TY - JOUR
AU - Kurisu,, Genji
AB - Abstract Plant-type ferredoxin (Fd) is an electron transfer protein in chloroplast. Redox-dependent structural change of Fd controls its association with and dissociation from Fd-dependent enzymes. Among many X-ray structures of oxidized Fd have been reported so far, very likely a given number of them was partially reduced by strong X-ray. To understand the precise structural change between reduced and oxidized Fd, it is important to know whether the crystals of oxidized Fd may or may not be reduced during the X-ray experiment. We prepared the thin plate-shaped Fd crystals from Chlamydomonas reinhardtii and monitored its absorption spectra during experiment. Absorption spectra of oxidized Fd crystals were clearly changed to that of reduced form in an X-ray dose-dependent manner. In another independent experiment, the X-ray diffraction images obtained from different parts of one single crystal were sorted and merged to form two datasets with low and high X-ray doses. An Fo–Fo map calculated from the two datasets showed that X-ray reduction causes a small displacement of the iron atoms in the [2Fe-2S] cluster. Both our spectroscopic and crystallographic studies confirm X-ray dose-dependent reduction of Fd, and suggest a structural basis for its initial reduction step especially in the core of the cluster. absorption spectroscopy, ferredoxin, protein–protein interaction, redox-dependent structural change, X-ray crystallography Ferredoxin (Fd) is a relatively small iron-sulphur protein with a FeS cluster(s) as its redox centre and serves as an electron carrier in a variety of redox reactions within the cell. Fds are classified into four types based on the type of FeS cluster; [1Fe-0S], [2Fe-2S], [3Fe-4S] and [4Fe-4S]. Among them, [2Fe-2S]-type Fds have specific names referring to the metabolic function of each; Plant-type Fd of photosynthetic electron transport chain (1), Putidaredoxin in the oxidation pathway of camphor by cytochrome P450-CAM (2), the Rieske iron sulphur protein of reciprocal chain (3), and Adrenodoxin (adrenal Fd) in the metabolic pathway of the adrenal redox cascade (4). Each Fd has a [2Fe-2S] cluster with four particular coordinating amino acids and a characteristic redox potential. The plant-type Fd has a [2Fe-2S] cluster ligated by the highly conserved four cysteine residues. At around −400 mV the redox potential of plant-type Fd is extremely low. In chloroplasts, it accepts an electron from Photosystem I (PSI) reaction centre on the stromal side of the thylakoid membrane and delivers it to various Fd-dependent enzymes that reside in the stromal space of chloroplast (1). An electron transfer to or from Fd is mainly driven by the gradient of redox potentials between Fd and target proteins, but also controlled by its specific protein–protein interactions. Reduced and oxidized Fd showed a significant redox-dependent change of affinity to their respective target proteins. For example, oxidized Fd is preferentially bound to PSI and easily dissociates from PSI after reduction by PSI (5). Thus, structural changes upon reduction of Fd must be a trigger for dissociation from PSI. X-ray structures of oxidized and partly reduced Fds from cyanobacterium Anabaena PCC7119 (PDB IDs: 1QT9 and 1CZP) have been reported at 1.30 and 1.17 Å resolution, respectively (6). Structural comparison between them clearly showed a rearrangement of the NH-S hydrogen bond to the [2Fe-2S] cluster from the main chain amide hydrogen, accompanying a peptide bond flippage between Cys46 and Ser47. This redox-dependent structural change is now widely accepted also based on the further computational calculations (7, 8). The hydrogen bond network, including the NH-S hydrogen bond mentioned above, around the [2Fe-2S] cluster is believed to be important for tuning its redox potential. Several high-resolution X-ray structures of the plant-type Fd have been further published to visualize the hydrogen bond network around the [2Fe-2S] cluster; Fd from Mastigocladus laminosus at 1.25 Å resolution (PDB ID: 1RFK) (9), Fd from red algae Cyanidioschyzon melorae at 0.97 Å (3WCQ) (10), Fd2 from green algae Chlamydomonas reinhardtii at 1.18 Å (4ITK) (11) and Fd2 from Equisetum arvense at 1.2 Å (1WRI) (12). Notwithstanding, the positions of hydrogen atoms near the [2Fe-2S] cluster were not clearly visualized. All of the above described high-resolution structural studies of Fd relied on the use of high-dose X-ray from the modern synchrotron beamline for data collection. Crystallographers have already known that high-dose X-ray can reduce a native disulphide bond or a redox centre of the metalloproteins even in the frozen crystals (13). Therefore, the reported high-resolution data of oxidized Fd collected from a single crystal may be a mixture of oxidized and reduced form, which could hinder the precise structural analysis of the hydrogen atoms near the cluster. Then, it is of importance to know whether the crystals of oxidized Fd is or is not reduced during the X-ray diffraction experiment. In this study, we performed conventional X-ray structural analysis of Fd1 from C. reinhardtii (CrFd1) at 