Light-induced Conformational Changes of LOV1 (Light Oxygen Voltage-sensing Domain 1) and LOV2 Relative to the Kinase Domain and Regulation of Kinase Activity in Chlamydomonas Phototropin*

Background: The plant photoreceptor “phototropin” is a light-regulated kinase containing two photosensory domains named LOV. Results: Light-induced conformational change related to the kinase activation was detected in full-length phototropin of Chlamydomonas. Conclusion: LOV1 may interact with LOV2 and modify the photosensitivity of the kinase regulation by LOV2. Significance: Configuration of LOV1, LOV2, and kinase domain in a phot molecule is first demonstrated. Phototropin (phot), a blue light (BL) receptor in plants, has two photoreceptive domains named LOV1 and LOV2 as well as a Ser/Thr kinase domain (KD) and acts as a BL-regulated protein kinase. A LOV domain harbors a flavin mononucleotide that undergoes a cyclic photoreaction upon BL excitation via a signaling state in which the inhibition of the kinase activity by LOV2 is negated. To understand the molecular mechanism underlying the BL-dependent activation of the kinase, the photochemistry, kinase activity, and molecular structure were studied with the phot of Chlamydomonas reinhardtii. Full-length and LOV2-KD samples of C. reinhardtii phot showed cyclic photoreaction characteristics with the activation of LOV- and BL-dependent kinase. Truncation of LOV1 decreased the photosensitivity of the kinase activation, which was well explained by the fact that the signaling state lasted for a shorter period of time compared with that of the phot. Small angle x-ray scattering revealed monomeric forms of the proteins in solution and detected BL-dependent conformational changes, suggesting an extension of the global molecular shapes of both samples. Constructed molecular model of full-length phot based on the small angle x-ray scattering data proved the arrangement of LOV1, LOV2, and KD for the first time that showed a tandem arrangement both in the dark and under BL irradiation. The models suggest that LOV1 alters its position relative to LOV2-KD under BL irradiation. This finding demonstrates that LOV1 may interact with LOV2 and modify the photosensitivity of the kinase activation through alteration of the duration of the signaling state in LOV2.

Plants sense and respond to variations in environmental conditions. Light is one of the most essential environmental signals for photosynthetic plants. Through evolution, plants have acquired three major light sensors: phytochrome (1), cryptochrome (2) and phototropin (3). Phototropin (phot) 3 is a BL sensor and was determined to be a receptor for the phototropic response (4). Afterward, phot was shown to mediate chloroplast relocation (5), stomata opening (6), leaf flattering (7), and leaf photomorphogenesis (8). All of these responses serve to optimize the efficiency of photosynthetic activities. Most plants, including Arabidopsis thaliana (At), 3 have two isoforms of phot (phot1 and phot2) that share roles depending on the light intensity (9).
The phot molecule is composed of ϳ1000 amino acid residues and has two photoreceptive domains named light-oxygenvoltage 1 (LOV1) and LOV2 (10) in the N-terminal region; the C-terminal half forms a Ser/Thr kinase domain (KD) (Fig. 1). LOV binds one flavin mononucleotide (FMN) noncovalently on a five-stranded ␤-sheet, which is referred to as the ␤-scaffold in the ␣/␤-fold (11). Upon BL irradiation, the FMN in the ground state (D450) transiently forms a covalent bond with a Cys residue conserved among LOVs to form an adduct state (S390) (12) via a triplet-excited state (L660) (13). Within seconds to minutes, the S390 reverts to D450, thermally depending on the types of LOV (14), and forms a characteristic photocycle with LOV. S390 is a signaling state capable of activating the kinase (15). Of the two LOV domains, LOV2 has been shown to play a major role in kinase activation by BL through both autophosphorylation (16,17) and substrate phosphorylation in vitro (18). Phot acts, therefore, as a light-regulated protein kinase (3). BL-mediated phosphorylation of some signaling components is reported (19,20). Recently, a Ser/Thr protein kinase, BLUS1, was found to be a substrate of phot and to mediate a primary step for phototropin signaling in stomata opening of guard cells in Arabidopsis (21).
