Chemical Ligation: Tools for Biomolecule Synthesis and Modification

Presenting a wide array of information on chemical ligation – one of the more powerful tools for protein and peptide synthesis – this book helps readers understand key methodologies and applications that protein therapeutic synthesis, drug discovery, and molecular imaging.

•    Moves from fundamental to applied aspects, so that novice readers can follow the entire book and apply these reactions in the lab
•    Presents a wide array of information on chemical ligation reactions, otherwise scattered across the literature, into one source
•    Features comprehensive and multidisciplinary coverage that goes from basics to advanced topics
•    Helps researchers choose the right chemical ligation technique for their needs

• Language: English
• Publisher: Wiley
• Release date: Mar 27, 2017
• ISBN: 9781119044130

Book preview

List of Figures

Figure 1.1 Schematic representation of the solid-phase peptide synthesis.

Figure 1.2 Chemical preparation of a polypeptide by convergent synthesis in solution phase.

Figure 1.3 The chemical ligation general concept. A single polypeptide chain is obtained by covalently joining two peptide segments through the reaction of two mutually reactive functional groups. The type of covalent bond generated at the junction site depends on the reactive groups employed for the ligation reaction.

Figure 1.4 Derivatization of the C-terminus of a recombinant protein with an oxyamine functional group using intein chemistry.

Figure 1.5 The native chemical ligation reaction mechanism.

Figure 1.6 Amino acid analogs used in native chemical ligation–desulfurization synthetic strategies.

Figure 1.7 The mechanism of intein splicing.

Figure 1.8 Preparation of recombinant α-thioester proteins (a) or N-terminal cysteinyl-proteins (b) using engineered inteins.

Figure 1.9 Schematic mechanism of protein trans-splicing. IntN and IntC are the halves, N- and C-terminal, respectively, of a split intein.

Figure 2.1 Structure of the bis(2-sulfanylethyl)amido (SEA) group.

Figure 2.2 Protein synthesis by N-to-C solid-phase sequential ligation of unprotected peptide segments using the SEAoff group as a latent thioester surrogate. The peptide segment elongation cycle consists in activating the SEAoff group by a SEA–thiol exchange reaction and then performing an NCL reaction in the presence of MPAA (see also Scheme 2.11). (a) Synthesis of a model 135 amino acid polypeptide. A sample was treated with NaOH after each elongation cycle and analyzed by LC–MS to verify the completeness of the coupling step (right). (b) Characterization of the purified 135-amino-acid model polypeptide.

Figure 2.3 Design of the multivalent semisynthetic NB and K1B-streptavidin scaffolds.

Figure 2.5 Binding of NB/S or K1B/S complexes to purified recombinant or endogenous MET receptor. (a) NB, K1B and MET-Fc binding assay: increasing concentrations of NB or K1B were mixed with extracellular MET domain fused with human IgG1-Fc (MET-Fc) and incubated with streptavidin AlphaScreen® donor beads and Protein A acceptor beads. Error bars correspond to standard error (±SD) of triplicates. (b) Endogenous MET capture. Streptavidin-coated beads loaded with NB or K1B were incubated with HeLa or CaPan1 total cell lysates. Input, flow-through, and elution fractions from NB or K1 loaded beads were analyzed by specific total MET Western blot.

Figure 2.6 HeLa cells were treated with increasing concentrations of mature HGF/SF, K1B/S, NK1, and K1B/Ab for 7 min. Activation levels of Akt (a) and ERK (b) were measured using HTRF technology and plotted as the 665/620 nm HTRF signal ratio. (c) Cell scattering assay. MDCK isolated cell islets were incubated for 18 h in culture media with HGF/SF (HGF), K1B, K1B/S, and NK1. Cells were then stained and observed under microscope (40×). (d) Angiogenesis. Mice were injected with a mixture of Matrigel and HGF/SF (HGF), VEGF, NK1, K1B/S, K1B, or S. Hemoglobin absorbance was measured and concentration was determined using a rate hemoglobin standard curve and plug weight.

Figure 5.1 Native chemical ligation reaction mechanism demonstrated for the synthesis of glycosylated XCL1. The first step is an intermolecular trans-thioesterification, where the thioester is cleaved from the N-terminal fragment. Next a spontaneous intramolecular S→N acyl transfer leads to the native peptide bond between both protein segments.

Figure 5.2 Generation of lipidated chemokines demonstrated for CXCL12. The first CXCL12 (1–33) fragment is synthesized by SPPS coupling the β-carboxy group of an Asp-Oallyl-amino acid at position 33 on Rink amide resin. Next, the allyl protection group is removed and thioesterification of the segment is performed with ethyl-2-mercaptopropionate. The CXCL12 fragment is cleaved from the resin as amide leading to the native Asn amino acid at position 33. The fragment CXCL12 (34–68) that is lipidated is synthesized by SPPS with a Dde protection at Lys68, the lipidation position. After Dde cleavage, Fmoc-Glu is attached at that position via its γ-carboxy group. Fmoc is removed by piperidine, and the fatty acid is coupled at the amino function of the carboxyl group of Glu. Finally, the peptide is cleaved from the resin and NCL is performed.

Figure 5.3 The approach of N ent-0F8FF C directed NCL demonstrated for the generation of [G49]-CXCL14. For the N ent-0F8FF C directed NCL, three fragments of CXCL14 have been synthesized, where the first and third segments are generated by Fmoc strategy. The second part is prepared by using the specific (N-Fmoc-glycyl-N-sulfanylmethyl)aminobenzoic acid linker-incorporated resin. It is now possible to cleave the peptide as cysteinyl peptide thioacid by NaSH. The first NCL occurs between the first and second fragments using thiophenol as activator. Next, the thioacid at the C-terminus is converted into a thioester by the application of Ellman's reagent and KHCO3. The second NCL is performed by using this thioester and the Cys of the N-terminus of the third part leading to the full length [G49]-CXCL14.

Figure 5.4 Solid-phase NCL demonstrated for the synthesis of the chemokine vCCL3 combined with a safety-catch linker (SCAL). The SCAL is colored in gray and allows the synthesis of an amidated C-terminus. Therefore, an agarose resin is coupled with a Cys, which can react with the thioester function of the SCAL linker. The three vCCL3 fragments have been prepared by Fmoc SPPS strategy, whereas the third fragment the SCAL is coupled to the resin, prior to the peptide synthesis. After chemical ligation of the SCAL with the modified agarose resin, C ent-0F8FF N directed NCL reactions can be performed using the fragment vCCL3 (13–35) and next vCCL3 (1–12). At the end, the full-length chemokine can be cleaved from the resin.

