Similarity score for screening phase-retrieved maps in X-ray diffraction imaging - characterization in reciprocal space.

X-ray diffraction imaging (XDI) is utilized for visualizing the structures of non-crystalline particles in material sciences and biology. In the structural analysis, phase-retrieval (PR) algorithms are applied to the diffraction amplitude data alone to reconstruct the electron density map of a specimen particle projected along the direction of the incident X-rays. However, PR calculations may not lead to good convergence because of a lack of diffraction patterns in small-angle regions and Poisson noise in X-ray detection. Therefore, the PR calculation is still a bottleneck for the efficient application of XDI in the structural analyses of non-crystalline particles. For screening maps from hundreds of trial PR calculations, we have been using a score and measuring the similarity between a pair of retrieved maps. Empirically, probable maps approximating the particle structures gave a score smaller than a threshold value, but the reasons for the effectiveness of the score are still unclear. In this study, the score is characterized in terms of the phase differences between the structure factors of the retrieved maps, the usefulness of the score in screening the maps retrieved from experimental diffraction patterns is demonstrated, and the effective resolution of similarity-score-selected maps is discussed.

Protocol using similarity score and improved shrink-wrap algorithm for better convergence of phase-retrieval calculation in X-ray diffraction imaging.

In X-ray diffraction imaging (XDI), electron density maps of a targeted particle are reconstructed computationally from the diffraction pattern alone using phase-retrieval (PR) algorithms. However, the PR calculations sometimes fail to yield realistic electron density maps that approximate the structure of the particle. This occurs due to the absence of structure amplitudes at and near the zero-scattering angle and the presence of Poisson noise in weak diffraction patterns. Consequently, the PR calculation becomes a bottleneck for XDI structure analyses. Here, a protocol to efficiently yield realistic maps is proposed. The protocol is based on the empirical observation that realistic maps tend to yield low similarity scores, as suggested in our prior study [Sekiguchi et al. (2017), J. Synchrotron Rad. 24, 1024-1038]. Among independently and concurrently executed PR calculations, the protocol modifies all maps using the electron-density maps exhibiting low similarity scores. This approach, along with a new protocol for estimating particle shape, improved the probability of obtaining realistic maps for diffraction patterns from various aggregates of colloidal gold particles, as compared with PR calculations performed without the protocol. Consequently, the protocol has the potential to reduce computational costs in PR calculations and enable efficient XDI structure analysis of non-crystalline particles using synchrotron X-rays and X-ray free-electron laser pulses.

Domain Organization in Plant Blue-Light Receptor Phototropin2 of Arabidopsis thaliana Studied by Small-Angle X-ray Scattering.

Phototropin2 (phot2) is a blue-light (BL) receptor protein that regulates the BL-dependent activities of plants for efficient photosynthesis. Phot2 is composed of two light-oxygen-voltage sensing domains (LOV1 and LOV2) to absorb BL, and a kinase domain. Photo-activated LOV domains, especially LOV2, play a major role in photo-dependent increase in the phosphorylation activity of the kinase domain. The atomic details of the overall structure of phot2 and the intramolecular mechanism to convert BL energy to a phosphorylation signal remain unknown. We performed structural studies on the LOV fragments LOV1, LOV2, LOV2-linker, and LOV2-kinase, and full-length phot2, using small-angle X-ray scattering (SAXS). The aim of the study was to understand structural changes under BL irradiation and discuss the molecular mechanism that enhance the phosphorylation activity under BL. SAXS is a suitable technique for visualizing molecular structures of proteins in solution at low resolution and is advantageous for monitoring their structural changes in the presence of external physical and/or chemical stimuli. Structural parameters and molecular models of the recombinant specimens were obtained from SAXS profiles in the dark, under BL irradiation, and after dark reversion. LOV1, LOV2, and LOV2-linker fragments displayed minimal structural changes. However, BL-induced rearrangements of functional domains were noted for LOV2-kinase and full-length phot2. Based on the molecular model together with the absorption measurements and biochemical assays, we discuss the intramolecular interactions and domain motions necessary for BL-enhanced phosphorylation activity of phot2.

