AlphaFold predictions are valuable hypotheses and accelerate but do not replace experimental structure determination.

Artificial intelligence-based protein structure prediction methods such as AlphaFold have revolutionized structural biology. The accuracies of these predictions vary, however, and they do not take into account ligands, covalent modifications or other environmental factors. Here, we evaluate how well AlphaFold predictions can be expected to describe the structure of a protein by comparing predictions directly with experimental crystallographic maps. In many cases, AlphaFold predictions matched experimental maps remarkably closely. In other cases, even very high-confidence predictions differed from experimental maps on a global scale through distortion and domain orientation, and on a local scale in backbone and side-chain conformation. We suggest considering AlphaFold predictions as exceptionally useful hypotheses. We further suggest that it is important to consider the confidence in prediction when interpreting AlphaFold predictions and to carry out experimental structure determination to verify structural details, particularly those that involve interactions not included in the prediction.

Outcomes of the EMDataResource cryo-EM Ligand Modeling Challenge.

The EMDataResource Ligand Model Challenge aimed to assess the reliability and reproducibility of modeling ligands bound to protein and protein-nucleic acid complexes in cryogenic electron microscopy (cryo-EM) maps determined at near-atomic (1.9-2.5 Å) resolution. Three published maps were selected as targets: Escherichia coli beta-galactosidase with inhibitor, SARS-CoV-2 virus RNA-dependent RNA polymerase with covalently bound nucleotide analog and SARS-CoV-2 virus ion channel ORF3a with bound lipid. Sixty-one models were submitted from 17 independent research groups, each with supporting workflow details. The quality of submitted ligand models and surrounding atoms were analyzed by visual inspection and quantification of local map quality, model-to-map fit, geometry, energetics and contact scores. A composite rather than a single score was needed to assess macromolecule+ligand model quality. These observations lead us to recommend best practices for assessing cryo-EM structures of liganded macromolecules reported at near-atomic resolution.

Structures of apo Cas12a and its complex with crRNA and DNA reveal the dynamics of ternary complex formation and target DNA cleavage.

Cas12a is a programmable nuclease for adaptive immunity against invading nucleic acids in CRISPR-Cas systems. Here, we report the crystal structures of apo Cas12a from Lachnospiraceae bacterium MA2020 (Lb2) and the Lb2Cas12a+crRNA complex, as well as the cryo-EM structure and functional studies of the Lb2Cas12a+crRNA+DNA complex. We demonstrate that apo Lb2Cas12a assumes a unique, elongated conformation, whereas the Lb2Cas12a+crRNA binary complex exhibits a compact conformation that subsequently rearranges to a semi-open conformation in the Lb2Cas12a+crRNA+DNA ternary complex. Notably, in solution, apo Lb2Cas12a is dynamic and can exist in both elongated and compact forms. Residues from Met493 to Leu523 of the WED domain undergo major conformational changes to facilitate the required structural rearrangements. The REC lobe of Lb2Cas12a rotates 103° concomitant with rearrangement of the hinge region close to the WED and RuvC II domains to position the RNA-DNA duplex near the catalytic site. Our findings provide insight into crRNA recognition and the mechanism of target DNA cleavage.

Improved AlphaFold modeling with implicit experimental information.

Machine-learning prediction algorithms such as AlphaFold and RoseTTAFold can create remarkably accurate protein models, but these models usually have some regions that are predicted with low confidence or poor accuracy. We hypothesized that by implicitly including new experimental information such as a density map, a greater portion of a model could be predicted accurately, and that this might synergistically improve parts of the model that were not fully addressed by either machine learning or experiment alone. An iterative procedure was developed in which AlphaFold models are automatically rebuilt on the basis of experimental density maps and the rebuilt models are used as templates in new AlphaFold predictions. We show that including experimental information improves prediction beyond the improvement obtained with simple rebuilding guided by the experimental data. This procedure for AlphaFold modeling with density has been incorporated into an automated procedure for interpretation of crystallographic and electron cryo-microscopy maps.

CRLF3 plays a key role in the final stage of platelet genesis and is a potential therapeutic target for thrombocythemia.