Figure 5.5 Mechanism of the tandem peptide ligation represented for the generation of the vCCL3 chemokine. The tandem peptide ligation is a combination of the formation of a pseudoproline and a subsequent NCL. For the generation of the pseudoproline bond, the N-terminal amine of the second fragment reacts with the aldehyde function of the C-terminal part of the first peptide segment. After this imine capture in acidic environment, the thiol interacts and a ring chain tautomerization occurs. The pseudoproline reaction is finished by an intramolecular O ent-0F8FF N acyl transfer. The subsequent NCL is performed under mild conditions at pH ≥⃒ 7.0. The typical reaction mechanism takes place leading to a CCL3 variant including a pseudoproline bond.

Figure 5.6 Reaction mechanism of the Ag+-catalyzed (left) and non-Ag+-catalyzed (right) chemical ligation illustrated for the generation of CCL27. CCL27 is synthesized into three fragments, where the second and the first segments contain C-terminal linear thioesters. The sulfur interacts with the Ag+ and at the same time the hydroxyl function of the HOOBt with the carbonyl C-atom during the Ag+-catalyzed reaction. The thioester is cleaved from the second fragment, and the amino function of the third fragment can react with the activated carbonyl C-atom leading to a native peptide bond. After Fmoc cleavage, the reaction is repeated with the first peptide fragment. In comparison to the Ag+-catalyzed reaction, the Ag+-free chemical ligation is performed with the new thioesterification at the C-terminus using 4-mercaptophenylacetic acid (MPAA). By the connection of MPAA to the C-terminus, HOOBt can directly react with the carbonyl C-atom using the same reaction mechanism that leads to the same product as for Ag+-catalyzed reaction. The difference to the silver-catalyzed reaction is the necessity to additionally cleave the Acm protection groups from the side chains of Cys.

Figure 5.7 Expressed protein ligation mechanism demonstrated for the labeling of CXCL8. (a) One part of the protein is obtained by the use of the IMPACT™ system, where the target protein is linked to the intein-CBD. The CBD can bind to the chitin column and subsequently purified. After an intramolecular N ent-0F8FF S acyl transfer, the cleavage is performed by the addition of an excess of thiols such as MESNa. The second fragment [K69(CF)]-CXCL8 (56–77) is synthesized by SPPS with an N-terminal Cys. In the final step, both fragments react with each other by NCL leading to a fluorescently labeled CXCL8 chemokine.

Figure 5.8 Mechanism of the controlled photoactivation of chemokines illustrated for CXCL12. The CXCL12 fragment M-[A49]-CXCL12 (1–49) was expressed and purified by the IMPACT™ system, whereas the second fragment CXCL12 (50–68) is synthesized by SPPS and includes the exchange of position 56 or 58 by Nvoc-serine and Nvoc-homoserine, respectively. Coupling of amino acid 55 or 57 via the side chain leads to a depsipeptide bond. After NCL of both fragments and protein refolding, the Nvoc can be removed by UV irradiation. The free amino group attacks the carbonyl C-atom and leads to a native peptide bond, where the protein changes into its native active tertiary structure. Thus, a photoactivatable version of CXCL12 has been achieved.

Figure 6.1 Protein synthesis by the thioester method.

Figure 6.2 Synthetic route of glycoprotein by the thioester method.

Figure 6.3 Synthesis of the Ig domain of emmprin.

Figure 6.4 Synthesis of mucin model by the thioester method.

Figure 6.5 Synthesis of glycopeptide dendrimer.

Figure 6.6 Synthesis of TIM-3 Ig domain by the one-pot ligation.

Figure 6.7 Resynthesis of the Ig domain of emmprin by the one-pot ligation method.

Figure 7.1 Zheng et al.'s two-step conversion of peptides with removable Arg-tagged backbone modification to the native peptide [32].

Figure 8.2 UAAs used in residue-specific UAA mutagenesis.

Figure 8.3 UAAs used in site-specific UAA mutagenesis.

Figure 9.1 Conjugation of proteins modified with hydrazines or alkoxyamines.

Figure 9.2 Pharmacokinetic analysis of circulating CT-tagged α-HER2 thioPz-Glu-PEG2-Maytansine following a single 5 mg kg−1 i.v. bolus injection in mice. Analyte concentrations were determined by ELISA. Total antibody was captured with anti-human IgG and detected with anti-human Fc. Total ADC was captured with anti-human Fab and detected with a mouse anti-maytansine primary followed by an anti-mouse IgG1 secondary. Calculated ADC half-life: 7.6 days; error bars represent 1 SD; n = 3 mice/time point.

Figure 10.1 Synthesis of thioacylating monomers from commercially available Fmoc-protected amino acids. (a) IBCF, NMM, THF, 0 °C, 15 min; N-Boc-OPD or 4-nitro-OPD; (b) Lawesson's reagent or P4S10, CH2Cl2, or THF, 2–18 h; (c) R2═F, R3═H; phosgene or triphsogene, CH2Cl2; (d) R2═NO2, R3═H; NaNO2, 95:5 AcOH/H2O, 30 min; (e) R2═H, R3═Boc; 50:50 TFA/CH2Cl2, 0 °C, 2 h; then NaNO2, 95:5 AcOH/H2O, 30 min. IBCF = isobutyl chloroformate.

Figure 10.2 Synthesis of doubly labeled thioproteins. (a) Expressed protein fragment containing Cnf from Uaa mutagenesis and an N-terminal Cys. Ligation with a thiopeptide yields double-labeled protein containing a free cysteine. (b) Expressed protein fragment containing Cnf as a Uaa and an N-terminal Lys or Arg. Hcm is transferred to the N-terminal residue via AaT. The resulting protein can undergo NCL with a thiopeptide and subsequent methylation to yield double-labeled protein with a native methionine at the ligation site.

Figure 10.4 Photocontrol of RNase S hydrolysis of cytidine 2′,3′-cyclic monophosphate. Binding of the thioamide containing RNase S peptide to the RNase S protein produces the active enzymatic complex. Irradiation of the thioamide results in trans to cis isomerization (represented by the lighter segment of the peptide) and halts catalytic hydrolysis. Thermal relaxation of the thioamide bond to the trans conformation revives activity.

Figure 10.5 Fluorogenic protease-sensing peptides. (a) The LSLKAAμ peptide, where μ represents 7-methoxycoumarin-4-ylalanine, is quenched by the proximal thioamide. Proteolytic cleavage at the Lys-Ala bond liberates the coumarin-containing peptide resulting in a turn-on of fluorescence. (b) Dithioamide-containing peptides can be used to enhance the turn-on of protease-sensing peptides via increased fluorescence quenching in the uncleaved peptide.

Figure 10.6 FRET/PET quenching during α-synuclein compaction. With increasing concentration of TMAO, α-synuclein undergoes successive compation. During this process, changes in the distance between two regions of the protein have been monitor by FRET (a) where quenching of Cnf fluorescence by thioamide occurs in a distance-dependent manner and PET (b) where van der Waals contact between fluorescein and thioamide is required for quenching.

