Thymotropism of murine leukemia virus is conferred by its long terminal repeat

Rearrangements and insertions in the Moloney murine leukemia virus long terminal repeat alter biological properties in vivo and in vitro

The effects of rearrangement and insertion of sequences in the Moloney murine leukemia virus (M-MuLV) long terminal repeat (LTR) were investigated. The alterations were made by recombinant DNA manipulations on a plasmid subclone containing an M-MuLV LTR. Promoter activity of altered LTRs was measured by fusion to the bacterial chloramphenicol acetyltransferase gene, followed by transient expression assay in NIH 3T3 cells. M-MuLV proviral organizations containing the altered LTRs were also generated, and infectious virus was recovered by transfection. Infectivity of the resulting virus was quantified by XC plaque assay, and pathogenicity was determined by inoculating neonatal NIH Swiss mice. Inversion of sequences in the U3 region containing the tandemly repeated enhancer sequences (-150 to -353 base pairs [bp]) reduced promoter activity approximately fivefold in the transient-expression assays. Infectious virus containing the inverted sequences (Mo- M-MuLV) showed a 20-fold reduction in relative infectivity compared with wild-type M-MuLV, but the virus still induced thymus-derived lymphoblastic lymphoma or leukemia in mice, with essentially the same kinetics as for wild-type M-MuLV. We previously derived an M-MuLV which carried inserted enhancer sequences from the F101 strain of polyomavirus (Mo + PyF101 M-MuLV) and showed that this virus is nonleukemogenic. In Mo + PyF101 M-MuLV, the PyF101 sequences were inserted between the M-MuLV promoter and the M-MuLV enhancers (at -150 bp). A new LTR was generated in which the PyF101 sequences were inserted to the 5' side of the M-MuLV enhancers (at -353 bp, PyF101 + Mo M-MuLV). The PyF101 + Mo LTR exhibited promoter activity similar (40 to 50%) to that of wild-type M-MuLV, and infectious PyF101 + Mo M-MuLV had high infectivity on NIH 3T3 cells (50% of wild type). In contrast to the nonleukemogenic Mo + PyF101 M-MuLV, PyF101 + Mo M-MuLV induced leukemia with kinetics similar to that of wild-type M-MuLV. Thus, the position of the PyF101 sequences relative to the M-MuLV LTR affected the biological behavior of the molecular construct. Furthermore, PyF101 + Mo M-MuLV induced a different spectrum of neoplastic disease. In comparison with wild-type M-MuLV, which induces a characteristic thymus-derived lymphoblastic lymphoma with extremely high frequency, PyF101 + Mo M-MuLV was capable of inducing both acute myeloid leukemia or thymus-derived lymphoblastic lymphoma, or both. Tumor DNA from both the PyF101 + Mo- and Mo- M-MuLV-inoculated animals contained recombinant proviruses with LTRs that differed from the initially inoculated virus.

Murine leukemia virus long terminal repeat sequences can enhance gene activity in a cell-type-specific manner

We tested the ability of sequences in the long terminal repeat (LTR) of a mink cell focus-forming (MCF) murine leukemia virus to function as an enhancer in a cell-type-specific manner. In a stable transformation assay, the MCF or Akv LTR and the simian virus 40 enhancer had similar activities in murine fibroblasts. In contrast, the MCF LTR had a significantly greater activity in murine T lymphoid cells than did either the simian virus 40 enhancer or the Akv LTR.

The neurovirulent determinants of ts1, a paralytogenic mutant of Moloney murine leukemia virus TB, are localized in at least two functionally distinct regions of the genome

To better understand the molecular mechanism involved in retrovirus ts1-induced paralytic disease in mice, we constructed a panel of recombinant viruses between ts1 and the wild-type viruses Moloney murine leukemia virus (MoMuLV) and MoMuLV-TB, a strain of MoMuLV. These recombinant viruses were constructed in an attempt to identify the sequence(s) in the genome of ts1 which contains the critical mutation(s) responsible for the neurovirulence of ts1. Two functionally distinct sequences in the genome of ts1 were found to be responsible for its paralytogenic ability. One of these sequences, the 0.77-kilobase-pair XbaI-BamHI (nucleotides 5765 to 6537) fragment which encodes the 5' half of gp70 and 11 base pairs upstream of the env gene coding sequence, determines the inability of ts1 to process Pr80env. The other sequence, the 2.30-kilobase-pair BamHI-PstI (nucleotides 538 to 8264 and 1 to 567) fragment, which comprises nearly two-thirds of the env gene, the long terminal repeat, and the 5' noncoding sequence, determines the enhanced neurotropism of ts1. Replacement of any one of these two regions with the homologous region from either one of the two wild-type viruses resulted in recombinant viruses which either totally failed to induce paralysis or induced a greatly attenuated form of paresis in some of the infected mice.

