Refinement of protein structures by iterative comparative modeling and CryoEM density fitting

We developed a method for structure characterization of assembly components by iterative comparative protein structure modeling and fitting into cryo-electron microscopy (cryoEM) density maps. Specifically, we calculate a comparative model of a given component by considering many alternative alignments between the target sequence and a related template structure while optimizing the fit of a model into the corresponding density map. The method relies on the previously developed Moulder protocol that iterates over alignment, model building, and model assessment. The protocol was benchmarked using 20 varied target-template pairs of known structures with less than 30% sequence identity and corresponding simulated density maps at resolutions from 5A to 25A. Relative to the models based on the best existing sequence profile alignment methods, the percentage of C(alpha) atoms that are within 5A of the corresponding C(alpha) atoms in the superposed native structure increases on average from 52% to 66%, which is half-way between the starting models and the models from the best possible alignments (82%). The test also reveals that despite the improvements in the accuracy of the fitness function, this function is still the bottleneck in reducing the remaining errors. To demonstrate the usefulness of the protocol, we applied it to the upper domain of the P8 capsid protein of rice dwarf virus that has been studied by cryoEM at 6.8A. The C(alpha) root-mean-square deviation of the model based on the remotely related template, bluetongue virus VP7, improved from 8.7A to 6.0A, while the best possible model has a C(alpha) RMSD value of 5.3A. Moreover, the resulting model fits better into the cryoEM density map than the initial template structure. The method is being implemented in our program MODELLER for protein structure modeling by satisfaction of spatial restraints and will be applicable to the rapidly increasing number of cryoEM density maps of macromolecular assemblies.     

BCL::EM-Fit: rigid body fitting of atomic structures into density maps using geometric hashing and real space refinement

Cryo-electron microscopy (cryoEM) can visualize large macromolecular assemblies at resolutions often below 10? and recently as good as 3.8-4.5 ?. These density maps provide important insights into the biological functioning of molecular machineries such as viruses or the ribosome, in particular if atomic-resolution crystal structures or models of individual components of the assembly can be placed into the density map. The present work introduces a novel algorithm termed BCL::EM-Fit that accurately fits atomic-detail structural models into medium resolution density maps. In an initial step, a "geometric hashing" algorithm provides a short list of likely placements. In a follow up Monte Carlo/Metropolis refinement step, the initial placements are optimized by their cross correlation coefficient. The resolution of density maps for a reliable fit was determined to be 10 ? or better using tests with simulated density maps. The algorithm was applied to fitting of capsid proteins into an experimental cryoEM density map of human adenovirus at a resolution of 6.8 and 9.0 ?, and fitting of the GroEL protein at 5.4 ?. In the process, the handedness of the cryoEM density map was unambiguously identified. The BCL::EM-Fit algorithm offers an alternative to the established Fourier/Real space fitting programs. BCL::EM-Fit is free for academic use and available from a web server or as downloadable binary file at http://www.meilerlab.org.     

Finding rigid bodies in protein structures: Application to flexible fitting into cryoEM maps

We present RIBFIND, a method for detecting flexibility in protein structures via the clustering of secondary structural elements (SSEs) into rigid bodies. To test the usefulness of the method in refining atomic structures within cryoEM density we incorporated it into our flexible fitting protocol (Flex-EM). Our benchmark includes 13 pairs of protein structures in two conformations each, one of which is represented by a corresponding cryoEM map. Refining the structures in simulated and experimental maps at the 5–15 Å resolution range using rigid bodies identified by RIBFIND shows a significant improvement over using individual SSEs as rigid bodies. For the 15 Å resolution simulated maps, using RIBFIND-based rigid bodies improves the initial fits by 40.64% on average, as compared to 26.52% when using individual SSEs. Furthermore, for some test cases we show that at the sub-nanometer resolution range the fits can be further improved by applying a two-stage refinement protocol (using RIBFIND-based refinement followed by an SSE-based refinement). The method is stand-alone and could serve as a general interactive tool for guiding flexible fitting into EM maps.     

