Phenix Software Research Articles

The Phenix-AlphaFold webservice: Enabling AlphaFold predictions for use in Phenix.

Advances in machine learning have enabled sufficiently accurate predictions of protein structure to be used in macromolecular structure determination with crystallography and cryo-electron microscopy data. The Phenix software suite has AlphaFold predictions integrated into an automated pipeline that can start with an amino acid sequence and data, and automatically perform model-building and refinement to return a protein model fitted into the data. Due to the steep technical requirements of running AlphaFold efficiently, we have implemented a Phenix-AlphaFold webservice that enables all Phenix users to run AlphaFold predictions remotely from the Phenix GUI starting with the official 1.21 release. This webservice will be improved based on how it is used by the research community and the future research directions for Phenix.

Accounting for nonuniformity of bulk-solvent: A mosaic model.

A flat mask-based model is almost universally used in macromolecular crystallography to account for disordered (bulk) solvent. This model assumes any voxel of the crystal unit cell that is not occupied by the atomic model is occupied by the solvent. The properties of this solvent are assumed to be exactly the same across the whole volume of the unit cell. While this is a reasonable approximation in practice, there are a number of scenarios where this model becomes suboptimal. In this work, we enumerate several of these scenarios and describe a new generalized approach to modeling the bulk-solvent which we refer to as mosaic bulk-solvent model. The mosaic bulk-solvent model allows nonuniform features of the solvent in the crystal to be accounted for in a computationally efficient way. It is implemented in the computational crystallography toolbox and the Phenix software.

Efficient structure-factor modeling for crystals with multiple components.

Diffraction intensities from a crystallographic experiment include contributions from the entire unit cell of the crystal: the macromolecule, the solvent around it and eventually other compounds. These contributions cannot typically be well described by an atomic model alone, i.e. using point scatterers. Indeed, entities such as disordered (bulk) solvent, semi-ordered solvent (e.g. lipid belts in membrane proteins, ligands, ion channels) and disordered polymer loops require other types of modeling than a collection of individual atoms. This results in the model structure factors containing multiple contributions. Most macromolecular applications assume two-component structure factors: one component arising from the atomic model and the second one describing the bulk solvent. A more accurate and detailed modeling of the disordered regions of the crystal will naturally require more than two components in the structure factors, which presents algorithmic and computational challenges. Here an efficient solution of this problem is proposed. All algorithms described in this work have been implemented in the computational crystallography toolbox (CCTBX) and are also available within Phenix software. These algorithms are rather general and do not use any assumptions about molecule type or size nor about those of its components.

Conformational space exploration of cryo-EM structures by variability refinement

Cryo-EM observation of biological samples enables visualization of sample heterogeneity, in the form of discrete states that are separable, or continuous heterogeneity as a result of local protein motion before flash freezing. Variability analysis of this continuous heterogeneity describes the variance between a particle stack and a volume, and results in a map series describing the various steps undertaken by the sample in the particle stack. While this observation is absolutely stunning, it is very hard to pinpoint structural details to elements of the maps. In order to bridge the gap between observation and explanation, we designed a tool that refines an ensemble of structures into all the maps from variability analysis. Using this bundle of structures, it is easy to spot variable parts of the structure, as well as the parts that are not moving. Comparison with molecular dynamics simulations highlights the fact that the movements follow the same directions, albeit with different amplitudes. Ligand can also be investigated using this method. Variability refinement is available in the Phenix software suite, accessible under the program name phenix.varref.

Sleuthing biochemical evidence to elucidate unassigned electron density in a CBL-SLAP2 crystal complex.

The Src-like adaptor proteins (SLAP/SLAP2) bind to CBL E3 ubiquitin ligase to downregulate antigen, cytokine and tyrosine kinase receptor signalling. In contrast to the phosphotyrosine-dependent binding of CBL substrates through its tyrosine kinase-binding domain (TKBD), CBL TKBD associates with the C-terminal tail of SLAP2 in a phospho-independent manner. To understand the distinct nature of this interaction, a purification protocol for SLAP2 in complex with CBL TKBD was established and the complex was crystallized. However, determination of the complex crystal structure was hindered by the apparent degradation of SLAP2 during the crystallization process, such that only the CBL TKBD residues could initially be modelled. Close examination of the CBL TKBD structure revealed a unique dimer interface that included two short segments of electron density of unknown origin. To elucidate which residues of SLAP2 to model into this unassigned density, a co-expression system was generated to test SLAP2 deletion mutants and define the minimal SLAP2 binding region. SLAP2 degradation products were also analysed by mass spectrometry. The model-building and map-generation features of the Phenix software package were employed, leading to successful modelling of the C-terminal tail of SLAP2 into the unassigned electron-density segments.