AKR ecotropic murine leukemia virus SL3-3 forms envelope gene recombinants in vivo

The ecotropic AKR virus SL3-3 was injected into neonatal mice of the high-leukemia strains HRS/J and CWD/J and the low-leukemia strains CBA/J, SEA/J, and NIH Swiss. SL3-3 was highly leukemogenic in each strain, and 90% of the inoculated animals died by 6 months of age. T1 oligonucleotide fingerprint analysis of the genomic RNAs of viruses recovered from 9 of 13 leukemic animals revealed the presence of the SL3-3 virus and recombinant viruses with polytropic virus-related envelope gene sequences. Recombinant proviruses were detected by the Southern blot technique in the DNAs of 17 of 18 tumors. The pattern of substitutions within the envelope genes of the SL3-3 recombinant viruses appeared to be dependent on the strain of the animal. These observations indicate that the SL3-3 virus formed envelope gene recombinants in vivo in each of the strains that were studied. However, the role of these recombinants during leukemogenesis remains to be defined.

Nucleotide sequences of a feline leukemia virus subgroup A envelope gene and long terminal repeat and evidence for the recombinational origin of subgroup B viruses

Molecular clones of the subgroup A feline leukemia virus FeLV-A/Glasgow-1 have been obtained. Nucleotide sequence analysis of the 3' end of the proviral genome and comparison with the published sequence of FeLV-B/Gardner-Arnstein showed that the most extensive differences are located within the 5' domain of the env gene. Within this domain, several divergent regions of env are separated by more conserved segments. The 3' end of env is highly conserved, with only a single amino acid coding difference in p15env. The proviral long terminal repeats are also highly conserved, differing by only eight base substitutions and one base insertion. Specific probes constructed from the FeLV-A or FeLV-B env genes were used to compare the env genes of various exogenous FeLV isolates and the endogenous FeLV-related proviruses of normal cat DNA. An FeLV-A-derived env probe showed no hybridization to normal cat DNA but detected all FeLV-A and FeLV-C isolates tested. In contrast, an FeLV-B env probe detected independent FeLV-B isolates and a family of endogenous FeLV-related proviruses. Our observations provide strong evidence to support the hypothesis that FeLV-B viruses have arisen by recombination between FeLV-A and endogenous proviral elements in cat DNA.

Isolation of a retroviruslike sequence from the TL locus of the C57BL/10 murine major histocompatibility complex

Two retroviruslike sequences have been isolated from the TL locus of the major histocompatibility complex of C57BL/10 mice. One sequence (TLev2) hybridizes only with probes derived from the pol region of the murine leukemia provirus AKR; the other sequence (TLev1) hybridizes with gag, pol, and env AKR region probes. This 9-kilobase endogenous, TL region-associated virus (TLev1) has been further characterized. The TLev1 genome has been shown to contain murine leukemia virus-related sequences bounded by retroviruslike, VL30 long terminal repeats. Hybridization of TLev1-derived probes to mouse genomic digests reveals multiple copies which show distinct patterns compared with those observed with murine leukemia virus probes. The study of TLev1 may prove significant with respect to the interaction of retroviral sequences within the genome, expression of genes within the TL locus, and polymorphisms within the major histocompatibility complex.

Studies on emerging radiation leukemia virus variants in C57BL/Ka mice

To analyze the emergence of radiation leukemia virus (RadLV) variants in primary X-ray-induced C57BL/Ka thymoma and to identify the virus responsible for the very high leukemogenic potential of passaged Kaplan strain BL/VL3 preparation, we cloned several primary and passaged ecotropic RadLV infectious genomes. By restriction analysis, we found that BL/VL3 cells harbor three related but different ecotropic RadLVs. Their restriction map differs significantly from those of primary RadLVs. Hybridization analysis also indicated that BL/VL3 and primary RadLVs differ in their p15E and long terminal repeat (LTR) regions. As compared with the LTR sequence of the putative parental endogenous ecotropic provirus, the LTR sequence of primary weakly leukemogenic RadLV has only one change, a C-rich sequence, generating a 6-base-pair direct repeat just in front of the promotor. The LTR of the primary nonleukemogenic RadLV only showed few base changes, mainly clustered in R and U5. The LTR from a moderately leukemogenic passaged BL/VL3 RadLV had conserved the C-rich sequence and acquired a 43-base-pair direct repeat in U3 and several other point mutations, small insertions, and deletions scattered in U3, R, and U5. All cloned primary RadLVs were fibrotropic, and some were weakly leukemogenic. All cloned BL/VL3 RadLVs were thymotropic and nonfibrotropic. The block of their replication was found to be after the synthesis of unintegrated linear and supercoiled viral DNA. Most of the BL/VL3 RadLVs were moderately leukemogenic, and one (V-13) was highly leukemogenic, being as virulent as the Moloney strain. We propose a model for the emergence of the RadLV variants and show that the virus responsible for the high leukemogenic potential of BL/VL3 preparation is a nondefective, ecotropic, lymphotropic, nonfibrotropic, unique retrovirus which most likely arose from a parental primary RadLV similar to those studied here.

