Enzymatic process for producing n-acetyl galactosamine clusters

Abstract

The invention relates to a novel process for generating N-acetylgalactosamine (GalNAc) clusters using a nitrilase and conjugating said GalNAc clusters to oligonucleotides, such as oligonucleotides for use in therapy. In particular, the nitrilase is used to catalyse the conversion of a trinitrile intermediate to a triacid intermediate.

Description

[0001] ENZYMATIC PROCESS FIELD OF THE INVENTION The invention relates to a novel process for generating N-acetylgalactosamine(GalNAc) clusters using a nitrilase and conjugating said GalNAc clusters to oligonucleotides,such as oligonucleotides for use in therapy.BACKGROUND Oligonucleotides have emerged as an important new class of therapeutic drugswhich have the ability to treat a wide range of diseases. They can silence mRNA via anantisense mechanism, triggering exquisite selectivity for genes of interest and blocking the transfer of genetic information from DNA to protein, or recruit the RNA-induced silencingcomplex and downregulate protein expression, e.g., siRNAs (Stein and Castanotto, 2017,Mol. Ther.25(5), 1069-1075; Mullard, 2018, Nat. Rev. Drug Discov.17(9), 613; Roberts et al, 2020, Nat. Rev. Drug Discov.19, 673-694). Unlike small molecular agents which are lipophilic and readily taken up by tissuesand cells, the majority of oligonucleotide therapeutics are larger in size, more hydrophilic andtheir cellular uptake remains challenging.The liver is the largest internal organ in the human body, playing vital functions in metabolism, detoxification, and iron homeostasis, as well as synthesis and secretion of major plasma proteins, with over 80% of its mass composed of hepatocytes. It is therefore not surprising that large amounts of disease targets reside within the liver hepatocytes (Cui et al, 2021, ACS Omega, 6, 16259-16265). N-acetylgalactosamine (GalNAc) binds to asialoglycoprotein receptors (ASGPRs)(KD ≈ 2.5 nM) on hepatocytes and conjugation of oligonucleotides to clusters of GalNAcunits more effectively targets the drug to the liver and enables hepatocyte uptake via ASGPreceptor-mediated endocytosis (Huang, 2017, Mol. Ther. Nucl. Acids 6, 116-132; Prakash etal, 2014, 42(13), 8796-8807). This can lead to lower doses, thereby increasing thetherapeutic index and improving the affordability of the medicine by reducing cost of goods (Wang et al, 2019, Excerpt Opin. Drug Met.15(6), 475-485; Weng et al, 2019 Biotechnol.Adv. 37(5), 801-825).Alnylam’s Givosiran was the first GalNAc oligonucleotide conjugate to be registeredand was approved for use in the treatment of acute hepatic porphyria (AHP) in the US in2019 and in Europe in 2020. Lumasiran and Inclisiran were subsequently approved forhyperoxaluria type 1 (PH1) and hypercholesterolemia, respectively, with many otherGalNAc-conjugated oligonucleotides in late-stage clinical trials for liver-based diseases, including hepatitis B infection, transthyretin-mediated amyloidosis and α-1-antitypsindeficiency (Cui et al, 2021, ACS Omega, 6, 16259-16265).Structurally, GalNAc clusters are designed to give the appropriate spatial arrangement for optimal binding to the hepatocyte receptor, with the optimal presentation distance between each GalNAc thought to be 15–20 Å (Debacker et al, 2020, Mol. Ther. 28(8), 1759-1771). GalNAc clusters may be divalent, trivalent or tetravalent, although trivalent GalNAcs are the most commonly used. GalNAc clusters have a high degree of entropy and are therefore not typicallycrystalline. Crystallisation is often exploited by synthetic chemists to reduce the levels ofimpurities but, for GalNAc clusters, alternative strategies are required to ensure quality,especially if it is desirable to avoid chromatographic purification. Current methods for makingGalNAc clusters are multi-step chemical reaction schemes with various difficult to separateby-products (Prakash et al, 2016, J. Med. Chem., 59(6), 2718-2733; Kim et al, 2023, Org,Process. Res. Dev., https: / / doi.org / 10.1021 / acs.oprd.3c00281). Accordingly, there is a need for a simpler, more cost-effective means of generatingGalNAc clusters. SUMMARY OF THE INVENTION In the first aspect, the present invention provides use of a nitrilase in the manufacture of an N-acetylgalactosamine (GalNAc) cluster. In another aspect, the present invention provides use of a nitrilase to form a compound of formula (I): from a compound of formula (II): , wherein R is a linker group suitable for attaching to a therapeutic oligonucleotide. In another aspect, the present invention provides a method of producing a compound of formula (I): by reacting a compound of formula (II): with a nitrilase under suitable conditions, wherein R is a linker group suitable for attaching toa therapeutic oligonucleotide. In another aspect, the present invention provides a method of producing a compound of formula (III): by reacting a compound of formula (IV): with a nitrilase under suitable conditions, wherein L is a leaving group such as C1-20alkyl, C1- 20alkoxy or benzyl. In another aspect, the present invention provides a method of manufacturing aGalNAc-conjugated oligonucleotide comprising conjugating a GalNAc cluster to said oligonucleotide under suitable conditions, wherein said GalNAc cluster is made using a nitrilase. In another aspect, the present invention provides a GalNAc-conjugatedoligonucleotide for use in therapy, wherein the GalNAc-conjugated oligonucleotide comprises a GalNAc cluster made using a nitrilase. In another aspect, the present invention provides a method of producing a tris-based GalNAc cluster compound of formula (V): wherein R is a linker group suitable for attaching to a therapeutic oligonucleotide and A is a group comprising GalNAc, the method comprising reacting a compound of formula (II): with a nitrilase under suitable conditions to obtain a triacid intermediate, then reacting the triacid intermediate with a compound A-L2 wherein L2 is a leaving group. In another aspect, the invention provides a GalNAc-conjugated therapeuticoligonucleotide obtained as described herein. In another aspect, the present invention provides an engineered nitrilase comprising a sequence that is at least 90% identical to SEQ ID NO:1 and includes a catalytic triad consisting of E, K and C at positions equivalent to E57, K147 and C181 in SEQ ID NO:1. In another aspect, the present invention provides an engineered nitrilase comprisingor consisting of an amino acid sequence of SEQ ID NO:2 or SEQ ID NO:3. In another aspect, the present invention provides an engineered nitrilase comprising a sequence that is at least 90% identical to SEQ ID NO:1 and includes a catalytic triad consisting of Glu, Lys, and Cys, wherein the engineered nitrilase exhibits improved nitrilase activity relative to the nitrilase of SEQ ID NO:1 for converting a trinitrile substrate to a triacid product. In another aspect, the present invention provides a polynucleotide encoding theengineered nitrilase according to the present invention. In another aspect, the present invention provides an expression vector comprising the polynucleotide according to the present invention. In another aspect, the present invention provides a host cell comprising theexpression vector according to the present invention. In another aspect, the present invention provides a method of preparing anengineered nitrilase comprising culturing a host cell according to the present invention under suitable conditions. In another aspect, the present invention provides a method of screening a panel ofnitrilases to identify an enzyme capable of catalysing the conversion of a compound of wherein R is a linker group suitable for attaching to a therapeutic oligonucleotide; comprising: a. identifying sequences encoding said panel of nitrilases;b. synthesizing said nitrilases;c. isolating said nitrilases;d. mixing each of said isolated nitrilases with a compound of formula (II) undersuitable conditions, to form a reaction mixture; and e. analysing each of the reaction mixtures for the presence of a compound offormula (I) to determine which enzyme(s) are able to catalyse the conversion of compound (II) to compound (I). DESCRIPTION OF DRAWINGS / FIGURES FIG.1 is a general reaction scheme for the nitrilase catalysed hydrolysis of organicnitriles (-C≡N) into carboxylic acids and free amine (and amides under certain conditions). FIG.2 is reaction scheme for generating an exemplary GalNAc cluster depictingchemical versus enzymatic routes (Scheme 1).FIG.3 is a reaction scheme for the nitrilase hydrolysis of trinitriles 3 and 4 and theidentified products (Scheme 2).FIG.4 (A) shows the amino acid distribution within the nitrilase active site from hitsNIT1-D12, NIT1-F11, NIT1-G2, and NIT1-H8, 1524 homologues, the selected NIT2 panel and NIT2-D1. (B) shows the location of W158 in the active site of NIT1-D12 homology modelon 3WUY. Catalytic residues (E57, K147, C181) are shown as spheres and residues liningthe active site are shown as sticks. Position 158, is predominantly tryptophan in identified homologues. Mutation of this position away from tryptophan has been observed to drive formation of the amide side product. FIG. 5 shows the largest fold improvement (FI) found for each position in the nitrilaseafter one round of directed evolution (Rd01) with and fatter tubes indicating larger FIs.DETAILED DESCRIPTION OF THE INVENTION DEFINITIONS “Derived from” as used herein in the context of nitrilases, identifies the originating nitrilase, and / or the gene encoding such nitrilase, upon which engineering was based. “Engineered enzyme” or “engineered nitrilase” as used herein refers to a nitrilase in which the natural or native form of the enzyme, known as the “wild-type”, has been modified in a manner that would not otherwise exist in nature. The modification may comprise, for example, one or more deletions, insertions and / or substitutions of amino acid(s) in the wild- type sequence of the enzyme. “GalNAc” as used herein is short for N-acetylgalactosamine, also known as 2-(acetylamino)-2-deoxy-D-galactose (IUPAC), the structure of which is shown below. As the skilled person would appreciate, in any of the embodiments described herein, the three hydroxy groups in the N-acetylgalactosamine moiety may be independently acetylated. For example, the N-acetylgalactosamine group may be: (where none of the hydroxy groups are acetylated), or (when all three hydroxy groups are acetylated), or any other combination of acetylation. In an embodiment, the GalNAc cluster is trivalent. “GalNAc cluster” as used herein comprises a chemical scaffold with two or moreGalNAc molecules attached thereto. A GalNAc cluster is a multivalent ligand for anasialoglycoprotein receptor (ASGPR). GalNAc clusters may be divalent (two GalNAcmolecules attached to the scaffold), trivalent (three GalNAc molecules attached to thescaffold, also known as triantennary GalNAc) or tetravalent (four GalNAc molecules attachedto the scaffold). The affinity of the ASGPR for a trimer of GalNAc is 1,000-fold higher than adimer, and 1,000-fold higher than a monomer, while a tetramer has just a slightly higheraffinity for the receptor than a trimer. The optimal presentation distance between eachGalNAc is thought to be 15–20 Å (Debacker et al, 2020, Mol. Ther.28(8), 1759-1771). Areview article of the history of GalNAc targeting from early academic studies to current oligo nucleotide therapeutics authored by Kumar and Turnbull was recently published andexamples of GalNAc clusters, with different scaffolds, are provided therein (Chem. Soc.Rev., 2023, 52, 1273-1287).“Tris-based GalNAc cluster” is a GalNAc cluster derived from tris buffer (2-amino-2-(hydroxymethyl)-1,3-propanediol). “Nitrilase” as used herein is an enzyme, also known as a nitrile aminohydrolase, that catalyses the hydrolysis of organic nitriles (-C≡N) into carboxylic acids and free amine (and amides under certain conditions): RCN + 2H2O → RCO2H + NH3(see also Figure 1). Nitrilases are covered in EC 3.5.5.1. A nitrilase is usually a single polypeptide ranging from 32-45 kDa with an α-β-β-α fold structure. The most favoured form is a filament consisting of 6-26 subunits. Nitrilase enzymes belong to the thiol hydrolases class possessing the Lys (K)- Cys (C)-Glu (E) catalytic triad which is essential for its active site function. “Oligonucleotide”, or “oligo” for short, as used herein means a polymer of nucleotide residues. The term “oligonucleotide” is usually used for shorter sequences than the term “polynucleotide”, generally in the range of 3 to 30 nucleotides. These may incorporate deoxyribonucleotides (wherein the resulting oligonucleotide is DNA), ribonucleotides (wherein the resulting oligonucleotide is RNA), modified nucleotides, or a mixture thereof. Oligos can be single-stranded, such as antisense molecules, or double-stranded, such assiRNA molecules. An oligonucleotide may be conjugated to another molecule, e.g., N-Acetylgalactosamine (GalNAc) or multiples thereof (e.g., a GalNAc cluster).