Subgenomic DNA Research Articles

CHARACTERISATION AND USE OF AN INTRAGENIC POLYMORPHIC MARKER FOR DETECTION OF CARRIERS OF HAEMOPHILIA B (FACTOR IX DEFICIENCY)

DNA from 33 healthy White subjects was analysed with a 2·5 kilobase subgenomic DNA probe derived from the gene for coagulation factor IX, containing the exon "d" region of that gene. Intragenic Taq I restriction-fragment length polymorphism was revealed, with allelic frequencies estimated at 0·65 and 0·35 (SE = 0·06), also detectable by a cDNA probe. The genomic DNA probe is technically superior to the cDNA probe and has been used in three families with haemophilia B (factor IX deficiency). The polymorphic marker segregates with the deleterious mutation, allowing the identification or exclusion of carriers. The allelic frequencies of the Taq I polymorphism are virtually ideal. Therefore, such a polymorphism should be helpful both in genetic counselling of approximately 40% of affected families and in prenatal diagnosis.

Coestablishment of persistent infection and oncogenic transformation of hamster embryo cells by equine cytomegalovirus

Semipermissive, primary hamster embryo (HE) cells were morphologically transformed in vitro by infection with UV-irradiated equine cytomegalovirus (equine herpesvirus type 2; ECMV). Cell lines (designated EC-1-3) were established independently from foci and were shown to exhibit growth and biological properties typically associated with transformed cells: altered morphology, loss of contact inhibition, increased saturation density, decreased generation time, immortality in culture, normal growth in low concentrations of serum, colony formation in soft agar, and resistance to ECMV superinfection. All ECMV transformed cells were restrictive for the replication of equine herpesvirus type 1(EHV-1) which shares 2.9% homology with ECMV and replicated to high titers in normal HE cells. All EC cell lines were oncogenic in immunocompetent syngeneic LSH hamsters. Tumor cell lines were established from selected malignant fibrosarcomas that developed at the site of injection. Two of the transformed cell lines (EC-2 and EC-3) were found to be persistently infected and to release infectious ECMV from 0.5 to 2% (EC-2) and 0.8 to 5% (EC-3) of the total cell populations. The transformed cell line EC-1 as well as all tumor cell lines were virus nonproducers. Data from DNA-DNA reassociation analyses indicated the presence of 8–32 ECMV genome equivalents per productive EC-2 cell and greater than 300 ECMV genome equivalents per productive EC-3 cell. Small amounts of subgenomic ECMV DNA sequences were detected in the nonproducer EC-1 transformed cells and in all tumor cell lines (EC-1T, -2T, -3T). Some of these DNA sequences must be expressed since ECMV-specific polypeptides were demonstrated in all transformed and tumor cell lines by indirect immunofluorescence using antiserum to ECMV-infected cell extracts and since the sera of tumor-bearing hamsters contained ECMV antibody as detected by an ECMV plaque reduction assay.

MRNA transcripts related to full-length endogenous retroviral DNA in human cells.

Mammalian cells contain multiple copies of endogenous type C retroviral DNA sequences. Among these sequences are complete, potentially infectious proviruses, proviral DNA that is expressed only in the form of viral antigens, retroviral segments that may contribute portions of envelope (env) genes during the generation of recombinant polytropic viruses, and many subgenomic viral DNA segments that may not be expressed at all. We have previously reported the identification and molecular cloning of type C retroviral sequences from human DNA and have shown that the partial nucleotide and deduced amino acid sequences of one of the clones obtained (lambda 51) are homologous to Moloney MuLV (MoMuLV) in the gag and pol regions. The lambda 51 clone as well as several others isolated from a human DNA library contained approximately 4.3 kilobases (kb) of retroviral sequences, were deleted in the env region, and were flanked by tandem repeats unlike the long terminal repeats (LTRs) typically found in proviral DNAs (P.E.S., in preparation). We describe here the characterization of a full-length human retroviral clone (lambda 4-1) containing LTR elements as well as a putative env region. DNA-RNA hybridization experiments reveal that human cells contain species of poly(A)+ RNA that anneal to segments of the full-length retroviral DNA clone.

Discontinuities in the DNA synthesized by an avian retrovirus.