Two elements in the bovine leukemia virus long terminal repeat that regulate gene expression

The bovine leukemia virus, like the human T-cell leukemia viruses (HTLV-I and HTLV-II), are unusual biologically in that viral transcripts are not detected in tumors or infected tissues. The bovine leukemia virus long terminal repeat (BLV LTR) functions as a transcriptional promoter only in cell lines productively infected with BLV. Deletion mapping indicated that at least two regions of the LTR, on the 5' and 3' sides of the RNA start site, influenced gene expression. An analysis has now been made of the effects of coupling sequences from these LTR regions to a heterologous core promoter derived from the SV40 early promoter unit. Through the use of the transient expression of the bacterial chloramphenicol acetyltransferase (CAT) gene to monitor transcriptional activity in vivo, two independent, regulatory elements were identified in the BLV LTR. One was present in a fragment of 75 base pairs derived from the U3 region of the LTR and behaved much like other enhancer elements. It may be a major determinant of BLV expression in productively infected cell lines, since it enhanced transcription controlled by the heterologous core promoter only in these cells. The second element was contained in a 250-bp fragment derived from LTR sequences in the R region, located downstream from the RNA start site. Its activation of CAT expression was not dependent on BLV infection and was evident only when the fragment was located immediately downstream from the RNA start site. BLV expression thus appears to be regulated in part by a cell-specific enhancer element upstream from the core promoter and a novel sequence downstream from the RNA initiation site in the viral LTR.

Ecotropic and mink cell focus-forming murine leukemia viruses integrate in mouse T, B, and non-T/non-B cell lymphoma DNA

Structures of somatically acquired murine leukemia virus (MuLV) genomes present in the DNA of a large panel of MuLV-induced C57BL and BALB/c B and non-T/non-B cell lymphomas were compared with those present in MuLV-induced T-cell lymphomas induced in the same low-"spontaneous"-lymphoma-incidence mice. Analyses were performed with probes specific for the gp70, p15E, and U3-long terminal repeat (LTR) regions of ecotropic AKV MuLV and a mink cell focus-forming virus (MCF)-LTR probe annealing with U3-LTR sequences of a unique endogenous xenotropic MuLV, which also hybridizes with U3-LTR sequences of a substantial portion of somatically acquired MCF genomes in spontaneous AKR thymomas. The DNAs of both T- and B-cell tumors induced by neonatal inoculation with the highly oncogenic C57BL-derived MCF 1233 virus predominantly contain integrated MCF proviruses. In contrast, the DNAs of more slowly developing B and non-T/non-B cell lymphomas induced by poorly oncogenic ecotropic or MCF C57BL MuLV isolates mostly contain somatically acquired ecotropic MuLV genomes. Approximately 50% of the spontaneous C57BL lymphoma DNAs contain somatically acquired MuLV genomes. None of the integrated MuLV proviruses annealed with the MCF-LTR probe, which indicates a clear difference in LTR structure with a substantial portion of the somatically acquired MuLV genomes present in the DNA of spontaneous AKR thymomas. This study stresses a dominant role of MuLV with ecotropic gp70 and LTR sequences in the development of slowly arising MuLV-induced B and non-T/non-B cell lymphomas.

The pathophysiology of murine retrovirus-induced leukemias

Murine leukemia viruses (MuLVs) are retroviruses which induce a broad spectrum of hematopoietic malignancies. In contrast to the acutely transforming retroviruses, MuLVs do not contain transduced cellular genes, or oncogenes. Nonetheless, MuLVs can cause leukemias quickly (4 to 6 weeks) and efficiently (up to 100% incidence) in susceptible strains of mice. The molecular basis of MuLV-induced leukemia is not clear. However, the contribution of individual viral genes to leukemogenesis can be assayed by creating novel viruses in vitro using recombinant DNA techniques. These genetically engineered viruses are tested in vivo for their ability to cause leukemia. Leukemogenic MuLVs possess genetic sequences which are not found in nonleukemogenic viruses. These sequences control the histologic type, incidence, and latency of disease induced by individual MuL Vs.

Retrovirus-induced spongiform encephalopathy: the 3'-end long terminal repeat-containing viral sequences influence the incidence of the disease and the specificity of the neurological syndrome

Using chimeric murine leukemia viruses (MuLVs) constructed in vitro with parental viral genomes from the neurotropic Cas-BR-E MuLV and the nonneurotropic amphotropic 4070-A MuLV, we previously mapped the paralysis-inducing determinant of Cas-BR-E MuLV within a pol-env region. To assess the role of the long terminal repeats (LTRs) in influencing the neurological disease, we constructed another chimeric MuLV (pNEMO-1)m harboring the gag-pol-env from Cas-BR-E MuLV and the LTR region from the strongly T-cell tropic Moloney MuLV. Although the Cas-BR-E MuLV induced mainly nonthymic leukemia, pNEMO-1 MuLV induced a thymic form of leukemia, as the parental Moloney MuLV. The pNEMO-1 MuLV induced neurological diseases less frequently than Cas-BR-E MuLV when inoculated intraperitoneally into NIH/Swiss, SIM.S, and SWR/J mice. However, it induced neurological disorders more frequently and with a shorter latency than Cas-BR-E MuLV when inoculated intrathymically. Most mice with a neurological disorder induced with pNEMO-1 MuLV showed a new clinical syndrome not usually seen with the parental Cas-BR-E MuLV: They had no lower limb paralysis but were excessively tremulous, spastic, and immobile. The topographical distribution of the spongiform degeneration in the brain of mice with this new syndrome was different from that seen in mice with lower limb paralysis induced by Cas-BR-E MuLV. These results indicate that the 1.0-kilobase-pair Cla I-Pvu I LTR-containing fragment harbors sequences influencing the incidence and the clinical manifestation of the neurological disease and suggest a specificity of LTR sequences for a new tissue (brain).