“Percent identity” or “% identity” between a query amino acid sequence and a subject amino acid sequence is the “Identities” value, expressed as a percentage, that is calculated using a suitable algorithm (e.g. BLASTP, FASTA, Needleman-Wunsch, Smith-Waterman, LALIGN, or GenePAST / KERR) or software (e.g. DNASTAR Lasergene, GenomeQuest, EMBOSS needle or EMBOSS infoalign), over the entire length of the query sequence after a pair-wise global sequence alignment has been performed using a suitable algorithm (e.g. Needleman-Wunsch or GenePAST / KERR) or software (e.g. DNASTAR Lasergene or GenePAST / KERR). Importantly, a query amino acid sequence may be described by an amino acid sequence disclosed herein, in particular in one or more of the claims, forexample SEQ ID NO:1, SEQ ID NO:2 or SEQ ID NO:3. The query sequence may be100% identical to the subject sequence, or it may include up to a certain integer number of amino acid alterations as compared to the subject sequence such that the % identity is lessthan 100%. For example, the query sequence is at least 50, 60, 70, 75, 80, 85, 90, 95, 96,97, 98, or 99% identical to the subject sequence. Such alterations include at least one amino acid residue deletion, substitution (including conservative and non-conservativesubstitutions), or insertion, wherein said alterations may occur at the amino- or carboxy-terminal positions of the query sequence or anywhere between those terminal positions, interspersed either individually among the amino acid residues in the query sequence or in one or more contiguous groups within the query sequence.“Scaffold” as used herein is the core chemical structure within a GalNAc cluster.Examples of scaffolds include those derived from tris buffer (2-amino-2-(hydroxymethyl)-1,3-propanediol). “Therapeutic oligonucleotide” or “therapeutic oligo” as used herein means anoligonucleotide that has a therapeutic application, e.g., in the prevention or treatment of acondition or disease in a human or animal. Therapeutic oligos typically contains one or moremodified nucleotide residues or linkages. Therapeutic oligos act via one of several different mechanisms, including, but not limited to, antisense, splice-switching or exon-skipping,immunostimulation, RNA interference (RNAi), e.g., via microRNA (miRNA) or smallinterfering RNA (siRNA), and recruitment and guiding of DNA and RNA editing enzymes, for example an A to I RNA base-editing oligonucleotide (AIMer). A therapeutic oligo may be an aptamer. Therapeutic oligonucleotides will usually, but not always, have a defined sequence. “Substitution” of an amino acid refers to a change in the amino acid residue at a specified position of a polypeptide sequence when compared to a reference sequence, such as the wild-type sequence or another engineered variant. “Conservative” amino acid substitutions or mutations refer to the interchangeability of residues having similar side chains, and thus typically involves the substitution of the amino acid in the polypeptide with amino acids within the same or similar defined class of amino acids. However, as used herein, in some embodiments, conservative mutations do not include substitutions from a hydrophilic to hydrophilic, hydrophobic to hydrophobic, hydroxyl- containing to hydroxyl-containing, or small to small residue, if the conservative mutation can instead be a substitution from an aliphatic to an aliphatic, non-polar to non-polar, polar to polar, acidic to acidic, basic to basic, aromatic to aromatic, or constrained to constrained residue. Further, as used herein, A, V, L, or I can be conservatively mutated to eitheranother aliphatic residue or to another non-polar residue. Table 1 below shows exemplaryconservative substitutions. Table 1 Residue Possible Conservative Mutations A, L, V, I Other aliphatic (A, L, V, I)Other non-polar (A, L, V, I, G, M) G, M Other non-polar (A, L, V, I, G, M)D, E Other acidic (D, E)K, R Other basic (K, R)P NoneN, Q, S, T Other polarH, Y, W, F Other aromatic (H, Y, W, F)C NoneThe term "pharmaceutically acceptable" refers to those compounds (including salts), materials, compositions, and dosage forms which are, within the scope of sound medical judgment, suitable for use in contact with the tissues of human beings and animals without excessive toxicity, irritation, or other problem or complication, commensurate with a reasonable benefit / risk ratio. EMBODIMENTS The inventors have surprisingly found that nitrilases are useful in the manufacture ofN-acetylgalactosamine (GalNAc) clusters. The inventors have designed a convergentprocess for the synthesis of GalNAc clusters. In some embodiments, the GalNAc cluster produced in this process are crystalline, which aids the control of impurities. Nitrilases are enzymes that catalyse the hydrolysis of organic nitriles (-C≡N) intocarboxylic acids and free amine, and amides under certain conditions (see Figure 1). Theelectrophilic carbon of the nitrile is subject to nucleophilic attack by one of the two SH groupson the nitrilase (EnzSH). The thioimidate formed is subsequently hydrolyzed to theacylenzyme (SEnz) and ammonia is created as a by-product. The acylenzyme can undergoone of two pathways depending on the conditions: (i) further hydrolyzation of the acylenzymewith water produces the carboxylic acid and the regenerated enzyme, or (ii) the acylenzyme is hydrolyzed by ammonia, displacing the enzyme and forming the amide product.Conditions that promote amide formation include early release of the enzyme-boundsubstrate after the first water hydrolysis followed by delayed addition of the second watermolecule, low temperature and increased pH conditions (for bioconversions by nitrilase formost bacteria and fungi, the optimal pH range is between 7.0-8.0 and the optimal temperature range is between 30-50℃) and electron withdrawing groups at the ⍺-position. Nitrilases have previously been successfully used to manufacture targets rangingfrom commodity chemicals, such as acrylic acid, to intermediates in the synthesis ofatorvastatin (LIPITOR) and pregabalin (LYRICA). However, despite their well-establishedutility in the synthesis of simple intermediates, implementation of nitrilases in the preparation of bulky substrates has been limited. The published method of chemically synthesising a tris based GalNAc cluster (seee.g., Prakash et al, 2016, J. Med. Chem., 59(6), 2718-2733) involves a Michael addition ofTHAM (tris(hydroxymethyl)aminomethane) with tert-butyl acrylate. This reaction can produce various by-products if it proceeds via 1,2-addition rather than the desired 1,4-addition. In order to minimize the number of by-products, the tert-butyl acrylate electrophile can bereplaced with acrylonitrile (CH₂=CHCN); effectively shutting down the 1,2-additionmechanism (see Scheme 1, Figure 2). In this route the crystallisation of intermediate triacid6 (3,3'-((2-(5-(Benzyloxy)-5-oxopentanamido)-2-((2-carboxyethoxy)methyl)propane-1,3-diyl)bis(oxy))dipropionic acid) is an effective control point. When generating the triacid from the trinitrile it is desirable to use mild conditions toprevent side reactions, but selective nitrile hydrolysis is difficult to achieve. Strong acids are well known to carry out nitrile hydrolysis, but these reaction conditions are incompatible withthe other functional groups in the trinitrile and can generate undesirable by-products, e.g.,via concomitant ester hydrolysis. The inventors have replaced chemical synthesis with enzymatic synthesis to achieveselective nitrile hydrolysis. They have used protein engineering tools to modify a nitrilase tosuccessfully accept bulky substrates, such as trinitrile 3 or 4, and catalyse the formation ofthe corresponding triacid, such as triacid 5 or 6, respectively (See Figure 3). The trinitrilesubstrates are easily produced from commercially available materials. The triacid may becrystallisable. In an aspect of the invention, use of a nitrilase in the manufacture of a tris-basedN-acetylgalactosamine (GalNAc) cluster is provided. In an aspect of the invention, there is provided use of a nitrilase to form a compound of formula (I): from a compound of formula (II): , wherein R is a linker group suitable for attaching to a therapeutic oligonucleotide. In an aspect of the invention, there is provided a method of producing a compound of formula (I): by reacting a compound of formula (II): with a nitrilase under suitable conditions, wherein R is a linker group suitable for attaching to a therapeutic oligonucleotide.The skilled person will appreciate that “a linker group suitable for attaching to a therapeuticoligonucleotide” will inevitably also be suitable for attaching to a protected version of thetherapeutic oligonucleotide, or to a portion of the therapeutic oligonucleotide, for example, anoligonucleotide monomer (e.g. a phosphoroamidite) that will be incorporated into thetherapeutic oligonucleotide via solid phase synthesis, or a “shortmer” (a portion of the oligonucleotide, which forms the therapeutic oligonucleotide once ligated to one or more othershortmers), or to protected versions thereof.The linker group R (attached to the GalNAc cluster) enables the ultimate attachment of theGalNAc cluster to therapeutic oligonucleotides, for hepatic delivery of the attachedoligonucleotide via the asialoglycyprotein receptor on hepatocytes. As mentioned above, thisneed not be direct and encompasses situations where the formed GalNAc cluster is addedafter manufacture of the therapeutic oligonucleotide, as well as to situations where theGalNAc cluster is incorporated into the therapeutic oligonucleotide at earlier stages of itsmanufacture by either solid phase synthesis or enzymatic methods i.e. into oligonucleotidemonomers or shortmers. It will also be apparent to the skilled reader that in addition toincorporating a “complete” GalNAc cluster, it would be possible to incorporate the nitrile (i.e.the compounds of formulae II, IV or VII) or acid (i.e. the compound of formulae I, III or VI)precursors into the therapeutic oligonucleotide or to portions thereof (i.e. oligonucleotide monomers or shortmers) and to convert the precursors into a GalNAc cluster at a later stage of manufacture. For example, the precursor, for example on the oligonucleotide monomer, shortmer or therapeutic oligonucleotide. Accordingly, R may be any moiety suitable for attaching the GalNAc cluster to a therapeuticoligonucleotide, by acting as a linker between the GalNAc cluster and the oligonucleotide ora portion thereof. R as used herein may comprise part of, or all of, the linker between the GalNAc cluster and the oligonucleotide, depending on the final structure of the conjugated oligonucleotide. Depending on the reactions used to attach the GalNAc cluster, nitrile or acidprecursor to an oligonucleotide or portion thereof, R may end in a hydrogen or any suitableleaving group such as alkyl or aryl, e.g. C1-20alkyl or benzyl. As would be understood by theskilled person, since the R group is present on the other side of the compound to where thenitrilases of the present invention act on, the exact structure of R is not significant to thefunction of the nitrilase and will vary depending on the design of the final conjugatedoligonucleotide. In one embodiment, the compound of formula (I) according to the present invention isselected from: ,wherein L is a leaving group, such as C1-20alkyl, C1-20alkoxy or benzyl. In one embodiment Lis C1-6alkyl or benzyl.In an aspect of the invention, there is provided a method of producing a compound offormula (III): by reacting a compound of formula (IV): with a nitrilase under suitable conditions, wherein L is a leaving group such as C1-20alkyl, C1-20alkoxy or benzyl. In one embodiment L is C1-6alkyl or benzyl. In an embodiment, the reaction temperature of the nitrilase as described herein is from 30 ºC to 45 ºC. In an embodiment, the reaction temperature is from 30 ºC to 40 ºC. In an embodiment, the reaction temperature is about 30 ºC. In an embodiment, the reaction pH is between 4 and 10, between 5 and 9, or between 6 and 8. In an embodiment, the reaction pH is about 7. In an embodiment, the triacid of formula (I) is crystallisable. In an embodiment, the compound of formula (II) is: , wherein L1is ethyl or benzyl, and the compound of formula (I) is: . In an embodiment, the compound of formula (II) is: and the compound of formula (I) is: . In an embodiment, the compound of formula (II) is: and the compound of formula (I) is: In an embodiment, the compound of formula (II) is: and the compound of formula (I) is: . Once the triacid compound of formula (I) is formed as described herein, it may bereacted with any appropriate moiety comprising GalNAc using standard chemistry known tothe skilled person to form a GalNAc cluster suitable for ultimately conjugating to atherapeutic oligonucleotide or a portion thereof.In one aspect, there is provided a method of producing a tris-based GalNAc cluster compound of formula (V): wherein R is a linker group suitable for attaching to a therapeutic oligonucleotide and A is a group comprising GalNAc, the method comprising reacting a compound of formula (II): with a nitrilase under suitable conditions to obtain a triacid intermediate, then reacting thetriacid intermediate with a compound A-L2 wherein L2 is a leaving group. In one embodiment, the compound A-L2is: wherein L2 is a leaving group such as benzyl and each of R1, R2 and R3 is independentlyhydrogen or acyl. In one embodiment, R1, R2 and R3 are each hydrogen. In one embodiment, the compound A-L2 is: wherein L2 is a leaving group such as benzyl and each of R1, R2 and R3 is independently hydrogen or acyl. In one embodiment, R1, R2 and R3 are each hydrogen.In an embodiment, the trivalent GalNAc cluster formed from the triacid compound of formula(I) is: wherein G is N-acetylgalactosamine, n is 0 to 10, and R is a linker group suitable forattaching to a therapeutic oligonucleotide.