The process of platelet production has so far been understood to be a 2-stage process: megakaryocyte maturation from hematopoietic stem cells followed by proplatelet formation, with each phase regulating the peripheral blood platelet count. Proplatelet formation releases into the bloodstream beads-on-a-string preplatelets, which undergo fission into mature platelets. For the first time, we show that preplatelet maturation is a third, tightly regulated, critical process akin to cytokinesis that regulates platelet count. We show that deficiency in cytokine receptor-like factor 3 (CRLF3) in mice leads to an isolated and sustained 25% to 48% reduction in the platelet count without any effect on other blood cell lineages. We show that Crlf3-/- preplatelets have increased microtubule stability, possibly because of increased microtubule glutamylation via the interaction of CRLF3 with key members of the Hippo pathway. Using a mouse model of JAK2 V617F essential thrombocythemia, we show that a lack of CRLF3 leads to long-term lineage-specific normalization of the platelet count. We thereby postulate that targeting CRLF3 has therapeutic potential for treatment of thrombocythemia.

An oligomeric state-dependent switch in the ER enzyme FICD regulates AMPylation and deAMPylation of BiP.

AMPylation is an inactivating modification that alters the activity of the major endoplasmic reticulum (ER) chaperone BiP to match the burden of unfolded proteins. A single ER-localised Fic protein, FICD (HYPE), catalyses both AMPylation and deAMPylation of BiP. However, the basis for the switch in FICD's activity is unknown. We report on the transition of FICD from a dimeric enzyme, that deAMPylates BiP, to a monomer with potent AMPylation activity. Mutations in the dimer interface, or of residues along an inhibitory pathway linking the dimer interface to the enzyme's active site, favour BiP AMPylation in vitro and in cells. Mechanistically, monomerisation relieves a repressive effect allosterically propagated from the dimer interface to the inhibitory Glu234, thereby permitting AMPylation-competent binding of MgATP. Moreover, a reciprocal signal, propagated from the nucleotide-binding site, provides a mechanism for coupling the oligomeric state and enzymatic activity of FICD to the energy status of the ER.

Improvement of cryo-EM maps by density modification.

A density-modification procedure for improving maps from single-particle electron cryogenic microscopy (cryo-EM) is presented. The theoretical basis of the method is identical to that of maximum-likelihood density modification, previously used to improve maps from macromolecular X-ray crystallography. Key differences from applications in crystallography are that the errors in Fourier coefficients are largely in the phases in crystallography but in both phases and amplitudes in cryo-EM, and that half-maps with independent errors are available in cryo-EM. These differences lead to a distinct approach for combination of information from starting maps with information obtained in the density-modification process. The density-modification procedure was applied to a set of 104 datasets and improved map-model correlation and increased the visibility of details in many of the maps. The procedure requires two unmasked half-maps and a sequence file or other source of information on the volume of the macromolecule that has been imaged.

Assessing the utility of CASP14 models for molecular replacement.

The assessment of CASP models for utility in molecular replacement is a measure of their use in a valuable real-world application. In CASP7, the metric for molecular replacement assessment involved full likelihood-based molecular replacement searches; however, this restricted the assessable targets to crystal structures with only one copy of the target in the asymmetric unit, and to those where the search found the correct pose. In CASP10, full molecular replacement searches were replaced by likelihood-based rigid-body refinement of models superimposed on the target using the LGA algorithm, with the metric being the refined log-likelihood-gain (LLG) score. This enabled multi-copy targets and very poor models to be evaluated, but a significant further issue remained: the requirement of diffraction data for assessment. We introduce here the relative-expected-LLG (reLLG), which is independent of diffraction data. This reLLG is also independent of any crystal form, and can be calculated regardless of the source of the target, be it X-ray, NMR or cryo-EM. We calibrate the reLLG against the LLG for targets in CASP14, showing that it is a robust measure of both model and group ranking. Like the LLG, the reLLG shows that accurate coordinate error estimates add substantial value to predicted models. We find that refinement by CASP groups can often convert an inadequate initial model into a successful MR search model. Consistent with findings from others, we show that the AlphaFold2 models are sufficiently good, and reliably so, to surpass other current model generation strategies for attempting molecular replacement phasing.

Germline mutations in the transcription factor IKZF5 cause thrombocytopenia.

