Search

The binding properties of seven mouse monoclonal antibodies (McAbs) raised against cowpea mosaic virus (CPMV) were characterized by conventional and inhibition enzyme-linked immunosorbent assay (ELISA) technique. McAb binding to CPMV on electron microscope (EM) grids was also assayed with gold-labeled anti-mouse antibodies. Two of the seven McAbs (5B2 and 10B7) were found to bind tighter to CPMV than the others in the inhibition ELISA and the EM assay. F(ab) fragments from both of these McAbs were prepared and complexed with CPMV in solution. Electron micrographs of flash frozen (vitrified) samples of native CPMV and CPMV complexed with F(ab) fragments from McAbs 5B2 and 10B7 as well as IgGs from 5B2 were recorded and reconstructions were computed at 23 A resolution for the CPMV/F(ab) complexes and 30 A resolution for the CPMV/IgG complex. Structures of all three complexes clearly displayed the F(ab) fragments distributed with icosahedral symmetry on the surface of CPMV. The IgG bound in a monodentate fashion with only one F(ab) attached to the virus surface. F(ab) fragments from 5B2 and 10B7 bound to nearly identical positions. The refined 2.8 A X-ray structure of CPMV was used to identify the roughly 30 amino acids covered by the F(ab) fragments. The "footprint" spans a subunit interface and appears spatially similar to antigenic site 3B on poliovirus. In a previous, preliminary report of the CPMV/F(ab) 5B2 complex (Wang et al., 1992, Nature 355, 275-278) the wrong enantiomorph of the reconstruction was chosen. This was corrected and, since the F(ab) binds close to a mirror plane, the change in the footprint was minor.

Crystals of cowpea chlorotic mottle virus (CCMV) that diffract X-rays to 3.1 A resolution were grown in a succinate-PEG solution buffered at pH 3.3. The crystals are in space group P2(1)2(1)2(1) with unit cell dimensions of a = 381.26 A, b = 381.26 A, and c = 408.59 A. Four particles occupy the unit cell, placing a single virion in the crystallographic asymmetric unit. Diffraction intensities measured from 196 films collected at the Cornell High Energy Synchrotron Source accounted for 55% of the theoretically possible data to 3.2 A. Unit cell dimensions and rotation function analyses of the X-ray data revealed that the particles were organized in a pseudo-tetragonal relationship with the pseudo-fourfold axis along the crystal c axis. Analysis of electron micrographs of two-dimensional crystals of CCMV revealed a remarkable similarity between these and planes of particles perpendicular to the crystallographic c axis in the three-dimensional crystal.

Tetraviruses are small, non-enveloped, RNA viruses that exclusively infect lepidopteran insects. Their particles comprise 240 copies of a single capsid protein precursor (CP), which undergoes autoproteolytic cleavage during maturation. The molecular mechanisms of capsid assembly and maturation are well understood, but little is known about the viral infectious lifecycle due to a lack of tissue culture cell lines that are susceptible to tetravirus infection. We show here that binding and entry of the alphatetravirus, Helicoverpa armigera stunt virus (HaSV), is triggered by alkaline pH. At pH 9.0, wild-type HaSV virus particles undergo conformational changes that induce membrane-lytic activity and binding to Spodoptera frugiperda Sf9 cells. Binding is followed by entry and infection, with virus replication complexes detected by immunofluorescence microscopy within 2h post-infection and the CP after 12h. HaSV particles produced in S. frugiperda Sf9 cells are infectious. Helicoverpa armigera larval virus biofeed assays showed that pre-treatment with the V-ATPase inhibitor, Bafilomycin A1, resulted in a 50% decrease in larval mortality and stunting, while incubation of virus particles at pH 9.0 prior to infection restored infectivity. Together, these data show that HaSV, and likely other tetraviruses, requires the alkaline environment of the lepidopteran larval midgut for binding and entry into host cells., (Copyright © 2016 Elsevier Inc. All rights reserved.)

