Retrovirus integration and chromatin structure: Moloney murine leukemia proviral integration sites map near DNase I-hypersensitive sites.

The chromatin conformation of mouse genome regions containing Moloney murine leukemia proviral intergration sites in two Mov mouse strains and randomly selected integration sites in virus-infected mouse 3T3 fibroblasts was analyzed. All integrations have occurred into chromosomal regions containing several DNase-hypersensitive sites, and invariably the proviral integration sites map within a few hundred base pairs of a DNase-hypersensitive site. The probability that this close association between proviral integration sites and DNase-hypersensitive sites was due to chance was calculated to be extremely low (2 X 10(-4]. Because the proviral integrations analyzed were not selected for an altered phenotype, our results suggest that DNase-hypersensitive regions are preferred targets for retrovirus integration.     

High molecular weight RNAs from Rous sarcoma virus and Moloney murine leukemia virus contain two subunits.

The molecular weights of the large genomic RNAs from Rous sarcoma and Moloney murine leukemia viruses were determined by a combination of sedimentation coefficients and retardation coefficients from gel electrophoresis. Six RNA standards, ranging from 0.7 X 10(6) to 5.3 X 10(6) daltons, were employed. Studies in the presence of varying concentrations of Mg2+ showed that the method provided valid molecular weights for RNAs of differing amounts of ordered structure. The molecular weight (X 10(-6)) of the high molecular weight RNA complexe from Rous sarcoma virus was 7.6 (+/-0.3) and from murine leukemia virus was 6.9 (+/-0.3). The molecular weights (X 10 (-6) of their Subunits were 3.3 (+/-0.1) and 2.8 (+/-0.2), respectively. Hence, the large complexes consisted of two, not three or more, subunits plus small associated RNAs. The high molecular weight RNA from cloned Rous sarcoma virus was heterogenous in molecular weight although the apparent molecular radius was constant; stuides were performed on subfractions of the RNA as well as on RNA from virus harvested at various time intervals. The preparations with lowest molecular weight approached a mass equal to twice that of the subunit, with hydrodynamic properties approaching those expected of normal single-stranded RNA.     

Highly inducible cell lines derived from mice genetically transmitting the Moloney murine leukemia virus genome.

Permanent, non-virus-producing cell lines have been established from a mouse embryo carrying an endogenous, genetically transmitted Moloney murine leukemia virus (M-MuLV) genome. These cells carry the M-MuLV genome, as demonstrated by hybridization of cellular DNA to M-MuLV complementary DNA, but do not express it at the levels of virus production, accumulation of intracellular viral p30, or M-MuLV-specific RNA. Treatment with bromodeoxyuridine (50 microgram/ml for 24 h) resulted in induction of XC-positive NB-tropic virus, although only a small fraction of the cells released virus (less than 0.1% after 48 h). Immunofluorescent staining and flow microfluorometry indicated that a wave of p30 accumulation occurs in the induced cells, with a maximum at 24 to 48 h after the addition of bromodeoxyuridine. Furthermore, most, if not all, cells were induced to produce p30 protein. Similar kinetics were found for the accumulation of M-MuLV-specific RNA in the cytoplasm of induced cells. This rapid induction of virus expression in a majority of cells was dependent on the presence of the M-MuLV genome and probably represents primarily the expression of this endogenous virus since induction was not observed in cells similarly derived from a sibling embryo lacking the M-MuLV genome.     

Organization and stability of endogenous xenotropic murine leukemia virus proviral DNA in mouse genomes.

In a series of blot hybridization experiments, using a xenotropic envelope probe and restriction enzymes known to cut xenotropic proviral DNA a single time (EcoRI) or not at all (HindIII), we have studied the organization and relationship of endogenous xenotropic env-related sequences in various mouse strains. Multiple copies (18 to 28) of xenotropic env-reactive fragments were found in all mouse DNAs after digestion with either HindIII or EcoRI, and the majority of fragments were of sizes compatible with their origin from full-length proviral DNA. Five HindIII and five EcoRI restriction fragments were common to all inbred mouse DNAs tested. In addition, each strain exhibited unique characteristic xenotropic env-reactive bands; these bands were remarkably stable during many years of inbreeding. The cleavage patterns characteristic of each strain were also useful for showing genealogical relatedness among the various inbred mice.     

Dsi-1, a region with frequent proviral insertions in Moloney murine leukemia virus-induced rat thymomas.