Molecular and biological characterization of the endogenous ecotropic provirus of BALB/c mice

We have isolated two identical molecular clones of the single, endogenous ecotropic provirus of BALB/c mice. The BALB/c clones are approximately 1/10 as infectious as an exogenous proviral clone derived from AKR mice, p623. Transfection of mouse cells with each BALB/c proviral clone yielded XC-negative, N-tropic, ecotropic virus. Cotransfection of subgenomic fragments of p623 and the BALB/c provirus did not increase infectivity to the level observed for p623; however, a 292-base-pair fragment of the p623 env gene was found to rescue XC-plaque formation. Sequence analysis showed that the XC-negative BALB/c provirus differed from the XC-positive AKR-derived provirus at a single nucleotide at the junction of the gp70 and p15E envelope proteins. Extensive sequence analysis of the BALB/c endogenous provirus showed that it differed from the sequence of the AKR-derived provirus at approximately 0.5% of 4,500 sequenced nucleotides. In addition, the BALB/c long terminal repeat contains a single copy of the enhancer-containing sequences that are repeated twice in p623. The limited variation between the ecotropic proviruses of BALB/c mice and AKR mice suggests that few cycles of reverse transcription separate these viral genomes.

Cellular DNA regions involved in the induction of rat thymic lymphomas (Mlvi-1, Mlvi-2, Mlvi-3, and c-myc) represent independent loci as determined by their chromosomal map location in the rat

The induction of thymic lymphomas by Moloney murine leukemia virus in the rat is linked to provirus integration in at least four independent cellular DNA regions (Mlvi-1, Mlvi-2, Mlvi-3, and c-myc). Because sequences homologous to at least three of these regions (Mlvi-1, Mlvi-2, and c-myc) map to chromosome 15 in the mouse, the question was raised whether they are closely linked in the rat genome and whether provirus integration in any one of these regions affects the same functional domain in rat DNA. In this study, we identified the chromosomal map location of Mlvi-1, Mlvi-2, and Mlvi-3 in the rat by using mouse-rat somatic cell hybrids that lose the rat chromosomes. The results showed that Mlvi-1 maps similarly to c-myc to chromosome 7, and Mlvi-2 maps to chromosome 2. Mlvi-3 probably maps to chromosome 15. We conclude that Mlvi-1, Mlvi-2, and Mlvi-3 are separate and independent genetic loci. Although Mlvi-1 and c-myc map to the same chromosome, they are not related, as determined by hybridization and restriction endonuclease mapping. The chromosomal map location of Mlvi-1 to chromosome 7 and Mlvi-2 to chromosome 2 is interesting, since chromosomal aberrations involving these two chromosomes are reproducibly observed in rat neoplasias induced by a variety of agents.

Concerted DNA rearrangements in Moloney murine leukemia virus-induced thymomas: a potential synergistic relationship in oncogenesis

Rat thymic lymphomas induced by Moloney murine leukemia virus carry DNA rearrangements due to provirus integration in at least five independent cellular DNA domains (Mlvi-1, Mlvi-2, Mlvi-3, RMoInt-1, and c-myc). We had previously shown that rearrangements in more than one of these domains could occur in the same tumor. In this report we extend these findings by showing that, with one exception, tumors containing provirus insertions in Mlvi-1 always contained provirus insertions in a second locus, Mlvi-2. To determine whether both events occurred in the same population of tumor cells, we examined the clonal nature of these tumors by taking advantage of allelic polymorphisms that occur naturally in both Mlvi-1 and Mlvi-2. Tumors with provirus insertions in both Mlvi-1 and Mlvi-2 arising in rats heterozygous at one of these loci were identified. DNA from these tumors was analyzed by restriction endonuclease digestion and hybridization to DNA probes derived from both Mlvi-1 and Mlvi-2. Thus, we determined the clonal nature of three thymomas and showed that in these tumors both insertion events occurred in the same population of tumor cells. The concomitant appearance of provirus insertions in Mlvi-1 and Mlvi-2 suggests a synergism of these two events that may be important in tumor induction and progression.

Murine leukemia virus long terminal repeat sequences can enhance gene activity in a cell-type-specific manner

We tested the ability of sequences in the long terminal repeat (LTR) of a mink cell focus-forming (MCF) murine leukemia virus to function as an enhancer in a cell-type-specific manner. In a stable transformation assay, the MCF or Akv LTR and the simian virus 40 enhancer had similar activities in murine fibroblasts. In contrast, the MCF LTR had a significantly greater activity in murine T lymphoid cells than did either the simian virus 40 enhancer or the Akv LTR.

Efficient insertion of genes into the mouse germ line via retroviral vectors

We present a general strategy for the efficient insertion of recombinant retroviral vector DNA into the mouse germ line via infection of preimplantation mouse embryos. Transgenic mice were generated that harbor a replication-competent recombinant retrovirus (delta Mo + Py M-MuLV) that lacks the Moloney murine leukemia virus (M-MuLV)-type enhancer sequence in the long terminal repeat (LTR). Instead, the LTR contains an enhancer element that permits polyoma virus F101 to grow in undifferentiated F9 embryonal carcinoma cells. Expression studies in different tissues of animals transgenic for delta Mo + Py M-MuLV indicate possibilities to target and modulate expression of retroviral recombinants in mice via their LTR enhancer sequences. In addition, 16 transgenic mice were generated that harbor proviral DNA of a defective recombinant retrovirus carrying a mutant dihydrofolate reductase gene.

