Rat sequences of the Kirsten and Harvey murine sarcoma virus genomes: nature, origin, and expression in rat tumor RNA.

Two murine sarcoma viruses, the Kirsten and the Harvey, were isolated by passage of mouse type C leukemia viruses through rats. These sarcoma viruses have genomes containing portions of their parental type C mouse leukemia virus genomes, in stable association with specific rat cellular sequences that we find to be quite likely not those of a rat type C leukemia virus. To determine if these murine sarcoma viruses provide a model relevant to the events occurring in spontaneous tumors, we have hybridized DNA and RNA prepared from rat tumors and normal rat tissues to [3H]DNA prepared from the Kirsten murine sarcoma virus. We have also hybridized these rat tissue nucleic acids to [3H]DNA prepared from a respresentative endogenous rat type C leukemia virus, the WFU (Wistar-Furth). Sarcoma-viral rat cellular sequences and endogenous rat leukemia viral sequences were detected in the DNA of both tumor and normal tissues, with no evidence of either gene amplification or additional sequences being present in tumor DNA. Sarcoma-viral rat cellular sequences and endogenous rat leukemia viral sequences were detected at elevated concentrations in the RNA of many rat tumors and in specific groups of normal tissues.     

Rate of divergence of cellular sequences homologous to segments of Moloney sarcoma virus.

The RNA genome of the Moloney isolate of murine sarcoma virus (M-MSV) consists of two parts--a sarcoma-specific region with no homology to known leukemia viral RNAs, and a shared region present also in Moloney murine leukemia virus RNA. Complementary DNA was isolated which was specific for each part of the M-MSV genome. The DNA of a number of mammalian species was examined for the presence of nucleotide sequences homologous with the two M-MSV regions. Both sets of viral sequences had homologous nucleotide sequences present in normal mouse cellular DNA. MSV-specific sequences found in mouse cellular DNA closely matched those nucleotide sequences found in M-MSV as seen by comparisons of thermal denaturation profiles. In all normal mouse cells tested, the cellular set of M-MSV-specific nucleotide sequences was present in DNA as one to a few copies per cell. The rate of base substitution of M-MSV nucleotide sequences was compared with the rate of evolution of both unique sequences and the hemoglobin gene of various species. Conservation of MSV-specific nucleotide sequences among species was similar to that of mouse globin gene(s) and greater than that of average unique cellular sequences. In contrast, cellular nucleotide sequences that are homologous to the M-MSV-murine leukemia virus "common" nucleotide region were present in multiple copies in mouse cells and were less well matched, as seen by reduced melting profiles of the hybrids. The cellular common nucleotide sequences diverged very rapidly during evolution, with a base substitution rate similar to that reported for some primate and avian endogenous virogenes. The observation that two sets of covalently linked viral sequences evolved at very different rates suggests that the origin of M-MSV may be different from endogenous helper viruses and that cellular sequences homologous to MSV-specific nucleotide sequences may be important to survival.     

Nucleic Acid Homology of Murine Type-C Viral Genes

The nucleic acid sequence homology between various murine, endogenous type-C viruses (three host range classes of BALB/c virus, the AT-124 virus, and the CCL 52 virus) and two laboratory strains of murine leukemia virus (Rauscher and Kirsten) was determined by DNA:RNA hybridization. The viral sequences exhibit varying degrees of partial homology. DNA:DNA hybridizations were performed between [³H]DNA probes prepared from N- and X-tropic BALB/c endogenous viruses and cellular DNAs from BALB/c, NIH Swiss, and AKR inbred mouse strains as well as from California feral mice and the Asian mouse subspecies Mus musculus molossinus and M. musculus castaneus. All of these strains of mice are shown to possess multiple (six to seven per haploid genome), partially related copies of type-C virogenes in their DNAs. Thermal melting profiles of the DNA:RNA and DNA:DNA hybrids suggest that the partial homology of the viral nucleic acid sequences is the result of base alterations throughout the viral genomes, rather than the loss of discrete segments of viral sequences.     

