Multivalent CXCR4-targeting nanobody formats differently affect affinity, receptor clustering, and antagonism.

ABSTRACT

The chemokine receptor CXCR4 is involved in the development and migration of stem and immune cells but is also implicated in tumor progression and metastasis for a variety of cancers. Antagonizing ligand (CXCL12)-induced CXCR4 signaling is, therefore, of therapeutic interest. Currently, there are two small-molecule CXCR4 antagonists on the market for the mobilization of hematopoietic stem cells. Other molecules with improved potencies and safety profiles are being developed for different indications, including cancer. Moreover, multiple antagonistic nanobodies targeting CXCR4 displayed similar or better potencies as compared to the CXCR4-targeting molecule AMD3100 (Plerixafor), which was further enhanced through avid binding of bivalent derivatives. In this study, we aimed to compare the affinities of various multivalent nanobody formats which might be differently impacted by avidity. By fusion to a flexible GS-linker, Fc-region of human IgG1, different C4bp/CLR multimerization domains, or via site-directed conjugation to a trivalent linker scaffold, we generated different types of multivalent nanobodies with varying valencies ranging from bivalent to decavalent. Of these, C-terminal fusion, especially to human Fc, was most advantageous with a 2-log-fold and 3-log-fold increased potency in inhibiting CXCL12-mediated Gαi- or ß-arrestin recruitment, respectively. Overall, we describe strategies for generating multivalent and high-potency CXCR4 antagonistic nanobodies able to induce receptor clustering and conclude that fusion to an Fc-tail results in the highest avidity effect irrespective of the hinge linker.