Long terminal repeat sequences impart hematopoietic transformation properties to the myeloproliferative sarcoma virus

The myeloproliferative sarcoma virus not only transforms fibroblasts but also causes extensive expansion of the hematopoietic stem cell compartment on infection of adult mice. Similar to the Moloney sarcoma virus, it carries the mos oncogene. Moloney sarcoma virus, however, does not induce myeloproliferation and leukemia in adult mice. The difference between the two viruses was explored by using their molecularly cloned genomes and the cellular mos oncogene to construct recombinant genomes. It was shown that the U3 region of the viral long terminal repeat (LTR) has a decisive function in determining the target cell specificity of the myeloproliferative sarcoma virus. Any mos gene, whether of cellular or viral origin, is sufficient in conjunction with the proper LTR to induce myeloproliferation. Our results indicate that the pathogenicity of acutely transforming viruses is determined not only by the oncogene but also by sequences in the viral LTR.

Tissue selectivity of murine leukemia virus infection is determined by long terminal repeat sequences

Here we show that the tissue specificity of murine retrovirus infections is determined by the long terminal repeat (LTR) of an otherwise isogenic set of viruses. The isogenic viruses used for this study contain the coding gag, pol, and env genes of the avirulent Akv virus. Recombinant viruses that contain the LTR of a virus that induces T-cell leukemia lymphoma preferentially infect T lymphocytes. Viruses that carry the LTR of a virus that induces erythroleukemia preferentially infect non-T lymphoblastoid cell lines in the marrow and spleen. The Akv virus itself displays no tissue preference for hematopoietic cells. These experiments suggest that retroviruses that carry appropriate enhancer-promoters can be used to infect selectively specific target cells in animals.

Envelope gene and long terminal repeat determine the different biological properties of Rauscher, Friend, and Moloney mink cell focus-inducing viruses

The nucleotide sequence of the envelope (env) gene and the long terminal repeat (LTR) of an infectious clone of Rauscher mink cell focus-inducing (R-MCF) virus has been determined and compared with the published env gene and LTR sequences of Friend (F)- and Moloney (M)-MCF viruses. The sequence shows that R-MCF virus, like other MCF viruses, is a recombinant virus. Its env gene contains sequences which were acquired from an env gene in the mouse genome and which confer on the MCF virus its dualtropic host range. Unlike F-MCF and M-MCF viruses, R-MCF virus will not replicate in NIH 3T3 cells. The deduced amino acid sequence for the gp70 of R-MCF differs from that of F- and M-MCF viruses by 15 amino acids between residues 49 and 138 of gp70. These differences in amino acid sequences may be responsible for the inability of R-MCF virus to replicate in NIH 3T3 cells. The host range of two hybrid viruses constructed in vitro is consistent with this hypothesis. R-MCF virus and Friend murine leukemia virus (F-MLV) show 98% identity in their env gene 3' from the acquired env sequences. This contrasts with 82% identity between the env gene of R-MCF virus and M-MLV. The LTR of R-MCF shows 98% identity with the LTR of F-MCF as compared to 88% identity with the LTR of M-MCF. This striking similarity between the sequences of R-MCF, F-MCF, and F-MLV is surprising since the Rauscher virus and the Friend virus are thought to have originated independently. The high degree of similarity suggests that Rauscher and Friend viruses have a common origin. In contrast to M-MLV, which induces predominantly a lymphoid disease, R- and F-MCF viruses induce an erythroproliferative disease in NIH Swiss mice. A hybrid R-MCF virus with a genome derived primarily from R-MCF virus and a 3' end including the U3 region derived from M-MLV induces a lymphoid disease instead of an erythroid disease. This result indicates that it is the U3 region which determines the tissue specificity of the MCF virus-induced disease. It is suggested that the putative viral enhancers in the U3 region play two roles in the process of leukemogenesis: in the Friend and Rauscher disease, the viral enhancers act by increasing the transcription of the MCF env gene; in the thymic lymphoma, the enhancers activate mainly the expression of cellular genes.

Proviral genome of radiation leukemia virus: molecular cloning of biologically active proviral DNA and nucleotide sequence of its long terminal repeat

The proviral genome of a leukemogenic and thymotropic C57BL/Ka mouse retrovirus, RadLV/VL3(T+L+), was cloned as a biologically active PstI insert in the bacterial plasmid pBR322. Its restriction map was compared with those, already known, of two nonthymotropic and nonleukemogenic viruses of the same mouse strain: the ecotropic BL/Ka(B) virus and the xenotropic constituent of the radiation leukemia virus complex. Differences were observed around the gag-pol gene junction, in the pol gene, and in the env gene. Moreover, the nucleotide sequence of the RadLV/VL3(T+L+) long terminal repeat revealed the existence of two copies of a 43-base-pair sequence, of which BL/Ka(B) possesses only one copy.