Although vesicular stomatitis virus (VSV) neurovirulence and pathogenicity in rodents have been well studied, little is known about VSV pathogenicity in non-human primates. To address this question, we measured VSV viremia, shedding, and neurovirulence in macaques. Following intranasal inoculation, macaques shed minimal recombinant VSV (rVSV) in nasal washes for 1 day post-inoculation; viremia was not detected. Following intranasal inoculation of macaques, wild type (wt) VSV, rVSV, and two rVSV-HIV vectors showed no evidence of spread to CNS tissues. However, macaques inoculated intrathalamically with wt VSV developed severe neurological disease. One of four macaques receiving rVSV developed clinical and histological signs similar to the wt group, while the remaining three macaques in this group and all of the macaques in the rVSV-HIV vector groups showed no clinical signs of disease and reduced severity of histopathology compared to the wt group. The implications of these findings for rVSV vaccine development are discussed.

The assembly of monomeric protein subunits into a viral capsid is a finely tuned molecular process. In response to subtle changes in environmental conditions, this supramolecular complex can dramatically reorganize. Defining the forces that control this structure and the cooperative action of subunits has implications for biology and nanotechnology. The small icosahedral RNA tetravirus family members Nudaurelia omega capensis (NomegaV) and Helicoverpa armigera stunt virus (HaSV) can be purified as provirions, and maturation to capsids can be induced by a drop in pH. In this study, a comparison of capsid secondary structure using FT-IR revealed that the procapsid has more alpha-helical content than the capsid, supporting the proposal that helix to coil transition may be important for maturation. The dynamic properties of the two states were probed using limited proteolysis and peptide mass mapping to identify regions of significant flexibility. Interestingly, the initial sites of protease cleavage were the N and C terminal domains that are internal in high-resolution models, and to inter-subunit surfaces. Further comparison of the two particle forms using FT-IR revealed that in response to thermal stress, the provirion disassembles and unfolds in a cooperative manner over a narrow temperature range (approximately 5 degrees C). Paradoxically, the capsid form, which is stable in a wide range of pH and ionic conditions and is more resistant to proteolysis, responds to thermal stress at a lower temperature than the procapsid form. This suggests that a metastable state is the end product of assembly.

Large-scale reorganization of protein interactions characterizes many biological processes, yet few systems are accessible to biophysical studies that display this property. The capsid protein of Nudaurelia capensis omega Virus (NomegaV) has previously been characterized in two dramatically different T = 4 quasi-equivalent assembly states when expressed as virus-like particles (VLPs) in a baculovirus system. The procapsid (pH 7), is round, porous, and approximately 450 A in diameter. It converts, in vitro, to the capsid form at pH 5 and the capsid is sealed shut, shaped like an icosahedron, has a maximum diameter of 410 A and undergoes an autocatalytic cleavage at residue 570. Residues 571-644, the gamma peptide, remain associated with the particle and are partially ordered. The interconversion of these states has been previously studied by solution X-ray scattering, electron cryo microscopy (CryoEM), and site-directed mutagenesis. The particle structures appear equivalent in authentic virions and the low pH form of the expressed and assembled protein. Previously, and before the discovery of the multiple morphological forms of the VLPs, we reported the X-ray structure of authentic NomegaV at 2.8 A resolution. These coordinates defined the fold of the protein but were not refined at the time because of technical issues associated with the approximately 2.5 million reflection data set. We now report the refined, authentic virus structure that has added 29 residues to the original model and allows the description of the chemistry of molecular switching for T = 4 capsid formation and the multiple morphological forms. The amino and carboxy termini are internal, predominantly helical, and disordered to different degrees in the four structurally independent subunits; however, the refined structure shows significantly more ordered residues in this region, particularly at the amino end of the B subunit that is now seen to invade space occupied by the A subunits. These additional residues revealed a previously unnoticed strong interaction between the pentameric, gamma peptide helices of the A and B subunits that are largely proximal to the quasi-6-fold axes. One C-terminal helix is ordered in the C and D subunits and stabilizes a flat interaction in two interfaces between the protein monomers while the other, quasi-equivalent, interactions are bent. As this helix is arginine rich, the comparable, disordered region in the A and B subunits probably interacts with RNA. One of the subunit-subunit interfaces has an unusual arrangement of carboxylate side chains. Based on this observation, we propose a mechanism for the control of the pH-dependent transitions of the virus particle.

