The architecture of adenoviruses: recent views and problems: Brief review

Shapes of minimal-energy DNA ropes condensed in confinement

Shapes of a single, long DNA molecule condensed in a confinement of a virus capsid are described as conformations optimizing a model free energy functional accounting for the interplay between the bending energy of the DNA and the surface energy of the DNA bundled in a “rope”. The rope is formed by bundled DNA brought together by (self-)attractive interactions. The conformations predicted by the model depend on the shape of the confinement, the total amount of the packed DNA but also on the relative contributions of the bending and surface energies. Some of the conformations found were not predicted previously, but many previously proposed DNA conformations, some of which are seemingly contradictory, were found as the solutions of the model. The results show that there are many possible packing conformations of the DNA and that the one which realizes in a particular virus depends on the capsid geometry and the nature of condensing agents.

Infectious pancreatic necrosis virus: identification of a VP3-containing ribonucleoprotein core structure and evidence for O-linked glycosylation of the capsid protein VP2

Virions of infectious pancreatic necrosis virus (IPNV) were completely disintegrated upon dialysis against salt-free buffers. Direct visualization of such preparations by electron microscopy revealed 5.0- to 6.5-nm-thick entangled filaments. By using a specific colloidal gold immunolabeling technique, these structures were shown to contain the viral protein VP3. Isolation by sucrose gradient centrifugation of the filaments, followed by serological analysis, demonstrated that the entire VP3 content of the virion was recovered together with the radiolabeled genomic material forming the unique threadlike ribonucleoprotein complexes. In a sensitive blotting assay, the outer capsid component of IPNV, i.e., the major structural protein VP2, was shown to specifically bind lectins recognizing sugar moieties of N-acetylgalactosamine, mannose, and fucose. Three established metabolic inhibitors of N-linked glycosylation did not prevent addition of sugar residues to virions, and enzymatic deglycosylation of isolated virions using N-glycosidase failed to remove sugar residues of VP2 recognized by lectins. However, gentle alkaline beta elimination clearly reduced the ability of lectins to recognize VP2. These results suggest that the glycosylation of VP2 is of the O-linked type when IPNV is propagated in RTG-2 cells.

Linear adenovirus DNA is organized into supercoiled domains in virus particles

Electron microscopic analysis of bis-psoralen crosslinked adenovirus type 5 virion DNA revealed supercoiled domains in an otherwise linear DNA. The existence of supercoiled arrangement in all the virion DNA was demonstrated by the sensitivity of Ad5 DNA in pentonless virus particles to the supercoiling-dependent endonucleolytic activity of Bal31 and S1 nucleases. These nucleases were found to cleave Ad5 virion DNA at specific sites. The observation of stable cleavage sites in the limit digestion of virion DNA by Bal31 suggests that cleavage sites represent boundaries of core proteins which impede the exonuclease activity of Bal31. These data suggest that specific arrangement of core proteins on Ad5 virion DNA. Based on this analysis we determined positions of core proteins in viral genome using indirect end labeling technique. The size of supercoiled domains of virion DNA was estimated by electron microscopy and also by boundaries of mutually exclusive Bal31 cleavage sites at limit digestion condition. Our data suggest each supercoiled domain is equal to about 12% of Ad5 genome length and about 8 loops can be accommodated in Ad5 virion. However sequences at two extreme ends of the viral genome were found to be outside of supercoiled domains. An interesting correlation between supercoiled domains and gene domains of Ad5 genome was noticed.

Comparison of the interactions of the adenovirus type 2 major core protein and its precursor with DNA

The interactions of the major core protein of adenovirus type 2 (Ad2) protein VII, and its precursor, protein pre-VII, with viral DNA, were studied using UV light induced crosslinking of 32P-labelled oligonucleotides to the proteins. Proteolytic fragments of these two proteins that contain DNA-binding domains were identified by virtue of their covalently attached, alkali-resistant 32P-radioactivity. The overall efficiency of crosslinking of protein pre-VII to DNA, in H2ts1 virions assembled at 39 degrees C, was comparable to that of the crosslinking of protein VII to DNA in Ad2 virions. However, a protease V8 fragment comprising the N-terminal half of protein pre-VII crosslinked to DNA at least ten times more efficiently than the corresponding N-terminal fragment of protein VII, which is truncated by the removal of 23 amino acids from the N-terminus of protein pre-VII during virion maturation.