Figure 11.1 Chemical structures of cyclopeptide natural products generate from nonribosomal (NRPS) and ribosomal (RiPPs) biosynthetic pathways and representative structures of macrocyclic organo-peptide hybrids obtained according to the intein-based methodologies highlighted in the text.

Figure 11.2 General scheme for the modular synthesis of macrocyclic organo-peptide hybrids (MOrPHs) via a dual ligation reaction between a genetically encoded biosynthetic precursor (BP) and a chemically synthesized synthetic precursor (SP). UAA: unnatural amino acid.

Figure 11.3 Overview of the copper-dependent and the catalyst-free methods for the synthesis of macrocyclic organo-peptide hybrids (MOrPHs). (a) CuAAC/hydrazide-mediated ligation method. (b) Oxime/AMA-mediated ligation method. The different structures of the nonpeptidic components are illustrate in the boxes. CuAAC: copper-catalyzed alkyne/azide cycloaddition; AMA: 2-amino-mercaptomethyl-aryl.

Figure 11.6 MOrPH-based inhibitors of p53:HDM2/X interaction. (a) Crystal structure of the complex between p53 transactivation domain and the p53-binding domain of HDM2. The triad of cofacial residues in p53 helix which insert into HDM2 binding cleft are labeled. The model of a representative MOrPH-based HDM2/X inhibitor is provided (i/i + 10 peptide cyclization with SP8). (b) Sequence and in vitro inhibitory activities for the linear p53-derived peptide and the MOrPH-based inhibitors. The structures of the linker SP6, SP8, and SP4 (negative control) are indicated.

Figure 12.1 Schematic presentation of the chemical mechanism of protein splicing and expressed protein ligation (EPL). (a) The four reaction steps involved in protein splicing by canonical inteins. Step 1, N ent-0F8FF S(O) acyl shift: thiol or hydroxyl group of the first residue of the intein attacks the preceding peptide bond forming a (thio)ester bond; step 2, trans-(thio)esterification: thiol or hydroxyl group of the nucleophilic +1 residue replaces the (thio)ester bond formed by the first residue of the intein resulting in a branched intermediate; step 3, Asn cyclization: the intein is cleaved off from the exteins; step 4, S(O) ent-0F8FF N acyl shift: the released branched (thio)ester intermediate undergoes spontaneous rearrangement to form a more energetically stable peptide bond between N- and C-exteins. (b) Chemical reaction steps of EPL. The precursor protein containing a protein of interest (POI) fused to the inactivated intein is immobilized on the column by an affinity tag for purification. Step 1, N ent-0F8FF S acyl shift: the first step of EPL is N ent-0F8FF S acyl shift induced by the partially inactivated intein; step 2, thiol modification and cleavage: thiol agent cleaves the thioester bond in the precursor to release POI with the C-terminal thioester; step 3, elution: cleaved POI with the C-terminal thioester is eluted from the column; step 4, NCL step: the released POI with the C-terminal thioester reacts with an N-terminal cysteinyl peptide to form a peptide bond by NCL. ExteinN and ExteinC stand for N- and C-exteins, respectively. Asterisks indicate the mutated intein.

Figure 12.2 Protein trans-splicing (PTS) and the split sites on the intein structure. (a) Schematic presentation of PTS. An intein is split into interacting halves, termed N- and C-inteins (IN and IC). (b) Cartoon presentation of the three-dimensional structure of a HINT domain (NpuDnaE intein, PDB ID: 2KEQ) showing the natural split site (C35) and two artificially engineered split sites (C6 and N11). C35 site is located at the conserved HEN insertion site before 35 residues from the C-terminus. The C6 and N11 split inteins were created by shifting the split site toward either C- or N-termini to have 6-residue C-intein and 11-residue N-intein, respectively. The shorter fragments of the split inteins are shown in black. Black arrows indicate the split sites and the conserved HEN insertion site. N and C show the N- and C-termini, respectively.

Figure 12.3 Segmental isotope-labeling and protein semisynthesis by PTS. (a) Segmental isotopic labeling in vitro. The N- and C-terminal fragments of POI are fused to N- and C-inteins, respectively. One of the precursors is expressed in isotopic (¹⁵N)-labeled culture medium and the other precursor in unlabeled culture medium. The two precursors are purified separately and mixed together for PTS by refolding. (b) Protein semisynthesis. An N-terminal fragment of POI is fused with N-intein and recombinantly expressed. The C-terminal precursor containing C-intein and a chemical label is chemically prepared. The two precursor fragments are mixed to initiate PTS, thereby producing the semisynthetically ligated product. IN and IC stand for N- and C-inteins, respectively. POIN and POIC stand for the N- and C-terminal fragments, respectively, of POI.

Figure 12.4 Intra- and intermolecular applications with protein cis-splicing (PS) and trans-splicing (PTS). (a) Small-molecule induced CPS: Split intein halves with low affinity are fused with two interacting domains in the presence of a small molecule. Addition of a small molecule induces PTS upon association of the two precursors. (b) Light-induced CPS: light removes a protective group stalling PTS. (c) Three-fragment ligation. POI is split into three pieces (POIN, POIM, and POIC) and fused with two orthogonal split inteins (IN1/IC1 and IN2/IC2). (d) Intein-mediated protein alternative splicing (iPAS). Intermolecular protein splicing by protein 3D-domain swapping produces alternatively ligated products instead of cis-spliced products. (e) Cyclization and polymerization. Circularly permutated precursor bearing a split intein ligates the N- and C-termini of a protein by intramolecular protein splicing, thereby resulting in the backbone cyclization. The identical precursor protein could result in polymerization when intermolecular interactions dominate intramolecular interactions. IN and IC stand for N- and C-inteins, respectively.

Figure 12.5 HINT domains and protein splicing by BIL domain. (a) A superposition of the structures of DnaE intein from Synechocystis sp. PCC6803 (SspDnaBΔ275 intein, PDB ID: 1MI8) colored in light gray and BIL4 domain from Clostridium thermocellum (CthBIL4, PDB ID: 2LWY) colored in black. N- and C indicate the N- and C-termini. (b) The mechanism of protein splicing by BIL domains involves the following steps: step 1, N ent-0F8FF S(O) acyl shift forming a (thio)ester bond; step 2, Asn cyclization cleaving off the C-flank from the BIL domain; step 3, aminolysis by the N-terminal amine group of the cleaved C-flank. FlankN and FlankC stand for the N- and C-terminal flanking sequences, respectively.

