Catcher/Tag Toolbox: Biomolecular Click‐Reactions For Protein Engineering Beyond Genetics

Manipulating protein architectures beyond genetic control has attracted widespread attention. Catcher/Tag systems enable highly specific conjugation of proteins in vivo and in vitro via an isopeptide‐bond. They provide efficient, robust, and irreversible strategies for protein conjugation and are simple yet powerful tools for a variety of applications in enzyme industry, vaccines, biomaterials, and cellular applications. Here we summarize recent development of the Catcher/Tag toolbox with a particular emphasis on the design of Catcher/Tag pairs targeted for specific applications. We cover the current limitations of the Catcher/Tag systems and discuss the pH sensitivity of the reactions. Finally, we conclude some of the future directions in the development of this versatile protein conjugation method and envision that improved control over inducing the ligation reaction will further broaden the range of applications.


Introduction
Biological networks are inherently modular, enabling them to be dissected into almost independently functioning units. [1]odularity offers an opportunity to construct highly complex systems by manipulating the assembly of biological units. [2]roteins and peptides are versatile units that are widely utilized in biological networks, due to their numerous biological functions.Owing to their potential for applications in protein and peptide therapeutics, biomaterials, and more, there has been a great interest in technologies that enable manipulating protein and peptide architectures. [3,4]In addition to manipulation at the genetic level, strategies have been developed that enable modifications at protein level.
An ideal conjugation method would be an efficient yet highly selective reaction that created a specific amide bond under a wide range of conditions. [5]Chemical methods of protein ligation are the most widely applied methods for protein ligation.In native chemical ligation (NCL) and its semisynthetic counterpart, expressed protein ligation (EPL), the thiol group of an N-terminal cysteine attacks a C-terminal thioester, forming a peptide bond.The applications are, however, limited by inherently restricted biocompatibility and specificity, and by the requirement for high concentrations of reactants. [6,7]Recent studies have demonstrated overcoming the concentration and specificity constraints by highly sophisticated engineering of peptide substrates [8,9] thus leading the way to overcome these limitations in in vitro ligation applications.
In addition to the chemical ligation methods, several strategies based on reactions catalyzed by split protein domains and enzymes have been reported.Split inteins enable scarless conjugation of the flanking proteins at high yields both in vivo and in vitro. [10]12][13] Sortases, on the other hand, catalyze specific peptide bond formation between two protein targets with a C-terminal motif, which is typically LPXTG, and an N-terminal glycine.While this method is limited by low efficiency and dependency on calcium ions, the recognition sequence remaining as a scar is relatively short. [6,14,15]Transglutaminases are another type of protein conjugation enzymes, which catalyse isopeptide bond formation between lysine and glutamine side chains.They are biocompatible and widely used, although the non-specificity of the reactions hinders their applications. [16,17]Finally, asparaginyl endopeptidases (AEPs) and peptide asparaginyl ligases (PALs) catalyze peptide bond formation between short and varying Cand N-terminal recognition sequences, [14] in which the only invariable residue is Asx in the N-terminus of the protein to be ligated.Ligation methods relying on (AEPs) and (PALs) have rapidly evolved over the past years, [18] most noteworthy with several strategies to control the catalytic activity and to overcome the limitations in efficiency being reported [19][20][21] The Catcher/Tag system enables rapid, efficient, and specific protein conjugation in vivo and in vitro.The conjugation is mediated by a covalent isopeptide bond that forms autocatalytically between Catcher and Tag, thus covalently connecting also the fusion proteins.The Catcher/Tag pairs developed create a unique protein conjugation toolbox that appears to overcome many of the challenges of other ligation techniques.Catcher/Tag pairs function in a variety of mild reaction conditions within a wide concentration range, catalyse spontaneous and efficient ligation, are robust to various proteins of interest (POIs) as fusion proteins, as well as have no side reactions. [1,22,23]In this review, we summarize the existing Catcher/Tag systems and their key applications.Finally, we focus on the current limitations and future prospects of this inspiring method.