To identify novel causes of hereditary thrombocytopenia, we performed a genetic association analysis of whole-genome sequencing data from 13 037 individuals enrolled in the National Institute for Health Research (NIHR) BioResource, including 233 cases with isolated thrombocytopenia. We found an association between rare variants in the transcription factor-encoding gene IKZF5 and thrombocytopenia. We report 5 causal missense variants in or near IKZF5 zinc fingers, of which 2 occurred de novo and 3 co-segregated in 3 pedigrees. A canonical DNA-zinc finger binding model predicts that 3 of the variants alter DNA recognition. Expression studies showed that chromatin binding was disrupted in mutant compared with wild-type IKZF5, and electron microscopy revealed a reduced quantity of α granules in normally sized platelets. Proplatelet formation was reduced in megakaryocytes from 7 cases relative to 6 controls. Comparison of RNA-sequencing data from platelets, monocytes, neutrophils, and CD4+ T cells from 3 cases and 14 healthy controls showed 1194 differentially expressed genes in platelets but only 4 differentially expressed genes in each of the other blood cell types. In conclusion, IKZF5 is a novel transcriptional regulator of megakaryopoiesis and the eighth transcription factor associated with dominant thrombocytopenia in humans.

Structural basis for the specificity of renin-mediated angiotensinogen cleavage.

The renin-angiotensin cascade is a hormone system that regulates blood pressure and fluid balance. Renin-mediated cleavage of the angiotensin I peptide from the N terminus of angiotensinogen (AGT) is the rate-limiting step of this cascade; however, the detailed molecular mechanism underlying this step is unclear. Here, we solved the crystal structures of glycosylated human AGT (2.30 Å resolution), its encounter complex with renin (2.55 Å), AGT cleaved in its reactive center loop (RCL; 2.97 Å), and spent AGT from which the N-terminal angiotensin peptide was removed (2.63 Å). These structures revealed that AGT undergoes profound conformational changes and binds renin through a tail-into-mouth allosteric mechanism that inserts the N terminus into a pocket equivalent to a hormone-binding site on other serpins. These changes fully extended the N-terminal tail, with the scissile bond for angiotensin release docked in renin's active site. Insertion of the N terminus into this pocket accompanied a complete unwinding of helix H of AGT, which, in turn, formed key interactions with renin in the complementary binding interface. Mutagenesis and kinetic analyses confirmed that renin-mediated production of angiotensin I is controlled by interactions of amino acid residues and glycan components outside renin's active-site cleft. Our findings indicate that AGT adapts unique serpin features for hormone delivery and binds renin through concerted movements in the N-terminal tail and in its main body to modulate angiotensin release. These insights provide a structural basis for the development of agents that attenuate angiotensin release by targeting AGT's hormone binding pocket.

Ab initio solution of macromolecular crystal structures without direct methods.

The majority of macromolecular crystal structures are determined using the method of molecular replacement, in which known related structures are rotated and translated to provide an initial atomic model for the new structure. A theoretical understanding of the signal-to-noise ratio in likelihood-based molecular replacement searches has been developed to account for the influence of model quality and completeness, as well as the resolution of the diffraction data. Here we show that, contrary to current belief, molecular replacement need not be restricted to the use of models comprising a substantial fraction of the unknown structure. Instead, likelihood-based methods allow a continuum of applications depending predictably on the quality of the model and the resolution of the data. Unexpectedly, our understanding of the signal-to-noise ratio in molecular replacement leads to the finding that, with data to sufficiently high resolution, fragments as small as single atoms of elements usually found in proteins can yield ab initio solutions of macromolecular structures, including some that elude traditional direct methods.

How serpins transport hormones and regulate their release.

The adaptation of the serpin framework and its mechanism to perform diverse functions is epitomised in the hormone carriers of the blood. Thyroxine and the corticosteroids are transported bound in a 1:1 ratio on almost identical sites in the two homologous binding-globulins, TBG and CBG. Recent structural findings show an equilibrated, rather than on-and-off, release of the hormones from the carriers, reflecting small reversible movements of the hinge region of the reactive loop that modify the conformational flexibility of the underlying hormone-binding site. Consequently, contrary to previous concepts, the binding affinities of TBG and CBG are not fixed but can be allosterically modified to allow differential hormone delivery. Notably, the two carriers function like protein thermocouples with a surge in hormone release as body temperatures rise in fevers, and conversely a large diminution in free hormone levels at hibernation temperatures. By comparison angiotensinogen, the source of the angiotensin peptides that control blood pressure, does not appear to utilise the serpin mechanism. It has instead evolved a 63 residue terminal extension containing the buried angiotensin cleavage site, which on interaction moves into the active cleft of the renin. The conformational shift involved is critically linked by a labile disulphide bridge. The observation of changes in the redox status of this S-S bridge, in the hypertensive complication of pregnancy, pre-eclampsia, has opened an unexpected level of regulation at what is the initial stage in the control of blood pressure.

