Cryo-electron Tomography （CryoET） can visualize near-native 3D structures of heterogenous specimen, such as organelles or cells. Recent advances in equipment and methods of sample preparation greatly expand the application ranges of the CryoET. In this review, we will briefly introduce these advances and workflow of CryoET.

The scattering powers of electron and X-ray differ by 4 - 5 orders of magnitude. Thanks to this property, the electron beam yields high-resolution diffraction spots from undersized crystals of various samples, which are hard to grow to a suitable size for X-ray diffraction even with a high-intensity synchrotron radiation beam. Thus, the technique known as electron 3D crystallography/3D ED/MicroED is recognized as being important especially in synthetic chemistry, material sciences and related areas, while single particle analysis can be used for larger-sized proteins. Here I review this technology including our recent developments and results.

Radial distribution functions are commonly used to study the structures of many non-crystalline materials. The present author proposed a completely different method to describe the solution structure by expressing the inhomogeneity in distribution of molecules and in concentration as “density fluctuation” and “concentration fluctuation”, respectively; namely the structure of a solution is described in terms of the “mixing state” or “mixing scheme.” This paper introduces the fluctuations, as well as Kirkwood-Buff Integrals. Fluctuations of solutions become more pronounced in the mesoscale region. The relationship with solution thermodynamics, which represents the macroscopic limit, is also discussed. The features and cautions of experiments to measure the fluctuations are described. Finally, as analytical examples, temperature and concentration dependences of mixing schemes for two solution systems with upper critical and lower critical solution temperatures are presented.

There are five macromolecular crystallography beamlines at the Photon Factory. All beamlines support fully automated unattended data collection and remote interactive data collection, and users can select them when they submit beamtime requests. The beamlines also support browsing of results of the past experiments as well as the automated data analysis pipelines. The interactive data processing and raw data download from remote will be supported in the next autumn.

We developed a fully automated protein crystallization and monitoring system （PXS） in 2003 and have been operating the system. We improved PXS with step-by-step and describes the summary of the improvements of PXS, now named PXS2. The minimum sample volume reduces to 0.1 μL, the resolution of crystallization drop images increases to 5 M pixels, and a low temperature incubator is added. In addition to the vapor diffusion method, PXS2 can handle bicelle and LCP methods for membrane protein crystallization. These improvements expand the applicability of PXS2 and could reduce the bottleneck of X-ray protein crystallography.

Synchrotron beamlines are now required to provide an experimental environment that supports digital transformation. Many of the beamlines for macromolecular crystals around the world now employ the systems for automated data acquisition and remote access. At SPring-8, both systems have been newly established with brilliant undulator beamlines to provide rapid and accurate measurements. Here, we will introduce these systems that can be used without visiting the site.

Automated data collection by the ZOO system has realized effective data collection for protein structure determination at SPring-8. Using the ZOO system, it is now available to collect diffraction data from hundreds of single crystals within a day. In order to analyze crystal structures using the huge amount of data, we have been developing an automated structure analysis pipeline. Furthermore, we have applied the highly efficient data collection systems of SPring-8 to high-throughput ligand screening using protein crystals.

Here we report our recent developments on the automated structure analysis pipeline NABE system and the ligand screening pipeline using an acoustic liquid handler ECHO.

The recent developments of automated data collection and remote access technologies at synchrotron facility are highly remarkable. Here I describe our recent attempt to identify phosphates using their anomalous signals without visiting synchrotron facility. Crystals were screened using automated data collection with the X-ray at the wavelength of 1.0 Å. Then the solvent around the qualified crystals was removed by deep-UV laser ablation. The diffraction datasets were subsequently collected from the solvent-removed crystals using the X-ray at the wavelength of 2.7 Å. Anomalous Fourier map from the dataset collected with 2.7 Å X-ray was superposed on the electron density map calculated from the dataset collected with 1.0 Å X-ray. The two maps clearly indicated the positions of non-Carbon/Nitrogen/Oxygen atoms in the crystal. Our trial proved that considerable parts of the experiments in the synchrotron facility could be performed from the remote site.

High-speed, fully automated data collection allows us to conduct DNA Monozukuri （manufacturing） based on a large number of 3D structural information. In the process of designing and synthesizing prototype DNAs by mimicking structures deposited in Protein Data Bank （PDB）, brushing up their functions, and obtaining the finished products, each of the previously- and newly-solved crystal structures will be respected and reviewed/observed in more detail than before.

