Investigation of RNA structure in satellite panicum mosaic virus.

Three new crystal forms of satellite panicum mosaic virus (SPMV) were grown and their structures solved from X-ray diffraction data using molecular replacement techniques. The crystals were grown under conditions of pH and ionic strength that were appreciably different then those used for the original structure determination. In rhombohedral crystals grown at pH 8.5 and low ionic strength PEG 3350 solutions, Fourier syntheses revealed segments, ten amino acid residues long, of amino-terminal polypeptides not previously seen, as well as masses of electron density within concavities on the interior of the capsid, which appeared in the neighborhoods of icosahedral five- and threefold axes. The densities were compatible with secondary structural domains of RNA, and they included a segment of double helical RNA of about four to five base pairs oriented, at least approximately, along the fivefold axes. The distribution of RNA observed for SPMV appears to be distinctly different than the encapsidated nucleic acid conformation previously suggested for another satellite virus, satellite tobacco mosaic virus. This study further shows that analysis of viruses in crystals grown under different chemical conditions may reveal additional information regarding the structure of encapsidated RNA.

Crystallization of Brome mosaic virus and T = 1 Brome mosaic virus particles following a structural transition.

Brome mosaic virus (BMV), a T = 3 icosahedral plant virus, can be dissociated into coat protein subunits and subunit oligomers at pH 7.5 in the presence of concentrated salts. We have found that during the course of this treatment the coat protein subunits are cleaved, presumably by plant cell proteases still present in the preparation, between amino acids 35 and 36. The truncated protein subunits will then reorganize into T = 1 icosahedral particles and can be crystallized from sodium malonate. Quasi elastic light scattering and atomic force microscopy results suggest that the transition from T = 3 to T = 1 particles can occur by separate pathways, dissociation into coat protein subunits and oligomers and reassembly into T = 1 particles, or direct condensation of the T = 3 virions to T = 1 particles with the shedding of hexameric capsomeres. The latter process has been directly visualized using atomic force microscopy. Native T = 3 virions have been crystallized in several different crystal forms, but neither a rhombohedral form nor either of two orthorhombic forms diffract beyond about 3.4 A. Tetragonal crystals of the T = 1 particles, however, diffract to at least 2.5 A resolution. Evidence suggests that the T = 1 particles are more structurally uniform and ordered than are native T = 3 virions. A variety of anomalous virus particles having diverse sizes have been visualized in preparations of BMV used for crystallization. In some cases these aberrant particles are incorporated into growing crystals where they are frequently responsible for defect formation.

Structural transitions of satellite tobacco mosaic virus particles.

Satellite tobacco mosaic virus (STMV) can undergo at least two physical transitions that significantly alter its mechanical and structural characteristics. At high pH the 17-nm STMV particles expand radially by about 5 A to yield particles having diameters of about 18 nm. This pH-induced transition is further promoted by aging of the virions and degradation of the RNA, so that swollen particles ultimately appear even at neutral pH. While the native 17-nm particles crystallize as orthorhombic or monoclinic crystals which diffract to high resolution (1.8 A), the enlarged 18-nm particles crystallize in a cubic form which diffracts to no better than 5 A. In the transition, not only do the capsid protein subunits move radially outward, but the helical RNA segments with which they interact do as well. This is noteworthy because it demonstrates that the RNA and the protein shell are capable of coordinated movement, and that neither structure is rigidly defined or independent of the other. Using atomic force microscopy, it can be shown that STMV particles, upon drying, lose their mechanical rigidity and undergo deformation. Virions initially 17 nm in diameter shrink to more uniform final sizes than do 18 nm, initially swollen particles. This transition appears to be irreversible, as the particles do not reassume their former size nor structural rigidity upon rehydration. Evidence is also presented that preparations of native virus and their crystals are naturally somewhat heterogeneous and contain a variety of particles of anomalous size.

Biophysical studies on the RNA cores of satellite tobacco mosaic virus.