0.90 Å and analysed the anisotropic temperature factors that may be related to the [2Fe-2S] cluster which is partially reduced by X-ray irradiation. Then, we monitored absorption spectra from CrFd1 crystals during X-ray data collection, and found that X-ray dose-dependent spectral changes of Fd do indeed occur. Furthermore, we collected two independent diffraction datasets from one single crystal at 1.4 Å resolution using either low or high X-ray doses by sorting and merging many X-ray diffraction images derived from a variety of crystal orientations. Here, we discuss the observed initial structural changes between these two datasets of the [2Fe-2S] cluster upon reduction by X-ray irradiation in terms of the difference Fourier technique. Materials and Methods Preparation of recombinant Fd1 from C. reinhardtii We used a codon-optimized synthetic DNA of Fd1 from C. reinhardtii for protein expression in Escherichia coli. The synthesized gene lacking the N-terminal transit peptide and adding the N-terminal Met-Ala- was inserted into the pIDTSMART vector. CrFd1 encoding sequences were amplified using primers CRPETF_FW (5′-GCCCATGGCATACAAGGTCACC-3′) and CRPETF_RE (5′-GCGGATCCTTAGTACAGGGCC-3′). The amplified DNA fragment was cloned into the NcoI and BamHI restriction site on a pTrc99A vector. Recombinant CrFd1 was expressed in E. coli and purified as described previously (1). Crystallization of CrFd1 Crystallization was performed by the batch method. 1 µl of CrFd1 solution (36.6 mg/ml) dissolved in the buffer (150 mM NaCl and 50 mM Tris-HCl pH 7.5) and 2 µl of reservoir solution (2.77 M (NH4)2SO4, 3.3% benzamidine hydrochloride, 99 mM sodium citrate pH 5.6, 0.21 M NaSCN) was mixed on an MRC crystallization plate (Molecular dimensions) and covered with 30 µl of fluorinert. This crystallization drop was placed at 4°C for 2 days, during which red rod-shaped crystals appeared. Since the colour of the obtained rod-shaped Fd crystals was too deep, we explored better crystallization conditions to obtain crystals sufficiently thin for micro-spectrophotometry using the micro-seeding technique. The seed crystal solution was created by crushing rod-shaped crystals with Seed BeadTM (Hampton research) in a seed stabilization solution (3.20 M (NH4)2SO4, 2.0% benzamidine hydrochloride, 67 mM sodium citrate pH 5.6, 23 mM NaSCN, 50 mM NaCl, 17 mM Tris-DCl pH 7.5, solvent: D2O). Crystals thin enough for micro-spectrophotometry were obtained as follows: 1 µl of CrFd1 (70 mg/ml), 2 µl of reservoir solution (2.40 M (NH4)2SO4, 3.0% benzamidine hydrochloride, 100 mM sodium citrate pH 5.6, 35 mM NaSCN, solvent: D2O) and 0.6 µl of seed crystal solution were mixed on an MRC plate and covered with 30 µl of paraffin oil. This crystallization drop was placed at 20°C for 2 days during which plate-like crystals with a variety of thickness appeared. We chose the thinnest one (∼100 × 100 × 2 µm3) and substituted the solvent around the crystal with paraffin oil. Other cryo-protectants prohibited us from measuring clear absorption spectra. The thus obtained crystals were frozen by dipping them into liquid nitrogen. Crystals for X-ray diffraction experiment with low and high X-ray doses were obtained in the same way as thin crystal except for the concentration of the reservoir (2.20 M (NH4)2SO4, 2.0% benzamidine hydrochloride, 100 mM sodium citrate pH 5.6, 0.11 M NaSCN, solvent: D2O). A plate-shaped crystal (400 × 400 × 10 µm3) was soaked into cryoprotectant (37% D-Trehalose, 2.77 M (NH4)2SO4, 2.0% benzamidine hydrochloride, 67 mM sodium citrate pH 5.6, 23 mM NaSCN, 50 mM NaCl, 17 mM Tris-DCl pH 7.5, solvent: D2O) for 10 s and dipped into liquid nitrogen. Micro-spectrophotometry from a mounted CrFd1 crystal Absorption spectra of the CrFd1 crystal mounted on the cryo-loop was measured using an off-line Micro-spectrophotometer at SPring-8/RIKEN equipped with DT-Mini and SD2000 (Ocean Optics Inc.) as a light source and a detector, respectively. Spectral changes induced by X-ray dose in megagrays (MGy) or kilograys (kGy) level were observed from independent crystals using different beamlines for each crystal. To detect spectral changes induced by X-ray dose in MGy level, X-ray beam with a wavelength of 0.9000 Å from BL44XU at SPring-8 was used for X-ray irradiation to the mounted crystals. The beam was casted by a 70 µm pinhole. For this beam size, the flux of the beam was 8.43 × 1012 photons s−1. The full width at half maximum (FWHM) was ∼68.9 µm for horizontal and ∼60.3 µm for vertical. Micro-spectrophotometry was taken after a total of 0, 1 and 10 s X-ray irradiation. To detect spectral changes induced by X-ray dose in kGy level, X-ray beam with a wavelength of 1.0000 Å from BL26B2 at SPring-8 (14, 15) was used for X-ray irradiation. The beam was casted by a 100 µm pinhole. For this beam size, the flux of the beam was 3.84 × 1010 photons s−1. The FWHM was ∼86 µm for horizontal and ∼93 µm for vertical. Micro-spectrophotometry was taken after a total of 0, 1, 2, 3, 4, 5, 6, 8, 10, 15, 20, 30, 40 and 60 s of X-ray irradiation. The dose of X-ray absorbed by the crystals was calculated with RADDOSE-3D (16). Since the change of absorption