One of the most interesting issues concerning the structurefunction relation of phot is how LOV2 regulates the kinase activity. It has been postulated that an ␣-helix named J␣ is involved in this connection. Based on the results of an NMR study with Avena sativa (As) phot1 (22), J␣, which locates next to the C terminus of LOV2 in the linker region between LOV2 and KD ( Fig. 1), is suggested to interact with the ␤-scaffold in the dark and to unfold and dissociate from the ␤-scaffold upon BL excitation. Localization of J␣ onto the ␤-scaffold was visualized by a crystallographic study (23). At phot1 also showed structural changes in the J␣ region that were detectable by SAXS (24) and Fourier transform infrared (FTIR) spectroscopy (25). However, the molecular mechanism for the regulation of the kinase by LOV2 is still obscure. One of the main reasons for this problem is the difficulty in preparing phot samples consisting of both LOV and KD with a high enough purity and in large enough quantities for biophysical analyses. To overcome this difficulty, we previously established a preparation system for pure LOV2-KD of At phot1 and phot2 (15,26). The LOV2-KDs showed BL-regulated kinase activity on the N-terminal fragment of At phot1 (AtP1Nt), including the autophosphorylation sites around the LOV1 region that was created as an artificial substrate for the phot kinase (15,26). Because the LOV2-KDs have BL-regulated kinase activity, structural study of these molecules provides useful information regarding the molecular basis for the BL-dependent regulation of the kinase. In this regard we previously measured the SAXS of a D720N substitute of At phot2 LOV-KD and reported its molecular models, which revealed the topological organization of LOV2 and KD and its BL-induced alteration (27).
The next question is where LOV1 resides and how it interacts with LOV2 and KD in a phot molecule. Currently, the phot of the unicellular green alga Chlamydomonas reinhardtii (Cr) (28) is the only species available to answer this question. Cr phot is proposed to mediate sexual differentiation (29), expression of several photosynthetic genes (30), and size of eyespot and phototaxis (31). Cr phot is supposed to exist in a monomeric form in contrast to higher plant phots, which are proposed to be in a dimeric form in vivo (32). Furthermore, its primary structure differs from those of higher plants in the N-terminal extension and the hinge between LOV1 and LOV2 regions (Fig. 1). Despite these differences, Cr phot is able to complement the physiological responses mediated by phot1 and phot2 in Arabidopsis (33), suggesting that the Cr phot molecule has similar machinery for the regulation of signal transduction by BL when compared with those of higher plant phots. Recently, we have established a large scale preparation system for highly pure fulllength Cr phot and reported that the Cr phot has BL-dependent kinase activity and that the N-terminal region of LOV2 is involved in BL signaling (34).
In the present study we investigated the photochemistry, kinase activities, and molecular structures of the highly purified Cr phot and its LOV2-KD fragment. The results revealed distinct functional roles of LOV1 in the regulation of kinase activity and, for the first time, the organization of LOV1, LOV2, and KD in a full-length phot molecule. Based on these results, we discuss the regulatory mechanism of Cr phot kinase in comparison with those of higher plants.

EXPERIMENTAL PROCEDURES
Sample Preparation-The vector for overexpression of fulllength Cr phot (CrPFul) was described in a previous paper (34). For site-directed mutants of CrPFul, PCR-based mutageneses were performed using PrimeStar GXL DNA polymerase (TaKaRa) with primers for C57A, C250A, and D545N (Table 1). For the LOV2-KD fragment of Cr phot (CrPL2K), DNA of the LOV2 ϩ linker ϩ kinase (192-749) was synthesized by PCR with primers (Table 1), and the amplified fragment was inserted into the NdeI/SalI site of the pET28a bacterial expression vector (GE Healthcare). CrPFul and CrPL2K were purified as described in the previous paper (27). The Escherichia coli BL21 (DE3) cells overexpressed the protein, which were suspended in the purification buffer and lysed by sonication. After centrifugation (100,000 ϫ g, for 30 min), the sample was purified from the supernatant by nickel affinity (HisTrap, GE Healthcare) and size exclusion column (Superdex 200 pg, GE Healthcare) chromatography. Roughly purified sample was loaded onto an anion exchange column (Mono Q, GE Healthcare). Purified sample was fractionated in the flow-through. Purified sample was stored at Ϫ80°C in a buffer (20 mM Tris-HCl, pH 7.8, 200 mM NaCl, 10% (w/v) glycerol, 1 mM Na 2 EGTA) after concentration by spin column (Amicon Ultra, Millipore).