Design of Catcher/Tag systems
The Catcher/Tag conjugation strategy is based on two insightful findings.[26][27][28][29] Kang, et al. [24] solved the crystal structure of the major pilin protein Spy0128 from Streptococcus pyogenes (S. pyogenes) and discovered an intramolecular isopeptide bond connecting the side chains of Lys179 and Asn303.(Figure 1A).Later Hagan, et al. [30] analyzed the structure of the much smaller CnaB domain of the fibronectin-binding protein FbaB from S. pyogenes and found the isopeptide bond between Lys31 and Asp117 side chains (Figure 1B).These isopeptide bonds form autocatalytically upon the correct folding of the protein domain containing the bond, the reaction being catalyzed by a highly conserved Glu. [30]The second key finding was the realization that a CnaB domain can be split into two fragments, first shown by Zakeri, et al. [31] who dissected Spy0128 into two fragments at the loop between the two β-sheets closest to the C-terminus (Figure 1A).The larger fragment of this first generation isopeptide-protein assembly system is called Catcher and the smaller Tag.To develop a more efficient protein-peptide pair, Zakeri, et al. [23] split the smaller CnaB2 domain of FbaB from S. pyogenes which, accompanied by rational optimization, resulted in the second generation isopeptide-protein assembly system, SpyCatcher/SpyTag (Figure 1B).The two fragments, Catcher and Tag, can associate together into a correctly folded CnaB domain via protein fragment complementation. [32]Impor-Ruxia Fan has obtained an M.Sc. in chemical engineering and technology from Shanghai Jiao Tong University and is now a Ph.D. student at Aalto University.Her main research focus is the self-assembly of spider silk proteins, the construction of high-performance artificially produced spider silk and posttranslational modification of silk protein.
Sesilja Aranko is Senior Scientist and group leader at Aalto University, Finland.She did her Ph.D. at the Institute of Biotechnology, University of Helsinki, Finland, in 2011-2014, focusing on protein splicing.After finishing her Ph.D., she gained experience in biomaterials and X-ray crystallography at Aalto University and EMBL Hamburg.Currently, Dr. Aranko and her research group develop methods for protein modifications and functionalization.Lys-Asp isopeptide bond in CnaB2 domain from S. pyogenes (based on PDB ID: 2X5P and 4MLI). [22,30]The residues forming the isopeptide bond (Lys and Asn/Asp tantly, the association and folding of the two fragments result in the autocatalytic formation of an isopeptide bond connecting the two halves, without the need for any external enzymes or cofactors.Fusion of the Catcher and Tag with two POIs will lead to covalent conjugation of the POIs, which is the general principle of using Catcher/Tag pairs (Figure 1C).
Inspired by the SpyCatcher/Tag system, a number of protein-peptide pairs capable of spontaneously forming isopeptide bonds have been constructed by either dissecting different isopeptide-bond containing domains, by shifting the split site, or by creating variants of previously reported Catcher/Tag pairs, for example, by directed evolution (Table 1, 2).
To expand the Catcher/Tag toolbox, several isopeptide bonds containing domains were identified by structural similarity searches.All of them are CnaB domains of outermembrane proteins from Gram-positive bacteria. [33]The CnaB domain is composed of typically seven β-strands, of which the first strand and the last strand are linked by an isopeptide bond. [34,35]Splitting of nine different CnaB domains to create Catcher/Tag pairs has been reported so far (Table 1).The Catcher/Tag pairs can be divided into three classes according to the location of the split site (Figure 1D-F, Table 1).The most common choice for a split site is between the two β-strands closest to the C-terminus, called here Type C (Figure 1D).This class includes the most widely used SpyCatcher/Tag pair. [23]lthough systematic study on choosing the split site is lacking, Type C split site appears to be promiscuous for splitting.Indeed, our results on testing different split variants of the Lactiplantibacillus plantarum (L.plantarum) CnaB1 domain, including Silk-Catcher/Tag pair, demonstrated that dissecting the protein chain at this site resulted in the most efficient protein conjugation reaction (Table 1). [36]Another commonly reported choice for a split site is to cut between the first two β-strands starting from the N-terminus, called here Type N split site (Figure 1E, Table 1).SnoopCatcher/Tag and DogCatcher/Tag pairs are examples of Catcher/Tag pairs created from the same CnaB domain by splitting at the Type N and Type C split sites, respectively. [37,38][41] Importantly, the ligase will not be present in the ligation product, and this method can thus greatly reduce the scar remaining in the final conjugation product.Finally, although a number of Catcher/tag pairs engineered from different split CnaB domains have been reported, not all isopeptide-bond containing domains will remain functional after splitting.For instance, Pröschel, et al. [34] probed the potential for design of Catcher/Tag pairs of four similar CnaB domains, only two of them can be split into active Catcher/Tag pairs.In addition to engineering Catcher/Tag variants from different CnaB domains, a series of variants based on the widely used SpyCatcher/Tag pair has been constructed through directed evolution, [44,45] site-directed mutation, [46,47] minimization, [22] and insertion of functional tags [48] with the aim being to create pairs suitable for specific applications (Table 2).Two accelerated variants, SpyCatcher/Tag (002) and SpyCatcher/Tag (003), were developed by a combination of phage-display platform screening and rational site-directed mutation.The reaction rate of SpyCatcher/Tag (003) is close to the diffusion limit, making it an attractive choice for most standard protein-conjugation applications. [45]The increasingly varied Catcher/Tag toolbox offers a wide range of options targeted for different applications.For example, by inserting a protease recognition site, SpyCatcher-N TEV can be cleaved by the protease after the reaction is completed allowing the scar to be minimized [48] whereas the super negatively charged SpyCatcher has a controllable reaction in response to pH, temperature, and ionic strength. [46]