spectrum from a mounted crystal depends on the absorbed X-ray dose, the size of light from the optic fibre was used for calculation with RADDOSE-3D. Because RADDOSE-3D didn’t support circle beam, the beam shape was approximated by a square which have same area as the cross-section of the light for dose estimation. X-ray data collection and data analysis Rod-shaped crystals were mounted in a nylon loop and flash-cooled by dipping them into liquid nitrogen. X-ray diffraction images for conventional analysis were collected at BL44XU, SPring-8 (Harima, Japan) under cryo-genic temperature. Diffraction images from crystals were collected by the oscillation method (Δψ = 0.5°) covering a total rotation angle of 180° at 100 K using an MX300HE (Ryonix) CCD detector. The diffraction images were integrated and merged using the XDS programme package (17). Resultant statistics of the data are summarized in Table I. The crystal structure was solved by molecular replacement using PHASER (18). The crystal structure of Fd from Chlorella fusca (PDB ID: 1AWD) was used as a search model. Refinement was performed at 0.90 Å using SHELXL (19) applying isotropic temperature factors initially and anisotropic temperature factors later to all atoms including water molecules. Table I. Crystallographic statistics of CrFd1 structures . Ultrahigh resolution . Low dose . High dose . Data collection  Light source SPring-8 BL44XU SPring-8 BL26B1 SPring-8 BL26B1  Wavelength/Å 0.9000 1.0000 1.0000  Space group P21 P1 P1  Lattice constants  a, b, c/Å 25.95, 60.09, 50.09 25.88, 32.33, 49.91 25.82, 32.30, 49.93  α, β, γ/degree 90, 102.40, 90 72.38, 78.06, 72.39 72.27, 78.09, 72.22  Resolution/Å 37.96–0.90 (0.92–0.90)a 28.74–1.40 (1.42–1.40)a 29.71–1.40 (1.42–1.40)a  Number of reflections 519,886 (25,485)a 105,463 (5,108)a 106,438 (5,348)a  Number of unique reflection 109,765 (7,995)a 27491 (1328)a 27,286 (1,368)a  Rmarge 6.1% (72.9%)a 9.4% (25.3%)a 8.2% (35.6%)a  I/σ(I) 12.9 (1.7)a 7.9 (3.8)a 8.1 (2.8)a  Completeness 99.2% (97.6%)a 95.7 % (93.0 %)a 95.4 % (92.9 %)a Refinement  Resolution/Å 37.96–0.90 28.75–1.40 (1.43–1.40)a 28.75–1.40 (1.43–1.40)a  Rwork 14.49% 17.13% (20.16%)a 17.51% (24.71%)a  Rfree 17.03% 18.97% (18.07%)a 18.74% (26.33%)a  RMS bond/Å 0.0192 0.011 0.008  RMS angle/degree 2.46 1.435 1.234 . Ultrahigh resolution . Low dose . High dose . Data collection  Light source SPring-8 BL44XU SPring-8 BL26B1 SPring-8 BL26B1  Wavelength/Å 0.9000 1.0000 1.0000  Space group P21 P1 P1  Lattice constants  a, b, c/Å 25.95, 60.09, 50.09 25.88, 32.33, 49.91 25.82, 32.30, 49.93  α, β, γ/degree 90, 102.40, 90 72.38, 78.06, 72.39 72.27, 78.09, 72.22  Resolution/Å 37.96–0.90 (0.92–0.90)a 28.74–1.40 (1.42–1.40)a 29.71–1.40 (1.42–1.40)a  Number of reflections 519,886 (25,485)a 105,463 (5,108)a 106,438 (5,348)a  Number of unique reflection 109,765 (7,995)a 27491 (1328)a 27,286 (1,368)a  Rmarge 6.1% (72.9%)a 9.4% (25.3%)a 8.2% (35.6%)a  I/σ(I) 12.9 (1.7)a 7.9 (3.8)a 8.1 (2.8)a  Completeness 99.2% (97.6%)a 95.7 % (93.0 %)a 95.4 % (92.9 %)a Refinement  Resolution/Å 37.96–0.90 28.75–1.40 (1.43–1.40)a 28.75–1.40 (1.43–1.40)a  Rwork 14.49% 17.13% (20.16%)a 17.51% (24.71%)a  Rfree 17.03% 18.97% (18.07%)a 18.74% (26.33%)a  RMS bond/Å 0.0192 0.011 0.008  RMS angle/degree 2.46 1.435 1.234 a Values in parentheses are for the highest resolution shells. Open in new tab Table I. Crystallographic statistics of CrFd1 structures . Ultrahigh resolution . Low dose . High dose . Data collection  Light source SPring-8 BL44XU SPring-8 BL26B1 SPring-8 BL26B1  Wavelength/Å 0.9000 1.0000 1.0000  Space group P21 P1 P1  Lattice constants  a, b, c/Å 25.95, 60.09, 50.09 25.88, 32.33, 49.91 25.82, 32.30, 49.93  α, β, γ/degree 90, 102.40, 90 72.38, 78.06, 72.39 72.27, 78.09, 72.22  Resolution/Å 37.96–0.90 (0.92–0.90)a 28.74–1.40 (1.42–1.40)a 29.71–1.40 (1.42–1.40)a  Number of reflections 519,886 (25,485)a 105,463 (5,108)a 106,438 (5,348)a  Number of unique reflection 109,765 (7,995)a 27491 (1328)a 27,286 (1,368)a  Rmarge 6.1% (72.9%)a 9.4% (25.3%)a 8.2% (35.6%)a  I/σ(I) 12.9 (1.7)a 7.9 (3.8)a 8.1 (2.8)a  Completeness 99.2% (97.6%)a 95.7 % (93.0 %)a 95.4 % (92.9 %)a Refinement  Resolution/Å 37.96–0.90 28.75–1.40 (1.43–1.40)a 28.75–1.40 (1.43–1.40)a  Rwork 14.49% 17.13% (20.16%)a 17.51% (24.71%)a  Rfree 17.03% 18.97% (18.07%)a 18.74% (26.33%)a  RMS bond/Å 0.0192 0.011 0.008  RMS angle/degree 2.46 1.435 1.234 . Ultrahigh resolution . Low dose . High dose . Data collection  Light source SPring-8 BL44XU SPring-8 BL26B1 SPring-8 BL26B1  Wavelength/Å 0.9000 1.0000 1.0000  Space group P21 P1 P1  Lattice constants  a, b, c/Å 25.95, 60.09, 50.09 25.88, 32.33, 49.91 25.82, 32.30, 49.93  α, β, γ/degree 90, 102.40, 90 72.38, 78.06, 72.39 72.27, 78.09, 72.22  Resolution/Å 37.96–0.90 (0.92–0.90)a 28.74–1.40 (1.42–1.40)a 29.71–1.40 (1.42–1.40)a  Number of reflections 519,886 (25,485)a 105,463 (5,108)a 106,438 (5,348)a  Number of unique reflection 109,765 (7,995)a 27491 (1328)a 27,286 (1,368)a  Rmarge 6.1% (72.9%)a 9.4% (25.3%)a 8.2% (35.6%)a  I/σ(I) 12.9 (1.7)a 7.9 (3.8)a 8.1 (2.8)a  Completeness 99.2% (97.6%)a 95.7 % (93.0 %)a 95.4 % (92.9 %)a Refinement  Resolution/Å 37.96–0.90 