Blue light-excited LOV1 and LOV2 domains cooperatively regulate the kinase activity of full-length phototropin2 from Arabidopsis.

Phototropin2 (phot2) is a blue-light (BL) receptor that regulates BL-dependent activities for efficient photosynthesis in plants. phot2 comprises two BL-receiving light-oxygen-voltage-sensing domains (LOV1 and LOV2) and a kinase domain. BL-excited LOV2 is thought to be primarily responsible for the BL-dependent activation of the kinase. However, the molecular mechanisms by which small BL-induced conformational changes in the LOV2 domain are transmitted to the kinase remain unclear. Here, we used full-length wild-type and mutant phot2 proteins from Arabidopsis to study their molecular properties in the dark and under BL irradiation. Phosphorylation assays and absorption measurements indicated that the LOV1 domain assists the thermal relaxation of BL-excited LOV2 and vice versa. Using small-angle X-ray scattering and electron microscopy, we observed that phot2 forms a dimer and has a rod shape with a maximum length of 188 Å and a radius of gyration of 44 Å. Under BL, phot2 displayed large conformational changes that bent the rod shape. By superimposing the crystal structures of the LOV1 dimer, LOV2, and a homology model of the kinase to the observed changes, we inferred that the BL-dependent change consisted of positional shifts of both LOV2 and the kinase relative to LOV1. Furthermore, phot2 mutants lacking the photocycle in LOV1 or LOV2 still exhibited conformational changes under BL, suggesting that LOV1 and LOV2 cooperatively contribute to the conformational changes that activate the kinase. These results suggest that BL-activated LOV1 contributes to the kinase activity of phot2. We discuss the possible intramolecular interactions and signaling mechanisms in phot2.

Growth of Cuprous Oxide Particles in Liquid-Phase Synthesis Investigated by X-ray Laser Diffraction.

Cuprous oxide (Cu2O) particles obtained by surfactant-assisted liquid-phase synthesis have cuboid shapes but the internal structures are difficult to be visualized by electron microscopy. Herein, we investigated the internal structures of numerous individual Cu2O particles with submicrometer dimensions by X-ray diffraction imaging (XDI) using X-ray free-electron laser (XFEL) pulses. The reconstructed two-dimensional electron density maps, which displayed inhomogeneous internal structures, were divided into five classes characterized by the positions and shapes of high and low electron density areas. Further analysis of the maps in each class by a manifold learning algorithm revealed that the internal structures of Cu2O particles varied in correlation with total electron density while retaining the characteristics within each class. On the basis of the analyses, we proposed a growth mechanism to yield the inhomogeneity in the internal structures of Cu2O particles in surfactant-mediated liquid-phase synthesis.

Classification of ab initio models of proteins restored from small-angle X-ray scattering.

In structure analyses of proteins in solution by using small-angle X-ray scattering (SAXS), the molecular models are restored by using ab initio molecular modeling algorithms. There can be variation among restored models owing to the loss of phase information in the scattering profiles, averaging with regard to the orientation of proteins against the direction of the incident X-ray beam, and also conformational fluctuations. In many cases, a representative molecular model is obtained by averaging models restored in a number of ab initio calculations, which possibly provide nonrealistic models inconsistent with the biological and structural information about the target protein. Here, a protocol for classifying predicted models by multivariate analysis to select probable and realistic models is proposed. In the protocol, each structure model is represented as a point in a hyper-dimensional space describing the shape of the model. Principal component analysis followed by the clustering method is applied to visualize the distribution of the points in the hyper-dimensional space. Then, the classification provides an opportunity to exclude nonrealistic models. The feasibility of the protocol was examined through the application to the SAXS profiles of four proteins.

Diffraction apparatus and procedure in tomography X-ray diffraction imaging for biological cells at cryogenic temperature using synchrotron X-ray radiation.

