Down regulation of gene expression by the vaccinia virus D10 protein

Vaccinia virus genes are expressed in a sequential fashion, suggesting a role for negative as well as positive regulatory mechanisms. A potential down regulator of gene expression was mapped by transfection assays to vaccinia virus open reading frame D10, which encodes a protein with no previously known function. Inhibition was independent of the promoter type used for the reporter gene, indicating that the mechanism did not involve promoter sequence recognition. The inhibition was overcome, however, when the open reading frame of the reporter gene was preceded by the encephalomyocarditis virus internal ribosome entry site, which excludes the possibility of nonspecific metabolic or other antiviral effects and suggests that capped mRNAs or cap-dependent translation might be the target of the D10 product. The inducible overexpression of the D10 gene by a recombinant vaccinia virus severely inhibited viral protein synthesis, decreased the steady-state level of viral late mRNA, and blocked the formation of infectious virus.

NF-kappa B-mediated chromatin reconfiguration and transcriptional activation of the HIV-1 enhancer in vitro

NF-kappa B is a potent inducible transcription factor that regulates many genes in activated T cells. In this report we examined the ability of different subunits of NF-kappa B to enhance HIV-1 transcription in vitro with chromatin templates. We find that the p65 subunit of NF-kappa B is a strong transcriptional activator of nucleosome-assembled HIV-1 DNA, whereas p50 does not activate transcription, and that p65 activates transcription synergistically with Sp1 and distal HIV-1 enhancer-binding factors (LEF-1, Ets-1, and TFE-3). These effects were observed with chromatin, but not with nonchromatin templates. Furthermore, binding of either p50 or p65 with Sp1 induces rearrangement of the chromatin to a structure that resembles the one reported previously for integrated HIV-1 proviral DNA in vivo. These results suggest that p50 and Sp1 contribute to the establishment of the nucleosomal arrangement of the uninduced provirus in resting T cells, and that p65 activates transcription by recruitment of the RNA polymerase II transcriptional machinery to the chromatin-repressed basal promoter.

Modification of the cascade model for regulation of vaccinia virus gene expression: purification of a prereplicative, late-stage-specific transcription factor

In vivo and in vitro studies have provided evidence that vaccinia virus late gene transcription factors are intermediate gene products synthesized exclusively after DNA replication. Here, we describe an additional transcription factor (P3 factor) that stimulates late gene transcription between 10- and 40-fold but is made in the absence of viral DNA replication. P3 factor activity was not detected either in uninfected cells or in purified virions. A > 1,500-fold purification of P3 factor was achieved by column chromatography of cytoplasmic extracts prepared from cells infected with vaccinia virus in the presence of a DNA replication inhibitor. P3 factor was stage specific, since it could not substitute for early or intermediate transcription factors. Evidence that late stage-specific transcription factors are made both before and after DNA replication necessitates a modification of the cascade model for vaccinia virus gene regulation.

Overexpression, purification, and late transcription factor activity of the 17-kilodalton protein encoded by the vaccinia virus A1L gene

The A1L, A2L, and G8R open reading frames (ORFs) were previously shown by transfection assays to encode transactivators of late gene expression. We now present evidence that the 17-kDa protein product of the A1L gene can function in vitro as a transcription factor. Simultaneous overexpression of the transactivators was achieved by coinfecting HeLa cells with one recombinant vaccinia virus that encodes the bacteriophage T7 RNA polymerase and three recombinant vaccinia viruses that contain copies of A1L, A2L, and G8R ORFs regulated by T7 promoters. Extracts from the recombinant virus-infected cells exhibited greatly enhanced late in vitro transcription activity and served as a source of factors. The 17-kDa product of the A1L ORF represented approximately 8% of the ammonium sulfate-precipitated cell protein and copurified with a late transcription factor activity. The transcription factor activity could be specifically immunodepleted with immobilized antibody to the bacterially expressed A1L-encoded protein, providing additional evidence for its identity and role. A sequence encoding six consecutive histidines was added to the A1L ORF, which was then incorporated into the genome of a baculovirus expression vector. The 17-kDa protein, synthesized in insect cells and purified by binding to an Ni(2+)-chelating affinity column, could replace the vaccinia virus-overexpressed 17-kDa protein in transcription assays. In addition to the 17-kDa product of the A1L gene, which was named vaccinia virus late transcription factor 2, the proteins that stimulate specific transcription of late promoter-regulated templates included the viral multisubunit RNA polymerase, vaccinia virus late transcription factor 1 (the product of the G8R ORF), and at least one other partially purified protein.