Satellite tobacco mosaic virus (STMV) was probed using a variety of proteases. Consequences of the degradation were analyzed using gel electrophoresis, quasi-elastic light scattering (QELS), and atomic force microscopy (AFM). Proteolysis rates of 30 minutes for complete degradation of the protein capsid, up to many hours, were investigated. With each protease, degradation of virions 17 nm in diameter was shown by QELS to result in particles of 10 nm diameter, which is that of the RNA core observed in the virion by x-ray diffraction analysis. This was verified by direct visualization with atomic force microscopy. Using QELS, it was further shown that freshly prepared RNA cores remain as individual, stable, 10-nm condensed particles for 12 to 24 h. Clusters of particles then formed, followed by very large aggregates of 500 to 1000 nm diameter. AFM showed that the aggregates were composed of groups of the condensed RNA cores and were not due to unfolding of the nucleic acid. No unfolding of the core particles into extended conformation was seen by AFM until the samples were heated well beyond 90 degrees C. Mass spectrometry of RNA core particles revealed the presence of a major polypeptide whose amino acid sequence corresponded to residues 2 through 25 of the coat protein. Amino acids 13 through 25 were previously observed to be in direct contact with the RNA and are presumably protected from protease digestion. Low resolution difference Fourier analyses indicated the courses of the remainders of the amino terminal strands (amino acids 2-12) in intact virions. Any individual strand appears to have several choices of path, which accounts for the observed disorder at high resolution. These positively charged strands, serving as virtual polyamines, engage the helical segments of RNA. The intimate association of amino acid residues 2 through 25 with RNA likely contributes to the stability of the condensed conformation of the nucleic acid cores.

Satellite tobacco mosaic virus RNA: structure and implications for assembly.

The initial appearance of 45% of the single-stranded RNA of satellite tobacco mosaic virus in electron density maps suggested the entire RNA conformation could be delineated. Subsequent work has localized nearly 80% of the RNA as stem-loop elements. Connection of the stem-loops in the most efficient manner produces a persuasive model for the encapsidated RNA. This arrangement has significant implications for virus assembly and for the essential role of RNA.

Refined structure of desmodium yellow mottle tymovirus at 2.7 A resolution.

Desmodium yellow mottle virus is a 28 nm diameter, T=3 icosahedral plant virus of the tymovirus group. Its structure has been solved to a resolution of 2.7 A using X-ray diffraction analysis based on molecular replacement and phase extension methods. The final R value was 0.151 (R(free)=0.159) for 134,454 independent reflections. The folding of the polypeptide backbone is nearly identical with that of turnip yellow mosaic virus, as is the arrangement of subunits in the virus capsid. However, a major difference in the disposition of the amino-terminal ends of the subunits was observed. In turnip yellow mosaic virus, those from the B and C subunits comprising the hexameric capsomeres formed an annulus about the interior of the capsomere, while the corresponding N termini of the pentameric capsomere A subunits were not visible at all in electron density maps. In Desmodium yellow mottle tymovirus, amino termini from the A and B subunits combine to form the annuli, thereby resulting in a much strengthened association between the two types of capsomeres and an, apparently, more stable capsid. The first 13 residues of the C subunit were invisible in electron density maps. Two ordered fragments of single-stranded RNA, seven and two nucleotides in length, were observed. The ordered water structure of the virus particle was delineated and required 95 solvent molecules per protein subunit.

Comparison of the conformations of two intact monoclonal antibodies with hinges.

Structures of two intact monoclonal antibodies were solved by X-ray diffraction analysis revealing, in both cases, the dispositions of all segments, as well as the structures of the hinge polypeptides. An IgG1, whose antigen is the drug phenobarbital, assumed a completely different conformation when compared with an IgG2a specific for canine lymphoma cells. Though neither IgG displays global two-fold symmetry, both maintain two pseudo dyad axes, one relating Fab segments, and the other the halves of the Fc. In both IgGs, the Fc segment is obliquely disposed with respect to the plane of the Fabs, making an angle of 128 degrees in the IgG2a, and 107 degrees in the IgG1. Hinge angles of the IgG1 are notably different at 78 degrees and 123 degrees, and unique as well from IgG2a values of 66 degrees and 113 degrees. Elbow angles within the IgG1 Fabs are the same at 155 degrees, but non-identical in IgG2a where they took on values of 143 degrees and 159 degrees. The IgG2a has an angle of 172 degrees between Fabs so that it exhibits a "distorted T" shape, whereas that angle in the IgG1 is a much more acute 115 degrees producing a "distorted Y". CH2 domains appear, in both antibodies, to be the most independently mobile of the paired IgG domains. This perhaps reflects their importance in modulating effector functions through exposure of binding loci on the CH2, at the CH2-CH3 interface, and on lower hinge polypeptides. Hinges in both antibodies contain disulfide-linked cores bounded by fluid regions above and below, which provide mobility to the Fabs and Fc respectively. The conformations seen in these two structures are undoubtedly only two among many but they illustrate the modes of flexibility inherent to the IgG architecture.

