Avian magnetoreception is assumed to occur in the retina. Although its molecular mechanism is unclear, magnetic field-dependent formation and the stability of radical-containing photointermediate(s) are suggested to play key roles in a hypothesis called the radical pair mechanism. Chicken cryptochrome4 (cCRY4) has been identified as a candidate magnetoreceptive molecule due to its expression in the retina and its ability to form stable flavin neutral radicals (FADH●) upon blue light absorption. Herein, we used millisecond flash photolysis to investigate the cCRY4 photocycle, in both the presence and absence of dithiothreitol (DTT); detecting the anion radical form of FAD (FAD●–) under both conditions. Using spectral data obtained during flash photolysis and UV–visible photospectroscopy, we estimated the absolute absorbance spectra of the photointermediates, thus allowing us to decompose each spectrum into its individual components. Notably, in the absence of DTT, approximately 37% and 63% of FAD●– was oxidized to FADOX and protonated to form FADH●, respectively. Singular value decomposition analysis suggested the presence of two FAD●– molecular species, each of which was destined to be oxidized to FADOX or protonated to FADH●. A tyrosine neutral radical was also detected; however, it likely decayed concomitantly with the oxidation of FAD●–. On the basis of these results, we considered the occurrence of bifurcation prior to FAD●– generation, or during FAD●– oxidization, and discussed the potential role played by the tyrosine radical in the radical pair mechanism.

Cryptochromes (CRYs) are multifunctional molecules that act as a circadian clock oscillating factor, a blue-light sensor, and a light-driven magnetoreceptor. Cry genes are classified into several groups based on the evolutionary relationships. Cryptochrome 6 gene (Cry6) is present in invertebrates and lower vertebrates such as amphibians and fishes. Here we identified a Cry6 ortholog in Xenopus tropicalis (XtCry6). XtCRY6 retains a conserved long N-terminal extension (termed CRY Nterminal extension; CNE) that is not found in any CRY in the other groups. A structural prediction suggested that CNE contained unique structures; a tetrahelical fold structure topologically related to KaiA/RbsU domain, overlapping nuclear-and nucleolar-localizing signals (NLS/NoLS), and a novel motif (termed DI-UIM) overlapping a double-sided ubiquitin-interacting motif (DUIM) and an inverted ubiquitininteracting motif (IUIM). Potential activities of the NLS/NoLS and DI-UIM were examined to infer the molecular function of XtCRY6. GFP-NLS/NoLS fusion protein exogenously expressed in HEK293 cells was mostly observed in the nucleolus, while GFP-XtCRY6 was observed in the cytoplasm. A glutathione S-transferase (GST) pulldown assay suggested that the DI-UIM physically interacts with polyubiquitin. Consistently, protein docking simulations implied that XtCRY6 DI-UIM binds two ubiquitin molecules in a relationship of a twofold rotational symmetry with the symmetry axis parallel or perpendicular to the DI-UIM helix. These results strongly suggested that XtCRY6 does not function as a circadian transcriptional repressor and that it might have another function such as photoreceptive molecule regulating lightdependent protein degradation or gene expression through a CNE-mediated interaction with ubiquitinated proteins in the cytoplasm and/or nucleolus.