino acid sequence at a high yield, this peptide was obtained in a very low yield (see Example 1). As illustrated by Example 2, the yield by enzymatic preparation of H-Liraglutide-1 -31 -OH from the corresponding 9-mer C-terminal ester and 22-mer peptide nucleophile was even less although it was expected that the presence of a hydrophobic Phe residue on the P4 position would make these fragments particularly good fragments for an enzymatic condensation reaction.

It is an object of the present invention to provide a novel method of enzymatically synthesizing GLP-1 or an analogue thereof, in particular Liraglutide or Semaglutide. There is a need for alternative enzymatic peptide synthesis processes for these peptides in general, in particular in order to broaden the palette of tools for making them. In particular it is an object to provide such a process that overcomes one or more of the problems mentioned above or discussed in the above cited prior art, more in particular an improved overall yield or an improved selectivity.

One or more other objects that may be subject of the invention follow from the description below.

It has now surprisingly been found that one or more of these objects are met by a method wherein GLP- 1 or an analogue thereof is prepared in a method comprising the enzymatic synthesis of a peptide by fragment condensation, wherein two specific fragments of said peptide are coupled in the presence of a ligase, in particular a subtilisin variant or homologue.

Accordingly, the present invention relates to a method for synthesising a peptide comprising the sequence His-X-Glu-Gly-Thr-Phe-Thr-Ser-Asp-Val-Ser-Ser-Tyr-Leu-Glu-Gl y-Gln-Ala-Ala-Y-Glu-Phe-lle-Ala-Trp-Leu- Val-Z-Gly-Arg-Gly, wherein

X is Ala or an a-amino-isobutyric acid residue (Aib);

Y is Lys, which Lys has a free side-chain e-amino group (i.e. a non-derivatised lysine residue), or of which the side-chain e-amino group is protected with a protective group, or which is functionalised with an amino acid or another functional group;

Z is Arg or Lys;

the method comprising enzymatically coupling

(a) a peptide C-terminal ester or thioester comprising a first peptide fragment comprising the sequence His-X-Glu-Gly-Thr-Phe-Thr-Ser-Asp-Val-Ser-(thio)ester, and

(b) a peptide nucleophile having an N-terminally unprotected amine comprising a second peptide fragment comprising the sequence H-Ser-Tyr-Leu-Glu-Gly-Gln-Ala-Ala-Y-Glu-Phe-lle-Ala- T rp-Leu-Val-Z-Gly-Arg-Gly;

which enzymatic coupling is catalysed by a ligase.

It is in particular surprising that a method according to the invention allows the synthesis of the peptide of interest (the ligation product) in a high yield, also when carried out in an aqueous reaction medium. After all, the Ser at the P4 position of the C-terminal (thio)ester fragment is polar, which from a scientific perspective was expected to be unfavourable for the enzymatic coupling reaction.

This has been accomplished without needing any side-chain protective groups on the peptide fragments, and without needing one or both of the fragments to be provided with a functional group to increase solubility ( e.g . a 2-hydroxy-4- methoxybenzyl amide group on the peptide backbone amide functionality or a peptide-tag of polar amino acids at the terminal ends of the respective fragments not taking part in the coupling reaction), although in a specific embodiment use may be made of protective groups or solubility-enhancing groups. The high S/H ratio without needing a solubilityenhancing group is surprising because the solubility of the peptide nucleophile is low.

In a method according to the invention, the Y of the peptide nucleophile can be a Lys having a free side- chain e-amino group (i.e. a non-derivatised lysine residue). However, the invention also allows the coupling of a peptide nucleophile wherein Y is a Lys of which the side-chain e-amino group comprises a functional group, in particular wherein Y is Lysy-Glu, Lys(AEEA-AEEA-y-Glu, Lys(Pal-y-Glu-OH) or Lys(AEEA-AEEA-y-Glu-17-carboxyheptadecanoyl-OH). It is in particular surprising that efficient enzymatic coupling is also possible using such peptide nucleophiles, despite the presence of such groups that are sterically very demanding. In this respect it is further surprising that a high synthesis over hydrolysis ratio (S/H-ratio) is achieved also in an aqueous reaction medium while the peptide nucleophile has a high hydrophobicity, which is increased by the presence of an hydrophobic group (such as a fatty acid tail) at the side-chain e-amino group of the Lys. Generally, it is the inventors finding that the S/H ratio is directly related to the concentration of the nucleophile and thus to the solubility of the nucleophile in aqueous solution. Since the presence of a hydrophobic functional group renders the solubility of the resulting molecules very low, it is thus very surprising that the S/H ratio’s are still very high, when coupling the (thio)ester to the peptide nucleophile as defined in the present invention, i.e. comprising the 11-mer peptide (thio)ester and the 20-mer peptide nucleophile. Reference methods wherein it was attempted to obtain the same ligation product by coupling a different (thio)ester and peptide nucleophile provided with or without the side-chain functionality at the Y position, e.g. the corresponding 9-mer (thio)ester and 22-mer peptide nucleophile were not successful.

Coupling with a peptide nucleophile wherein Y is a Lys of which the side-chain e-amino group has been functionalised with an amino acid or another functional group has in particular been found possible with subtilisin BPN’ variants, as described in further detail elsewhere herein. Preferred embodiments of methods wherein the coupling is carried out using a peptide nucleophile wherein Y is a Lys of which the side-chain e-amino group is functionalised will also be described in further detail below.

For the purpose of this invention, with“synthesis over hydrolysis ratio” (S/H ratio) is meant the amount of enzymatically synthesised (oligo)peptide product divided by the amount of (oligo)peptide C-terminal ester or thioester of which the ester or thioester group has been hydrolysed. For further details on determining an S/H ratio, reference is made to WO 2016/056913.

The term“or” as used herein is defined as“and/or” unless it is specified otherwise or it follows from the context that it means‘either ....or...’.

The term“a” or“an” as used herein is defined as“at least one” unless it is specified otherwise or it follows from the context that it should refer to the singular only.

When referring to a noun (e.g. a compound, an additive, etc.) in the singular, the plural is meant to be included, unless it follows from the context that it should refer to the singular only.

The term‘pH’ is used herein for the apparent pH, i.e. the pH as measured with a standard, calibrated pH electrode.

For the purpose of this invention, with“peptides” is meant any chain composed of two or more amino acids. Thus, peptides are generally amides at least conceptually composed of two or more amino carboxylic acid molecules (i.e. amino acids) by formation of a covalent bond from the carbonyl carbon of one to the nitrogen atom of another with formal loss of water. The term‘peptide’ is usually applied to structures formed from a-amino acids, although a peptide may comprise other amino acids, such as one or more beta-amino acids and/or one or more y- amino acids. The amino acid sequence of a peptide is referred to as the primary structure. In an embodiment, the peptide is essentially free of a secondary structure and essentially free of a tertiary structure.

In an embodiment, a peptide that has been synthesised or that is to be coupled in a method according to the invention essentially consists of amino acid residues. E.g. GLP-1 consists of amino acid residues. In a further embodiment, the peptide essentially consists of amino acid units and protective groups.

In a further embodiment, a peptide that has been synthesised or that is to be coupled in a method according to the invention is a conjugate of a peptide chain and another residue, such as a fatty acid. These peptides are called lipopeptides. Fatty acids can e.g. be used to change the solubility. Examples of suitable fatty acids, are C8- C24 saturated fatty acids and C8-C24 unsaturated fatty acids. If desired, a polar linker is provided between the peptide and the fatty acid, e.g. to increase the solubility in an aqueous environment. Liraglutide and Semaglutide are peptides that are conjugates of a peptide chain and a fatty acid. Semaglutide comprises a polar linker between the peptide and the fatty acid residue.

Typically, peptides - which term includes oligopeptides, proteins and chimeric peptides - comprise up to about 35 000 amino acid units, in particular 3-20 000, more in particular 4-1000 or 5-500 amino acid units. The ligase according to the invention may be used for the synthesis of other peptides than His-X-Glu-Gly-Thr-Phe-Thr-Ser-Asp- Val-Ser-Ser-Tyr-Leu-Glu-Gly-Gln-Ala-Ala-Y-Glu-Phe-lle-Ala-Tr p-Leu-Val-Z-Gly-Arg-Gly. Such peptides preferably comprise 500 amino acid units or less, in particular 200 or less, more in particular 100 or less. In a specifically preferred embodiment, the synthesised peptide comprises at least 10 amino acid units, more specifically at least 15 amino acids, at least 25 amino acids or at least 40 amino acids. The fragments from which such peptide can be chosen within wide ranges; the length of a fragment can be at least 2, in particular at least 5, more in particular at least 10, with the upper limit determined by the length of the synthesised peptide.

With“oligopeptides” is meant within the context of the invention, a peptide composed of 2-200 amino acid units, in particular composed of 5-100 amino acid units, more in particular composed of 10-50 amino acid units.

For the purpose of this invention, with“peptide bond” is meant the amide bond between (i) either the a- amino terminus of one a-amino acid or the beta-amino terminus of one beta-amino acid and (ii) either the a-carboxyl terminus of one other a-amino acid or the beta-carboxyl terminus of one other beta-amino acid. Preferably, the peptide bond is between the a-amino terminus of one a-amino acid and the a-carboxyl terminus of another a-amino acid.

In the context of the invention with“amino acid side-chain” is meant any proteinogenic or non- proteinogenic amino acid side-chain.

Proteinogenic amino acids are the amino acids that are encoded by the genetic code. Proteinogenic amino acids include: alanine (Ala), valine (Val), leucine (Leu), isoleucine (lie), serine (Ser), threonine (Thr), methionine (Met), cysteine (Cys), asparagine (Asn), glutamine (Gin), tyrosine (Tyr), tryptophan (Trp), glycine (Gly), aspartic acid (Asp), glutamic acid (Glu), histidine (His), lysine (Lys), arginine (Arg), proline (Pro) and phenylalanine (Phe). Selenocysteine (Sec, U) is an amino acid, whose structure corresponds to cysteine, with the proviso that it contains a selenium instead of a sulphur atom. Proteinogenic amino acids are the L-stereo isomers of said amino acids (except for glycine, which does not have a stereo-isomeric form). The non-proteinogenic amino acid of particular interest in a method according to the present invention is 2-aminoisobutyric acid (Aib), which forms part of the peptide chain of Semaglutide.

The term“(thio)ester” is used herein as short-hand for the phrase“ester or thioester”.

The term“W-terminal protection” is used herein to indicate that an W-terminal amine group of a peptide, typically the W-terminal a-amine group, is provided with a protective group, generally at least substantially protecting the W-terminal amine group from being coupled to a C-terminal carboxylic group of another peptide or of the same peptide molecule.

The term“C-terminal protection” is used herein to indicate that a C-terminal carboxylic group of a peptide, typically the C-terminal a-carboxylic group is provided with a protective group, generally substantially protecting the carboxylic group from being coupled to an N-terminal amine group of another peptide or of the same peptide molecule.

The term "mutated" or“mutation” as used herein regarding proteins or polypeptides - in particular enzymes such as ligases - means that at least one amino acid in the wild-type or naturally occurring protein or polypeptide sequence has been replaced with a different amino acid, inserted into, appended to, or deleted from the sequence via mutagenesis of nucleic acids encoding these amino acids. Mutagenesis is a well-known method in the art, and includes, for example, site-directed mutagenesis by means of PCR or via oligonucleotide-mediated mutagenesis as described in Sambrook et al., Molecular Cloning-A Laboratory Manual, 2nd ed., Vol. 1-3 (1989). The term "mutated" or“mutation” as used herein regarding genes means that at least one nucleotide in the nucleic acid sequence of that gene or a regulatory sequence thereof, has been replaced with a different nucleotide, has been inserted into, has been appended to, or has been deleted from the sequence via mutagenesis, resulting in the transcription of a protein sequence with a qualitatively of quantitatively altered function or resulting in the knock-out of that gene.

In the present specification, a shorthand for denoting amino acid substitutions employs the single letter amino acid code of the amino acid that is substituted, followed by the number designating where in the protein amino acid sequence the substitution is made. This number is the amino acid position of the wild-type amino acid sequence. Thus for the mutated amino acid sequence it is the amino acid position corresponding to the position with that number in the wild type enzyme. Due to one or more other mutations at a lower position (additions, insertions, deletions, etc.) the actual position does not need to be the same. The skilled person will be able to determine the corresponding positions using a generally known alignment technique, such as NEEDLE. The number is followed by the single letter code of the amino acid that replaces the wild-type amino acid therein. For example, S221 C denotes the substitution of serine at the position corresponding to position 221 to cysteine. X is used to indicate any other proteinogenic amino acid than the amino acid to be substituted. For example, S221X denotes the substitution of serine at the position corresponding to position 221 to any other proteinogenic amino acid.

The term“ligase” is used herein for an enzyme having catalytic activity in the coupling of two peptides by catalysing the formation of a peptide bond by coupling the C-terminus of a first peptide and the N-terminus of another peptide. Generally, the ligase (used in a method) according to the invention has ligase activity with respect to coupling an 11-mer His-X-Glu-Gly-Thr-Phe-Thr-Ser-Asp-Val-Ser-(thio)ester, and a 20-mer peptide nucleophile H-Ser-Tyr-Leu- Glu-Gly-Gln-Ala-Ala-Y-Glu-Phe-lle-Ala-Trp-Leu-Val-Z-Gly-Arg- Gly.

As defined by Schechter and Berger, the active site residues in proteases, including ligases are composed of contiguous pockets termed subsites. Each subsite pocket binds to a corresponding residue in the peptide substrate sequence, referred to here as the sequence position. According to this definition, amino acid residues in the substrate sequence are consecutively numbered outward from the cleavage sites as.-P4-P3-P2-P1-P1'-P2'-P3'-P4'- .(the scissile bond is located between the P1 and P1' positions), while the subsites (pockets) in the active site are correspondingly labelled as.-S4-S3-S2-S1-S1'-S2'-S3'-S4'-. (Schechter and Berger, Biochem Biophys Res Commun. 1967 Apr 20;27(2): 157-62.)). It should be noted that not all proteases have all of said subsites. E.g. an S3’ and/or an S4’ pocket may be absent in a subtilisin BPN’ variant or homologue thereof according to the invention.