Figure 12.6 Junction sequence dependencies of different inteins. (a) Numbering of junction residues of an intein-containing precursor. The −1 and −2 positions are the first and second residues preceding the intein sequence. The +1 and +2 positions are the first and second residues following the intein sequence. The first residue of an intein is indicated as 1. The +1 residue is the nucleophilic residue and either Cys or Ser or Thr in inteins. (b) Cis-splicing efficiencies of DnaE intein from Nostoc punctiforme (NpuDnaE intein) with 20 different amino-acid types at the N- and C-terminal junctions (the −1 and +2 positions). (c) Cis-splicing efficiencies of minimized DnaB intein from Nostoc punctiforme (NpuDnaBΔ²⁹⁰ intein) with 20 different amino-acid types. The left and right panels show the data for the −1 position and the +2 position, respectively. Amino-acid types are indicated by one-letter codes at the bottom. The standard deviations of the mean (n = 3) are shown as error bars.

Figure 13.1 (a) General principle of bioorthogonal reactions. X and Y moieties are orthogonal to the functional groups found in biological environment. (b) Chemical ligation reactions, reagents, and ligation products generated with the respective reaction.

Figure 13.2 (a) Staudinger reaction [12]. (b) Reaction mechanism for Staudinger ligation [13]. (c) Traceless Staudinger ligation [14–17].

Figure 13.5 (a) An example of SPAAC-based synthesis of PET probes. 18F-DIBO (1) is conjugated to an azido-modified targeting moiety such as geldanamycin (2), an inhibitor of Hsp90 protein, and TATE peptide (3), a somatostatin 2 agonist. Examples are taken from Bouvet et al. and Arumugam et al. [117, 118]. (b, c) Antibody-based in vivo pretargeting methodology using IED DA reaction for PET imaging. (b) IED DA reaction for conjugation of a TCO-modified antibody (A33-TCO) as targeting moiety and ⁶⁴Cu–NOTA–Tetrazine construct (⁶⁴Cu–Tz–Bn–NOTA) as tracer. (c) Schematic of the antibody-based in vivo pretargeting methodology using IED DA reaction between A33-TCO and ⁶⁴Cu–Tz–Bn–NOTA. Examples are taken from Zeglis et al. [119].

Figure 14.1 Segmental isotopic labeling. An isotopic-labeled domain and an unlabeled domain are separately prepared and then joined through a native chemical ligation reaction to form the native protein of interest.

Figure 14.3 Preparation of protein with an NMR-invisible solubility tag by protein trans-splicing. The protein of interest is fused to the solubility tag and a split intein (IntC) portion and expressed using a minimal medium with heavy isotopes (¹³C, ¹⁵N, etc.). The other split intein (IntN) portion is fused to the solubility tag and expressed in standard medium. Intein assembly and splicing exchange the solubility tag domains, appending an unlabeled tag domain to the labeled protein of interest.

Figure 14.4 Schematic procedure to prepare double-modified proteins combining EPL and chemoselective labeling in solution.

Figure 14.5 Schematic mechanism of intein splicing.

List of Plates

Figure 2.4 Chemical synthesis of NB and K1B proteins using the one-pot three peptide segment assembly process.

Figure 2.7 (a) FVB mice were injected intravenously with anti-Fas monoclonal antibody (AbFAS) mixed with K1B, K1B/S, NK1, or HGF/SF or PBS. A second injection without anti-Fas was performed 90 min later. Livers were extracted and fixed in formalin after three additional hours. (b) FVB mice were injected with an increased concentration of K1B/S complex, K1B, or NK1. After 10 min, livers were extracted, snap-frozen, and crushed. Cell lysates were analyzed by specific total MET, Akt and ERK or phospho-MET, phospho-Akt, and phospho-ERK Western blot.

Figure 7.2 Retrosynthesis scheme for the synthesis of DAGK. Figure was kindly provided by C. F.W. Becker.

Figure 7.3 (a) Ribbon diagram of the crystal structure of the transmembrane part of the NpSRII/NpHtrII complex [2]. (b) Scheme of the synthesis of the NpHtrII1-114 domain containing 114 amino acids.

Figure 7.4 The selectivity filter of K+ channels. (a) Ribbon representation of two opposite subunits of the KcsA channel. Green spheres illustrate the potassium ions in their binding sites. (b) Close-up view of the selectivity filter. In a functional channel, sites 1 and 3 are in equilibrium with sites 2 and 4.

Figure 8.1 Reagents and products of 1,2-aminothiol condensations.

Figure 10.3 Future directions for incorporation of thioamides into full-length proteins. (a) Three-part ligation. The intein generated protein–thioester 1 undergoes ligation with 2 to generate thioprotein fragment 3 after oxidation and treatment with a thiol. The C-terminal fragment is expressed with an N-terminal Lys, and Hcm is transferred by AaT to yield 5. Ligation proceeds between 3 and 5, followed by methylation of the two Hcs residues with CH3I to yield a full-length thioprotein with Met residues at each ligation site. (b) Expressed protein containing a side-chain thioamide and N-terminal Cys is ligated with a backbone thiopeptide to result in a side chain–backbone dithioamide protein. (c) A peptide fragment containing an N-terminal cysteine surrogate undergoes NCL with a thiopeptide. Subsequent desulfurization yields a thiopeptide with a non-Cys amino acid at the ligation site.

Figure 11.4 Representative structure-reactivity data for oxime/AMA-mediated ligation method. (a) Percentage of MOrPH product upon reaction of different oxyamino/amino-thiol synthetic precursors (SP4–SP7) with biosynthetic precursors containing peptide target sequences from 4 to 15-amino acid residues long (12 hour time point). (b) Reactivity of the 20 biosynthetic precursor I-1 variants. Left: Plot of the amount of full-length protein after purification from Escherichia coli versus yield of SP4-induced macrocyclization (5 h). Right: Overall yield of MOrPH product as given by the (% of full-length protein) × (% of SP-induced macrocyclization) yield from full-length precursor protein. (c) Amount of SP4-induced protein splicing for the proline-scan and glycine-scan biosynthetic precursor libraries and the reference X5T library. The target sequence, in which X corresponds to a fully randomized position (NNK codon), is specified.

Figure 11.5 Bicyclic MOrPHs. (a) Strategy for synthesis of bicyclic organo-peptide hybrids via post-cyclization oxidation of MOrPHs derived from cysteine-containing target sequences. The oxyamine/amino-thiol synthetic precursor, the unnatural amino acid (p-acetyl-Phe), and the reactive cysteine residue within the target peptide sequence are highlighted in blue, green, and red, respectively. (b) Amino acid sequences of the precursor polypeptides utilized for synthesis of bicyclic MOrPHs. (c) Representative MALDI-MS spectra illustrating the clean formation of the macrocyclic (m) and bicyclic product (b) before and after the oxidation step.

Figure 13.3 (a) Metabolic introduction of chemical reporters on biomolecules in cell. Cells are incubated with unnatural metabolic precursors bearing the chemical reported (e.g., azide) and are incorporated in biomolecules using the metabolic machinery of the cells. The reporter is selectively reacted with an exogenous probe using a compatible bioorthogonal reaction. (b) Fluorescence microscopy of CHO cells labeled with Cy5.5–phosphine conjugate. Cells, incubated with Ac4ManNAz (100 μM) for 3 days, were treated with Cy5.5–phosphine conjugate (200 μM) for 2 h at 37 °C. The cells were then fixed and permeabilized with MeOH and stained with DAPI before imaging. Red: Cy5.5 channel. Blue: DAPI channel. Scale bar: 20 µm.