X-ray diffraction imaging is a technique for visualizing the structure of biological cells. In X-ray diffraction imaging experiments using synchrotron radiation, cryogenic conditions are necessary in order to reduce radiation damage in the biological cells. Frozen-hydrated biological specimens kept at cryogenic temperatures are also free from drying and bubbling, which occurs in wet specimens under vacuum conditions. In a previous study, the diffraction apparatus KOTOBUKI-1 [Nakasako et al. (2013), Rev. Sci. Instrum. 84, 093705] was constructed for X-ray diffraction imaging at cryogenic temperatures by utilizing a cryogenic pot, which is a cooling device developed in low-temperature physics. In this study a new cryogenic pot, suitable for tomography experiments, has been developed. The pot can rotate a biological cell over an angular range of ±170° against the direction of the incident X-ray beam. Herein, the details and the performance of the pot and miscellaneous devices are reported, along with established experimental procedures including specimen preparation. The apparatus has been used in tomography experiments for visualizing the three-dimensional structure of a Cyanidioschyzon merolae cell with an approximate size of 5 µm at a resolution of 136 nm. Based on the experimental results, the necessary improvements for future experiments and the resolution limit achievable under experimental conditions within a maximum tolerable dose are discussed.

Blue Light-excited Light-Oxygen-Voltage-sensing Domain 2 (LOV2) Triggers a Rearrangement of the Kinase Domain to Induce Phosphorylation Activity in Arabidopsis Phototropin1.

Phototropin1 is a blue light (BL) receptor in plants and shows BL-dependent kinase activation. The BL-excited light-oxygen-voltage-sensing domain 2 (LOV2) is primarily responsible for the activation of the kinase domain; however, the molecular mechanism by which conformational changes in LOV2 are transmitted to the kinase domain remains unclear. Here, we investigated BL-induced structural changes of a minimum functional fragment of Arabidopsis phototropin1 composed of LOV2, the kinase domain, and a linker connecting the two domains using small-angle x-ray scattering (SAXS). The fragment existed as a dimer and displayed photoreversible SAXS changes reflected in the radii of gyration of 42.9 Å in the dark and 48.8 Å under BL irradiation. In the dark, the molecular shape reconstructed from the SAXS profiles appeared as two bean-shaped lobes in a twisted arrangement that was 170 Å long, 80 Å wide, and 50 Å thick. The molecular shape under BL became slightly elongated from that in the dark. By fitting the crystal structure of the LOV2 dimer and a homology model of the kinase domain to their inferred shapes, the BL-dependent change could be interpreted as the positional shift in the kinase domain relative to that of the LOV2 dimer. In addition, we found that lysine 475, a functionally important residue, in the N-terminal region of LOV2 plays a critical role in transmitting the structural changes in LOV2 to the kinase domain. The interface between the domains is critical for signaling, suitably changing the structure to activate the kinase in response to conformational changes in the adjoining LOV2.

A protocol for searching the most probable phase-retrieved maps in coherent X-ray diffraction imaging by exploiting the relationship between convergence of the retrieved phase and success of calculation.

Coherent X-ray diffraction imaging (CXDI) is a technique for visualizing the structures of non-crystalline particles with size in the submicrometer to micrometer range in material sciences and biology. In the structural analysis of CXDI, the electron density map of a specimen particle projected along the direction of the incident X-rays can be reconstructed only from the diffraction pattern by using phase-retrieval (PR) algorithms. However, in practice, the reconstruction, relying entirely on the computational procedure, sometimes fails because diffraction patterns miss the data in small-angle regions owing to the beam stop and saturation of the detector pixels, and are modified by Poisson noise in X-ray detection. To date, X-ray free-electron lasers have allowed us to collect a large number of diffraction patterns within a short period of time. Therefore, the reconstruction of correct electron density maps is the bottleneck for efficiently conducting structure analyses of non-crystalline particles. To automatically address the correctness of retrieved electron density maps, a data analysis protocol to extract the most probable electron density maps from a set of maps retrieved from 1000 different random seeds for a single diffraction pattern is proposed. Through monitoring the variations of the phase values during PR calculations, the tendency for the PR calculations to succeed when the retrieved phase sets converged on a certain value was found. On the other hand, if the phase set was in persistent variation, the PR calculation tended to fail to yield the correct electron density map. To quantify this tendency, here a figure of merit for the variation of the phase values during PR calculation is introduced. In addition, a PR protocol to evaluate the similarity between a map of the highest figure of merit and other independently reconstructed maps is proposed. The protocol is implemented and practically examined in the structure analyses for diffraction patterns from aggregates of gold colloidal particles. Furthermore, the feasibility of the protocol in the structure analysis of organelles from biological cells is examined.