Virus-Like 30S RNA in Mouse Cells

Uninfected JLS-V9 mouse cells are known to express high levels of viral sequences that hybridize to complementary DNA made by the BrdU-induced virus of JLS-V9 cells. The genome in the BrdU-induced virus has been found to consist mainly of an RNA species that migrates as 30S RNA material during electrophoresis through agarose gels. This virus-like 30S RNA, designated VL30 RNA, apparently represents a new class of endogenous defective retroviruses that are not generally evident because of their defectiveness and lack of biological function. Fingerprint analysis and hybridization studies show that VL30 RNA does not have homology with the standard nondefective murine leukemia viruses. Upon superinfection with a nondefective murine leukemia virus, or upon induction of endogenous virus with BrdU, VL30 RNA is rescued into virions by phenotypic mixing. When VL30 RNA is rescued by BrdU induction, the VL30 RNA is mainly organized as a 50S complex, but when VL30 is rescued by superinfection, VL30 is also found in 70S RNA. Rescued VL30 RNA sequences can be reverse transcribed by the virion-associated DNA polymerase in an endogenous reaction. Many mouse cells express the sequences, whereas heterologous cells such as rat or rabbit cells do not contain them. By using hybridization of a complementary DNA probe to cellular RNA immobilized on paper, no subgenomic RNA related to the VL30 RNA could be found in cells expressing the VL30 sequences. From 20 to 50 copies of these sequences were found to be contained in the mouse genome. VL30 RNA is probably present in most stocks of leukemia and sarcoma viruses made in mouse cells.     

Structure of the Abelson murine leukemia virus genome and the homologous cellular gene: studies with cloned viral DNA

Circular double-stranded DNA produced after infection of mouse cells with Abelson murine leukemia virus (A-MuLV) was isolated and cloned in the phage vector Charon 21A. The resulting clones of the A-MuLV genome show homology to the ends of Moloney MuLV and to a 3.5 kb central region containing sequences unique to Abelson virus. A 2.3 kb restriction fragment containing only A-MuLV-specific sequences was subcloned in the plasmid vector pBR322 and used as a probe for the cellular gene that had been acquired by the virus. DNA from all inbred mouse lines examined contains an identical region of homology spread out over 11 to 20 kb. The cellular gene contains intervening sequences which are lacking in the viral genome. Rat, Chinese hamster, rabbit, chicken and human DNA also show homology to the viral probe.     

Biochemical characterization of the amphotropic group of murine leukemia viruses.

The recently described amphotropic group of murine leukemia viruses constitutes a distinct biological group, differing from the ecotropic and xenotropic groups in host range, cross interference, and serological reactivity. Viruses of this group have been detected only in wild mice from certain areas in California. By using a [3H]DNA probe synthesized in an endogenous reaction from detergent-lysed amphotropic virus (strain 1504-A), it was demonstrated that the amphotropic murine leukemia viruses are distinct biochemically, in that 20% of the viral genome sequences are not shared by AKR-type ecotropic or nay of three types of xenotropic murine leukemia virus tested. A subset of these amphotropic unique sequences, comprising one half of them, is present in the genome of wild mouse ecotropic viruses and in Moloney and Rauscher viruses as well. Sequences homologous to the entire genome of 1504-A amphotropic virus are present in the cellular DNA of all eight inbred mouse strains tested, as well as in wild Mus in Asia, in amounts varying from three to six complete viral genomes per haploid cell genome. Evidence is presented that at least 20% of the DNA sequences in both mouse- and mink-grown murine leukemia virus probes are of host-cell origin.     

Construction of a mammalian transducing vector from the genome of Moloney murine leukemia virus.

A 0.9-kilobase DNA fragment from the genome of Moloney murine leukemia virus, including the viral long terminal repeat, was covalently linked to the herpes simplex virus I thymidine kinase (tk) gene whose promoter was previously removed. The hybrid DNA structure was introduced into the chromosome of tk- mouse cells at single copy numbers, via transfection procedures. Cells expressing the newly introduced tk gene were identified by the HAT selection procedure and analyzed for tk- and moloney murine leukemia virus-specific DNA and RNA sequences by blot hybridization procedures. Expression of the tk gene is dependent on function(s) provided in cis by the viral DNA fragment. Vectors derived from this region are termed rGag (rG) vectors.     

Structural organization and biological activity of molecular clones of the integrated genome of a BALB/c mouse sarcoma virus.

