LAUSR

In plants, cryptochromes (CRYs) are blue light receptors that regulate various developmental progress. Photoactivated CRYs undergo a series of molecular events, including tetramerization, photobody formation and interactions with signal-transduction proteins, to transduce blue light signals (Wang and Lin, 2020). Owing to their light-induced self-oligomerization and proteininteracting features, CRYs are widely used in optogenetics (Seong and Lin, 2021; Figure 1a). Photoactivated CRY2 forms photobodies through phase separation (Wang et al., 2021). The CRY2 C-terminal extension (CCE) domain and its phosphorylation are required for maintaining the liquid property of CRY2 photobody (Wang et al., 2021). However, the residues essential for CRY2 phase separation and the underlying mechanism remain elusive. Consistent with the previous report (Wang et al., 2021), our results also suggest that photoactivated CRY2 undergoes phase separation to form photobodies (Figure 1b–f and Figure S1). We then investigated the determinants that influence CRY2 photobody formation and function. CRY2 contains an N-terminal photolyase homology region (PHR) domain (residues 1–489) and a CRY CCE domain (residues 490–612). In line with the previous findings (Wang et al., 2021), we also found that the CCE domain modulates the size and property of CRY2 phase separation (Figures S2 and S3). We divided the CCE into 20 parts and explored which residues influence photobody formation through alanine-substitution analyses (Figure S4A). Except for CRY2 and CRY2, the rest 18 mutants formed photobodies similar to the ones observed in CRY2 (Figure S4B). We further fused the TALE-effector NLS RVKRPRTR (T-NLS) to CRY2 and CRY2 as the CRY2 nuclear localization sequence (NLS) was partially disrupted in CRY2 and CRY2. The sizes of CRY2+T-NLS and CRY2+T-NLS speckles were smaller than those of CRY2 (Figure S4C), suggesting that the CRY2 NLS (541–557) might modulate CRY2 phase separation. CRY2+T-NLS (N538-R557 truncation) formed small speckles in the nucleus (Figure 1g and Figure S4D), which is similar to those observed for CRY2+T-NLS (Figure S2). These together suggest that the CRY2 NLS not only mediates nuclear importation but also affects photobodies size, marking a step forward in understanding the specific residues that modulates the previously discovered CRY2 phase separation (Wang et al., 2021). We further explored how the CCE domain modulates CRY2 phase separation. Consistent with a previous report (Partch et al., 2005), our nuclear magnetic resonance (NMR) results suggest that the CCE domain is highly dynamic and flexible (Figure S5). As phase separation can be driven by multivalent interactions, we investigated the inter-molecular interactions between CCE domains via NMR (Figures S6 and S7). The observed chemical shift perturbations indicate that the CCE G558-C611 region underwent slight conformational changes upon blue light irradiation (Figure 1h and Figure S7D), which is in line with a previous report (Partch et al., 2005). Signal intensities of many CCE residues (including V535, V543, E546, G558 and V579) decreased, possibly owing to the weak CCE-CCE and/or CCE-PHR-tetramer interactions of the photoactivated CRY2 (Figure 1i). Furthermore, the paramagnetic relaxation enhancement profiles indicated that the Ipara/Idia ratios decreased around CCE residues G511-F559 and S587-T604 after blue light illumination (Figure 1j and Figure S8). Taken together, these data suggest the occurrence of intermolecular CCE-CCE interactions in CCE G511-F559 and S587-T604 regions. Many of these residues were within or close to the CCE N538-R557 region, which have a role in regulating CRY2 photobody size (Figure 1g). Thus, we speculate that these interactions might contribute to phase separation of photoactivated CRY2. Whether phosphorylation affects the intermolecular interactions between the CCE domains remains to be elucidated. Cysteines can modulate phase separation (Wang et al., 2019); we wondered whether CRY2 cysteines participate in photobody formation. Eight and three cysteines were found in the PHR and CCE domains, respectively. We generated the CRY2, CRY2 and CRY2 mutants harbouring the mutations for the 8 PHR cysteines, 3 CCE cysteines and all 11 cysteines, respectively (Figure S9A). These mutants retained the photoactivated tetramerization and photoreduction capacity (Ma et al., 2020) (Figure S9B,C), suggesting that mutating these cysteines had little impact on CRY2 structure. However, CRY2 and CRY2 completely lost the photobody formation capability (Figure 1k and Figure S9D). Moreover, compared to PHR, the blue light-induced soft gel-like phase transition of PHR was significantly decreased in vitro (Figure S9E), suggesting that the PHR cysteines play crucial roles in CRY2 phase separation. Four of the eight cysteines locate on the surface of the photoactivated CRY2 tetramer (Figure S9F); we thus speculate that these residues might play roles in sensing the redox state of surrounding conditions and modulating the oligomerization of CRY2 tetramer. How these cysteines influence