Refined structure of satellite tobacco mosaic virus at 1.8 A resolution.

The molecular structure of satellite tobacco mosaic virus (STMV) has been refined to 1.8 A resolution using X-ray diffraction data collected from crystals grown in microgravity. The final R value was 0.179 and Rfree was 0.184 for 219,086 independent reflections. The final model of the asymmetric unit contained amino acid residues 13 to 159 of a coat protein monomer, 21 nucleotides, a sulfate ion, and 168 water molecules. The nucleotides were visualized as 30 helical segments of nine base-pairs with an additional base stacked at each 3' end, plus a "free" nucleotide, not belonging to the helical segments, but firmly bound by the protein. Sulfate ions are located exactly on 5-fold axes and each is coordinated by ten asparagine side-chains. Of the 10,080 structural waters, 168 per asymmetric unit, about 20% serve to bridge the macromolecular components at protein-protein and protein-nucleic acid interfaces. Binding of RNA to the protein involves some salt linkages, particularly to the phosphate of the free nucleotide, but the major contribution is from an intricate network of hydrogen bonds. There are numerous water molecules in the RNA-protein interface, many serving as intermediate hydrogen bond bridges. The sugar-phosphate backbone contributes most of the donors and acceptors for the RNA. The helical RNA conformation is nearest that of A form DNA. The central region of a helical segment is most extensively involved in contacts with protein, and exhibits low thermal parameters which increase dramatically toward the ends. The visible RNA represents approximately 59% of the total nucleic acid in the virion and is derived from the single-stranded genome, which has folded upon itself to form helical segments. Linking of the helices and the free nucleotides in a contiguous and efficient manner severely restricts the disposition of the remaining, unseen nucleic acid. Using the remaining nucleotides it is possible to fold the RNA according to motifs that provide a periodic distribution of RNA structural elements compatible with the icosahedrally symmetrical arrangement seen in the crystallographic structure. The intimate relationship between protein and nucleic acid in STMV suggests an assembly pathway based on the cooperative and coordinated co-condensation of RNA with capsid protein dimers.

Crystallization of xylanase from Erwinia chrysanthemi: influence of heat and polymeric substrate.

Xylanase from the bacterial plant pathogen Erwinia chrysanthemi (E.C. 3.2.1.8), expressed in E. coli, has been crystallized for X-ray diffraction analysis both in the presence and the absence of its polymeric substrate 4-O-methyl glucuronoxylan. In all cases it was found that the quality, time of appearance, and reproducibility of both the native and complex crystals were significantly enhanced by heating of the protein to 323 K prior to dispensing the crystallization trials. Crystals of the native protein are ideal for X-ray diffraction analysis, producing Bragg reflections beyond 1.5 A resolution with virtually no degradation with time. The native crystals are in space group P2(1), with a = 39.33, b = 49.46, c = 90.85 A and beta = 101.58 degrees. Other polymorphs have also been obtained and their cell parameters determined. Crystallization of the enzyme in the presence of polymeric substrate yields two distinctly different crystals at different concentrations of the xylan. These are thought to be complexes of the protein with stable products of the enzymatic reaction. A similar result had been obtained previously with pancreatic alpha-amylase and its substrate glycogen.

The open conformation of a Pseudomonas lipase.

BACKGROUND: . The interfacial activation of lipases results primarily from conformational changes in the enzymes which expose the active site and provide a hydrophobic surface for interaction with the lipid substrate. Comparison of the crystallization conditions used and the structures observed for a variety of lipases suggests that the enzyme conformation is dependent on solution conditions. Pseudomonas cepacia lipase (PCL) was crystallized in conditions from which the open, active conformation of the enzyme was expected. Its three-dimensional structure was determined independently in three different laboratories and was compared with the previously reported closed conformations of the closely related lipases from Pseudomonas glumae (PGL) and Chromobacterium viscosum (CVL). These structures provide new insights into the function of this commercially important family of lipases. RESULTS: . The three independent structures of PCL superimpose with only small differences in the mainchain conformations. As expected, the observed conformation reveals a catalytic site exposed to the solvent. Superposition of PCL with the PGL and CVL structures indicates that the rearrangement from the closed to the open conformation involves three loops. The largest movement involves a 40 residue stretch, within which a helical segment moves to afford access to the catalytic site. A hydrophobic cleft that is presumed to be the lipid binding site is formed around the active site. CONCLUSIONS: . The interfacial activation of Pseudomonas lipases involves conformational rearrangements of surface loops and appears to conform to models of activation deduced from the structures of fungal and mammalian lipases. Factors controlling the conformational rearrangement are not understood, but a comparison of crystallization conditions and observed conformation suggests that the conformation of the protein is determined by the solution conditions, perhaps by the dielectric constant.