[0002] In an embodiment, the compound of formula (V) is selected from: wherein R is a linker group suitable for attaching to a therapeutic oligonucleotide.In an embodiment, the compound of formula (V) is selected from: ,

[0003] 5

[0004] wherein the symbol indicates the attachment point to an oligonucleotide or a linker to an oligonucleotide. In an embodiment, the nitrilase is NIT2-D1 (SEQ ID NO:1). In an embodiment, the nitrilase is M1 (SEQ ID NO:2). In an embodiment, the nitrilase is M2 (SEQ ID NO:3). In an embodiment, the nitrilase comprises an amino acid sequence that is at least 95%, 96%, 97%, 98%, or 99% identical to SEQ ID NO:1, SEQ ID NO:2 or SEQ ID NO:3. In an embodiment, the nitrilase comprises an amino acid sequence that is at least 95%, 96%, 97%, 98%, or 99% identical to SEQ ID NO:1, SEQ ID NO:2 or SEQ ID NO:3, wherein the difference is a conservative substitution of the amino acid(s).In an embodiment, the nitrilase comprises an amino acid sequence that is at least 97%identical to SEQ ID NO:1. In an embodiment, the nitrilase comprises an amino acidsequence that is at least 97% identical to SEQ ID NO:2. In an embodiment, the nitrilasecomprises an amino acid sequence that is at least 97% identical to SEQ ID NO:3.In an embodiment, the nitrilase is an engineered nitrilase. In an embodiment, the nitrilase is present at <10% w / w and complete conversion ofsubstrate to products is obtained in <24 hours with <0.20% a / a impurities.Methods of engineering enzymes are known to persons skilled in the art and includemutagenesis and / or DNA shuffling, as described in Stemmer, 1994, Proc. Natl. Acad. Sci.USA 91:10747-10751; WO 95 / 22625; WO 97 / 0078; WO 97 / 35966; WO 98 / 27230; WO00 / 42651; WO 01 / 75767; and U.S. Patent No. 6,537,746. Other procedures that can beused include, among others, staggered extension process (StEP), in vitro recombination (Zhao, et al., 1998, Nat. Biotechnol.16:258–261), mutagenic PCR (Caldwell, et al., 1994, PCR Methods Appl.3:S136-S140), and cassette mutagenesis (Black, et al., 1996, Proc Natl Acad Sci USA 93:3525-3529). Mutagenesis and directed evolution techniques useful for the purposes herein are also described in the following references: Ling, et al., 1997, “Approaches to DNA mutagenesis: an overview,” Anal. Biochem.254(2):157-78; Vidal, et al., 2023, “A primer to directed evolution: current methodologies and future directions”, RSC Chem Biol, 4:271-291; Currin, et al., 2021, “The evolving art of creating genetic diversity: From directed evolution to synthetic biology”, Biotechnology Advances, 50:107762; Currin et al., 2015, “Synthetic biology for the directed evolution of protein biocatalysts: navigating sequence space intelligently”, Chem Soc rev, 44:1172-1239; Dale, et al., 1996, “Oligonucleotide-directed random mutagenesis using the phosphorothioate method,” Methods Mol. Biol.57:369-74; Smith, 1985, “In vitro mutagenesis,” Ann. Rev. Genet.19:423- 462; Botstein, et al., 1985, “Strategies and applications of in vitro mutagenesis,” Science 229:1193-1201; Carter, 1986, “Site-directed mutagenesis,” Biochem. J.237:1-7; Kramer, et al., 1984, “Point Mismatch Repair,” Cell 38:879-887; Wells, et al., 1985, “Cassette mutagenesis: an efficient method for generation of multiple mutations at defined sites,” Gene 34:315-323; Minshull, et al., 1999, “Protein evolution by molecular breeding,” Curr Opin Chem Biol 3:284-290; Christians, et al., 1999, “Directed evolution of thymidine kinase for AZT phosphorylation using DNA family shuffling,” Nature Biotech 17:259-264; Crameri, et al., 1998, “DNA shuffling of a family of genes from diverse species accelerates directed evolution,” Nature 391:288-291; Crameri, et al., 1997, “Molecular evolution of an arsenate detoxification pathway by DNA shuffling,” Nature Biotech 15:436-438; Zhang, et al., 1997, “Directed evolution of an effective fructosidase from a galactosidase by DNA shuffling and screening,” Proc Natl Acad Sci USA 94:45-4-4509; Crameri, et al., 1996, “Improved green fluorescent protein by molecular evolution using DNA shuffling,’ Nature Biotech 14:315-319; and Stemmer, 1994, “Rapid evolution of a protein in vitro by DNA shuffling,” Nature 370:389- 391. In an aspect of the invention, there is provided an engineered nitrilase comprising a sequence that is at least 90% identical to SEQ ID NO:1 and includes a catalytic triad consisting of E, K and C at positions equivalent to E57, K147 and C181 in SEQ ID NO:1. In an embodiment, the engineered nitrilase comprises a sequence that is at least 90% identical to SEQ ID NO:1 and includes a catalytic triad consisting of E, K and C atpositions equivalent to E57, K147 and C181 in SEQ ID NO:1. In an embodiment, the aminoacid at a position equivalent to position 158 in SEQ ID NO:1 is an aromatic amino acid. In anembodiment, the aromatic amino acid is selected from the group consisting of: F, H, W andY. In one embodiment, any other amino acid substitutions are conservative substitutions.In an embodiment, one or more amino acids at positions equivalent to the following positions in SEQ ID NO:1 have been mutated: 9, 56, 58, 61, 94, 137, 138, 141, 183, 192,242, 243, 258, 275, 278, 305. In an embodiment, 1, 2, 3 or 4 amino acids at any positionequivalent to the following positions in SEQ ID NO:1 have been mutated: 9, 56, 58, 61, 94,137, 138, 141, 183, 192, 242, 243, 258, 275, 278, 305. In an embodiment, amino acidposition 9 is L or Y; position 56 is C, F or V; position 58 is K, N or W; position 59 is V; position 61 is E, H, I, L or M; position 94 is R; position 137 is A, C, D, E, F, G, L, M, S, T, W, or Y; position 138 is A, C, E, F, L, M, R, W or Y; position 141 is I; position 183 is C; position 192 is G; position 209 is A; position 242 is I; position 243 is S; position 258 is A; position 275 is N; position 278 is M; and / or position 205 is L; wherein the foregoing amino acid positions are numbered with reference to SEQ ID NO:1. In an embodiment, the engineered nitrilase comprises G at a position equivalent to position 192 in SEQ ID NO:1 and a set of mutations as set out in any row of Table 4. In an aspect of the invention, an engineered nitrilase comprising or consisting of anamino acid sequence of SEQ ID NO:2 or SEQ ID NO:3 is provided.In an aspect of the invention, an engineered nitrilase is provided comprising asequence that is at least 90% identical to SEQ ID NO:1 and includes a catalytic triadconsisting of Glu, Lys, and Cys, wherein the engineered nitrilase exhibits improved nitrilaseactivity relative to the nitrilase of SEQ ID NO:1 for converting a trinitrile substrate to a triacidproduct. In one embodiment, the improved nitrilase activity is at least 1.1 fold improvedactivity for converting a trinitrile substrate to a triacid product. In an embodiment, the engineered nitrilase has reduced side product formationrelative to SEQ ID NO:1. In an embodiment, the side product level is <0.2% a / a.In an embodiment, the improved activity is complete conversion of the trinitrile product to triacid substrate in less than 24 hours. In an embodiment, the engineered nitrilase has reduced side product formationrelative to SEQ ID NO:1 and complete conversion of the trinitrile product to triacid substratein less than 24 hours. In an aspect of the invention, there is provided a method of screening a panel of nitrilases to identify an enzyme capable of catalysing the conversion of a compound of formula (II): wherein R is a linker group suitable for attaching to a therapeutic oligonucleotide; comprising: a. identifying sequences encoding said panel of nitrilases;a. synthesizing said nitrilases;b. isolating said nitrilases;c. mixing each of said isolated nitrilases with a compound of formula(II) under suitable conditions, to form a reaction mixture; and d. analysing each of the reaction mixtures for the presence of acompound of formula (I) to determine which enzyme(s) are able to catalyse the conversion of compound (II) to compound (I). Nitrilases disclosed herein may be prepared by any of a number of conventional techniques. For example, nitrilases may be purified from cells that naturally express them or produced in recombinant expression systems. In an aspect of the invention, a polynucleotide encoding an engineered nitrilase as disclosed herein is provided. In an aspect of the invention, an expression vector comprising a polynucleotide encoding an engineered nitrilase as disclosed herein is provided. In an aspect of the invention, a host cell comprising the aforementioned expressionvector is provided. A number of different expression systems and purification regimes can be used to generate the nitrilases of the disclosure. Generally, host cells are transformed with a recombinant expression vector encoding the desired nitrilase. The expression vector may be maintained by the host as a separate genetic element or integrated into the host chromosome depending on the expression system. A wide range of host cells can be employed, including Prokaryotes (including Gram negative or Gram positive bacteria, for example Escherichia coli, Bacilli sp., Pseudomonas sp., Corynebacterium sp.), Eukaryotes including yeast (for example Saccharomyces cerevisiae, Pichia pastoris), fungi (for example Aspergilus sp.), or higher Eukaryotes, including insect cells and cell lines of mammalian origin (for example, CHO, NS0, PER.C6, HEK293, and HeLa). In an embodiment, the vector is a pET vector. In an embodiment, the vector is pET24b(+). The host cell may be an isolated host cell. The host cell is usually not part of a multicellular organism (e.g., plant or animal). The host cell may be a non-human host cell. In an embodiment, the host cell is a bacterial cell. In an embodiment, the host cell is a gram- positive bacterial cell. In an embodiment, the host cell is E. coli. In an embodiment, the hostcell is an E. coli strain derived from W3110. In an embodiment, the host cell is E. coli BL21.In an embodiment, the host cell is E. coli BL21(DE3). Appropriate cloning and expression vectors for use with bacterial, fungal, yeast, and mammalian host cells are known in the art. In an aspect of the invention, a method of preparing an engineered nitrilasecomprising culturing a host cell under suitable conditions is provided.The cells can be cultured under conditions that promote expression of the nitrilase using a variety of equipment such as shake flasks, spinner flasks, and bioreactors. A crude cell lysate or lyophilised powder may be used in experiments to demonstratenitrilase activity. In an embodiment, cells containing the nitrilase are harvested bycentrifugation and the supernatant is discarded. The pellets are then resuspended, e.g., inpotassium phosphate buffer, and re-suspended cultures are filtered, e.g., through two layersof muslin, lysed, e.g., by two passes through a microfluidizer, and cooled, e.g., with ice, withcell debris being removed, e.g., by centrifugation. In an embodiment, the lysate supernatantis collected, pooled, and lyophilized. Alternatively, the nitrilase polypeptide is recovered byconventional protein downstream procedures and may be used in purified form.In an aspect of the invention, a method of manufacturing a GalNAc-conjugated oligonucleotide is provided, comprising a step of making a GalNAc cluster using a nitrilase.In one embodiment, the invention provides a method of manufacturing a GalNAc-conjugatedoligonucleotide is provided, comprising conjugating a GalNAc cluster to said oligonucleotideunder suitable conditions, wherein said GalNAc cluster is made using a nitrilase. In oneembodiment, the GalNAc is directly conjugated to a therapeutic oligonucleotide. In anotherembodiment, the GalNAc is conjugated to a protected oligonucleotide that can be deprotected to provide a GalNAc conjugated therapeutic oligonucleotide. In one embodiment, a method of manufacturing a GalNAc-conjugated therapeutic oligonucleotide is provided, comprising: a. manufacturing a GalNAc cluster using a nitrilase;b. conjugating said GalNAc cluster to a step to an oligonucleotide or a portion thereof;and c. manufacture of the GalNAc-conjugated therapeutic oligonucleotide.Steps a. and b. may be conducted as described herein. In one embodiment, the oligonucleotide in step b. is a protected version of the therapeutic oligonucleotide and step c. comprises deprotection. In one embodiment, step b. comprises conjugation of the