Understanding CRY2 interactions for optical control of intracellular signaling

Abstract

Arabidopsis cryptochrome 2 (CRY2) can simultaneously undergo light-dependent CRY2-CRY2 homo-oligomerization and CRY2-CIB1 hetero-dimerization, both of which have been widely used to optically control intracellular processes. Applications using CRY2-CIB1 interaction desire minimal CRY2 homo-oligomerization to avoid unintended complications, while those utilizing CRY2-CRY2 interaction prefer robust homo-oligomerization. However, selecting the type of CRY2 interaction has not been possible as the molecular mechanisms underlying CRY2 interactions are unknown. Here we report CRY2-CIB1 and CRY2-CRY2 interactions are governed by well-separated protein interfaces at the two termini of CRY2. N-terminal charges are critical for CRY2-CIB1 interaction. Moreover, two C-terminal charges impact CRY2 homo-oligomerization, with positive charges facilitating oligomerization and negative charges inhibiting it. By engineering C-terminal charges, we develop CRY2high and CRY2low with elevated or suppressed oligomerization respectively, which we use to tune the levels of Raf/MEK/ERK signaling. These results contribute to our understanding of the mechanisms underlying light-induced CRY2 interactions and enhance the controllability of CRY2-based optogenetic systems.Cryptochrome 2 (CRY2) can form light-regulated CRY2-CRY2 homo-oligomers or CRY2-CIB1 hetero-dimers, but modulating these interactions is difficult owing to the lack of interaction mechanism. Here the authors identify the interactions facilitating homo-oligomers and introduce mutations to create low and high oligomerization versions.

Fig. 1

Electrostatic charges at N- and C-termini are important for CRY2–CIB1 and CRY2–CRY2 interactions, respectively. a Surface charge distribution of CRY2wt predicted by homology modeling. Charged residues are labeled in the insets of both the N-terminus and C-terminus. b Electrostatic charges at the N-terminus of CRY2 are removed by having charged amino acids replaced with neutral ones (CRY2(neutral2–6)) or deleted (CRY2(Δ2–6)). c The light-induced CRY2–CIB1 binding is much weaker for CRY2(neutral2-6) or CRY2(Δ2–6), as compared to CRY2wt. COS7 cells were co-transfected with CIB1-GFP-Sec61 and each mCh-tagged CRY2, respectively. Blue light was delivered at 2-s intervals for 100 s. d The amino acid sequences for the C-terminus of CRY2wt, CRY2(E490G) and each truncated CRY2. Truncated CRY2 variants, including CRY2(Δ491–498), CRY2(Δ490–498), CRY2(Δ489–498), CRY2(Δ488–498) and CRY2(Δ487–498), were constructed by sequentially removing residues from CRY2(487–490). e Cytosolic CRY2(E490G) and CRY2(Δ490–498) formed clusters upon light stimulation, while the other C-terminally truncated derivatives did not. COS7 cells were transfected with each mCh-tagged CRY2 and stimulated with blue light at 5-s intervals for 500 s. Scale bars, 5 µm (c), 10 µm (e)

Fig. 2

Double-positive charges at residues 489 and 490 enhance CRY2 homo-oligomerization. a A series of mutations were introduced at positions 489 and 490 of CRY2. b Cytosolic CRY2(R489D) and CRY2(R489E) did not form noticeable clusters, while the other mutants oligomerized into clusters drastically. Blue light was delivered at 5-s intervals for a total of 500 s. c The cluster mass was quantified using a custom Matlab program and normalized against that of CRY2(E490G). d Half-maximal clustering time of each CRY2 variant. e Dissociation half time of each CRY2 variant. Results are presented as means ± s.e.m. (n = 13, 12, 14, 21, 14). Results were analyzed using one-way ANOVA with Dunnett’s post hoc test. (*P < 0.05, ****P = 0.0001). Scale bars, 10 µm

Fig. 3

Negative charges at the C-terminus of CRY2 inhibit CRY2 homo-oligomerization. COS7 cells were co-transfected with CIB1-GFP-Sec61 and each mCh-tagged CRY2, respectively. Blue light of 200-ms duration was delivered at 2-s intervals for 20 s (a–c) or one 100-ms pulse of blue light was delivered (d). a Increasing numbers of negative charges were introduced to the C-terminus of CRY2(1–488). b After recruitment to the ER membrane via CRY2–CIB1 hetero-dimerization, CRY2 derivatives with more C-terminal negative charges formed fewer clusters as compared to CRY2wt. c The cluster mass of each CRY2 mutant on ER membrane was separately quantified and then normalized to that of CRY2wt. Results are presented as means ± s.e.m. (n = 13, 12, 15, 14, 16, 12, 11, 13, 11, 15, 12) and analyzed using one-way ANOVA with Dunnett’s post hoc test. (*P < 0.05, ****P = 0.0001). d CRY2high, CRY2wt and CRY2low formed different amount of clusters on plasma membrane at different CRY2 concentrations. (CRY2high n = 48, CRY2wt n = 48, CRY2low n = 49) Scale bars, 5 µm

Fig. 4

Fusion with a large fluorescent protein, tandem dimeric Tomato, further decreases CRY2low homo-oligomerization. Blue light was delivered at 5-s intervals for 100 s (a, b) or at 2-s intervals for 20 s (c, d). a In cells expressing both CRY2wt-GFP-Sec61 and CIB1-tdTom, blue light induced fewer clusters formation compared to cells expressing only CRY2wt-GFP-Sec61. b Quantification of the cluster mass of CRY2wt-GFP-Sec61, with or without co-expressing CIB1 fusion proteins, shows that CIB1-tdTom suppresses CRY2 cluster formation. (n = 11, 16, 11, 10). c After recruitment to the ER membrane via CRY2-CIB1 hetero-dimerization, CRY2wt-tdTom formed fewer clusters than CRY2wt-mCh, and CRY2low-tdTom did not form visible clusters on the ER network. d Illustration of CRY2 homo-oligomerization suppressed by the large protein tdTom. e The cluster mass of CRY2 on ER membrane after recruitment via CRY2–CIB1 hetero-dimerization was quantified and normalized against that of CRY2wt-mCh. (n = 16, 12, 15, 14). Results are presented as means ± s.e.m. and analyzed using one-way ANOVA with Dunnett’s post hoc test. (**P < 0.005, ***P < 0.0005). Scale bars, 5 µm

Fig. 5

Optical control of Raf activation using CRY2 constructs with differential clustering capacities. a Illustrative scheme for light-induced activation of MAPK signaling pathway. b Light-activatable Raf1 with different CRY2 oligomerization propensities. c Cells expressing CRY2high-mCh-Raf1 yielded the highest level of pERK after light stimulation, while cells expression CRY2low-tdTom-Raf1 produced the lowest level of pERK after light illumination. Western blot analysis of phosphorylated ERK (pERK, Thr202 and Tyr204) was conduceted after blue light illumination at 0.2 mW cm⁻² for 5 min. d The pERK value decreased from CRY2high-mCh-Raf1, CRY2wt-mCh-Raf1, CRY2low-mCh-Raf1 to CRY2low-tdTom-Raf1 after light stimulation. The pERK value in each transfection was averaged from two technical replicates from two independent experiments. The values were normalized against pERK from cells expressing CRY2wt-mCh-Raf1. Results are presented as means ± s.e.m. and analyzed using one-way ANOVA with Dunnett’s post hoc test. (*P < 0.05)