Evaluation of template-based modeling in CASP13.

Performance in the template-based modeling (TBM) category of CASP13 is assessed here, using a variety of metrics. Performance of the predictor groups that participated is ranked using the primary ranking score that was developed by the assessors for CASP12. This reveals that the best results are obtained by groups that include contact predictions or inter-residue distance predictions derived from deep multiple sequence alignments. In cases where there is a good homolog in the wwPDB (TBM-easy category), the best results are obtained by modifying a template. However, for cases with poorer homologs (TBM-hard), very good results can be obtained without using an explicit template, by deep learning algorithms trained on the wwPDB. Alternative metrics are introduced, to allow testing of aspects of structural models that are not addressed by traditional CASP metrics. These include comparisons to the main-chain and side-chain torsion angles of the target, and the utility of models for solving crystal structures by the molecular replacement method. The alternative metrics are poorly correlated with the traditional metrics, and it is proposed that modeling has reached a sufficient level of maturity that the best models should be expected to satisfy this wider range of criteria.

Evaluation of model refinement in CASP13.

Performance in the model refinement category of the 13th round of Critical Assessment of Structure Prediction (CASP13) is assessed, showing that some groups consistently improve most starting models whereas the majority of participants continue to degrade the starting model on average. Using the ranking formula developed for CASP12, it is shown that only 7 of 32 groups perform better than a "naïve predictor" who just submits the starting model. Common features in their approaches include a dependence on physics-based force fields to judge alternative conformations and the use of molecular dynamics to relax models to local minima, usually with some restraints to prevent excessively large movements. In addition to the traditional CASP metrics that focus largely on the quality of the overall fold, alternative metrics are evaluated, including comparisons of the main-chain and side-chain torsion angles, and the utility of the models for solving crystal structures by the molecular replacement method. It is proposed that the introduction of these metrics, as well as consideration of the accuracy of coordinate error estimates, would improve the discrimination between good and very good models.

Measurement of the total angiotensinogen and its reduced and oxidised forms in human plasma using targeted LC-MS/MS.

Angiotensinogen (AGT) is a critical protein in the renin-angiotensin-aldosterone system and may have an important role in the pathogenesis of pre-eclampsia. The disulphide linkage between cysteines 18 and 138 has a key role in the redox switch of AGT which modulates the release of angiotensin I with consequential effects on blood pressure. In this paper, we report a quantitative targeted LC-MS/MS method for the reliable measurement of the total AGT and its reduced and oxidised forms in human plasma. AGT was selectively enriched from human plasma using two-dimensional chromatography employing concanavalin A lectin affinity and reversed phase steps and then deglycosylated using PNGase F. A differential alkylation approach was coupled with targeted LC-MS/MS method to identify the two AGT forms in the plasma chymotryptic digest. An additional AGT proteolytic marker peptide was identified and used to measure total AGT levels. The developed MS workflow enabled the reproducible detection of total AGT and its two distinct forms in human plasma with analytical precision of ≤ 15%. The LC-MS/MS assay for total AGT in plasma showed a linear response (R2 = 0.992) with a limit of quantification in the low nanomolar range. The method gave suitable validation characteristics for biomedical application to the quantification of the oxidation level and the total level of AGT in plasma samples collected from normal and pre-eclamptic patients.

A critical examination of the recently reported crystal structures of the human SMN protein.

A recent publication by Seng et al. in this journal reports the crystallographic structure of refolded, full-length SMN protein and two disease-relevant derivatives thereof. Here, we would like to suggest that at least two of the structures reported in that study are incorrect. We present evidence that one of the associated crystallographic datasets is derived from a crystal of the bacterial Sm-like protein Hfq and that a second dataset is derived from a crystal of the bacterial Gab protein. Both proteins are frequent contaminants of bacterially overexpressed proteins which might have been co-purified during metal affinity chromatography. A third structure presented in the Seng et al. paper cannot be examined further because neither the atomic coordinates, nor the diffraction intensities were made publicly available. The Tudor domain protein SMN has been shown to be a component of the SMN complex, which mediates the assembly of RNA-protein complexes of uridine-rich small nuclear ribonucleoproteins (UsnRNPs). Importantly, this activity is reduced in SMA patients, raising the possibility that the aetiology of SMA is linked to RNA metabolism. Structural studies on diverse components of the SMN complex, including fragments of SMN itself have contributed greatly to our understanding of the cellular UsnRNP assembly machinery. Yet full-length SMN has so far evaded structural elucidation. The Seng et al. study claimed to have closed this gap, but based on the results presented here, the only conclusion that can be drawn is that the Seng et al. study is largely invalid and should be retracted from the literature.