Refined structure of an intact IgG2a monoclonal antibody.

The structure of an intact, anti-canine lymphoma monoclonal antibody (Mab231) was determined by molecular replacement and refined in a triclinic cell to an R-value of 20.9%, using synchrotron diffraction data from 2.8 to 20 A resolution. All segments of the antibody, including the hinge region and carbohydrate component, are visible in electron density maps. There is no overall symmetry to the antibody, as the Fc is disposed in an entirely oblique manner with respect to the Fabs. The CH2 and CH3 domains do, however, possess a nearly exact, local 2-fold relationship. The Fab segments are related by a second, independent, local dyad axis, exact only with respect to constant domains. Variable domains exhibit no symmetry relationship as a consequence of the 16 degrees difference in Fab elbow angles. Variable domain pair associations VL:VH for the Fabs are virtually the same, and corresponding CDRs of the two Fabs also are nearly identical in structure. CDR-H3 displays the greatest difference. Hypervariable loops of both Fabs are involved in contacts with symmetry-related Fc segments at the CH2-CH3 switch junction, suggesting a "complex" structure. The hinge segment connecting Fabs with the Fc is quite extended and exhibits thermal factors indicative of a high degree of mobility. It consists of a well-defined upper hinge that partially maintains dyad symmetry and a fairly rigid core bounded above and below by fluid polypeptides that provide segmental flexibility. This structure represents the first visualization by X-ray analysis of a murine Fc segment, and its CH2 domains exhibit substantial rigid body conformational changes with respect to the human Fc used as an initial molecular replacement model. The oligosaccharides were found by difference Fourier syntheses to be very similar to those of the free human Fc fragment, although differences are present in the terminal residues. The detailed structure of the IgG presented here, and the distribution of effector binding sites, appears consistent with effector activation mechanisms involving translocation and/or aggregation of the Fc following antigen binding by the Fabs.

Crystal structure of turnip yellow mosaic virus.

The structure of turnip yellow mosaic virus (TYMV) has been solved to 3.2 A resolution and an R-value of 18.7%. The structure is consistent with models based on low resolution X-ray and electron microscopy studies, with pentameric and hexameric protein aggregates protruding from the surface and forming deep valleys at the quasi three-fold axes. The N-terminal 26 residues of the A-subunit are disordered, while those of the B- and C-subunits are seen to interact around the interior of the quasi six-fold cluster where they form an annulus. The three histidine residues of each protein subunit are located in the interior and accessible for interaction with the RNA genome. The appearance of the interior surface of the virus capsid, along with buried surface area calculations, suggest that a pentameric unit is lost during decapsidation.

Structural comparison of the plant satellite viruses.

Detailed structures are now available for three plant satellite viruses, satellite tobacco necrosis virus (STNV), satellite tobacco mosaic virus (STMV), and satellite panicum mosaic virus (SPMV). It is, therefore, possible to compare the tertiary structure of viral protein subunits, their quaternary interactions, and the interactions of protein subunits with the RNA genome. This analysis indicates that, in spite of common function and preservation of a "jelly-roll" motif in the protein monomer, the three viruses are remarkably different. The differences include the arrangement of secondary structural elements, interactions of adjacent subunits, and the disposition of subunits relative to icosahedral symmetry axes. In each of the three viruses, however, the narrow end of the jelly roll forms fivefold contacts. The fivefold protein interactions are organized about a Ca2+ ion for STNV, an anion for STMV, and, apparently, neither of these for SPMV. Low-resolution neutron diffraction studies using H2O/D2O solvent contrast variation revealed the general location of the RNA genome within the STNV. In the case of SPMV, regions of electron density on the interior of the capsid could be assigned to RNA, although it was not possible to model the nucleic acid. Only for STMV was nucleic acid visible in election density maps, and this was manifested as double-helical RNA segments associated with each coat protein dimer. The observations presented here provide no support for any common evolutionary relationship.