CERES: a cryo-EM re-refinement system for continuous improvement of deposited models.

The field of electron cryomicroscopy (cryo-EM) has advanced quickly in recent years as the result of numerous technological and methodological developments. This has led to an increase in the number of atomic structures determined using this method. Recently, several tools for the analysis of cryo-EM data and models have been developed within the Phenix software package, such as phenix.real_space_refine for the refinement of atomic models against real-space maps. Also, new validation metrics have been developed for low-resolution cryo-EM models. To understand the quality of deposited cryo-EM structures and how they might be improved, models deposited in the Protein Data Bank that have map resolutions of better than 5 Å were automatically re-refined using current versions of Phenix tools. The results are available on a publicly accessible web page (https://cci.lbl.gov/ceres). The implementation of a Cryo-EM Re-refinement System (CERES) for the improvement of models deposited in the wwPDB, and the results of the re-refinements, are described. Based on these results, contents are proposed for a `cryo-EM Table 1', which summarizes experimental details and validation metrics in a similar way to `Table 1' in crystallography. The consistent use of robust metrics for the evaluation of cryo-EM models and data should accompany every structure deposition and be reported in scientific publications.

Barium blockade of the KcsA channel in open and closed conformation datasets.

Barium is a potent blocker of the KcsA potassium channel. A strategy using x-ray crystallography and molecular dynamics (MD) simulation has been used to understand this phenomenon as described in Rohaim et al. [1]. Wild type KcsA is purified to homogeneity and crystallized in low and high K+ conditions. Crystals are grown using the hanging drop vapor diffusion method. To examine barium binding in the selectivity filter of KcsA, the crystals are systemically soaked in various concentrations of barium chloride solution. X-ray crystallography datasets are collected at the Advanced Photon Source. A total of 10 datasets are collected for various barium ion concentrations. Diffraction data are processed using the crystallography pipeline software RAPID. The crystal structures are solved by molecular replacement methods. The structure models are visualized using COOT and refined using REFMAC. Anomalous map coefficients are calculated using the phenix.maps tool in the PHENIX software suite. The datasets are deposited in the Protein Data Bank. The data provides a detailed picture of barium ion interaction with potassium channels. Structural analysis of the KcsA channel reveals two distinct configurations, open- and closed- state. Further MD simulation analysis suggests an energetically favorable binding mechanism for barium ion in the selectivity filter. The data could be used to interpret functional experiments related to barium blockade for potassium channels. Also, it is valuable for comparison and cross validation with other relevant potassium channel structures.

Deriving and refining atomic models in crystallography and cryo-EM: the latest Phenix tools to facilitate structure analysis.

In structural biology, deriving and refining atomic models into maps obtained from X-ray crystallography or cryo electron microscopy (cryo-EM) is essential for the detailed interpretation of a structure and its functional implications through interactions so that for example hydrogen bonds, drug specificity and associated molecular mechanisms can be analysed. This commentary summarizes the latest features of the Phenix software and also highlights the fact that cryo-EM increasingly contributes to data depositions in the PDB and EMDB.

Cryo_fit: Democratization of flexible fitting for cryo-EM

Cryo-electron microscopy (cryo-EM) is becoming a method of choice for describing native conformations of biomolecular complexes at high resolution. The rapid growth of cryo-EM in recent years has created a high demand for automated solutions, both in hardware and software. Flexible fitting of atomic models to three-dimensional (3D) cryo-EM reconstructions by molecular dynamics (MD) simulation is a popular technique but often requires technical expertise in computer simulation. This work introduces cryo_fit, a package for the automatic flexible fitting of atomic models in cryo-EM maps using MD simulation. The package is integrated with the Phenix software suite. The module was designed to automate the multiple steps of MD simulation in a reproducible manner, as well as facilitate refinement and validation through Phenix. Through the use of cryo_fit, scientists with little experience in MD simulation can produce high quality atomic models automatically and better exploit the potential of cryo-EM.

MolProbity: More and better reference data for improved all-atom structure validation.