GalNAc cluster to anoligonucleotide monomer or a protected version thereof (e.g. a phosphoroamidite). Theoligonucleotide monomer or protected version can be used in the manufacture of theGalNAc-conjugated therapeutic oligonucleotide using techniques well known in the art. In an alternative embodiment, step b. comprises conjugation of the GalNAc cluster to ashortmer or a protected version thereof. The shortmer or protected version thereof can be used in the manufacture of the GalNAc-conjugated therapeutic oligonucleotide by enzymatic methods known in the art. In one embodiment, a method of manufacturing a GalNAc-conjugated therapeutic oligonucleotide is provided, comprising: a. manufacturing the carboxylic acid precursor to the GalNAc cluster using a nitrilase;b. conjugating said carboxylic acid precursor to an oligonucleotide or a portion thereof;and c. manufacture of the GalNAc-conjugated therapeutic oligonucleotide.Steps a. and b. may be conducted as described herein. In an embodiment in which the oligonucleotide is a protected version of the therapeutic oligonucleotide, step c. would comprise formation of the GalNAc cluster and deprotection to form the therapeutic oligonucleotide.In one embodiment, step b. comprises conjugation of the carboxylic acid precursor to anoligonucleotide monomer or a protected version thereof. The oligonucleotide monomer orprotected version thereof can be used in the manufacture of the GalNAc-conjugatedtherapeutic oligonucleotide using techniques well known in the art. In one embodiment, the GalNAc cluster is formed on the therapeutic oligonucleotide. In an alternative embodiment, step b. comprises conjugation of the carboxylic acid precursorto a shortmer or a protected version thereof. The shortmer or protected version thereof canbe used in the manufacture of the GalNAc-conjugated therapeutic oligonucleotide by enzymatic methods known in the art. In one embodiment, the GalNAc cluster is formed on the shortmer. In one embodiment, a method of manufacturing a GalNAc-conjugated therapeutic oligonucleotide is provided, comprising: a. conjugating the nitrile precursor to a GalNAc cluster onto an oligonucleotide or aportion thereof; b. manufacture of the GalNAc-conjugated therapeutic oligonucleotide in a method usinga nitrilase. Step a. may be conducted as described herein. Step b. comprises a step of using a nitrilase to prepare the carboxylic acid precursor to theGalNAc cluster from the nitrile precursor. This step may take place at any time in themanufacture of the GalNAc-conjugated therapeutic oligonucleotide. In one embodiment in which the nitrile precursor is conjugated onto a protected version of the therapeutic oligonucleotide, step b would further comprise formation of the GalNAc cluster and deprotection to form the therapeutic oligonucleotide.In one embodiment in which the nitrile precursor is conjugated to an oligonucleotidemonomer or protected version thereof (e.g. a phosphoroamidite), the step of formation of the carboxylic acid precursor may take place after the phosphoroamidite is used inoligonucleotide synthesis. In addition, it will be appreciated that the subsequent step offormation of the GalNAc cluster does not need to take place immediately following the step of preparing the carboxylic acid precursor. For example, it would be possible for the nitrileprecursor to be conjugated to a phosphoroamidite, which is then used in shortmer synthesis.The carboxylic acid conversion could take place on the shortmer, with the shortmer beingused in oligonucleotide synthesis, with the step of formation of the GalNAc cluster occurringonly as the final step.In one embodiment in which the nitrile precursor is conjugated to a shortmer or a protectedversion thereof, the step of formation of the carboxylic acid precursor may take place before or after the shortmer or protected version thereof is used in oligonucleotide synthesis. In addition, the subsequent step of formation of the GalNAc cluster does not need to take place immediately following the step of preparing the carboxylic acid precursor. For example, it would be possible for the nitrile precursor to be conjugated to a shortmer, which is then used in oligonucleotide synthesis, with the step of formation of the GalNAc cluster occurring after oligonucleotide synthesis. GalNAc clusters as disclosed herein may be conjugated to oligonucleotides by anumber of different methods known in the art, including amide coupling, phosphoramiditecoupling and click-chemistry. For example, see conjugation methodology in Prakash et al,Nucleic Acids Res., 2014, 42, 8796-8807; Nair et al, J. Am. Chem. Soc., 2014, 136, 16958-16961; Prakash et al, J. Med. Chem., 2016, 59, 2718-2733; Matsuda et al, ACS Che Biol,2015, 10, 1181-1187; Farzan et al, Bioconjug. Chem, 2017, 28, 10, 2599-2607; Østergaardet al, Bioconjug, Chem., 2015, 26, 1451-1455; Migawa et al, Bioorg. Med. Chem. Lett., 2016,26, 2194-2197; and Sharma et al, Bioconjug. Chem., 2018, 29, 2478-2488).In one aspect, the invention provides a GalNAc-conjugated therapeutic oligonucleotide obtained as described herein. In an embodiment, the oligonucleotide is selected from the group consisting of: an antisense oligonucleotide, siRNA, miRNA, aptamer, CRISPR guide RNA, and an oligonucleotide used to recruit and guide DNA and RNA editing enzymes. In an embodiment, the oligonucleotide used to recruit and guide DNA and RNA editing enzymes is an A to I RNA base-editing oligonucleotide (AIMer). In an embodiment, the oligonucleotide is a therapeutic oligonucleotide. In an aspect of the invention, a GalNAc-conjugated oligonucleotide for use in therapy is provided, wherein the GalNAc-conjugated oligonucleotide comprises a GalNAc cluster made using a nitrilase. In an embodiment, the oligonucleotide is an antisense oligonucleotide. In anembodiment, the oligonucleotide is an siRNA. In an embodiment, the oligonucleotide is anmiRNA. In an embodiment, the oligo nucleotide is an aptamer. In an embodiment, theoligonucleotide is a CRISPR guide RNA. In an embodiment, the oligonucleotide is anoligonucleotide used to recruit and guide DNA and RNA editing enzymes. In an embodiment,the oligonucleotide is an A to I RNA base-editing oligonucleotide (AIMer). In an embodiment, the GalNAc-conjugated oligonucleotide is selected from the groupconsisting of: cimdelirsen, donidalorsen, eplontersen, givosiran, inclisiran, lumasiran,olezarsen, pelacarsen, sapablursen, vupanorsen, vutrisiran, AZD8233, H-020, IONIS-FB-LRx / RG6299, IONIS-FXI-LRx, RG-125, SR-059, STP-135G, VIR-2218, WVE-006 and AB-729.In an embodiment, the GalNAc-conjugated oligonucleotide is AZD8233. In an embodiment, the GalNAc-conjugated oligonucleotide is selected from the groupconsisting of: givosiran, inclisiran, lumasiran, vutrisiran, H-020, RG-125, SR-059, VIR-2218and WVE-006. In an embodiment, the GalNAc-conjugated oligonucleotide is selected from the groupconsisting of: cimdelirsen, donidalorsen, eplontersen, olezarsen, pelacarsen, sapablursen,vupanorsen, IONIS-FB-LRx / RG6299, and IONIS-FXI-LRx.In an embodiment, the GalNAc-conjugated oligonucleotide is STP-135G. In an embodiment, the GalNAc-conjugated oligonucleotide is selected from the groupconsisting of: plozasiran, fazirsiran, olpasiran, zodasiran, GSK4532990, daplosiran andtomligisiran. Numbered Embodiments The invention may be described by means of the following numbered embodiments:1. Use of a nitrilase in the manufacture of an N-acetylgalactosamine (GalNAc) cluster.2. The use according to embodiment 1, wherein the GalNAc cluster is a trivalent cluster.3. The use according to embodiment 2, wherein the trivalent cluster is a tris-basedtrivalent cluster.4. The use according to embodiment 3, wherein the nitrilase catalyses the conversion ofa compound of formula (II): (I): , wherein R is a linker group suitable for attaching to a therapeutic oligonucleotide.5. The use according to embodiment 4, wherein the compound of formula (I) iscrystalline.6. The use according to embodiment 4, wherein the trivalent cluster is: , wherein G is N-acetylgalactosamine, n is 0 to 10, and R is a linker group suitable for attaching to a therapeutic oligonucleotide.7. Use of a nitrilase to form a compound of formula (I): from a compound of formula (II): , wherein R is a linker group suitable for attaching to a therapeutic oligonucleotide.8. The use according to embodiment 7, wherein the compound of formula (I) is: , wherein L is a leaving group, such as C1-20alkyl, C1-20alkoxy or benzyl. In one embodiment L is C1-6alkyl or benzyl.9. The use according to embodiment 8, wherein the compound of formula (II) is: , and the compound of formula (I) is: , wherein R is ethyl or benzyl.10. The use according to embodiment 8, wherein the compound of formula (II) is: and the compound of formula (I) is: .11. The use according to embodiment 8, wherein the compound of formula (II) is: and the compound of formula (I) is: .12. The use according to embodiment 8, wherein the compound of formula (II) is: and the compound of formula (I) is: .13. A method of producing a compound of formula (I): by reacting a compound of formula (II): with a nitrilase under suitable conditions, wherein R is a linker group suitable for attaching to a therapeutic oligonucleotide.14. A method of producing a compound of formula (III): by reacting a compound of formula (IV): with a nitrilase under suitable conditions, wherein L is a leaving group such as C1-20alkyl, C1-20alkoxy or benzyl.15. The method according to embodiment 14, wherein the reaction temperature is 30 ºCto 45 ºC.16. The method according to embodiment 15, wherein the reaction temperature is 30 ºCto 40 ºC.17. The method according to embodiment 16, wherein the reaction temperature is about30 ºC.18. The method according to any one of embodiments 14 to 17, wherein the reaction pHis between 4 and 10, between 5 and 9, or between 6 and 8.19. The method according to embodiment 18, wherein the reaction pH is about 7.20. The method according to any one of embodiments 14 to 19, wherein the impuritylevel is <0.20 a / a.21. The method according to any one of embodiments 14 to 20, wherein completeconversion of substrate to product takes ≤ 24 hours.22. A method of manufacturing a GalNAc-conjugated oligonucleotide comprisingconjugating a GalNAc cluster to said oligonucleotide under suitable conditions, wherein said GalNAc cluster is made using a nitrilase. 23. A GalNAc-conjugated oligonucleotide for use in therapy, wherein the GalNAc-conjugated oligonucleotide comprises a GalNAc cluster made using a nitrilase. 24. A method of producing a tris-based GalNAc cluster compound of formula (V): wherein R is a linker group suitable for attaching to a therapeutic oligonucleotide and A is a group comprising GalNAc, the method comprising reacting a compound of formula (II): with a nitrilase under suitable conditions to obtain a triacid intermediate, then reacting the triacid intermediate with a compound A-L2 wherein L2 is a leaving group. 