Mutational analysis of a predicted zinc-binding motif in the 26-kilodalton protein encoded by the vaccinia virus A2L gene: correlation of zinc binding with late transcriptional transactivation activity

Transient transfection assays indicated that A2L is one of three vaccinia virus intermediate genes that are required for the transcriptional transactivation of viral late genes. We have expressed the A2L open reading frame in Escherichia coli and shown by blotting experiments that the 26-kDa protein binds zinc, a property predicted by the presence of a CX2CX13CX2C zinc finger motif. The specificity for zinc binding was demonstrated by competition with other metals. The role of the sequence motif in zinc binding was established by analysis of a series of mutations, including truncations and conservative single amino acid substitutions. Mutations that reduced zinc binding in vitro prevented the ability of A2L to transactivate late genes in vivo.

Transcription of viral late genes is dependent on expression of the viral intermediate gene G8R in cells infected with an inducible conditional-lethal mutant vaccinia virus

There are three temporal classes of vaccinia virus genes: early, intermediate, and late. The object of this study was to determine the effects on virus replication of regulating the expression of G8R, an intermediate gene that encodes a late transcription factor. We inserted the lac operator adjacent to the RNA start site of the G8R gene in a recombinant vaccinia virus that constitutively expresses the Escherichia coli lac repressor to make expression of the G8R gene dependent on the inducer isopropyl-beta-D-thiogalactopyranoside (IPTG). In case repression would not be complete, we also weakened the promoter of the G8R gene by making a single-nucleotide substitution designed to reduce its basal level of transcription. The mutant virus replicated well in the presence of the inducer, although synthesis of the G8R-encoded 30,000-M(r) protein was only 10% of that of the wild-type virus. In the absence of IPTG, (i) synthesis of the G8R protein was inhibited by more than 99% relative to that of the wild-type virus, (ii) synthesis of early and intermediate mRNAs appeared to be unaffected, (iii) intermediate proteins accumulated to higher than normal levels, (iv) synthesis of late mRNA and protein was reduced by about 90%, (v) viral DNA was replicated but incompletely resolved concatemeric molecules accumulated, (vi) not even the earliest stages of virion assembly were detectable by transmission electron microscopy, and (vii) virus yield under one-step growth conditions and plaque formation were 10(-3) and 10(-4) times the wild-type values, respectively. The defect in late gene expression could be overcome by transfection of a G8R gene that was not under lac operator control, as well as by addition of IPTG, further demonstrating the specificity of the repression. The correlation between decreased expression of the G8R intermediate gene and inhibition of late mRNA synthesis is consistent with the notion that the G8R product serves as an essential late transcription factor and supports a cascade mechanism of vaccinia virus gene regulation. In addition, the inducer-dependent vaccinia virus mutant provided a tool for selective inhibition of late gene expression while allowing synthesis of early and intermediate mRNAs and proteins.

Mutational analysis of the core, spacer, and initiator regions of vaccinia virus intermediate-class promoters

Activation of vaccinia virus late gene transcription is dependent on DNA replication and the expression of three genes: A1L, A2L, and G8R (J. G. Keck, C. J. Baldick, Jr., and B. Moss, Cell 61:801-809, 1990). To fully characterize the promoter elements of these trans-activator genes, we prepared more than 140 plasmid vectors containing natural and mutated DNA segments ligated to the Escherichia coli lacZ or chloramphenicol acetyltransferase reporter gene. Expression of the reporter genes occurred when the plasmids were transfected into vaccinia virus-infected cells and was enhanced when DNA replication was prevented, indicating that the A1L, A2L, and G8R promoters belong to the intermediate regulatory class. Deletional mutagenesis demonstrated that the regulatory elements of all three promoters extended between 20 and 30 nucleotides upstream of their RNA start sites. Single-base substitutions of the G8R promoter revealed two critical elements located from -26 to -13 (the core element) and -1 to +3 (the initiator element). Mutations in these regions drastically affected expression, as determined by beta-galactosidase and mRNA analyses. Additional mutations defined the TAAA sequence as the critical initiator element. The length, but not the nucleotide sequence, of the segment between the core and initiator regions was crucial. The requirement for the spacer to be 10 or 11 nucleotides was consistent with a single turn of a double helix. The A1L and A2L promoters resembled the G8R promoter, and mutations in the conserved bases had the predicted effects on expression. We concluded that the three intermediate promoters are composed of a 14-bp A+T-rich core sequence separated by one turn of the double helix from the TAAA initiator element.

