Because cyanobacteriochrome photoreceptors need only a single compact domain for chromophore incorporation and for absorption of visible spectra including the long-wavelength far-red region, these molecules have been paid much attention for application to bioimaging and optogenetics. Most cyanobacteriochromes, however, have a drawback to incorporate phycocyanobilin that is not available in the mammalian cells. In this study, we focused on biliverdin (BV) that is a mammalian intrinsic chromophore and absorbs the far-red region and revealed that replacement of only four residues was enough for conversion from BV-rejective cyanobacteriochromes into BV-acceptable molecules. We succeeded in determining the crystal structure of one of such engineered molecules, AnPixJg2_BV4, at 1.6 Å resolution. This structure identified unusual covalent bond linkage, which resulted in deep BV insertion into the protein pocket. The four mutated residues contributed to reducing steric hindrances derived from the deeper insertion. We introduced these residues into other domains, and one of them, NpF2164g5_BV4, produced bright near-infrared fluorescence from mammalian liver in vivo. Collectively, this study provides not only molecular basis to incorporate BV by the cyanobacteriochromes but also rational strategy to open the door for application of cyanobacteriochromes to visualization and regulation of deep mammalian tissues.
Keywords: bilin; in vivo imaging; protein engineering.