Ordered nucleic acid in an icosahedral virus was first visualized in the X-ray structure of the Picorna-like plant virus, Bean pod mottle virus (BPMV). Virus particles containing the 3500 nucleotide segment of the BPMV bipartite RNA genome (middle component) had nearly 20% of the genome ordered. Here we report the refined structures of the middle component, bottom component (particles containing the 5800 nucleotide segment of the genome), and top component (empty particles of BPMV capsid protein). The bottom component particles contain ordered RNA in the same location as middle component. Although the ordered RNA density in both nucleoprotein particles is the average of the contents of 60 icosahedral asymmetric units, both nucleoprotein components show that the base density for the first two nucleotides is predominantly purine, while the next five appear to be predominantly pyrimidine. The empty capsid demonstrates that RNA dictates the order of the N-terminal 19 residues of the large subunit because these residues are invisible in the top component.

The X-ray crystal structure of Cowpea chlorotic mottle bromovirus (CCMV) revealed a unique tubular structure formed by the interaction of the N-termini from six coat protein subunits at each three-fold axis of the assembled virion. This structure, termed the beta-hexamer, consists of six short beta-strands. The beta-hexamer was postulated to play a critical role in the assembly and stability of the virion by stabilizing hexameric capsomers. Mutational analyses of the beta-hexamer structure, utilizing both in vitro and in vivo assembly assays, demonstrate that this structure is not required for virion formation devoid of nucleic acids in vitro or for RNA-containing virions in vivo. However, the beta-hexamer structure does contribute to virion stability in vitro and modulates disease expression in vivo. These results support a model for CCMV assembly through pentamer intermediates.

Betanodaviruses are causative agents of neurological disorders in several species of fish. We cloned and sequenced the RNA2 segment of two grouper viruses isolated from Epinephelus malabaricus (malabaricus grouper nervous necrosis virus, MGNNV) and Epinephelus lanceolatus (dragon grouper nervous necrosis virus, DGNNV). The sequences of the two RNAs were 99% identical and comparison with previously sequenced RNA2 segments of fish nodaviruses striped jack nervous necrosis virus, Atlantic halibut virus, sea bass encephalitis virus, and greasy grouper nervous necrosis virus (GGNNV) revealed that MGNNV and DGNNV were most closely related to GGNNV. No correlation of sequence with geographical habitat was detected. The MGNNV coat protein, the gene product of RNA2, was expressed in Sf21 cells with a recombinant baculovirus system and virus-like particles (VLPs) spontaneously formed. Two types of VLPs were observed: a slower sedimenting particle was RNase-sensitive and stain-permeable, while the faster sedimenting particle survived RNase treatment and was not stain-permeable. An image reconstruction of the latter, obtained with electron cryomicroscopy data, revealed a morphology consistent with T = 3 quasi-symmetry but with features significantly different from insect nodavirus structures at the same resolution. This assembly system allows the first biophysical comparisons of fish and insect nodavirus structure, assembly, and stability.