Figure 13.4 (a) Principle of caged-luciferin approach. d-Luciferin can be modified with caging groups on the phenol, amino, or carboxylic acid group prohibiting the interaction with Fluc until caging moiety is removed. (b) Use of a phosphine–luciferin conjugate (caged luciferin) for bioluminescence imaging of cell-surface azido-sugars. The phosphine–luciferin conjugate releases d-luciferin upon Staudinger ligation with azides. Then, free d-luciferin diffuses into cells, and luciferase-catalyzed conversion of d-luciferin to oxyluciferin yields a photon of light, which is detected using CCD camera. (c) Imaging of cell-surface azido-sugars with phosphine–luciferin conjugate. LNCaP-luc cells, a prostate cancer cell line stably transfected with luciferase, were incubated for 2 days at various concentrations of the following azido-modified sugars: Ac4ManNAz, Ac4GalNAz, Ac4GlcNAz, or media. The cells were washed thrice with 200 μL of PBS and then treated with the phosphine–luciferin probe (100 μM) for 120 min. Error bars represent the standard deviation of the mean for three replicate experiments. Source: (b, c) Cohen et al. (2010) [26]. Reproduced with the permission of American Chemical Society.(d) The split luciferin reaction. Condensation reaction between 6-amino or 6-hydroxy-2-cyanobenzothiazole and d-cysteine forming d-aminoluciferin or d-luciferin, respectively, as product. (e) Caged d-cysteine for caspase-3/7 imaging using the split luciferin reaction. (f) Imaging caspase-3/7 in FVB-Luc+ mice using the DEVD-(d-Cys) peptide and H2N-CBT. Graph represents the total luminescence over 1 h from transgenic reporter mice treated with either PBS (control group) or combination of LPS (100 µg kg−1 in 50 μL of PBS) and d-GalN (267 mg kg−1 in 50 μL of PBS). Six hours posttreatment, the animals received IP injections of a combination of DEVD-(d-Cys) peptide (22.6 mg kg−1 in 100 μL of PBS) and H2N-CBT (6.8 mg kg−1 in 20 μL of DMSO). Statistical analyses were performed with a two-tailed Student's t-test. **P < 0.01 (n = 4). Error bars are ±SD. (g) Representative image of mice treated with LPS and d-GalN or vehicle, 15 min postinjection of a combination of DEVD-(d-Cys) and H2N-CBT reagents. Source: (e, f, g) Godinat et al. (2013) [85]. Reproduced with the permission of American Chemical Society.

Figure 13.6 (a) Imaging sialic acid on tumor glycans via SL. Sialic acid in the tumor glycans was metabolically labeled in living mice upon intraperitoneal injection of Ac4ManNAz (1), which passively diffuses into cells and is deacetylated by intracellular carboxyesterases. The deacetylated product, ManNAz, is then incorporated into glycans by the cellular metabolic machinery (2). Next, a biotinylated phosphine (bPhp) is injected intraperitoneally (3) reacting specifically with the azido-labeled sialic acid on the cell-surface glycans (4) via SL. Biotin–sialic acid conjugate is then detected by subsequent intravenous injection of a radionuclide-labeled neutral and deglycosilated avidin (NA) (5). Example is taken from Neves et al. [126]. (b–d) Tumor pretargeting methodology using IED DA reaction. (b) Schematic of the antibody-based in vivo pretargeting methodology. (c) Pretargeting components in (b): tetrazine–DOTA (9) for ¹¹¹In chelation and transcyclooctene–NHS (10) for CC49 antibody functionalization via lysine chemistry. (d) Small-animal SPECT/CT imaging of live mice bearing colon carcinoma xenografts: posterior projection of mice, which were preinjected with (i) CC49-TCO, (ii) CC49, or (iii) Rtx-TCO, followed 1 day later by [¹¹¹In]-9; (iv–vi) single transverse slices (2 mm) passing through the tumors in (i–iii). Examples are taken from Rossin et al. [127].

Figure 13.7 (a) In vitro and in vivo clicking NPs. Two complementary IONPs were designed to undergo a bioorthogonal reaction after cleavage by MMP enzymes, which exposes the azide or alkyne moieties on either set of NPs (MNP = magnetic nanoparticle, PEG = polyethylene glycol). (b) Molecular structures of the final NP (1) and controls; nontargeted NP (azide family) (2), NP without self-assembling properties (azide family) (3). (c) Hydrodynamic size measurements. Blue = 1:1 mixture of alkyne and azide NPs. Red = 1:1 mixture of the alkyne and azide NPs in the presence of hrMMP9. Purple = alkyne NPs in the presence of hrMMP9. Green = azide NPs in the presence of hrMMP9. (d) Representative tumor T2 maps are shown from a series of spin echo images acquired with different TEs before (4) and 4 h after (5) the intravenous injection of targeted NPs. Color bar scale in milliseconds.

Figure 14.2 Protein trans-splicing. The two protein fragments are expressed as fusion protein with a split intein partner. After intein reconstitution, the splicing reaction takes place and the two protein fragments are joined through a native peptide bond.

Figure 14.6 Molecular detail of site-specifically labeled MxeGyrA intein with a single ¹³C atom. Source: De Rosa, Lucia; Russomanno, Anna; Romanelli, Alessandra; D'Andrea, Luca D. 2013. Semi-synthesis of labeled proteins for spectroscopic applications. Molecules 18, no. 1: 440–465.

