The Laboratory of Structural Biology Research seeks to elucidate structure-function-assembly relationships of macromolecular complexes. To this end, cryo-electron microscopy and cryo-electron tomography are integrated with other experimental approaches. Systems currently under study include virus assembly and maturation; amyloid filaments and prions; and several macromolecular complexes.
The structural basis of virus replication has been a central interest of the LSBR with particular focus on the roles of conformational changes in regulating two critical steps in the cycle - assembly and maturation of nucleocapsids; and recognition and entry of susceptible cells.
Capsid structure and assembly of Hepatitis B Virus. The HBV virion consists of a DNA genome enclosed within an icosahedral capsid, in turn enclosed within a lipoprotein envelope. Two forms of the capsid protein are expressed, one of which – core antigen (cAg) – assembles into capsids; the other - e-antigen (eAg) - remains soluble. eAg is a marker for the chronic form of the disease that leads in many cases to liver carcinoma. With P. Wingfield (PEL-NIAMS), we have investigated many aspects of this capsid: its structure; its two sizes corresponding to triangulation numbers of T=3 and T=4; and the conformational epitopes that are distributed over the capsid surface. eAg has a short N-terminal propeptide that induces a radically altered mode of dimerization relative to cAg - a 140° relative rotation between the two subunits. This switch precludes capsid assembly and engenders a distinct antigenic repertoire, explaining why the two antigens are cross-reactive at the T-cell level (through sequence identity) but not at the B-cell level (through conformation).
Assembly and maturation of herpes simplex virus type 1. Herpesviruses, a family of animal viruses with large (125 - 250 kbp) linear DNA genomes, are highly diversified in terms of host range; nevertheless, their virions conform to a common architecture. The genome is confined at high density within a thick-walled icosahedral capsid with T=16. Multiple viral glycoproteins are implanted in the envelope. Between the capsid and the envelope is a compartment called the tegument that accommodates viral proteins destined for delivery into a host cell. A strong body of evidence now supports the hypothesis that herpesvirus capsids and those of tailed bacteriophages stem from a distant common ancestor, whereas their radically different infection apparatuses - envelope on one hand and tail on the other - reflect co-evolution with divergent hosts. We have carried out a series of cryo-EM studies (with J. Brown, U. Virginia) addressing the molecular components of herpesvirus capsids and the mechanisms that regulate their assembly. Our most recent projects have targeted the building blocks used to assemble the procapsid; the location of the maturational protease in the mature capsid (mostly, under the vertices); and the structure of the Primary Enveloped Virion – the particle produced when the mature, DNA-filled capsid buds out of the nucleus where it assembles into the perinuclear space.
Assembly and maturation of double-stranded DNA bacteriophages. In capsid assembly and maturation, this class of viruses has in common: icosahedral capsids with a 12-fold portal complex substituting for a pentamer of capsid protein at one vertex; the fold of the capsid protein; initial assembly of a precursor particle tha matures via a radical conformational change; and packaging of DNA to high density by a viral motor protein, the terminase. Assembly proceeds in three phases: first, a precursor procapsid is formed; next, the genome is packaged. The final phase—maturation, which overlaps with packaging—consists of programmed structural changes that convert the provirion into an infectious, virion. In many but not all systems, maturation is controlled by the activity of a viral protease. We have worked on several phages including: T7, for which the concentric spool model for packaged DNA was conclusively demonstrated; HK97, in which a pathway of steps was found to guide maturation; phiKZ, which was found to encapsidate a large cylindrical “inner body” in addition to the genome; and T4, which binds a reinforcement protein to the outer surface of the mature capsid. Our most recent work is directed towards exceptionally large and complex phages of this kind.
Assembly and maturation of a double-stranded RNA virus, cystovirus phi6. DsRNA viruses share unusual properties in terms of multiple nestd capsids, the unorthodox icosahedral T-number of 2, and use of the innermost capsid as a compartment for genome replication and transcription. We have studied phi6 with L. Mindich (Rutgers), and characterized the novel fold of its inner capsid protein and the maturational transformation that its procapsid undergoes.