Comoviruses are a group of plant viruses in the picornavirus superfamily. The type member of comoviruses, cowpea mosaic virus (CPMV), was crystallized in the cubic space group I23, a = 317 A and the hexagonal space group P6(1)22, a = 451 A, c = 1038 A. Structures of three closely similar nucleoprotein particles were determined in the cubic form. The roughly 300-A capsid was similar to the picornavirus capsid displaying a pseudo T = 3 (P = 3) surface lattice. The three beta-sandwich domains adopt two orientations, one with the long axis radial and the other two with the long axes tangential in reference to the capsid sphere. T = 3 viruses display one or the other of these two orientations. The CPMV capsid was permeable to cesium ions, leading to a disturbance of the beta-annulus inside a channel-like structure, suggesting an ion channel. The hexagonal crystal form diffracted X rays to 3 A resolution, despite the large unit cell. The large ( approximately 200 A) solvent channels in the lattice allow exchange of CPMV cognate Fab fragments. As an initial step in the structure determination of the CPMV/Fab complex, the P6(1)22 crystal structure was solved by molecular replacement with the CPMV model determined in the cubic cell., (Copyright 1999 Academic Press.)

Previous studies showed that the HIV-1 envelope (Env) protein was not incorporated into vesicular stomatitis virus (VSV) virions unless its cytoplasmic tail was replaced with that of the VSV glycoprotein (G). To determine whether the G tail provided a positive incorporation signal for Env, or if sequences in the Env tail prevented incorporation, we generated mutants of Env with its 150-amino-acid tail shortened to 29, 10, or 3 amino acids (Envtr mutants). Cells infected with VSV recombinants expressing these proteins or an Env-G tail hybrid showed similar amounts of Env protein at the surface. The Env-G tail hybrid or the Envtr3 mutant were incorporated at the highest levels into budding VSV virions. In contrast, the Envtr29 or Envtr10 mutants were incorporated poorly. These results defined a signal preventing incorporation within the 10 membrane-proximal amino acids of the Env tail. Confocal microscopy revealed that this signal functioned by causing localization of human immunodeficiency virus type 1 Env to plasma membrane domains distinct from the VSV budding sites, where VSV proteins were concentrated., (Copyright 1998 Academic Press.)

Cryoelectron microscopy and three-dimensional image reconstruction analysis has been used to determine the structure of native and in vitro assembled cowpea chlorotic mottle virus (CCMV) virions and capsids to 25-A resolution. Purified CCMV coat protein was used in conjunction with in vitro transcribed viral RNAs to assemble RNA 1 only, RNA 2 only, RNA 3/4 only, and empty (RNA lacking) virions. The image reconstructions demonstrate that the in vitro assembled CCMV virions are morphologically indistinguishable from native virions purified from infected plants. The viral RNA (vRNA) is packaged similarly within the different types of virions. The centers of all assembled particles are generally devoid of density and the vRNA packs against the interior surface of the virion shell. The vRNA appears to adopt an ordered conformation at each of the quasi-threefold axes.

HK97 is a temperate dsDNA bacteriophage of Escherichia coli that is structurally similar to phage lambda, with an icosahedral head of triangulation (7) number 7. Although the capsids of several large dsDNA phages have been studied extensively using a variety of biophysical approaches, no high-resolution structure is available. We have grown crystals of mature but empty bacteriophage HK97 capsids that diffract to at least 3.5 A using synchrotron radiation. The HK97 Head II crystals are the first capsid crystals from a dsDNA bacteriophage that diffract X-rays to high resolution. A capsid precursor (prohead) was made in vivo by expressing capsid proteins in E. coli. This prohead was purified, converted to Head II in vitro, and used to grow crystals. The empty Head II has the same form as the mature HK97 capsid, but without DNA. The crystals were grown in a mixture of ammonium sulfate and PEG 8000, directly in an X-ray capillary to minimize crystal handling. The unit cell is monoclinic, with dimensions a = 580 A, b = 625 A, c = 790 A, beta = 90.0 degrees and two particles per unit cell.