Classification and assessment of retrieved electron density maps in coherent X-ray diffraction imaging using multivariate analysis.

Coherent X-ray diffraction imaging (CXDI) is one of the techniques used to visualize structures of non-crystalline particles of micrometer to submicrometer size from materials and biological science. In the structural analysis of CXDI, the electron density map of a sample particle can theoretically be reconstructed from a diffraction pattern by using phase-retrieval (PR) algorithms. However, in practice, the reconstruction is difficult because diffraction patterns are affected by Poisson noise and miss data in small-angle regions due to the beam stop and the saturation of detector pixels. In contrast to X-ray protein crystallography, in which the phases of diffracted waves are experimentally estimated, phase retrieval in CXDI relies entirely on the computational procedure driven by the PR algorithms. Thus, objective criteria and methods to assess the accuracy of retrieved electron density maps are necessary in addition to conventional parameters monitoring the convergence of PR calculations. Here, a data analysis scheme, named ASURA, is proposed which selects the most probable electron density maps from a set of maps retrieved from 1000 different random seeds for a diffraction pattern. Each electron density map composed of J pixels is expressed as a point in a J-dimensional space. Principal component analysis is applied to describe characteristics in the distribution of the maps in the J-dimensional space. When the distribution is characterized by a small number of principal components, the distribution is classified using the k-means clustering method. The classified maps are evaluated by several parameters to assess the quality of the maps. Using the proposed scheme, structure analysis of a diffraction pattern from a non-crystalline particle is conducted in two stages: estimation of the overall shape and determination of the fine structure inside the support shape. In each stage, the most accurate and probable density maps are objectively selected. The validity of the proposed scheme is examined by application to diffraction data that were obtained from an aggregate of metal particles and a biological specimen at the XFEL facility SACLA using custom-made diffraction apparatus.

Specimen preparation for cryogenic coherent X-ray diffraction imaging of biological cells and cellular organelles by using the X-ray free-electron laser at SACLA.

Coherent X-ray diffraction imaging (CXDI) allows internal structures of biological cells and cellular organelles to be analyzed. CXDI experiments have been conducted at 66 K for frozen-hydrated biological specimens at the SPring-8 Angstrom Compact Free-Electron Laser facility (SACLA). In these cryogenic CXDI experiments using X-ray free-electron laser (XFEL) pulses, specimen particles dispersed on thin membranes of specimen disks are transferred into the vacuum chamber of a diffraction apparatus. Because focused single XFEL pulses destroy specimen particles at the atomic level, diffraction patterns are collected through raster scanning the specimen disks to provide fresh specimen particles in the irradiation area. The efficiency of diffraction data collection in cryogenic experiments depends on the quality of the prepared specimens. Here, detailed procedures for preparing frozen-hydrated biological specimens, particularly thin membranes and devices developed in our laboratory, are reported. In addition, the quality of the frozen-hydrated specimens are evaluated by analyzing the characteristics of the collected diffraction patterns. Based on the experimental results, the internal structures of the frozen-hydrated specimens and the future development for efficient diffraction data collection are discussed.

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.

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.

Coherent X-Ray Diffraction Imaging of Chloroplasts from Cyanidioschyzon merolae by Using X-Ray Free Electron Laser.

Coherent X-ray diffraction imaging (CXDI) is a lens-less technique for visualizing the structures of non-crystalline particles with the dimensions of submicrometer to micrometer at a resolution of several tens of nanometers. We conducted cryogenic CXDI experiments at 66 K to visualize the internal structures of frozen-hydrated chloroplasts of Cyanidioschyzon merolae using X-ray free electron laser (XFEL) as a coherent X-ray source. Chloroplast dispersed specimen disks at a number density of 7/(10×10 µm(2)) were flash-cooled with liquid ethane without staining, sectioning or chemical labeling. Chloroplasts are destroyed at atomic level immediately after the diffraction by XFEL pulses. Thus, diffraction patterns with a good signal-to-noise ratio from single chloroplasts were selected from many diffraction patterns collected through scanning specimen disks to provide fresh specimens into the irradiation area. The electron density maps of single chloroplasts projected along the direction of the incident X-ray beam were reconstructed by using the iterative phase-retrieval method and multivariate analyses. The electron density map at a resolution of 70 nm appeared as a C-shape. In addition, the fluorescence image of proteins stained with Flamingo™ dye also appeared as a C-shape as did the autofluorescence from Chl. The similar images suggest that the thylakoid membranes with an abundance of proteins distribute along the outer membranes of chloroplasts. To confirm the present results statistically, a number of projection structures must be accumulated through high-throughput data collection in the near future. Based on the results, we discuss the feasibility of XFEL-CXDI experiments in the structural analyses of cellular organelles.