Assembly and Maturation of HIV. We currently focus on the modes of action of maturation inhibitors and on the Rev protein that mediates the nuclear export of genomic RNAs. Retroviral capsids are unusual in that they assemble inside the maturing virion, not in the cytoplasm or nucleus. Capsid protein is incorporated into the immature virion in a spherical shell of the Gag polyprotein. After budding off from the host cell, the viral protease dissects Gag into its matrix (MA), capsid (CA), and nucleocapsid (NC) domains and two spacer peptides, SP1 and SP2. The CA subunits assemble into the conical capsid, housing the RNA and NC. Maturation inhibitors (MI) deny the protease access to its cleavage site in the SP1 peptide. We have studied the mode of action of the MI Beviramat (BVM), with E. Freed, NCI. We found that BVM-treated virions retain a Gag-derived shell that lacks the innermost layer that is associated with NC/RNA. Another inhibition involves the viral integrase IN. We found that this enzyme also has an important role in capsid assembly (with A. Engelman, Harvard). Certain IN mutations and compounds (allosteric IN inhibitors, ALLINIs) produce virions that contain "eccentric condensates outside empty capsids. The same mutations and ALLINIs greatly increase the fraction of virions with malformed capsids.
HIV Rev protein mediates nuclear export of viral mRNAs, an essential step in HIV replication. In this process, Rev oligomerizes in association with the RNA motif, the Rev response element (RRE). This complex engages the nuclear export machinery of the host cell. Crystallographic work on Rev was long thwarted by its tendency to aggregate but we were able to devise a monoclonal antibody whose Fab forms a stable complex with Rev, and solve these co-crystals (with P. Wingfield, NIAMS and D. Stuart, Oxford). We subsequently extended the resolution of this analysis and solved crystals with four different space groups. All revealed the same structure for the Rev monomer, but the crossing angle of the Rev dimer varies widely. We also performed cryo-EM studies of helical tubes that Rev assembles into in vitro. They exhibit polymorphism, with the tube diameter varying between 11 nm and 13 nm. These variations in tube width correlated with the variations in crossing-angle seen in the crystals. Our data revealed a third interface between Revs which offers an explanation for how the arrangement of Rev subunits is matched to the A-shaped architecture of the RRE in export-active complexes.
Amyloids and Prions.
Amyloids are polymers of aberrantly folded proteins distinguished by their cross-beta structures. Accumulation of amyloid occurs in some 20 human diseases, including Alzheimer's, type-2 diabetes, and rheumatoid arthritis. Amyloids fall into two broad categories: infectious and non infectious. Infectious amyloids are called prions. We started studying yeast prion structures in 1998 (with R. Wickner, NIDDK), focusing initially on Ure2p, a negative regulator of nitrogen catabolism. Its N-terminal (“amyloid”) domain is responsible for prionogenesis, while its C-terminal domain is involved in regulating gene expression. In the wild-type protein, the N-domain is unfolded and in the prion, it forms amyloid fibrils surrounded by the C-domains that retain their native fold but are inactivated by steric blocking. We (with A. Kajava, CNRS) proposed the parallel superpleated beta-structure model for the amyloid backbone. This envisages arrays of parallel beta-sheets generated by stacking monomers with beta-serpentine folds. Similar structures are good candidates for some other amyloid fibrils. Ongoing work is aimed at testing and refining this model and investigating fibril polymorphism
We also investigate amyloid fibrillation of the protein alfa-synuclein (aS) associated with Parkinson disease (PD), a neurodegenerative disease affecting motor function. The140 aa-long protein is alfa-helical when associated with lipid and a random coil in solution. In fibril formation, the protein polymerizes into cross-beta structure. We have studied aS fibrils assembled in vitro from recombinant protein by cryo-EM in our laboratory and by dark-field STEM at the Brookhaven STEM facility (with J. Wall). Like those of other aS researchers our fibrils are polymorphic. We focused on a twisting fibril with an axial repeat of 77 nm. Their reconstructed cross-section showed two asymmetrically associated protofibrils. STEM mass-per-length measurements gave a value equivalent to two subunits per 0.47 nm axial rise, i.e. one subunit per protofibril, consistent with a superpleated structure (see above). The STEM images showed two thread-like densities running along each fibril that we interpret as ladders of metal ions. Similar threads were observed after doping metal-free fibrils with copper. These observations support the idea that metal binding promotes fibrillation in PD.