For the purpose of this invention, with“S1 , S2, S3 and S4 pocket” is meant the amino acids of a protease (in particular a ligase) which interact with the amino acids of a peptide acyl donor. The C-terminal amino acid (1 ^st amino acid; P1) of the acyl donor peptide interacts with the amino acids in the S1 pocket of the protease. The penultimate amino acid (2 ^nd amino acid from the C-terminal end; P2) of the acyl donor peptide interacts with the amino acids in the S2 pocket of the protease, the third amino acid (P3) with the S3 and the fourth amino acid (P4) with the S4 pocket. The S1-S4 binding pockets of a protease are defined by several amino acids which can be distant in the primary structure of the protease, but are close in the three dimensional space. For the purpose of this invention, with ST and S2’ pockets are meant the amino acids of a protease which interact with the N-terminal amino acids of a peptide nucleophile. The N-terminal amino acid of the peptide nucleophile interacts with the amino acids in the ST pocket of the protease. The N-terminal penultimate amino acid of the peptide nucleophile interacts with the amino acids in the S2’ pocket of the protease. The ST and S2’ binding pockets of a protease are defined by several amino acids which can be distant in the primary structure of the protease, but are close in the three dimensional space.

When an enzyme is mentioned with reference to an enzyme class (EC) between brackets, the enzyme class is a class wherein the enzyme is classified or may be classified, on the basis of the Enzyme Nomenclature provided by the Nomenclature Committee of the International Union of Biochemistry and Molecular Biology (NC- IUBMB), which nomenclature may be found at http://www.chem.qmul.ac.uk/iubmb/enzyme/. Other suitable enzymes that have not (yet) been classified in a specified class but may be classified as such, are meant to be included.

Homologues typically have an intended function in common with the peptide or enzyme, of which it is a homologue, such as being capable of catalyzing the same reaction, in particular an enzymatic coupling of a method according to the invention.

Amino acid or nucleotide sequences are said to be homologous when exhibiting a certain level of similarity. Whether two homologous sequences are closely related or more distantly related is indicated by“percent identity” or“percent similarity”, which is high or low respectively.

The terms “homology”, “percent homology”, “percent identity” or “percent similarity” are used interchangeably herein. For the purpose of this invention, it is defined here that in order to determine the percent identity of two amino acid sequences, the complete sequences are aligned for optimal comparison purposes. In order to optimise the alignment between the two sequences gaps may be introduced in any of the two sequences that are compared. Such alignment is carried out over the full length of the sequences being compared. Alternatively, the alignment may be carried out over a shorter length, for example over about 20, about 50, about 100 or more nucleic acids or amino acids. The percentage identity is the percentage of identical matches between the two sequences over the reported aligned region.

A comparison of sequences and determination of percent identity between two sequences can be accomplished using a mathematical algorithm. The skilled person will be aware of the fact that several different computer programs are available to align two sequences and determine the homology between two sequences (Kruskal, J. B. (1983) An overview of sequence comparison In D. Sankoff and J. B. Kruskal, (ed.), Time warps, string edits and macromolecules: the theory and practice of sequence comparison, pp. 1-44 Addison Wesley). The percent identity between two amino acid sequences can be determined using the Needleman and Wunsch algorithm for the alignment of two sequences. (Needleman, S. B. and Wunsch, C. D. (1970) J. Mol. Biol. 48, pp 443-453). The Needleman-Wunsch algorithm has been implemented in the computer program NEEDLE. For the purpose of this invention the NEEDLE program from the EMBOSS package was used (version 2.8.0 or higher, EMBOSS: The European Molecular Biology Open Software Suite (2000) Rice,P. LongdenJ. and BleasbyA Trends in Genetics 16, (6) pp 276—277, http://emboss.bioinformatics.nl/). For protein sequences, EBLOSUM62 is used for the substitution matrix. Other matrices can be specified. The optional parameters used for alignment of amino acid sequences are a gap-open penalty of 10 and a gap extension penalty of 0.5. The skilled person will appreciate that all these different parameters will yield slightly different results but that the overall percentage identity of two sequences is not significantly altered when using different algorithms.

The homology or identity between the two aligned sequences is calculated as follows: the number of corresponding positions in the alignment showing an identical amino acid in both sequences divided by the total length of the alignment after subtraction of the total number of gaps in the alignment. The identity defined as herein can be obtained from NEEDLE by using the NOBRIEF option and is labelled in the output of the program as“longest-identity”. For purposes of the invention the level of identity (homology) between two sequences is calculated according to the definition of“longest-identity” as can be carried out by using the program NEEDLE.

The polypeptide sequences, in particular enzyme sequences, can further be used as a“query sequence” to perform a search against sequence databases, for example to identify other family members or related sequences. Such searches can be performed using the BLAST programs. Software for performing BLAST analyses is publicly available through the National Center for Biotechnology Information (http://www.ncbi.nlm.nih.gov). BLASTP is used for amino acid sequences. The BLAST program uses as defaults:

-Cost to open gap: default = 11 for proteins

-Cost to extend gap: default = 1 for proteins

-Expect value: default = 10

-Wordsize: default = 28 for megablast/ 3 for proteins Furthermore the degree of local identity (homology) between the amino acid sequence query and the retrieved homologous sequences is determined by the BLAST program. However only those sequence segments are compared that give a match above a certain threshold. Accordingly the program calculates the identity only for these matching segments. Therefore the identity calculated in this way is referred to as local identity.

The term“homologue” is used herein in particular for peptides, more in particular enzymes, having a sequence identity of at least 50 %, preferably at least 60 %, more preferably at least 70 %, at least 80 %, at least 90 %, at least 95 %, at least 96 %, at least 97 %, at least 98 % or at least 99 % with the peptide, in particular enzyme, with which the homologue peptide or enzyme is compared. Evidently, the sequence identity will be less than 100 %. The percentage of sequence identity will depend on the number of mutations and the length of the peptide (enzyme) with which the homologue is compared. In‘longest identity’ alignment deletions are not taken into account.

For the purpose of this invention, with“condensation” is meant the formation of a new amide bond between the C-terminal carboxylic function of a peptide) with the N-terminal amine function of a nucleophile, in particular another peptide.

The term“analogue” of a peptide is used in particular for peptides that are structural analogues and/or functional analogues of said peptide. Functional analogues have a same in vivo target ( e.g . the same target receptor on a cell membrane); structural analogues have a high similarity in amino acid sequence. Functional analogues of a peptide may have a relatively low amino acid sequence identity, e.g. of about 50 % or less over the full amino acid sequence, yet a high sequence identity (and thus a high structural similarity) with the peptide of which they are an analogue in a segment of the amino acid sequence, such as near the N-terminal part or near the C-terminal part. A structural analogue, in particular comprises an amino acid sequence that has at least 60 %, more in particular at least 70 %, preferably at least 80 %, more preferably at least 90 % sequence identity, more preferably at least 95 % sequence identity with the amino acid sequence of the peptide of which a peptide is an analogue. For the purpose of clarity and a concise description features are described herein as part of the same or separate embodiments, however, it will be appreciated that the scope of the invention may include embodiments having combinations of all or some of the features described. Terms used herein that are not specifically defined herein are as defined in WO 2016/056913, or - if not defined therein - used in accordance with common general knowledge.

The peptide C-terminal ester or thioester comprises a first peptide fragment comprising the amino acid sequence His-X-Glu-Gly-Thr-Phe-Thr-Ser-Asp-Val-Ser-(thio)ester, wherein X is Ala or an a-amino-isobutyric acid unit (Aib). Particularly good results have been achieved with a first peptide fragment, wherein this is the amino acid sequence of the first peptide fragment. In a specific embodiment, the N-terminal end is extended with at least one amino acid (W _v , see below), e.g. Gly or Phe, and the N-terminal end of the extended peptide is provided with a protective group, typically an Edman-type protective group (see also below). If desired, the N-terminal end can be provided with a peptide tag, e.g. to modify solubility in the reaction medium. However, this is generally not required, especially not in an aqueous reaction medium.

The peptide C-terminal ester or thioester typically is an activated (thio)ester, i.e. it contains a carboxy ester or carboxy thioester group that can take part in the enzymatic coupling reaction. In principle, any (substituted or unsubstituted) alkyl or (substituted or unsubstituted) aryl (thio)estercan be used. Typical examples of (thio)esters which can take part in the enzymatic coupling reaction are methyl-, ethyl, propyl-, isopropyl-, phenyl-, benzyl- (such as p- carboxy-benzyl-), 2,2,2-trichloroethyl-, 2,2,2-trifluoroethyl-, cyanomethyl- and carboxyamidomethyl-(thio)esters.

Particularly good results have been obtained with carboxyamidomethyl-type esters (Cam-esters) represented by the formula peptide-(C=0)-0-CXiX ₂ -C(=0)N-RiR ₂ . Herein, each Xi and X2 independently represents a hydrogen atom or an alkyl group. Good results have been achieved when both X1 and X2 are a hydrogen atom (peptide-(C=0)-0-CH ₂ -C(=0)N-RiR ₂ ). Herein R1 represents a hydrogen atom or an alkyl group and R2 represents a hydrogen atom or an alkyl group or an amino acid or a peptide residue with a C-terminal carboxyamide or carboxylic acid functionality, optionally protected on the side-chain functionality of the amino acid or on one or more of the side- chain functionalities of the amino acids. Herein, each alkyl group may independently represent a (substituted or unsubstituted) C1-C7 alkyl group, preferably a (substituted or unsubstituted) linear C1-C6 alkyl group, more preferably a (substituted or unsubstituted) linear C1-C3 alkyl group, and most preferably a methyl group. Good results have in particular been achieved in a method of the invention wherein both R1 and R2 represent a hydrogen atom or wherein R1 represents a hydrogen atom and R2 represents an amino acid or peptide residue with a C-terminal carboxyamide or carboxylic acid functionality, optionally protected on the side-chain functionality of the amino acid or on one or more of the side-chain functionalities of the amino acids.

It is especially advantageous to use a Cam-AA1-AA2- ester, wherein AA1 is a first amino acid residue and AA2 is a second amino acid residue. Herein AA1 is a hydrophobic amino acid residue, such as an alanine, valine, leucine, isoleucine, phenylalanine, methionine or tryptophan unit. AA2 is a basic amino acid residue, such as an arginine or a lysine unit. Particularly preferred are Cam-Phe-Arg and Cam-Phe-Lys. The AA1 and the AA2 typically have a free side-chain functionality, i.e. that is free of a protective group or another residue.

Particularly good results have also been obtained with carboxyl substituted benzyl esters, in particular with p-carboxyl substituted benzyl esters represented by the formula peptide-(C=0)-0-CH ₂ -C ₆ H ₄ -C0 ₂ E wherein E represents a hydrogen atom, a positively charged salt ion such as an ammonium ion, or an amino acid or a peptide residue with a C-terminal carboxyamide or carboxylic acid functionality, optionally protected on the side-chain functionality of the amino acid or on one or more of the side-chain functionalities of the amino acids. Good results have also been obtained with p-carboxyl substituted benzyl esters represented by the formula peptide-(C=0)-0-CH ₂ -C ₆ H ₄ - CO2E wherein E is defined as above and in which one or more hydrogen atoms in the phenyl ring (C6H4 in the above formula) are replaced by a substituent, such as hydroxy, alkoxy, aryloxy or halogen.

The peptide C-terminal (thio)ester can be N-terminally unprotected or N-terminally protected.

Suitable N-terminal protecting groups are those N-protecting groups which can be used for the synthesis of peptides. Such groups are known to the person skilled in the art. Examples of suitable N-protecting groups include carbamate or acyl type protecting groups, for instance‘Cbz’ (benzyloxycarbonyl),‘Boc’ (tert-butyloxycarbonyl),‘For’ (formyl),‘Fmoc’ (9-fluorenylmethoxycarbonyl),‘PhAc’ (phenacetyl) and‘Ac’ (acetyl). The groups For, PhAc and Ac may be introduced and cleaved enzymatically using the enzymes Peptide Deformylase, PenG acylase or Acylase, respectively. The inventors further realized that a substituted thiocarbamoyl group, such as a phenylthiocarbamoyl (PTC) group is a useful protective group for the N-terminal a-amine function of the C-terminal (thio)ester in a method for enzymatically synthesising peptides by fragment condensation. The use of such groups as such is e.g. well known from Edman degradation processes. Protective agents that can be used for Edman-type amino acid sequencing are also referred to herein as Edman-type protective agents, and likewise agents coupled to an amine group (in particular the N-terminal amine) of a peptide are referred to herein as Edman-type protective groups. The inventors found that an Edman-type protective agent - attached to the N-terminal a-amino function via a linking additional amino acid, e.g. glycine - forms an effective protecting group, when after the enzymatic coupling reaction the obtained peptide needs to be provided with a functional group at position Y to obtain Liraglutide or Semaglutide, e.g. when coupling Pal to the Lys-y-Glu (Y) or when coupling 17-carboxyheptadecanoic acid to the Lys-AEEA-AEEA-y-Glu (Y). Substituted thiocarbamoyl groups can be provided to the N-terminal a-amino function by reacting said amine function with the corresponding isothiocyanate under (slightly) alkaline conditions. Hence, a phenylthiocarbamoyl (PTC) group can be introduced using phenylisothiocyanate (PITC) and a methylthiocarbamoyl (MTC) group can be introduced using methylisothiocyanate (MITC). Under acidic conditions such substituted thiocarbamoyl groups are cleaved from the peptide together with the a-amino acid to which they are attached in the form of a thiazolinone derivative.

This new way of providing N-terminal protection has been found advantageous over, e.g., Fmoc with respect to solubility in an aqueous reaction system. It has been found advantageous over, e.g. Boc, in terms of compatibility when using solid-phase synthesis. An Edman-type protective group, such as a substituted thiocarbamoyl moiety, functions particularly well as a protective group at neutral or alkaline pH and can be easily removed at acidic pH. Thus, such group is usually employed in a coupling reaction at neutral or alkaline pH, using a ligase having a good S/H ratio at such pH, like a Subtilisin BPN’ variant or homologue, as described in more detail elsewhere herein.