BALB/c mouse sarcoma virus (BALB-MSV) is a spontaneously occurring transforming retrovirus of mouse origin. The integrated form of the viral genome was cloned from the DNA of a BALB-MSV-transformed nonproducer NRK cell line in the Charon 9 strain of bacteriophage lambda. In transfection assays, the 19-kilobase-pair (kbp) recombinant DNA clone transformed NIH/3T3 mouse cells with an efficiency of 3 X 10(4) focus-forming units per pmol. Such transformants possessed typical BALB-MSV morphology and released BALB-MSV after helper virus superinfection. A 6.8-kbp DNA segment within the 19-kbp DNA possessed restriction enzyme sites identical to those of the linear BALB-MSV genome. Long terminal repeats of approximately 0.6 kbp were localized at either end of the viral genome by the presence of a repeated constellation of restriction sites and by hybridization of segments containing these sites with nick-translated Moloney murine leukemia virus long terminal repeat DNA. A continuous segment of at least 0.6 and no more than 0.9 kbp of helper virus-unrelated sequences was localized toward the 3' end of the viral genome in relation to viral RNA. A probe composed of these sequences detected six EcoRI-generated DNA bands in normal mouse cell DNA as well as a smaller number of bands in rat and human DNAs. These studies demonstrate that BALB-MSV, like previously characterized avian and mammalian transforming retroviruses, arose by recombination of a type C helper virus with a well-conserved cellular gene.     

Biochemical properties of the bromodeoxyuridine-induced guinea pig virus.

The biophysical and biochemical properties of the virus particles released by guinea pig embryo cells treated with 5-bromo-2'-deoxyuridine (BUdR) have been compared to those of the B-type mouse mammary tumor virus (MMTV) and the C-type Rauscher murine leukemia virus. The high-molecular-weight (60 to 70S) RNA of the BUdR-induced guinea pig virus (GPV) has a molecular weight of 8 X 106 when measred by mixed agarose polyacylamide gel electrophoresis. The virus particles isolated from the tissue culture medium of BUdR-induced guniea pig cells have the following properties in common with MMTV: (i) a buoyant density of 1.18 g/ml in sucrose and 1.21 g/ml in CsCl, and (ii) a DNA polymerase that prefers Mg2+ over Mn2+ in an assay using the synthetic template poly(rC):oligo(dG). No nucleic acid sequence homology between GPV RNA and the viral RNAs of the MMTV, murine leukemia virus, hamster sarcoma virus, or Mason-Pfizer monkey virus could be observed in a competition hybridization assay using the radioactive-labeled GPV 60 to 70S RNA. By this same competition by hybridization assay the frequency of GPV proviral sequences was estimated to be at least 83 per haploid cellular genome of guniea pig cells. No nucleic acid sequences related to be GPV RNA were detected in the DNA of normal tissues of mice, rats, cats, dogs, baboons, or humans by direct RNA-DNA hybridization using radioactive GPV60 to 70S RNA.     

Isolation of an integrated provirus of Moloney murine leukemia virus with long terminal repeats in inverted orientation: integration utilizing two U3 sequences.

We have previously described the construction of a mutant of Moloney murine leukemia virus, in594-2, which carries a 2-base-pair insertion in the U5 region of the genome and is partially defective in forming the integrated proviral DNA. We have now recovered a cloned copy of an unusual provirus from rat cells infected with this mutant. The viral genome is flanked by long terminal repeats in inverted orientation, with U3 sequences joined to cellular DNA at both of the outer edges. In addition, the provirus is a recombinant, containing a segment of a VL30 element in inverted orientation in place of the Moloney murine leukemia virus env region. The recovery of this provirus indicates that two U3 regions can be used for viral integration and suggests that there may be no absolute requirement in the reaction for those U5 sequences outside the 13-base-pair inverted repeats.     

Structure and expression of mouse VL30 genes.