Spectroscopy-Ultraviolet (UV)-visible absorption spectra were recorded with a spectrophotometer (model U3100, Hitachi-hitec) equipped with a thermostat controller (model 131We were unable to verify your reference in PubMed, please confirm that it is correct or change as needed. Also, please provide this article title.0305, Hitachi hitec). Samples were excited with a combination of a BL emitting diode (LED) (LUXEON star, Lumileds Lighting, maximum emission at 465 nm) and an electronic shutter (COPAL) (15). The BL-excited spectra were recorded under blue illumination. The reversion of S390 to D450 in the dark was monitored by the absorption changes at 450 nm at the temperatures indicated. Kinase Assay-CrPFul or CrPL2K was incubated with AtP1Nt as an artificial substrate in a kinase reaction buffer (20 mM Tris-HCl, pH 7.8, 100 mM NaCl, 10% (w/v) glycerol, 1 mM Na 2 EGTA, 10 mM MgCl 2 , 10 M ATP containing 3.7 kBq l Ϫ1 [␥-32 P]ATP) for 15 min at 20°C. The effect of blue light on phosphorylation was measured by either irradiation with a blue LED (ISL-150X150-88, CCS Inc. Japan, emission maximum at 475 nm) or mock irradiation. The intensity of the irradiation was varied with natural density filters. The reaction was stopped by the addition of SDS sample buffer followed by boiling for 3 min. Samples were run on SDS-PAGE. After Coomassie Brilliant Blue (CBB) staining, the gel was dried, and phosphorylated bands were visualized with an imaging plate (Fuji film) and a scanner (STORM, GE Healthcare) (15). The signal intensities were quantified by ImageJ software (rsbweb. nih.gov).
SAXS Measurement and Analysis-SAXS data of CrPL2K and CrPFul were collected at BL45XU of SPring-8 using a PILATUS3 detector (DECTRIS). The x-ray wavelength was tuned to 0.9000 Å, and the camera distance was ϳ2000 mm. The exposure time was 60 s, and each SAXS pattern was composed of 18 frames. Other details were described previously (27). The temperature of the sample cell was maintained at 293 K. Before measurements, concentrated samples were centrifuged at 150,000 ϫ g for 1 h using a Himac CP 85␤ ultracentrifuge (Hitachi, Japan) to remove aggregated proteins. Before SAXS measurement, the fluid dynamic diameter of CrPFul was measured by dynamic light scattering using a Zetasizer Nano (Malvern Instruments). The decay of the time-correlation function of the back-scattered light monotonously decreased and was well fitted by a theoretical curve calculated assuming that particles with the dimension of ϳ100 Å are monodispersively suspended. This observation indicates that CrPFul is in a monomeric form in solution.
SAXS profiles of both samples were collected in the concentration range from 1.2 to 2.3 mg ml Ϫ1 . For each sample, SAXS was measured sequentially in the dark, under BL after pre-irradiation for 5 min, and again in the dark after the dark adaptation for Ͼ5 min after the measurement under BL. The fluence rate of BL at the sample position was 450 M m Ϫ2 s Ϫ1 using a blue LED (27). SAXS data of hen egg white lysozyme was collected as a molecular weight reference. A small amount of radiation damage to all samples was confirmed by the stabilities of SAXS profiles, absorption spectra, and SDS-PAGE patterns after x-ray exposure.
SAXS Analysis-The two-dimensionally recorded SAXS patterns were reduced to one-dimensional profiles after subtraction of the background scattering from the buffer solution. Profiles in the small-angle region were analyzed by Guinier's plot (35) to obtain zero-angle scattering intensity and radius of gyration (R g ). Their concentration dependences were analyzed as described previously (27), and the distance distribution function P(r) was calculated using GNOM (37). The low resolution molecular models were restored as an assembly of small spheres with a diameter of 3.8 Å using GASBOR (38). Mathematics for Guinier's plot and the distance distribution function are indicated below.
The small-angle regions of background-subtracted SAXS profiles were analyzed by Guinier's approximation (35). At a scattering vector length S, the scattering intensity I(S,C) of a protein solution at a concentration C is approximated by the forward scattering intensity I(S ϭ 0,C) and the radius of gyration R g (C) as where 2 is the scattering angle, and is the x-ray wavelength. Under dilute conditions, the dependence of I(S ϭ 0,C) and R g (C) 2 upon the concentrations are approximated as where K is a constant, M r is the apparent molecular weight of the protein, A 2 is the second virial coefficient, and B if reflects interprotein interactions. The sign of B if is identical to that of A 2 . Assuming a partial specific volume of 0.74 cm 3 /g for soluble proteins, we can determine the apparent M r of a soluble protein using I(S ϭ 0, C ϭ 0) of a reference protein with a known M r . The S Ͻ 0.005 Å Ϫ1 intensity profiles were extrapolated to the infinite dilution limit to correct for the concentration effects on the scattering profiles. The corrected profiles were merged with the S Ͼ 0.005 Å Ϫ1 profile measured from 2.3 mg ml Ϫ1 solution.