A viral enhancer element specifically active in human haematopoietic cells

One particular class of DNA regulatory elements, the enhancers or activators, can, relatively independently of distance and orientation, dramatically increase the transcriptional activity of homologous and heterologous promoters located in cis (see refs 1-3 for reviews, also refs 4-6). Sequence differences between various heterologous enhancers may explain their apparent host- and/or tissue-specific action. Furthermore, differences in the transcriptional control elements may contribute to viral tropism. At least for murine leukaemia virus isolates, thymotropism and leukaemogenicity have been attributed to alterations within the viral long terminal repeat, which harbours their enhancers and other transcriptional control elements. We report here the identification of a viral enhancer element possessing a very restricted tissue range. The enhancer is active in all human cells of the haematopoetic system tested, but not in cells of fibroblast or epithelial origin.

Genetic mapping of a cellular DNA region involved in induction of thymic lymphomas (Mlvi-1) to mouse chromosome 15

Mlvi-1 defines a genetic locus representing a common domain for proviral DNA integration in Moloney murine leukemia virus-induced rat thymic lymphomas. Cellular sequences homologous to Mlvi-1 are present in mouse DNA, and we have used hamster-mouse somatic cell hybrids to chromosomally map Mlvi-1 in the mouse genome. Results demonstrated that Mlvi-1 maps to mouse chromosome 15 and that it is distinct from the Mlvi-2 integration region and from the cellular oncogenes c-myc and c-sis, which also map to this chromosome. Therefore, Mlvi-1 may contain novel sequences involved in the establishment and maintenance of virus-induced murine tumors, many of which contain abnormalities of chromosome 15.

Rearrangements in the long terminal repeat of extra mouse mammary tumor proviruses in T-cell leukemias of mouse strain GR result in a novel enhancer-like structure

Male GR mice develop T-cell leukemia at low frequency late in life. These leukemia cells invariably contain large amounts of mouse mammary tumor virus (MMTV) RNA and MMTV proteins and have extra MMTV proviruses integrated in their DNA. We show here that the extra MMTV proviruses are all derived from the endogenous MMTV provirus associated with the Mtv-2 locus and that the T-cell leukemias are clonal with respect to the acquired MMTV proviruses. The extra MMTV proviruses in six transplantable T-cell leukemia lines studied had rearranged, shortened long terminal repeats (LTRs); each T-cell leukemia, however, had a different LTR rearrangement within its extra MMTV provirus. The alteration within the extra LTRs of T-cell leukemia line 42 involved deletion of 453 nucleotides and generation of a tandem repeat region consisting of regions flanking the deletion. This alteration generated a sequence similar to the adenovirus enhancer core sequence. The viral RNAs in the T-cell leukemias contained corresponding alterations in their U3 regions. These results demonstrate that expression of MMTV in T-cell leukemias of GR mice may be the consequence of the generation of a novel enhancer, which could also stimulate expression of any adjacent cellular oncogene.

Generation of a recombinant Moloney murine leukemia virus carrying the v-src gene of avian sarcoma virus: transformation in vitro and pathogenesis in vivo

A Moloney murine leukemia virus (M-MuLV) recombinant carrying the v-src gene of avian sarcoma virus was generated by the introduction of a cloned portion of v-src from Schmidt-Ruppin A avian sarcoma virus into a molecular clone of M-MuLV provirus at the recombinant DNA level. The v-src sequences (lacking a portion of the 5' end of v-src) were inserted into the p30 region of the M-MulV gag gene so that M-MuLV gag and v-src were in the same reading frame. Transfection of this chimeric clone, pMLV(src), into NIH 3T3 cells which were constitutively producing M-MuLV gag and pol protein resulted in the formation of foci of transformed cells. Infectious and transforming virus could be recovered from the transformed cells. This virus was designated M-MuLV(src). M-MuLV(src)-transformed cells contained two novel proteins of 78 and 90 kilodaltons. The 78-kilodalton protein, p78gag-src, contained both gag and src determinants, exhibited kinase activity in an immune kinase assay, and is probably a fusion of Pr65gag and src. The 90-kilodalton protein, which is of the appropriate size to be the gPr80gag fused to src, contained gag determinants as well as a V8 protease cleavage fragment typical of the carboxy terminus of avian sarcoma virus pp60src. However, it could not be immunoprecipitated with an anti-v-src serum. M-MuLV(src)-transformed cells showed elevated levels of intracellular phosphotyrosine in proteins, although the elevation was intermediate compared with cells transformed with wild-type v-src. M-MuLV and amphotropic murine leukemia virus pseudotypes of M-MuLV(src) were inoculated into newborn NIH Swiss mice. Inoculated mice developed solid tumors at the site of inoculation after 3 to 6 weeks, with most animals dying by 14 weeks. Histopathological analysis indicated that the solid tumors were mesenchymally derived fibrosarcomas that were both invasive and metastatic.

Mutational analyses of the Moloney murine sarcoma virus enhancer

Retroviral enhancers seem to play an integral role in determining tissue tropism. In an attempt to define the sequences essential for Moloney murine sarcoma virus (Mo-MSV) enhancer activity, we have undertaken deletion and insertion mutational analyses of the 72-bp repeats. We demonstrate that, in contrast to simian virus 40 (SV40), the proximal 19 bp of the 72-bp unit of Mo-MSV are not essential for enhancer function in mouse fibroblasts. Our analysis further localizes the 5' boundary of the Mo-MSV enhancer to a 5'-CAGGAT-3' region.