Mechanism of isopeptide-bond formation
The mechanism of isopeptide-bond formation has been elucidated based on structural data obtained by X-ray crystallography [24] and nuclear magnetic resonance spectroscopy (NMR), [30] along with quantum mechanical/molecular mechanical (QM/MM) calculations. [27]Based on these works, the triad of reactive residues locates in the hydrophobic protein interior.The relative positions of the Lys, Asn/Asp, and Glu residues are strictly conserved (Figure 2).
These key residues are surrounded by a cluster of aromatic residues.The Glu residue is protonated by a hydrogen bond formed with the carbonyl group of a neighboring Asp or Asn residue.The protonated Glu polarizes the carbonyl bond, inducing a positive charge of Cγ.The ɛ-amino group of unprotonated Lys undergoes nucleophilic attack on Cγ, forming a zwitterionic intermediate (Step 1).Glu then acts as a proton shuttle for donating or receiving a proton to facilitate a reaction to form a neutral tetrahedral intermediate (Step 2).Finally, with  the release of ammonia or water, the isopeptide bond is formed (Step 3). [27,30,51]

Applications of the Catcher/Tag Systems -Expanding the Range of Protein Architectures
Given the spontaneous, efficient, and stable nature of the conjugation reaction and its robustness towards the reaction conditions, the Catcher/Tag system has become an ideal choice for a variety of applications.In this review, we summarize some of the key concepts of the applications of the Catcher/Tag system.A comprehensive list of publications and patents utilizing SpyCatcher/Tag and related technologies is collected on the SpyInfo webpage (available online at http://www.bioch.ox.ac.uk/howarth/info.htm),while the SpyBank database lists the corresponding sequence and expression pathways.

Cyclization of enzymes for thermal resilience
Superlative catalytic efficiency, mild reaction conditions, unique stereochemistry selectivity, and environmental compatibility have enabled enzymes to be extensively used in industry. [52]owever, the poor thermal resilience of enzymes has limited the de novo design of enzymes as well as fully exploiting the impact of enzymes beyond the laboratory conditions. [53,54]Some conventional strategies, such as directed evolution [55] structure-or library-based optimization, [56] have been employed to overcome these challenges.Despite the remarkable success of these enzyme modification approaches, given the rich structural and functional diversity of enzymes, a generic optimization strategy suitable for a range of different enzymes would be more promising than case-by-case solutions.Protein cyclization technology has been developed as a powerful method to enhance the thermal resilience of enzymes. [61]yclization of proteins by covalently linking the N-and Ctermini of the protein backbone leads to increased rigidity and a higher entropy barrier for protein unfolding. [62]Various approaches have been explored for protein cyclization, such as chemical synthesis, [63] sortase-mediated cyclization, [14] and split inteins. [64,65] protein cyclization strategy based on the SpyCatcher/Tag system, called SpyRing, [54] was reported to significantly improve the thermal resilience of various enzymes, such as βlactamase, [66] phytase, [67] luciferase [68] and xylanase. [69]Fusing the N-and C-termini of the enzyme of interest with SpyCatcher and SpyTag, respectively, resulted in a spontaneous cyclization of the enzyme (Figure 3A).The cyclization can not only improve the thermodynamic stability of enzymes, but some cyclized enzymes will also be endowed with improved tolerance to extreme pH, [68] and higher binding efficiency between enzymes and substrates. [70,71]In addition, ultra-high heat resistance can also simplify the enzyme isolation process. [66,67]Protein cyclization strategies based on other Catcher/Tag pairs, such as SnoopRing and PilinRing, [67] have also been reported to be effective in improving the thermostability of enzymes.In summary, the isopeptide bond-based protein cyclization strategy has the advantages of high efficiency, irreversibility, and a mild but wide range of reaction conditions, including reducing and oxidizing conditions, making it a competitive enzyme modification method for the enzyme industry.