The unintegrated linear DNA synthesized in cells infected by Rous sarcoma virus is a predominantly double-stranded structure in which most of the minus-strand DNA, complementary to the viral RNA genome, is genome sized, whereas the plus-strand DNA is present as subgenomic fragments. We previously reported the application of benzoylated naphthoylated DEAE-cellulose chromatography to demonstrate that of the linear viral DNA species synthesized in quail embryo fibroblasts infected with Rous sarcoma virus greater than 99.5% contain single-stranded regions and these regions are predominantly composed of plus-strand DNA sequences (T. W. Hsu and J. M. Taylor, J. Virol. 44:47-53, 1982). We now present the following additional findings. (i) There were on the average 3.5 single-stranded regions per linear viral DNA, and these single-stranded regions could occur at many locations. (ii) With a probe to the long terminal repeat, we detected, in addition to a heterogeneous size distribution of subgenomic plus-strand DNA species, at least three prominent discrete size classes. Each of these discrete species had its own specific initiation site, but all had the same termination site. Such species were analogous to those reported by Kung et al. (J. Virol. 37: 127-138, 1981). (iii) These discrete size classes of plus-strand DNA were present not only on the major size class of linear DNA but also on a heterogeneous of slower-sedimenting species, which we have called immature linears. Our interpretation is that we have thus detected several additional sites for the initiation of plus-strand DNA. (iv) The 340-base plus-strand strong-stop DNA was only found associated with the immature linears. (v) From a size and hybridization comparison of these discrete size classes of plus-strand DNA with minus-strand DNA species, as synthesized in the endogenous reaction of melittin-disrupted virions, it was found that the putative additional initiation sites for plus-strand DNA synthesis corresponded to many of the pause sites in the synthesis of minus-strand DNA.

Transcription in mouse embryo cells permissively infected by murine cytomegalovirus.

The sites of transcription and abundance of steady-state cytoplasmic viral RNA in murine cytomegalovirus (Smith strain) infected mouse embryo cells were analyzed. Cloned subgenomic DNA fragments were labeled with 32P and hybridized to filters containing polyadenylated RNA extracted from cells at immediate-early, early, and late times in the infection. The pattern of transcription was distinctive at each time point. Immediate-early transcription occurred primarily at 0.770-0.816 map units. Minor sites of immediate-early transcription were clustered at 0.671-0.861 map units and at the termini of the genome (0.944-0.002 map units). Early transcripts were detected from the entire genome with the exception of a fragment at 0.278-0.305 map units. The site of major immediate-early transcription at 0.770-0.816 map units was represented less abundantly while the concentration of RNA from most other sites of immediate-early transcription was increased at the early time point. The most abundant site of early transcription was at 0.840-0.861 map units. RNA from late in the infection hybridized to all subgenomic fragments. The highest concentration of late RNA transcripts was detected with fragments located at 0.444-0.770 map units. In contrast, late RNA transcripts from both ends of the genome were present at a concentration equal to or lower than that seen at the early time point, and the concentration of late RNA from the major early site (0.840-0.861 map units) was significantly decreased. We also detected uninfected mouse cell RNA with three separate subgenomic EcoRI fragments.

Mutant feline sarcoma proviruses containing the viral oncogene (v-fes) and either feline or murine control elements.