Suitable protection/deprotection conditions when using Edman-type protective moieties include those that are generally known in the art for using such moiety in Edman-type degradation methods. A substituted thiocarbamoyl group has been found particularly effective, in combination with contributing to good solubility, also in an aqueous reaction medium. The substituted thiocarbamoyl group can be aromatic or aliphatic. Preferably, the substituted thiocarbamoyl group is an aryl-substituted thiocarbamoyl group, or an alkyl-substituted thiocarbamoyl group. Particularly preferred aryl-substituted thiocarbamoyl groups are C6-C12-aryl-substituted thiocarbamoyl groups, more in particular phenylthiocarbamoyl (PTC). Particularly preferred alkyl-substituted thiocarbamoyl groups are C1- C6-alkyl-substituted thiocarbamoyl groups, more in particular methylthiocarbamoyl (MTC). Further examples of preferred isothiocyanates to be used for the introduction of substituted thiocarbamoyl groups are those mentioned in H. Matsunaga, T. Santa, K. Hagiwara, H. Homma, K. Imai, S. Uzu, K. Nakashima, S. Akiyama, Anal. Chem.1995, 67, 4276, such as FITC, BAMPITC, DNTC, DNSAPITC, dansylamino-PITC, 3-POPICs, 4-POPICs, CIPIC and 7-[(N,N- dimethylamino)sulphonyl]-2,1 ,3-benzoxadiazol-4-yl isothiocyanate (DBD-NCS), see the paragraph bridging the left- hand and right hand column of page 4276, incorporated by reference. Yet another preferred example is 7- aminosulphonyl-4-(2,1 ,3-benzoxadiazolyl)-isothiocyanate (ABD-NCS). As an alternative to a substituted thiocarbamoyl moiety, another moiety suitable for sequencing amino acids in a peptide with an Edman-type degradation method can be used as a protective group in a similar fashion, i.e. by labelling the N-terminal end of the peptide C-terminal ester with said moiety via a linking amino acid and, after enzymatic coupling with the peptide nucleophile - cleaving the moiety together with the linking amino acid residue from the remainder of the coupling product. Suitable protective moieties that can be labelled to the N-terminal end of a peptide via a linking amino acid residue and cleaved off together with the linking amino acid residue are therefore also referred to herein as‘Edman-type protective groups’.

Further, it is possible to link an Edman-type protective group to the peptide C-terminal (thio)ester via more than one amino acid (i.e. via a peptide chain). The linking amino acids can then be removed by a number of cycles of labeling with the moiety and cleaving off the moiety plus amino acid, in a similar way as is done in a peptide sequencing method. The use of additional linking-amino acids is not necessary, but they can be used - if desired - e.g. to modify the solubility of the peptide C-terminal (thio)ester in a reaction medium of choice.

Thus, in a specific preferred embodiment, a method according to the invention comprises enzymatically coupling

(a) a peptide C-terminal ester or thioester represented by the formula P-W _v -His-X-Glu-Gly-Thr-Phe-Thr- Ser-Asp-Val-Ser-(thio)ester with (b) the peptide nucleophile. Herein P represents the Edman-type protective group, preferably a phenylthiocarbamoyl (PTC) or a methylthiocarbamoyl (MTC) moiety. Herein, v is an integer of at least 1 , usually of 1-10, preferably 1-4, more preferably 1 , 2 or 3, most preferably 1 and v represents the number of amino acid residues W. Each W can be the same or different. Usually each W is selected from the group of proteinogenic amino acids, although in principle another amino acid could be used, provided it can be cleaved off as a P-W under Edman-type cleaving conditions.

This N-terminally protected peptide C-terminal (thio)ester is coupled with the nucleophile (b), whereby the N-terminally protected peptide P-W _v -His-X-Glu-Gly-Thr-Phe-Thr-Ser-Asp-Val-Ser-Ser-Tyr-Leu -Glu-Gly-Gln-Ala- Ala-Y-GI u-Phe-lle-Ala-T rp-Leu-Val-Z-Gly-Arg-Gly is formed. Thereafter, this peptide is subjected to a cleavage reaction wherein the peptide

W _v -i-His-X-Glu-Gly-Thr-Phe-Thr-Ser-Asp-Val-Ser-Ser-Tyr-L eu-Glu-Gly-Gln-Ala-Ala-Y-Glu-Phe-lle-Ala-Trp-Leu-Val-Z- Gly-Arg-Gly (which peptide may be elongated at the C-terminus) is formed.

If v-1 > 0, a group P is coupled to the N-terminal a-amino function of W of the peptide, after which P-W is removed by cleavage. A coupling and cleaving cycle is then repeated till the peptide ¹ His-X-Glu-Gly-Thr-Phe-Thr-Ser-Asp-Val-Ser- Ser-Tyr-Leu-Glu-Gly-Gln-Ala-Ala-Y-Glu-Phe-lle-Ala-Trp-Leu-Va l-Z-Gly-Arg-Gly is obtained.

Labelling of P to the N-terminal end of the peptide is accomplished in a manner known perse, based on the Edman type methodology known for said P, typically at mildly alkaline conditions, e.g. about pH 8. Cleavage of P- W from the N-terminal end of the peptide is accomplished in a manner known per se, based on the Edman type methodology known for said P, typically at acidic conditions, usually at a pH of about 4 or less, in particular in the range of about 3 or less, e.g. of 0-2. For instance trifluoroacetic acid (TFA) may be used. N-terminal protection of the peptide (thio)ester is in particular useful in a method wherein Y comprises a Lys(Y-Glu-OH) moiety bearing a free a-amino function which needs to be coupled to palmitic acid or if Y comprises a Lys(AEEA-AEEA-y-Glu-OH) moiety bearing a free a-amino function which needs to be coupled to 17-carboxy- heptadecanoic acid.

In particular, good results have been achieved with a peptide C-terminal (thio)ester without protected side-chain functionalities. However, in an embodiment, one or more side-chain functionalities, e.g. all side-chain functionalities, are provided with a protecting group. Suitable protecting groups are known to the person skilled in the art. Carboxylic acid groups can for instance be protected with a cyclohexyl, benzyl or allyl group.

The activated C-terminal (thio)ester group of the peptide C-terminal (thio)ester can be synthesized using solid phase synthesis in high yield and purity without racemization. An additional advantage of the use of (thio)esters of the carboxyamidomethyl type wherein Ri represents a hydrogen atom and f¾ represents an amino acid or peptide residue with a C-terminal carboxylic acid functionality, optionally protected on the side-chain functionality of the amino acid or on one or more of the side-chain functionalities of the amino acids is, that their activated C-terminal ester or thioester group can be synthesized using the cheap and industrially available 2-chlorotritylchloride resin.

The activated C-terminal (thio)ester group of the peptide C-terminal (thio)ester can also be synthesized by solution phase synthesis or by fermentation, i.e. using a microorganism. As generally known in the art fermentative processes include production of a compound, i.e. a peptide under aerobic or anaerobic conditions. A reliable method to obtain peptide (thio)esters using fermentation is via so-called intein expression (see for instance E.K. Lee, Journal of Chemical Technology and Biotechnology, 2010, 9, 11-18). Different intein expression system kits are commercially available (for instance the IMPACT™ kit). Other methods for the fermentative production of peptide (thio)esters are known in the art.

The peptide nucleophile having an N-terminally unprotected amine comprises the amino acid sequence H-Ser-Tyr-Leu-Glu-Gly-Gln-Ala-Ala-Y-Glu-Phe-lle-Ala-Trp-Leu- Val-Z-Gly-Arg-Gly (‘the second peptide fragment’). Particularly good results have been achieved with a peptide nucleophile, wherein this is the amino acid sequence of the peptide nucleophile. It is in particular an important advantage of the present invention that - also in an aqueous system - the enzymatic coupling works well without needing the C-terminal end to be extended with a peptide tag or another derivative to enhance solubility or reactivity of the peptide nucleophile.

In an embodiment, the peptide nucleophile is C-terminal protected. In another embodiment it is free of C-terminal protection.

In particular, good results have been achieved with peptide nucleophiles without protected side-chain functionalities.

In an embodiment, one or more side-chain functionalities (in particular one or more hydroxyl, carboxyl or amine groups) of the peptide nucleophile are provided with a protecting group. Suitable protecting groups are known to the person skilled in the art. Carboxylic acid groups can for instance be protected with a cyclohexyl, benzyl or allyl group; amine functionalities can for instance be protected with an allyloxycarbonyl group or a trifluoroacetyl group. The peptide nucleophile may be synthesized using methods known in the art, such as solid phase synthesis, solution phase synthesis or by fermentation.

As mentioned above, Y is Lys, of which Lys the side-chain e-amino group may be protected with a protective group. However, it is generally not necessary for a satisfactory coupling yield and rate to protect the side- chain e-amino group, in particular not if a subtilisin or homologue thereof is used as the ligase. In particular, a subtilisin BPN’ variant or homologue as described herein is suitable to couple both fragments also when the e-amino group of Lys at position Y is free of a protective group.

Accordingly, usually the Y of the peptide nucleophile is a lysine residue having a free e-amino side chain or having a functionalised e-amino side chain. The product obtained by the enzymatic coupling can be the peptide of interest (optionally after removal of protective groups, if any), e.g. if GLP-1 is the peptide of interest to be synthesised or if Y of the peptide nucleophile already comprises the needed functionalisation to obtain Liraglutide of Semaglutide. Alternatively, the product obtained by enzymatic coupling can subsequently be subjected to further reactions to functionalise it, in particular with an amino acid or another functional group, more in particular a functional group selected from the group consisting of Pal-y-Glu-OH, and AEEA-AEEA-y-Glu-N-17-carboxyheptadecanoyl-OH, wherein Pal is palmitoyl and AEEA-AEEA is 2-[2-(2-aminoethoxy)ethoxy]acetyl-2-[2-(2-aminoethoxy)ethoxy ]acetyl. Ways to functionalise the free amino acid side chain of Y to yield Liraglutide or Semaglutide or to provide the peptide nucleophile suitable for synthesising Liraglutide or Semaglutide can be based on methodology generally known in the art or may be based on the present examples or on the technology described in the literature referred to in the references cited herein. In particular, a functionalisation protocol may be used based on US 6,451 ,974 B1.

In a particularly preferred embodiment, the peptide that is synthesised in a method according to the invention is Liraglutide.

The present invention is advantageous in that it allows enzymatically coupling (a) the peptide C-terminal ester or thioester comprising the first peptide fragment, said first fragment comprising the sequence His-Ala-Glu-Gly- Thr-Phe-Thr-Ser-Asp-Val-Ser-(thio)ester, and (b) the peptide nucleophile comprising the second peptide fragment, said second fragment comprising the sequence H-Ser-Tyr-Leu-Glu-Gly-Gln-Ala-Ala-Y-Glu-Phe-lle-Ala-Trp-Leu- Val- Arg-Gly-Arg-Gly, wherein Y is Lys(Pal-y-Glu-OH). Thus, the enzymatic coupling may be carried out coupling the corresponding 11-mer peptide C-terminal (thio)ester and 20-mer N-terminal nucleophile, thereby obtaining Liraglutide. Although the N-terminal a-amino function of the peptide (thio)ester may be provided with a protective group such as a group represented by P-W _v , as defined elsewhere herein when describing Edman-type protective groups, particularly good results have inter alia been achieved with a method wherein the peptide nucleophile having Y = Lys(Pal-y-Glu- OH) is coupled to the peptide C-terminal ester or thioester represented by the formula His-Ala-Glu-Gly-Thr-Phe-Thr- Ser-Asp-Val-Ser-(thio)ester without any protection of the N-terminal end of the peptide (thio)ester and without needing any further protective groups.

In a further preferred embodiment, Liraglutide is prepared in a method comprising enzymatically coupling (a) the peptide C-terminal ester or thioester comprising the first peptide fragment, said first fragment comprising the sequence His-Ala-Glu-Gly-Thr-Phe-Thr-Ser-Asp-Val-Ser-(thio)ester, and (b) the peptide nucleophile comprising the second peptide fragment, said second fragment comprising the sequence H-Ser-Tyr-Leu-Glu-Gly-Gln-Ala-Ala-Lys(Y - Glu-OH) -Glu-Phe-lle-Ala-Trp-Leu-Val-Arg-Gly-Arg-Gly; and thereafter providing said Lys (g-Glu-OH) residue with a palmitoyl group (Pal), thereby obtaining the Liraglutide. In this embodiment, the N-terminal a-amino function of the peptide C-terminal ester or thioester is usually protected during enzymatic coupling, preferably with an Edman-type protective group.

In a further particularly preferred embodiment, Liraglutide is prepared by a method comprising enzymatically coupling (a) the peptide C-terminal ester or thioester comprising the first peptide fragment, said first fragment comprising the sequence His-Ala-Glu-Gly-Thr-Phe-Thr-Ser-Asp-Val-Ser-(thio)ester, and (b) the peptide nucleophile comprising the second peptide fragment, said second fragment comprising the sequence H-Ser-Tyr-Leu- Glu-Gly-Gln-Ala-Ala-Y-Glu-Phe-lle-Ala-T rp-Leu-Val-Arg-Gly-Arg-Gly, wherein Y is a lysine residue having a free e- amino side chain; and thereafter providing said e-amino side chain with Pal-y-Glu-OH, thereby obtaining the Liraglutide. In this embodiment, particularly good results have been achieved without using any protective group at the N-terminal a-amino function of the peptide (thio)ester.

In an alternative embodiment, a method according to the invention comprises the synthesis of

Semaglutide.

In an advantageous embodiment, the synthesis of Semaglutide comprises enzymatically coupling (a) the peptide C-terminal ester or thioester comprising the first peptide fragment, said first fragment comprising the sequence His-Aib-Glu-Gly-Thr-Phe-Thr-Ser-Asp-Val-Ser-(thio)ester, and (b) the peptide nucleophile comprising the second peptide fragment, said second fragment comprising the sequence H-Ser-Tyr-Leu-Glu-Gly-GIn-Ala-Ala-Y-Glu- Phe-lle-Ala-Trp-Leu-Val-Arg-Gly-Arg-Gly, wherein Y is Lys(AEEA-AEEA-y-Glu-N-17-carboxyheptadecanoyl-OH). Thereby the Semaglutide can be directly obtained. Although the N-terminal a-amino function of the peptide (thio)ester may be provided with a protective group such as a group represented by P-W _v , as defined elsewhere herein when describing Edman-type protective groups, particularly good results have inter alia been achieved with a method wherein the peptide nucleophile with Y = Lys(AEEA-AEEA-y-Glu-N-17-carboxyheptadecanoyl-OH is coupled to the peptide C- terminal ester or thioester represented by the formula His-Ala-Glu-Gly-Thr-Phe-Thr-Ser-Asp-Val-Ser-(thio)ester without any protection of the N-terminal a-amino function of the peptide (thio)ester and without needing any further protective groups.