List of Contributors

Aaron E. Albers

Catalent Biologics

Emeryville

CA

USA

A. Sesilja Aranko

Research Program in Structural Biology and Biophysics

Institute of Biotechnology

University of Helsinki

Helsinki

Finland

Annette G. Beck-Sickinger

Faculty of Biosciences

Pharmacology and Psychology

Institute of Biochemistry

University Leipzig

Leipzig

Germany

Ghyslain Budin

School of Basic Sciences

Institute of Chemical Sciences and Engineering

Swiss Federal Institute of Technology of Lausanne

Lausanne

Switzerland

Xi Chen

Chemical Genomics Centre of the Max Planck Society

DortmundGermany

Max Planck Institute for Molecular Physiology

Dortmund

Germany

Luca Domenico D'Andrea

Istituto di Biostrutture e Bioimmagini

CNR

Napoli

Italy

Lucia De Rosa

Istituto di Biostrutture e Bioimmagini

CNR

Napoli

Italy

Marc Dittman

Department of Structural Biochemistry

Max Planck Institute for Molecular Physiology

Dortmund

Germany

Penelope M. Drake

Catalent Biologics

Emeryville

CA

USA

Elena A. Dubikovskaya

School of Basic Sciences

Institute of Chemical Sciences and Engineering

Swiss Federal Institute of Technology of Lausanne

Lausanne

Switzerland

Martin Engelhard

Department of Structural Biochemistry

Max Planck Institute for Molecular Physiology

Dortmund

Germany

Rudi Fasan

Department of Chemistry

University of Rochester

Rochester

NY

USA

John J. Ferrie

Department of Chemistry

University of Pennsylvania

Philadelphia

PA

USA

John R. Frost

Department of Chemistry

University of Rochester

Rochester

NY

USA

Aurélien Godinat

School of Basic Sciences

Institute of Chemical Sciences and Engineering

Swiss Federal Institute of Technology of Lausanne

Lausanne

Switzerland

Hironobu Hojo

Institute for Protein Research

Osaka University

Osaka

Japan

Hideo Iwaï

Research Program in Structural Biology and Biophysics

Institute of Biotechnology

University of Helsinki

Helsinki

Finland

Hacer Karatas

School of Basic Sciences

Institute of Chemical Sciences and Engineering

Swiss Federal Institute of Technology of Lausanne

Lausanne

Switzerland

Tianlu Li

Department of Chemistry

The University of Hong Kong

Hong Kong SAR

PR China

Xuechen Li

Department of Chemistry

The University of Hong Kong

Hong Kong SAR

PR China

Oleg Melnyk

Institut Pasteur de Lille

UMR CNRS 8161

Université de Lille

Lille

France

Nydia Panitz

Faculty of Biosciences

Pharmacology and Psychology

Institute of Biochemistry

University Leipzig

Leipzig

Germany

Richard J. Payne

School of Chemistry

The University of Sydney

Sydney

Australia

E. James Petersson

Department of Chemistry

University of Pennsylvania

Philadelphia

PA

USA

Bhavesh Premdjee

School of Chemistry

The University of Sydney

Sydney

Australia

Department of Protein and Peptide Chemistry

Novo Nordisk A/S

Måløv

Denmark

David Rabuka

Catalent Biologics

Emeryville

CA

USA

Alessandra Romanelli

Dipartimento di Farmacia

Università di Napoli Federico II

Napoli

Italy

Claire Simonneau

Institut Pasteur de Lille

UMR CNRS 8161

Université de Lille

Lille

France

Jérôme Vicogne

Institut Pasteur de Lille

UMR CNRS 8161

Université de Lille

Lille

France

Stephanie Voss

Chemical Genomics Centre of the Max Planck Society

DortmundGermany

Max-Planck Institute for Molecular Physiology

Dortmund

Germany

Christopher R. Walters

Department of Chemistry

University of Pennsylvania

Philadelphia

PA

USA

Yao-Wen Wu

Chemical Genomics Centre of the Max Planck Society

Dortmund

Germany

Max Planck Institute for Molecular Physiology

Dortmund

Germany

Preface

The ability to synthesize and modify a biological macromolecule has been at the basis of several achievements in chemistry and biology. The possibility to handle a natural molecule, such as a protein, is crucial to characterize its chemical and biological properties, verify theoretical hypotheses, and develop novel molecules with optimized properties. These activities mainly require a significant amount of homogeneous and highly pure material, which is hardly isolated by natural fonts. A significant advancement has been made with the advent of recombinant protein expression and synthetic methodologies, in particular, the solid-phase procedure developed by Merrifield. Since the Merrifield work, the peptide synthesis field grew exponentially, and many improvements were reported. The main achievement was made in 1994 by Kent and coworkers with the development of the chemical ligation strategy to build long polypeptide chains. With this weapon in their hands, scientists started to investigate sophisticated biological problems and to develop finest applications. At the beginning, chemical ligation reactions were devised to join polypeptide chains by an amide (native) or a different (unnatural) bond. As soon as scientists started to appreciate the beauty of these chemical reactions, they envisioned that, similarly, these reactions could be fruitfully used to decorate/immobilize biomolecules in site-specific manner. Nowadays, a wide array of chemoselective reactions are known, and the development of new ones is one of the frontiers of the bioorganic chemistry. The combination of ligation techniques to assemble a polypeptide chain harnessing unnatural amino acids with chemoselective reactions allows to create in the laboratory a tailor-made protein endowed, in principle, with any specific property.

This book provides scientists, even not synthetic chemists, with an up-to-date reference to the possibilities offered by chemical ligation reactions in building biological macromolecules and presents an overview of the so-far-reported applications, showing a plethora of different strategies/tools to approach biophysical and mechanicistic studies with proteins. At the same time, large space is devoted to the chemistry behind the specific reactions, so this book could also be a reference for students interested in modern protein chemistry.

The book starts with an introductive chapter on the chemistry of native chemical ligation reactions. This chapter is an overview on the major ligation strategies reported so far and represents the bases of the ligation strategies presented in the successive chapters. Chapters 2, 3, and 4 deal with the different synthetic strategies to prepare a polypeptide and report some successful examples. Chapters 5, 6, and 7 focus on chemical aspects and applications of chemical ligation to three classes of proteins (chemokine, glycoprotein, and membrane proteins). Chapters 8, 9, and 10 explore protein-labeling techniques and applications. Finally, Chapters 11–14 are thematic, highlighting how the use of molecules made through chemical ligation approaches contributed to the advancement of numerous research field including macrocycle synthesis, molecular imaging, and structural biology. All chapters have been written by scientist leaders in their field and active in the ligation realms.

Last but not least, we would like to thank all the authors who contributed to this work.

We hope that this work will contribute to the expansion of the audience of scientists working with chemical ligation reactions and for people already active in this field will represent a useful reference book to look at to find novel solutions or ideas.

Luca D. D'Andrea

Alessandra Romanelli

Napoli, Italy

July 2016

Chapter 1

Introduction to Chemical Ligation Reactions

Lucia De Rosa¹, Alessandra Romanelli² and Luca Domenico D'Andrea¹

¹Istituto di Biostrutture e Bioimmagini, CNR, Napoli, Italy

²Dipartimento di Farmacia, Università di Napoli Federico II, Napoli, Italy

1.1 Introduction

Unraveling the molecular basis that controls the protein structure, function, folding dynamics, and interactions outstands as one of the most ambitious challenges of the post-genomic era. A complete understanding of protein properties will definitively put these biological macromolecules at the service of scientists, teaching about how to use nature's rules to design new proteins endowed with specific features and geared for specific functions. The ability to synthesize a protein by chemical route and selectively modify its primary structure provides a powerful tool to fulfill such requests. Until now, the DNA recombinant techniques have succeeded in preparing a wide number of proteins and protein mutants, allowing investigation of their properties and behavior. However, protein engineering approaches suffer some restrictions, mainly the genetic code barrier, which only tolerates the introduction of the natural amino acids and has been only in part overcome by the introduction of engineered tRNA/aminoacyl-tRNA synthetase (aaRS) [1, 2], and the poor control over site-specific protein modifications. For these reasons, novel stratagems are needed to make protein synthesis and site-selective modification a realistic goal. Nowadays, chemical synthesis is unanimously recognized as a key strategy in protein preparation and modification, which ensures open access to any site of any protein sequence with unique level of specificity, allowing a limitless and surgically precise modification of protein covalent structure [3, 4].