Scoring functions for cryoEM density fitting

In fitting atomic structures into cryoEM density maps of macromolecular assemblies, the cross-correlation function (CCF) is the most prevalent method of scoring the goodness-of-fit. However, there are still many possible, less studied ways of scoring fits. In this paper, we introduce four scores new to cryoEM fitting and compare their performance to three known scores. Our benchmark consists of (a) 4 protein assemblies with simulated maps at 5-20 ? resolution, including the heptameric ring of GroEL; and (b) 4 experimental maps of GroEL at ～6-23 ? resolution with corresponding fitted atomic models. We perturb each fit 1000 times and assess each new fit with each score. The correlation between a score and the Cα RMSD of each fit from the "correctly" fitted structure shows that the CCF is one of the best scores, but in certain situations could be augmented or even replaced by other scores. For instance, our implementation of a score based on mutual information outperforms or is comparable to the CCF in almost all test cases, and our new "envelope score" works as well as the CCF at sub-nanometer resolution but is an order of magnitude faster to calculate. The results also suggest that the width of the Gaussian function used to blur the atomic structure into a density map can significantly affect the fitting process. Finally, we show that our score-testing method, when combined with the Laplacian CCF or the mutual information scores, can be used as a statistical tool for improving cryoEM density fitting.     

Structural characterization of components of protein assemblies by comparative modeling and electron cryo-microscopy

We explore structural characterization of protein assemblies by a combination of electron cryo-microscopy (cryoEM) and comparative protein structure modeling. Specifically, our method finds an optimal atomic model of a given assembly subunit and its position within an assembly by fitting alternative comparative models into a cryoEM map. The alternative models are calculated by MODELLER [J. Mol. Biol. 234 (1993) 313] from different sequence alignments between the modeled protein and its template structures. The fitting of these models into a cryoEM density map is performed either by FOLDHUNTER [J. Mol. Biol. 308 (2001) 1033] or by a new density fitting module of MODELLER (Mod-EM). Identification of the most accurate model is based on the correlation between the model accuracy and the quality of fit into the cryoEM density map. To quantify this correlation, we created a benchmark consisting of eight proteins of different structural folds with corresponding density maps simulated at five resolutions from 5 to 15 angstroms, with three noise levels each. Each of the proteins in the set was modeled based on 300 different alignments to their remotely related templates (12-32% sequence identity), spanning the range from entirely inaccurate to essentially accurate alignments. The benchmark revealed that one of the most accurate models can usually be identified by the quality of its fit into the cryoEM density map, even for noisy maps at 15 angstroms resolution. Therefore, a cryoEM density map can be helpful in improving the accuracy of a comparative model. Moreover, a pseudo-atomic model of a component in an assembly may be built better with comparative models of the native subunit sequences than with experimentally determined structures of their homologs.     

Differences between poliovirus empty capsids formed in vivo and those formed in vitro: a role for the morphopoietic factor.

Empty capsid species formed from the self- and extract-mediated assembly of poliovirus type 1 14S particles in vitro and procapsids isolated from virus-infected cells were subjected to isoelectric focusing in charge-free agarose gels. The empty capsid formed in the self-assembly reaction had an isoelectric point (pI) of 5.0, whereas procapsids and extract-assembled empty capsids focused at pH 6.8. Unreacted 14S particles focused at pH 4.8 to 5.0. The sedimentation coefficient (s20,w) and density of the empty capsid species were also determined. Procapsids had a density in CsCl of 1.31 g/cm3, whereas empty capsids formed by self- or extract-mediated assembly had a density of 1.29 g/cm3. Both extract-assembled empty capsids and procapsids had an s20,w of 75S, whereas self-assembled empty capsids had an s20,w of 71S. Self-assembled empty capsids were not converted to pI 6.8 empty capsids by incubation with poliovirus-infected HeLa cell extracts. The dissociated polypeptides of self-assembled empty capsids (pI 5.0) and procapsids (pI 6.8) behaved identically when analyzed by isoelectric focusing in the presence of 9 M urea and by polyacrylamide gel electrophoresis in the presence of sodium dodecyl sulfate. These results suggest that infected cell extracts possess a factor that influences the final conformation of the empty shell (pI 6.8, 75S) formed from 14S particles and that this influences is exerted at the initiation step or during the polymerization reaction. A small amount of this activity (less than or equal to 20% of infected extracts) was detected in uninfected cells; the significance of this remains unknown.     