Data processing software suite SITENNO for coherent X-ray diffraction imaging using the X-ray free-electron laser SACLA.

Coherent X-ray diffraction imaging is a promising technique for visualizing the structures of non-crystalline particles with dimensions of micrometers to sub-micrometers. Recently, X-ray free-electron laser sources have enabled efficient experiments in the `diffraction before destruction' scheme. Diffraction experiments have been conducted at SPring-8 Angstrom Compact free-electron LAser (SACLA) using the custom-made diffraction apparatus KOTOBUKI-1 and two multiport CCD detectors. In the experiments, ten thousands of single-shot diffraction patterns can be collected within several hours. Then, diffraction patterns with significant levels of intensity suitable for structural analysis must be found, direct-beam positions in diffraction patterns determined, diffraction patterns from the two CCD detectors merged, and phase-retrieval calculations for structural analyses performed. A software suite named SITENNO has been developed to semi-automatically apply the four-step processing to a huge number of diffraction data. Here, details of the algorithm used in the suite are described and the performance for approximately 9000 diffraction patterns collected from cuboid-shaped copper oxide particles reported. Using the SITENNO suite, it is possible to conduct experiments with data processing immediately after the data collection, and to characterize the size distribution and internal structures of the non-crystalline particles.

IDATEN and G-SITENNO: GUI-assisted software for coherent X-ray diffraction imaging experiments and data analyses at SACLA.

Using our custom-made diffraction apparatus KOTOBUKI-1 and two multiport CCD detectors, cryogenic coherent X-ray diffraction imaging experiments have been undertaken at the SPring-8 Angstrom Compact free electron LAser (SACLA) facility. To efficiently perform experiments and data processing, two software suites with user-friendly graphical user interfaces have been developed. The first is a program suite named IDATEN, which was developed to easily conduct four procedures during experiments: aligning KOTOBUKI-1, loading a flash-cooled sample into the cryogenic goniometer stage inside the vacuum chamber of KOTOBUKI-1, adjusting the sample position with respect to the X-ray beam using a pair of telescopes, and collecting diffraction data by raster scanning the sample with X-ray pulses. Named G-SITENNO, the other suite is an automated version of the original SITENNO suite, which was designed for processing diffraction data. These user-friendly software suites are now indispensable for collecting a large number of diffraction patterns and for processing the diffraction patterns immediately after collecting data within a limited beam time.

Dark-field phase retrieval under the constraint of the Friedel symmetry in coherent X-ray diffraction imaging.

Coherent X-ray diffraction imaging (CXDI) is a lensless imaging technique that is suitable for visualizing the structures of non-crystalline particles with micrometer to sub-micrometer dimensions from material science and biology. One of the difficulties inherent to CXDI structural analyses is the reconstruction of electron density maps of specimen particles from diffraction patterns because saturated detector pixels and a beam stopper result in missing data in small-angle regions. To overcome this difficulty, the dark-field phase-retrieval (DFPR) method has been proposed. The DFPR method reconstructs electron density maps from diffraction data, which are modified by multiplying Gaussian masks with an observed diffraction pattern in the high-angle regions. In this paper, we incorporated Friedel centrosymmetry for diffraction patterns into the DFPR method to provide a constraint for the phase-retrieval calculation. A set of model simulations demonstrated that this constraint dramatically improved the probability of reconstructing correct electron density maps from diffraction patterns that were missing data in the small-angle region. In addition, the DFPR method with the constraint was applied successfully to experimentally obtained diffraction patterns with significant quantities of missing data. We also discuss this method's limitations with respect to the level of Poisson noise in X-ray detection.

