In advanced cancer patients, malignant cells invade and disseminate within normal cells and develop resistance to therapy with additional genetic mutations, which makes radical cure very difficult. Precision medicine against advanced cancer is hampered by the lack of systems aimed at multiple target molecules within multiple loci. Here, we report the development of a versatile diagnostic and therapeutic system for advanced cancer, named the Cupid and Psyche system. Based on the strong non-covalent interaction of streptavidin and biotin, a low immunogenic mutated streptavidin, Cupid, and a modified artificial biotin, Psyche, have been designed. Cupid can be fused with various single-chain variable fragment antibodies and forms tetramer to recognize cancer cells precisely. Psyche can be conjugated to a wide range of diagnostic and therapeutic agents against malignant cells. The Cupid and Psyche system can be used in pre-targeting therapy as well as photo-immunotherapy effectively in animal models supporting the concept of a system for precision medicine for multiple targets within multiple loci.

We previously created a low-immunogenic core streptavidin mutant No. 314 (LISA-314) by replacing six amino-acid residues for use as a delivery tool for an antibody multistep pre-targeting process (Yumura et al., Protein Sci., 22, 213-221, 2013). Here, we performed high-resolution X-ray structural analyses of LISA-314 and wild-type streptavidin to investigate the effect of substitutions on the protein function and the three-dimensional structure. LISA-314 forms a tetramer in the same manner as wild-type streptavidin. The binding mode of d-biotin in LISA-314 is also completely identical to that in wild-type streptavidin, and conformational changes were observed mostly at the side chains of substituted sites. Any large conformational changes corresponding to the reduction of B factors around the substituted sites were not observed. These results demonstrated the LISA-314 acquired low immunogenicity without losing structural properties of original wild-type streptavidin.

For a multistep pre-targeting method using antibodies, a streptavidin mutant with low immunogenicity, termed low immunogenic streptavidin mutant No. 314 (LISA-314), was produced previously as a drug delivery tool. However, endogenous biotins (BTNs) with high affinity (Kd < 10(-10) M) for the binding pocket of LISA-314 prevents access of exogenous BTN-labelled anticancer drugs. In this study, we improve the binding pocket of LISA-314 to abolish its affinity for endogenous BTN species, therefore ensuring that the newly designed LISA-314 binds only artificial BTN analogue. The replacement of three amino acid residues was performed in two steps to develop a mutant termed V212, which selectively binds to 6-(5-((3aS,4S,6aR)-2-iminohexahydro-1H-thieno[3,4-d]imidazol-4-yl)pentanamido)hexanoic acid (iminobiotin long tail, IMNtail). Surface plasmon resonance results showed that V212 has a Kd value of 5.9 × 10(-7) M towards IMNtail, but no binding affinity for endogenous BTN species. This V212/IMNtail system will be useful as a novel delivery tool for anticancer therapy.

(2015) Structurebased design and synthesis of a bivalent iminobiotin analog showing strong affinity toward a low immunogenic streptavidin mutant, Bioscience, Biotechnology, and Biochemistry, 79:4,[640][641][642]

Fragment-based ligand discovery was successfully applied to histone deacetylase HDAC2. In addition to the anticipated hydroxamic acid- and benzamide-based fragment screening hits, a low affinity (∼1 mM) α-amino-amide zinc binding fragment was identified, as well as fragments binding to other regions of the catalytic site. This alternative zinc-binding fragment was further optimized, guided by the structural information from protein–ligand complex X-ray structures, into a sub-μM, brain penetrant, HDAC2 inhibitor (17) capable of modulating histone acetylation levels in vivo.