The sequences required for transformation by the Gardner-Arnstein (GA) strain of feline sarcoma virus (GA-FeSV) were defined by site-directed, in vitro mutagenesis of molecularly cloned proviral DNA. Portions of the Ga-FeSV provirus, subcloned in the plasmid pBR322, were mutagenized by deletion or frameshift at XhoI restriction sites flanking the nucleotide sequences presumed to encode the GA-FeSV transforming polyprotein (P108(gag-fes)). The biological activity of subgenomic and reconstructed full-genome-length molecules was assayed by transfection and focus induction in NIH 3T3 cells. Both mutant and wild-type molecules containing the intact P108(gag-fes) coding region induced foci of transformed cells at efficiencies between 10(4) and 10(5) focus-forming units per pmol of DNA; a deletion mutant lacking 3'-terminal v-fes sequences was completely nontransforming in parallel assays. Representative subcloned foci of transformed NIH 3T3 cells synthesized P108(gag-fes) with associated in vitro protein kinase activity. Focus-forming viruses could be rescued from transformed subclones induced by full-length proviral DNA, but not from cells transformed by subgenomic DNA lacking a 3' long terminal repeat (LTR). It was concluded that: (i) nucleotide sequences encoding P108(gag-fes) and its associated kinase activity are responsible for transformation, (ii) the GA-FeSV 3' env and LTR sequences are not required for focus induction, and (iii) the 3' LTR is necessary for rescue of infectious FeSV RNA. A chimeric DNA containing the 5' LTR and P108(gag-fes) coding region of GA-FeSV joined to the 3' LTR of Moloney murine sarcoma virus was both transforming and rescuable at high efficiency. Restriction analysis showed that passaged stocks of rescued transforming virus contained Moloney murine sarcoma virus U3 sequences at both proviral DNA termini, consistent with generally accepted models for LTR formation during reverse transcription. Wild-type GA-FeSV and the chimeric virus (here designated as GAHT), each rescued from NIH 3T3 cells with the same amphotropic murine leukemia virus, yielded approximately equal numbers of foci when titrated on CCL 64 mink cells. By contrast, on mouse NIH 3T3 cells, the focus-forming titer of GAHT was 1 to 2 log higher than that of FeSV. The foci induced on NIH 3T3 cells by GAHT appeared earlier and were reproducibly larger than those induced by GA-FeSV. Differences in transforming activity on NIH 3T3 cells were also found using colony formation in agar, showing that the more rapid appearance and larger size of foci formed in liquid media were not due to virus spread. These data suggest that transcriptional control signals within the viral LTR regulate the levels of the transforming gene product in a species-specific manner.

Transcription in human fibroblasts permissively infected by human cytomegalovirus strain AD169

We used cloned subgenomic DNA fragments of human cytomegalovirus strain AD169 as hybridization reagents to analyze the sites of transcription and abundance of viral RNA in permissively infected human embryonic lung cells. RNA extracted from immediate early, early, middle, and late times in the infection was attached to filters and hybridized with excess cloned subgenomic fragments. Each hybridization was performed with an internal standard to allow quantitation of the RNA concentration and standardization for the variation in probe complexity and specific activity. We found immediate early transcription in the AD169 strain occurring primarily from three regions in the long unique segment at 0.061–0.110, 0.117–0.142, and 0.588–0.616 map units. A low level of transcription was also detected from the long unique segment at 0.230–0.372, 0.419–0.437 and 0.703–0.707 map units and the short unique segment and terminal repeats at 0.00–0.046, 0.776–0.854, and 0.892–1.00 map units. At 8 hr after infection in the presence of de novo protein synthesis, the transcription pattern was changed. Two of the major immediate early sites at 0.061–0.110 and 0.117–0.142 map units were represented less abundantly in the absence of cycloheximide while the site at 0.588–0.616 map units was represented at the same level. The most abundant early RNA was synthesized from the terminal repeat sequences and part of the short unique segment at 0.00–0.046, 0.776–0.822, and 0.892–1.00 map units. Steady-state RNA from the midpoint of the infection hybridized with most regions of the genome. Abundantly transcribed regions included the abundant early sites mentioned above (0.00–0.046, 0.588–0.616, 0.776–0.822, and 0.892–1.00 map units), several new sites in the long unique segment at 0.104–0.110, 0.576–0.588, 0.653–0.664, 0.703–0.707, and 0.766–0.776 map units and the L-S junction fragment at 0.804–0.854 map unit. RNA from late in the infection hybridized to all subgenomic fragments. The sites of de novo RNA synthesis along the genome were determined by hybridizing pulse-labeled RNA to individual cloned subgenomic fragments attached to filters. Transcription from the terminal repeat sequences and the long unique segment at 0.576–0.588 map unit accounted for 40% of the total viral RNA synthesized from 22 to 28 hr after infection. Of the total de novo RNA synthesis from 71 to 77 hr after infection, 4.3% was virus specific and 47% of the total de novo synthesized viral RNA hybridized to the terminal repeat sequences.

Nuclei purified from cauliflower mosaic virus-infected turnip leaves contain subgenomic, covalently closed circular cauliflower mosaic virus DNAs.