The distance distribution function P(r) was calculated using GNOM (37) software. The low resolution molecular models of P2L2K were restored as an assembly of small spheres with a diameter of 3.8 Å Ϫ1 , called dummy residues, using GASBOR (38) software. GASBOR minimizes the discrepancy between experimental and calculated scattering profiles by keeping a compactly interconnected configuration of dummy residues approximating a molecular shape. The discrepancy in the observed and the calculated scattering profiles was monitored via the 2 value, which is defined as where n is the number of experimental data points, S j is the scattering vector of the jth data point, c(S j ) is a correction factor, K is a scale factor, and (S j ) is the statistical error in the experimental scattering profile I exp (S j ). I model (S j ) represents the scattering profile of the predicted structural model. As GASBOR analysis does not provide a unique solution for three-dimensional structures, 20 independent calculations were performed for a targeted profile, and the obtained molecular models were aligned manually (39).

UV-visible Spectrum and Photoreaction of CrPFul and
CrPL2K-The purity of the two preparations used in the experiments, CrPFul and CrPL2K (Fig. 1), was estimated at Ͼ95% by CBB staining of the SDS-PAGE gel ( Fig. 2A). CrPFul showed the same UV-visible absorption spectrum in the dark and under BL irradiation as those reported previously (Fig. 2B in Ref. 34). The UV-visible absorption spectrum of CrPL2K in the D450 state exhibited peaks at 471, 445, 369, and 354 nm (Fig. 2B), which is almost the same as those of CrPFul (see Fig. 6 in Ref. 34) and the truncated LOV2 of Cr phot (40). Under BL irradiation, the absorption at 450 nm decreased, whereas absorption at 390 nm increased (Fig. 2B), indicating the formation of a cysteinyl-flavin adduct. The S390 of both CrPFul and CrPL2K reverted to D450 in the dark (data not shown), showing a characteristic photocycle with LOVs.
We have previously shown that the S390 in LOV2 is involved in the photosensitivity of the activation of phot kinase by BL in At phot (15); then the dark reversion of S390 to D450 was measured. In CrPFul, the time course of the dark reversion was well approximated with a double-exponential curve. The half-lives (t1 ⁄ 2 ) of the two components were calculated as 13.0 and 54.8 s (Fig. 3, upper panel, and Table 2), which is similar to previously reported values (34). The faster and the slower components were attributed to LOV2 and LOV1, respectively (15,40). Dark reversion of S390 to D450 in LOV2 of CrPL2K was also fitted with a double-exponential curve despite the lack of LOV1. Similar double-exponential kinetics was reported with the truncated LOV2, although the origins were unclear (40). The major (91%) and the minor (9%) components showed t1 ⁄ 2 of 5.1 and 104 s, respectively (Fig. 3, lower panel, Table 2). When comparing the t1 ⁄ 2 in LOV2 of CrPFul (Fig. 3, dashed line) with that of the major component of CrPL2K, it was observed that the reversion of S390 in LOV2 is accelerated by ϳ2.5 times in the presence of LOV1. It is interesting to see if the acceleration requires the photoreaction in LOV1. Therefore, we measured the dark reversion of S390 in LOV2 of the C57A substitute of CrPFul (CrPFul_C57A), which had lost the ability to form adducts in LOV1. The reversion was well simulated with a single-exponential curve. The t1 ⁄ 2 was calculated as 10.2 s (Table 2), which is slightly shorter than that of CrPFul; however, it is 2 times longer than that of CrPL2K. This indicates that the presence of LOV1 itself is the predominant contributor to the acceleration of the dark reversion in LOV2.