In a further embodiment, a method for the synthesis of Semaglutide comprises enzymatically coupling (a) the peptide C-terminal ester or thioester comprising the first peptide fragment, said first fragment comprising the sequence His-Aib-Glu-Gly-Thr-Phe-Thr-Ser-Asp-Val-Ser-(thio)ester, and (b) the peptide nucleophile comprising the second peptide fragment, said second fragment comprising the sequence H-Ser-Tyr-Leu-Glu-Gly-GIn-Ala-Ala-Y-Glu- Phe-lle-Ala-Trp-Leu-Val-Arg-Gly-Arg-Gly, wherein Y is Lys(AEEA-AEEA-y-Glu-OH), and thereafter providing the Lys(AEEA-AEEA-y-Glu-OH) moiety with a 17-carboxyheptadecanoyl group. Thus, Semaglutide can be obtained by further functionalisation after enzymatic coupling. In this embodiment, the N-terminal a-amino function of the peptide C-terminal ester or thioester is usually protected during enzymatic coupling, preferably with an Edman-type protective group.

In a further embodiment, the Semaglutide is obtained by a method comprising enzymatically coupling (a) the peptide C-terminal ester or thioester comprising the first peptide fragment, said first fragment comprising the sequence His-Aib-Glu-Gly-Thr-Phe-Thr-Ser-Asp-Val-Ser-(thio)ester, and (b) the peptide nucleophile comprising the second peptide fragment, said second fragment comprising the sequence H-Ser-Tyr-Leu-Glu-Gly-GIn-Ala-Ala-Y-Glu- Phe-lle-Ala-Trp-Leu-Val-Arg-Gly-Arg-Gly, wherein Y is a lysine residue having a free e-amino side chain, and thereafter providing said e amino side chain with an AEEA-AEEA-y-Glu-N-17-carboxyheptadecanoyl-OH group, thereby obtaining the Semaglutide. Thus, Semaglutide can be obtained by functionalisation after enzymatic coupling. In this embodiment, particularly good results have been achieved without using any protective group at the N-terminal a-amino function of the peptide (thio)ester (or any side-chain functionalities).

In yet another alternative embodiment, the peptide that is synthesised is GLP-1.

The ligase used to catalyse the coupling of the peptide C-terminal (thio)esterand the peptide nucleophile can be any ligase having catalytic activity in coupling both peptides by catalysing the formation of a peptide bond between the C-terminus of the peptide C-terminal (thio)ester and the N-terminus of the peptide nucleophile, wherein the S/H ratio for the coupling vs. the hydrolysis of the coupling product in the used reaction medium is larger than 1. Usually, the ligase can be classified as a serine protease which can generally be classified in EC 3.4.21. Generally, it has a catalytic triad in the order Asp, His and Ser.

In particular, a ligase used in a method according to the invention is an isolated enzyme. Thus, it is isolated from the organism wherein it has been expressed, typically a recombinant organism, if it has been produced in an organism, respectively isolated from the reaction medium in which it has been synthesized.

In particular, an enzyme of the invention is considered isolated for the purpose of the invention either in the crude form or substantially purified by any suitable technique such as, for example, the single-step purification method disclosed in Smith and Johnson, Gene 67:31-40 (1988).

In particular, the ligase can be a serine endoprotease. The ligase typically has an S/H ratio larger than 1, preferably 2 or more, in particular 5 or more in the used reaction medium, in particular in a reaction medium comprising water, more in particular an aqueous medium. The upper value of this quotient is not critical; in practice it may e.g. be 100 or less, in particular 20 or less. The ligase used in a method according to the invention generally has an improved“synthesis over hydrolysis ratio” (S/H ratio), at least compared to subtilisin BPN’.

The S/H ratio of the ligases (used in a method) according to the invention divided by the S/H ratio of subtilisin BPN’- at least under the conditions described in the examples - is usually more than 100, preferably 250 or more, more preferably 500 or more, in particular 1000 or more. The upper value of this quotient is not critical; it may approximate infinity.

In particular, very good results have been achieved with a subtilisin BPN’ variant ora homologue thereof. Especially when carrying out the enzymatic coupling in a reaction medium comprising water as a major solvent ( e.g . 50-100 wt.% based on total liquid) a subtilisin BPN’ variant or a homologue thereof according to WO 2016/056913 has been found particularly suitable. The contents of the publication are incorporated by reference, in particular with respect to the details about the subtilisin BPN’ variant or a homologue, as present in the claims thereof.

Thus, usually, the ligase used for the coupling reaction is a subtilisin BPN’ variant or a homologue thereof comprising the following mutations compared to subtilisin BPN’ represented by SEQ ID NO: 2 or a homologue sequence thereof :

- a deletion of the amino acids corresponding to positions 75-83;

- a mutation at the amino acid position corresponding to S221 , the mutation being S221C or S221selenocysteine;

- preferably a mutation at the amino acid position corresponding to P225;

- wherein the amino acid positions are defined according to the sequence of subtilisin BPN’ represented by SEQ ID NO: 2.

Further preferred ligases for use in a method according to the invention may comprise one or more additional mutations, in particular one or more further mutations as identified elsewhere herein or in WO 2016/056913, incorporated herein by reference.

The mutation at the amino acid position corresponding to S221 of the ligase, in particular the subtilisin BPN’ variant or homologue thereof, preferably is S221C.

The mutation at the amino acid position corresponding to P225 is usually advantageous for the S/H ratio for the enzymatic coupling. The mutation is usually selected from the group of P225N, P225D, P225S, P225C, P225G, P225A, P225T, P225V, P225I, P225L, P225H, P225Q, preferably from the group of P225N, P225D, P225S, P225C and P225G, more preferably P225N or P225D, most preferably P225N.

For a good enzyme stability, the ligase, in particular the subtilisin BPN’ variant or homologue thereof , preferably comprises one or more mutations selected from the group of mutations at an amino acid position corresponding to Q2, S3, P5, S9, 131 , K43, M50, A73, S188, Q206, N212, N218, T254 and Q271 of SEQ ID NO 2.

A preferred mutation at the position corresponding to Q2 corresponds to Q2K.

A preferred mutation at the position corresponding to S3 corresponds to S3C.

A preferred mutation at the position corresponding to P5 corresponds to P5S.

A preferred mutation at the position corresponding to S9 corresponds to S9A.

A preferred mutation at the position corresponding to 131 corresponds to 131 L

A preferred mutation at the position corresponding to K43 corresponds to K43N.

A preferred mutation at the position corresponding to M50 corresponds to M50F.

A preferred mutation at the position corresponding to A73 corresponds to A73L

A preferred mutation at the position corresponding to S188 corresponds to S188P.

A preferred mutation at the position corresponding to Q206 corresponds to Q206C.

A preferred mutation at the position corresponding to N212 corresponds to N212G. A preferred mutation at the position corresponding to T254 corresponds to T254A.

A preferred mutation at the position corresponding to Q271 corresponds to Q271 E.

In a particularly preferred embodiment, the ligase, in particular the subtilisin BPN’ variant or homologue thereof, comprises at least six, preferably at least eight, more preferably at least 10, in particular 12, 13 or 14 of said mutations selected from the group of mutations at positions corresponding to Q2, S3, P5, S9, 131 , K43, M50, A73, S188, Q206, N212, T254 and Q271. This is in particular preferred for enzyme stability in a reaction medium comprising water as a major or the only solvent. The ligase may have further mutations compared to subtilisin BPN’, provided that it has enzymatic fragment condensation activity (coupling activity) in the preparation of the peptides according to the present invention, in particular one or more further mutations as described in the references cited herein.

Alternatives to subtilisin BPN’, as template enzymes from which an enzyme according to the invention, in particular a homologue of a subtilisin BPN’ variant of the invention, can be derived by mutagenesis, are other subtilisins, in particular subtilisins having at least 50 % homology with subtilisin BPN’.

Sequences of suitable subtilisins can be retrieved from the UNIPROT sequence database (http://www.uniprot.org/), as available on 11 August 2014, by BLASTing the database with subtilisin BPN’ (SEQ ID 2) as a query. However, sequence retrieval is not limited to UNIPROT nor to the date. The skilled person in the art knows how to query alternative sequence depositories or to collect additional homologue sequences by sequencing (see for example Zooming in on metagenomics: molecular microdiversity of Subtilisin Carlsberg in soil. Gabor E, Niehaus F, Aehle W, Eck J.J Mol Biol. 2012 Apr 20;418(1-2):16-20).

In particular, the invention further relates to variants, having at least said deletions of the amino acids corresponding to L75 till and including G83 of subtilisin BPN’, cysteine or selenocystein at a position corresponding to position 221 in subtilisin BPN’ and at least one of said further mutations in present claim 1.

The sequence of subtilisin BPN’ is given in SEQ ID NO 2 (mature form). The gene encoding for subtilisin BPN’ amino acids -107 to 275 is given in SEQ ID NO 1. The subtilisin BPN’ variant or homologue can be based on the enzymes according to WO2016/056913, with the proviso that it has the above-mentioned mutations.

In an advantageous embodiment, the ligase is a subtilisin BPN’ variant having a deletion of the amino acids corresponding to positions 75-83, the mutation S221C and one or more further mutations, preferably at least 3 further mutations, in particular 5-8 further mutations, at amino acid positions corresponding to M222, Y217, P225, F189, N218, E156, G166 and N62 of wild-type subtilisin BPN’ (mature).

Of these mutations, in particular good results have been achieved with the mutations corresponding to: M222P, Y217H, P225N, F189W, N218D, E156N, G166E, N62A. SEQ ID NO: 3 shows a subtilisin BPN’ variant (for use) according to the invention with deletion of the Ca ²⁺ binding loop, S221C and having said further mutations. The His tag was included for facilitating purification and is not needed for ligase activity. Further preferred enzymes may comprise one or more additional mutations, in particular one or more further mutations as identified elsewhere herein or in WO 2016/056913, incorporated herein by reference.

In a method of the invention the enzymatic reaction is typically performed in a fluid comprising water. Preferably the reaction is performed in a buffered fluid. The water content usually is 10-100 vol %, based on total liquids, preferably 20 vol. % or more, preferably 40 vol. % or more, in particular 50 vol. % or more, more in particular 60 vol. % or more. In particular good results have been achieved in a reaction medium, comprising 70-100 vol % water, more in particular 90-100 vol. %, 95-100 vol. % or 98-100 vol. % water. The term‘aqueous’ is used for media at least substantially consisting of water.

In principle, any buffer is suitable. Good buffers are known to a person skilled in the art. See for instance David Sheehan in Physical Biochemistry, 2 ^nd Ed. Wiley-VCH Verlag GmbH, Weinheim 2009; http://www.sigmaaldrich.com/life-science/core-bioreagents/bi ological-buffers/learning-center/buffer-calculator.html. Particularly good results have e.g. been achieved with a Good's buffer, such as tricine. The concentration of the buffer may be chosen within wide limits, e.g. in the range of 10-1000 mM, in particular in the range of 25-500 mM, more in particular in the range of 50-250 mM. A relatively low molarity of the buffer has been found advantageous for coupling a peptide nucleophile wherein Y is Lys(Pal-y-Glu-OH) or the like.

The pH of the buffer for a coupling reaction in a method according to the invention may be at least 5, in particular at least 6, preferably at least 7. A desired pH is usually less than 11 , in particular less than 10, even more preferably less than 9. Usually the optimal pH for the enzymatic coupling is between 7 and 9.

Due to the high S/H ratio, a large excess of the peptide C-terminal ester or thioester or of the peptide nucleophile is generally not needed to reach a high yield in the condensation reaction. Generally, they are contacted in an about stoichiometric ratio or in an excess of the peptide C-terminal ester, in particular in a molar ratio of (a) the peptide C-terminal ester or thioester to (b) the peptide nucleophile in the range of 1 :1 to 5:1. Although satisfactory results are achieved with a stoichiometric ratio, an excess of the peptide C-terminal (thio)ester has been found advantageous for the reaction rate. Thus, preferably the molar ratio of (a) the peptide C-terminal ester or thioester to (b) the peptide nucleophile is in the range of 1.05:1.0 to 4:1 , more preferably in the range of 1.1 :1.0 to 3:1 , even more preferably in the range of 1.2:1.0 to 2.5:1.0, in particular in the range of 1.2:1.0 to 2.0:1.0.

In a method of the invention, it may be advantageous to add additives to the fluid wherein the reaction is carried out to improve the solubility of the peptide fragments or to improve the reaction yield. Such additives may be a salt or an organic molecule, for instance guanidinium hydrochloride, urea, sodium dodecasulphate or Tween. However, good results have been achieved without such additive, also in a fully aqueous reaction medium, e.g. in an embodiment wherein the Y is Lys(Pal-y-Glu-OH) or the like.

The reaction may be carried out in a fully aqueous liquid or in a mixture of water and a water miscible co-solvent such as N,N-dimethylformamide (DMF), N-methyl-pyrrolidinone (NMP), N,N-dimethylacetamide (DMA), dimethylsulphoxide (DMSO), acetonitrile, an ether, such as tetrahydrofuran (THF), 2-methyl-tetrahydrofuran (Me-THF) or 1 ,2-dimethoxyethane, or a (halogenated) alcohol, such as methanol, ethanol, isopropanol, tert-butanol, 2,2,2- trifluoroethanol (TFE), 1 ,1 ,1 ,3,3,3-hexafluoroisopropanol, or a mixture of these organic solvents. Depending on the stability of the subtilisin BPN’ variant and the solubility of the peptide substrates, the amount of co-solvent is preferably below 70 vol%, more preferably below 60 vol%, even more preferably below 50 vol%, and most preferably below 40%.