Characterization of crystals of satellite panicum mosaic virus.

Satellite panicum mosaic virus (SPMV) has been purified from pearl millet and obtained in a variety of different crystal forms, at least two of which appear suitable for high resolution X-ray diffraction analysis. The first is of cubic space group P4(2)32 with a = b = c = 183.1 A and two virus particles in the unit cell. The second is of a primitive orthorhombic space group with a = 166.1 A, b = 266.7 A and c = 269.1 A, with four virus particles in the unit cell. While the cubic crystal has as its asymmetric unit one twelfth of the icosahedron, or five capsid protein subunits, the asymmetric unit of the orthorhombic crystals is an entire particle. The cubic crystals diffract to at least 2.8 A resolution. We have also succeeded in crystallizing, but not yet characterizing the master virus, PMV.

Refined molecular structure of pig pancreatic alpha-amylase at 2.1 A resolution.

The structure of pig pancreatic alpha-amylase has been determined by X-ray diffraction analysis using multiple isomorphous replacement in a crystal of space group P2(1)2(1)2(1) (a = 70.6 A, b = 114.8 A, c = 118.8 A) containing nearly 75% solvent. The structure was refined by simulated annealing and Powell minimization, as monitored by 2Fo-Fc difference Fourier syntheses, to a conventional R of 0.168 at 2.1 A resolution. The final model consists of all 496 amino acid residues, a chloride and a calcium ion, 145 water molecules and an endogenous disaccharide molecule that contiguously links protein molecules related by the 2(1) crystallographic operator along x. The protein is composed of a large domain (amino acid residues 1 to 403) featuring a central alpha ta-barrel of eight parallel strands and connecting helices with a prominent excursion between strand beta 3 and helix alpha 3 (amino acid residues 100 to 168). The final 93 amino acid residues at the carboxyl terminus form a second small domain consisting of a compact Greek key beta-barrel. The domains are tightly associated through hydrophobic interfaces. The beta 3/alpha 3 excursion and portions of the central alpha/beta-barrel provide four protein ligands to the tightly bound Ca ion; three water molecules complete the coordination. The Cl- ion is bound within one end of the alpha/beta-barrel by two arginine residues in a manner suggesting a plausible mechanism for its allosteric activation of the enzyme. A crystalline complex of the pancreatic alpha-amylase with alpha-cyclodextrin, a cyclic substrate analog of six glucose residues, reveals, in difference Fourier maps, three unique binding sites. One of the alpha-cyclodextrin sites is near the center of the long polysaccharide binding cleft that traverses one end of the alpha/beta-barrel, another is at the extreme of this cleft. By symmetry this can also be considered as two half sites located at the extremes of the active site cleft. This latter alpha-cyclodextrin displaces the endogenous disaccharide when it binds and, along with the first sugar ring, delineates the extended starch binding site. The third alpha-cyclodextrin binds at an "accessory site" near the edge of the protein and is quite distant from the polysaccharide binding cleft. Its presence explains the multivalency of alpha-amylase binding to dextrins in solution. The extended active site cleft is formed by large, sweeping, connecting loops at one end of the alpha/beta-barrel. These include three sequence segments that are highly conserved among alpha-amylases.(ABSTRACT TRUNCATED AT 400 WORDS)

Three-dimensional structure of satellite tobacco mosaic virus at 2.9 A resolution.

The crystal structure of satellite tobacco mosaic virus (STMV) has been solved by a combination of multiple isomorphous replacement and molecular replacement methods and refined at 2.9 A resolution to a conventional R-factor of 0.215. STMV, a T = 1 icosahedral virus, is the smallest whose structure has been determined. The coat protein is an eight-stranded "Swiss roll" beta-barrel with an amino-terminal strand that extends away from the beta-barrel by more than 60 A. This strand is primarily responsible for quaternary interactions within the capsid. The most arresting feature of the virus structure is the intimate association of each capsid protein dimer with a Watson-Crick base-paired segment of RNA double helix on the interior of the virion. The icosahedral 2-fold axis of each dimer pair is coincident with that of the central base-pair of each helical RNA segment whose helical axis is along the edge of the icosahedron. The helical RNA segments are seven base-pairs in length with a stacked base at each 3' end so that a total of 16 nucleotides is clearly visible. The character of the RNA helix is somewhat different than any of the canonical forms. Assuming full occupancy, then approximately 45% of the total RNA genome is present in the electron density map. The close association of capsid with highly structured nucleic acid suggests that assembly of STMV is likely to be a highly co-operative process involving both protein and RNA. The nucleic acid is distributed within the virion with a high degree of order. The capsid protein is a true double helical RNA binding protein and a number of prominent interactions between protein and RNA can be clearly seen.