In turn, we also measured the dark reversion in LOV1 of the C250A substitute of CrPFul (CrPFul_C250A), which had lost its ability to form an adduct in LOV2. The reversion in CrPFul_C250A was also well approximated with a single-expo-   nential curve and showed the same t1 ⁄ 2 as LOV1 with that of CrPFul (54.1 s) ( Table 2). Hence, the dark reversion in LOV1 of CrPFul is not affected by the photoreaction in LOV2. Similar vectorial effects of the photoreactions upon the decay behavior of S390 between LOV1 and LOV2 were reported with truncated LOV1-LOV2 fragments of Cr phot (40). In addition, the effect of kinase activity on the dark reversion was measured using a D545N substitute (CrPFul_D545N), which lacks ATP binding ability and kinase activity. CrPFul_D545N showed a double-exponential reversion profile with a t1 ⁄ 2 of 13.8 and 54.1 s ( Table 2), indicating that the disruption of the kinase activity does not affect the dark reversion in both of the LOVs. These results provided useful information regarding the effects of the interdomain interactions between LOV2 and LOV1 or KD on the duration of the signaling state in LOV2. Activation of Kinase by BL in CrFul and CrPL2K-We have reported previously that CrPFul was able to phosphorylate AtP1Nt (34). We verified that the present CrPFul preparation was able to phosphorylate AtP1Nt in a manner that was dependent on the kinase activity in the KD (Fig. 4A). CrPL2K was also revealed to phosphorylate AtP1Nt in a light-dependent manner (Fig. 4B). The fluence response of the photoactivation of the kinase of CrPFul and CrPL2K was compared (Fig.  5); CrPFul exhibited 50% (see the legend to Fig. 5B) kinase activation at 5 mol m Ϫ2 s Ϫ1 , and the activation exceeded 90% at 20 mol m Ϫ2 s Ϫ1 . In contrast, CrPL2K showed 50% kinase activation at 10 mol m Ϫ2 s Ϫ1 , and the activation was saturated at 50 mol m Ϫ2 s Ϫ1 . Based on the BL intensity for 50% activation, CrPFul is ϳ2 times more sensitive to BL than CrPL2K in terms of the photosensitivity of the kinase activation.
Strangely, CrPFul_C250A, which does not exhibit adduct formation in LOV2, also maintained BL kinase activation ability, although to a lesser extent than CrPFul (Fig. 5A). The leaky kinase activity of the C250A substitute was also observed with a Cr phot by yeast kinase assay; however, the cause was unclear (34). Involvement of photoreaction of LOV1 in this leaky kinase activity is negated because CrPFul CrPFul_C57/250A, which has both C57A and C250A substitutions, showed a similar lesser degree of BL-dependent kinase activation than CrPFul to that of the CrPFul_250A (Fig. 6). It has been shown that the C250S substitute of LOV2 of Cr phot forms a reduced form of FMN instead of cysteinyl-flavin adduct upon BL irradiation.
The reduced form is not reversible to the oxidized form in the dark. In contrast, C57S substitute of LOV1 forms a neutral FMN semiquinone, FMNH ⅐ , which is reversible to the oxidized form (41). Formation of the reduced form has not been reported with phot of higher plants (42). One of the possible explanations for this leaky kinase activity is that BL-dependent reduction of FMN in LOV2 may induce conformational changes similar to those that activate the kinase through the adduct formation in Cr phot.
Structural Parameters of CrFul and CrPL2K Deduced from SAXS-In the concentration range measured, the SAXS profiles of CrPFul displayed an ϳ5% increase in two S regions, S Ͻ 0.003 Å Ϫ1 and 0.01 Ͻ S Ͻ 0.02 Å Ϫ1 and an ϳ3% decrease at S ϳ 0.007 Å Ϫ1 by BL irradiation (Fig. 7), which suggests BL-induced conformational changes. Even 10 min after BL was turned off, the changes did not completely relax. This most likely reflects the delay of the global molecular changes after completion of the photocycle of the FMN chromophore. The Guinier plots (35) of all the samples were approximated by straight lines in the S 2 region of 1 ϫ 10 Ϫ5 Å Ϫ2 Ͻ S 2 Ͻ 4 ϫ 10 Ϫ5 Å Ϫ2 (Fig. 7A,  inset). With the exception of small variations, the calculated    JANUARY 3, 2014 • VOLUME 289 • NUMBER 1

JOURNAL OF BIOLOGICAL CHEMISTRY 417
C/I(S ϭ 0,C) and the R g 2 (C) were proportional to the concentration of CrPFul (Fig. 7B). These observations indicated that the CrPFul solutions were almost monodispersive in the measured concentration range under both dark and light conditions. Based on the C/I(S ϭ 0, C), the apparent M r of CrPFul was estimated at 70,000, which indicates that the concentrated solution consisted of a monomeric form of CrPFul. This result correlates with the monomeric form of CrPFul in the diluted solution, which was determined by size exclusion chromatography (34). R g values at infinitely diluted conditions were 42.7 Å in the dark and 43.4 Å under BL irradiation. The maximum dimension estimated from the P(r) function in the dark was 157 Å smaller than the 167 Å under BL (Fig. 7C). The changes in structural parameters regarding molecular dimensions indicate a BL-induced extension of CrPFul.