The three-dimensional structure of cucumber mosaic virus (CMV) was analyzed at 23 A resolution by cryoelectron microscopy and image reconstruction, demonstrating structural similarity to cowpea chlorotic mottle virus (CCMV), another member of the Bromoviridae family. The CMV structure was determined at 8 A resolution by X-ray crystallography with phases determined by single isomorphous replacement and refined by fivefold noncrystallographic symmetry averaging. The X-ray structure agreed with the electron microscopy reconstruction; the electron density is consistent with beta-barrel subunits arranged with T = 3 quasi-symmetry in an orientation similar to that observed in CCMV. Strong density surrounding the icosahedral threefold axes (quasi sixfold axes in the T = 3 particle) between 80 and 100 A from the particle center formed a cylinder of radius 11 A, similar to the density observed in the same region of CCMV. This density corresponds to the beta-annulus of CCMV, which differentiates hexamers from pentamers and determines the formation of the T = 3 particles. The CMV and CCMV amino acid sequences were aligned, providing information (based on the CCMV atomic model) about the probable distribution of residues in the three-dimensional structure of CMV.

Expression of a soluble CD4 molecule (sCD4-KDEL) containing a specific retention signal for the endoplasmic reticulum was shown previously to block propagation of the HIV-1MN prototype strain in a transformed T cell line. However, the virus present in HIV-1-infected individuals is more closely represented by primary HIV-1 isolates which, unlike the HIV-1MN strain, have not been adapted to growth in cell lines. To determine if sCD4-KDEL could block replication of primary isolates we used the PM1 cell line that has been shown to propagate primary isolates without adaptation. Here we show that the replication of four primary HIV-1 isolates was strongly inhibited in PM1 cells that expressed sCD4-KDEL under control of the HIV-1 LTR. Infection with primary HIV-1 isolates induced sCD4-KDEL expression driven by the LTR, HIV-1 spread was dramatically reduced, and reverse transcriptase activity in the cell culture supernatants was greatly diminished sCD4-KDEL, therefore, represents a potent inhibitor of HIV-1 replication for gene therapy-based approaches for the treatment of AIDS.

We have expressed the gene encoding the coat protein (CP) of Nudaurelia capensis omega virus in insect cells with a baculovirus vector. Expression of CP resulted in formation of virus-like particles (VLPs) having a size consistent with the T = 4 quasi-symmetry observed for native virions. This is the first demonstration of assembly for a T = 4 particle, with chemically identical subunits present in four distinct environments, by heterologous expression. Initial yields of VLPs were low, and an efficient one step nondenaturing procedure involving separation on discontinuous glycerol gradients was developed. Using this method, VLPs were obtained in quantities sufficient for further characterization. Electron microscopic observation revealed 40-nm particles that were morphologically similar to native virus. SDS-PAGE revealed that these particles were composed of a 62-kDa major protein and a minor 70-kDa protein. Pulse-chase experiments revealed that the larger species was processed into the smaller one very slowly over the course of an infection. It was also determined that this cleavage was apparently dependent on release of these particles from the cell. Furthermore, these particles were found to encapsidate the polyhedrin promoter-directed CP mRNA with an apparently striking degree of specificity and selectivity. This investigation establishes that a specific encapsidation signal exists within the CP coding sequences and that the components required for reconstructing most, if not all, steps in the morphogenesis of this virus can be accomplished in baculovirus-infected cells. The results presented here are consistent with the belief that similar biological strategies are utilized by T = 3 nodaviruses and T = 4 tetraviruses in particle assembly.