1.1.1 Chemical Synthesis of Proteins: From the Stepwise Synthesis to the Chemical Ligation Approach

Since the first peptide syntheses performed in solution by Emil Fisher in the early years of the eighteenth century, chemical synthesis appeared as the most powerful approach for preparing peptides and proteins. An important milestone in the development of chemical strategies for full-protein synthesis is represented by the introduction of an original synthetic method that revolutionized the chemical synthesis of peptides and is still in use, the stepwise solid-phase peptide synthesis (SPPS) [5] (Figure 1.1).

Image described by caption and surrounding text.

Figure 1.1 Schematic representation of the solid-phase peptide synthesis.

Essentially, in SPPS, the peptide is built up on a solid support, and each amino acid is used with its side chain and α-amino group masked by removable protecting groups. Peptide synthesis is performed in the C- to N-terminus direction and starts by anchoring the C-terminal amino acid of the target peptide sequence to an insoluble polymeric support through the covalent binding of its α-carboxyl group to a linker moiety. Repeated cycles of deprotection of the α-amino group of the resin-bound amino acid and coupling with the carboxyl-activated form of the next amino acid allow the assembly on the solid support of the full peptide with all the side chains of the amino acid residues properly protected. After the assembly of the full peptide, all side-chain-protecting groups are removed, and simultaneously, the covalent link to the polymer support is cleaved to give the fully unprotected peptide product in solution. As in SPPS, the nascent peptide chain remains bound to the resin for the entire synthesis, after each reaction step of the synthesis, a facile purification by filtration and washing with organic solvents can be performed, enabling the use of large excess of reactants and limiting side reactions. Consequently, reactions are rapid and near-quantitative, handling losses and side-product formation are strongly limited with respect to the classical synthesis in solution phase. As a result, the desired peptide is obtained in high yield and quality. Many impressive protein syntheses have been performed using SPPS in a stepwise manner, such as ribonuclease A (124 amino acids) [6], HIV-1 aspartyl protease (99 amino acids per chain) [7–10], and bovine pancreatic trypsin inhibitor and its analogs (58 amino acids) [11–13]. The ability to assemble a full protein in a linear and straightforward way by SPPS is, however, limited by the synthetic efficiency of each reaction step. Although each reaction step is almost quantitative, inevitable accumulation of by-products limits the ultimate size of high-purity polypeptides that can be effectively prepared by this way to chains of ~50 amino acids in length. In order to get beyond this limit and access longer polypeptide chains, SPPS was combined with the logic of the convergent synthesis (Figure 1.2).

A process diagram with schematics for chemical preparation of a polypeptide by convergent synthesis in solution phase.

Figure 1.2 Chemical preparation of a polypeptide by convergent synthesis in solution phase.

Convergent synthesis consists of the sequential condensation of short protected peptide segments in organic solvent to obtain longer polypeptide chains. Human insulin protein (51 amino acids) and a series of analogs of insulin [14], the enzyme ribonuclease A (124 amino acids) [15], and a consensus lysozyme enzyme molecule (129 amino acids) [16] were some of the first full proteins chemically prepared by using convergent synthesis. Despite convergent synthesis allowed the preparation of many proteins, it faces laboriousness and arduousness. This strategy was hampered by low yield and technically demanding reaction steps for the preparation and purification of the protected segments and their condensation. The high level of skills required, the lack of chiral homogeneity in peptide bond formation, and the inability to manipulate fully protected peptides because of their low solubility led to the abandonment of the convergent synthesis in solution. The right answer came in the early 1990s, when a novel and revolutionary methodological advance, named chemical ligation, was proposed [17] (Figure 1.3).

A process diagram with schematics for the chemical ligation general concept.

Figure 1.3 The chemical ligation general concept. A single polypeptide chain is obtained by covalently joining two peptide segments through the reaction of two mutually reactive functional groups. The type of covalent bond generated at the junction site depends on the reactive groups employed for the ligation reaction.

The chemical ligation approach consists of preparation of a polypeptide chain by sequentially joining short and unprotected peptide segments. To this aim, chemical ligation exploits chemoselective reactions between two mutually reactive functional groups placed at C- and N-terminus of two contiguous peptide segments, respectively. The functional groups react by forming a stable chemical bond without reacting with any other reactive group present in the peptides, thus permitting the use of peptide segments in their fully unprotected form. The use of unprotected peptides ensures ligation reactions to be performed under mild conditions and usually in aqueous buffers, thus overcoming many issues of convergent synthesis. The ingenious synthetic stratagem of chemical ligation allowed chemical protein synthesis and modification to become a reality. Synthetic peptide segments may harbor a plethora of possible chemical modifications allowing site-selective introduction of unnatural elements into proteins. Chemically synthesized proteins found widespread application in biochemistry, biotechnology, biophysics, and chemical biology, allowing the study and the characterization of many proteins and leading to a great number of scientific discoveries. In pharmacology, chemical synthesis of protein therapeutics offers the critical benefit of sample homogeneity and purity [18, 19]. Examples of the first proteins and unnatural protein analogs prepared by chemical synthesis include HIV-1 protease [17], a tethered dimer of HIV-1 protease [20], the covalent heterodimer of b/HLH/Z transcription factor cMyc-Max [21], a four α-helix template-assembled synthetic protein (TASP) molecule [22], and a folded β-sandwich fibronectin domain model [23].