The SAXS profiles of CrPL2K under BL irradiation showed a significant increase of ϳ10% in a S region of S Ͻ 0.003 Å Ϫ1 and a decrease of ϳ3% at S ϳ 0.01 Å Ϫ1 , which demonstrates BLinduced conformational changes (data not shown). These profile changes are similar to those observed in the D720N substitute of the LOV2-KD fragment of At phot2 (AtP2L2K_D720N) (27). The Guinier plots deviated from straight lines in the S 2 region of 5 ϫ 10 Ϫ6 Å Ϫ2 Ͻ S 2 Ͻ 3 ϫ 10 Ϫ5 Å Ϫ2 because of aggregated components that were unable to be removed by ultracentrifugation. Molecular dimensions of CrPL2K were, therefore, estimated by applying the Guinier approximation to the S 2 region of 3 ϫ 10 Ϫ5 Å Ϫ2 Ͻ S 2 Ͻ 5 ϫ 10 Ϫ5 Å Ϫ2 , in which region the Guinier approximation worked well with AtP2L2K_D720N (27). The M r , the R g at 1.7 mg ml Ϫ1 , and the D max of P(r) (data not shown) were estimated as 50,000, 34.7 Å, and 127 Å, respectively; these values were close to the values obtained for At-P2L2K_D720N (27). Under BL irradiation, the R g and the D max increased to 36.1 and 131 Å, respectively, indicating an expansion of the molecule.
Molecular Models for CrPFul and CrPL2K-Low resolution molecular models of CrPL2K in the dark and under BL irradiation were restored using the SAXS profiles of S Ͼ 0.006 Å Ϫ1 . The restored models appeared as elongated shapes with the dimensions of 110 ϫ 40 ϫ 40 Å 3 (Fig. 8, left) and reproduced the experimental profiles as indicated by the small 2 index of less than 2.0. The models of CrPL2K in the dark and under BL irradiation were similar in shape, but the model under BL is larger at ϳ10 Å greater than the model in the dark. The molecular shapes and dimensions of CrPL2K are quite similar to those of AtP2L2K_D720N (27) (Fig. 8, center and left), suggesting that the molecular structures in the LOV2-KD region do not vary between Cr and At phots. The constructed molecular models of CrPFul in the dark appeared as an elongated shape with the dimensions of 140 ϫ 40 ϫ 40 Å 3 (Fig. 8, right). The molecular model of CrPFul was roughly divided into a main portion with an elongated spheroid shape and a protrusion, manually. The size and the shape of the main portion with an adjacent part of the protrusion are closely similar to those of CrPL2K (Fig. 6, center) and AtP2L2K_D720N (27) (Fig. 8, left), and the crystal structure of Cr LOV1 (43) fits well with the tip of the protrusion (Fig. 8, right). The molecular model of CrPFul under BL irradiation had nearly the same dimension as the model in the dark; however, the density assignable to LOV1 could be put at a different position relative to the KD from the position in the dark model. The differences in the molecular models suggest that the positions and orientations of the LOV2-LOV1 region and/or LOV1 might be lightdependent (Fig. 8, right).

DISCUSSION
The Role of LOV1 in the BL-dependent Kinase Activation of Cr Phot-Higher plant phots are thought to transmit signals downstream of the signal transduction paths via autophosphorylation (3,16) and/or phosphorylation of some signal mediators (19 -21). In this study we used Cr phot instead of At phot due to limitations in the ability to prepare a full-length sample. The results showed that AtP1Nt serves as a substrate for the kinase of Cr phot as well as At phots, suggesting that Cr phot has similar substrate recognition ability and kinase activity as At phots.
Fluence response curves showed that the photosensitivity of the BL-dependent kinase activation in Cr phot decreased to less than half when LOV1 was truncated (Fig. 5B). This suggests that LOV1 serves as an amplifier of photosensitivity. We have reported previously that the photosensitivity of the kinase activation by BL is positively correlated with the duration of S390 in LOV2 by using LOV2-KD fragments of At phot1 and phot2 with different durations and an amino acid substitution of phot1 that has a prolonged duration (15). Similarly, the fact that CrPFul has a 2ϫ higher photosensitivity than CrPL2K is well explained by the observation that LOV2 has a greater than 2ϫ longer t1 ⁄ 2 in CrPFul (13.0 s) than in CrPL2K (5.1 s) ( Table 2). Likewise, the only slightly shorter t1 ⁄ 2 of LOV2 in CrPFul_C57A than that in CrPFul (10.2s and 13.8 s, respectively) ( Table 2) may explain the finding that BL-dependent activation is almost unchanged with the C57A substitute (Fig. 4A). In contrast, a small decrease of the activation was observed with the C250A substitution (Fig. 4A), in which photoreaction of LOV1 was not involved. These findings suggest that the duration of S390 in LOV2 also plays one of the key roles in activation of the kinase by BL in Cr phot as with At phot. This is shown schematically in Fig. 9A; formation of S390 by BL in LOV2 (dark blue center circle) induces a conformational change in the protein moiety. This change leads to signal transmission, i.e. activation of kinase (red outer arc).