28.75–1.40 (1.43–1.40)a 28.75–1.40 (1.43–1.40)a  Rwork 14.49% 17.13% (20.16%)a 17.51% (24.71%)a  Rfree 17.03% 18.97% (18.07%)a 18.74% (26.33%)a  RMS bond/Å 0.0192 0.011 0.008  RMS angle/degree 2.46 1.435 1.234 a Values in parentheses are for the highest resolution shells. Open in new tab X-ray diffraction images for low and high X-ray dose datasets were collected at BL26B1 of SPring-8 (14, 15) operated in the top-up mode under 100 K with a flux of 8.92 × 109 photons s−1 at a wavelength of 1.0000 Å. The beam was casted by a 50 µm pinhole. For this beam size, the FWHM was ∼33 µm for the horizontal and ∼38 µm for the vertical. The CCD-based X-ray detector, MX225HE (Rayonix), was used for collecting the diffraction images. The images obtained from one crystal covered a 360° rotation with an oscillation of 0.5° and 100 ms exposure per flame as shown in Fig. 1. We assigned nine exposure points in one crystal without any overlaps, and first eight and last one of nine points was used for low- and high-dose datasets, respectively. Fig. 1 Open in new tabDownload slide X-ray data collection strategy for low- and high-dose datasets. Nine data collection points separated enough were assigned. First eight were used starting from different angles of 0, 45, 90, 135, 180, 225, 270 and 315 degrees. The final point was used for high-dose X-ray dataset after 10 min exposure of X-ray. Fig. 1 Open in new tabDownload slide X-ray data collection strategy for low- and high-dose datasets. Nine data collection points separated enough were assigned. First eight were used starting from different angles of 0, 45, 90, 135, 180, 225, 270 and 315 degrees. The final point was used for high-dose X-ray dataset after 10 min exposure of X-ray. Low-dose data were formed by eight non-overlapping angular shells from one crystal. X-ray data collection started from one end of the crystal. A total of 90 images covering 45° rotation angles were collected from one assigned point. Subsequently, the measurement point was moved 60 µm towards the other end and the next 90 images were collected from the fresh part of the crystal. This crystal movement and diffraction experiment was repeated until all angular range was covered by eight assigned exposure points. High-dose data were collected as follows using the remaining exposure point of a same crystal above, and the parameter of X-ray beam are same as the case of low-dose data collection. At first, X-ray was focussed on the last exposure point which was not used for low-dose data collection and irradiated for 10 min rotating the crystal in 360°. Subsequently, a full dataset covering a rotation of 360° was collected at the same position. All diffraction data for low and high X-ray doses were processed with MOSFILM (20) and merged with AIMLESS (21). To obtain an initial set of phases for the low- and high-dose dataset using a long plate-shaped crystal, molecular replacement was performed in PHASER (18). A crystal structure of CrFd1 as mentioned above (PDB ID: 6LK1) was used as a search model. Iterative refinement with XYZ coordinates, real space refinement, isotropic B-factor refinement, TLS refinement and manual model correction was performed starting from the initial model using phenix.refine (22) and COOT (23). Crystallographic restraints file for the [2Fe-2S] cluster was removed in the final stage of refinement calculation and the B-factors of the [2Fe-2S] cluster atoms were refined anisotropically. The Fo (low dose) – Fo (high dose) maps were generated with FFT (24) in the CCP4 programme package (21) using phase angles calculated from the final refined structure against the low-dose dataset to confirm the structural difference between two datasets. X-ray dose absorbed by crystals was estimated with RADDOSE-3D (16). The beam shape was approximated by a square which have same area as the cross-section of the light for dose estimation. Results Recombinant protein preparation and crystallization of CrFd1 Recombinant CrFd1 protein was successfully produced using a synthetic gene with altered codon usages optimized for expression in E. coli. The purified CrFd1 protein solution showed typical spectra of [2Fe–2S] Fd, and migrated in SDS-PAGE as homogeneous bands. Rod- and plate-shaped crystals used for X-ray and spectral data collections were reproducibly obtained by optimized crystallization conditions. Conventional crystal structure analysis of CrFd1 at 0.90 Å resolution A crystal structure of CrFd1, whose diffraction data were collected conventionally from a rod-shaped crystal, was determined by molecular replacement method, and refined with anisotropic temperature factors of all atoms at 0.90 Å (Fig. 2A). The refinement statistics are shown in Table I. The working and free R-factors were reasonably low enough to discuss about the bond lengths, bond angles and the anisotropy of the temperature factors. There were no bond length outliers in the model, and the ellipsoidal representation of the [2Fe-2S] cluster showed the orientational thermal vibration of the vicinity of the cluster (Fig. 2B). Fig. 2 Open in new tabDownload slide X-ray structure of CrFd1 determined by conventional X-ray data