Dsi-1 is a region of chromosomal DNA that underwent proviral insertion in 3 of 24 Moloney murine leukemia virus-induced rat thymomas. In one of these tumors, a provirus is also integrated adjacent to the proto-oncogene c-myc. The proviruses in Dsi-1 have been characterized and appear to be complete. The proviruses were located within a 2-kilobase region that contained four prominent DNase I-hypersensitive sites. These hypersensitive sites were observed in Moloney murine leukemia virus-induced thymomas but not in NRK cells. The region of Dsi-1 immediately 3' to the insertions cross-hybridized with human and chicken DNA, indicating that it contains highly conserved sequences. No evidence could be found for the expression of this highly conserved region. Dsi-1 was mapped to mouse chromosome 4. This location demonstrates that Dsi-1 is different from 16 of the known proto-oncogenes (c-abl, c-erbA c-erbB, c-ets-1, c-ets-2, c-fes, c-fos, c-myb, c-myc, c-raf, A-raf, c-Ha-ras, c-Ki-ras, N-ras, c-sis, and c-src) and 12 cellular regions of tumor-associated integrations in retrovirus-induced tumors (c-erbB, Fis-1, int-1, int-2, Mis-1/pvt-1, Mlvi-1, Mlvi-2, c-mos, c-myb, c-myc, Pim-1, and c-Ha-ras). Hybridization experiments indicated that Dsi-1 is probably different from five additional proto-oncogenes (c-fgr, c-fms, c-mos, neu, and c-yes) and from two additional frequent integration regions (lck and Mlvi-3).     

Characterization of endogenous and exogenous mouse mammary tumor virus proviral DNA with site-specific molecular clones.

Restriction fragments of the mouse mammary tumor virus (MMTV) proviral DNA were obtained by molecular cloning procedures. A 4-kilobase fragment delimited by two PstI sites was isolated from unintegrated, linear MMTV DNA and amplified in the pBr322 plasmid vector. EcoRI fragments of proviral DNA, integrated into the genome of a GR mammary tumor cell line, were isolated as lambda recombinant molecules. Five different recombinant phages which contained the 3' region of the MMTV proviral DNA and adjacent host DNA sequences were isolated. Heteroduplex analysis and S1 nuclease digestion suggested that there is no extensive sequence homology in the host DNA flanking the different proviral genes. The cloned DNA was fractionated into site-specific restriction fragments which served as molecular probes in the analysis of the endogenous MMTV proviral copies of C3H, GR, BALB/c, and feral mice. This allowed the correlation of MMTV-specific EcoRI fragments obtained from genomic DNA of these strains with the 5' and 3' ends of the proviral gene. Restriction fragments of two clones which contained the proviral sequences adjacent to the flanking host DNA as well as 1 to 2 kilobases of host DNA were used as hybridization probes, and the results allow the following conclusions: the proviral DNA of both clones contains nucleotide sequences complementary to the 5' and 3' ends of proviral DNA; and the host DNA flanking one clone belongs to the unique class of genomic DNA, whereas the DNA flanking the second clone is reiterated at least 15 times within the mouse genome.     

The palindromic LTR-LTR junction of Moloney murine leukemia virus is not an efficient substrate for proviral integration.

We generated viral constructs to test the hypothesis that the major substrate on retroviral DNA that is utilized for proviral DNA integration is the palindromic sequence, termed the LTR-LTR junction, normally present in circular molecules formed by joining the two termini of linear proviral DNA. Recombinant viral genomes were built which carried a selectable marker and an extra copy of the LTR-LTR junction from a cloned circular provirus. The junction sequence in each case was positioned such that its use during integration would lead to an easily detected, aberrantly integrated proviral DNA. Analysis of DNA from cells infected with the virus constructs showed that the introduced junction sequence is used at least 1,000-fold less efficiently than the natural sequences at the ends of the genome. This suggests that a linear or more exotic DNA intermediate is most likely the true precursor for the integration reaction.     

Gene product of Moloney murine leukemia virus required for proviral integration is a DNA-binding protein

The 3'-terminal portion of the retroviral pol gene encodes a function required for the formation of the integrated provirus soon after infection of sensitive cells. To permit the isolation of large quantities of the gene product, we expressed various portions of the pol gene of Moloney murine leukemia virus (M-MuLV) as trpE fusion proteins in Escherichia coli. The proteins were found to exhibit strong DNA-binding activity after extraction and renaturation by two different procedures. In the first method, proteins separated by polyacrylamide gel electrophoresis were blotted to nitrocellulose and assayed when bound to the support. The second procedure involved the isolation of proteins in an insoluble fraction, solubilization with guanidine, and renaturation. The characteristics of the binding activity are described and compared with those of authentic viral protein.     