Nuclei isolated from cauliflower mosaic virus (CaMV) infected turnip leaves contain subgenomic CaMV DNA species in addition to the genome length CaMV DNA. These subgenomic CaMV DNA species are present as covalently closed circles (form I), relaxed circles (form II) and linear (form III) molecules. The subgenomic form I DNA species range in size from about 10% of genome length to nearly genome length. These subgenomic DNA species appear in tissue infected with cloned CaMV DNA, indicating that they arise rapidly and have not accumulated in the virus population from serial propagation of CaMV. No specific region of the CaMV genome appears to be preferentially deleted to form the subgenomic CaMV DNA species. At least three distinct subgenomic species appear to accumulate preferentially in nuclei isolated from infected tissue. Two of these abundant subgenomic CaMV DNA species are form I and the other one is form III. Some of the subgenomic CaMV DNA species appear to be minichromosomes.

Tumor induction by direct injection of cloned v-src DNA into chickens.

The in vitro transforming potential of Rous sarcoma viruses has been shown to reside in a gene segment, v-src. To provide direct proof that the isolated v-src sequence contains all the information necessary for sarcoma induction in vivo, we have injected subgenomic and genomic cloned DNA carrying v-src directly into chickens. We report that v-src, without the complements of other viral genes, can effectively induce sarcomas in the host within weeks after injection. The sarcoma tumors generated by subgenomic v-src DNA are of mono- or oligoclonal origin. They have acquired v-src sequences in the genome and expressed high levels of src messages. Intact long terminal repeats do not seem to be essential in the tumorigenic process.

Transformation of human embryonic fibroblasts by BK virus, BK virus DNA and a subgenomic BK virus DNA fragment.

Human embryonic fibroblasts (HEF) have been transformed by BK virus (BKV) DNA and by u.v.-inactivated or live BKV alone or in association with methyl-cholanthrene (MTC). The transformed cells produced BKV large T and small t antigens as well as the cellular 53 kdal protein, detected by immunofluorescence and immunoprecipitation. After an initial phase of lysis and virus shedding, virus or its coat protein antigen could not be detected in transformed cells. All human transformed cell lines could be superinfected by BKV or BKV DNA, but their susceptibility to superinfection was 20- to 500-fold lower than normal HEF. BKV could be rescued by fusion of transformed cells with normal HEF or Vero cells and by transfection of normal HEF with total DNA and DNA extracted from the Hirt supernatant of transformed cells. Blot hybridization analysis of DNA from transformed cells showed a considerable amount of free BKV DNA in monomeric and polymeric forms. Integrated BKV DNA was absent in most cell lines but present in only small amounts in BKV-transformed cells treated with MTC. Analysis of free BKV DNA with various restriction endonucleases and by blot hybridization showed that monomeric forms were complete BKV genomes, whereas polymers contained both complete and defective or rearranged BKV DNA. Transformation of HEF was also obtained with a 3.7 kilobase (kb) fragment of the BKV genome, produced by sequential digestion of BKV with the restriction endonucleases HhaI and EcoRI. This fragment extends clockwise on the virus genome from 0 to 72.2 map units and contains the entire early region. Blot hybridization analysis of cells transformed by the HhaI/EcoRI 3.7 kb fragment showed two separate integrations of BKV sequences without free virus DNA.

Biological activity of cloned rat endogenous C-type virus DNA transferred by microinjection.

The biological activity of a molecularly cloned DNA of a rat endogenous C-type leukaemia helper virus, RHHV, was assessed by intranuclear microinjection into normal rat kidney cells (NRK153). Release of rat C-type leukaemia helper viruses by the microinjected cells was examined by superinfection of Kirsten-transformed non-producer cells (K-NRK). Immediate release of helper leukaemia viruses at a very low level was observed only in the NRK153 m3 . 5/cir cells microinjected with the supercoiled form of RHHV DNA in toto, suggesting that the circular form of the virus DNA might have expedited the replication and expression of virus particles. Genome rescue experiments were also performed by co-cultivating the microinjected NRK153 m cells carrying various linear RHHV DNAs, in toto or of subgenomic sizes, with K-NRK cells. The results indicated that both the total and the 5.8 to 6.2 kb DNA fragment proximal to the 5' terminus of the cloned RHHV 8.8 kb DNA were able to rescue successfully a transforming replication-competent pseudotype virus. Subgenomic DNA fragments derived from the centre or the 3' end of the RHHV DNA were ineffective in the genome rescue experiments.