Macromolecular X-ray structure determination using weak, single-wavelength anomalous data.

We describe a likelihood-based method for determining the substructure of anomalously scattering atoms in macromolecular crystals that allows successful structure determination by single-wavelength anomalous diffraction (SAD) X-ray analysis with weak anomalous signal. With the use of partial models and electron density maps in searches for anomalously scattering atoms, testing of alternative values of parameters and parallelized automated model-building, this method has the potential to extend the applicability of the SAD method in challenging cases.

Molecular Mechanism of Z α1-Antitrypsin Deficiency.

The Z mutation (E342K) of α1-antitrypsin (α1-AT), carried by 4% of Northern Europeans, predisposes to early onset of emphysema due to decreased functional α1-AT in the lung and to liver cirrhosis due to accumulation of polymers in hepatocytes. However, it remains unclear why the Z mutation causes intracellular polymerization of nascent Z α1-AT and why 15% of the expressed Z α1-AT is secreted into circulation as functional, but polymerogenic, monomers. Here, we solve the crystal structure of the Z-monomer and have engineered replacements to assess the conformational role of residue Glu-342 in α1-AT. The results reveal that Z α1-AT has a labile strand 5 of the central ß-sheet A (s5A) with a consequent equilibrium between a native inhibitory conformation, as in its crystal structure here, and an aberrant conformation with s5A only partially incorporated into the central ß-sheet. This aberrant conformation, induced by the loss of interactions from the Glu-342 side chain, explains why Z α1-AT is prone to polymerization and readily binds to a 6-mer peptide, and it supports that annealing of s5A into the central ß-sheet is a crucial step in the serpins' metastable conformational formation. The demonstration that the aberrant conformation can be rectified through stabilization of the labile s5A by binding of a small molecule opens a potential therapeutic approach for Z α1-AT deficiency.

Improved molecular replacement by density- and energy-guided protein structure optimization.

Molecular replacement procedures, which search for placements of a starting model within the crystallographic unit cell that best account for the measured diffraction amplitudes, followed by automatic chain tracing methods, have allowed the rapid solution of large numbers of protein crystal structures. Despite extensive work, molecular replacement or the subsequent rebuilding usually fail with more divergent starting models based on remote homologues with less than 30% sequence identity. Here we show that this limitation can be substantially reduced by combining algorithms for protein structure modelling with those developed for crystallographic structure determination. An approach integrating Rosetta structure modelling with Autobuild chain tracing yielded high-resolution structures for 8 of 13 X-ray diffraction data sets that could not be solved in the laboratories of expert crystallographers and that remained unsolved after application of an extensive array of alternative approaches. We estimate that the new method should allow rapid structure determination without experimental phase information for over half the cases where current methods fail, given diffraction data sets of better than 3.2 Å resolution, four or fewer copies in the asymmetric unit, and the availability of structures of homologous proteins with >20% sequence identity.

A redox switch in angiotensinogen modulates angiotensin release.

Blood pressure is critically controlled by angiotensins, which are vasopressor peptides specifically released by the enzyme renin from the tail of angiotensinogen-a non-inhibitory member of the serpin family of protease inhibitors. Although angiotensinogen has long been regarded as a passive substrate, the crystal structures solved here to 2.1 Å resolution show that the angiotensin cleavage site is inaccessibly buried in its amino-terminal tail. The conformational rearrangement that makes this site accessible for proteolysis is revealed in our 4.4 Å structure of the complex of human angiotensinogen with renin. The co-ordinated changes involved are seen to be critically linked by a conserved but labile disulphide bridge. Here we show that the reduced unbridged form of angiotensinogen is present in the circulation in a near 40:60 ratio with the oxidized sulphydryl-bridged form, which preferentially interacts with receptor-bound renin. We propose that this redox-responsive transition of angiotensinogen to a form that will more effectively release angiotensin at a cellular level contributes to the modulation of blood pressure. Specifically, we demonstrate the oxidative switch of angiotensinogen to its more active sulphydryl-bridged form in the maternal circulation in pre-eclampsia-the hypertensive crisis of pregnancy that threatens the health and survival of both mother and child.