Suppression of leukaemia virus pathogenicity by polyoma virus enhancers

The long terminal repeats (LTRs) of retroviruses contain sequences necessary for the initiation and termination of retroviral transcription. These sequences include promoter elements, transcriptional termination signals and transcriptional enhancer elements. The enhancer elements of Moloney murine leukaemia virus (M-MuLV) are localized in a tandemly repeated region (approximately 75 base pairs (bp) long), which lies 5' to the CAT and TATA promoter elements in the U3 region of the LTR (see Fig. 1). We have shown that the tandem repeats are required both for LTR promoter activity, as measured by transient expression assays, and for biological activity, as measured by production of infectious virus. Furthermore, they can be replaced by transcriptional enhancers from the F101 host-range mutant of polyoma virus without loss of function. We report here that the addition of the polyoma (PyF101) enhancers to the M-MuLV LTRs (either with or without the M-MuLV tandem repeats) results in complete loss of viral leukaemogenicity, even though the virus can replicate to high titres in tissue culture fibroblasts and can establish infection in animals.

Genetic mapping of a cellular DNA region involved in induction of thymic lymphomas (Mlvi-1) to mouse chromosome 15

Mlvi-1 defines a genetic locus representing a common domain for proviral DNA integration in Moloney murine leukemia virus-induced rat thymic lymphomas. Cellular sequences homologous to Mlvi-1 are present in mouse DNA, and we have used hamster-mouse somatic cell hybrids to chromosomally map Mlvi-1 in the mouse genome. Results demonstrated that Mlvi-1 maps to mouse chromosome 15 and that it is distinct from the Mlvi-2 integration region and from the cellular oncogenes c-myc and c-sis, which also map to this chromosome. Therefore, Mlvi-1 may contain novel sequences involved in the establishment and maintenance of virus-induced murine tumors, many of which contain abnormalities of chromosome 15.

Rearrangements in the long terminal repeat of extra mouse mammary tumor proviruses in T-cell leukemias of mouse strain GR result in a novel enhancer-like structure

Male GR mice develop T-cell leukemia at low frequency late in life. These leukemia cells invariably contain large amounts of mouse mammary tumor virus (MMTV) RNA and MMTV proteins and have extra MMTV proviruses integrated in their DNA. We show here that the extra MMTV proviruses are all derived from the endogenous MMTV provirus associated with the Mtv-2 locus and that the T-cell leukemias are clonal with respect to the acquired MMTV proviruses. The extra MMTV proviruses in six transplantable T-cell leukemia lines studied had rearranged, shortened long terminal repeats (LTRs); each T-cell leukemia, however, had a different LTR rearrangement within its extra MMTV provirus. The alteration within the extra LTRs of T-cell leukemia line 42 involved deletion of 453 nucleotides and generation of a tandem repeat region consisting of regions flanking the deletion. This alteration generated a sequence similar to the adenovirus enhancer core sequence. The viral RNAs in the T-cell leukemias contained corresponding alterations in their U3 regions. These results demonstrate that expression of MMTV in T-cell leukemias of GR mice may be the consequence of the generation of a novel enhancer, which could also stimulate expression of any adjacent cellular oncogene.

Characterization of long terminal repeat sequences of HTLV-III

The nucleotide sequence of the long terminal repeat sequence (LTR) of the human T-cell leukemia (lymphotropic) virus type III (HTLV-III) was determined. This virus is associated etiologically with the acquired immune deficiency syndrome. The LTR was found to be 634 base pairs in length with U3, R, and U5 regions of 453, 98, and 83 bp, respectively. The proviral DNA is flanked by a 7-base-pair direct repeat. The promoter and polyadenylation signals are situated 27 and 24 base pairs upstream from the respective transcriptional initiation and polyadenylation sites. The primer binding site is complementary to transfer RNA-lysine. The LTR of HTLV-III, like that of HTLV-I, showed a limited homology to enhancer-like sequences within two genes expressed specifically in T lymphocytes, T-cell growth factor, and gamma-interferon. Structural comparisons revealed that the LTR of HTLV-III is distantly related to those of HTLV-I, HTLV-II, and bovine leukemia virus.

At least four viral genes contribute to the leukemogenicity of murine retrovirus MCF 247 in AKR mice