Discrete subgenomic DNA fragments in incomplete particles of adenovirus type 2.

Discrete subgenomic DNA fragments were found in three out of thirty-two preparations of adenovirus type 2 incomplete particles grown in human Hep-2 cells and examined over the course of 1 year. One preparation contained three fragments corresponding to 5, 14 and 19% of the genome, another contained a 37% fragment and the third a 40% fragment. Each fragment hybridized exclusively to the left end of the genome. Digestion of the nick-translated 37% fragment with HindIII confirmed that it contained the left 37.3% of the genome. Synthesis of these fragments was not dependent on high input multiplicity of infection. Comparable fragments were not found in unpackaged DNA from the corresponding infected cells. This is consistent with the hypothesis that such fragments are generated during virus assembly or, alternatively, may reflect the very small proportion of these fragments relative to the pool of unpackaged DNA within the cells. The possibility that they are generated by errors in DNA replication is discussed.

Envelope gene sequences which encode the gp52 protein of spleen focus-forming virus are required for the induction of erythroid cell proliferation.

A series of insertion-deletion mutants was constructed in a molecularly cloned DNA copy of the Friend strain of spleen focus-forming virus (SFFV). The mutants were produced by inserting a synthetic oligonucleotide linker containing the recognition sequence of SalI endonuclease into several different locations of the SFFV DNA. Three classes of mutants were isolated: insertion-deletion mutants in the 5' half of the SFFV genome, in the long terminal repeat of the SFFV genome, and in the env gene of the SFFV genome. The env gene mutant has a deletion of sequences shared in common between the env gene of SFFV and the env genes of mink cell focus-inducing murine leukemia viruses. From analyses of the biological activity of the various mutants and a biologically active subgenomic SFFV DNA fragment described herein, we can deduce that the coding sequence encompassing the env gene of SFFV is required for the biological activity. This region, required for the pathogenic phenotype, cannot be larger than 1.5 kilobase pairs, a size only slightly more than that sufficient to encode the nonglycosylated precursor of the gp52 env gene product.

Physical mapping of human adenovirus type 7 DNA using a simple and sensitive method of terminal labeling

The tyrosine-containing peptide covalently attached to each 5′-terminus of adenovirus type 7 (Ad 7; Greider) DNA was labeled with 125I. The 5′-labeled DNA was subjected to digestion with several restriction endonucleases and the size of the labeled terminal fragments was determined. Partial hydrolysis by these endonucleases generated a series of labeled fragments which were fused to the terminal fragments and could, therefore, be detected by autoradiography. From the sizes of the partial products the location of the cleavage sites of the enzymes on Ad7 DNA could be determined. The subgenomic DNA extracted from incomplete particles by protease treatment could also be labeled with 125I, since it was found to contain the tyrosine-containing peptide covalently attached to the preferentially packaged left end of the genome.

Deletions within the transformation-specific RNA sequences of acute leukemia virus MC29 give rise to partially transformation-defective mutants.

