JOURNAL OF VIROLOGY, Apr. 1976, p. 85-91 Copyright © 1976 American Society for Microbiology
Vol. 18, No. 1 Printed in US A.
Fate of Input Oncornavirion RNA -Biological Studies M. M. SVEDA, B. N. FIELDS, AND R. SOEIRO* Departments of Medicine and Cell Biology, Albert Einstein College of Medicine, Bronx, New York 10461
Received for publication 5 November 1975
The fate of input Friend leukemia virus RNA was studied using labeled input virus. The appearance of nuclear RNA-DNA hybrid molecules and the apparent integration of input virion RNA with host cell DNA was studied using a series of inhibitors of DNA or protein synthesis, cell growth conditions, and an intercalating agent. Under all these conditions of infection, little to no viral-specific RNA-DNA hybrid molecules were formed. These data demonstrate that the formation of such RNA-DNA hybrid structures requires conditions of infection that allow provirus synthesis and integration. Furthermore, they suggest that at least a fraction of input virion RNA may transiently become integrated with host cell DNA.
Early events in the infection of cells by oncornaviruses involve DNA synthesis both for viral replication and for cell transformation (4, 12). The appearance of new DNA copies homologous to infecting viral RNA in productively infected or transformed cells has suggested a role for the virion-associated reverse transcriptase (2, 26). Virions lacking this enzyme are noninfectious (8, 9). Although aspects of viral genome RNAdirected DNA synthesis have been studied in vitro (25), the full details of the mechanism by which genome RNA is transcribed into viralspecific DNA and subsequently integrated into the host chromosome are not completely understood. Analysis of these early events by two different methodologies has resulted in some confusion in the literature, since this has led to two different models. A double-stranded closed circular form of proviral DNA has been found both in the cytoplasm and the nucleus of newly infected cells by means of molecular hybridization experiments. The observation that this molecule appears first in the cytoplasm and only later in the nucleus has led to the hypothesis that the earliest events in viral replication are cytoplasmic (6, 27, 28). On the other hand, we and others found, by following the fate of radiolabeled input genome RNA, that input genomic RNA appears to form viral-specific RNA-DNA covalently linked hybrids, a presumed intermediate in the formation of a double-stranded DNA copy of genome RNA. Our results indicated that virtually all of these molecules were in the nucleus, and therefore we suggested that the nucleus appears to be the initial site for provirus synthesis (J. Leis, J. Hurwitz, A. Schincariol, and W. Jok-
lik, in 34th Annual Biology Colloquy, Oregon State University, in press; 22). Our earlier results showed this labeled material to be singlestranded RNA covalently linked to DNA. The RNA has been shown to be viral specific by means of molecular annealing of radiolabeled RNA with viral-specific DNA. Analysis in Cs2SO4 gradients revealed genome RNA covalently linked to DNA in two forms. The first form to appear banded at p = 1.55 in Cs2SO4, and it was proposed that this material represents viral RNA covalently linked to viral DNA. Later in infection, genome RNA appears linked to essentially pure DNA (p = 1.44) in Cs2SO4. The material at p = 1.44 represents viral RNA apparently covalently linked to host cell DNA, since this RNA sediments with very high-molecular-weight DNA and anneals specifically with viral DNA. Both these forms of RNA-DNA hybrid molecules appear to be localized exclusively in the nucleus and appear to be synthesized by the virion reverse transcriptase (16). In an effort to provide further evidence that the formation of covalent linkage between radiolabeled input genome RNA and viral or cellular DNA is biologically significant, we have carried out experiments under conditions specifically designed to prevent provirus formation. It was reasoned that if the appearance of viral specific RNA-DNA hybrids represented a valid pathway in replication, then biological conditions of virus infection known to prevent provirus formation should prevent the covalent linkage of viral genomic RNA with host cell DNA. We have found that all the conditions we have tested that interfere with either DNA or protein synthesis prevent the appearance of an 85