Mouse hepatitis virus nucleocapsid protein-specific cytotoxic T lymphocytes are Ld restricted and specific for the carboxy terminus

Infection of mice with the JHM strain of mouse hepatitis virus (MHV) results in an acute encephalomyelitis associated with primary demyelination of the central nervous system. Efforts at understanding the components of the immune response in the development of chronic MHV-induced demyelination have implicated the antibody response and both the CD4+ and CD8+ T cell responses. In this report, we demonstrate that Balb/c (H-2d) mice immunized with the JHM (JHMV) strain of MHV develop a CD8+ cytotoxic T lymphocyte (CTL) response. One population of these virus-specific CTL recognize the nucleocapsid (N) protein. Recombinant vaccinia viruses expressing either the entire N protein or carboxy-terminal deletions were used to determine the number and location of the epitope(s) recognized. The CTLs were found to recognize a peptide contained within the carboxy-terminal 149 amino acids of the N protein. Analysis of infected cell lines expressing transfected major histocompatibility genes demonstrated that the anti-N protein CTLs were restricted exclusively to the Ld molecule. These data provide the first definition of a MHV-specific CTL response directed to a viral protein and suggest that the anti-N protein CTL response is one potential mechanism used by the host to clear JHMV from the central nervous system.

A transcription factor for expression of vaccinia virus late genes is encoded by an intermediate gene

A factor, designated VLTF-1, that is required in vitro for specific transcription of vaccinia virus late genes was previously isolated from vaccinia virus-infected cells. Subsequent genetic experiments identified three vaccinia virus genes, encoding proteins of 17, 26, and 30 kDa, that together trans activate late gene expression in vivo. The purpose of this study was to determine whether VLTF-1 corresponded to one of the three trans activators. Toward this end, VLTF-1 was further purified, the trans-activator genes were expressed in Escherichia coli, and antisera were made to the native and recombinant proteins. Antibody to the 30-kDa recombinant protein reacted on Western immunoblots with a protein of approximately Mr 30,000 that cochromatographed and cosedimented with VLTF-1 activity from virus-infected cells. Conversely, antibody to purified VLTF-1 bound to products produced by in vitro transcription and translation of the open reading frame encoding the 30-kDa trans-activator protein. Both antisera depleted VLTF-1 activity and blocked late gene transcription by partially purified extracts of vaccinia virus-infected cells. Taken together, these data demonstrate that the 30-kDa trans activator comprises part, if not all, of VLTF-1 activity.

Role of DNA replication in vaccinia virus gene expression: a naked template is required for transcription of three late trans-activator genes

The DNA replication requirement for vaccinia virus late gene expression was mimicked by transfecting a late promoter-controlled reporter gene into infected cells in the presence of a DNA synthesis inhibitor. This late promoter activation block was overcome by cotransfecting either naked linear vaccinia virion DNA or three cloned viral genes encoding trans-activator polypeptides of 17, 26, and 30 kd. These newly identified trans-activator genes were independently transcribed only from replicated or transfected DNA. These data suggest a regulatory cascade in which the parental viral genome serves as a template for the RNA polymerase and early promoter-specific transcription factors that are packaged in the infectious particle; the newly replicated DNA is accessible to sequentially synthesized intermediate promoter- and late promoter-specific trans-activators.