In principle the temperature during the enzymatic fragment condensation is not critical, as long as a temperature is chosen at which the ligase to be used shows sufficient activity and stability. Such a temperature can be routinely determined. Generally, the temperature may be at least -10 °C, in particular at least 0 °C or at least 10 °C. Generally, the temperature may be 70°C or less, in particular 60°C or less or 50 °C or less. Optimal temperature conditions can easily be identified for a specific ligase for a specific enzymatic fragment condensation by a person skilled in the art through routine experimentation based on common general knowledge and the information disclosed herein. In general, the temperature advantageously is in the range of 20-50°C.

The invention further relates to the use of an Edman-type agent to provide a protective group in the synthesis of a peptide in a method comprising enzymatic coupling of peptides by fragment condensation. Accordingly, the invention further relates to a method for synthesizing a peptide, comprising enzymatically coupling (a) a peptide C- terminal ester or thioester represented by the formula P-W _v -AA _n -(thio)ester with a peptide nucleophile represented by the formula AA _m , which coupling is catalysed by a ligase, preferably a subtilisin BPN’ variant or homologue, such as described elsewhere herein.

Herein P represents the Edman-type protective group, as defined above, preferably a thiocarbamoyl group. Coupling of P to the N-terminal end of the peptide is accomplished in a manner known per se, based on the Edman type methodology known for said P, typically at mildly alkaline conditions, e.g. about pH 8. Herein v is an integer of at least 1 , usually preferably 1-10, preferably 1-5, more preferably 1 , 2 or 3, most preferably 1 and v representing the number of amino acid residues W, wherein each W can be the same or different, and preferably is as defined as above. Each AA stands for an amino acid residue, n is an integer representing the number of amino acid residues of the peptide C-terminal ester or thioester, and m is an integer representing the number of amino acid residues of the peptide nucleophile. Typically, the sum of n and v is at least 4 in order to allow recognition by the ligase. Preferably, n is in the range of 3-200, in particular in the range of 3-50, more in particular in the range of 3-25. In a specific embodiment, n is at least 4, at least 6, at least 8, at least 10, at least 15 or at least 20. Preferably, m is in the range of 3-200, in particular in the range of 5-50, more in particular in the range of 8-30. In a specific embodiment, m is at least 4, at least 10, at least 15 or at least 20.

The coupling product, P-WAA _n -AA _m is subject to a cleavage reaction wherein the peptide W _v -i-AA _n - AA _m is formed. Typically, cleavage is accomplished under acidic conditions. If v-1 > 0, thereafter a group P is coupled to the W at the N-terminal position of the peptide W _v -i-AA _n -AA _m , to form P-W _v -i-AA _n -AA _m after which P-W is cleaved. This is then repeated till the peptide represented by formula AA _n -AA _m is obtained.

Accordingly, the present invention relates to a method for synthesising a peptide comprising the sequence His-X-Glu-Gly-Thr-Phe-Thr-Ser-Asp-Val-Ser-Ser-Tyr-Leu-Glu-Gl y-Gln-Ala-Ala-Y-Glu-Phe-lle-Ala-Trp-Leu- Val-Z-Gly-Arg-Gly, wherein

X is Ala or an a-amino-isobutyric acid (Aib) residue;

Y is Lys, which Lys has a free side-chain e-amino group (i.e. a non-derivatised lysine residue), or whose Lys side-chain e-amino group is protected with a protective group, or whose Lys side-chain e-amino group is functionalised with an amino acid or another functional group; Z is Arg or Lys;

the method comprising enzymatically coupling

(c) a first peptide C-terminal ester or thioester fragment comprising the sequence His-X-Glu-Gly- Thr-Phe-Thr-Ser-Asp-Val-Ser-(thio)ester, and

(d) a second peptide nucleophile fragment having an N-terminally unprotected amine comprising the sequence H-Ser-Tyr-Leu-Glu-Gly-Gln-Ala-Ala-Y-Glu-Phe-lle-Ala-Trp-Leu- Val-Z-Gly-Arg- Gly;

which enzymatic coupling is catalysed by a ligase, wherein said ligase is a subtilisin BPN’ variant, or a homologue thereof having at least 80%, or 85%, or 90%, preferably 95%, sequence identity. Preferably the ligase is a subtilisin BPN’ variant having a deletion of the amino acids corresponding to positions 75-83, the mutation S221 C and one or more further mutations, preferably at least 3 further mutations, in particular 5-8 further mutations, at amino acid positions corresponding to M222, Y217, P225, F189, N218, E156, G166 and N62 of wild-type subtilisin BPN’ (mature), whereof most preferably one mutation is at P225.

In a particularly advantageous embodiment, the ligase used in the method according to the invention is a subtilisin BPN’ variant, or a homologue thereof having at least 80%, or 85%, or 90%, preferably 95%, sequence identity with SEQ ID N0 14, comprising the mutations Q2K, S3C, P5S, S9A, 131 L, K43N, M50F, A73L, D75-83, E156S, G166S, G169A, S188P, Q206C, N212G, Y217H, S221C, M222P, P225N, T254A, and Q271 E, optionally comprising a His tag.

In another preferred embodiment, the ligase used in the method according to the invention is a subtilisin BPN’ variant with SEQ ID NO 3, or a homologue thereof having at least 80%, or 85%, or 90%, preferably 95%, sequence identity, comprising the mutations Q2K, S3C, P5S, S9A, 131 L, K43N, M50F, N62A, A73L, D75-83, E156N, G166E, G169A, S188P, F189W, Q206C, N212G, Y217H, N218D, S221C, M222P, P225N, T254A, Q271 E, optionally comprising a His tag.

SEQ ID NO 14 shows a subtilisin BPN’ variant (for use) according to the invention with deletion of the Ca ²⁺ binding loop, S221C and further mutations. The His tag was included for facilitating purification and is not needed for ligase activity.

All the ligases described are also embodiments of the invention, preferably those specified by SED ID NO 14 and SEQ ID NO 3 and homologues thereof having at least 80%, or 85%, or 90%, preferably 95%, sequence identity.

In a preferred embodiment the method is further characterised by Y being a Lys, of which the side-chain e-amino group is functionalised with a functional group, selected from the group consisting of g-Glu-OH, Pal-y-Glu-OH, AEEA-AEEA-g-Glu-OH and AEEA-AEEA-y-Glu-N-17-carboxyheptadecanoyl-OH, wherein Pal is palmitoyl and AEEA- AEEA is -2-[2-(2-aminoethoxy)ethoxy]acetyl-2-[2-(2-aminoethoxy)ethox y]acetyl.

In an even more preferred embodiment said Y is Lys(y-Glu-OH), Lys(AEEA-AEEA-y-Glu-OH), Lys(Pal- g-Glu-OH) or Lys(AEEA-AEEA-y-Glu-17-carboxyheptadecanoyl-OH). With the term“peptide fragment” or“fragment” is meant a peptide with a partial amino acid sequence, with reference to a longer peptide with a defined sequence.

The invention will now be illustrated by the following examples, without being limited thereto.

EXAMPLES

Production of liqases

Mutagenesis, Cloning and Expression

SEQ ID NO 1 shows the wild type gene coding for subtilisin BPN’ amino acids -107 to 275. Herein the codons coding for amino acids -107 to -1 are present. These amino acids comprise the signal sequence, the presequence and a pro sequence which are cleaved off upon full maturation. SEQ ID NO 2 shows the mature wild type subtilisin BPN’ (i.e. without the amino acids -107 to -1). The ligase used for the Examples was as shown in SEQ ID NO: 3. Compared to the mature wild type subtilisin BPN’, this ligase had the mutations Q2K, S3C, P5S, S9A, 131 L, K43N, M50F, N62A, A73L, D75-83, E156N, G166E, G169A, S188P, F189W, Q206C, N212G, Y217H, N218D, S221C, M222P, P225N, T254A, and Q271 E. Further, in order to facilitate a fast and efficient purification after amino acid 275 a C-terminal His-tag is attached as shown in SEQ ID NO 3. The corresponding amino acid sequence is numbered according to the subtilisin BPN’ numbering scheme. Thus, in order to maintain the subtilisin BPN’ numbering for used ligases the numbering jumps from 74 to 83.

The gene coding for the ligase used for the following synthesis examples was obtained from GenScript. The genes were cloned (by GenScript) into a pUB-110 E. coli-B. subtilis shuttle vector (pBES) using the Mlul and BamHI site based vector. In the shuttle vector, the expression of the gene is under the control of the aprE promoter. The vector contained the pUB ori of replication for Bacillus and a kanamycin resistance marker. The vector also contained the ColE1 ori of replication and an ampicillin resistance marker for maintenance in E. coli. The resulting plasmid pBES-ligaseHIS was propagated in E.coli TOP10 and transformed into B. subtilis GX4935 ( trpC2 metBIO lys- 3 AnprE AaprE).

Production and purification of the ligases

A single microbial colony of B. subtilis containing a plasmid with the subtilisin variant gene of interest was inoculated in 5 mL LB with kanamycin (10 pg/mL) at 37°C in a shaking incubator. To the 30 mL Terrific Broth supplemented with antibiotic (kanamycin 10 pg/mL) and amino acids (100 mg/L Trp, 100 mg/L Met and 100 mg/L Lys) 0.6 mL of the overnight culture was added. The cells were grown for 48h at 37°C in a shaking incubator (200 rpm). The cells were harvested by centrifugation (15 min, 4,000 rpm, 4°C). The medium (30 mL) was decanted and concentrated on an Sartorius Vivaspin 15R unit (15 mL, 10 kDa MW cut-off) in two centrifugation steps (15 min, 4000 rpm, 4°C). The concentrated medium (0.5 mL) was then exchanged for buffer A (25 mM Tricine, pH 7.5, 0.5 M NaCI) in three washing/concentrating steps (14 mL buffer A, 10 min, 4,000 rpm, 4°C). For His-tag purification Talon resin (2.5 mL, Clonetech) was added to a plastic column cartridge. The resin was washed with 20 mL MilliQ water and equilibrated with 20 mL of buffer A. The crude enzyme was loaded on the column and washed with 5 mL buffer A. The enzyme was eluted with 15mL buffer B (25 mM Tricine, pH 7.5, 0.5 M NaCI, 500 mM imidazole). The elute was concentrated on Sartorius Vivaspin 15R (15 mL, 10 kDa MW cut-off) by centrifugation (15 min, 4000 rpm, 4°C) and the buffer was exchanged to 25 mM Tricine, pH 7.5 in three washing/concentrating steps (15 mL buffer, 10 min, 4, 000 rpm, 4°C).

The purity of the protein was analyzed by SDS-PAGE and the enzyme concentration was determined as described in WO2016056913 (A1). The purity was more than 90 %. The obtained aqueous solution (25 mM Tricine, pH 7.5) containing about 2 mg/mL of the obtained enzyme was used as such for the oligopeptide fragment condensations.

Enzymatic fragment condensation examples

Materials and methods

Unless stated otherwise, chemicals were obtained from commercial sources and used without further purification. In all enzymatic fragment condensations the ligase of SEQ ID:3 was used. Analytical HPLC was performed on an Agilent 1260 infinity Liquid Chromatograph, using a reversed-phase column (Phenomenex, C18, 5 mhi particle size, 250 x 4.6 mm) at 40°C. UV detection was performed at 220 nm using a UV-VIS 204 Linear spectrometer. The gradient program was: 0-25 min linear gradient ramp from 5% to 98% eluent B and from 25.1-30 min 5% eluent B (eluent A: 0.5 mL/L methane sulfonic acid (MSA) in H2O, eluent B 0.5 mL/L MSA in acetonitrile). The flow was 1 mL/min from 0-25.1 min and 2 mL/min from 25.2-29.8 min, then back to 1 mL/min until stop at 30 min. Injection volumes were 10 pL. Preparative HPLC was performed on a Varian PrepStar system using a stationary-phase column (Phenomenex, C18, 10 mhi particle size, 250 x 50 mm). LC-MS was performed on an Agilent 1200 series Liquid Chromatograph, using a reversed-phase column (Phenomenex, C18, 5 mhi particle size, 150 x 4.6 mm) at 40°C. UV detection and gradient program were as described for analytical HPLC. The molecular weights were determined using an Agilent 6130 quadrupole LC/MS system.

Protocol 1: Synthesis of Fmoc-glycolic acid

Tert-butyl 2-hydroxy-acetate (2.5 g) was dissolved in a mixture of pyridine (15 ml) and dichloromethane (DCM, 30 ml). Then Fmoc-chloride (5 g) in dry DCM (15ml) was added dropwise at 0 °C. The reaction mixture was stirred at room temperature for 24 hours. The solvent was removed under vacuum and the residue was redissolved in DCM (40 ml), washed with 1M sodium bicarbonate solution (20 mL) twice, brine solution (20 ml) twice, dried over anhydrous magnesium sulfate and concentrated. The obtained Fmoc-glycolic acid tert-butyl ester (4 g) was dissolved in trifluoroacetic acid (TFA), triisopropylsilane (TIS) and water (95/2.5/2.5, v/v/v, 15 mL) and stirred for 120 min. The solvent was removed under vacuum and the viscous residue was redissolved in 5% sodium bicarbonate solution (150 ml), washed with diethyl ether (75 ml) 3 times. The aqueous solution was then mixed with ethyl acetate (45 mL) and acidified with 40% phosphoric acid to pH = 2 at 0 °C. The organic layer was collected and dried with anhydrous magnesium sulfate. The solvent was removed under vacuum to give the final product Fmoc-glycolic acid (Fmoc-GA).

Protocol 2: Synthesis of oligopeptide-OCam-Leu-OH esters

1 gram of preloaded Fmoc-Leu-Wang resin (with a loading of 0.81 mmol/gram) was washed with DCM (2 x 2 min, 10 mL) and L/,L/'-dimethylformamide (DMF, 2 x 2 min, 10 mL) and Fmoc-deprotected using piperidine/DMF (1/5, v/v, 2 x 8 min, 10 mL). After washing with DMF (6 x 2 min, 10 mL), Fmoc-GA (4 equiv.) was coupled to the resin using 2-(1 /-/-benzotriazol-1 -yl)-1 , 1 ,3,3-tetramethyl uronium hexafluorophosphate (HBTU, 4 equiv.), OxymaPure (4 equiv.) and di-isopropylethylamine (DIPEA, 8 equiv.) in DMF (45 min, 10 mL). After washing with DMF (2 x 2 min, 10 mL) the resin was Fmoc-deprotected using piperidine/DMF (1/5, v/v, 2 x 8 min, 10 mL). The Cam-Leu-OH ester was formed by coupling of the first Fmoc-protected amino acid using 4 equiv. Fmoc-Xxx-OH, 4 equiv. N,N'- diisopropylcarbodiimide (DIC) and 0.1 equiv. 4-dimethylaminopyridine (DMAP) in DMF (2 x 60 min, 10 mL). Here and in other parts of this disclosure‘Xxx’ stands for one amino acid (variable as indicated in the sequences in the examples below). For the Semaglutide starting material a commercially available Fmoc-Aib-OH building block was used.