Immobilization
An ideal vaccine platform for pandemic situations combines rapid production, cost-effectiveness, scalability, and widespread accessibility. [72]Modular vaccine construction enables rapid customization and combination of various antigen and adjuvant modules, streamlining the vaccine development process. [73]irus-like particles (VLPs) are self-assembling structures that mimic the shape and surface features of viruses but lack genetic material, making them non-infectious and safe. [72]Moreover, VLPs enhance the immune response by promoting antigen presentation and inducing strong antibody and cellular immune responses, making them a promising vaccine platform. [74]Classic approaches to create chimeric VLPs are either by genetic-level engineering of fusion proteins or by chemical conjugation of the nucleophilic amino-acid side chains on the VLPs and the antigen. [72]Although these methods are biocompatible and simple, the coupling of antigens to VLPs is time-consuming and frequently non-uniform thus resulting in sub-optimal immunogenicity. [57,72]LPs modularly decorated by SpyCatcher/Tag pairs to construct vaccines represent a promising approach to enhance vaccine development and customization.Brune, et al. [57] created a Plug-and-Display device based on SpyCatcher/Tag pairs to develop a modular vaccine platform where different antigens can be rapidly and precisely attached to VLPs (Figure 3B).In this approach, antigens from various pathogens or tumor cells are engineered to contain the SpyTag sequence, while the VLPs are modified to carry the complementary SpyCatcher sequence.[77] The Plug-and-Display highlights the separate production and modular assembly of antigens and scaffolds, which streamlines vaccine assembly, enabling the generation of multivalent vaccines that can target multiple pathogens or cancer antigens. [78]The method enables the incorporation of different antigens without the need for complex genetic manipulation or protein purification. [74,79]This approach has significant potential in addressing emerging infectious diseases, developing personalized cancer vaccines, and facilitating rapid responses to new threats. [57,72,75]ther immobilization applications in which Catcher/Tag systems have been demonstrated to be useful include immobilization of antibodies for diagnostics and developing enzyme cascades.[82] Finally, immobilization of enzymes on Escherichia coli (E.coli) biofilm curli by taking advantage of Catcher/Tag pairs was demonstrated to be an effective strategy to create cascades of enzymes.The specific immobilization strategies were demonstrated to enable one-pot conversion of chitin to glucosamine [83] and starch to trehalose. [84]

Overcoming the size-limitation in recombinant production
Protein-based biomaterials have been used in our daily lives for thousands of years.They are biocompatible and biodegradable, yet carrying excellent mechanical properties.Biomaterials with extremely competitive mechanical properties, such as spider silk, have received profound attention from researchers due to their potential to replace chemically synthesized materials.However, it has proven to be challenging to perfectly replicate the excellent properties of natural biomaterials.High-performance protein-based natural materials typically obtain their remarkable mechanical strength from hierarchical assemblies of proteins with ultra-high molecular weight and extremely repetitive amino-acid sequences. [85]These complicated sequence features greatly limit their recombinant production.First, it is impossible to directly synthesize long and highly repetitive genes encoding the high-molecular weight and highly repetitive proteins.Second, these complex genes are often unstable in heterologous hosts and are prone to undergo gene recombination resulting in no protein expression or truncated protein expression.Third, long and repetitive complex genes usually impose a huge translation burden on the host, resulting in low protein yields that are not enough to support material research.
Polymerization of low-molecular-weight protein sequences using protein conjugation tools is an effective method for the synthesis of high molecular weight and highly repetitive proteins.Catcher/Tag conjugation systems have fast kinetics in a broad range of solution conditions.Various Catcher/Tag pairs with orthogonal activity can be used in combination to achieve precise control of protein polymerization, providing diverse options for the construction of ultra-high molecular weight proteins.We successfully achieved the construction of nativesized spider-silk proteins by fusing two orthogonal Catcher/Tag pairs at the N-and C-termini of smaller spider-silk repeat sequence fragments (Figure 3C). [36]Affibody polymerization using SpyLigase/SpyTag/KTag has been reported as an innovative way to enhance the capture of cancerous cells with high specificity. [39]In addition to linear molecules, also nonlinear  [54] (B) Plug-and-display decoration for modular vaccines. [57](C) Constructting high-molecular-weight protein polymers by orthogonal Catcher/Tag pairs. [36](D) Entirely protein-based hydrogel based on Spy network. [58](E) Fluorescent protein labeling for cell imaging. [59](F) Efficient protein purification using Spy & Go. [60]igh-molecular-weight proteins with diverse topology were constructed by SpyCatcher/Tag pairs. [86]