Identification of proteins and protein domains that contact DNA within adenovirus nucleoprotein cores by ultraviolet light crosslinking of oligonucleotides 32P-labelled in vivo

A new approach to the identification of DNA binding proteins has been developed and used to study the DNA-protein interactions within the nucleoprotein core of subgroup C adenoviruses. Virions labelled in vivo with [32P]orthophosphate were exposed to ultraviolet light and the DNA digested by chemical or enzymatic methods. Labelled phosphoamino acids of the virion phosphoproteins were selectively hydrolysed by alkali, permitting proteins crosslinked to DNA to be identified by virtue of their covalently attached, 32P-labelled nucleotides. In parallel experiments, [3H]arginine-labelled virions were crosslinked by exposure to ultraviolet light and analysed by more conventional methods. The results indicate that proteins VII and V lie in close contact with viral DNA within the core. The compact arrangement of the nucleoprotein core appears to be capable of trapping protein VII molecules that are not covalently attached to DNA after exposure to ultraviolet light, suggesting that viral DNA might be wrapped around clusters of protein VII molecules. The domains of protein VII that lie in contact with DNA were identified by partial proteolytic mapping of the sites of covalent-attachment of the 32P-labelled oligonucleotides. The implications of these data for the nature of the interactions that mediate the packaging of viral DNA within the nucleoprotein core of adenovirions are discussed.

Ionic effects on the structure of nucleoprotein cores from adenovirus

Nucleoprotein cores, prepared from adenovirus type 5 with a deoxycholate/heat treatment, consist of the viral DNA and two major internal proteins. The core particles exhibit structural characteristics that are highly reproducible and dependent on their ionic environment. In low-ionic-strength buffer, the cores had a sedimentation coefficient of 180 S and appeared in the electron microscope as homogeneous particles with distinct centers from which numerous arms and loops radiated. Condensation of the cores was induced by Mg2+ or Ca2+ over the range 0 to 1 mM. The sedimentation coefficient increased monotonically with divalent cation concentration, reaching a maximum of 405 S in 1 mM Mg2+. A corresponding condensation in the core structure was observed by electron microscopy. Increasing concentrations of NaCl also produced a conformational change in the cores, with an almost linear increase in sedimentation velocity up to 274 S in 0.04 M NaCl. Between 0.05 and 1.0 M NaCl, the cores were insoluble. In 2.0 M NaCl, the cores were again soluble with an s20,w of 228 S. Under all ionic strength conditions in which the cores were soluble, both core proteins remained bound to the DNA.

The structure of the adenovirus capsid. I. An envelope model of hexon at 6 A resolution

The structure of hexon, the major coat protein of adenovirus, has been determined by X-ray crystallography. Electron density maps were obtained with phases based on five heavy-atom derivatives at 2.9 A resolution. The main experimental finding derives from a low resolution electron density map calculated at 6.0 A resolution, but based on phases determined by the multiple isomorphous replacement method at 2.9 A resolution. Hexon consists of three subunits together forming two major components of different morphological symmetry. A triangular top with three towers of density is superimposed on a more bulky pseudo-hexagonal base. The symmetry of the top is in accord with the trimeric nature of hexon, but that of the base derives from the molecular function, which is to provide a densely packed impenetrable protective outer layer for the virion. A close-packed array of hexons forms a planar facet of the icosahedral capsid, with the tops presenting a spiky appearance that is consistent with electron micrographs of the adenovirus capsid. Hexon is hollow, permitting it to occupy a larger volume than normal for the same quantity of protein. The polypeptide chain has been traced in the 2.9 A electron density map for several non-contiguous stretches, allowing C alpha co-ordinates to be measured for 820 out of the 967 amino acid residues. The overall folding pattern confirms the assignment of shape, but the lack of connectivity so far precludes its complete description. The modest amount of alpha-helix and beta-sheet present is in accord with spectroscopic results.

Interactions among the three adenovirus core proteins

Interactions among the three adenovirus core polypeptides V, VII, and mu were examined, using the reversible chemical cross-linker dithiobis(succinimidyl propionate) and two-dimensional polyacrylamide gel electrophoresis. Cross-linked species obtained from gradient-purified adenovirus type 2 cores were well represented among the cross-linked products of pentonless virions and crude core preparations. The more efficiently formed cross-linked core species were also identified with the arginine-specific cross-linker, p-azidophenyl glyoxal. In addition to dimers of polypeptides V and VII, efficient cross-linking of V to VII, V to mu, and VII to V to mu was detected in adenovirus cores. Notably absent were cross-linked species corresponding to higher multimers of polypeptide VII. A major core-capsid interaction appeared to be via the association of polypeptide V with a dimer of polypeptide VI.