Cyclic β-amino acids, such as （1R,2R）-2-aminocyclohexane carboxylic acid （ACHC）, are known to induce compact conformation of peptides by forming turns and helices. Recently, a method to incorporate cyclic β-amino acids into peptides using ribosomes has been reported. Combining this method with the mRNA display method enabled the de novo discovery of macrocyclic peptides that contain cyclic β-amino acid residues and efficiently inhibit human coagulation factor XIIa （FXIIa）. One such peptide, F3, potently inhibits FXIIa （Ki: 1.5 nM） and contains two ACHC residues. In the co-crystal structure of FXIIa-F3, F3 adopts an antiparallel β-hairpin structure stabilized by two ACHC residues. Despite its smaller size, F3 engages its target in a manner similar to that of natural protein-based inhibitors of serine proteases, demonstrating the utility of foldamer library consisting of macrocyclic peptides containing cyclic β-amino acids.

Recent advance in the automatic data collection systems at the synchrotron facilities enable us to obtain high quality diffraction data without on-site operation. Structural determinations from the crystal clusters of small proteins were successfully achieved with microfocus beam, the automated data-collection system, and the data-processing pipeline in SPring-8.

In recent years, fragment-based drug discovery（FBDD） has revolutionised the development of potent lead compounds against protein targets. This “start small, elaborate efficiently” approach promises to address many deficiencies of traditional medicinal chemistry. It has been used successfully by industry and academics, delivering several drugs to the clinic, and many more to late-stage clinical trials. Probably the most useful and informative technique for FBDD is X-ray crystallography, employed most effectively as a primary screen to detect the binding of fragment hits to protein target sites, while simultaneously elucidating binding poise and identifying possibilities for fragment development. Here we report the development of the Fast Fragment and Compound Screening（FFCS） platform at the Swiss Light Source（SLS）, an integrated next generation pipeline for crystal soaking, handling and data collection, which allows crystallography-based screening of protein crystals against hundreds of potential drug leads. We also report the release of the Smart Digital User（SDU） software at our PXI X06SA and PXIII X06DA beamlines. In combination with existing beamline software infrastructure, SDU enables fully automated data collection, and is used both in concert with our FFCS platform and is a tool that is available for all beamline users.

お詫びと訂正

本誌59巻2・3号掲載のクリスタリット「電子線三次元結晶構造解析」（Vol.59 No.2・3 p.128）について, 本文を以下のとおり訂正いたします.

誤：分子が三次元に並んだ薄い結晶から，電子線回折パターンを測定，結晶構造を解析する技術．先に確立したタンパク質の電子線二次元結晶構造解析では，分子が一層にのみ並んだ二次元の結晶を対象とする．二次元結晶からの回折は繰り返しのない方向に連続的な格子線となるのに対して，三次元結晶ではすべての方向に離散的に分布する回折点となる．したがって，その回折点の強度情報をもれなく取得するためには，試料を回転させながら回折パターンを記録する回転測定や，ある角度ごとに，静止パターンや電子線を傾けたプレセッションパターンを撮影することが必要になる．電子線はX線に比べて4～5乗も強く物質と相互作用するため，X線結晶構造解析に適さない微小で薄い単結晶が使用できる．反対に，主に軽い元素からなるタンパク質の結晶でも，厚さ200～300 nmを超えるものからの解析は難しくなる．電子線の原子の散乱特性からは，電荷分布に関する情報が得られる．

正：分子が三次元に並んだ薄い結晶から，電子回折パターンを測定，結晶構造を解析する技術．先に確立したタンパク質の電子線二次元結晶構造解析では，分子が一層にのみ並んだ二次元の結晶を対象とする．二次元結晶からの回折は繰り返しのない方向に連続的な格子線となるのに対して，三次元結晶ではすべての方向に離散的に分布する回折点となる．したがって，その回折点の強度情報をもれなく取得するためには，試料を回転させながら回折パターンを記録する回転測定や，ある角度ごとに，静止パターンや電子線を傾けプリセッションパターンを撮影することが必要になる．電子はX線に比べて4～5乗も強く物質と相互作用するため，X線結晶構造解析に適さない微小で薄い単結晶が使用できる．反対に，主に軽い元素からなるタンパク質の結晶でも，厚さ数百nmを超えるものからの解析は難しくなる．電子の散乱特性からは，電荷に関する情報が得られる．