collection strategy at 0.90 Å. (A) Overall structure of CrFd1. (B) Zooming up view of the [2Fe-2S] cluster region shown in an ellipsoidal ball-and-stick model viewed from two different viewing angles. Fig. 2 Open in new tabDownload slide X-ray structure of CrFd1 determined by conventional X-ray data collection strategy at 0.90 Å. (A) Overall structure of CrFd1. (B) Zooming up view of the [2Fe-2S] cluster region shown in an ellipsoidal ball-and-stick model viewed from two different viewing angles. Micro-spectrophotometry of the thin plate-shaped CrFd1 crystal Unlike deep coloured rod-shaped crystals, plate-shaped crystals were thin enough to measure absorption spectra. Absorption spectra from a plate-shaped CrFd1 crystal showed absorption peaks at 420 and 470 nm as similar as that from a CrFd1 solution (Fig. 3A and B). In the spectra obtained from CrFd1 crystal, the peak height of 420 nm was lower than that of 470 nm; however, in solution the peak height of 420 nm was higher than that of 470 nm. The 470 nm peak in crystal was lowered with increasing X-ray dose and almost disappeared at a dose of 30 MGy. For the whole ultraviolet region, the crystal’s absorption increased concomitant with X-ray dose. The X-ray dose-dependent decrease of the peaks at 420  and 470 nm was observed for a dose of more than 20 kGy (Fig. 3C). It is interesting to note that, for a dose of under 20 kGy at the beginning of the X-ray exposure, the two peaks at 420 and 470 nm increased and decreased three times for some inexplicable reason. Fig. 3 Open in new tabDownload slide Absorption spectra of the plate-shaped CrFd1 crystals by micro-spectrophotometry. (A) Absorption spectra from CrFd1 solution as a reference with/without 20 mM dithionite. Ultraviolet region (<400 nm) were not shown due to the high absorbance from added dithionite. (B) Absorption spectra of the CrFd1 crystal with X-ray irradiation of 0, 3 and 30 MGy. (C) Relationship between X-ray dose and the two absorption peak heights. Fig. 3 Open in new tabDownload slide Absorption spectra of the plate-shaped CrFd1 crystals by micro-spectrophotometry. (A) Absorption spectra from CrFd1 solution as a reference with/without 20 mM dithionite. Ultraviolet region (<400 nm) were not shown due to the high absorbance from added dithionite. (B) Absorption spectra of the CrFd1 crystal with X-ray irradiation of 0, 3 and 30 MGy. (C) Relationship between X-ray dose and the two absorption peak heights. Crystal structures of CrFd1 with low and high X-ray dose Crystal structures of CrFd1 with low and high dose were successfully determined using the molecular replacement method. For each dataset, two molecules, Chains A and B related by a non-crystallographic symmetry, were contained in the crystallographic asymmetric unit. The structures were refined until the crystallographic R-factors converged. The average of B-factors of all atoms constituting protein molecules/clusters were 12.39/8.90 and 13.69/9.26 Å2 for the structure with low- and high-dose X-ray, respectively. Although the absorption spectra corresponding to these two datasets were significantly different, no obvious structural change in the coordinates was detected between the two structures. The geometry of the cluster refined without any crystallographic restraints is shown in Table II. Concomitant with the increase of the X-ray dose, the bond lengths between iron and sulphite atoms of the cluster were extended on the whole (0.019 Å on average). An exception is the shortened length between Fe2 and S1 atoms which was common to the Non-crystallographic symmetry (NCS)-related two molecules. Being consistent with the coordinates changed in the [2Fe-2S] cluster upon reduction by X-ray exposure, the calculated Fo (low dose) – Fo (high dose) difference Fourier map showed the positive and negative peaks around the [2Fe-2S] atoms (Fig. 4). Fig. 4 Open in new tabDownload slide Difference Fourier map calculated with low- and high-dose X-ray datasets. Fo (low dose) – Fo (high dose) positive map ( mesh, 5.5 σ) and negative map ( mesh, 5.5 σ) around the [2Fe-2S] cluster, Gln89 and Cys16. Fig. 4 Open in new tabDownload slide Difference Fourier map calculated with low- and high-dose X-ray datasets. Fo (low dose) – Fo (high dose) positive map ( mesh, 5.5 σ) and negative map ( mesh, 5.5 σ) around the [2Fe-2S] cluster, Gln89 and Cys16. Table II. The geometry of the [2Fe-2S] cluster in low- and high-dose dataset of CrFd1, oxidized and partially reduced Fd crystal structures from Anabaena PCC7119 . This work . PCC7119 [2Fe-2S] Fd . . Low dose . High dose . Oxidized . Reduced . . mol−1 . mol−2 . mol−1 . mol−2 . mol−1 . mol−2 . Bonds (Å)  Fe1-S1 2.21 2.22 2.24 2.24 2.746 2.749 2.747  -S2 2.14 2.17 2.18 2.19 2.278 2 .293 2.285 vFe2-S1 2.19 2.19 2.18 2.18 2.227 2.235 2.218   -S2 2.14 2.14 2.18 2.16 2.184 2.178 2.197 Angles (deg)  S1-Fe1-S2 102.2 102.4 101.5 101.8 102.2 101.8 102.2  Fe1-S1-Fe2 76.2 75.5 76.7 76.0 75.1 74.7 75.2  Fe1-S2-Fe2 78.6 77.3 78.0 77.4 76.8 76.5 76.6  S1-Fe2-S2 102.8 104.3 103.4 104.4 105.6 106.4 105.5 . This work . PCC7119 [2Fe-2S] Fd . . Low dose . High dose . Oxidized . Reduced . . mol−1 . mol−2 . mol−1 . mol−2 . mol−1 . mol−2 . Bonds (Å)  Fe1-S1 2.21 2.22 2.24 2.24 2.746 2.749 2.747  -S2 2.14 2.17 2.18 2.19 2.278 2 .293 2.285 vFe2-S1 2.19 2.19 2.18 2.18 2.227 2.235 2.218   -S2 2.14 2.14 2.18 2.16 2.184 2.178 2.197 Angles (deg)  S1-Fe1-S2 102.2 102.4 101.5 101.8 102.2 101.8 102.2  Fe1-S1-Fe2 76.2 75.5 76.7 76.0 75.1 74.7 75.2  Fe1-S2-Fe2 78.6 77.3 78.0 77.4 76.8 76.5 76.6  S1-Fe2-S2 102.8 104.3 103.4 104.4 105.6 106.4 105.5 Open in new tab Table II. The geometry of the [2Fe-2S] cluster in low- and high-dose dataset of CrFd1, oxidized and partially reduced Fd crystal structures from Anabaena PCC7119 . This work . PCC7119 [2Fe-2S] Fd . . Low dose . High dose . Oxidized . Reduced . . mol−1 . mol−2 . mol−1 . mol−2 . mol−1 . mol−2 . Bonds (Å)  Fe1-S1 2.21 2.22 2.24 2.24 2.746 2.749 2.747  -S2 2.14 2.17 2.18 2.19 2.278 2 .293 2.285 vFe2-S1 2.19 2.19 2.18 2.18 2.227 2.235 2.218   -S2 2.14 2.14 2.18 2.16 2.184 2.178 2.197 Angles (deg)  S1-Fe1-S2 102.2 102.4 101.5 101.8 102.2 101.8 102.2  Fe1-S1-Fe2 76.2 75.5 76.7 76.0 75.1 74.7 75.2  Fe1-S2-Fe2 78.6 77.3 78.0 77.4 76.8 76.5 76.6  S1-Fe2-S2 102.8 104.3 103.4 104.4 105.6 106.4 105.5 . This work . PCC7119 [2Fe-2S] Fd . . Low dose . High dose . Oxidized . Reduced . . mol−1 . mol−2 . mol−1 . mol−2 . mol−1 . mol−2 . Bonds (Å)  Fe1-S1 2.21 2.22 2.24 2.24 2.746 2.749 2.747  -S2 2.14 2.17 2.18 2.19 2.278 2 .293 2.285 vFe2-S1 2.19 2.19 2.18 2.18 2.227 2.235 2.218   -S2 2.14 2.14 2.18 2.16 2.184 2.178 2.197 Angles (deg)  S1-Fe1-S2 102.2 102.4 101.5 101.8 102.2 101.8 102.2  Fe1-S1-Fe2 76.2 75.5 76.7 76.0 75.1 74.7 75.2  Fe1-S2-Fe2 78.6 77.3 78.0 77.4 76.8 76.5 76.6  S1-Fe2-S2 102.8 104.3 103.4 104.4 105.6 106.4 105.5 Open in new tab Discussion Crystallographic analysis of the conventionally obtained X-ray data from the rod-shaped CrFd1 crystal exhibited the ellipsoidal temperature factors, as shown in Fig. 2B. Based on the obtained coordinates without any spectral measurement upon X-ray irradiation, it was difficult to say whether these ellipsoidal temperature factors originated from innate feature or were due to the partial reduction by X-ray even at the very high resolution of 0.90 Å. Regrettably, the original rod-shaped crystals showed a too strong deep red colour to measure their absorption spectra. However, the plate-shaped crystals obtained from new crystallization condition optimized for micro-spectrophotometry were thin enough to measure their spectra. The obtained absorption spectrum from this plate-shaped crystal before X-ray experiment was almost same as that from Fd solution, except for the relative peak heights of 420 and 470 nm specific for the [2Fe-2S] cluster. This is probably because the chemical environment around the Fd molecule in the crystalline state is slightly different from that of Fd in solution. Hence, we assume that the absorption spectrum from a Fd crystal before the X-ray exposure is that of fully oxidized Fd. After the X-ray irradiation to the mounted Fd crystals, the absorption spectrum of CrFd1 has obviously changed as shown in Fig. 3B. The absorption peaks at 420 and 470 nm characteristic for oxidized Fd were lowered by the increase of X-ray dose. This result indicates that Fd was reduced by X-ray irradiation even under the cryogenic condition. To estimate the rate of oxidize/reduced forms, it is better to refer to the reduction of the peak height at 470 nm only, because the increased absorption in whole ultraviolet region also affects the peak height at 420 nm. Based on this, roughly two-thirds of Fd were reduced by a dose of 3 MGy X-ray and mostly reduced by a dose of 30 MGy. Decrease of these peaks were observed only after 20 kGy dose and the absorption of the <20 kGy dose region shows unexpected behaviour as mentioned in the result section. For the moment, we do not know why the two peak heights fluctuated at the beginning of the X-ray experiment. This may need more accurate measurements using detailed online micro-spectrophotometry. To identify the structural change upon reduction of the [2Fe-2S] cluster by X-ray, we compared the two coordinates refined against the low- and high-dose datasets. There was no significant displacement of the atomic coordinates in the [2Fe-2S] cluster (0.019 Å on average), although the overall B-factors increased slightly as previously reported (25). Focussing on the [2Fe-2S] cluster, the bond length became slightly longer. However, the resolution of two structures with low and high X-ray dose was not high enough (1.4 Å) to discuss about it because the estimated coordination error is big (26). To visualize the structural change upon reduction by X-ray, we calculated the difference Fourier map between low- and high-dose datasets, which is independent to the atomic coordinates in real space but only related to the structure factors in reciprocal space. Both datasets were taken from the same crystal, and the resultant crystallographic statistics on data processing were similar enough to be compared. Most of the peaks in the difference Fourier map were located around the cluster and remarkably strong. This means that the cluster was influenced strongly upon reduction by X-ray compared with other region of the