Transglutaminase catalyses the formation of an isopeptide bond between a lysine (K) residue and the γ-carboxamide group of a glutamine (Q) residue. The microbial (mTGase) rather than the eukaryotic enzyme tends to be used in conjugation applications. Given the plurality of K and Q residues in antibodies, it is not immediately obvious how such an enzyme might be useful in site-directed conjugation. However, mTGase acts on specific recognition motifs in a context-dependent fashion, which opens a route for targeted conjugation after incorporating, by genetic means, favourable (Q)-containing motifs in pre-determined positions in the antibody chains.
Synthetic mTGase substrates containing either lysine or a lysine mimetic, often cadaverine, along with a label are employed to target Q-containing motifs. The enzyme appears to be quite promiscuous, which means that conjugates for a range of different applications can be made using mTGase conjugation technology.
Strop et al. (ref 1) inserted a LLQG motif at ~90 different sites in an anti-EGFR monoclonal antibody and identified twelve positions that permitted efficient mTGase-mediated conjugation of various substrate molecules and also resulted in good biophysical properties.
Other Q-containing recognition motifs have also been successfully used in bioconjugation applications. From a set of seven putative motifs from different natural sources or de-novo designed, Ebenig et al. (ref 2) found that the motif DIPIGQGMTG afforded the highest efficiency of conjugation.
Huggins et al. (ref 3) used an engineered anti-CD33 human IgG4 to explore the properties of siRNA conjugates prepared using an mTGase tagging strategy. In this approach, the H and L variable domains were cloned upstream of human IgG4 constant regions and K light chain domains. The C terminal H-chain lysine (K) was replaced with a short TGase motif, LLQGA, by PCR mutagenesis. After expression in ExpiCHO-S cells and purification by protein A and size-exclusion chromatography (SEC), the efficiency and specificity of the reaction between the tagged antibody and a fluorescent peptide KAYA-[PEG6]-K(Fluorescein) was assessed.
The authors showed by SDS-PAGE that H chains, but not L chains showed slightly reduced mobility, consistent with the attachment of a short peptide. Off target inter-mAb K-to-Q conjugation (polymerization) was not observed. Furthermore, no fluorescence was associated with the L chains. In flow cytometry, CD33-positive THP1 cells gave the expected staining pattern, indicating that the modified antibody was still functionally active.
Next, the authors attached K-[PEG6]-SGK(azide) to the Q-tagged antibody and separated the resulting conjugate from MTGase and excess small peptide by SEC. The azide group is a chemical handle that may be reacted with DBCO-tagged molecules, in this case DBCO-siRNA, in an overnight reaction 37oC. H chains were shifted on SDS-gels by ~15K as expected for the siRNA and >90% of chains carried the nucleic acid.
Using secondary antibodies this time, the antibody-siRNA conjugate was shown in flow cytometry to retain the antigen binding avidity of the parent un-conjugated mAb. Drug antibody ratios (DAR) were close to 2, as expected for a construct with a single tag on each H chain. The authors note that a DAR of 4 is possible by also tagging the L chains, although this requires extra genetic manipulation.
In conclusion, transglutaminase represents a powerful tool for site-directed antibody conjugation, offering specificity and precision, particularly in the development of AOC therapeutic agents.
References