This paper describes the current update on macromolecular model validation services that are provided at the MolProbity website, emphasizing changes and additions since the previous review in 2010. There have been many infrastructure improvements, including rewrite of previous Java utilities to now use existing or newly written Python utilities in the open-source CCTBX portion of the Phenix software system. This improves long-term maintainability and enhances the thorough integration of MolProbity-style validation within Phenix. There is now a complete MolProbity mirror site at http://molprobity.manchester.ac.uk. GitHub serves our open-source code, reference datasets, and the resulting multi-dimensional distributions that define most validation criteria. Coordinate output after Asn/Gln/His "flip" correction is now more idealized, since the post-refinement step has apparently often been skipped in the past. Two distinct sets of heavy-atom-to-hydrogen distances and accompanying van der Waals radii have been researched and improved in accuracy, one for the electron-cloud-center positions suitable for X-ray crystallography and one for nuclear positions. New validations include messages at input about problem-causing format irregularities, updates of Ramachandran and rotamer criteria from the million quality-filtered residues in a new reference dataset, the CaBLAM Cα-CO virtual-angle analysis of backbone and secondary structure for cryoEM or low-resolution X-ray, and flagging of the very rare cis-nonProline and twisted peptides which have recently been greatly overused. Due to wide application of MolProbity validation and corrections by the research community, in Phenix, and at the worldwide Protein Data Bank, newly deposited structures have continued to improve greatly as measured by MolProbity's unique all-atom clashscore.

Expression and crystallographic studies of D-glycero-β-D-manno-heptose-1-phosphate adenylyltransferase from Burkholderia pseudomallei.

The Gram-negative bacterium Burkholderia pseudomallei is the causative agent of melioidosis. D-glycero-β-D-manno-Heptose-1-phosphate adenylyltransferase (HldC) is the fourth enzyme of the ADP-L-glycero-β-D-manno-heptose biosynthesis pathway, which produces an essential carbohydrate comprising the inner core of lipopolysaccharide. Therefore, HldC is a potential target of antibiotics against melioidosis. In this study, HldC from B. pseudomallei has been cloned, expressed, purified and crystallized. Synchrotron X-ray data from a selenomethionine-substituted HldC crystal were also collected to 2.8 Å resolution. The crystal belonged to the primitive triclinic space group P1, with unit-cell parameters a = 74.0, b = 74.0, c = 74.9 Å, α = 108.4, β = 108.4, γ = 108.0°. Eight protomers are present in the unit cell and three out of five selenomethionines were found in each protomer using the PHENIX software suite. A full structural determination is in progress to elucidate the structure-function relationship of the protein.

Improved ligand geometries in crystallographic refinement using AFITT in PHENIX.

Modern crystal structure refinement programs rely on geometry restraints to overcome the challenge of a low data-to-parameter ratio. While the classical Engh and Huber restraints work well for standard amino-acid residues, the chemical complexity of small-molecule ligands presents a particular challenge. Most current approaches either limit ligand restraints to those that can be readily described in the Crystallographic Information File (CIF) format, thus sacrificing chemical flexibility and energetic accuracy, or they employ protocols that substantially lengthen the refinement time, potentially hindering rapid automated refinement workflows. PHENIX-AFITT refinement uses a full molecular-mechanics force field for user-selected small-molecule ligands during refinement, eliminating the potentially difficult problem of finding or generating high-quality geometry restraints. It is fully integrated with a standard refinement protocol and requires practically no additional steps from the user, making it ideal for high-throughput workflows. PHENIX-AFITT refinements also handle multiple ligands in a single model, alternate conformations and covalently bound ligands. Here, the results of combining AFITT and the PHENIX software suite on a data set of 189 protein-ligand PDB structures are presented. Refinements using PHENIX-AFITT significantly reduce ligand conformational energy and lead to improved geometries without detriment to the fit to the experimental data. For the data presented, PHENIX-AFITT refinements result in more chemically accurate models for small-molecule ligands.