Thus, LOV1 may alter the photosensitivity of the BL-dependent kinase activation by modifying the duration of S390 in LOV2. Because the photoreaction in LOV1 only has a small effect on the duration of S390 in LOV2, the modification may come from the structural presence of LOV1 itself rather than the photoreaction of LOV1. This is consistent with the report that the rate of the thermal back reaction from S390 in LOV2 depends on the presence of LOV1 in a Cr phot fragment con-  (43) are fitted to the density models. The orientation and positions of the kinase and the LOV2 models are fixed to those in the AtP2L2K_D720N models reported previously (27). Black bars are the references for the positions of each domain. sisting of LOV1 and LOV2 (40). This is depicted schematically in Fig. 9A (yellow circle and arrow).
Together with the report that Cr phot complements the phenotypes of At phots (33), the molecular mechanism of the kinase activation by LOV2 is suggested to be basically conserved between Cr phot and At phots, and the Cr phot can serve as a useful molecular model to uncover this mechanism. It should be noted, however, that the observed leaky BL-dependent kinase activity of CrPFul_C250A (Fig. 4A) indicates that a currently unknown mechanism specific to Cr phot turns on the LOV2 switch to activate the kinase in a different way from the adduct formation when LOV2 lacks a reactive cysteine.
Organization of LOV1, LOV2, and KD in a Cr Phot Molecule-It has been of great interest to identify how LOV1, LOV2, and KD are configured in a phot molecule. The present SAXS studies revealed a tandem configuration of LOV1, LOV2, and KD in a photoactive full-length Cr phot molecule existing in a monomeric form in solution under both dark and BL conditions (Fig.  8). This is the first report describing the position of LOV1 relative to LOV2-KD of phot; this is shown schematically in Fig.  7B. The reconstructed tandem models suggest that LOV1 and KD are located at opposite sides of LOV2. This suggests an indirect contribution of LOV1 to the photoactivation of the kinase. This proposal is consistent with the present interpretation of the effect of LOV1 truncation on the photoactivation of the kinase that LOV2 is a main regulator of the kinase activation, whereas LOV1 alters its photosensitivity. This alteration is possibly achieved by modification of the duration of the signaling state in the neighbor LOV2. This modification may come from the structural interaction of LOV1 with LOV2 in which LOV1 may prolong the t1 ⁄ 2 of S390 in LOV2 by acting as a weight on the photoreaction, therefore, preventing the conformational change of LOV2. A similar weight effect was observed with the truncated LOV2 of Cr phot fused with calmodulin-binding protein (44).

BL-induced Global Conformation Change in a Cr Phot
Molecule-It has been postulated that the unfolding and dissociation of J␣ from the LOV core may play important roles in the photoactivation of the kinase (22,45). Previously, based on the SAXS of LOV2-KD of At phot2, we reported that BL induced a 13 Å shift of the LOV2 domain from the KD. We further proposed that the light-activated LOV2 domain triggers conformational changes in the linker region between LOV2 and KD, causing them to separate (25) (Fig. 8, left). Because secondary structure prediction reveals the presence of an ␣-helix in the linker region and the FTIR study proposed an unfolding of an ␣-helix in the linker region (46), J␣ may exist and unfold upon BL excitation in Cr phot. However, our previous report indicated that the amino acids around the N terminus of LOV2 also play essential roles in the photoactivation of the kinase in Cr phot (34). Secondary structure prediction showed the presence of an ␣-helix consisting of 10 amino acids at that position in Cr phot. In the crystal structure of LOV2-J␣ of As phot1 (23), there is a small ␣-helix consisting of four amino acids due to the truncated construct of the sample; this ␣-helix is called AЈ␣. The AЈ␣ resides close to the C-terminal region of J␣. The two helices might form an intramolecular signaling module and serve cooperatively in the photoactivation of the kinase. BL-induced changes in the module may move LOV2 away from the KD and cancel the depression of the kinase activity. This is illustrated schematically in Fig. 7B. The observed differences in the molecular models of the dark and light states (Fig. 8) might reflect these conformational changes.
