The Mechanism of Viral Replication. Structure of Replication Complexes of Encephalomyocarditis Virus

Abstract

The structure of the purified replicative intermediate of encephalomyocarditis virus was determined by electron microscopy. Approximately 80% of the replicative intermediate complexes were characterized by a filament of double-stranded RNA of widely variable length, which had a "bush" of single-stranded RNA at one end. In many examples one or more additional singlestranded bushes were appended internally to the doublestranded RNA filament. These results support the view that before deproteinization, replicative intermediate contains little if any double-stranded RNA.Double-stranded RNA is formed in vivo during the replication of picorna and other RNA-containing viruses. This material has been implicated in the viral inhibition of host-cell protein biosynthesis, and it may play a role in the induction of interferon. While early work suggested that double-stranded RNA formed an integral part of the RNA replication complex (termed "replicative intermediate," RI) this idea has more recently been questioned (1-3). The possibility has been raised that double-stranded RNA may be formed only as a byproduct of replication (e.g., "replicative form," RF), and its presence in RI has been termed an artifact that occurs during purification. A similar view has been expressed regarding the role of double-stranded RNA in the replication of RNA bacteriophages (1,2). This theory has recently been confirmed for R17 phage by the demonstration that the replicating complexes, in particular those containing RI, are composed almost entirely of single-stranded RNA (4). Double-stranded RNA was formed in these structures only as a result of heating or deproteinization. The electron microscopic techniques used in that study have now been applied to the replication of encephalomyocarditis (EMC) virus, and the results are described in the present communication. As in the case of R17 RNA replication, the issue can be stated in terms of two conflicting models for the structure of RI. The older model (model A, Fig. 1), involving double-stranded RNA, and the newer singlestranded model (model B) predict slightly different structures for purified (deproteinized) RI. In model A, deproteinized RI (RI-A') contains a long "backbone" of double-stranded RNA of constant length with two free ends, and one or more "bushes" of single-stranded RNA appended internally; in the newer model (model B), however, deproteinized RI (RI-B') contains a shorter double-stranded region of variable length with only one free end, a bush at the other end, and any number of bushes in between. These two structures, RI-A' and RI-B', are readily distinguishable in the electron microscope by recently developed techniques (4). The results of the present analysis show that the structure of EMC RI is identical to the hypothetical RI-B'. From this it may be concluded that RI contains little or no double-stranded RNA before deproteinization. This result implies that the only extensively double-stranded viral RNA in infected cells must be in the form of RF. METHQDS EMC virus an...

Reaction of poliovirus RNAs with antibodies to double-stranded RNA demonstrated by an immunochemical binding assay

“…Recent electron microscope studies have shown variability in the size of double-stranded segments in the RI of encephalomyocarditis virus (18). In our experiments, the proportion of RI in sucrose gradients that reacted with the antibodies increased as the sedimentation constant decreased toward 18S.…”

An analysis of tobacco mosaic virus replicative structures synthesized in vitro

“…Although the RNA backbone of RI molecules is believed to be ss in vivo, it assumes a ds-form upon extraction and isolation (35,10). RF is considered to comprise fully doublestranded (ds) RNA molecules consisting of one full-length genomic (plus strand) molecule and one full-length complementary (minus strand) molecule.…”

Adsorption of DNA molecules to different support films

“…The fact that with the 'diffusion' method of Gordon (1973), who treats the surface of mica with aluminium chloride, and with the method of Highton & Whitfield (1974) no such 'thick and short' regions are found, is no argument against this hypothesis, since the supports used in their procedures bind nucleic acid molecules more weakly than does aluminium-beryllium. Also, the methods described earlier by Koller et al (1969), Dubochet et al (1971 and Thach & Thach (1973) cannot be taken into consideration in this connection, because of a possible unwinding interaction between the different ligands used and DNA is not excluded.…”

“…However, in order to obtain information with improved resolution, it is necessary to spread the nucleic acid strands without carrier protein and to adsorb the molecules to a smooth ('low noise') support film. Several protein-free nucleic acid preparation techniques have been described (Nanninga et al, 1972;Thach & Thach, 1973;Highton & Whitfield, 1974; for review Koller et al, 1971). All these techniques, however, use carbon as supports.…”