A clustering of RNA recombination sites adjacent to a hypervariable region of the peplomer gene of murine coronavirus

Coronaviruses undergo RNA recombination at a very high frequency. To understand the mechanism of recombination in murine coronavirus, we have performed RNA sequencing of viral genomic RNA to determine the precise sites of recombination in a series of recombinants which have crossovers within the gene encoding the peplomer protein. We found that all of the recombination sites are clustered within a region of 278 nucleotides in the 5'-half of the gene. This region in which all of the crossovers occurred represents a small fraction of the distance between the two selection markers used for the isolation of these recombinant viruses. This result suggests that this region may be a preferred site for RNA recombination. The crossover sites are located within and immediately adjacent to a hypervariable area of the gene. This area has undergone deletions of varying sizes in several virus strains which have been passaged either in vivo or in vitro. These results suggest that a similar RNA structure may be involved in the occurrence of both recombination and deletion events.

Design, synthesis, testing, and quantitative structure-activity relationship analysis of substituted salicylaldehyde Schiff bases of 1-amino-3-hydroxyguanidine tosylate as new antiviral agents against coronavirus

Further modifications of the structural features of Schiff bases of hydroxyaminoguanidines (SB-HAG) led to nine substituted salicylaldehyde Schiff bases of HAG (SSB-HAG) derivatives and three other SB-HAG derivatives. These new compounds were tested for the first time against infection by a coronavirus, mouse hepatitis virus (MHV). The most active compound, 2 [1-[(3'-allyl-2'-hydroxybenzylidene)amino]- 3-hydroxyguanidine], against the growth of MHV is about 376 times more active than hydroxyguanidine and about 564 times more active than HAG itself when the TCID50 values are compared. Plaque assays of MHV released from cells treated with these compounds suggest that SSB-HAG tosylate may inhibit the transcription of viral RNAs in virus-infected cells. Quantitative structure-activity relationship (QSAR) analyses of two subsets show that the inhibitory activities correlate well with the electronic and the lipophilic parameters. The structural requirements for the antiviral activity of substituted SSB-HAG tosylate against coronaviral infection are stringent according to the inhibitory activities and QSAR analysis of these new compounds.

Inhibition of murine coronavirus RNA synthesis by hydroxyguanidine derivatives

A series of hydroxyguanidine derivatives, which are substituted salicylaldehyde Schiff-bases of 1-amino-3- hydroxyguanidine tosylate, were tested for the inhibition of RNA synthesis of mouse hepatitis virus (MHV). It was shown that these compounds could selectively inhibit virus-specific RNA synthesis. Every aspect of viral RNA synthesis, including synthesis of negative-stranded RNA, subgenomic mRNA transcription and genomic RNA replication, was inhibited to roughly the same extent. These compounds are the first known inhibitors of coronaviral RNA synthesis and should prove useful for understanding the mechanism of viral RNA synthesis.

Relationship between cell membrane potential and natural killer cell cytolysis in human hepatocellular carcinoma cells

One of the body's natural defense mechanisms against tumor cells is lysis of the invading cell by cytotoxic T-cells and natural killer (NK) cells. Five human hepatocellular carcinoma cell lines were found to have different sensitivities to killing by peripheral blood monocytes in a 51Cr release assay. This killing was demonstrated to be due to NK cell lysis. Electrical recording measurements of the membrane potentials of these five cell lines showed different values for each line, all below values reported for normal hepatocytes. Correlation between mean cell membrane potential, and sensitivity to NK lysis, revealed an inverse relationship. In this study we demonstrate that the lower the mean membrane potential of a human hepatocellular carcinoma cell line, the more sensitive it is to NK cell cytolysis. Cell surface positive potential did not correlate with NK cytolysis and only a weak correlation was found between cell membrane negative potential and cell surface positive potential between cell lines.

Interactions between coronavirus nucleocapsid protein and viral RNAs: implications for viral transcription

The interaction of the mouse hepatitis virus (MHV) nucleocapsid protein (N) and viral RNA was examined. Monoclonal antibody specific for N protein coimmunoprecipitated MHV genomic RNA as well as all six MHV subgenomic mRNAs found in MHV-infected cells. In contrast, monoclonal antibodies to the MHV E2 or E1 envelope glycoproteins, an anti-I-A monoclonal antibody, and serum samples from lupus patients did not immunoprecipitate the MHV mRNAs. Moreover, the anti-N monoclonal antibody did not coimmunoprecipitate vesicular stomatitis virus RNA or host cell RNA under conditions which immunoprecipitated all MHV RNAs. These data suggest a specific interaction between the N protein and the virus-specific mRNAs. Both the membrane-bound and cytosolic small MHV leader-specific RNAs of greater than 65 nucleotides long were immunoprecipitated only by anti-N monoclonal antibody. These data suggest that an N binding site is present within the leader RNA sequences at a site at least 65 nucleotides from the 5' end of genomic RNA and all six subgenomic mRNAs. The larger leader-containing RNAs originating from mRNA 1 and mRNA 6, as well as the MHV negative-stranded RNA, were also immunoprecipitated by the anti-N monoclonal antibody. These data indicate that the MHV N protein is associated with MHV-specific RNAs and RNA intermediates and may play an important functional role during MHV transcription and replication.