DNA sequencing and blot hybridization analyses have been used to study the structure of a mouse VL30 gene and the molecular nature of VL30-related RNA which is induced upon the stimulation of cultured AKR mouse embryo cells with defined peptide growth factors. An integrated mouse VL30 gene was found to contain identical 601-base-pair long terminal repeats (LTRs) which were themselves terminated in short inverted repeats. The entire VL30 gene was flanked by a 4-base-pair direct repeat of cellular DNA. Thus, VL30 genes are structurally analogous to integrated forms of retrovirus proviruses and certain other classes of mobile genetic elements. The LTR sequence was found to contain putative promoter and polyadenylation signals and generally exhibited little sequence homology to murine leukemia virus proviral LTRs. Certain short regions of sequence conservation, however, were evident, including the inverted terminal repeat, LTR-adjacent regions corresponding to origins of murine leukemia virus proviral DNA synthesis, and a 36-base-pair direct repeat bearing homology to the 72-base-pair direct repeat (enhancer sequence) of the murine leukemia virus-related Moloney sarcoma virus. Upon mitogenic stimulation of quiescent cells with epidermal growth factor and insulin, a major 5.5-kilobase VL30-specific RNA complementary to both LTR and non-LTR sequences was rapidly induced. We conclude that a complete VL30 gene(s) is highly regulated by peptide growth factor binding to specific membrane receptors in these cells.     

Selective degradation of integrated murine leukemia proviral DNA by deoxyribonucleases

The sensitivity to micrococcal nuclease and DNAase I of the integrated proviral DNA sequences in Swiss mouse cells infected with Moloney murine leukemia virus has been studied. Chromatin was separated into micrococcal nuclease-sensitive and -resistant regions, and the amount of proviral sequences in these DNA preparations was estimated by kinetic hybridization with single-stranded complementary DNA of Moloney murine leukemia virus. At least two thirds of the proviral DNA sequences were found in the open regions of chromatin, and only one third was resistant to nuclease. The proviral DNA sequences are even more sensitive to deoxyribonuclease I. When intact nuclei were treated with limited amounts of enzyme, only 5% of the nuclear DNA was digested, whereas 48% of the proviral DNA was degraded.The proviral DNA sequences in cells which do not produce virus are more resistant to nuclease digestion, as compared to virus producer cells. Thus the endogenous proviral sequences, in normal uninduced Swiss mouse cells, are randomly distributed between resistant and sensitive portions of chromatin when tested with either micrococcal nuclease or pancreatic deoxyribonuclease I. The effect of cell cycle synchronization on the accessibility of the proviral sequences to pancreatic deoxyribonuclease I was investigated with rat cells infected with Moloney murine leukemia virus. The amount of proviral DNA sensitive to pancreatic deoxyribonuclease I is higher in actively dividing cells than in cells arrested at Go phase, which produce only small amounts of virus.     

Characterization of the feline c-abl proto-oncogene

Analysis of total feline DNA by genomic blot hybridization, using the viral oncogene of Abelson murine leukemia virus as a specific probe, has led to the identification of multiple v-abl homologous genetic sequences in the cat genome. Upon restriction endonuclease BamHI digestion, the combined size of the v-abl homologous DNA fragments was about 31 kbp. To characterize these sequences further, four independent v-abl homologous cosmid clones with overlapping cellular inserts have been isolated from a gene library of cat lung genomic DNA. These inserts represent a contiguous region of cellular DNA sequences of 56 kbp in length. Within this region of the feline genome, the v-abl homologous sequences are discontinuously dispersed over a region of about 34 kbp. They represent the complete feline v-abl cellular homolog and are colinear with the viral v-abl oncogene. Nine regions of highly repetitive DNA sequences have been mapped in close proximity to v-abl homologous sequences. These results establish the presence of only a single c-abl proto-oncogene in the cat genome and present its genetic organization.     

Endogenous nonecotropic proviruses mapped with oligonucleotide probes from the long terminal repeat region

We characterized endogenous proviruses in C57BL/6J, DBA/2J, and C3H/HeJ mouse strains with oligonucleotide probes derived from long terminal repeat (LTR) sequences of three classes of nonecotropic murine leukemia virus. The segregation of proviral-host DNA junction fragments was followed in BXH and BXD recombinant inbred (RI) strain sets, and most fragments mapped readily to defined chromosomal regions. Most of the LTR fragments appear to correspond to proviruses mapped previously with oligonucleotide env region probes of the same viral class. At least 22 elements represent new proviral loci, no more than half of which may be solo LTRs, and an additional six may correspond to proviruses identified previously with less specific hybridization probes. Together with proviruses identified previously with env probes, the LTR probe-reactive elements represent the majority of endogenous murine leukemia proviruses in the mouse genome.     

Structural relationship between a normal chicken DNA locus and the transforming gene of the avian acute leukemia virus MC29.