FEM: Feature Enhanced Map

A method is proposed that modifies 2Fo-Fc σA-weighted map such that the resulting map has significantly reduced amount of noise and model bias, and at the same time the weak signal is enhanced so it is visible as good as the strong signal. We call the new map Feature Enhanced Map (FEM). An atomic model is a result of interpretation of the electron density map. There are at least three fundamentally different sources of map imperfections that affect quality of this interpretation. 1) Errors. Fourier map artifacts due to finite resolution and completeness of data, errors in experimental data and crystal model parameters result in noise peaks in map that oftentimes appear very much similar to the true atomic structure and in turn may be erroneously interpreted as such. Also, the noise may obscure or corrupt the signal making its meaningful interpretation difficult or even impossible. 2) Signal strength. Not all the signal presented in the map has the same strength: a strong signal arising from a heavy atom derivative may obscure weak signal arising from a partially occupied very mobile ligand or residue side chain alternative conformation or hydrogen atoms. It is common that such weak signals are near or below the noise level. 3) Model bias. Crystallographic maps are calculated using model phases alone or a combination of model and experimental phases. Once placed, correctly or not, atoms feed back revealing themselves in the map. Therefore erroneously placed atoms may be positively confirmed against the map resulting in an incorrect structure. All three sources of map imperfections are always present together. They make interpretation of crystallographic maps never unambiguous, trivial, or unique at typical macromolecular resolutions. The new method aims to overcome these problems and is available in Phenix software. More information: http://www.phenix-online.org/presentations/fem_SEP2013.pdf

New tools for automated model completion and refinement

Although macromolecular crystallography has been greatly accelerated by the development of automated software for data processing, phasing, and model building, most structures require significant manual intervention to yield a truly final model. In addition to missing individual protein or nucleic residues, this may include the addition of alternate conformations, ligands (both free and covalently bound), elemental ions, or modified amino acids. We have developed a number of tools to streamline several of these steps within the Phenix software suite (Adams et al. 2010): 1) an automated pipeline for the determination of ligand-bound structures by molecular replacement (Echols et al. 2014a); 2) placement of elemental ions during refinement (Echols et al. 2014b), as an extension of solvent placement; 3) fitting of additional conformations of protein residues into difference density. These tools reliably reproduce published structures in a majority of test cases, and in several instances identify details omitted by the original authors. Their low false positive rate makes them suitable for use in high-throughput workflows.

Anomalies in the refinement of isoleucine.

A study of isoleucines in protein structures solved using X-ray crystallography revealed a series of systematic trends for the two side-chain torsion angles χ1 and χ2 dependent on the resolution, secondary structure and refinement software used. The average torsion angles for the nine rotamers were similar in high-resolution structures solved using either the REFMAC, CNS or PHENIX software. However, at low resolution these programs often refine towards somewhat different χ1 and χ2 values. Small systematic differences can be observed between refinement software that uses molecular dynamics-type energy terms (for example CNS) and software that does not use these terms (for example REFMAC). Detailing the standard torsion angles used in refinement software can improve the refinement of protein structures. The target values in the molecular dynamics-type energy functions can also be improved.

Protein structure determination via X‐ray diffraction methods in an undergraduate biophysical laboratory course

Understanding the relationship between structure and function of biological macromolecules is a commonly emphasized concept in biochemistry and molecular biology. Teaching experimental aspects of structure determination is increasingly important so that students can navigate structure databases, understand method limitations, and critically interpret the growing body of structural information. Practical experience also has potential advantages to students starting graduate school or careers in industry. We sought to test the feasibility of undergraduate students in a laboratory class solving a protein structure, starting from purified protein, using X‐ray diffraction methods. We chose the serine protease trypsin as a target protein and used an in‐house X‐ray diffractometer recently acquired for teaching and research. We assessed strategies to grow crystals and collect high quality, high resolution (1.1 – 1.3 Å) data sets within a short time frame. Using the Phenix software program, we solved the structure via sulfur‐single wavelength anomalous diffraction and/or molecular replacement methods, and built and refined molecular models. Our results indicate that all aspects of this process can be implemented in a laboratory class. There is potential for altering the target protein to address hypothesis‐driven questions, and also for sharing data with other institutions for instructional purposes.

Structures of the PelD Cyclic Diguanylate Effector Involved in Pellicle Formation in Pseudomonas aeruginosa PAO1

The second messenger bis-(3'-5')-cyclic dimeric guanosine monophosphate (c-di-GMP) plays a vital role in the global regulation in bacteria. Here, we describe structural and biochemical characterization of a novel c-di-GMP effector PelD that is critical to the formation of pellicles by Pseudomonas aeruginosa. We present high-resolution structures of a cytosolic fragment of PelD in apo form and its complex with c-di-GMP. The structure contains a bi-domain architecture composed of a GAF domain (commonly found in cyclic nucleotide receptors) and a GGDEF domain (found in c-di-GMP synthesizing enzymes), with the latter binding to one molecule of c-di-GMP. The GGDEF domain has a degenerate active site but a conserved allosteric site (I-site), which we show binds c-di-GMP with a K(d) of 0.5 μm. We identified a series of residues that are crucial for c-di-GMP binding, and confirmed the roles of these residues through biochemical characterization of site-specific variants. The structures of PelD represent a novel class of c-di-GMP effector and expand the knowledge of scaffolds that mediate c-di-GMP recognition.