Human hepatitis delta antigen is a nuclear phosphoprotein with RNA-binding activity

The genetic origin, structure, and biochemical properties of the delta antigen (HDAg) of a human hepatitis delta virus (HDV) were investigated. A cDNA fragment containing the open reading frame encoding the HDAg was transcribed into RNA and used for in vitro translation in rabbit reticulocyte lysates. The HDAg open reading frame was also inserted into an expression vector containing a simian virus 40 T-antigen promoter and expressed into COS 7 cells. In both systems, a protein species of 26 kilodaltons was synthesized from this open reading frame and could be specifically immunoprecipitated with antisera obtained from patients with delta hepatitis. A similar protein was also synthesized from antigenomic-sense monomeric HDV RNA in both systems, although the efficiency of translation was lower than that of the isolated open reading frame. This protein was found to be phosphorylated at the serine residues. Immunoperoxidase studies with anti-HDV sera demonstrated that the HDAg was expressed mainly in the nuclei of the transfected COS 7 cells. Moreover, the HDAg was shown to bind the genomic RNA of HDV. These studies indicate that HDAg is encoded by the antigenomic-sense RNA of HDV and is a nuclear phosphoprotein associated with an RNA-binding activity.

RNA recombination of murine coronaviruses: recombination between fusion-positive mouse hepatitis virus A59 and fusion-negative mouse hepatitis virus 2

It has previously been shown that the murine coronavirus mouse hepatitis virus (MHV) undergoes RNA recombination at a relatively high frequency in both tissue culture and infected animals. Thus far, all of the recombination sites had been localized at the 5' half of the RNA genome. We have now performed a cross between MHV-2, a fusion-negative murine coronavirus, and a temperature-sensitive mutant of the A59 strain of MHV, which is fusion positive at the permissive temperature. By selecting fusion-positive viruses at the nonpermissive temperature, we isolated several recombinants containing multiple crossovers in a single genome. Some of the recombinants became fusion negative during the plaque purification. The fusion ability of the recombinants parallels the presence or absence of the A59 genomic sequences encoding peplomers. Several of the recombinants have crossovers within 3' end genes which encode viral structural proteins, N and E1. These recombination sites were not specifically selected with the selection markers used. This finding, together with results of previous recombination studies, indicates that RNA recombination can occur almost anywhere from the 5' end to the 3' end along the entire genome. The data also show that the replacement of A59 genetic sequences at the 5' end of gene C, which encodes the peplomer protein, with the fusion-negative MHV-2 sequences do not affect the fusion ability of the recombinant viruses. Thus, the crucial determinant for the fusion-inducing capability appears to reside in the more carboxyl portion of the peplomer protein.

In vivo RNA-RNA recombination of coronavirus in mouse brain

RNA-RNA recombination between different strains of the murine coronavirus mouse hepatitis virus (MHV) occurs at a very high frequency in tissue culture. To demonstrate that RNA recombination may play a role in the evolution and pathogenesis of coronaviruses, we sought to determine whether MHV recombination could occur during replication in the animal host of the virus. By using two selectable markers, i.e., temperature sensitivity and monoclonal antibody neutralization, we isolated several recombinant viruses from the brains of mice infected with two different strains of MHV. The recombination frequency was very high, and recombination occurred at multiple sites on the viral RNA genome. This finding suggests that RNA-RNA recombination may play a significant role in natural evolution and neuropathogenesis of coronaviruses.