The viral RNAs of three nonconditional mutants of avian myelocytomatosis virus MC29 were analyzed. These mutants, which were originally isolated from the quail producer line Q10 and were designated 10A, 10C, and 10H, have lost most of the ability to transform hematopoietic cells in vitro and to induce tumors in vivo, but they still transform cultured fibroblasts with the same efficiency as wild-type (wt) MC29. Electrophoretic analyses showed that the mutant genomic RNAs were smaller than the 5.7-kilobase genome of wt MC29; the genomes of mutants 10A, 10C, and 10H were about 5.5, 5.3, and 5.1 kilobases long, respectively. Analyses of the transformation-specific sequences of these mutant RNAs by a combination of T(1) oligonucleotide fingerprinting and hybridization with cDNA from the transformation-specific sequences myc of wt MC29 or competition hybridization including wt MC29 RNA revealed that deletions of myc-specific sequences had occurred. The deletions in all three mutants overlapped, since they all had lost one particular myc-specific oligonucleotide. In agreement with the size of the genomic RNAs, mutants 10C and 10H had lost two additional myc oligonucleotides, and mutant 10A contained a modified myc oligonucleotide. The locations of the deletions were deduced from comparisons with previously established oligonucleotide maps of several members of the MC29 subgroup of acute leukemia viruses and by hybridization of wt and mutant RNAs to molecularly cloned subgenomic fragments of wt MC29 proviral DNA, representing the 5' and 3' domains of the myc sequence. We found that the deleted sequences represented overlapping internal segments of the myc sequence and that the borders of myc with the partial complements of the virion genes gag and env appeared to be conserved in mutant and wt MC29 RNAs. The correlation between the altered transforming potential for hematopoietic cells and the partial deletion of myc in the mutant RNAs provided direct genetic evidence for the involvement of myc in oncogenesis. However, the unaffected efficiency of these mutants in fibroblast transformation suggested that the deleted sequences are not essential for the fibroblast-transforming potential of the onc gene of MC29.

Mouse cells transformed by bovine papillomavirus contain only extrachromosomal viral DNA sequences.

The viral DNA sequences in mouse C127 cells transformed by bovine papillomavirus type 1 (BPV-1) virions, by full-length linear BPV-1 DNA, or by a defined transforming subgenomic DNA segment of BPV-1 were examined by reassociation kinetics and blot hybridization. In all cases, the transformed cells contained multiple copies of BPV-1 DNA, present exclusively as supercoiled or nicked circular extrachromosomal molecules or as a slowly migrating complex of circular viral DNA molecules. In the transformed cell lines established from cells transfected with full-length linear BPV-1 DNA, there was recircularization of the input DNA which in some cases resulted in the loss of the restriction site used in the linearization of the DNA. In the transformed cell lines established with the defined subgenomic segment there was circularization of the DNA accompanied by the acquisition of new sequences or duplication and rearrangement of the BPV-1 sequences. In contrast to other well-studied virus transformation systems, no integration of the BPV-1 genome into the host chromosome could be detected under conditions sensitive enough to detect 0.1-0.2 viral genome equivalent. It was concluded that maintenance of transformation may be mediated by nonintegrated viral DNA.

Biological activity of the spleen focus-forming virus is encoded by a molecularly cloned subgenomic fragment of spleen focus-forming virus DNA.

A biologically active subgenomic DNA fragment of a polycythemia-inducing strain of the replication-defective spleen focus-forming virus (SFFV) has been molecularly cloned. The SFFV DNA fragment includes 2.0 kilobase pairs (kbp) from the 3' end of SFFV, the long terminal repeat sequences of SFFV, and 0.4 kbp from the 5' end of SFFV. The fragment contains the previously described env-related gene of SFFV. All the properties associated with SFFV can be assigned to this SFFV DNA fragment by using a two-stage DNA transfection assay with infectious helper virus DNA. The virus recovered from the transfection assays can induce erythroblastosis, splenic foci, and polycythemia in infected mice. Fibroblast cultures transfected with the SFFV DNA fragment synthesize gp52, the known intracellular product of the env-related gene of SFFV. gp52 can also be detected in spleens from diseased mice infected with the virus recovered in the two-stage transfection. The results are consistent with the hypothesis that the env-related gene sequences of SFFV and their product gp52 are required for the initiation of SFFV-induced disease.

Cellular transformation by subgenomic feline sarcoma virus DNA.