25. A method as described in embodiment 24, wherein A-L2 is: wherein L2is a leaving group such as benzyl and each of R1, R2and R3is independently hydrogen or acyl.26. An engineered nitrilase comprising a sequence that is at least 90% identical to SEQID NO:1 and includes a catalytic triad consisting of E, K and C at positions equivalent to E57, K147 and C181 in SEQ ID NO:1.27. The engineered nitrilase according to embodiment 26, wherein the amino acid at aposition equivalent to position 158 in SEQ ID NO:1 is an aromatic amino acid.28. The engineered nitrilase according to embodiment 27, wherein the aromatic aminoacid is selected from the group consisting of: F, H, W and Y.29. The engineered nitrilase according to any one of embodiments 26 to 28, wherein oneor more amino acids at positions equivalent to the following positions in SEQ ID NO:1 have been mutated: 9, 56, 58, 61, 94, 137, 138, 141, 183, 192, 242, 243, 258, 275, 278, 305.30. The engineered nitrilase according to any one of embodiments 26 to 29, whereinamino acid position 9 is L or Y; position 56 is C, F or V; position 58 is K, N or W; position 59 is V; position 61 is E, H, I, L or M; position 94 is R; position 137 is A, C, D, E, F, G, L, M, S, T, W, or Y; position 138 is A, C, E, F, L, M, R, W or Y; position 141 is I; position 183 is C; position 192 is G; position 209 is A; position 242 is I; position 243 is S; position 258 is A; position 275 is N; position 278 is M; and / or position 205 is L; wherein the foregoing amino acid positions are numbered with reference to SEQ ID NO:1.31. The engineered nitrilase according to any one of embodiments 26 to 30, wherein thenitrilase comprises G at a position equivalent to position 192 in SEQ ID NO:1 and a set of mutations as set out in any row of Table 5.32. An engineered nitrilase comprising or consisting of an amino acid sequence of SEQID NO:2 or SEQ ID NO:3.33. An engineered nitrilase comprising a sequence that is at least 90% identical to SEQID NO:1 and includes a catalytic triad consisting of Glu, Lys, and Cys, wherein the engineered nitrilase exhibits improved nitrilase activity relative to the nitrilase of SEQ ID NO:1 for converting a trinitrile substrate to a triacid product.34. The engineered nitrilase according to embodiment 33, wherein the trinitrile substrateis a compound of formula (II): is a compound of formula (I): wherein R is a linker group suitable for attaching to a therapeutic oligonucleotide.35. The engineered nitrilase according to embodiment 33 or embodiment 34, wherein theengineered nitrilase has reduced side product formation relative to SEQ ID NO:1, optionally wherein the side product level is <0.2% a / a.35. The engineered nitrilase according to any one of embodiments 33 to 35, wherein theimproved activity is complete conversion of the trinitrile product to triacid substrate in less than 24 hours.36. A polynucleotide encoding the engineered nitrilase according to any one ofembodiments 26 to 35.37. An expression vector comprising the polynucleotide according to embodiment 36.38. A host cell comprising the expression vector according to embodiment 37.39. A method of preparing an engineered nitrilase comprising culturing a host cellaccording to embodiment 38 under suitable conditions. 40. A method of screening a panel of nitrilases to identify an enzyme capable of catalysing the conversion of a compound of formula (II): wherein R is a linker group suitable for attaching to a therapeutic oligonucleotide; comprising: a. identifying sequences encoding said panel of nitrilases;b. synthesizing said nitrilases;c. isolating said nitrilases;d. mixing each of said isolated nitrilases with a compound of formula (II) under suitableconditions, to form a reaction mixture; and e. analysing each of the reaction mixtures for the presence of a compound of formula (I)to determine which enzyme(s) are able to catalyse the conversion of compound (II) to compound (I). EXAMPLES Abbreviations ACN acetonitrileAPI active pharmaceutical ingredientCHES N-cyclohexyl-2-aminoethanesulfonic acidDMSO dimethylsulfoxideESI electrospray ionisationFI fold improvementLC-MS liquid chromatography-mass spectrometryMES 2-(N-morpholino)ethanesulfonic acid MS mass spectrometry MS-ES+positive electrospray ionisation mass spectrometry PLIF protein-ligand interaction fingerprintingLC-MS Method 1: high throughput analytical LC-MS method to determine conversion oftrinitrile 3 to triacid 5Conversion of the trinitrile 3 to the triacid 5 was determined using a Waters AcquityBinary Solvent Manager, Acquity Column Manager at 45 °C equipped with a Leap CTC PAL HTS Autosampler (2 µL injection); ELS: Waters Acquity ELSD (50 °C) and MS: Waters Acquity SQD, Polarity (positive or negative); Mode (continuum); Scan Time (0.1 s) Capillary V (3500); Cone V (25-35). The LC-MS was equipped with a Waters CSH C18 (2.1 mm x 30 mm, 1.7 µm) column and 0.02% trifluoroacetic acid in water (mobile phase A) and 0.02% trifluoroacetic acid in ACN (mobile phase B) at flow rate of 1.3 mL / min at a column temperature of 45 °C.The run time was 2 min with gradient Table 2 as below. Compound elution was monitoredusing Evaporative Light-Scattering Detector and Mass Ions. Table 2 Time (minutes) Phase APhase B (%) (%)- 97 31.8 2 981.9 2 981.91 97 32.0 97 3LC-MS Method 2 Conversion of the trinitrile to the triacid was determined using a Waters AcquitySystem with a QDa MS using UPLC ACQUITY UPLC CSH C18 Column (130 Å, 2.1 mm x 50 mm x 1.7 µm) and the following gradient (see Table 2) with 10 mM ammonium bicarbonate pH 10 (mobile phase A) and acetonitrile (mobile phase B) at a flow rate of 1.0 mL / min and a column temperature of 40 °C. UV was monitored (210 to 350 nm) and the MS was scanned (100 to 1250 m / z) in positive and negative mode. The capillary voltage was 1.6 kV for positive mode and 0.8 kV for negative mode. The cone voltage was 7 V for ESI+ and 15 V for ESI-. Table 3 Time (minutes) Phase APhase B (%) (%) -97 31.5 5 952.0 97 32.3 97 3LC-MS Method 3 Conversion of the trinitrile to the triacid was determined using a Waters AcquitySystem with a QDa MS using UPLC ACQUITY UPLC CSH C18 Column (130 Å, 2.1 mm x 50 mm x 1.7 µm) and the gradient set out in Table 2 with 0.1% Formic acid in water (mobile phase A) and formic acid in acetonitrile (mobile phase B) at a flow rate of 1.0 mL / min and a column temperature of 40 °C. UV was monitored (210 to 350 nm) and the MS was scanned (100 to 1250 m / z) in positive and negative mode. The capillary voltage was 1.6 kV for positive and 0.8 kV for negative mode. The cone voltage was 7 V for ESI+ and 15 V for ESI-.Example 1: Screening nitrilases for activity in trinitrile to triacid transformation toproduce 3,3'-((2-((2-Carboxyethoxy)methyl)-2-(5-ethoxy-5-oxopentanamido)propane- 1,3-diyl)bis(oxy))dipropionic acidAn in-house nitrilase collection, nitrilase panel version 1 (NIT1), was screened for thereaction of trinitrile 3 substrate (ethyl 5-((1,3-bis(2-cyanoethoxy)-2-((2-cyanoethoxy)methyl)propan-2-yl)amino)-5-oxopentanoate), which was easily prepared fromcommercially available materials, to triacid 5 (3,3'-((2-((2-Carboxyethoxy)methyl)-2-(5-ethoxy-5-oxopentanamido)propane-1,3-diyl)bis(oxy))dipropionic acid). Figure 3 is a reactionscheme (Scheme 2) for the nitrilase-catalysed hydrolysis of trinitriles 3 and 4 and identifiedproducts. The following general screening protocol was used. The NIT1 panel of nitrilases were cloned into and expressed in E. coli BL-21(DE3)and crude cell lysates were used to test nitrilase activity.Potassium phosphate buffer (100 μL / well, 100 mM, pH 7.0) was added to a nitrilase deep-well panel plate (NIT1, containing 100 μL of crude clarified cell lysate per well).Substrate 3 in DMSO (final concentration 1.4 mg / mL 3, 3.3% v / v DMSO) was then added tothis plate. The reaction plate was then sealed using aluminium / polypropylene laminate heat seal tape and incubated at 30 °C, 500 RPM in an Infors MultiTron shaker. At 17 h and 47 h, 20 µL / well aliquots were taken and diluted into 180 µL / well 1:1ACN:H2O in a shallow-well polypropylene plate (Corning 96). Diluted reaction aliquots were incubated at room temperature with shaking 900 RPM for 5 min, then clarified by centrifugation (4000 RPM, 10 min) and 100 µL / well of supernatant transferred to a new shallow-well polypropylene plate before analysis by LC-MS Method 1. The top four enzymes identified after performing the general screening protocol forhydrolysis of the trinitrile 3 (NIT1-D12, NIT1-F11, NIT1-G2, and NIT1-H8) were prepared aslyophilised powder for further reaction analysis. NIT1-D12 provided excellent selectivity for the formation of mono acid 8 (80%) anddiacid 10 (15%) (see Scheme 2, Figure 3). However, the desired triacid 5 and side productprimary amides were not detected after 48 h at pH 7.0 in the initial screen. NIT1-H8 showed higher activity but delivered primary amides 12 (45%) and 14 (5%)in addition to the desired carboxylic acids 5 (45%) and 10 (5%) at pH 7.0 after 24 h (seeScheme 2, Figure 3). Primary amides are common side-products of nitrilase reactions andthey accumulate in reactions, as they are not substrates for further hydrolysis by the enzyme (Jiang et al, 2017, Catal. Sci Technol., 7 (5), 1122-1128; Sosedov and Stolz, 2015,Microbiol. Biot.99(6), 2623-2635). Minimizing the formation of amides 12 and 14 (seeScheme 2, Figure 3) was therefore identified as a key factor in selecting an appropriate enzyme, as these impurities could be considered critical quality attributes as derivatives of these impurities would be likely to contaminate the API if they could not be removed bydownstream purification. Since the primary amide by-products could potentially arise via theattack of acyl-enzyme intermediate by ammonia / ammonium generated during the reaction course, it was envisaged that controlling pH of the buffer system could bias the productdistribution in favour of acids 5, 8 and 10.Accordingly, each of the four nitrilases were screened from pH 4.0 to pH 10.0 for thehydrolysis of trinitrile ethyl ester 3 to investigate activity and product distribution. Trinitrileethyl ester 3 (10 mg) was charged to wells of a deep-well plate (final concentration 50mg / mL 3).100 μl / well of the appropriate buffer (either sodium citrate (200 mM, pH 4.0), sodium citrate (200 mM, pH 5.0), MES (200 mM, pH 6.0), potassium phosphate buffer (200mM, pH 7.0), potassium phosphate buffer (200 mM, pH 8.0), CHES (200 mM, pH 9.0), orCHES (200 mM, pH 10.0)) were then charged to the plate, followed by 100 µL / well of alyophilised enzyme powder solution in deionized water (20 mg / mL stock concentration) of the nitrilases NIT1-D12, -F11, -G2, and -H8, such that each enzyme / pH combination was sampled once. The reaction plate was sealed with aluminium / polypropylene laminate heat seal tapeand incubated at 30 °C, 300 RPM in a Kuhner shaker. At 24 h and 48 h, 20 µL / well aliquotswere diluted into 180 µL / well 1:1 ACN:H2O in a shallow-well polypropylene plate. Diluted reaction aliquots were incubated at room temperature with shaking 800 RPM for 10 min, then clarified by centrifugation (4000 RPM, 10 min) and 100 µL / well of supernatant was used in analysis by LC-MS Method 1. Attenuated reactivities were observed at the extreme pHs for both for NIT1-D12 and NIT1-H8. Similar pH trends were observed for NIT1-F11 and NIT-G2 while conversion waslower. No significant decrease in amides 12 and 14 was observed for NIT1-H8. Hencefurther experimental work focused on optimization of the NIT1-D12 catalysed reaction, toaccelerate hydrolysis of 8 and 10 to generate the triacid 5.NIT1-D12 was screened at different temperatures ranging from 30 °C to 45 °C.Trinitrile ethyl ester 3 (10 mg) was charged to 8 x 1.5 mL HPLC vials, and 500 µL of nitrilase NIT1-D12 in potassium phosphate buffer (200 mM, pH 7.0) was added (final concentration 20 mg / mL 3, either 10 mg / mL or 25 mg / mL nitrilase NIT1-D12). Vials were incubated at 30°C, 35 °C, 40 °C, or 45 °C, with shaking at 500 RPM in an Eppendorf thermomixer. After 22h, a 10 µL aliquot was taken from each reaction and diluted by addition to 350 µL 1:1ACN:H2O in filter vials and the filtrate analysed using LC-MS Method 1.An increase in temperature to 40 °C positively impacted activity of NIT-D12 and athigh NIT1-D12 loading (25 mg / mL) triacid 5 (15%) was still observed in modest amountswhile diacid 8 (85%) was the major product, with no side amide products being observed.In summary, the initial hits from screening the nitrilase panel version 1 delivered enzymes able to conduct the trinitrile to triacid transformation, but activity and selectivity were far from those desired for industrial manufacture, e.g., of API.Example 2: ‘Fingerprinting’ nitrilase computational hit expansion strategyDirected evolution can be inefficient and time consuming, especially when the starting and desired end points are sufficiently different that multiple rounds of evolution are required.Smarter ways of identifying better starting points for evolution than simply utilising the outputfrom Example 1 were needed.Having successfully identified a collection of nitrilases that catalyse the desiredtransformation, the inventors recognised this as a unique opportunity to use these results tosearch for improved wild-type nitrilases with the aim of reducing the number of rounds ofevolution necessary to generate a fit for purpose variant. A computational hit expansionstrategy called ‘fingerprinting’ was employed. The inventors looked into finding similarities between the enzyme hits, focusing onthe positions and compositions of residues or motifs leading to higher conversions or reduced impurities. Since activity and side product formation are likely a consequence of theresidues in and adjacent to the active site, they