The capsid protein (CP) of alfalfa mosaic virus (AIMV) is required for viral replication when susceptible plants are inoculated with purified viral genomic RNA. The discovery of AIMV CP in the zinc activated RNA-dependent RNA polymerase complex prompted our further investigation of AIMV virions and the potential involvement of AIMV CP in metal binding. AIMV CP, isolated from nucleoprotein components, fractionated into four distinct ionic species when purified by cation exchange fast protein liquid chromatography. The CP existed as zinc complexed homodimers, metal-free homodimers, and two forms of proteolyzed heterodimers, as analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis, gel filtration chromatography, amino-terminal sequencing, and atomic absorption spectroscopy. Although the relative amounts of proteolyzed heterodimers varied, the ratio of zinc complexed homodimers to metal-free homodimers (1:10) was constant between virus and protein isolations for the strains 425 and WISC14. Purified metal-free and zinc-complexed homodimers could be interconverted in vitro by incubation with zinc chloride or with the metal chelator, sodium diethyldithiocarbamate (NaDDC). The potential role of zinc in AIMV nucleoprotein structure and infectivity was investigated by treatment of the virions with NaDDC. Electron microscopy and sucrose density gradient studies failed to detect any gross structural changes for zinc depleted virus; however, a decrease in infectivity was observed with local lesion leaf assays, suggesting a functional role for zinc in viral replication.

It has recently been shown that cowpea plants can be infected with a cowpea mosaic virus (CPMV) chimera containing an antigenic site from foot-and-mouth disease virus (Usha et al., Virology 197, 366-374, 1993). Analysis of progeny RNA produced during such an infection has revealed that the inserted sequence is rapidly lost during serial passaging, probably by a process of homologous recombination. Using the information gained from this analysis, we have redesigned the chimeras in such a way that they are now genetically stable. The modified constructs have been used to obtain large quantities of purified virus particles expressing epitopes derived from human rhinovirus 14 (HRV-14) and human immunodeficiency virus type 1 (HIV-1). The chimeric virus particles possess the antigenic properties of the inserted sequence and, in the case of the HRV-14-derived construct, it has been shown that the inserted epitope is immunogenic in rabbits. These results demonstrate that CPMV can be used as a high-yielding system for the presentation of foreign peptide sequences.

To investigate if cowpea mosaic virus (CPMV) particles can be used to express foreign protein sequences, oligonucleotides encoding an epitope derived from VP1 of foot-and-mouth disease virus (FMDV) were cloned into the region of the CPMV genome encoding the small (S) coat protein. The chimeras were designed so that the foreign sequence was expressed either as an insertion or as a replacement for part of the wild-type sequence. While RNA from both chimeras was able to replicate in cowpea protoplasts only the construct containing the FMDV sequence as an insertion was able to direct capsid formation and infect whole cowpea plants. The modified S protein produced in plants infected with the insertion derivative reacted with FMDV-specific antiserum. These results show that CPMV can be used as an antigen presentation system and raises the possibility of producing vaccines in plants using a RNA virus-based vector.

Isolated tobacco streak virus strain mild (TSV-M) top nucleoprotein component (TV) was crystallized by vapor diffusion with polyethylene glycol (PEG) and methyl pentanediol. The morphology of the crystal suggested a hexagonal space group, however, the crystals were disordered as analyzed by X-ray diffraction. Capsid protein from TSV (strains M and white clover) was treated with trypsin to remove 87 amino terminal residues. The modified capsid protein was purified by ion exchange chromatography and crystallized in both hexagonal and triclinic forms. Hexagonal crystals grown by vapor diffusion in PEG diffract X rays to 4.5 A. The crystals, in the space group P6(2) or its enantiomorph, possess unit cell dimensions of a = 98.3 A, c = 108.7 A and probably contain 12 dimers/unit cell. Triclinic crystals grown by vapor diffusion in ammonium sulfate diffracted X rays to 2.4 A resolution when exposed to X-ray synchrotron radiation. The crystals have unit cell dimensions of a = 60.1 A, b = 77.5 A, c = 73.9 A with alpha = 67.8 degrees, beta = 130 degrees, and gamma = 106 degrees and probably contain 4 dimers/unit cell.