After washing with DMF (6 x 2 min, 10 mL), standard SPPS protocols were followed to elongate the peptide (Weng C. Chan and Peter White, OUP Oxford, 2000). Cleavage from the resin and side-chain deprotection was performed using a mixture of TFA/TIS/water (95/2.5/2.5, v/v/v, 15 mL) for 120 min. The crude peptide was precipitated using methyl ferf-butyl ether (MTBE)/n-heptanes (1/1 , v/v, 50 mL). The precipitated peptide was collected by centrifugation and washed twice with MTBE/n-heptanes (1/1 , v/v, 50 mL) followed by lyophilization from acetonitrile/water (1/1 , v/v, 50 mL). The crude products were purified by preparative HPLC followed by lyophilization of the pure fractions.

As described above, but then on a preloaded Fmoc-Arg(Pbf)-Rink resin (with a loading of 0.62 mmol/gram), several alternative peptide Cam-esters were prepared, i.e. H- ¹ His- ² Ala- ³ Glu- ⁴ Gly- ⁵ Thr- ⁶ Phe- ⁷ Thr- ⁸ Ser- ⁹ Asp- ¹⁰ Val- ¹¹ Ser-OCam-Phe-Arg-NH ₂ , H- ¹ His- ² Ala- ³ Glu- ⁴ Gly- ⁵ Thr- ⁶ Phe- ⁷ Thr- ⁸ Ser- ⁹ Asp- ¹⁰ Val- ¹¹ Ser-OCam-Leu-Arg- NH ₂ and H- ¹ His- ² Ala- ³ Glu- ⁴ Gly- ⁵ Thr- ⁶ Phe- ⁷ Thr- ⁸ Ser- ⁹ Asp- ¹⁰ Val- ¹¹ Ser-OCam-Trp-Arg-NH ₂ .

Protocol 3: Synthesis of oligopeptide C-terminal acid nucleophiles

1 gram of preloaded Fmoc-Gly-Wang resin (with a loading of 0.30 mmol/gram) was washed with DCM (2 x 2 min, 10 mL) and DMF (2 x 2 min, 10 mL) and Fmoc-deprotected using piperidine/DMF (1/5, v/v, 2 x 8 min, 10 mL). Standard SPPS protocols were followed to elongate the peptide (Weng C. Chan and Peter White, OUP Oxford, 2000). Cleavage from the resin and side-chain deprotection was performed using a mixture of TFA/TIS/water (95/2.5/2.5, v/v/v, 15 mL) for 120 min. The crude peptide was precipitated using MTBE/n-heptanes (1/1 , v/v, 50 mL). The precipitated peptide was collected by centrifugation and washed twice with MTBE/n-heptanes (1/1 , v/v, 50 mL) followed by lyophilization from acetonitrile/water (1/1 , v/v, 50 mL). The crude products were purified by preparative HPLC followed by lyophilization of the pure fractions. Protocol 4: PTC protection of H-Gly- ¹ His- ² Ala- ³ Glu- ⁴ Gly- ⁵ Thr- ⁶ Phe- ⁷ Thr- ⁸ Ser- ⁹ Asp- ¹⁰ Val- ¹¹ Ser-OCam-

Leu-OH

100 mg of H-Gly- ¹ His- ² Ala- ³ Glu- ⁴ Gly- ⁵ Thr- ⁶ Phe- ⁷ Thr- ⁸ Ser- ⁹ Asp- ¹⁰ Val- ¹¹ Ser-OCam-Leu-OH was dissolved in 10 mL pyridine/water (1/1 , v/v). To this mixture 25 mg of phenylisothiocyanate was added and the solution was stirred at ambient temperature for 14 hours. The crude reaction mixture was diluted with 50 mL water and washed three times with 50 mL dichloromethane (DCM). The water layer was purified by preparative HPLC followed by lyophilization of the pure fractions giving the PTC-Gly protected peptide

Protocol 5: Synthesis of y-Giu or Pai-y-Giu containing peptides

General protocol 3 was followed using commercially available Fmoc-Lys(Boc-y-Glu-0'Bu)-OH or Fmoc- Lys(Pal-y-Glu-0'Bu)-OH building blocks.

Protocol 6: Synthesis of the Semaglutide fragment H- ¹² Ser- ¹³ Tyr- ¹⁴ Leu- ¹⁵ Glu- ¹⁶ Gly- ¹⁷ Gln- ¹⁸ Ala- ¹⁹ Ala- ²⁰ Lys(AEEA-AEEA-Y-Glu-N-17-carboxyheptadecanoyl-OH)- ²¹ Glu- ²² Phe- ²³ lle- ²⁴ Ala- ²⁵ Trp- ²⁶ Leu- ²⁷ Val- ²⁸ Arg- ²⁹ Gly- ³⁰ Arg- ³¹ Gly-OH

General protocol 3 was followed using a commercially available Fmoc- ²⁰ Lys(Mtt)-OH and Boc- ¹² Ser('Bu)- OH building block. After SPPS of the Boc-12-31 -Wang fragment the Mtt protecting group was removed using 10 mL of TIS/TFA/DCM (1/1/48, v/v/v, 3x 15 min). Standard SPPS procedures were used for the coupling of Fmoc-AEEA-OH (twice), Fmoc-Glu-O'Bu, and 17-carboxyheptadecanoyl-O'Bu. Cleavage from the resin and side-chain deprotection was performed using a mixture of TFA/TIS/water (95/2.5/2.5, v/v/v, 15 mL) for 120 min. The crude peptide was precipitated using MTBE/n-heptane (1/1 , v/v, 50 mL). The precipitated peptide was collected by centrifugation and washed twice with MTBE/n-heptane (1/1 , v/v, 50 mL) followed by lyophilization from acetonitrile/water (1/1 , v/v, 50 mL). The crude products were purified by preparative HPLC followed by lyophilization of the pure fractions.

EXAMPLE 1 (reference):

Enzymatic synthesis of the Liraqlutide precursor H-Liraqlutide-1-31-OH using a 13-mer + 18-mer approach.

In an HPLC vial, 6 mg of H- ¹ His- ² Ala- ³ Glu- ⁴ Gly- ⁵ Thr- ⁶ Phe- ⁷ Thr- ⁸ Ser- ⁹ Asp- ¹⁰ Val- ¹¹ Ser- ¹² Ser- ¹³ Tyr-OCam- Leu-OH and 9 mg of H- ¹⁴ Leu- ¹⁵ Glu- ¹⁶ Gly- ¹⁷ Gln- ¹⁸ Ala- ¹⁹ Ala- ²⁰ Lys- ²¹ Glu- ²² Phe- ²³ lle- ²⁴ Ala- ²⁵ Trp - ²⁶ Leu- ²⁷ Val- ²⁸ Arg- ²⁹ Gly- ³⁰ Arg- ³¹ Gly-OH were dissolved in 800 mί 2M aqueous guanidinium chloride. To this mixture, 50 mί 1 M tricine buffer pH 9.0 was added and the pH was adjusted to 8.3 using a 4M NaOH solution. Subsequently, 10 mί of TCEP (tris(2- carboxyethyl)phosphine) solution (100 mg/mL in water) and 100 mί of the ligase solution (10 mg/mL) were added. The mixture was left to react at ambient temperature. Every 15 minutes, 10 mί of the reaction mixture was withdrawn and quenched in 980 mί 5 vol% MSA in acetonitrile/water (2/1 , v/v) and analysed using LC-MS.

After 90 minutes all Cam-ester starting material had been consumed, and the product and hydrolysis peaks were integrated. The ligation product H- ¹ His- ² Ala- ³ Glu- ⁴ Gly- ⁵ Thr- ⁶ Phe- ⁷ Thr- ⁸ Ser- ⁹ Asp- ¹⁰ Val- ¹¹ Ser- ¹² Ser- ¹³ Tyr- ¹⁴ Leu- ¹⁵ Glu- ¹⁶ Gly- ¹⁷ Gln- ¹⁸ Ala- ¹⁹ Ala- ²⁰ Lys- ²¹ Glu- ²² Phe- ²³ lle- ²⁴ Ala- ²⁵ Trp- ²⁶ Leu- ²⁷ Val- ²⁸ Arg- ²⁹ Gly- ³⁰ Arg- ³¹ Gly-OH was 14 area% and the hydrolysed Cam-ester H- ¹ His- ² Ala- ³ Glu- ⁴ Gly- ⁵ Thr- ⁶ Phe- ⁷ Thr- ⁸ Ser- ⁹ Asp- ¹⁰ Val- ¹¹ Ser- ¹² Ser- ¹³ Tyr-OH was

86 area%.

The product H-Liraglutide-1 -31 -OH could be obtained by preparative HPLC followed by lyophilization of the pure fractions.

EXAMPLE 2 (reference):

Enzymatic synthesis of the Liraqlutide precursor H-Liraqlutide-1-31-OH using a 9-mer + 22-mer approach.

In an HPLC vial, 6 mg of H- ¹ His- ² Ala- ³ Glu- ⁴ Gly- ⁵ Thr- ⁶ Phe- ⁷ Thr- ⁸ Ser- ⁹ Asp-OCam-Leu-OH and 10 mg of H- ¹⁰ Val- ¹¹ Ser- ¹² Ser- ¹³ Tyr- ¹⁴ Leu- ¹⁵ Glu- ¹⁶ Gly- ¹⁷ Gln- ¹⁸ Ala- ¹⁹ Ala- ²⁰ Lys- ²¹ Glu- ²² Phe- ²³ lle- ²⁴ Ala- ²⁵ Trp- ²⁶ Leu- ²⁷ Val- ²⁸ Arg- ²⁹ Gly- ³⁰ Arg- ³¹ Gly-OH were dissolved in 800 mί 2M aqueous guanidinium chloride. To this mixture, 50 mί 1 M tricine buffer pH 9.0 was added and the pH was adjusted to 8.3 using a 4M NaOH solution. Subsequently, 10 mί of TCEP solution (100 mg/mL in water) and 100 mί of ligase solution (10 mg/mL) were added. The mixture was left to react at ambient temperature. Every 15 minutes, 10 mί of the reaction mixture was quenched in 980 mί 5 vol% MSA in acetonitrile/water (2/1 , v/v) and analysed using LC-MS.

After 300 minutes all Cam-ester starting material was consumed, and the product and hydrolysis peaks were integrated. The ligation product H- ¹ His- ² Ala- ³ Glu- ⁴ Gly- ⁵ Thr- ⁶ Phe- ⁷ Thr- ⁸ Ser- ⁹ Asp- ¹⁰ Val- ¹¹ Ser- ¹² Ser- ¹³ Tyr- ¹⁴ Leu- ¹⁵ Glu- ¹⁶ Gly- ¹⁷ Gln- ¹⁸ Ala- ¹⁹ Ala- ²⁰ Lys- ²¹ Glu- ²² Phe- ²³ lle- ²⁴ Ala- ²⁵ Trp- ²⁶ Leu- ²⁷ Val- ²⁸ Arg- ²⁹ Gly- ³⁰ Arg- ³¹ Gly-OH was 8 area% and the hydrolysed Cam-ester H- ¹ His- ² Ala- ³ Glu- ⁴ Gly- ⁵ Thr- ⁶ Phe- ⁷ Thr- ⁸ Ser- ⁹ Asp-OH was 92 area%.

The product H-Liraglutide-1 -31 -OH could be obtained by preparative HPLC followed by lyophilization of the pure fractions.

EXAMPLE 3: Enzymatic synthesis of H-Liraqlutide-1-31-OH using an 11-mer + 20-mer approach.

In an HPLC vial, 6 mg of H- ¹ His- ² Ala- ³ Glu- ⁴ Gly- ⁵ Thr- ⁶ Phe- ⁷ Thr- ⁸ Ser- ⁹ Asp- ¹⁰ Val- ¹¹ Ser-OCam-Leu-OH and 10 mg of H- ¹² Ser- ¹³ Tyr- ¹⁴ Leu- ¹⁵ Glu- ¹⁶ Gly- ¹⁷ Gln- ¹⁸ Ala- ¹⁹ Ala- ²⁰ Lys- ²¹ Glu- ²² Phe- ²³ lle- ²⁴ Ala- ²⁵ Trp- ²⁶ Leu- ²⁷ Val- ²⁸ Arg- ²⁹ Gly- ³⁰ Arg- ³¹ Gly-OH were dissolved in 800 mί 2M aqueous guanidinium chloride. To this mixture, 50 mί 1 M tricine buffer pH 9.0 was added and the pH was adjusted to 8.3 using a 4M NaOH solution. Subsequently, 10 mί of TCEP solution (100 mg/mL in water) and 100 mί of ligase solution (10 mg/mL) were added. The mixture was left to react at ambient temperature. Every 15 minutes, 10 mί of the reaction mixture was withdrawn and quenched in 980 mί 5 vol% MSA in acetonitrile/water (2/1 , v/v) and analysed using LC-MS.

After 180 minutes all Cam-ester starting material had been consumed, and the product and hydrolysis peaks were integrated. The ligation product H- ¹ His- ² Ala- ³ Glu- ⁴ Gly- ⁵ Thr- ⁶ Phe- ⁷ Thr- ⁸ Ser- ⁹ Asp- ¹⁰ Val- ¹¹ Ser- ¹² Ser- ¹³ Tyr- ¹⁴ Leu- ¹⁵ Glu- ¹⁶ Gly- ¹⁷ Gln- ¹⁸ Ala- ¹⁹ Ala- ²⁰ Lys- ²¹ Glu- ²² Phe- ²³ lle- ²⁴ Ala- ²⁵ Trp- ²⁶ Leu- ²⁷ Val- ²⁸ Arg- ²⁹ Gly- ³⁰ Arg- ³¹ Gly-OH was 96 area% and the hydrolysed Cam-ester H- ¹ His- ² Ala- ³ Glu- ⁴ Gly- ⁵ Thr- ⁶ Phe- ⁷ Thr- ⁸ Ser- ⁹ Asp- ¹⁰ Val- ¹¹ Ser- ¹² Ser- ¹³ Tyr-OH was 4 area%. The product H-Liraglutide-1 -31 -OH could be obtained by preparative HPLC followed by lyophilization of the pure fractions.