focused on identifying homologues havingsimilar active site compositions. To examine the thousands of potential nitrilases and identify those having similaractive sites, a technique akin to protein-ligand interaction fingerprinting (PLIF) was used.PLIFs have been used successfully in cheminformatics and structure-based drug design toidentify similarities in binding poses of different small molecules when bound to their protein targets. While PLIFs are generally used to identify similar ligands that bind to the same protein target, they can also be used to identify similar binding pockets or residue motifsacross protein targets that will bind the same ligand. A PLIF at its most basic is acompressed representation of three dimensional protein-ligand interactions and often takes the form of a one-dimensional list of positions in a protein that interact with a bound ligand. Each interaction is further annotated with chemical information to help classify it and can include atom types, charges, hydrogen bonds, contacting protein residue types, as well asother well-known molecular descriptors. Since the fingerprint captures the protein-ligandinteractions unique to the complex, i.e., the ligand’s binding mode, it follows that enzymeswith similar fingerprints will bind the similar ligand. To identify nitrilases that catalyze the hydrolysis of trinitriles 3 and 4 (benzyl 5-((1,3-bis(2-cyanoethoxy)-2-((2-cyanoethoxy)methyl)propan-2-yl)amino)-5-oxopentanoate),enzymes that have similar PLIFs to NIT1-D12 when bound to the substrate wereinvestigated. Unfortunately, there were no available crystal structures with bound ligandsthat have similar molecular structures to compounds 3 and 4. In addition, docking studieswere inconclusive in identifying specific interactions due to the size and conformational flexibility of both the ligand and the nitrilase active site loops. Accordingly, the approach wassimplified and focus shifted to just the amino acid composition of the active site, i.e., from analignment of our hits, the amino acid identities at each of the positions lining the active site was extracted and compared. To define the enzyme’s active site and potential substrate interacting residues, ahomology model using 3WUY from Synechocystis sp. as the template was constructed.3WUY was selected as the homology template since it has a medium to high-qualityalignment matching 306 of 327 positions with minimal insertions and deletions which were regulated to regions of loop structure.3WUY was resolved at 3.1 Å resolution, contains two copies of the nitrilase in the asymmetric unit, and has an identity to NIT1-D12 of 36%. The nitrilase model’s active site does not contain a ligand, thus Chemical Computing Group’sSiteFinder was used to identify the active site via cavity detection which was subsequentlycorroborated by the presence of the well-known catalytic triad of E57, K147 and C181. The homology model was used to identify positions lining the nitrilase active site by selecting any residues having atoms within 6.0 Å of the catalytic residues E57, K147 and C181. Twenty-two such positions were identified: 56, 57, 58, 63, 130, 146, 147, 148, 149, 150, 151, 154, 158, 180, 181, 182, 183, 204, 205, 206, 207 and 232. A comparison of the top four NIT1 enzymes revealed identities between NIT1-D12 and -F11 of 78%, and -H8 and -G2 of 55% and 30%, respectively. The alignment of the NIT1 enzymes producing the desired product (NIT1-D12, -F11 and -G2) and the most activeenzyme (NIT1-H8) and the homology template, showed -D12 and -F11 as having identicalactive sites while -H8 contains 3 mutations with respect to -D12: A58S, V232T and W158H. The catalytic residues E57, K147, and C181 are all conserved across the five enzymes ofinterest (including 3WUY), as well as positions Y63, L148, P150, T151, E154, and E183. Afurther comparison of residue conservation across 1524 closely related homologues of NIT- D12 identified low sequence conservation at positions 58 and 232, while position 158 clearly prefers an aromatic residue type of either W, Y, F, or H. Based on the active site composition of the NIT1 hits and the distribution of amino acid types across the homologues, the aim was then to identify additional enzymes thatcould 1) sample the aromatic preference at position 158, which the inventors postulate leadsto an increase in amide formation, and 2) sample diversity across the rest of the enzyme with minimal perturbation to the already substrate-tailored active site. Ideally, protein-ligand annotations like similarities in interaction types, pocket hydrophobicities and pocket volumes collated across the nitrilase family are also beneficial in narrowing the list of genes to be selected for the expansion. To accomplish this, the diversityof the active site fingerprinting was expanded to include the amino acid types at the 22 active site positions from not only the NIT1 hits, but the homologues and the homologymodel as well. A two-step hierarchical clustering leveraging this expanded set of diversitywas then performed. In the first step, the clustering was limited to only the fingerprint positions with a minimum cluster of 80% identity, essentially allowing active sites with up to 4 mutations to be clustered together, as an attempt to isolate the various active site motifs. This resulted in the initial hits being segmented into two clusters, one with NIT1-D12, -F11 and -H8 along with192 other enzymes, and another with NIT1-G2 and 39 additional enzymes, for a total of 231.A second, full-length sequence clustering step was then performed on the isolated 231 nitrilases at 70% identity, from which 48 disparate exemplars were selected from eachcluster, enabling enzymes to be successfully identified that retain a high sequence identity inthe active site, evenly sample aromatic diversity at position 158, in addition to varying levels of diversity at 257 other positions across the enzyme. The fingerprint approach resulted in identification of 48 homologous wild-type enzymes active in the trinitrile to triacid transformation (nitrilase panel version 2, NIT2). Notably, none of the 48 selected enzymes has previously been characterized in the literature to perform nitrile hydrolysis. The active site diversity sampled from the 48 identified homologues with respect to NIT1-D12 is shown in Table 4. Table 4 NIT1-D12 NIT1-D12 MOST MUTATIONS TO # MUTATIONS POSITIO AMINO FREQUENT NIT1-D12 AT OBSERVED AT N ACID AMINO ACID POSITION POSITION 56 P P S; T 257* E E 058 A S L; T; V; S; Q 563 Y Y 0130 N N L; I; M 3146 R R Q 1147* K K 0148 L L 0149 M M K 1150 P P 0151 T T 0154 E E 0158 W H H; F 2180 I I L; V; C; M 4181* C C 0182 W W A 1183 E E 0204 P P A; S 2205 T T S 1206 Y Y F 1207 D D S; T 2232 V T A; C; N; T; S 5*Denotes catalytic residue. Codon-optimized homologue genes for the new nitrilase collection (NIT2), were synthesized and cloned into pET24b(+). Plasmids were transformed into BL-21(DE3) Goldcells (Agilent) for expression and used for homologue screening against benzyl ester 4,which was selected over the ethyl ester 3 to facilitate a more facile deprotection stepdownstream in the synthesis. Six enzymes that performed better than NIT1-D12 were identified. Further screeninghelped select the wild-type nitrilase NIT2-D1 (SEQ ID NO:1) from Acidithiobacillalesbacterium SG8_45 (GenBank: KPK12037.1) as the best candidate.Figure 4 (A) shows the amino acid distribution within the nitrilase active site from hitsNIT1-D12, -F11 -G2, and -H8, 1524 homologues, the selected NIT2 panel and NIT2-D1. Figure 4(B) shows the location of W158 in the active site of NIT1-D12 homology model on3WUY. Catalytic residues E57, K147, C181 are shown by yellow spheres and residues liningthe active site are cyan sticks. Position 158, highlighted in green, is predominantly tryptophan in identified homologues. Mutation of this position away from tryptophan has been observed to drive formation of the amide side product. An optimized process with NIT2-D1 at 50% w / w biocatalyst loading, using 50 g / L oftrinitrile 4 and toluene as co-solvent to ensure substrate solubility, afforded crude triacid 6(3,3'-((2-(5-(Benzyloxy)-5-oxopentanamido)-2-((2-carboxyethoxy)methyl)propane-1,3-diyl)bis(oxy))dipropionic acid) in 86.8% yield, and amide 13 levels of ~0.50% a / a (%area / area 13 / 6 not corrected for response factors) (see Table 4). No additional amide sideproducts were detected using these conditions. Whilst these results were very encouraging, biocatalyst performance was still far from the desired manufacturing process: <10% w / wbiocatalyst loading, complete conversion in <24 h, and 13 <0.2% a / a (a proposedspecification to ensure API quality).It was therefore decided to evolve NIT2-D1 with two clear goals in mind: increasebiocatalyst activity to enable full conversion in <24 h using <10% w / w biocatalyst andreduction of the amide impurity 13 formation to <0.20% a / a. The amount of catalyst targeted10% w / w was the maximum amount at which this process was economically and technically viable. Immobilization was not considered for this scenario due to the low due to the low volumes required of this intermediate.Example 3: Directed evolution of nitrilase for the production of 3,3'-((2-(5-(benzyloxy)-5-oxopentanamido)-2-((2-carboxyethoxy)methyl)propane-1,3-diyl)bis(oxy))dipropionic acid Acandidate nitrilase (NIT2-D1, SEQ ID NO:1) identified using the fingerprintapproach as set out in Example 2, having improved activity and selectivity over the initial hitsfrom Example 1, was further improved using directed evolution. Site saturation was chosenas the evolution strategy with the goal of identifying beneficial diversity that could be recombined later. This focused approach was preferred to other approaches such as untargeted and codon biased error prone PCR methodology to ensure quick success. Agene encoding NIT2-D1 codon optimized for expression in E. coli was designedbased on the reported amino acid sequence of the Acidithiobacillales bacterium SG8_45 homologue (GenBank: KPK12037.1). The gene was cloned into the expression vector pCK110900 under control of a lac promoter. Round 1 of evolution (Rd01) The main objective of the first round of evolution was to identify nitrilase variants withincreased activity for the conversion of trinitrile 4 to triacid 6 which also reduce the formationof impurity 13. The homology model of NIT2-D1 generated using a nitrilase fromSynechocystis sp as a template (3WUY, see Example 2) was used for bioinformaticspurposes. This model was considered reasonable for the core of the structure, with acceptable conservation to template, but less so for the loops surrounding the entrance to the active site. The model suggests that the active site is mostly formed by residues from the monomer, but the entrance to the active site lies at the dimer interface. Site saturation libraries based on the location of residues in this structure weresubsequently designed, with 288 amino acids being targeted.4536 different variants weregenerated, expressed in E. coli and screened for conversion of 4 to 6 at 50 g / L substrateloading, using toluene as co-solvent, in order to closely resemble the desired process conditions. Several reaction conditions were tested including clarified and lyophilisedbiocatalyst lysates, to allow for the selection of the best variants. The screening led to the identification of 339 improved variants (calculated by FI: fold improvement defined as variant activity divided by the average activity of positive controls in the same plate). A library comprising mutations lining the active site delivered the highest FI for triacid6 formation (up to 26.7 for variant N137C), with a few other variants following closely(D192G, N243S or R209A). Position N137 was identified as a hotspot with N137F and N137T both showing excellent FI. The second active shell gave excellent hits such as R56C,P57C and R56F, while other second shell positions also provided promising variants - withW61 being a clear hotspot. The outer most active site loop also delivered a few mutationssuch as N64V and E65F which boosted nitrile hydrolysis activity. The top 174 variants were then screened under more challenging conditions. In parallel to running UPLC assays to determine enzyme activity, Rapidfire was also investigated to accurately determine the amount of primary amide impurity 13. Due