Monospecific bivalent scFv-SH: effects of linker length and location of an engineered cysteine on production, antigen binding activity and free SH accessibility.

Development of tumor targeting pharmaceuticals on a modular platform is an attractive paradigm. Design choices for bispecific (anti-tumor and anti-chelate) pretargeting molecules are increased by the use of scFvs. Because a scFv is monovalent and small in size, its functional affinity and in vivo residence time can be improved through multimerization. ScFv multimers can be covalent or non-covalent. In vivo studies indicate that covalent scFv multimers are preferable. Attachment of scFv modules to scaffolds offers a wide range of possibilities for size and valency. A free thiol introduced at the C terminal end of a scFv (scFv-SH) allows for site-specific covalent attachment to a PEG scaffold without interfering with its antigen (Ag) binding. Although in theory, multimerization of 3 or 4 scFvs can be achieved by direct conjugation, as scFv-SH, to a tri or tetrafunctionalized PEG, it is not a practical option since homogeneous tri and tetrafunctionalized PEG are not readily available. However, the generation of (scFv)(3-4)-PEG molecules through attachment of combinations of di-scFv-SH (tandemly expressed scFvs) and scFv-SH or 2 di-scFv-SH to a bifunctional PEG is a sound approach that also allows for better control of the scFv-PEG conjugate molecular composition. Optimization of the molecular format of the di-scFv-SH module for production as soluble proteins in E. coli, Ag binding and conjugation is reported in this study. ScFvs in the VH-VL format were used for the di-scFv constructs since Fv domain inversion to VL-VH, while not yielding more protein, also abolished Ag binding. The effects on production yield, Ag binding and conjugation potential of the scFv joining linker length and the presence and location of an engineered cysteine were assessed in vitro. Our data indicate that for di-scFv-SH, an increase of the scFv joining linker length results in higher production and better Ag binding; a 20 aa long linker (G(4)S)(4) was the longest linker tested. For the engineered cysteine, three locations were tested; within the scFv joining linker, at the C terminus upstream of the E Tag and as the carboxy terminal aa. The accessibility of the free SH assessed by conjugation of di-scFv-SH to HRP-Mal demonstrated that di-scFv-HRP conjugates are formed with comparable efficiencies when the cysteine is located at the scFv carboxy end. This empirical work provides a framework for the development of bispecific scFv multimers via site-specific attachment of scFv-SH and di-scFv-SH modules to a scaffold.