The hand, foot and mouth disease virus capsid: sequence analysis and prediction of antigenic sites from homology modelling

Enterovirus 71 (EV71) is the most common aetiological agent detected in cases of hand, foot and mouth disease (HFMD) resulting in incidences of neurological complications and fatality in recent years. A comparison of the capsid proteins implicated in the pathogenicity of the fatal and non-fatal strains of EV71, reveals a high degree of homology (93%-100% identity). To facilitate diagnostic immunoassays and vaccine development, a consensus structural model for the EV71 coat protein has been developed based primarily on the homologous structure of the bovine enterovirus. The overall architecture of the virion closely resembles those of related icosahedral picornaviruses. Detailed atomic modelling of the fatal 5865/SIN/00009 strain has been carried out, and the functional regions (known and predicted) from closely related viruses mapped onto the surface of the predicted structure. From the model, we have identified two putative immunogenic regions, one of which is unique to EV71. The hydrophobic pocket within VP1, found in bovine enterovirus, poliovirus and rhinovirus, is also conserved in EV71.     

Visualization of the externalized VP2 N termini of infectious human parvovirus B19

The structures of infectious human parvovirus B19 and empty wild-type particles were determined by cryoelectron microscopy (cryoEM) to 7.5-Å and 11.3-Å resolution, respectively, assuming icosahedral symmetry. Both of these, DNA filled and empty, wild-type particles contain a few copies of the minor capsid protein VP1. Comparison of wild-type B19 with the crystal structure and cryoEM reconstruction of recombinant B19 particles consisting of only the major capsid protein VP2 showed structural differences in the vicinity of the icosahedral fivefold axes. Although the unique N-terminal region of VP1 could not be visualized in the icosahedrally averaged maps, the N terminus of VP2 was shown to be exposed on the viral surface adjacent to the fivefold β-cylinder. The conserved glycine-rich region is positioned between two neighboring, fivefold-symmetrically related VP subunits and not in the fivefold channel as observed for other parvoviruses.     

Interaction of coxsackievirus A21 with its cellular receptor, ICAM-1

Coxsackievirus A21 (CAV21), like human rhinoviruses (HRVs), is a causative agent of the common cold. It uses the same cellular receptor, intercellular adhesion molecule 1 (ICAM-1), as does the major group of HRVs; unlike HRVs, however, it is stable at acid pH. The cryoelectron microscopy (cryoEM) image reconstruction of CAV21 is consistent with the highly homologous crystal structure of poliovirus 1; like other enteroviruses and HRVs, CAV21 has a canyon-like depression around each of the 12 fivefold vertices. A cryoEM reconstruction of CAV21 complexed with ICAM-1 shows all five domains of the extracellular component of ICAM-1. The known atomic structure of the ICAM-1 amino-terminal domains D1 and D2 has been fitted into the cryoEM density of the complex. The site of ICAM-1 binding within the canyon of CAV21 overlaps the site of receptor recognition utilized by rhinoviruses and polioviruses. Interactions within this common region may be essential for triggering viral destabilization after attachment to susceptible cells.     