to the low levels of 13 (<1% a / a) and low conversion targeted in screening reactions (<5%, to allowfor improved variants to be identified), it was not possible to detect amide 13 by UPLC.Rapidfire uses LC-MS / MS to allow detection of very low analyte concentrations, therebyallowing 13 to be detected in screening reactions.Variant D192G (M1, SEQ ID NO:2) was selected as the Rd02 backbone as it gavethe highest conversion of trinitrile 4 to triacid 6 at 2% & 10% w / w biocatalyst loading after 20h on 0.5 g (of 4) scale. It also gave the second lowest amount of impurity 13 at fullconversion. Following process development 10% w / w loading of M1 was used on 47.5 g (of 4) scale and 50 g / L 4, affording full conversion after 21 h with 71.4% yield and 0.23% a / a amide 13. Despite the improved properties of M1 over the initial NIT2-D1, another round was required to meet the process criteria. Figure 5 shows the largest fold improvement (FI) found for each position in thenitrilase after one round of directed evolution (Rd01) with warmer colours and fatter tubesindicating larger FIs. Round 2 of evolution (Rd02) Using the positive diversity from Rd01, 4 combinatorial libraries and 2 libraries targeting 2 specific amino acid combinations (double NNK) were designed. Three librariesexplored the combination of mutants that gave high to moderate conversion of 4 to 6 andhigh to moderate acid / amide (6 / 13) ratio FIs in Rd01 screens. A fourth library combined the best mutants from 11 discrete regions of the nitrilase sequence in order to look at combinations across the entirety of the nitrilase structure. In addition to the recombination libraries, two double NNK libraries were designed based on the observation that mutantswith high 6 FI and 6 / 13 ratios tended to lie in two regions; residues 56-61 and 137-141. Within these regions, residues 56 and 61 and 137 and 138 showed the greatest number ofvariants with a positive result for 6 / 13 ratio FI and on this basis were selected for furtherinvestigation. Because M1 appeared to satisfy the target for biocatalyst loading at 40 °C, Rd02screening was carried out at 30 °C as it had been observed that lower quantities of 13 wereproduced at this temperature suggesting reduced reaction temperature could be used tofurther control formation of the critical impurity. Although lower biocatalyst activity was observed at reduced reaction temperature, it was thought that the activity improvementsgained in Rd02 would allow increased activity to within the final biocatalyst loading target.After screening 2688 different variants, 145 showed higher activity FI. Table 5 showsthe top 50 variants identified from the second round of evolution.Table 5 FI[a]Entry Variant Mutations over M1FI of 6 area FI of 6 FI of 6 / 13 Unique x FI 6 / 13 area area counts[b]area1 M1 - 1.0 1 1.02 E9L; G58W; N137L; V242I 9.94 0.91 9.03 13 M2 G58W; N137L 7.98 1.22 9.75 14 E9L; G58N; N137L; V242I 7.39 1.18 8.73 15 N137Y; P138F 6.04 0.86 5.22 16 R56C; W61E 5.7 1.23 6.98 17 G58W; N137L; Q305L 5.69 1.32 7.52 18 N137L; E258A 5.48 1.21 6.63 19 N137G 5.34 1 5.19 210 W61M; N137L 5.27 1.29 6.77 111 E258A; A278M 5.03 0.98 4.9 112 E9Y; N137L 4.86 1.27 6.23 213 Q94R; A278M 4.67 1.15 5.35 114 E9L; N137L 4.58 1.19 5.41 215 N137L 4.25 1.14 4.9 516 G58W; N137L; V183C; V242I 4.17 1.04 4.34 117 N137F; P138F 4.12 1.04 4.28 1R56C; W61Y 4.04 1.3 5.27 1N137W; P138F 3.86 0.93 3.57 1G58W; W61M; N137L; V242I; 3.63 1.23 4.44 1K275NN137L; N243S 3.46 1.01 3.5 2E9L; W61M; N137L; K275N 3.29 1.29 4.23 1N137Y; P138L 3.11 0.97 3.03 1G58W; N137L; V183C 3.08 1.12 3.44 1N137W; P138L 3.03 0.98 2.99 3N137L; P138C 3.02 1.04 3.15 1N137G; P138Y 2.97 0.78 2.31 1N137A; P138A 2.93 0.93 2.73 1N137S; P138R 2.85 0.91 2.58 1N137C 2.74 1.11 2.94 6N137G; P138A 2.69 0.96 2.56 4W61M; N137C; R209A 2.68 0.7 1.88 1N137E; P138M 2.64 0.89 2.35 1N137D 2.6 0.98 2.53 1W61E 2.54 1.04 2.67 2N137D; P138R 2.49 0.97 2.41 1N137G; P138W 2.45 0.67 1.63 1N137W; P138W 2.43 0.92 2.23 1N137M; P138E 2.41 0.98 2.35 1R209A 2.4 1.24 2.97 1G58W; W61M; N137L; K275N 2.4 1.36 3.26 1R56V; P138L; M141I 2.4 0.85 2.03 1G58K; W61M; N137L 2.39 1.23 2.93 1N137E; P138L 2.36 0.8 1.89 2N137S 2.35 1.2 2.81 1N137A 2.3 1.3 2.99 1N243S 2.2 0.78 1.72 1G59V 2.18 1.03 2.25 1R56C 2.16 1.09 2.34 3N137D; P138C 2.09 0.84 1.76 1N137G; P138V 2.07 0.99 2.06 4[a] FI – Fold improvement. Calculated based on area of 6 and / or 13 detected by Rapidfire.Where variants with the same sequence were identified, an average is listed. [b] Number of unique variants which were found to have the same sequence. The most active enzyme was E9L; G58W; N137L; V242I, but Rapidfire analysisindicated a 10% increase in amide 13 formation. Variants G58W; N137L; Q305L; and E9L;W61M; N137L; K275N showed reduced amide formation compared to M1, in addition togood activity increase. As the primary reaction condition encompassed the enzyme properties desired, several variants were selected which showed the highest combined FIsfor 6 formation and 6 / 13 ratio for validation. Mutant G58W; N137L (M2) was selected as itshowed highest acid / amide ratio (6 / 13) of the variants tested in reactions at 40 ºC in addition to increased activity using 2% w / w of biocatalyst. Process development run with M2 at 2%w / w biocatalyst loading and 50 g / L of 4, gave complete conversion of 4 in <24 h andafforded 6 in 72.4% isolated yield after crystallisation, with 13 as the single detectableimpurity at <0.20% a / a (Table 6). Table 6: Overview of scale-up reactions using enzyme variants NIT2-D1, M1 and M2[a]Variant BufferTrinitrile 4 Nitrilase Tim Isolated yield Impurity 13 pH [mg / mL] loading e [h] Triacid 6 (relative to [% w / w][g] [%]Triacid 6)[b][% a / a]NIT2-D1 6.5[a] 50 50 21.5 13.1 86.8[c] ~0.50M1 (D192G)[d] 8.0 50 10 22 33.9 71.4[c] ~0.23M2 (G58W; 8.0 50 2 23 13.9 72.4[c] <0.20N137L)[e][a]Reactions were run at a concentration of 4 (50 g / L). Buffer pH was increased from 6.5 to8.0 as there was evidence that the pH dropped during reaction potentially slowing thehydrolysis rate (pH was not controlled). For all three enzymes all the nitrile 4 was consumed,conversion to products was >99%. [b]Determined from peak areas (13 / (6+13)x100 measured using HPLC. [c]Isolated yield for M1 and M2 smaller than NIT2-D1 as for M1 andM2 product 6 had been recrystalised.[d]Mutations over NIT2-D1. [e]Mutations over M1.Conclusions In summary, the inventors developed protein-based pharmacophores also known as “fingerprints” which allowed them to generate new nitrilase variants with a minimally perturbed active site and diversity throughout the remainder of the enzyme. After only two rounds of directed evolution, a new fit for purpose enzyme wasdeveloped, which showed 25-fold improvement in activity under process conditions as whencompared to the parent and importantly also possessed improved selectivity for triacid synthesis over primary amide critical impurity formation. Accordingly, the inventors have shown that the combination of active site fingerprinting with focused directed evolution is a technique that quickly delivers fit for purpose variants. The engineered nitrilase enzyme has enabled a new scalable route to the GalNAccluster. Nitrilase M2 is able to produce high-purity (>99.8%) triacid 6, a key crystallizableintermediate in the manufacture of the GalNAc cluster, from trinitrile 4.Further experiments Engineered enzyme M2 was subsequently utilised in Examples 4, 5 and 6 todemonstrate its wider utility in accepting other bulky trinitriles and converting them to theircorresponding crystallizable triacids, which in turn can be used in the production of furthertrivalent GalNAc clusters. The suitability of the combination of active site fingerprinting withfocused directed evolution to further increase the activity and selectivity of the nitrilase for agiven trinitrile has been demonstrated herein.Example 4: Preparation of 3,3'-((2-(12-(benzyloxy)-12-oxododecanamido)-2-((2-carboxyethoxy)methyl)propane-1,3-diyl)bis(oxy))dipropionic acid using a nitrilase Nitrilase M2 (220 mg) was weighed into a Falcon tube and resuspended in potassiumphosphate buffer (10 mL, 100 mM, pH 7.0) to a concentration of 22 mg / mL.900 µL of this solution was transferred to an Eppendorf Tube. Toluene (100 µL) was then added to the reaction mixture. The reaction was started by the addition of benzyl 12-((1,3-bis(2-cyanoethoxy)-2-((2-cyanoethoxy)methyl)propan-2-yl)amino)-12-oxododecanoate trinitrile (4 µL) and incubatingat 30 °C with shaking at 900 RPM in an Eppendorf Thermomixer. After 16 hours an aliquot (100 µL) was taken and diluted into 900 µL 1:1 ACN:H2O in an Eppendorf tube. The sample was clarified by centrifugation (10000 RPM for 1 minute) and the supernatant analysed using LC-MS Method 2. Analysis of the LC-MS data indicated the formation of the 3,3'-((2-(12-(benzyloxy)-12-oxododecanamido)-2-((2-carboxyethoxy)methyl)propane-1,3-diyl)bis(oxy))dipropionic acid (~3% conversion). MS-ES+(M+H) 640.39; MS-ES- (M-H) 638.41.Example 5: Preparation of 3,3'-((2-((6-(benzylamino)hexyl)amino)-2-((2-carboxyethoxy)methyl)propane-1,3-diyl)bis(oxy))dipropionic acid using a nitrilase Nitrilase M2 (220 mg) was weighed into a Falcon tube and resuspended in potassium phosphate buffer (10 mL, 100 mM, pH 7.0) to a concentration of 22 mg / mL.900 µL of this solution was transferred to an Eppendorf Tube. Toluene (100 µL) was then added to the reaction mixture. The reaction was started by the addition of 3,3'-((2-((6-(benzylamino)hexyl)amino)-2-((2-cyanoethoxy)methyl)propane-1,3-diyl)bis(oxy))dipropanenitrile (4 mg) and incubating at30 °C with shaking at 900 RPM in an Eppendorf Thermomixer. After 16 hours an aliquot(100 µL) was taken and diluted into 900 µL 1:1 ACN:H2O in an Eppendorf tube. The sample was clarified by centrifugation (10000 RPM for 1 minute) and the supernatant analysed using LC-MS Method 3. Analysis of the LC-MS data indicated the formation of the 3,3'-((2-((6-(benzylamino)hexyl)amino)-2-((2-carboxyethoxy)methyl)propane-1,3- diyl)bis(oxy))dipropionic acid (~4% conversion). MS-ES+(M+H) 527.30; MS-ES- (M-H) 525.37.Example 6: Preparation of 25,25-bis((2-carboxyethoxy)methyl)-23-oxo-1-phenyl-5,8,11,14,17,20,27-heptaoxa-2,24-diazatriacontan-30-oic acid using a nitrilaseNitrilase M2 (220 mg) was weighed into a Falcon tube and resuspended in potassiumphosphate buffer (10 mL, 100 mM, pH 7.0) to a concentration of 22 mg / mL.900 µL of this solution was transferred to an Eppendorf Tube. Toluene (100 µL) was then added to the reaction mixture. The reaction was started by the addition of N-(1,3-bis(2-cyanoethoxy)-2-((2- cyanoethoxy)methyl)propan-2-yl)-1-phenyl-5,8,11,14,17,20-hexaoxa-2-azatricosan-23-amide (4 µL) and incubating at 30 °C with shaking at 900 RPM in an Eppendorf Thermomixer. After16 hours an aliquot (100 µL) was taken and diluted into 900 µL 1:1 ACN:H2O in anEppendorf tube. The sample was clarified by centrifugation (10000 RPM for 1 minute) and the supernatant analysed using LC-MS Method 3. Analysis of the LC-MS data indicated the formation of the 25,25-bis((2-carboxyethoxy)methyl)-23-oxo-1-phenyl-5,8,11,14,17,20,27-heptaoxa-2,24-diazatriacontan- 30-oic acid (~32% conversion). MS-ES+(M+H) 763.36; MS-ES- (M-H) 761.41. SEQUENCE LISTING SEQ ID Sequence Identifier NO 1Acidithiobacillales bacterium SG8_45 Nitrilase NIT2-D1 (wild-type)2 Engineered Nitrilase M1 (Rd01)3 Engineered Nitrilase M2 (Rd02)SEQ ID NO:1 MPKIAVVQEAPKYLDKQGTIEKAVSLVEQAAANGAELVIFPEAFVPGYPAWIWRLRPGGDW SLNEEIHSRLLDNAVNIDSDDLAPLLASAKKHQVTIVCGMHERDNQLSQSTLYNTVVTIGPA GELINRHRKLMPTNPERMVWGFGDASGLKAIDTPVGKIGSLICWENYMPLARYAMYSQGV ELYIAPTYDSGDGWIGTMQHIAREGRCWVISSGVAIEASDLPEDFPGKKDLYPDAKEWVNP GDSTVIAPGGEIVEGPMRNEKGILYAEIDSKRAATAKRALDITGHYARPDIFQLHVNTEPQSP VKFSGKLE SEQ ID NO:2 MPKIAVVQEAPKYLDKQGTIEKAVSLVEQAAANGAELVIFPEAFVPGYPAWIWRLRPGGDW SLNEEIHSRLLDNAVNIDSDDLAPLLASAKKHQVTIVCGMHERDNQLSQSTLYNTVVTIGPA GELINRHRKLMPTNPERMVWGFGDASGLKAIDTPVGKIGSLICWENYMPLARYAMYSQGV ELYIAPTYGSGDGWIGTMQHIAREGRCWVISSGVAIEASDLPEDFPGKKDLYPDAKEWVNP GDSTVIAPGGEIVEGPMRNEKGILYAEIDSKRAATAKRALDITGHYARPDIFQLHVNTEPQSP VKFSGKLE SEQ ID NO:3 MPKIAVVQEAPKYLDKQGTIEKAVSLVEQAAANGAELVIFPEAFVPGYPAWIWRLRPWGDW SLNEEIHSRLLDNAVNIDSDDLAPLLASAKKHQVTIVCGMHERDNQLSQSTLYNTVVTIGPA GELINRHRKLMPTLPERMVWGFGDASGLKAIDTPVGKIGSLICWENYMPLARYAMYSQGVE LYIAPTYGSGDGWIGTMQHIAREGRCWVISSGVAIEASDLPEDFPGKKDLYPDAKEWVNPG DSTVIAPGGEIVEGPMRNEKGILYAEIDSKRAATAKRALDITGHYARPDIFQLHVNTEPQSPV KFSGKLE