Structural organization and polypeptide composition of the avian adenovirus core

CELO virus (fowl adenovirus 1) contained three core polypeptides of molecular weights 20,000, 12,000, and 9,500. The core was similar to that of human adenoviruses, with some evidence of compact subcore domains. Micrococcal nuclease digestion of CELO virus cores produced a smear of DNA fragments of gradually decreasing size, with no nucleosome subunit or repeat pattern. Moreover, when digested cores were analyzed without protease treatment, there was again no evidence of a nucleosome substructure; neither DNA fragments nor core proteins entered a 4% polyacrylamide gel. The organization of the core is thus quite unlike that of chromatin. Restriction endonuclease analysis of the DNA from digested cores showed that the right end was on the outside of the core. We suggest that adenovirus DNA is condensed into the core by cross-linking and neutralization by the core proteins, beginning with the packaging sequence at the center of the core and ending with the right end of the DNA on the outside.

Protein composition of adenovirus nucleoprotein complexes extracted from infected cells

A viral nucleoprotein complex was extracted from the nuclei of human cells 20 hr after infection with adenovirus type 2 or several of its temperature-sensitive mutants. In its sedimentation property, density in CsCl, and digestion pattern with micrococcal nuclease, the complex resembled viral cores. The polypeptides V, PVII, 11K, and 36K were found associated with this complex which is formed prior to or in the absence of virus assembly. The results suggest that this nucleoprotein complex is a direct precursor to virus assembly.

Structural studies of adenovirus type 2 by neutron and X-ray scattering

Small-angle neutron and X-ray scattering have been used to investigate various aspects of the structural organization of adenovirus type 2. Neutron scattering allows the determination of the radial distribution of DNA and protein, which because of the highly icosahedral form of the virus allows it to be described in terms of three icosahedral shells. X-ray scattering shows that the distance between the major coat proteins (hexons) in the capsid is 100 +/- A. Evidence was also observed for an organization in the nucleoprotein core that gives rise to a maximum in the X-ray scattering at 1/29 A-1.

Demonstration of stepwise coiling of nucleoprotein in adenovirus core by application of critical point drying

Mild destruction of a virus particle to observe the organized structure of the nucleoprotein complex in a virion was achieved by application of the critical point drying method. Adenovirus type 12 (ad12) virions have been treated by this method after the particles had been fixed with glutaraldehyde on an electron microscope grid. With 15 min prefixation, the capsids (shells) and the cores were in various stages of unfolding. The core was unfolded in the filamentous structure. The thickness of these filaments was 6.7, 13.3, +23 nm, or more. Some pictures showed that the thicker filaments consisted of super-coiling of two thinner filaments, for example two 6.7-nm filaments coiled up to give the 13.3-nm filaments. This suggests that the nucleoprotein complex of a circular double-stranded DNA and inner proteins of ad12 virus was folded in a stepwise fashion to produce the compacted form of the core.

Native molecular weight of adenovirus proteins: on the oligomeric structure of the fiber

Fluorescamine-modification of amino groups was used to eliminate the influence of basic charge on the final migration position of protein's in alkaline pH polyacrylamide gradient gel electrophoresis. As applied to adenovirus structural components, this type of analysis suggested the fiber to be composed of three identical subunits. The trimeric nature of both penton base and fiber therefore displaces the problem of symmetry mismatching to penton base and surrounding hexons at each vertex of the adenovirus icosahedron.

S1 sensitive sites in adenovirus DNA

S1 nuclease has been used as a probe for regions of DNA secondary structure in supercoiled recombinant plasmids containing adenovirus (Ad) DNA sequences. In the sequences examined two S1 sensitive sites were identified in the left-terminal 16.5% of Ad 12 DNA, one of which aligned approximately with an inverted repeat region in the DNA sequence. In addition an S1 sensitive site was dictated by a potential cruciform structure in the region of the Ad 2 major late promoter. In contrast to the expected cleavage site at the loop of the cruciform, cleavage occurred at the base of the stem in the region of the TATA box. All three S1 sensitive sites identified were more sensitive to S1 than the endogenous sites in the parent plasmids.