Defective-interfering particles of murine coronavirus: mechanism of synthesis of defective viral RNAs

The mechanism of synthesis of the defective viral RNAs in cells infected with defective-interfering (DI) particles of mouse hepatitis virus was studied. Two DI-specific RNA species, DIssA of genomic size and DIssE of subgenomic size, were detected in DI-infected cells. Purified DI particles, however, were found to contain predominantly DIssA and only a trace amount of DIssE RNA. Despite its negligible amount, the DIssE RNA in virions appears to serve as the template for the synthesis of DIssE RNA in infected cells. This conclusion was supported by two studies. First, the uv target size for DIssE RNA synthesis is significantly smaller than that for DIssA. Second, when purified DIssE RNA was transfected into cells which had been infected with a helper virus, DIssE RNA could replicate itself and became a predominant RNA species in the infected cells. Thus, DIssE RNA was not synthesized from the genomic RNA of DI particles. By studying the relationship between virus dilution and the amount of intracellular viral RNA synthesis, we have further shown that DIssE RNA synthesis requires a helper function, but it does not utilize the leader sequence of the helper virus. In contrast, DIssA synthesis appears to be helper-independent and can replicate itself. Thus DIssA codes for a functional RNA polymerase.

Temporal regulation of bovine coronavirus RNA synthesis

The structure and synthesis of bovine coronavirus (BCV)-specific intracellular RNA were studied. A genome-size RNA and seven subgenomic RNAs with molecular weights of approximately 3.3 X 10(6), 3.1 X 10(6), 2.6 X 10(6), 1.1 X 10(6), 1.0 X 10(6), 0.8 X 10(6) and 0.6 X 10(6) were detected. Comparisons of BCV intracellular RNAs with those of mouse hepatitis virus (MHV) demonstrated the presence of an additional RNA for BCV, species 2a, of 3.1 X 10(6) daltons. BCV RNAs contain a nested-set structure similar to that of other coronaviruses. This nested-set structure would suggest that the new RNA has a capacity to encode a protein of approximately 430 amino acids. Kinetic studies demonstrated that the synthesis of subgenomic mRNAs and genomic RNA are differentially regulated. At 4 to 8 h post-infection (p.i.), subgenomic RNAs are synthesized at a maximal rate and represent greater than 90% of the total viral RNA synthesized, whereas genome-size RNA accounts for only 7%. Later in infection, at 70 to 72 h p.i., genome-size RNA is much more abundant and accounts for 88% of total RNA synthesized. Immunoprecipitations of [35S]methionine-pulse-labeled viral proteins demonstrated that viral protein synthesis occurs early in the infection, concurrent with the peak of viral subgenomic RNA synthesis. Western blot analysis suggests that these proteins are stable since the proteins are present at high level as late as 70 to 72 h p.i. The kinetics of production of virus particles coincides with the synthesis of genomic RNA. These studies thus indicate that there is a differential temporal regulation of the synthesis of genomic RNA and subgenomic mRNAs, and that the synthesis of genomic RNA is the rate-limiting step regulating the production of virus particles.

RNA recombination of coronaviruses: localization of neutralizing epitopes and neuropathogenic determinants on the carboxyl terminus of peplomers

Murine coronaviruses undergo RNA recombination at a very high frequency. We have obtained a series of recombinant viruses using neutralizing monoclonal antibodies in conjunction with temperature-sensitive markers. All of the recombinants obtained have a crossover within gene C, which encodes the peplomer protein of the virus. The genetic structure of these recombinants suggests that the antigenic regions recognized by these neutralizing monoclonal antibodies are localized on the carboxyl-terminal one-third of the peplomer protein. Since the two monoclonal antibodies used are also associated with the critical determinants of virus neuropathogenicity, we conclude that both the neutralizing antibody binding sites and determinants of pathogenicity are localized at the carboxyl-terminal one-third of the peplomer. The variation of crossover sites in different recombinant viruses also allowed precise mapping of additional antigenic sites. RNA recombination thus presents a powerful genetic tool, and the carboxyl-terminal localization of the biological functions of peplomers suggests a distinct conformation of these viral membrane proteins.