Replication of lengthened Moloney murine leukemia virus genomes is impaired at multiple stages

It has been assumed that RNA packaging constraints limit the size of retroviral genomes. This notion of a retroviral "headful" was tested by examining the ability of Moloney murine leukemia virus genomes lengthened by 4, 8, or 11 kb to participate in a single replication cycle. Overall, replication of these lengthened genomes was 5- to 10-fold less efficient than that of native-length genomes. When RNA expression and virion formation, RNA packaging, and early stages of replication were compared, long genomes were found to complete each step less efficiently than did normal-length genomes. To test whether short RNAs might facilitate the packaging of lengthy RNAs by heterodimerization, some experiments involved coexpression of a short packageable RNA. However, enhancement of neither long vector RNA packaging nor long vector DNA synthesis was observed in the presence of the short RNA. Most of the proviruses templated by 12 and 16 kb vectors appeared to be full length. Most products of a 19. 2-kb vector contained deletions, but some integrated proviruses were around twice the native genome length. These results demonstrate that lengthy retroviral genomes can be packaged and that genome length is not strictly limited at any individual replication step. These observations also suggest that the lengthy read-through RNAs postulated to be intermediates in retroviral transduction can be packaged directly without further processing.     

Fv-1 N- and B-tropism-specific sequences in murine leukemia virus and related endogenous proviral genomes

Oligonucleotide probes specific for the Fv-1 N- and B-tropic host range determinants of the gag p30-coding sequence were used to analyze DNA clones of various murine leukemia virus (MuLV) and endogenous MuLV-related proviral genomes and chromosomal DNA from four mouse strains. The group of DNA clones consisted of ecotropic MuLVs of known Fv-1 host range, somatically acquired ecotropic MuLV proviruses, xenotropic MuLV isolates, and endogenous nonecotropic MuLV-related proviral sequences from mouse chromosomal DNA. As expected, the prototype N-tropism determinant is carried by N-tropic viruses of several different origins. All seven endogenous nonecotropic MuLV-related proviral sequence clones derived from RFM/Un mouse chromosomal DNA, although not recognized by the N probe, showed positive hybridization with the prototype B-tropism-specific probe. The two xenotropic MuLV clones derived from infectious virus (one of BALB:virus-2 and one of AKR xenotropic virus) failed to hybridize with the N- and B-tropic oligonucleotide probes tested and with one probe specific for NB-tropic Moloney MuLV. One of two endogenous xenotropic class proviruses derived from HRS/J mouse chromosomal DNA (J. P. Stoye and J. M. Coffin, J. Virol. 61:2659-2669, 1987) also failed to hybridize to the N- and B-tropic probes, whereas the other hybridized to the B-tropic probe. In addition, analysis of mouse chromosomal DNA from four strains indicates that hybridization with the N-tropic probe correlates with the presence or absence of endogenous ecotropic MuLV provirus, whereas the B-tropic probe detects abundant copies of endogenous nonecotropic MuLV-related proviral sequences. These results suggest that the B-tropism determinant in B-tropic ecotropic MuLV may arise from recombination between N-tropic ecotropic MuLV and members of the abundant endogenous nonecotropic MuLV-related classes including a subset of endogenous xenotropic proviruses.     

Endogenous Moloney leukemia virus in nonviremic Mov substrains of mice carries defects in the proviral genome

Substrains of mice carrying Moloney murine leukemia virus as a Mendelian gene (Mov locus) have been derived previously. Some of these strains, i.e., Mov-3 and Mov-9, develop viremia, whereas others, i.e., Mov-2, Mov-7, and Mov-10, do not regularly activate virus. We previously have molecularly cloned the respective Mov loci and shown that proviral clones derived from the different viral loci were either infectious (Mov-3, Mov-9) or failed to induce infectious virus (Mov-2, Mov-7, Mov-10) in a transfection assay. To analyze the sites affecting infectivity of the latter clones, complementation assays, in vitro recombinations, and marker rescue experiments were performed. Our results show that the three endogenous Moloney murine leukemia virus clones derived from Mov-2, Mov-7, and Mov-10 carry different mutations in the gag-pol region of the proviral genome. No inhibitory effect of flanking mouse sequences on provirus infectivity was observed.     

Nucleotide sequence analysis of endogenous murine leukemia virus-related proviral clones reveals primer-binding sites for glutamine tRNA.