Tunable networks for biomaterials
Protein-based hydrogels are promising alternatives to hydrogels made from synthetic polymers owing to their biocompatibility, biodegradability, and the potential to incorporate bioactive motifs or functional groups. [87]Protein-based hydrogels are widely used in various biomedical applications, such as tissue engineering and regeneration medicine.These hydrogels are usually constructed by physical cross-linking, chemical crosslinking, or a combination of the two.Physical crosslinking relies on non-covalent interactions, such as hydrogen bonds, and electrostatic interactions, to form a hydrogel network.On the other hand, chemical crosslinking involves the formation of covalent bonds between protein molecules, resulting in a more stable and durable structure.Due to the spontaneous formation of robust covalent isopeptide bonds without introducing other auxiliary components or chemical modifications, the Catcher/ Tag system has become an attractive choice to construct chemically crosslinked protein-based hydrogels.Sun, et al. [58] first constructed an entirely protein-based hydrogel through tailored molecular networks, "networks of spies".In their design, SpyCatcher and SpyTag were fused with Elastin-like-proteins (ELPs) followed by covalently crosslinking the block copolymers into hydrogel by spontaneous isopeptide bonds (Figure 3D).Compared to the physically crosslinked hydrogels, the protein hydrogels constructed by chemical crosslinking are robust and durable.In addition, various functional groups can be readily integrated into hydrogel networks by genetic engineering to produce multifunctional protein hydrogels according to specific applications.Examples of this include fusing two metalloproteins into Spy network to create hydrogel for selective heavymetal sequestration [88] as well as constructing smart stimuliresponsive protein hydrogels by introducing either switchable oligomeric protein blocks, [89,90] or metal/ligand coordinated interactions [91] into the network.A current limitation of the Spy network hydrogels is their weak mechanical properties owing to the low crosslink density.A potential strategy to overcome this is to introduce other cross-linked networks into Spy network to construct double-network hydrogels, such as chemical-physical double-networks, [92] covalent doublenetworks, [93] or coordination-covalent double-networks. [88,94]

Cell labeling and imaging
Fluorescence microimaging allows the study of biological phenomena, such as receptor proteins in their native biological environments. [95]Highly specific and efficient labeling of POIs is a fundamental prerequisite for the microscopic visualization of subcellular protein structures and interactions. [96]Catcher/Tag system is a promising tool for efficient and specific labeling of cells.The general method for specifically labeling cells by Catcher/Tag pairs is to fuse Catcher and Tag with fluorescent protein and cellular protein, respectively.The reaction between Catcher and Tag will result in labeling the cells rapidly and specifically (Figure 3E).Importantly, the reaction will lead into covalent conjugation of the fluorescent protein with the cellular target protein.In addition, the small size of Tag (typically 10-20 aa), reduces the likelihood that the functionality of the cellular target protein it is fused to is disturbed. [96]Zakeri, et al. [23] reported the first successful example of cellular specific labeling using SpyCatcher/Tag pair.Fusing SpyCatcher with the protein intimin on the bacterial outer membrane enabled realtime monitoring and analysis of the dynamic process of cell division. [44]Moreover, Bedbrook, et al. [59] used the SpyCatcher/ Tag system to label proteins in live Caenorhabditis elegans to track interactions between cells on the extracellular matrix in response to physiological changes in live animals.