We have determined the nucleotide sequence of the RNA2 segment of the Nudaurelia capensis omega virus genome. It was found to consist of 2448 nucleotides and contained one long open reading frame (ORF) encoding the 644 residue capsid protein. The deduced amino acid sequence of this protein reveals a positively charged amino terminus, a characteristic exhibited by several other viral capsid proteins, that is thought to be important for interactions between the capsid and the genomic RNA. There are 366 and 150 bases of untranslated sequence on the 5' and 3' ends, respectively. The ORF encoding the capsid protein initiates at the second AUG from the 5' end. The 5' proximal AUG specifies a short ORF (30 codons) which terminates 1 base before the initiation codon for the coat protein. Our analysis also revealed the presence of a second, previously unidentified polypeptide associated with Nudaurelia capensis omega virus particles. The amino terminal sequence of this protein corresponds to a portion of the long ORF beginning at codon 571. The lack of an initiation codon near this sequence indicates that the small polypeptide is most likely produced as a carboxy terminal cleavage product from a 70-kDa capsid protein precursor, yielding the previously identified 62-kDa protein and the 8-kDa protein that we have observed. The putative cleavage site would be at an Asn/Phe pair, somewhat resembling known cleavage sites (Asn/Ala) in the T = 3 Nodaviridae. In addition, we have found that there is also a second polypeptide similar in size to that from Nudaurelia capensis omega virus associated with particles of Nudaurelia capensis beta virus, the type member of Tetraviridae.

An in vitro murine T cell proliferation assay was used to determine whether an antigenic epitope(s) recognized by enterovirus-immune T cells is held in common between plant comoviruses and human enteroviruses. Splenocytes isolated from C3H/HeJ mice infected with coxsackievirus B3 (CVB3) proliferated in vitro not only against a variety of enterovirus (CVB2, CVB3, CVB6, CVA16, PV1) antigens, but against comovirus (CPMV, BPMV) antigens as well. Splenocytes from mice inoculated with bean pod mottle virus (BPMV) also proliferated in response to comoviral and enteroviral antigens in vitro. However, if the viral inocula were highly purified prior to inoculation, then the splenocyte response was generated only against the group used to inoculate, suggesting that the epitope shared between the comoviruses and the enteroviruses resided in the nonstructural region. B (nonstructural) and M (structural) genomic segments of CPMV were translated in rabbit reticulocyte lysates and used as in vitro antigens. Splenocytes from mice inoculated with live CVB3 proliferated in response to the B-RNA-encoded but not the M-RNA-encoded polypeptides, confirming the nonstructural coding region location of the common epitope. Comparison of predicted amino acid sequences in the nonstructural coding regions of the comoviruses and picornaviruses suggested a potentially immunogenic linear epitope in protein 2C. The consensus peptide LEEKGI was synthezized and shown to be immunogenic for both BPMV- and CVB3-immune splenocytes.

Crystals of Cowpea mosaic virus displaying a high degree of order have been grown in two different forms. Hexagonal crystals, space group P6(1)22 or P6(5)22 with a = 450 A and c = 1038 A containing six virus particles per cell, have been grown in sodium citrate buffer at pH 4.9. Each particle has four nearest neighbors at 284 A. Cubic crystals, with space group I23 and a = 308 A , containing two particles per cell were grown from ammonium sulfate at pH 4.9. In the cubic crystals, the tetrahedral subgroup of the icosahedral particle symmetry is expressed as part of the crystal lattice, with each particle occupying tetrahedral crystallographic positions. Contacts of 268 A occur between particles in contact along threefold icosahedral symmetry axes.

Black beetle virus, propagated in cultured Drosophila cells, crystallized into rhombic dodecahedra which diffracted X rays to 3.0 A resolution. Center-to-center spacing of particles in the unit cell was 305 A, while the spherically averaged diameter obtained from small-angle X-ray scattering from virus in solution was 312 A. Low resolution diffraction patterns from single crystals showed that the protein subunits are distributed centrosymmetrically, while electron microscopy indicated the particles are icosahedral in shape. The size of the particle is sufficient to accommodate about 180 protein subunits (44 kDa) consistent with T = 3 quasisymmetry.