Nucleotide sequences of the region that corresponds to the site of tRNA primer binding for a functional retrovirus were determined in five murine leukemia virus-related sequence clones from mouse chromosomal DNA, which contain a unique 170 to 200-base-pair additional internal segment in the long terminal repeats. The 3'-terminal 18-nucleotide sequence of a major glutamine tRNA isoacceptor was found to match well with the putative primer binding site: 18 of 18 in three clones, 17 of 18 in one clone, and 16 of 18 in one clone. This implies that most of these endogenous proviral sequences of the mouse genome, if replicated as retroviruses, will be different from ecotropic murine leukemia viruses and most mammalian type C retroviruses in using glutamine tRNA, rather than proline tRNA, as a primer.     

An additional dimer linkage structure in Moloney murine leukemia virus RNA.

We have identified an additional dimerization linkage structure in the genome of Moloney murine leukemia virus (MoMLV). Retroviral genomes have long been known to be linked at their 5' ends to form dimers. In MoMLV, a hairpin loop functioning as a dimer linkage structure (DLS) has previously been identified at nucleotides 278-303. Here, we describe RNA dimers formed from sections of the MoMLV 5' untranslated region that do not contain the previously described MoMLV DLS. These dimers exhibit the distinctive characteristics previously described for whole genome dimers. We have mapped this novel region to nucleotides 199-243. This sequence contains a stem-loop structure (nucleotides 204-227) much like the 278-303 region. We describe the chemical and thermal stability of dimers containing the 204-227 stem-loop as well as kinetics and salt-dependence of dimer formation. Our results show that dimerization of MoMLV RNA can be nucleated at multiple sites and suggest that the 5' untranslated region may contain separately folding and dimerizing domains.     

Two genetically transmitted BALB/c mouse mammary tumor virus genomes located on chromosomes 12 and 16.

We have examined EcoRI-restricted cellular DNA from BALB/c mouse-hamster somatic cell hybrids by blot hybridization for the presence of mouse mammary tumor virus-related sequences. Results of this analysis show that mouse mammary tumor virus-related proviral copies are located on chromosomes 16 (16-kilobase-pair fragment) and 12 (10.5- and 7.7-kilobase-pair fragments).     

Isolation of recombinant DNA clones carrying complete integrated proviruses of Moloney murine leukemia virus

EcoRI DNA fragments from a Moloney murine leukemia virus (M-MuLV)-infected mouse fibroblast line (M-MuLV clone A9) were cloned in lambda phage Charon 4A cloning vector to derive clones containing integrated M-MuLV proviral DNA. A 10- to 16-megadalton class of EcoRI fragments was chosen for cloning, based on (i) its ability to induce XC-positive virus upon transfection of NIH/3T3 cells, and (ii) its content of a 0.8-megadalton viral KpnI fragment diagnostic for M-MuLV. Six recombinant DNA clones were isolated which contain a complete M-MuLV provirus, as judged by (i) restriction endonuclease mapping and (ii) the fact that all of the clones gave rise to XC-positive, NB-tropic virus upon DNA infection in NIH/3T3 cells. The sizes of the inserts were 12.0 (for three clones) or 12.5 megadaltons (for three clones). Restriction mapping indicated that these six clones represent five different M-MuLV proviral integrations into different cellular DNA sites.     

Four Moloney murine leukemia virus-infected rat cell clones producing replication-defective particles: protein and nucleic acid analyses.

Four cloned rat cell lines (NX-1 to -4) infected with Moloney murine leukemia virus and defective in virus replication were found to be all different by viral protein and nucleic acid analyses. All four clones produced noninfectious particles and, except for NX-2, at about the same level as wild type. Compared with wild-type virions these defective particles contained larger amounts of gag precursor proteins and very little or no p30 or p15. Analysis of intracellular precursor proteins revealed that NX-2 to -4 synthesized normal Pr65gag, whereas NX-1 produced a slightly smaller precursor. Both NX-1 and NX-4 synthesized an intracellular polyprotein with a size similar to that of wild-type Pr180 gag-pol. Restriction endonuclease analysis of NX-1 to -4 cellular DNA showed that each clone contained a single integrated provirus which possessed large terminal repeat sequences at both the 5' and 3' ends. The proviruses of NX-1 to -3 appeared normal by restriction endonuclease analysis, but NX-4 provirus had a deletion of 1,700 base pairs comprising part of the polymerase region. The noninfectious particles produced by all four clones packaged Moloney viral RNAs and rat RNAs of two different sizes.