To date, information concerning the conformational change in a whole molecule of phot has not been available, except for the only FTIR study of Cr phot (46). The present SAXS study presented the image of the BL-induced global conformational changes of a whole Cr phot molecule; these changes could be interpreted by the moving away and possible tilting of LOV1 relative to the LOV2-KD (Fig. 8). Because the linker region between LOV1 and LOV2 has high disorder probability, major conformational changes might take place in this linker region. A cleavage of 20 amino acids within the KD of Cr phot has been reported under high light irradiation in vivo (31); this cleavage was also observed during the preparation of Cr phot using an E. coli expression system (45) and suggests that some BL-induced conformational changes occur within the KD. This may be correlated with the helical unfolding in the activation loop of KD detected by FTIR (45). These local changes are not included in Fig. 9B. To confirm the models presented in Fig. 9B, more structural studies, including crystallography, are required.
The Divergent Role of LOV1 between Cr Phot and At Phots-In Cr phot, LOV1 was shown to amplify the kinase photoactivation; however, LOV1 attenuated the photosensitivity in At phot2 (18). When we consider these opposite functions, it should be noted that despite the high homology of the amino acid sequence in LOV1 between Cr phot and higher plant phots, the lengths (see Fig. 1), and the amino acid sequences of the hinge between LOV1 and LOV2 and the N-terminal extension differ significantly between the two. Furthermore, their LOV1 domains play different roles in the formation of oligomeric structures. Crystal structures of the LOV1 of At phot1 and phot2 showed a dimeric structure contacting in the ␤-scaffold (47). This structure is consistent with the results of SAXS, size exclusion chromatography (24), and chemical cross-linking (48) studies in solution, suggesting that LOV1 act as a dimeric site. Because light-dependent but LOV1-independent dimerization of At phots in vivo was reported (32), the involvement of LOV1 in in vivo dimerization still needs further investigation. In contrast, crystal structure of the LOV1 in Cr phot indicated a monomeric form (43) that agrees with the current SAXS results on full-length Cr phot in solution and, therefore, suggests that the LOV1 of Cr phot cannot be a dimerization site. At phot was shown to phosphorylate each other in a dimeric form (32), resulting in autophosphorylation. Cr phot, however, showed less autophosphorylation activity (33), which may be explained by the reduced efficiency of the autophosphorylation in the monomeric Cr phot. If LOV1 forms a dimer, the weight effect discussed above will be altered, and the modification mode of the photosensitivity of the kinase activation by LOV1 may be affected. This might explain the antagonistic effects of LOV1 upon the photosensitivity of the kinase activation between Cr and At phots.
Thus, the structural differences in the LOV1-containing regions are capable of producing the diversity of LOV1 functions, including the modification mode of the photosensitivity of the kinase activation. With all of the different oligomeric structures and the partially different roles of LOV1 of Cr phot compared with those of higher plant phots, the results provided useful information regarding the light-regulation mechanism of the kinase by LOV1 and LOV2. To determine the precise mechanism of this regulation, further structure/function studies of At phots as well as Cr phot are required.
In contrast to higher plant phots, Cr phot is able to transduce light signal in a different way from phosphorylation. Overexpression of a LOV1-LOV2 fragment affected the eyespot size and phototaxis in a light-dependent manner, suggesting that Cr phot transduced BL signal through protein-protein interaction (31). This may come from the different primary and quaternary structures at around LOV1 of Cr phot from those of higher plant phots discussed above. Because LOV belongs to a subset of a PAS superfamily that acts as a protein-protein interaction module in eukaryotic cellular signaling (49), it is probable that the LOV1-LOV2 of Cr phot may interact with an unidentified signaling partner(s) in a light-dependent manner. Actually, two members of LOV, F-box, and Kelch repeat family proteins, ZTL and FKF1, have been shown to interact with TOC1 and GI, respectively, through the LOV in a light-dependent manner (50,51). The BL-induced changes in the LOV1-LOV2 region seen in the SAXS models of Cr phot might be involved in the interaction. The light-inducible protein-protein interaction of Cr phot reminds us of the application to the light-regulated molecular switch in the cell signaling. Such an approach has been attempted using LOV2-J␣ construct of As phot1 (52,53) and the interaction between FKF1 and GI (36). Further structure/function studies will provide useful information regarding the application of Cr phot to this kind of research.