Functional Comparison of SCARB2 and PSGL1 as Receptors for Enterovirus 71

Human scavenger receptor class B, member 2 (SCARB2), and P-selectin glycoprotein ligand-1 (PSGL1) have been identified to be the cellular receptors for enterovirus 71 (EV71). We compared the EV71 infection efficiencies of mouse L cells that expressed SCARB2 (L-SCARB2) and PSGL1 (L-PSGL1) and the abilities of SCARB2 and PSGL1 to bind to the virus. L-SCARB2 cells bound a reduced amount of EV71 compared to L-PSGL1 cells. However, EV71 could infect L-SCARB2 cells more efficiently than L-PSGL1 cells. The results suggested that the difference in the binding capacities of the two receptors was not the sole determinant of the infection efficiency and that SCARB2 plays an essential role after attaching to virions. Therefore, we examined the viral entry into L-SCARB2 cells and L-PSGL1 cells by immunofluorescence microscopy. In both cells, we detected internalized EV71 virions that colocalized with an early endosome marker. We then performed a sucrose density gradient centrifugation analysis to evaluate viral uncoating. After incubating the EV71 virion with L-SCARB2 cells or soluble SCARB2 under acidic conditions below pH 6.0, we observed that part of the native virion was converted into an empty capsid that lacked both genomic RNA and VP4 capsid proteins. The results suggested that the uncoating of EV71 requires both SCARB2 and an acidic environment and occurs after the internalization of the virus-receptor complex into endosomes. However, the empty capsid formation was not observed after incubation with L-PSGL1 cells or soluble PSGL1 under any of the tested pH conditions. These results indicated that SCARB2 is capable of viral binding, viral internalization, and viral uncoating and that the low infection efficiency of L-PSGL1 cells is due to the inability of PSGL1 to induce viral uncoating. The characterization of SCARB2 as an uncoating receptor greatly contributes to the understanding of the early steps of EV71 infection.     

Structural analysis of coxsackievirus A7 reveals conformational changes associated with uncoating

Coxsackievirus A7 (CAV7) is a rarely detected and poorly characterized serotype of the Enterovirus species Human enterovirus A (HEV-A) within the Picornaviridae family. The CAV7-USSR strain has caused polio-like epidemics and was originally thought to represent the fourth poliovirus type, but later evidence linked this strain to the CAV7-Parker prototype. Another isolate, CAV7-275/58, was also serologically similar to Parker but was noninfectious in a mouse model. Sequencing of the genomic region encoding the capsid proteins of the USSR and 275/58 strains and subsequent comparison with the corresponding amino acid sequences of the Parker strain revealed that the Parker and USSR strains are nearly identical, while the 275/58 strain is more distant. Using electron cryomicroscopy and three-dimensional image reconstruction, the structures of the CAV7-USSR virion and empty capsid were resolved to 8.2-Å and 6.1-Å resolutions, respectively. This is one of the first detailed structural analyses of the HEV-A species. Using homology modeling, reconstruction segmentation, and flexible fitting, we constructed a pseudoatomic T = 1 (pseudo T = 3) model incorporating the three major capsid proteins (VP1 to VP3), addressed the conformational changes of the capsid and its constituent viral proteins occurring during RNA release, and mapped the capsid proteins'' variable regions to the structure. During uncoating, VP4 and RNA are released analogously to poliovirus 1, the interfaces of VP2 and VP3 are rearranged, and VP1 rotates. Variable regions in the capsid proteins were predicted to map mainly to the surface of VP1 and are thus likely to affect the tropism and pathogenicity of CAV7.     

Quasi-atomic model of bacteriophage t7 procapsid shell: insights into the structure and evolution of a basic fold

The existence of similar folds among major structural subunits of viral capsids has shown unexpected evolutionary relationships suggesting common origins irrespective of the capsids' host life domain. Tailed bacteriophages are emerging as one such family, and we have studied the possible existence of the HK97-like fold in bacteriophage T7. The procapsid structure at approximately 10 A resolution was used to obtain a quasi-atomic model by fitting a homology model of the T7 capsid protein gp10 that was based on the atomic structure of the HK97 capsid protein. A number of fold similarities, such as the fitting of domains A and P into the L-shaped procapsid subunit, are evident between both viral systems. A different feature is related to the presence of the amino-terminal domain of gp10 found at the inner surface of the capsid that might play an important role in the interaction of capsid and scaffolding proteins.     

Structure and protein distribution for the capsid of Caulobacter crescentus bacteriophage phiCbK

Electron microscope images of negatively stained empty capsids of Caulobacter crescentus bacteriophage φCbK have been analyzed by computer Fourier methods. Two-dimensional computer density maps for the distribution of material in the capsid wall have been obtained by computer Fourier filtering which made possible the complete separation of contributions from the front and back of the capsid including overlapping Fourier coefficients. After scaling for the capsid thickness, using a one-dimensional reconstruction of data from edge-on views of the capsid wall and using topographical information provided by low-angle metal shadowing, a three-dimensional density distribution has been derived. A model for the distribution of subunits in the capsid is proposed which places each of the two major capsid subunit proteins (which are present in a 2:1 ratio) in quasi-equivalent bonding environments. A tentative model is presented in which assembly of the capsid (an elongated T = 7l icosadeltahedron) is regulated by the bonding geometry of the penton proteins.     