Spy & GO purification
Although most applications of the Catcher/Tag system are based on conjugation via the covalent isopeptide bond, also non-reactive variants have been engineered and applied for protein purification.Stepwise engineering of SpyCatcher resulted in a non-reactive Catcher variant "pseudo-Catcher", named SpyDock. [60]It was designed with three goals: to block the formation of covalent bonds with SpyTag, to enhance the affinity of SpyDock binding to SpyTag, and to immobilize the SpyDock on the resin.The Spy & GO purification platform based on SpyDock was reported to be used for protein purification with higher purity (98.9 � 0.5 %) of elution product than normal His-Tag mediated Ni-NTA purification (66.4 � 1.9 %).However, due to the strong binding affinity between SpyDock and SpyTags, stringent conditions (2.5 M imidazole) were required for the elution step, which may result in disruption or even denaturation of the target protein.The second-generation tool, known as SpySwitch, [97] was optimized to offer enhanced capability for protein purification.Numerous histidine mutations were rationally introduced at the binding interface of SpyTag/SpyCatcher through a phage display library selection.The obtained SpySwitch allows the purification of proteins with high purity (> 95 %) under mild conditions (e. g., weak acidic buffer with low ionic concentration), thus preserving the activity of the protein of interest.Furthermore, SpySwitch is thermally responsive, which allows elution also by switching temperature from 4 °C to 37 °C, thus further expanding the range of potential applications for the method.Both SpyDock and SpySwitch resin can be regenerated multiple times and stored in 20 % ethanol for prolonged periods, comparable to the commercial Ni-NTA resin, making the Spy & GO platform a promising new method for protein purification.

Overcoming Current Limitations and Unlocking Future Possibilities
The Catcher/Tag system is a powerful tool for protein engineering and bioconjugation and has been demonstrated to be applicable to numerous applications.In the following sections, we cover some of the remaining limitations hindering the use of the method, as well as discuss possible solutions and future outlooks.

Large scar after conjugation
In contrast to the split intein-mediated ligation and enzymemediated ligation reactions, the Catcher/Tag system is not a traceless ligation tool. [23]The ligation with the Catcher/Tag pair will typically result in a scar of more than 10 kDa in the final ligation product. [48]The large scar limits, for example, the application of Catcher/Tag systems in the engineering of protein-based hydrogels.Catcher/Tag pairs, occupy a large proportion in the protein copolymers and thus result in low cross-linking density and poor mechanical performance of the final hydrogel. [87]To minimize the impact of this scar, three strategies have been applied to reduce the molecular weight of the Catcher/Tag pairs.One approach is to split a CnaB domain into three fragments, as in the case of Type C/N (Table 1, Figure 1F).[41] An alternative strategy is to introduce enzyme cleavage sites within the Catcher.The variant retains the high reactivity of the intact Catcher but can be cleaved off by protease upon completion of the reaction with Tag. [48,49]Finally, moderate effect on the scar size can be obtained by exploring mini-Catcher/Tag pairs.For instance, the SilkCatcher/Tag pair is 20 % smaller in size than the widely used SpyCatcher/Tag pair (99 aa versus 121 aa). [36]n practice, it is preferable to select Catcher/Tag pairs with appropriate size or to combine Catcher/Tag system with additional protein ligation methods.

Limitations of reaction conditions and rate
Most Catcher/Tag pairs can react both in vitro and in vivo under a very broad and mild range of reaction conditions.However, some extreme conditions, such as extreme pH, high temperature, or denaturing agents, might inhibit or even block the isopeptide-bond formation.It is a prerequisite that the Catcher and Tag fold correctly and adopt the native structure for the effective formation of the covalent bond.The extreme conditions will perturb the protein folding and thus the relative positioning of the reactive residues and/or disrupt the microenvironment surrounding them.The available Catcher/Tag pairs cover the following ranges of reaction conditions: 4-45 °C, (even 100 °C for SpyStapler [42] ), pH 3.0-9.0,and are tolerant towards most common non-ionic detergents, such as 1 % Triton X-100, 1 % Tween20, and 0.5 % Nonidet P-40, [23] and protein denaturant, urea. [41,44]Despite the use of Catcher/Tag system is limited in some extreme conditions, the current toolbox of Catcher/Tag pairs enables selection of an enzyme suitable for a wide range of reactions conditions.Future studies are expected to broaden the spectra of possible reaction conditions further.
Most of the reported "native" Catcher/Tag pairs have a limited reaction rate and yield (Table 1).In an ideal reaction the covalent interaction occurs at the diffusion limit. [45]Although most Catcher/Tag pairs react in minutes with high yield in micromolar protein concentrations, the reaction rates are several orders of magnitude lower than the diffusion limit. [5]The in vivo reactions of Catcher/Tag pairs are even slower, due to the nanomolar concentrations. [51]The design of SpyCatcher/ Tag 003 pair that has reaction rate close to the diffusion limit [45] offers guidance to overcome this limitation by constructing ultra-efficient Catcher/Tag pairs.