It is thus shown that a method according to the invention unexpectedly provides the desired coupled product (i.e. H-Liraglutide-1 -31 -OH, the peptide having the amino acid sequence of Liraglutide without derivatization at ²⁰ Lys) at a much higher yield than in a comparable method wherein the coupling site for the C-terminal (thio)ester and peptide nucleophile is two amide bonds towards the C-terminus or N-terminus of the H-Liraglutide-1-31-0H.

EXAMPLE 4: Enzymatic synthesis of PTC-Glv-Liraglutide-1-31-f ²⁰ Lvs(v-Glu)l-OH using an 1 1-mer + 20-mer approach

In an HPLC vial, 6 mg of PTC-Gly- ¹ His- ² Ala- ³ Glu- ⁴ Gly- ⁵ Thr- ⁶ Phe- ⁷ Thr- ⁸ Ser- ⁹ Asp- ¹⁰ Val- ¹¹ Ser-OCam-Leu- OH and 10 mg of H- ¹² Ser- ¹³ Tyr- ¹⁴ Leu- ¹⁵ Glu- ¹⁶ Gly- ¹⁷ Gln- ¹⁸ Ala- ¹⁹ Ala- ²⁰ Lys(Y-Glu)- ²¹ Glu- ²² Phe- ²³ lle- ²⁴ Ala- ²⁵ Trp- ²⁶ Leu- ²⁷ Val- ²⁸ Arg- ²⁹ Gly- ³⁰ Arg- ³¹ Gly-OH were dissolved in 800 mί 2M aqueous guanidinium chloride.

To this mixture, 50 mί 1 M tricine buffer pH 9.0 was added and the pH was adjusted to 8.3 using a 4M NaOH solution. Subsequently, 10 mί of TCEP solution (100 mg/mL in water) and 100 mί of ligase solution (10 mg/mL) were added. The mixture was left to react at ambient temperature. Every 15 minutes, 10 mί of the reaction mixture was withdrawn and quenched in 980 mί 5 vol% MSA in acetonitrile/water (2/1 , v/v) and analysed using LC-MS.

After 180 minutes all Cam-ester starting material had been consumed, and the product and hydrolysis peaks were integrated. The ligation product PTC-Gly- ¹ His- ² Ala- ³ Glu- ⁴ Gly- ⁵ Thr- ⁶ Phe- ⁷ Thr- ⁸ Ser- ⁹ Asp- ¹⁰ Val- ¹¹ Ser- ¹² Ser- ¹³ Tyr- ¹⁴ Leu- ¹⁵ Glu- ¹⁶ Gly- ¹⁷ Gln- ¹⁸ Ala- ¹⁹ Ala- ²⁰ Lys(Y-Glu)- ²¹ Glu- ²² Phe- ²³ lle- ²⁴ Ala- ²⁵ Trp- ²⁶ Leu- ²⁷ Val- ²⁸ Arg- ²⁹ Gly- ³⁰ Arg- ³¹ Gly-

OH was 93 area% and the hydrolysed Cam-ester PTC-Gly- ¹ His- ² Ala- ³ Glu- ⁴ Gly- ⁵ Thr- ⁶ Phe- ⁷ Thr- ⁸ Ser- ⁹ Asp- ¹⁰ Val- ¹¹ Ser- OH was 7 area%.

The product PTC-Gly-Liraglutide-1-31-[ ²⁰ Lys(y-Glu)]-OH could be obtained by preparative HPLC followed by lyophilization of the pure fractions.

EXAMPLE 5: Synthesis of H-Liraql utide-1 -31 -I I-OH using the PTC-' !-1-31-P°Lys(v-Glu)l-

OH precursor from EXAMPLE 4.

2 mg of PTC-Gly- ¹ His- ² Ala- ³ Glu- ⁴ Gly- ⁵ Thr- ⁶ Phe- ⁷ Thr- ⁸ Ser- ⁹ Asp- ¹⁰ Val- ¹¹ Ser - ¹² Ser- ¹³ Tyr- ¹⁴ Leu- ¹⁵ Glu- ¹⁶ Gly- ¹⁷ Gln- ¹⁸ Ala- ¹⁹ Ala- ²⁰ Lys(Y-Glu)- ²¹ Glu- ²² Phe- ²³ lle- ²⁴ Ala- ²⁵ Trp - ²⁶ Leu- ²⁷ Val- ²⁸ Arg- ²⁹ Gly- ³⁰ Arg- ³¹ Gly-OH was dissolved in 500 mί water and 500 mί pyridine. To this solution, 2 mg of palmitic acid N-hydroxy succinimide ester (Pal-OSu) was added and the mixture was left to react at ambient temperature for 5 hours followed by evaporation of the solvents in vacuo. The crude product PTC-Gly- ¹ His- ² Ala- ³ Glu- ⁴ Gly- ⁵ Thr- ⁶ Phe- ⁷ Thr- ⁸ Ser- ⁹ Asp- ¹⁰ Val- ¹¹ Ser- ¹² Ser- ¹³ Tyr- ¹⁴ Leu- ¹⁵ Glu- ¹⁶ Gly- ¹⁷ Gln- ¹⁸ Ala- ¹⁹ Ala- ²⁰ Lys(Pal-Y-Glu)- ²¹ Glu- ²² Phe- ²³ lle- ²⁴ Ala- ²⁵ Trp- ²⁶ Leu- ²⁷ Val- ²⁸ Arg- ²⁹ Gly- ³⁰ Arg- ³¹ Gly-OH was dissolved in 5 vol% trifluoroacetic acid in water for cleavage (deprotection) of the PTC-Gly group.

After completion (15 min), the product H- ¹ His- ² Ala- ³ Glu- ⁴ Gly- ⁵ Thr- ⁶ Phe- ⁷ Thr- ⁸ Ser- ⁹ Asp- ¹⁰ Val- ¹¹ Ser - ¹² Ser- ¹³ Tyr- ¹⁴ Leu- ¹⁵ Glu- ¹⁶ Gly- ¹⁷ Gln- ¹⁸ Ala- ¹⁹ Ala- ²⁰ Lys(Pal-Y-Glu)- ²¹ Glu- ²² Phe- ²³ lle- ²⁴ Ala- ²⁵ Trp- ²⁶ Leu- ²⁷ Val- ²⁸ Arg- ²⁹ Gly- ³⁰ Arg- ³¹ Gly-OH was obtained and purified by preparative HPLC followed by lyophilization of the pure fractions. EXAMPLE 6: Enzymatic synthesis of H-Liraqlutide-1-31-[ ²⁰ Lvs(Pal-v-Glu)1-OH using an 11-mer + 20-mer approach

(Pal = palmitoyl).

In an HPLC vial, 6 mg of H- ¹ His- ² Ala- ³ Glu- ⁴ Gly- ⁵ Thr- ⁶ Phe- ⁷ Thr- ⁸ Ser- ⁹ Asp- ¹⁰ Val- ¹¹ Ser-OCam-Leu-OH and 10 mg of H- ¹² Ser- ¹³ Tyr- ¹⁴ Leu- ¹⁵ Glu- ¹⁶ Gly- ¹⁷ Gln- ¹⁸ Ala- ¹⁹ Ala- ²⁰ Lys(Pal-Y-Glu)- ²¹ Glu- ²² Phe- ²³ lle- ²⁴ Ala- ²⁵ Trp- ²⁶ Leu- ²⁷ Val- ²⁸ Arg- ²⁹ Gly- ³⁰ Arg- ³¹ Gly-OH were dissolved in 950 mί water. To this mixture, 50 mί 1 M tricine buffer pH 9.0 was added and the pH was adjusted to 8.1 using a 3M NaOH solution. Subsequently, 10 mί of TCEP solution (100 mg/mL in water) and 100 mί of ligase solution (10 mg/mL) were added. The mixture was left to react at ambient temperature. Every 15 minutes, 10 mί of the reaction mixture was withdrawn and quenched in 980 mί 5 vol% MSA in acetonitrile/water (2/1 , v/v) and analysed using LC-MS.

After 180 minutes all Cam-ester starting material had been consumed and the product and hydrolysis peaks were integrated. The ligation product H- ¹ His- ² Ala- ³ Glu- ⁴ Gly- ⁵ Thr- ⁶ Phe- ⁷ Thr- ⁸ Ser- ⁹ Asp- ¹⁰ Val- ¹¹ Ser- ¹² Ser- ¹³ Tyr- ¹⁴ Leu- ¹⁵ Glu- ¹⁶ Gly- ¹⁷ Gln- ¹⁸ Ala- ¹⁹ Ala- ²⁰ Lys(Pal-Y-Glu)- ²¹ Glu- ²² Phe- ²³ lle- ²⁴ Ala- ²⁵ Trp- ²⁶ Leu- ²⁷ Val- ²⁸ Arg- ²⁹ Gly- ³⁰ Arg- ³¹ Gly- OH was 95 area% and the hydrolysed Cam-ester H- ¹ His- ² Ala- ³ Glu- ⁴ Gly- ⁵ Thr- ⁶ Phe- ⁷ Thr- ⁸ Ser- ⁹ Asp- ¹⁰ Val- ¹¹ Ser-OH was 5 area%.

This Example shows that the present invention allows the direct synthesis of Liraglutide in a single enzymatic coupling step in high yield.

An identical ligation reaction as described above was performed except for using several alternative peptide Cam-esters, i.e. (1) H- ¹ His- ² Ala- ³ Glu- ⁴ Gly- ⁵ Thr- ⁶ Phe- ⁷ Thr- ⁸ Ser- ⁹ Asp- ¹⁰ Val- ¹¹ Ser-OCam-Phe-Arg-NH2, (2) H- ¹ His- ² Ala- ³ Glu- ⁴ Gly- ⁵ Thr- ⁶ Phe- ⁷ Thr- ⁸ Ser- ⁹ Asp- ¹⁰ Val- ¹¹ Ser-OCam-Leu-Arg-NH ₂ and (3) H- ¹ His- ² Ala- ³ Glu- ⁴ Gly- ⁵ Thr- ⁶ Phe- ⁷ Thr- ⁸ Ser- ⁹ Asp- ¹⁰ Val- ¹¹ Ser-OCam-Trp-Arg-NH2. In general the reactions proceeded faster than for the H- ¹ His- ² Ala- ³ Glu- ⁴ Gly- ⁵ Thr- ⁶ Phe- ⁷ Thr- ⁸ Ser- ⁹ Asp- ¹⁰ Val- ¹¹ Ser-OCam-Leu-OH (180 min) ester resulting in a complete conversion after 80 min for (1), 100 min for (2) and 85 min for (3).

EXAMPLE 7: Synthesis of H-Liraqlutide-1-31- (Pal-v-Glu)1-OH from the enzymatically synthesized precursor H-

Liraqlutide-1-31-OH of EXAMPLE 3.

To the precursor H- ¹ His- ² Ala- ³ Glu- ⁴ Gly- ⁵ Thr- ⁶ Phe- ⁷ Thr- ⁸ Ser- ⁹ Asp- ¹⁰ Val- ¹¹ Ser- ¹² Ser- ¹³ Tyr- ¹⁴ Leu- ¹⁵ Glu- ¹⁶ Gly- ¹⁷ Gln- ¹⁸ Ala- ¹⁹ Ala- ²⁰ Lys- ²¹ Glu- ²² Phe- ²³ lle- ²⁴ Ala- ²⁵ Trp- ²⁶ Leu- ²⁷ Val- ²⁸ Arg- ²⁹ Gly- ³⁰ Arg- ³¹ Gly-OH, the Pal-y-Glu moiety is coupled to obtain the product H- ¹ His- ² Ala- ³ Glu- ⁴ Gly- ⁵ Thr- ⁶ Phe- ⁷ Thr- ⁸ Ser- ⁹ Asp- ¹⁰ Val- ¹¹ Ser- ¹² Ser- ¹³ Tyr- ¹⁴ Leu- ¹⁵ Glu- ¹⁶ Gly- ¹⁷ Gln- ¹⁸ Ala- ¹⁹ Ala- ²⁰ Lys(Pal-Y-Glu)- ²¹ Glu- ²² Phe- ²³ lle- ²⁴ Ala- ²⁵ Trp- ²⁶ Leu- ²⁷ Val- ²⁸ Arg- ²⁹ Gly- ³⁰ Arg- ³¹ Gly-OH using protocols described in US 6451974 B1.

EXAMPLE 8: Synthesis of H-Semaqlutide-1-31-[ ²⁰ Lvs(AEEA-AEEA-y-Glu-N-17-carboxyheptadecanoyl-OH)1-OH using an 11-mer + 20-mer approach.

In an HPLC vial, 6 mg of H- ¹ His- ² Aib- ³ Glu- ⁴ Gly- ⁵ Thr- ⁶ Phe- ⁷ Thr- ⁸ Ser- ⁹ Asp- ¹⁰ Val- ¹¹ Ser-OCam-Leu-OH and 10 mg of H- ¹² Ser- ¹³ Tyr- ¹⁴ Leu- ¹⁵ Glu- ¹⁶ Gly- ¹⁷ Gln- ¹⁸ Ala- ¹⁹ Ala- ²⁰ Lys(AEEA-AEEA-y-Glu-N-17-carboxyheptadecanoyl- OH)- ²¹ Glu- ²² Phe- ²³ lle- ²⁴ Ala- ²⁵ Trp - ²⁶ Leu- ²⁷ Val- ²⁸ Arg- ²⁹ Gly- ³⁰ Arg- ³¹ Gly-OH were dissolved in 950 mί water. To this mixture, 50 mί 1 M tricine buffer pH 9.0 was added and the pH was adjusted to 8.1 using a 3M NaOH solution. Subsequently, 10 mί of TCEP solution (100 mg/mL in water) and 100 mί of ligase solution (10 mg/mL) were added. The mixture was left to react at ambient temperature. Every 15 minutes, 10 m L of the reaction mixture was withdrawn and quenched in 980 mί 5 vol% MSA in acetonitrile/water (2/1 , v/v) and analysed using LC-MS.