Lytic replication of Kaposi's sarcoma-associated herpesvirus results in the formation of multiple capsid species: isolation and molecular characterization of A, B, and C capsids from a gammaherpesvirus

Despite the discovery of Epstein-Barr virus more than 35 years ago, a thorough understanding of gammaherpesvirus capsid composition and structure has remained elusive. We approached this problem by purifying capsids from Kaposi's sarcoma-associated herpesvirus (KSHV), the only other known human gammaherpesvirus. The results from our biochemical and imaging analyses demonstrate that KSHV capsids possess a typical herpesvirus icosahedral capsid shell composed of four structural proteins. The hexameric and pentameric capsomers are composed of the major capsid protein (MCP) encoded by open reading frame 25. The heterotrimeric complexes, forming the capsid floor between the hexons and pentons, are each composed of one molecule of ORF62 and two molecules of ORF26. Each of these proteins has significant amino acid sequence homology to capsid proteins in alpha- and betaherpesviruses. In contrast, the fourth protein, ORF65, lacks significant sequence homology to its structural counterparts from the other subfamilies. Nevertheless, this small, basic, and highly antigenic protein decorates the surface of the capsids, as does, for example, the even smaller basic capsid protein VP26 of herpes simplex virus type 1. We have also found that, as with the alpha- and betaherpesviruses, lytic replication of KSHV leads to the formation of at least three capsid species, A, B, and C, with masses of approximately 200, 230, and 300 MDa, respectively. A capsids are empty, B capsids contain an inner array of a fifth structural protein, ORF17.5, and C capsids contain the viral genome.     

Three-dimensional structures of the A, B, and C capsids of rhesus monkey rhadinovirus: insights into gammaherpesvirus capsid assembly, maturation, and DNA packaging

Rhesus monkey rhadinovirus (RRV) exhibits high levels of sequence homology to human gammaherpesviruses, such as Kaposi's sarcoma-associated herpesvirus, and grows to high titers in cell cultures, making it a good model system for studying gammaherpesvirus capsid structure and assembly. We have purified RRV A, B, and C capsids, thus for the first time allowing direct structure comparisons by electron cryomicroscopy and three-dimensional reconstruction. The results show that the shells of these capsids are identical and are each composed of 12 pentons, 150 hexons, and 320 triplexes. Structural differences were apparent inside the shells and through the penton channels. The A capsid is empty, and its penton channels are open. The B capsid contains a scaffolding core, and its penton channels are closed. The C capsid contains a DNA genome, which is closely packaged into regularly spaced density shells (25 A apart), and its penton channels are open. The different statuses of the penton channels suggest a functional role of the channels during capsid maturation, and the overall structural similarities of RRV capsids to alphaherpesvirus capsids suggest a common assembly and maturation pathway. The RRV A capsid reconstruction at a 15-A resolution, the best achieved for gammaherpesvirus particles, reveals overall structural similarities to alpha- and betaherpesvirus capsids. However, the outer regions of the capsid, including densities attributed to the Ta triplex and the small capsomer-interacting protein (SCIP or ORF65), exhibit prominent differences from their structural counterparts in alphaherpesviruses. This structural disparity suggests that SCIP and the triplex, together with tegument and envelope proteins, confer structural and potentially functional specificities to alpha-, beta-, and gammaherpesviruses.     

Time-resolved molecular dynamics of bacteriophage HK97 capsid maturation interpreted by electron cryo-microscopy and X-ray crystallography