1H and 13C NMR studies on cowpea mosaic virus (CpMV) revealed that polyamines are present in the middle (M) and upper bottom (BU) components obtained by CsCl density gradient centrifugation but not in the top (T) component; the lower bottom (BL) component contains trace amounts of polyamine. Dialysis of the BL component against spermidine led to incorporation of spermidine which gave rise to NMR peaks very similar to those observed with the natural M and BU components. NMR results conclusively demonstrate that polyamines in the M and BU components of CpMV are exchangeable with cesium ions and the exchange process is pH dependent. They also provide experimental support for the hypothesis that the BU to BL conversion results from the displacement of polyamines and possibly other natural counter ions of the RNA by cesium ions [G. Bruening, (1977), In "Comprehensive Virology" (H. Fraenkel-Conrat and R. R. Wagner, eds.), Vol. 11, pp. 55-141. Plenum, New York]. No sharp peaks, attributable to mobile amino acid side chains, were seen in spectra of an intact CpMV particle or its empty protein shell (T component). 31P NMR spin-lattice relaxation time and nuclear Overhauser effect parameters, which are sensitive to high-frequency motions, suggest that the RNA and, when present, the bound polyamine undergo internal motions with correlation times in the nanosecond range.

The X-ray diffraction patterns of the four components of cowpea mosaic virus isolated from a cesium chloride gradient were measured, using film methods, to 30 A resolution. Diffraction patterns were analyzed by fitting computed two-shell spherical models to the observed data. The fitting procedure was applied to data to 80 A resolution to avoid the nonspherical contribution to the pattern observed at higher resolution. At pH 7.0 all four components displayed the same external spherically averaged radius of 140 +/-2 A. The lowest density component (top), which contains no RNA, was best modeled by an empty shell with an outer radius of 140 A and an inner radius of 101 +/- 3 A. The middle component, containing 27% RNA by weight, was modeled with a uniform electron density sphere. The bottom upper and bottom lower components, which are biologically identical but display different buoyant densities in cesium chloride solutions, were analyzed independently. The bottom upper component was best modeled with a 101 A inner (RNA containing) sphere of mean electron density 0.453e-/A(3) and a 101 to 140 A outer (protein containing) shell of electron density 0.410e-/A(3). The bottom lower component was fit with the same model except that the RNA containing region displayed a mean electron density of 0.459e-/A(3). The implications of the spherically averaged component structures for the protein structure, RNA and protein hydration, and cesium binding to RNA are discussed.

Tobacco streak virus (TSV) RNA and alfalfa mosaic virus (AIMV) RNA will replicate only if a few copies of their coat proteins are bound to the RNA. To understand this phenomenon experiments were performed to find unique features of the TSV and AIMV subunits. Atomic absorption analysis showed that TSV and AIMV contained substantial quantities of zinc in native virions (approximately one zinc atom per four protein subunits in TSV and one zinc atom per two protein subunits in AIMV), while other plant viruses tested did not. Treatment of TSV with a zinc-extracting reagent resulted in partial degradation of all the TSV nucleoprotein components, although the top component was most effected. The sequence (Cys X2 Cys X10 Cys X2 His) was found between residues 28 and 45 in the TSV primary structure and it is similar to a sequence found in several nucleic acid-binding, gene-regulatory proteins, most notably transcription factor IIIA from Xenopus laevis. TSV subunits were found to be extensively crosslinked within the virions. TSV and AIMV contain sequences rich in basic residues in the amino-terminal portion of the subunit (residues 51 to 72 in TSV and 1 to 26 in AIMV) and helical predictions suggested modes of protein-nucleic acid interactions in these regions similar to those proposed for histones. Two potential sites for glycosylation were identified near the amino terminus of the TSV sequence. Controlled treatment of TSV with trypsin removed 87 residues from the amino terminus and produced a monomer of cleaved protein, as analysed by SDS-PAGE.(ABSTRACT TRUNCATED AT 250 WORDS)