86
J. VIROL.
SVEDA, FIELDS, AND SOEIRO
apparent covalent linkage of viral-specific genome RNA to host cell DNA. Furthermore, the presence of the intercalating agent ethidium bromide, which is known to interfere with the synthesis of double-stranded closed circular nucleic acids, also prevents the integration of input viral RNA with the cellular genome. These experiments confirm our earlier studies as well as those of Leis et al. (in press) and suggest that some early nuclear events in the formation of provirus involve RNA-DNA hybrid molecules. MATERIALS AND METHODS Cells and viruses. NIH Swiss and BALB/c mouse embryo cells were purchased from Microbiological Associates and maintained in F-11 medium (GIBCO) with 5% fetal calf serum (North American Biologicals). XC cells and S+L- cells were generous gifts of Wallace Rowe and George Todaro, respectively. Friend leukemia virus, both N- and B-tropic strains, was kindly donated by Frank Lilly as extracts of virus-infected mouse spleen. Each virus extract was passaged in the permissive secondary mouse embryo cell line, and the in vitro stock was purified and frozen at -70 C for use in these experiments. Labeling of Friend virus. Six days after infection of the permissive cell line-a time of maximum virus yield-cultures were rinsed with phosphate-free medium and then labeled with [32Plorthophosphate (3 mCi/ml; New England Nuclear) for 2 h. The medium was then removed and replaced by fresh medium every 2 h for a total of six harvests to produce released virus no more than 2 h old. Concentration of virus from the medium and sucrose density gradient purification procedures were as previously described (21). Plaque assay for Friend leukemia virus was carried out by the XC assay described by Rowe et al.
(19). Adsorption of virus. 32P-labeled virus was diluted to a final concentration of sucrose of less than 3%. Although the number of PFU present in the 32p_ labeled virus was not measured directly, the multiplicity of infection was controlled in the following manner. A culture of monolayer cells, identical with that used for producing 32P-labeled virus, was infected in parallel but with nonradioactive medium. A titer of virus harvested and purified in the same manner as for 32P-labeled virus was obtained from the nonlabeled culture. Since labeled virus could only have a lower titer due to 32p decay-induced nucleic acid damage, the multiplicity of infection was determined by titration of the unlabeled culture, and the multiplicity of infection of the 32p_ labeled virus was estimated to be
Vol. 18, No. 1 Printed in US A.
Fate of Input Oncornavirion RNA -Biological Studies M. M. SVEDA, B. N. FIELDS, AND R. SOEIRO* Departments of Medicine and Cell Biology, Albert Einstein College of Medicine, Bronx, New York 10461
Received for publication 5 November 1975
The fate of input Friend leukemia virus RNA was studied using labeled input virus. The appearance of nuclear RNA-DNA hybrid molecules and the apparent integration of input virion RNA with host cell DNA was studied using a series of inhibitors of DNA or protein synthesis, cell growth conditions, and an intercalating agent. Under all these conditions of infection, little to no viral-specific RNA-DNA hybrid molecules were formed. These data demonstrate that the formation of such RNA-DNA hybrid structures requires conditions of infection that allow provirus synthesis and integration. Furthermore, they suggest that at least a fraction of input virion RNA may transiently become integrated with host cell DNA.