Potential immunogenicity
All Catcher/Tag pairs described to date are derived from the surface proteins of Gram-positive bacteria, which are nonhuman and may introduce immunogenicity concerns.Although the Catcher/Tag system offers a promising opportunity to develop a multifunctional platform for vaccines, [57,98] immunogenicity will be a challenge to their application in therapeutics, especially in applications involving live organisms.A study of a modular vaccine using VLPs showed that SpyCatcher-VLPs elicited an immune response, but the immune response was masked after decorating the target antigen. [57]While the report is promising for the biomedical applications of the Catcher/Tag system, the full impact of its potential immune response remains to be resolved.

Towards controlled reactivity
The isopeptide-bond formation occurs when the Catcher and Tag encounter each other.This uncontrollability limits their applications.Numerous efforts have been made for developing controllable Catcher/Tag ligation.Currently, there are two main strategies, based on pH and light, employed to control the reaction.
The reported Catcher/Tag pairs can be divided into two categories according to the isopeptide-forming residues, the two possible combinations being Lys-Asp or Lys-Asn (Table 1).These two different isopeptide bonds have different optimal reaction pH.For Lys-Asp the optimum is between pH 5-6, while for Lys-Asn it is above pH 7. At a non-optimal pH, the ligation rate and yield approach zero. [36,37]Comparison of the pH dependency of SilkCatcher/Tag pair either with the native Asp as isopeptide forming residue or that of a variant with the reactive Asp mutated to Asn demonstrate that the optimal pH for Lys-Asp is pH 5.0 and the optimal pH for Lys-Asn is over pH 7.0 thus supporting the relationship between pH and reactive residues (Figure 4A).Given the pH-dependent reaction, a pH-activated reaction between Catcher and Tag can be established.For instance, the SilkCatcher/Tag pairs have Lys-Asp as reactive residues, efficiently forming isopeptide bonds at pH 5.0, whereas exhibiting minimal reactivity at pH 9.0 (Figure 4A, B).Mixing SilkCatcher and SilkTag at pH 9.0 resulted in almost no reaction, while the reaction was activated by lowering pH from 9.0 to 5.0.This provides a successful and powerful example of the development of pH-activated Catcher/ Tag pairs.Moreover, the possibility to control the reactions by adjusting pH opens up the possibility to develop orthogonal Catcher/ Tag pairs, such as Snoop and Spy Catcher/Tag pairs. [37]rthogonal Catcher/Tag pairs are, however, not limited to those with different pH optimal for the reaction.Indeed, also Catcher/ Tag pairs with the same reactive residues and pH range that are derived from different CnaB domains may be orthogonal, such as Silk and Spy Catcher/Tag pairs. [36]The availability of orthogonal Catcher/Tag systems has established a new pathway for multiplex control of protein assembly and construction of novel protein architectures.The correlation between optimal pH and reactive residues offers guidance in selecting the appropriate Catcher/Tag pairs for different applications.
Another "switch" used for inducing the reaction of Catcher/ Tag pairs is light.Hartzell et al. [99] created a blue-light inducible SpyTag system (BLISS) which enables the controlled formation of isopeptide bonds in a spatiotemporal manner.This system involves the fusion of the light-oxygen-voltage domain 2 of Avena sativa (AsLOV2) with SpyTag.The C-terminal Jα-helix of AsLOV2 unfolds in response to blue light.In the dark state, SpyTag is hidden within the folded Jα helix, preventing it from reacting with SpyCatcher.However, upon unfolding triggered by blue light induction, SpyTag becomes exposed and capable to react with SpyCatcher (Figure 5A).Ruskowitz et al. [100] developed another light-activated method, named SpyLigation, by introducing a photocaged Lys into the active site of SpyCatcher during protein translation, facilitated by an unnatural tRNA/tRNA synthetase pair.An oNB moiety was installed on the ɛ-amine of lysine to prevent the reaction.After light exposure, the native active site Lys residue is exposed and autocatalytically reacts to form the isopeptide bond with SpyTag (Figure 5B).In another approach, the catalytic residue of Catcher/Tag pairs, Glu, was photocaged by o-nitrobenzyl and nitropiperonyl groups.The Catcher/Tag pairs with a photoc-aged Glu allows a photoactivated ligation by UV irradiation without observable leaky activity. [101]n general, the switch is regulated by the control of protein conformation and the accessibility of reactive residues.In addition to the methods described above, controllable reactivity can also be achieved by mutations, [60,97,102] deletion, [22] further splitting, [39] and other conformational restrictions, such as redox-responsive conformational constraints. [103]