The bacteriophage HK97 capsid is a molecular machine that exhibits large-scale conformational rearrangements of its 420 identical protein subunits during capsid maturation. Immature empty capsids, termed Prohead II, assemble in vivo in an Escherichia coli expression system. Maturation of these particles may be induced in vitro, converting them into Head II capsids that are indistinguishable in conformation from the capsid of an infectious phage particle. One method of in vitro maturation requires acidification to drive the reaction through two expansion intermediates (EI-I, EI-II) to its penultimate particle state (EI-III), which has 86% more internal volume than Prohead II. Neutralization of EI-III produces the fully mature capsid, Head II. The three expansion intermediates and the acid expansion pathway were characterized by cryo-EM analysis and 3D reconstruction. We now report that, although large-scale structural changes are involved, the electron density maps for these intermediate states are readily interpreted in terms of quasi-atomic models based on subunit structures determined by prior crystallographic analysis of Head II. Progression through the expansion intermediate states primarily represents rigid-body rotations and translations of the subunits, accompanied by refolding of two small regions, the N-terminal arm and a beta-hairpin called the E-loop. Movies made with these pseudo-atomic coordinates and the Head II X-ray coordinates illuminate various aspects of the maturation pathway in the course of which the pattern of inter-subunit interactions is sequentially transformed while the integrity of the capsid is maintained.     

The Enterovirus 71 A-particle Forms a Gateway to Allow Genome Release: A CryoEM Study of Picornavirus Uncoating

Since its discovery in 1969, enterovirus 71 (EV71) has emerged as a serious worldwide health threat. This human pathogen of the picornavirus family causes hand, foot, and mouth disease, and also has the capacity to invade the central nervous system to cause severe disease and death. Upon binding to a host receptor on the cell surface, the virus begins a two-step uncoating process, first forming an expanded, altered “A-particle”, which is primed for genome release. In a second step after endocytosis, an unknown trigger leads to RNA expulsion, generating an intact, empty capsid. Cryo-electron microscopy reconstructions of these two capsid states provide insight into the mechanics of genome release. The EV71 A-particle capsid interacts with the genome near the icosahedral two-fold axis of symmetry, which opens to the external environment via a channel ∼10 Å in diameter that is lined with patches of negatively charged residues. After the EV71 genome has been released, the two-fold channel shrinks, though the overall capsid dimensions are conserved. These structural characteristics identify the two-fold channel as the site where a gateway forms and regulates the process of genome release.     

C terminus of infectious bursal disease virus major capsid protein VP2 is involved in definition of the T number for capsid assembly

Infectious bursal disease virus (IBDV), a member of the Birnaviridae family, is a double-stranded RNA virus. The IBDV capsid is formed by two major structural proteins, VP2 and VP3, which assemble to form a T=13 markedly nonspherical capsid. During viral infection, VP2 is initially synthesized as a precursor, called VPX, whose C end is proteolytically processed to the mature form during capsid assembly. We have computed three-dimensional maps of IBDV capsid and virus-like particles built up by VP2 alone by using electron cryomicroscopy and image-processing techniques. The IBDV single-shelled capsid is characterized by the presence of 260 protruding trimers on the outer surface. Five classes of trimers can be distinguished according to their different local environments. When VP2 is expressed alone in insect cells, dodecahedral particles form spontaneously; these may be assembled into larger, fragile icosahedral capsids built up by 12 dodecahedral capsids. Each dodecahedral capsid is an empty T=1 shell composed of 20 trimeric clusters of VP2. Structural comparison between IBDV capsids and capsids consisting of VP2 alone allowed the determination of the major capsid protein locations and the interactions between them. Whereas VP2 forms the outer protruding trimers, VP3 is found as trimers on the inner surface and may be responsible for stabilizing functions. Since elimination of the C-terminal region of VPX is correlated with the assembly of T=1 capsids, this domain might be involved (either alone or in cooperation with VP3) in the induction of different conformations of VP2 during capsid morphogenesis.     

Formation of empty B19 parvovirus capsids by the truncated minor capsid protein.

We previously reported that empty capsids of B19 parvovirus were formed by the major capsid protein (VP2) alone expressed in a baculovirus system, but the minor capsid protein (VP1), longer by 227 amino acids, alone did not form empty capsids. We report here further investigations of the constraints on capsid formation by truncated versions of VP1. Studies were performed with recombinant baculoviruses expressed in Sf9 cells. Severely shortened VP1, extended beyond the VP2 core sequence by about 70 amino acids of the unique region, formed capsids normal in appearance; longer versions of VP1 also formed capsids but did so progressively less efficiently and produced capsids of more markedly dysmorphic appearance as the VP1-unique region was lengthened.