Early events in the infection of cells by oncornaviruses involve DNA synthesis both for viral replication and for cell transformation (4, 12). The appearance of new DNA copies homologous to infecting viral RNA in productively infected or transformed cells has suggested a role for the virion-associated reverse transcriptase (2, 26). Virions lacking this enzyme are noninfectious (8, 9). Although aspects of viral genome RNAdirected DNA synthesis have been studied in vitro (25), the full details of the mechanism by which genome RNA is transcribed into viralspecific DNA and subsequently integrated into the host chromosome are not completely understood. Analysis of these early events by two different methodologies has resulted in some confusion in the literature, since this has led to two different models. A double-stranded closed circular form of proviral DNA has been found both in the cytoplasm and the nucleus of newly infected cells by means of molecular hybridization experiments. The observation that this molecule appears first in the cytoplasm and only later in the nucleus has led to the hypothesis that the earliest events in viral replication are cytoplasmic (6, 27, 28). On the other hand, we and others found, by following the fate of radiolabeled input genome RNA, that input genomic RNA appears to form viral-specific RNA-DNA covalently linked hybrids, a presumed intermediate in the formation of a double-stranded DNA copy of genome RNA. Our results indicated that virtually all of these molecules were in the nucleus, and therefore we suggested that the nucleus appears to be the initial site for provirus synthesis (J. Leis, J. Hurwitz, A. Schincariol, and W. Jok-
lik, in 34th Annual Biology Colloquy, Oregon State University, in press; 22). Our earlier results showed this labeled material to be singlestranded RNA covalently linked to DNA. The RNA has been shown to be viral specific by means of molecular annealing of radiolabeled RNA with viral-specific DNA. Analysis in Cs2SO4 gradients revealed genome RNA covalently linked to DNA in two forms. The first form to appear banded at p = 1.55 in Cs2SO4, and it was proposed that this material represents viral RNA covalently linked to viral DNA. Later in infection, genome RNA appears linked to essentially pure DNA (p = 1.44) in Cs2SO4. The material at p = 1.44 represents viral RNA apparently covalently linked to host cell DNA, since this RNA sediments with very high-molecular-weight DNA and anneals specifically with viral DNA. Both these forms of RNA-DNA hybrid molecules appear to be localized exclusively in the nucleus and appear to be synthesized by the virion reverse transcriptase (16). In an effort to provide further evidence that the formation of covalent linkage between radiolabeled input genome RNA and viral or cellular DNA is biologically significant, we have carried out experiments under conditions specifically designed to prevent provirus formation. It was reasoned that if the appearance of viral specific RNA-DNA hybrids represented a valid pathway in replication, then biological conditions of virus infection known to prevent provirus formation should prevent the covalent linkage of viral genomic RNA with host cell DNA. We have found that all the conditions we have tested that interfere with either DNA or protein synthesis prevent the appearance of an 85
86
J. VIROL.
SVEDA, FIELDS, AND SOEIRO
apparent covalent linkage of viral-specific genome RNA to host cell DNA. Furthermore, the presence of the intercalating agent ethidium bromide, which is known to interfere with the synthesis of double-stranded closed circular nucleic acids, also prevents the integration of input viral RNA with the cellular genome. These experiments confirm our earlier studies as well as those of Leis et al. (in press) and suggest that some early nuclear events in the formation of provirus involve RNA-DNA hybrid molecules. MATERIALS AND METHODS Cells and viruses. NIH Swiss and BALB/c mouse embryo cells were purchased from Microbiological Associates and maintained in F-11 medium (GIBCO) with 5% fetal calf serum (North American Biologicals). XC cells and S+L- cells were generous gifts of Wallace Rowe and George Todaro, respectively. Friend leukemia virus, both N- and B-tropic strains, was kindly donated by Frank Lilly as extracts of virus-infected mouse spleen. Each virus extract was passaged in the permissive secondary mouse embryo cell line, and the in vitro stock was purified and frozen at -70 C for use in these experiments. Labeling of Friend virus. Six days after infection of the permissive cell line-a time of maximum virus yield-cultures were rinsed with phosphate-free medium and then labeled with [32Plorthophosphate (3 mCi/ml; New England Nuclear) for 2 h. The medium was then removed and replaced by fresh medium every 2 h for a total of six harvests to produce released virus no more than 2 h old. Concentration of virus from the medium and sucrose density gradient purification procedures were as previously described (21). Plaque assay for Friend leukemia virus was carried out by the XC assay described by Rowe et al.
(19). Adsorption of virus. 32P-labeled virus was diluted to a final concentration of sucrose of less than 3%. Although the number of PFU present in the 32p_ labeled virus was not measured directly, the multiplicity of infection was controlled in the following manner. A culture of monolayer cells, identical with that used for producing 32P-labeled virus, was infected in parallel but with nonradioactive medium. A titer of virus harvested and purified in the same manner as for 32P-labeled virus was obtained from the nonlabeled culture. Since labeled virus could only have a lower titer due to 32p decay-induced nucleic acid damage, the multiplicity of infection was determined by titration of the unlabeled culture, and the multiplicity of infection of the 32p_ labeled virus was estimated to be