Summary and Outlook
The Catcher/Tag toolbox has now emerged as a simple yet powerful route for covalent protein conjugation in protein research.The astonishingly robust and efficient isopeptidebond mediated conjugation using Catcher/Tag pairs applies to applications spanning numerous fields, including the enzyme industry, surface immobilization, biomaterials, and cellular imaging.Although most of the applications are proof-ofconcept, they are expected to mature as more data and user experiences are accumulating.
Recently, the discovery of new domains containing Catcher/ Tag pairs and the optimization of existing Catcher/Tag pairs have significantly expanded the capabilities of the Catcher/Tag toolbox, especially in the development of orthogonal Catcher/ Tag pairs and the reporting of Catcher/Tag pairs designed for specific applications.The current limitations of the Catcher/Tag toolbox include large post-conjugation scars, uncontrollable initiation of the reactions, limited compatibility, and potential immunogenicity.These hamper their use in certain applications, including those in which the Catcher/Tag pair remaining in the conjugation junction disturbs the functions of the end-product.By addressing current limitations and embracing future perspectives, the Catcher/Tag toolbox is poised to make even more significant contributions both to biological research and to commercial applications.We envision that a potential strategy to expand to new application fields is the development of Catcher/Tag pairs suitable for pH-or light-inducible conjugation that will enable improved control over the reaction.) reacts optimally at around pH 5. [36] Repeating the experiment with a mutant (Lys10 + Asp93Asn) shows a shifting of the optimal pH towards > pH 7. Adapted from Ref. [36], Copyright (2023), with permission from Wiley-VCH.(B) Schematic representation of pH-dependent reaction of SilkCatcher/Tag.unfolding to activate Catcher-Tag reaction. [99](B) Photocaged Lysine activated by light exposure is used for controlled SpyLigation. [100]

Figure 1 .
Figure 1.(A) Intramolecular isopeptide bond, Lys-Asn isopeptide bond in Spy0128 from S. pyogenes (based on PDB ID: 3B2 M);[31] Lys-Asp isopeptide bond in CnaB2 domain from S. pyogenes (based on PDB ID: 2X5P and 4MLI).[22,30]The residues forming the isopeptide bond (Lys and Asn/Asp) and the conserved Glu catalyzing the reaction are shown in sticks in purple and green.(B) Cartoon of Splitting CnaB2 domain into SpyCatcher (pink) and SpyTag (blue), reactive residues are highlighted by green.(C) General scheme of Catcher/Tag application.(D, E, and F) Three different splitting methods to construct Catcher/Tag pairs.
Figure 1.(A) Intramolecular isopeptide bond, Lys-Asn isopeptide bond in Spy0128 from S. pyogenes (based on PDB ID: 3B2 M);[31] Lys-Asp isopeptide bond in CnaB2 domain from S. pyogenes (based on PDB ID: 2X5P and 4MLI).[22,30]The residues forming the isopeptide bond (Lys and Asn/Asp) and the conserved Glu catalyzing the reaction are shown in sticks in purple and green.(B) Cartoon of Splitting CnaB2 domain into SpyCatcher (pink) and SpyTag (blue), reactive residues are highlighted by green.(C) General scheme of Catcher/Tag application.(D, E, and F) Three different splitting methods to construct Catcher/Tag pairs.

Table 1 .
Summary of the reported Catcher/Tag pairs.

Table 2 .
Reported variants of the SpyCatcher/Tag pair.