Double-helical RNA in satellite tobacco mosaic virus.

Satellite tobacco mosaic virus (STMV) is the spherical satellite to an obligatory rod-shaped helper tobacco mosaic virus (TMV), which is required for replication. STMV has 60 protein subunits of M(r) 17,500 on a T = 1 icosahedral capsid containing a single-stranded RNA genome of 1,059 bases. STMV appears similar to another virus, STNV, but is approximately 20 per cent smaller. It shows no amino-acid homology or immunological cross-reactivity with either STNV or its host TMV. Here we report the X-ray crystal structure of STMV, which shows that the coat protein of STMV contains a 'Swiss roll' beta-barrel. An amino-terminal strand extends more than 60A and is primarily responsible for quaternary interactions. Each capsid dimer is associated with a segment of genomic RNA double helix comprising seven base pairs. The dyad of each protein dimer is coincident with that of the central base pair of the associated RNA segment whose helix axis is directed along an icosahedral edge. Protein-nucleic acid interactions are extensive. The RNA helices, which have additional stacked bases at their 3' termini, differ significantly from canonical nucleic acid helical forms.

The three-dimensional structure of an intact monoclonal antibody for canine lymphoma.

Crystal structures of Fab antibody fragments determined by X-ray diffraction characteristically feature four-domain, beta-barrel arrangements. A human antibody Fc fragment has also been found to have four beta-barrel domains. The structures of a few intact antibodies have been solved: in two myeloma proteins, the flexible hinge regions that connect the Fc to the Fab segments were deleted so the molecules were non-functional, structurally restrained, T-shaped antibodies; a third antibody, Kol, had no hinge residues missing but the Fc region was sufficiently disordered that it was not possible to relate its disposition accurately with respect to the Fab components. Here we report the structure at 3.5 A resolution of an IgG2a antitumour monoclonal antibody which contains an intact hinge region and was solved in a triclinic crystal by molecular replacement using known Fc and Fab fragments. The antibody is asymmetric, reflecting its dynamic character. There are two local, apparently independent, dyads in the molecule. One relates the heavy chains in the Fc, the other relates the constant domains of the Fabs. The variable domains are not related by this 2-fold axis because of the different Fab elbow angles of 159 degrees and 143 degrees. The Fc has assumed an asymmetric, oblique orientation with respect to loosely tethered yet almost collinear Fabs. Our study enables the two antigen-binding segments as well as the Fc portion of a functional molecule to be visualized and illustrates the flexibility of these immune response proteins.

Synthesis and broad-spectrum antiviral activity of 7,8-dihydro-7-methyl-8-thioxoguanosine.

2,6,8-Trichloro-7-methylpurine (3) was converted to 2-chloro-8,9-dihydro-7-methyl-8-thioxopurin-6(1H)-one (5) by utilizing the difference in reactivity of the 2-, 6-, and 8-positions in the trichloropurine ring system to nucleophilic displacement. Compound 5 was subsequently glycosylated with 1-O-acetyl-2,3,5-tri-O-benzoyl-D-ribofuranose according to the Vorbrüggen procedure to yield 2-chloro-8,9-dihydro-7-methyl-9-(2,3,5-tri-O-benzoyl-beta-D-ribofuranosy l)-8- thioxopurin-6(1H)-one (6). Removal of the benzoyl protecting groups, followed by amination of 7 with liquid ammonia at 150 degrees C, gave 7,8-dihydro-7-methyl-8-thioxoguanosine (2). The structure of compound 2 was confirmed by X-ray crystallographic analysis. Compounds 1 (7,8-dihydro-7-methyl-8-oxoguanosine) and 2 were evaluated for activity in various animal virus infection models. Against banzi, Semliki Forest, and San Angelo viruses in mice, 2 was highly active when administered before virus inoculation.