protein. Additionally, plus and minus peaks of the map were located at the opposite positions across every iron atoms of the cluster, which implies that the iron atoms were repositioned upon reduction by X-ray irradiation. Since the directions of the movement of irons were common to the two iron atoms in each cluster, we assumed that the cluster was slightly moved towards one direction upon reduction. The mechanism of this small but significant translation can be explained as follow. By the one electron reduction of the cluster, the interactions between the cluster and the peptide chain starts being rearranged. First, the reduced cluster moves to the chemically most stable position, whose directions of the movement as implicated from the Fo–Fo difference maps were similar but not exactly same in the NCS-related molecules. Next the structure of peptide region would change sequentially to accommodate the newly positioned [2Fe-2S] cluster. In this report, the relationship between X-ray dose and the absorption spectra change of CrFd1 crystal is shown for the first time. By comparison of two diffraction datasets with low and high X-ray dose, small displacements of the ion atoms in the [2Fe-2S] cluster were visualized. These new findings suggest the structural basis for the initial reduction step of the [2Fe-2S] cluster of Fd that triggers the redox-dependent structural change of Fd. A one electron can change the iron positions in the [2Fe-2S] cluster first as shown in this study, and it must be amplified and transmitted through the protein region eventually to tune the protein–protein interactions between Fd and Fd-dependent enzymes. The crystallographic and spectroscopic studies presented here can also be the basis for a strategic X-ray diffraction experiment of Fd crystals at high resolution that aim at the reduction of X-ray radiation damage. Atomic coordinates The atomic coordinates and structure factors (Accession code 6LK1, 6KUM and 6KV0 for the structure from conventional X-ray data, data with low and high X-ray dose, respectively) have been deposited to the Protein Data Bank Japan, a member of the worldwide PDB (https://www.pdbj.org). Acknowledgements We thank Dr Christoph Gerle for scientific suggestions and proof reading of the article and Ms Arisa Sato for helping preparation of CrFd1 crystals. We are grateful to the staff at beamlines BL26B1 and BL44XU at SPring-8, Japan, for their kind help during the data collection under the proposal numbers of 2014B6500, 2017A6500, 2018A2698 and 2019A2899. Funding This work was supported by the grants-in-aid from the Ministry of Culture, Education, Science and Sports of Japan (16H06560). Conflict of Interest None declared. References 1 Mutoh R. , Muraki N., Shinmura K., Kubota-Kawai H., Lee Y.-H., Nowaczyk M.M., Rögner M., Hase T., Ikegami T., Kurisu G. ( 2015 ) X-ray structure and nuclear magnetic resonance analysis of the interaction sites of the ga-substituted cyanobacterial ferredoxin . Biochemistry 54 , 6052 – 6061 Google Scholar Crossref Search ADS PubMed WorldCat 2 Guengerich F.P. ( 2001 ) Common and uncommon cytochrome P450 reactions related to metabolism and chemical toxicity . Chem. Res. Toxicol. 14 , 611 – 650 Google Scholar Crossref Search ADS PubMed WorldCat 3 Kolling D.J. , Brunzelle J.S., Lhee S.M., Crofts A.R., Nair S.K. ( 2007 ) Atomic resolution structures of rieske iron-sulfur protein: role of hydrogen bonds in tuning the redox potential of iron-sulfur clusters . Structure 15 , 29 – 38 Google Scholar Crossref Search ADS PubMed WorldCat 4 Pikuleva I.A. , Tesh K., Waterman M.R., Kim Y. ( 2000 ) The tertiary structure of full-length bovine adrenodoxin suggests functional dimers . Arch. Biochem. Biophys . 373 , 44 – 55 Google Scholar Crossref Search ADS PubMed WorldCat 5 Sétif P. , Mutoh R., Kurisu G. ( 2017 ) Dynamics and energetics of cyanobacterial photosystem I: ferredoxin complexes in different redox states . Biochim. Biophys. Acta Bioenerg . 1858 , 483 – 496 Google Scholar Crossref Search ADS PubMed WorldCat 6 Morales R. , Charon M.-H., Hudry-Clergeon G., Pétillot Y., Norager S., Medina M., Frey M. ( 1999 ) Refined X-ray structures of the oxidized, at 1.3 Å, and reduced, at 1.17 Å, [2Fe-2S] ferredoxin from the cyanobacterium Anabaena PCC7119 show redox-linked conformational changes . Biochemistry 38 , 15764 – 15773 Google Scholar Crossref Search ADS PubMed WorldCat 7 Morales R. , Frey M., Mouesca J.-M. ( 2002 ) An Approach based on quantum chemistry calculations and structural analysis of a [2Fe-2S*] ferredoxin that reveal a redox-linked switch in the electron-transfer process to the Fd-NADP + reductase . J. Am. Chem. Soc. 124 , 6714 – 6722 Google Scholar Crossref Search ADS PubMed WorldCat 8 Pizzitutti F. , Sétif P., Marchi M. ( 2003 ) Theoretical investigation of the “CO in”−“CO out” isomerization in a [2Fe−2S] ferredoxin: free energy profiles and redox states . J. Am. Chem. Soc. 125 , 15224 – 15232 Google Scholar Crossref Search ADS PubMed WorldCat 9 Fish A. , Danieli T., Ohad I., Nechushtai R., Livnah O. ( 2005 ) Structural basis for the thermostability of ferredoxin from the cyanobacterium Mastigocladus laminosus . J. Mol. Biol . 