After 180 minutes all Cam-ester starting material had been consumed and the product and hydrolysis peaks were integrated. The ligation product H- ¹ His- ² Aib- ³ Glu- ⁴ Gly- ⁵ Thr- ⁶ Phe- ⁷ Thr- ⁸ Ser- ⁹ Asp- ¹⁰ Val- ¹¹ Ser- ¹² Ser- ¹³ Tyr- ¹⁴ Leu- ¹⁵ Glu- ¹⁶ Gly- ¹⁷ Gln- ¹⁸ Ala- ¹⁹ Ala- ²⁰ Lys(AEEA-AEEA-y-Glu-N-17-carboxyheptadecanoyl-OH)- ²¹ Glu- ²² Phe- ²³ lle- ²⁴ Ala- ²⁵ Trp- ²⁶ Leu- ²⁷ Val- ²⁸ Arg- ²⁹ Gly- ³⁰ Arg- ³¹ Gly-OH was 96 area% and the hydrolysed Cam-ester H- ¹ His- ² Ala- ³ Glu- ⁴ Gly- ⁵ Thr- ⁶ Phe- ⁷ Thr- ⁸ Ser- ⁹ Asp- ¹⁰ Val- ¹¹ Ser-OH was 4 area%.

This Example shows that the present invention allows the direct synthesis of Semaglutide in a single enzymatic coupling step in high yield.

EXAMPLE 9: Enzymatic synthesis of H-l_iraqlutide-1-31-OH using an 11-mer + 20-mer approach with enzyme variant

SEQ ID NO 14.

In an HPLC vial, 10 mg of H- ¹ His- ² Ala- ³ Glu- ⁴ Gly- ⁵ Thr- ⁶ Phe- ⁷ Thr- ⁸ Ser- ⁹ Asp- ¹⁰ Val- ¹¹ Ser-OCam-Leu-OH and 10 mg of H- ¹² Ser- ¹³ Tyr- ¹⁴ Leu- ¹⁵ Glu- ¹⁶ Gly- ¹⁷ Gln- ¹⁸ Ala- ¹⁹ Ala- ²⁰ Lys- ²¹ Glu- ²² Phe- ²³ lle- ²⁴ Ala- ²⁵ Trp- ²⁶ Leu- ²⁷ Val- ²⁸ Arg- ²⁹ Gly- ³⁰ Arg- ³¹ Gly-OH were dissolved in 500 mί 50 mM tricine buffer pH 9.0. The pH was adjusted to 8.3 using a 3 M NaOH solution. Subsequently, 10 mί of TCEP solution (100 mg/mL in water) and 100 mί of ligase solution (10 mg/mL) were added. The mixture was left to react at ambient temperature. Every 15 minutes, 10 mί of the reaction mixture was withdrawn and quenched in 980 mί 5 vol% MSA in acetonitrile/water (2/1, v/v) and analysed using LC-MS.

After 240 minutes all Cam-ester starting material had been consumed, and the product and hydrolysis peaks were integrated. The ligation product H- ¹ His- ² Ala- ³ Glu- ⁴ Gly- ⁵ Thr- ⁶ Phe- ⁷ Thr- ⁸ Ser- ⁹ Asp- ¹⁰ Val- ¹¹ Ser- ¹² Ser- ¹³ Tyr- ¹⁴ Leu- ¹⁵ Glu- ¹⁶ Gly- ¹⁷ Gln- ¹⁸ Ala- ¹⁹ Ala- ²⁰ Lys- ²¹ Glu- ²² Phe- ²³ lle- ²⁴ Ala- ²⁵ Trp- ²⁶ Leu- ²⁷ Val- ²⁸ Arg- ²⁹ Gly- ³⁰ Arg- ³¹ Gly-OH was 92 area% and the hydrolysed Cam-ester H- ¹ His- ² Ala- ³ Glu- ⁴ Gly- ⁵ Thr- ⁶ Phe- ⁷ Thr- ⁸ Ser- ⁹ Asp- ¹⁰ Val- ¹¹ Ser-OH was 8 area%.

The product H-Liraglutide-1 -31 -OH could be obtained by preparative HPLC followed by lyophilization of the pure fractions.

EXAMPLE 10: Identification of fragments suitable for the enzymatic synthesis of liraqlutide using serine endoprotease derived from subtilisin BPN’ variant with SEQ ID NO: 3 (C221S mutated)

1 mg of liraglutide 1-31 (with and without the Pal-y-Glu at the lysine in position 20) was dissolved in 1 mL of tricine buffer (50 mM, pH = 8.0). To this mixture, 1 mί of endoprotease solution (1 mg/mL) was added and the reaction mixture was stirred at room temperature.

Hydrolytic activity was monitored by analysing samples every 30 min using LC-MS analysis.

For both peptides, the highest hydrolytic activity was observed at the bond between amino acid 25 and amino acid 26 (yielding the fragments 1-25 + 26-31), followed by hydrolytic activity at the bond between amino acid 5 and 6 (yielding the fragments 1-5 + 6-31). The goodness of a 25-mer + 6-mer approach was tested in comparison EXAMPLE 11.

EXAMPLE 11 : Enzymatic synthesis of the Liraqlutide precursor H-Liraqlutide-1-31-OH using in parallel a 25-mer + 6- mer approach (reference), 5-mer + 26-mer approach (reference), or 11-mer + 20-mer approach (according to the invention).

In an HPLC vial, 3 mol of ester (1 : H- ¹ His- ² Ala- ³ Glu- ⁴ Gly- ⁵ Thr- ⁶ Phe- ⁷ Thr- ⁸ Ser- ⁹ Asp- ¹⁰ Val- ¹¹ Ser ¹² Ser- ¹³ Tyr- ¹⁴ Leu- ¹⁵ Glu- ¹⁶ Gly- ¹⁷ Gln- ¹⁸ Ala- ¹⁹ Ala- ²⁰ Lys(Pal-Y-Glu)- ²¹ Glu- ²² Phe- ²³ lle- ²⁴ Ala- ²⁵ Trp-OCam-Leu-OH, 2: H- ¹ His- ² Ala- ³ Glu- ⁴ Gly- ⁵ Thr-OCam-Leu-OH or 3: H- ¹ His- ² Ala- ³ Glu- ⁴ Gly- ⁵ Thr- ⁶ Phe- ⁷ Thr- ⁸ Ser- ⁹ Asp- ¹⁰ Val- ¹¹ Ser-OCam-Leu-OH) and 2 Mmol amine (1 : H- ²⁶ Leu- ²⁷ Val- ²⁸ Arg- ²⁹ Gly- ³⁰ Arg- ³¹ Gly-OH, 2: H- ⁶ Phe- ⁷ Thr- ⁸ Ser- ⁹ Asp- ¹⁰ Val- ¹¹ Ser ¹² Ser- ¹³ Tyr- ¹⁴ Leu- ¹⁵ Glu- ¹⁶ Gly- ¹⁷ Gln- ¹⁸ Ala- ¹⁹ Ala- ²⁰ Lys(Pal-Y-Glu)- ²¹ Glu- ²² Phe- ²³ lle- ²⁴ Ala- ²⁵ Trp- ²⁶ Leu- ²⁷ Val- ²⁸ Arg- ²⁹ Gly- ³⁰ Arg- ³¹ Gly-OH or 3: H- ¹² Ser- ¹³ Tyr- ¹⁴ Leu- ¹⁵ Glu- ¹⁶ Gly- ¹⁷ Gln- ¹⁸ Ala- ¹⁹ Ala- ²⁰ Lys(Pal-Y-Glu)- ²¹ Glu- ²² Phe- ²³ lle- ²⁴ Ala- ²⁵ Trp- ²⁶ Leu- ²⁷ Val- ²⁸ Arg- ²⁹ Gly- ³⁰ Arg- ³¹ Gly-OH) were dissolved in 950 ML water. To this mixture, 50 ML 1 M tricine buffer pH 9.0 was added and the pH was adjusted to 8.1 using a 3M NaOH solution. Subsequently, 10 ML of TCEP solution (100 mg/mL in water) and 100 ML of ligase (according to SEQ ID NO: 3) solution (10 mg/mL) were added. The mixture was left to react at ambient temperature. Every 15 minutes, 10 ML of the reaction mixture was withdrawn and quenched in 980 ML 5 vol% MSA in acetonitrile/water (2/1 , v/v) and analysed using LC-MS.

After 180 minutes, the Cam-ester starting material had been consumed for each of the three reactions. Samples of the resultant product mixture were analyzed and the product and hydrolysis peaks were integrated. The ligation product for 1 : was 72 area%, for 2: 53 area% and for 3: 95 area%.

SEQUENCES

SEQ ID NO 1 : wild type gene encoding for subtilisin BPN’ amino acids -107 to 275

E N A| K024961 K02496.1 B. Subtilisin BPN' Bacillus amyloliquefaciens

GTGAGAGGCAAAAAAGTATGGATCAGTTTGCTGTTTGCTTTAGCGTTAATCTTTACGATG GCGTTCGGCAGCACA

TCCTCTGCCCAGGCGGCAGGGAAATCAAACGGGGAAAAGAAATATATTGTCGGGTTT AAACAGACAATGAGCAC

GATGAGCGCCGCTAAGAAGAAAGATGTCATTTCTGAAAAAGGCGGGAAAGTGCAAAA GCAATTCAAATATGTAG

ACGCAGCTTCAGCTACATTAAACGAAAAAGCTGTAAAAGAATTGAAAAAAGACCCGA GCGTCGCTTACGTTGAAG

AAGATCACGTAGCACATGCGTACGCGCAGTCCGTGCCTTACGGCGTATCACAAATTA AAGCCCCTGCTCTGCAC

TCTCAAGGCTACACTGGATCAAATGTTAAAGTAGCGGTTATCGACAGCGGTATCGAT TCTTCTCATCCTGATTTA

AAGGTAGCAGGCGGAGCCAGCAT GGTT CCTT CT GAAACAAAT CCTTT CCAAGACAACAACT CT CACGGAACT CA

CGTTGCCGGCACAGTTGCGGCTCTTAATAACTCAATCGGTGTATTAGGCGTTGCGCC AAGCGCATCACTTTACG

CT GTAAAAGTTCT CGGT GCT GACGGTT CCGGCCAATACAGCT GGAT CATTAACGGAAT CGAGT GGGCGAT CGCA

AACAATATGGACGTTATTAACATGAGCCTCGGCGGACCTTCTGGTTCTGCTGCTTTA AAAGCGGCAGTTGATAAA

GCCGTTGCATCCGGCGTCGTAGTCGTTGCGGCAGCCGGTAACGAAGGCACTTCCGGC AGCTCAAGCACAGTG

GGCTACCCTGGTAAATACCCTTCTGTCATTGCAGTAGGCGCTGTTGACAGCAGCAAC CAAAGAGCATCTTTCTC

AAGCGTAGGACCTGAGCTTGATGTCATGGCACCTGGCGTATCTATCCAAAGCACGCT TCCTGGAAACAAATACG

GGGCGTACAACGGTACGTCAATGGCATCTCCGCACGTTGCCGGAGCGGCTGCTTTGA TTCTTTCTAAGCACCC

GAACT GGACAAACACT CAAGT CCGCAGCAGTTTAGAAAACACCACTACAAAACTTGGT GATT CTTT CTACTATGG

AAAAGGGCTGATCAACGTACAGGCGGCAGCTCAGTAA

SEQ ID NO 2: wild type subtilisin BPN’ (mature)

>SUBT_BACAM Subtilisin BPN' Bacillus amyloliquefaciens mature 1 to 275

>sp|P00782| 108-382

AQSVPYGVSQIKAPALHSQGYTGSNVKVAVIDSGIDSSHPDLKVAGGASMVPSETNPFQD NNSHGTHVAGTVAALNN

SIGVLGVAPSASLYAVKVLGADGSGQYSWIINGIEWAIANNMDVINMSLGGPSGSAA LKAAVDKAVASGWWAAAGN

EGTSGSSSTVGYPGKYPSVIAVGAVDSSNQRASFSSVGPELDVMAPGVSIQSTLPGN KYGAYNGTSMASPHVAGAA

ALILSKHPNWTNTQVRSSLENTTTKLGDSFYYGKGLINVQAAAQ

SEQ ID NO 3: subtilisin BPN’ variant having mutations Q2K, S3C, P5S, S9A, 131 L, K43N, M50F, N62A, A73L, D75- 83, E156N, G166E, G169A, S188P, F189W, Q206C, N212G, Y217H, N218D, S221 C, M222P, P225N, T254A, and Q271 E and a His tag

AKCVSYGVAQIKAPALHSQGYTGSNVKVAVLDSGIDSSHPDLNVAGGASFVPSETNPFQD NASHGTHVAGTVLAVAP

SASLYAVKVLGADGSGQYSWIINGIEWAIANNMDVINMSLGGPSGSAALKAAVDKAV ASGWWAAAGNNGTSGSSS

TVEYPAKYPSVIAVGAVDSSNQRAPWSSVGPELDVMAPGVSICSTLPGGKYGAHDGT CPASNHVAGAAALILSKHPN

WTNTQVRSSLENTATKLGDSFYYGKGLINVEAAAQHHHHHH

SEQ ID NO 14: subtilisin BPN’ variant having mutations Q2K, S3C, P5S, S9A, 131 L, K43N, M50F, A73L, D75-83, E156S, G166S, G169A, S188P, Q206C, N212G, Y217H, S221 C, M222P, P225N, T254A, and Q271 E and a His tag

AKCVSYGVAQIKAPALHSQGYTGSNVKVAVLDSGIDSSHPDLNVAGGASFVPSETNP FQDNNSHGTHVAGTVLAVAP

SASLYAVKVLGADGSGQYSWIINGIEWAIANNMDVINMSLGGPSGSAALKAAVDKAV ASGWWAAAGNSGTSGSSS

TVSYPAKYPSVIAVGAVDSSNQRAPFSSVGPELDVMAPGVSICSTLPGGKYGAHSGT CPASNHVAGAAALILSKHPN

WTNTQVRSSLENTATKLGDSFYYGKGLINVEAAAQHHHHHH