1.1.2 Chemical Modification of Proteins: From Conventional Methods to Chemoselective Labeling by Chemical Ligation

The introduction of unnatural amino acids, biophysical probes such as fluorophores or isotopes, post-translational modifications (PTMs), or even backbone rearrangements into a protein sequence is fundamental to solve protein science issues with a molecular precision. Many insights were gained by combining protein engineering with site-specific modifications performed by the use of chemical reagents that selectively tag amino acid side chains. Conventional methods for protein derivatization exploit the reactivity of native or engineered Lys or Cys side chains and use molecular probes that react selectively with amino or thiol groups [24, 25]. However, these approaches are limited because of the presence of more than one reactive amino acid in the protein sequence, resulting in multiple labeling and heterogeneous protein preparations [26]. Specific reactions for protein labeling with fluorescent dyes have also been performed by introducing at protein N- or C-terminus an extra sequence recognized by specific reactants such as FlAsH, CrAsH, or ReAsH (biarsenical derivatives of fluorescein and resorufin) [27–29]. Enzymatic approaches for protein labeling have also been described, such as the SNAP-tag, CLIP-tag, and Halo-tag technologies, which use enzymes that are able to recognize and label specific protein partner fused at the N- or C-terminus of the target [30–32]. These latter procedures are limited to N- or C-protein termini and generally do not allow incorporation of multiple, different molecular probes into a protein. Similarly, recombinant expression of post-translationally modified proteins is not trivial. Proteins modified with PTMs are traditionally prepared by enzymatic treatments, which are often impossible to control precisely and not always proceed with high yields. Such approaches suffer from poor versatility, being often tailored on a specific protein target, and moreover, the incorporation of multiple, different types of molecular probes into discrete sites or regions of a protein is not possible. Recombinant biosynthesis using PTM machineries of the expression hosts often results in heterogeneous mixtures, especially in the case of glycosylated proteins [33]. Sample heterogeneity represents a strong limit for many scientific applications, in particular in the case of protein therapeutics. A significant portion of the post-translationally modified protein therapeutics, in particular glycoproteins, administered to patients are biosynthesized as complex mixture, with major drawbacks such as the irreproducible bioactivity from one batch to another [34]. For these reasons, even if, overall, the aforementioned protein modification approaches are original and sophisticated and led to great discoveries and innovations, more general approaches are needed to make site-selective modification of proteins a realistic goal. The great potential of protein chemical synthesis appeared as a straightforward way to incorporate any type and number of nonnative elements into proteins.

1.2 Chemical Ligation Chemistries

Chemical ligation methodology provides an excellent platform for preparing proteins and protein conjugates in high yield and good purity, which is currently a major frontier of the synthetic bioorganic chemistry. An ideal ligation chemistry should be highly selective, should be compatible with all the functional groups present in proteins, and should preferentially proceed under mild aqueous conditions. Considerable efforts led to the development of a portfolio of chemoselective ligation chemistries, satisfying the desire of harnessing several reactions that can be performed orthogonally to one another for the one-pot rapid assembly of aprotein or a modified protein. Chemical ligation chemistries differ for the type of chemical bond introduced at the site of junction between the two reacting moieties, which in turn is correlated to the reactive groups employed for the ligation reaction. Although chemistries generating an amide bond are usually preferred, those leading to the formation of nonpeptide bond at the ligation site are also attractive approaches in synthetic protein research as far as the unnatural bond does not strongly affect the protein structure. Next section will dissect the major chemistries proposed as ligation strategies for protein synthesis, which, in their overall, ensure a complete flexibility in the manipulation of protein primary structure.

1.3 Imine Ligations

Imine ligations exploit the chemoselective reaction between amines and aldehydes or ketones to generate an imine-type bond. Imine ligations are among the firstly exploited ligation chemistries and have shown great utility in protein synthesis thanks to their high chemoselectivity in the presence of every natural amino acid and a variety of PTMs. Depending on the nature of the amine employed, imine ligations give rise to a different kind of imine bond at the ligation junction. For instance, the reaction of aldehydes or ketones with oxyamines generates an oxime bond, while the reaction of aldehydes or ketones with hydrazines leads to the formation of an hydrazone bond. The following subsections will describe in more detail the different types of imine ligations described in the literature.

1.3.1 Oxime Ligation

Oxime ligation refers to the highly chemoselective reaction between an aldehyde or a ketone and an aminooxy group to give an oxime bond (Scheme 1.1). Since the mid-1980s, the oxime ligation has found widespread use [35], not only for protein synthesis by peptide fragment assembly [36] but also for preparation of chemical microarrays [37], cyclic peptide libraries [38, 39], glycoprotein [40], protein–polymer conjugates [41, 42], peptide–oligonucleotide complexes [43], and for viral particle functionalization [44]. The oxime ligation is a particularly attractive chemistry since it is very efficient and chemoselective, taking place in aqueous solution under mild acidic conditions (usually pH around 5.0), and water is the only side product formed in this process. In acid medium, the first step of the oxime ligation leads to the formation of a carbinolamine from the nucleophilic addition of the oxyamine to the carbonyl compound (Scheme 1.1a). The second step is triggered by the protonation of the hydroxyl group, which causes the dehydration of the carbinolamine into the oxime (Scheme 1.1b) [45]. Oxime ligations are usually slow at physiological pH; maximum reaction rates are typically reached near the pKa of the nucleophile while dropping off sharply at higher or lower pH values [46].

Image described by caption and surrounding text.

Scheme 1.1 The oxime ligation reaction.

In general, in all types of imine ligations, and thus even in the oxime ligation, aldehydes are substantially more reactive than ketones, mainly because of steric effects, and aromatic carbonyls are more reactive than the aliphatic ones. Buré et al. compared the reactivity of C-terminal aldehyde and ketone peptides with aminooxy-containing peptides in acido-catalyzed oxime reactions. The results obtained confirmed that oxime ligation proceeds less smoothly in the case of peptide ketones compared to peptide aldehydes. The first step of the oximation reaction was the limiting step when ketone peptides are used, while the second step was the determining step in the case of peptide aldehydes [47]. The modest rate of oxime ligation reaction, especially observed when a ketone is used, can be improved by adding nucleophilic catalyst, such as aniline [48, 49], or the new-generation catalysts, such as meta- and para-phenylenediamine [50–52]. Catalysts are amines that form a reactive imine intermediate with the target carbonyl group. Use of catalysts allows oxime ligation to be performed faster even at neutral pH, as necessary for biomolecules not soluble or unstable under acidic pH and using lower reactant concentrations as needed for applications using cell extracts [53]. All the imine bonds are generally susceptible to hydrolysis. Among the imine-based ligation reactions, oxime ligation affords the most stable imine product. Despite the oxime bond being identified as the less hydrolysis-prone imine linkage, it remains thermodynamically unstable and undergoes hydrolysis at an appreciable extent in aqueous solution [54]. Although this is generally inconvenient when oxime ligation is used to synthesize proteins or protein bioconjugates, this feature may become a straightforward advantage in applications where covalent capture and controlled release are needed [35, 55]. For example, oxime chemistry was used to efficiently conjugate different peptide species to hyaluronic acid polymer. Peptides synthesized with an aminooxy N-terminus were reacted under slightly acidic aqueous conditions and without a catalyst with hyaluronic acid functionalized with aldehyde groups. The resulting oxime bond was found to rapidly hydrolyze at pH 2, releasing the bound peptide molecules, but was stable at higher pH values. The strategy was applied to two different biologically active peptide species, a multiple sclerosis antigen and an ICAM-1 ligand, known to block immune cell stimulation [56]. Similarly, some quinone oxime molecules have been used as redox cleavable linkers [55].

Both functionalities required to perform oxime ligation, aldehydes and oxyamines, can be straightforwardly introduced into peptides and proteins. In the case of aldehydes, several tools are