350 , 599 – 608 Google Scholar Crossref Search ADS PubMed WorldCat 10 Ueno Y. , Matsumoto T., Yamano A., Imai T., Morimoto Y. ( 2013 ) Increasing the electron-transfer ability of C. merolae ferredoxin by a one-point mutation - a high resolution and Fe-SAD phasing crystal structure analysis of the Asp58Asn mutant . Biochem. Biophys. Res. Commun . 436 , 736 – 739 Google Scholar Crossref Search ADS PubMed WorldCat 11 Boehm M. , Alahuhta M., Mulder D.W., Peden E.A., Long H., Brunecky R., Lunin V.V., King P.W., Ghirardi M.L., Dubini A. ( 2016 ) Crystal structure and biochemical characterization of Chlamydomonas FDX2 reveal two residues that, when mutated, partially confer FDX2 the redox potential and catalytic properties of FDX1 . Photosynth. Res . 128 , 45 – 57 Google Scholar Crossref Search ADS PubMed WorldCat 12 Kurisu G. , Nishiyama D., Kusunoki M., Fujikawa S., Katoh M., Hanke G.T., Hase T., Teshima K. ( 2005 ) A structural basis of Equisetum arvense ferredoxin isoform II producing an alternative electron transfer with ferredoxin-NADP+ reductase . J. Biol. Chem. 280 , 2275 – 2281 Google Scholar Crossref Search ADS PubMed WorldCat 13 Garman E.F. ( 2010 ) Radiation damage in macromolecular crystallography: what is it and why should we care? Acta Crystallogr. D Biol. Crystallogr . 66 , 339 – 351 Google Scholar Crossref Search ADS PubMed WorldCat 14 Ueno G. , Kanda H., Hirose R., Ida K., Kumasaka T., Yamamoto M. ( 2006 ) RIKEN structural genomics beamlines at the SPring-8; high throughput protein crystallography with automated beamline operation . J. Struct. Funct. Genomics 7 , 15 – 22 Google Scholar Crossref Search ADS PubMed WorldCat 15 Ueno G. , Kanda H., Kumasaka T., Yamamoto M. ( 2005 ) Beamline Scheduling Software: administration software for automatic operation of the RIKEN structural genomics beamlines at SPring-8 . J. Synchrotron Radiat . 12 , 380 – 384 Google Scholar Crossref Search ADS PubMed WorldCat 16 Zeldin O.B. , Gerstel M., Garman E.F ( 2013 ) RADDOSE-3D: time- and space-resolved modelling of dose in macromolecular crystallography . J. Appl. Crystallogr . 46 , 1225 – 1230 Google Scholar Crossref Search ADS WorldCat 17 Kabsch W. ( 2010 ) XDS . Acta Crystallogr. D Biol. Crystallogr. . 66 , 125 – 132 Google Scholar Crossref Search ADS PubMed WorldCat 18 McCoy A.J. , Grosse-Kunstleve R.W., Adams P.D., Winn M.D., Storoni L.C., Read R.J. ( 2007 ) Phaser crystallographic software . J. Appl. Crystallogr . 40 , 658 – 674 Google Scholar Crossref Search ADS PubMed WorldCat 19 Sheldrick G.M. ( 2015 ) Crystal structure refinement with SHELXL . Acta Crystallogr. C Struct. Chem . 71 , 3 – 8 Google Scholar Crossref Search ADS PubMed WorldCat 20 Battye T.G.G. , Kontogiannis L., Johnson O., Powell H.R., Leslie A.G.W. ( 2011 ) iMOSFLM: a new graphical interface for diffraction-image processing with MOSFLM . Acta Crystallogr. D Biol. Crystallogr . 67 , 271 – 281 Google Scholar Crossref Search ADS PubMed WorldCat 21 Winn M.D. , Ballard C.C., Cowtan K.D., Dodson E.J., Emsley P., Evans P.R., Keegan R.M., Krissinel E.B., Leslie A.G.W., McCoy A., McNicholas S.J., Murshudov G.N., Pannu N.S., Potterton E.A., Powell H.R., Read R.J., Vagin A., Wilson K.S. ( 2011 ) Overview of the CCP 4 suite and current developments . Acta Crystallogr. D Biol. Crystallogr . 67 , 235 – 242 Google Scholar Crossref Search ADS PubMed WorldCat 22 Afonine P.V. , Grosse-Kunstleve R.W., Echols N., Headd J.J., Moriarty N.W., Mustyakimov M., Terwilliger T.C., Urzhumtsev A., Zwart P.H., Adams P.D. ( 2012 ) Towards automated crystallographic structure refinement with phenix.refine . Acta Crystallogr. D Biol. Crystallogr . 68 , 352 – 367 Google Scholar Crossref Search ADS PubMed WorldCat 23 Emsley P. , Lohkamp B., Scott W.G., Cowtan K. ( 2010 ) Features and development of Coot . Acta Crystallogr. D Biol. Crystallogr . 66 , 486 – 501 Google Scholar Crossref Search ADS PubMed WorldCat 24 Ten Eyck L.F. ( 1973 ) Crystallographic fast Fourier transforms . Acta Cryst. A . 29 , 183 – 191 Google Scholar Crossref Search ADS WorldCat 25 Shimizu N. , Hirata K., Hasegawa K., Ueno G., Yamamoto M. ( 2007 ) Dose dependence of radiation damage for protein crystals studied at various X-ray energies . J. Synchrotron Radiat . 14 , 4 – 10 Google Scholar Crossref Search ADS PubMed WorldCat 26 Luzzati V. ( 1952 ) Traitement statistique des erreurs dans la determination des structures cristallines . Acta Cryst . 5 , 802 – 810 Google Scholar Crossref Search ADS WorldCat Abbreviations Abbreviations Fd ferredoxin FWHM full width at half maximum PSI Photosystem I PDB Protein Data Bank SPring-8 Super Photon ring-8 GeV. © The Author(s) 2020. Published by Oxford University Press on behalf of the Japanese Biochemical Society. All rights reserved This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)
TI - X-ray dose-dependent structural changes of the [2Fe-2S] ferredoxin from Chlamydomonas reinhardtii
JF - The Journal of Biochemistry
DO - 10.1093/jb/mvaa045
DA - 2020-06-01
UR - https://www.deepdyve.com/lp/oxford-university-press/x-ray-dose-dependent-structural-changes-of-the-2fe-2s-ferredoxin-from-3DIkO72g0e
SP - 549
VL - 167
IS - 6
DP - DeepDyve
ER -