We study several large complexes, aiming to achieve insight into their structures, how they are assembled, and how they are regulated. Our first major commitment targeted the unfoldase-assisted proteases, ClpAP and ClpXP. More recently, our focus has turned to encapsulin, a bacterial iron-storage protein, and retinoschisin, a junctional extracellular protein in the retina.
Retinoschisin, a retinal junction protein. Retinoschisin (RS1) is required to maintain the structural and functional integrity of the retina. Mutations in RS1 lead to vision impairment in young males, in a condition called X-linked retinoschisis (XLRS. From prior work, RS1 was thought to form an octamer, each subunit having a discoidin (DS) domain and an RS1 domain. We used cryo-EM to determine the structure of RS1 at 4 Å resolution, finding the complex to consist of two octameric rings paired back-to-back. The DS domains form the rings while the RS1 domains occupy central positions. Our model is consistent with known intra-molecular and intermolecular disulfides. Interfaces within and between rings accommodate residues implicated in XLRS, indicating that correct assembly of the 16-mer is required for a correctly constituted junction. RS1 is an extracellular protein that appears to couple adjacent retinal membranes together through multiple octamer-octamer contacts. With P. Sieving, NEI.
Encapsulin and its role in iron homeostasis. Cells compartmentalize materials and enzymatic reactions to increase metabolic efficiency. For this purpose, bacteria rely to a large extent on protein-bound nanocompartments. Encapsulins are a recently discovered class of nanocompartments. We have determined the structure of encapsulins isolated from the bacterium Myxococcus xanthus (with E. Hoiczyk). It consists of a protein shell assembled from the EncA protein and three internal proteins (EncB; EncC; EncD). Recombinant EncA self-assembles into an icosahedral shell 32 nm in diameter, built from 180 subunits. Intriguingly, EncA has the fold first observed in the HK97 capsid and thought, until the recent work on encapsulins, to be exclusive to phages and herpesviruses. Native M. xanthus particles have iron-rich cores. Functionally, they resemble ferritins, cage-like iron storage proteins, but with a massively greater capacity (30,000 Fe atoms vs. 3,000 in ferritin). Physiological data reveal that few nanocompartments are assembled during vegetative growth, but they increase five-fold upon starvation, protecting cells from oxidative stress through iron sequestration.
Methodological Developments and Infrastructure.
Most projects undertaken in the LSBR include EM and image processing as major components. We have a long-standing practice of developing and applying novel methods in both areas, including programs for processing data to facilitate high resolution reconstructions from cryo-EM data; algorithms for symmetry detection; for the estimation of resolution; cryo-electron tomograms; and the Bsoft program developed and maintained by Bernard Heymann. In 1997, the LSBR was one of the first two laboratories to calculate 3D density maps of "single particles" to resolutions higher than 10Å, revealing alpha-helical sub-structure. This effort is ongoing andresolutions continue to improve. Our best resolutions to date are at the 4Å level and appear to be limited primarily by conformational heterogeneity. We also developed “4-dimensional electron microscopy” whereby images of dynamic complexes are sorted into classes which are reconstructed and then assembled into a time-ordered sequence, producing movies. We have also developed “bubblegram” imaging, in which the susceptibility of certain proteins to radiation damage allows them to be localized in the interior of macromolecular complexes.
We have a JEOL 2200FS electron microscope equipped with a field-emission gun, an in-column energy filter, a Gatan K2 camera, and a phase plate. We also have a 300-keV FEI Polara with a field-emission gun and a Falcon 2 camera. Starting in 2003, we implemented cryo-electron tomography and this line of investigation continues to be very productive. Looking ahead, we (NIAMS) are one of the four Institutes who have combined resources to purchase a Titan Krios which is expected to be delivered and commissioned by the end of this year (2017).