We screened a recombinant chicken DNA/lambda phage library for sequences homologous to the transformation-specific sequences of the avian acute leukemia virus MC29 by hybridization with molecularly cloned MC29 proviral DNA. Three cellular DNA clones were found and compared with each other and with the viral genome by physical mapping with restriction endonucleases and by heteroduplex analysis. These experiments indicated that the three cellular clones overlap and represent a single cellular locus. The RNA genome of MC29 and normal cell DNA share a homologous region of 1.6 kilobases which is interrupted in the cellular DNA by 1.0 kilobase of sequences not present in the viral genome. Hybridization of the cloned cellular DNA to viral RNA and analysis of the protected viral RNA by fingerprinting techniques indicated that there is extensive sequence homology between the helper virus-unrelated mcv sequences of the viral RNA and the cellular DNA, with only minor base differences. The cellular mcv locus, however, lacks all helper virus-related sequences of MC29, including those of the partial viral gag gene which, together with mcv, encodes the probable transforming protein of MC29. We conclude that although the mcv locus of the normal cell does not represent a complete structural homolog to the onc gene of MC29, it is probably the precursor to the onc-specific sequence in the virus.     

Free and integrated recombinant murine leukemia virus DNAs appear in preleukemic thymuses of AKR/J mice.

We studied the appearance and structure of murine leukemia viral genomes in preleukemic AKR/J mice by Southern hybridization. Up to an average of one to two copies per thymocyte of unintegrated murine leukemia virus DNA appears in the thymuses of preleukemic mice beginning at 4 to 5 months of age and disappears in leukemic thymuses. The free viral genomes are absent in the spleens, livers, and brains of preleukemic mice. Using a series of ecotropic and nonecotropic murine leukemia virus hybridization probes, we showed that the unintegrated viral genomes are structurally analogous to those of recombinant mink cell focus-forming viruses that appear as proviruses in leukemic AKR thymocytes, suggesting that these free viral DNAs are the direct precursors to the leukemia-specific proviruses. The mosaic of ecotropic and nonecotropic sequences within these unintegrated viral DNAs varies from one preleukemic thymus to another but often appears structurally homogeneous within individual thymuses, indicating that often each thymus was being infected by a unique mink cell focus-forming virus. Analysis of high-molecular-weight DNA shows that recombinant proviruses reside in the chromosomal DNA of thymocytes within the preleukemic thymus, with the number rising to an average of several copies per thymocyte, but we do not detect any preferred integration sites. These results suggest that, in general, before the development of thymic leukemias in AKR mice there is a massive infection by a unique mink cell focus-forming virus which then integrates into many different sites of individual thymocytes, one of which grows out to become a tumor.     

Two distinct endogenous type C viruses isolated from the asian rodent Mus cervicolor: conservation of virogene sequences in related rodent species.

The cocultivation of a lung cell line from the Southeast Asian mouse Mus cervicolor with cells from heterologous species has resulted in the isolation of two new distinct type C viruses. Both viruses are endogenous to M. cervicolor and are present in multiple copies in the cellular DNA of these mice. One of the viruses, designated M. cervicolor type CI, replicates readily in the SIRC rabbit cell line and is antigenically related to the infectious primate type C viruses isolated from a woolly monkey (simian sarcoma-associated virus) and gibbon apes (gibbon ape leukemia virus). This virus is also closely related by both immunological and nucleic acid hybridization criteria to a type C virus previously isolated from a second Asian murine species, Mus caroli. The isolation of the M. cervicolor type C I virus thus provides further evidence that the infectious primate type C viruses originated by trans-species infection of primates by an endogenous virus of mice. The second virus, designated M. cervicolor type C II, replicates well in various cell lines derived from the laboratory mouse Mus musculus. While antigenically related to type C viruses derived from M. musculus, the M. cervicolor type C II virus isolate can be readily distinguished from standard murine leukemia viruses. Both new type C viruses from M. cervicolor are unrelated to the previously described retrovirus (M432) isolated from the same Mus species. The DNA of M. cervicolor therefore contains multiple copies of at least three distinct classes of endogenous viral genes. An examination of the cellular DNA of other rodent species for nucleic acid sequences related to the genomes of both M. cervicolor type C I and II reveals that both viruses have been highly conserved evolutionarily, and that other species of rodents, such as laboratory mice and rats, contain endogenous virogenes related to those in the DNA of M. cervicolor.