Claims

CLAIMS1. Use of a nitrilase in the manufacture of an N-acetylgalactosamine (GalNAc) cluster.

2. The use according to embodiment 1, wherein the GalNAc cluster is a trivalent cluster.

3. The use according to embodiment 2, wherein the trivalent cluster is a tris-based trivalent cluster.

4. A method of producing a compound of formula (I):by reacting a compound of formula (II):with a nitrilase under suitable conditions, wherein R is a linker group suitable for attaching toa therapeutic oligonucleotide.

5. The method according to claim 4, wherein the compound of formula (I) is crystalline.

6. The method of claim 4 or claim 5, wherein the reaction temperature is from 30 ºC to 45 ºC.

7. The method of claim 6, wherein the reaction temperature is about 30 ºC.

8. The method of any one of claims 4 to 7, wherein the reaction pH is between 4 and10.

9. The method of claim 8, wherein the reaction pH is about 7.

10. The method of any one of claims 4 to 9, wherein the impurity level is <0.20 a / a.

11. The method of any one of claims 4 to 10, wherein complete conversion of substrateto product takes ≤ 24 hours.

12. A method of manufacturing of an N-acetylgalactosamine cluster comprising the methods of any one of claims 1 to 11.

13. The method of claim 12, wherein the N-acetylgalactosamine cluster is:, wherein G is N-acetylgalactosamine, n is 0 to 10, and R is a linker group suitable for attaching to a therapeutic oligonucleotide.

14. A method of manufacturing a GalNAc-conjugated oligonucleotide, comprising a stepof making a GalNAc cluster using a nitrilase.

15. A method according to claim 14, which comprises a step of comprising conjugating aGalNAc cluster to an oligonucleotide or a portion thereof under suitable conditions.

16. A method according to claim 15, which comprises conjugation of the GalNAc clusterto an oligonucleotide monomer or a protected version thereof.

17. A method according to claim 15, which comprises conjugation of the GalNAc clusterto a shortmer or a protected version thereof.

18. A method according to claim 15, which comprises conjugation of the GalNAc clusterto an oligonucleotide or a protected version thereof.

19. A GalNAc-conjugated therapeutic oligonucleotide obtained according to the methodof any one of claims 14 to 18.

20. A GalNAc-conjugated oligonucleotide for use in therapy, wherein the GalNAc-conjugated oligonucleotide comprises a GalNAc cluster made using a nitrilase.

21. An engineered nitrilase comprising a sequence that is at least 90% identical to SEQID NO:1 and includes a catalytic triad consisting of E, K and C at positions equivalent to E57, K147 and C181 in SEQ ID NO:1.

22. The engineered nitrilase according to claim 21, wherein the amino acid at a positionequivalent to position 158 in SEQ ID NO:1 is an aromatic amino acid.

23. The engineered nitrilase according to claim 20, wherein the aromatic amino acid isselected from the group consisting of: F, H, W and Y.

24. The engineered nitrilase according to any one of claims 21 to 23, wherein one ormore amino acids at positions equivalent to the following positions in SEQ ID NO:1 have been mutated: 9, 56, 58, 61, 94, 137, 138, 141, 183, 192, 242, 243, 258, 275, 278, 305.

25. The engineered nitrilase according to any one of claims 21 to 23, wherein amino acidposition 9 is L or Y; position 56 is C, F or V; position 58 is K, N or W; position 59 is V; position 61 is E, H, I, L or M; position 94 is R; position 137 is A, C, D, E, F, G, L, M, S, T, W, or Y; position 138 is A, C, E, F, L, M, R, W or Y; position 141 is I; position 183 is C; position 192 is G; position 209 is A; position 242 is I; position 243 is S; position 258 is A; position 275 is N; position 278 is M; and / or position 205 is L; wherein the foregoing amino acid positions are numbered with reference to SEQ ID NO:1.

26. The engineered nitrilase according to any one of claims 21 to 23, wherein thenitrilase comprises G at a position equivalent to position 192 in SEQ ID NO:1 and a set of mutations as set out in any row of Table 5.

27. An engineered nitrilase comprising or consisting of an amino acid sequence ofSEQID NO:2 or SEQ ID NO:3.

28. An engineered nitrilase comprising a sequence that is at least 90% identical to SEQID NO:1 and includes a catalytic triad consisting of Glu, Lys, and Cys, wherein the engineered nitrilase exhibits improved nitrilase activity relative to the nitrilase of SEQ ID NO:1 for converting a trinitrile substrate to a triacid product.

29. The engineered nitrilase according to claim 28, wherein the trinitrile substrate is acompound of formula (II):and the triacid product is a compound of formula (I):wherein R is a linker group suitable for attaching to a therapeutic oligonucleotide.

30. The engineered nitrilase according to claim 28 or claim 29, wherein the engineerednitrilase has reduced side product formation relative to SEQ ID NO:1, optionally wherein theside product level is <0.2% a / a.

31. The engineered nitrilase according to any one of claims 27 to 29, wherein theimproved activity is complete conversion of the trinitrile product to triacid substrate in less than 24 hours.

32. A polynucleotide encoding the engineered nitrilase according to any one of claims 21to 31.

33. An expression vector comprising the polynucleotide according to claim 32.

34. A host cell comprising the expression vector according to claim 33.

35. A method of preparing an engineered nitrilase comprising culturing a host cellaccording to claim 34 under suitable conditions.

Images