Coherent diffraction imaging analysis of shape-controlled nanoparticles with focused hard X-ray free-electron laser pulses.

We report the first demonstration of the coherent diffraction imaging analysis of nanoparticles using focused hard X-ray free-electron laser pulses, allowing us to analyze the size distribution of particles as well as the electron density projection of individual particles. We measured 1000 single-shot coherent X-ray diffraction patterns of shape-controlled Ag nanocubes and Au/Ag nanoboxes and estimated the edge length from the speckle size of the coherent diffraction patterns. We then reconstructed the two-dimensional electron density projection with sub-10 nm resolution from selected coherent diffraction patterns. This method enables the simultaneous analysis of the size distribution of synthesized nanoparticles and the structures of particles at nanoscale resolution to address correlations between individual structures of components and the statistical properties in heterogeneous systems such as nanoparticles and cells.

CryoEM-sampling of metastable conformations appearing in cofactor-ligand association and catalysis of glutamate dehydrogenase.

Kinetic aspects of enzymatic reactions are described by equations based on the Michaelis-Menten theory for the initial stage. However, the kinetic parameters provide little information on the atomic mechanism of the reaction. In this study, we analyzed structures of glutamate dehydrogenase in the initial and steady stages of the reaction using cryoEM at near-atomic resolution. In the initial stage, four metastable conformations displayed different domain motions and cofactor/ligand association modes. The most striking finding was that the enzyme-cofactor-substrate complex, treated as a single state in the enzyme kinetic theory, comprised at least three different metastable conformations. In the steady stage, seven conformations, including derivatives from the four conformations in the initial stage, made the reaction pathway complicated. Based on the visualized conformations, we discussed stage-dependent pathways to illustrate the dynamics of the enzyme in action.

Prediction of hydrophilic and hydrophobic hydration structure of protein by neural network optimized using experimental data.

The hydration structures of proteins, which are necessary for their folding, stability, and functions, were visualized using X-ray and neutron crystallography and transmission electron microscopy. However, complete visualization of hydration structures over the entire protein surface remains difficult. To compensate for this incompleteness, we developed a three-dimensional convolutional neural network to predict the probability distribution of hydration water molecules on the hydrophilic and hydrophobic surfaces, and in the cavities of proteins. The neural network was optimized using the distribution patterns of protein atoms around the hydration water molecules identified in the high-resolution X-ray crystal structures. We examined the feasibility of the neural network using water sites in the protein crystal structures that were not included in the datasets. The predicted distribution covered most of the experimentally identified hydration sites, with local maxima appearing in their vicinity. This computational approach will help to highlight the relevance of hydration structures to the biological functions and dynamics of proteins.

Coenzyme-binding pathway on glutamate dehydrogenase suggested from multiple-binding sites visualized by cryo-electron microscopy.

The structure of hexameric glutamate dehydrogenase (GDH) in the presence of the coenzyme nicotinamide adenine dinucleotide phosphate (NADP) was visualized using cryogenic transmission electron microscopy to investigate the ligand-binding pathways to the active site of the enzyme. Each subunit of GDH comprises one hexamer-forming core domain and one nucleotide-binding domain (NAD domain), which spontaneously opens and closes the active-site cleft situated between the two domains. In the presence of NADP, the potential map of GDH hexamer, assuming D3 symmetry, was determined at a resolution of 2.4 Å, but the NAD domain was blurred due to the conformational variety. After focused classification with respect to the NAD domain, the potential maps interpreted as NADP molecules appeared at five different sites in the active-site cleft. The subunits associated with NADP molecules were close to one of the four metastable conformations in the unliganded state. Three of the five binding sites suggested a pathway of NADP molecules to approach the active-site cleft for initiating the enzymatic reaction. The other two binding modes may rarely appear in the presence of glutamate, as demonstrated by the reaction kinetics. Based on the visualized structures and the results from the enzymatic kinetics, we discussed the binding modes of NADP to GDH in the absence and presence of glutamate.