The genome of the Snyder-Theilen strain of feline sarcoma virus (ST-FeSV) is a 4.3-kilobase-pair (kbp) RNA molecule that contains a 1.5-kbp cellular insertion (fes gene) flanked by feline leukemia virus sequences at its 5' end (1.6 kbp) and 3' end (1.2 kbp) (Sherr et al., J. Virol. 34:200-212, 1980). DNA transfection techniques have been utilized to determine the regions of the ST-FeSV genome involved in malignant transformation. I have found that the 3.7-kbp 5'-end fragment of the ST-FeSV provirus (which corresponds to the 3.4-kbp 5'-end fragment of the viral genome) is sufficient to transform NIH/3T3 fibroblasts. Enzymes that cleave the ST-FeSV provirus DNA within the feline leukemia virus gag gene sequences or within the fes gene abolished the transforming activity. Preservation of the proviral large terminal repeats was also required for transformation. Transformed NIH/3T3 cells obtained by transfection of total or subgenomic ST-FeSV DNA expressed normal levels of the ST-FeSV gene product ST P85 and of its associated protein kinase activity. Furthermore, these cells contained high levels of phosphotyrosine residues, a biochemical marker associated with cellular transformation induced by certain retroviruses including ST-FeSV. These results, taken together, strongly support the concept that only those ST-FeSV proviral sequences necessary for ST P85 expression are involved in malignant transformation.

Synthesis of plus strands of retroviral DNA in cells infected with avian sarcoma virus and mouse mammary tumor virus.

The vast majority of plus strands synthesized in quail cells acutely infected with avian sarcoma virus were subgenomic in size, generally less than 3 kilobases (kb). A series of discrete species could be identified after agarose gel electrophoresis by annealing with various complementary DNAs, indicating specificity in the initiation and termination of plus strands. The first plus strand to appear (within 2 h postinfection) was similar in length to the long redundancy at the ends of linear DNA (0.35 kb), and it annealed with complementary DNAs specific for the 3' and 5' termini of viral RNA (Varmus et al., J. Mol. Biol. 120:50-82, 1978). Several subgenomic plus-strand fragments (0.94, 1.38, 2.3, and 3.4 kb) annealed with these reagents. At least the 0.94- and 1.38-kb strands were located at the same end of linear DNA as the 0.35-kb strand, indicating that multiple specific sites for initiation were employed to generate strands which overlapped on the structural map. We were unable to detect RNA liked to plus strands isolated as early as 2.5 h postinfection; thus, the primers must be short (fewer than 50 to 100 nucleotides), rapidly removed, or not composed of RNA. To determine whether multiple priming events are a general property of retroviral DNA synthesis in vivo, we also examined plus strands of mouse mammary tumor virus DNA in chronically infected rat cells after induction of RNA and subsequent DNA synthesis with dexamethasone. In this case, multiple, discrete subgenomic DNA plus strands were not found when the same methods applied to avian sarcoma virus DNA were used; instead, the plus strands present in the linear DNA of mouse mammary tumor virus fell mainly into two classes: (i) strands of ca. 1.3 kb which appeared early in synthesis and were similar in size and genetic content to the terminally repeated sequence in linear DNA; and (ii) plus strands of the same length as linear DNA. A heterogeneous population of other strands diminished with time, was not found in completed molecules, and was probably composed of strands undergoing elongation. These two retroviruses thus appear to differ with respect to both the number of priming sites used for the synthesis of plus strands and the abundance of full-length plus strands. On the other hand the major subgenomic plus strand of mouse mammary tumor virus DNA (1.3 kb) is probably the functional homolog of a major subgenomic plus strand of avian sarcoma virus DNA (0.35 kb). The significance of this plus strand species is discussed in the context of current models which hold that it is used as a template for the completion of the minus strand, thereby generating the long terminal redundancy.

Identification of ecotropic proviral sequences in inbred mouse strains with a cloned subgenomic DNA fragment.

A specific probe for detecting ecotropic murine leukemia virus sequences was constructed by cloning a 500-base-pair DNA segment, corresponding to a portion of the env region of the AKR ecotropic virus, in a pBR322/Escherichia coli K-12 host/vector system. This probe was used to screen the cellular DNAs of six inbred strains of mice for the presence of ecotropic retroviral DNA sequences by the Southern blot hybridization procedure. Three copies of ecotropic viral DNA were detected in AKR/N (a high-ecotropic virus strain) and two were found in BALB/c (a low-ecotropic virus strain) DNAs. As expected, no sequences reactive with this probe were found in NFS mouse DNA (a virus-negative strain). However, cellular DNA sequences that reacted strongly with the ecotropic-specific DNA probe were detected in certain NZB, C57L, and 129 mice (all virus-negative strains). In contrast to the reactive sequences in AKR and BALB/c, the reactions were chiefly associated with EcoRI segments that were subgenomic in size.