Crystal Structure of Elongator Subcomplex Elp4–6

Elongator is a multiprotein complex composed of two subcomplexes, Elp1-3 and Elp4-6. Elongator is highly conserved between yeast and humans and plays an important role in RNA polymerase II-mediated transcriptional elongation and many other processes, including cytoskeleton organization, exocytosis, and tRNA modification. Here, we determined the crystal structure of the Elp4-6 subcomplex of yeast. The overall structure of Elp4-6 revealed that Elp6 acts as a bridge to assemble Elp4 and Elp5. Detailed structural and sequence analyses revealed that each subunit in the Elp4-6 subcomplex forms a RecA-ATPase-like fold, although it lacks the key sequence signature of ATPases. Site-directed mutagenesis and biochemical analyses indicated that the Elp4-6 subcomplex can assemble into a hexameric ring-shaped structure in vitro and in vivo. Furthermore, GST pulldown assays showed that the ring-shaped assembly of the Elp4-6 subcomplex is important for its specific histone H3 binding. Our results may shed light on the substrate recognition and assembly of the holo-Elongator complex.

Crystal Structure of Clustered Regularly Interspaced Short Palindromic Repeats (CRISPR)-associated Csn2 Protein Revealed Ca2+-dependent Double-stranded DNA Binding Activity

Clustered regularly interspaced short palindromic repeats (CRISPR) and their associated protein genes (cas genes) are widespread in bacteria and archaea. They form a line of RNA-based immunity to eradicate invading bacteriophages and malicious plasmids. A key molecular event during this process is the acquisition of new spacers into the CRISPR loci to guide the selective degradation of the matching foreign genetic elements. Csn2 is a Nmeni subtype-specific cas gene required for new spacer acquisition. Here we characterize the Enterococcus faecalis Csn2 protein as a double-stranded (ds-) DNA-binding protein and report its 2.7 Å tetrameric ring structure. The inner circle of the Csn2 tetrameric ring is ∼26 Å wide and populated with conserved lysine residues poised for nonspecific interactions with ds-DNA. Each Csn2 protomer contains an α/β domain and an α-helical domain; significant hinge motion was observed between these two domains. Ca(2+) was located at strategic positions in the oligomerization interface. We further showed that removal of Ca(2+) ions altered the oligomerization state of Csn2, which in turn severely decreased its affinity for ds-DNA. In summary, our results provided the first insight into the function of the Csn2 protein in CRISPR adaptation by revealing that it is a ds-DNA-binding protein functioning at the quaternary structure level and regulated by Ca(2+) ions.

Structural Basis of Poly(ADP-ribose) Recognition by the Multizinc Binding Domain of Checkpoint with Forkhead-associated and RING Domains (CHFR)

Cellular stress in early mitosis activates the antephase checkpoint, resulting in the decondensation of chromosomes and delayed mitotic progression. Checkpoint with forkhead-associated and RING domains (CHFR) is central to this checkpoint, and its activity is ablated in many tumors and cancer cell lines through promoter hypermethylation or mutation. The interaction between the PAR-binding zinc finger (PBZ) of CHFR and poly(ADP-ribose) (PAR) is crucial for a functional antephase checkpoint. We determined the crystal structure of the cysteine-rich region of human CHFR (amino acids 425-664) to 1.9 Å resolution, which revealed a multizinc binding domain of elaborate topology within which the PBZ is embedded. The PBZ of CHFR closely resembles the analogous motifs from aprataxin-like factor and CG1218-PA, which lie within unstructured regions of their respective proteins. Based on co-crystal structures of CHFR bound to several different PAR-like ligands (adenosine 5'-diphosphoribose, adenosine monophosphate, and P(1)P(2)-diadenosine 5'-pyrophosphate), we made a model of the CHFR-PAR interaction, which we validated using site-specific mutagenesis and surface plasmon resonance. The PBZ motif of CHFR recognizes two adenine-containing subunits of PAR and the phosphate backbone that connects them. More generally, PBZ motifs may recognize different numbers of PAR subunits as required to carry out their functions.