Multiple recombination sites at the 5'-end of murine coronavirus RNA

Mouse hepatitis virus (MHV), a murine coronavirus, contains a nonsegmented RNA genome. We have previously shown that MHV could undergo RNA-RNA recombination in crosses between temperature-sensitive mutants and wild-type viruses at a very high frequency (S. Makino, J.G. Keck, S.A. Stohlman, and M.M.C. Lai (1986) J. Virol. 57, 729-737). To better define the mechanism of RNA recombination, we have performed additional crosses involving different sets of MHV strains. Three or possibly four classes of recombinants were isolated. Recombinants in the first class, which are similar to the ones previously reported, contain a single crossover in either gene A or B, which are the 5'-most genes. The second class of recombinants contain double crossovers in gene A. The third class of recombinants have crossovers within the leader sequence located at the 5'-end of the genome. The crossover sites of the third class have been located between 35 and 60 nucleotides from the 5'-end of the leader RNA. One of these recombinants has double crossovers within the short region comprising the leader sequences. Finally, we describe one recombinant which may contain a triple crossover. The presence of so many recombination sites within the 5'-end of the genome of murine coronaviruses confirms that RNA recombination is a frequent event during MHV replication and is consistent with our proposed model of "copy-choice" recombination in which RNA replication occurs in a discontinuous and nonprocessive manner.

RNA recombination of coronavirus

We have previously shown that Mouse hepatitis virus (MHV) can undergo RNA-RNA recombination at a very high frequency (S. Makino, et al., J. Virol. 57, 729-737, 1986). To better define the mechanism of RNA recombination, we have performed additional crosses involving different MHV strains. We have obtained recombinant viruses with multiple cross-overs. The isolation of such recombinants further indicates the high frequency of coronavirus RNA recombination. By using cell fusion as a selection marker, we have also obtained recombinants between MHV-2 and A59 strains. Some of these recombinants have cross-overs in the 3'-end genes of the genome, thus demonstrating that recombination could occur along the entire genome. Finally, we have obtained recombinants by selecting with neutralizing monoclonal antibodies. These recombinants have cross-overs within gene C which encodes the peplomer protein. The genetic structure of these recombinants allowed us to determine the important domains of the peplomer proteins.

High-frequency RNA recombination of murine coronaviruses

The RNA genome of coronaviruses consists of a single species of nonsegmented RNA. In this communication, we demonstrate that the RNA genomes of different strains of murine coronaviruses recombine during mixed infection at a very high frequency. Susceptible cells were coinfected with a temperature-sensitive mutant of one strain of mouse hepatitis virus (MHV) and a wild-type virus of a different strain. Of 21 randomly isolated viruses released from the coinfected cells at the nonpermissive temperature, 2 were recombinants which differed in the site of recombination. After three serial passages of the original virus pool derived from the mixed infection, the majority of the progeny viruses were recombinants. These recombinant viruses represented at least five different recombination sites between the two parental MHV strains. Such a high-frequency recombination between nonsegmented RNA genomes of MHV suggests that segmented RNA intermediates might be generated during MHV replication. We propose that the RNA replication of MHV proceeds in a discontinuous and nonprocessive manner, thus generating free segmented RNA intermediates, which could be used in RNA recombination via a copy-choice mechanism.

Recombination between nonsegmented RNA genomes of murine coronaviruses

We have isolated a recombinant virus between the A59 and JHM strains of mouse hepatitis virus, which contain a single species of nonsegmented RNA genome. This recombinant was derived by mixed infection of DBT cells with temperature-sensitive mutants of A59 and JHM at nonpermissive temperature. Viruses recovered at this temperature were screened by oligonucleotide fingerprinting of their genomic RNAs. One recombinant virus, B1, was found to contain mostly A59-derived sequences, but the 3 kilobases at the 5' end of the genomic RNA was derived from JHM. Thus, the crossover point in the B1 genome is located within gene A, which codes for the viral RNA polymerases. The study of the intracellular RNA species of B1 virus revealed that probably all of the virus-specific subgenomic mRNA species contained the body sequences of strain A59 but the leader sequences of JHM. This result indicates that the JHM leader RNA, which differs from the A59 leader RNA, could be fused to the mRNAs of a different virus strain during RNA transcription. Furthermore, B1 virus-infected cells contain an additional subgenomic mRNA species which is transcribed from a new initiation site within gene C, suggesting that the leader RNA could determine the site of initiation for coronavirus mRNAs. These data represent a first report of RNA recombination between viruses, other than picornaviruses, which contain nonsegmented RNA genomes.