Nucleotide sequences encoding gp70, Prp15E, and the U3 region of the long terminal repeat (LTR) distinguish mink cell focus-forming (MCF) retroviruses that can induce leukemia in AKR mice from closely related MCF and ecotropic murine retroviruses that are nonleukemogenic in all inbred mouse strains tested (Lung et al., Cold Spring Harbor Symp. Quant. Biol. 44:1269-1274, 1979; Lung et al., J. Virol. 45:275-290, 1983). We used a set of recombinants constructed in vitro from molecular clones of leukemogenic MCF 247 and nonleukemogenic ecotropic Akv to separate and thereby directly test the role of these genetic elements in disease induction. Leukemogenicity tests of recombinants in AKR mice show that introduction of fragments containing either an MCF LTR or MCF gp70 coding sequences can confer only a very low incidence of disease induction on Akv virus, whereas an MCF type Prp15E alone is completely ineffective. Recombinants with an MCF 247 LTR in combination with MCF Prp15E are moderately oncogenic, whereas those with an MCF 247 LTR plus MCF gp70 coding segment are quite highly leukemogenic. Mice infected with the latter virus show a substantial increase in latent period of disease induction relative to MCF 247; this delay can be reduced when Prp15E, and hence the entire 3' half of the genome, is from MCF 247. Surprisingly, sequences in the 5' half of the genome can also contribute to disease induction. We found a good correlation between oncogenicity and recovery of MCF viruses from thymocytes of injected mice, with early recovery and high titers of MCF in the thymus being correlated with high oncogenicity. This correlation held for recombinants with either an MCF or ecotropic type gp70. Together, these results (i) demonstrate that at least four genes contribute to the oncogenicity of MCF viruses in AKR mice and (ii) suggest that recombinants with only some of the necessary MCF type genes induce leukemia because they recombine to generate complete MCF genomes. Although neither Akv nor MCF 247 is leukemogenic in NFS mice, recombinant viruses whose gp70 gene was derived from Akv but whose LTRs were derived from MCF 247 induced a low incidence of leukemia in this mouse strain.

Construction of recombinants between molecular clones of murine retrovirus MCF 247 and Akv: determinant of an in vitro host range property that maps in the long terminal repeat

The leukemogenic mink cell focus-forming (MCF) retroviruses such as MCF 247 have biological properties distinct from those of their ecotropic progenitors. Nucleotide sequences encoding portions of gp70, Prp15E, and the long terminal repeat differ between the two types of viruses. To investigate the role of each of these genetic elements in determining the biological properties of MCF viruses, we prepared infectious molecular clones of MCF 247 and generated a set of recombinants between these clones and a molecular clone of Akv, the ecotropic parent of MCF 247. Each molecular clone of MCF 247 was distinct. All the recombinants between Akv and MCF 247 yielded infectious virus upon transfection. Most interestingly, recombinants which contain the long terminal repeat of MCF 247 were found to have an in vitro host range property that has been correlated with high oncogenic activity and thymotropism of certain MCF isolates; namely, they plated with higher efficiency on SC-1 cells than on NFS mouse embryo cells. Nononcogenic MCF isolates showed a slight preference for NFS cells, whereas Akv virus plated with approximately equal efficiency on the two cell types.

JC virus enhancer-promoter active in human brain cells

A human papovavirus, JCV, is the etiologic agent of the fatal demyelinating disease, progressive multifocal leukoencephalopathy. The JCV 98-base-pair tandem repeats, located to the late side of the viral replication origin, were shown to be a transcriptional regulatory element with enhancer-like activity in human fetal glial cells. These tandem repeats share significant homology with the 82-nucleotide rat brain-specific identifier RNA sequence.

The tandem direct repeats within the long terminal repeat of murine leukemia viruses are the primary determinant of their leukemogenic potential

To map the viral sequences encoding the leukemogenic determinant(s) of nondefective murine leukemia viruses (MuLVs), we constructed chimeric viral genomes in vitro between cloned viral DNAs from the highly leukemogenic Gross passage A (Gross A) MuLV and from the related nonleukemogenic BALB/c N-tropic MuLV. Infectious chimeric MuLVs, recovered from murine cells microinjected with these DNAs, were inoculated into newborn mice to test the leukemogenic potential of these viruses. We found that the U3 long terminal repeat region from Gross A genomes was sufficient to confer an intermediate leukemogenic potential to chimeric MuLVs. Sequencing data indicated that the U3 tandem direct repeat was responsible for this effect. Adding most of the Gross A p15E-coding sequences to the Gross A U3 long terminal repeat enhanced the leukemogenic potential of chimeric viruses significantly. Adding a larger 3'-end env region (all p15E-coding sequences and 345 base pairs of the carboxy terminus of gp70) to the Gross A U3 long terminal repeat restored the full leukemogenic potential of Gross A MuLV. Chimeric viruses harboring only the Gross A 3'-end env region were, however, nonleukemogenic. Similar chimeric MuLVs, constructed with genomes from the parental weakly leukemogenic BALB/c B-tropic MuLVs and nonleukemogenic BALB/c N-tropic MuLVs, were also studied. Our data indicate that the U3 tandem direct repeat sequences appear to be necessary and sufficient to confer some leukemogenic potential to MuLV. However, env 3'-end sequences, mostly the p15E-encoding sequences, are required for the expression of fully leukemic phenotypes.

Tissue-specific transcription preference as a determinant of cell tropism and leukaemogenic potential of murine retroviruses

Inoculation of susceptible strains of mice with the SL3-3 strain of murine leukaemia viruses induces T-cell lymphomas, whereas injection of the Akv strain does not. Recombinant viruses that contain the long terminal repeat (LTR) of the SL3-3 virus and the gag, pol and env genes of the Akv virus are also leukaemogenic. The cell-type specificity of leukaemias induced by viruses containing different LTR sequences is due in part to the ability of the virus to replicate in the appropriate cellular environment. One explanation of the role of the LTR in determination of both cell tropism and leukaemogenic potential is that the LTR encodes tissue-permissive transcriptional elements. We report here that there are differences in the transcriptional activity of the SL3-3 and Akv LTR sequences in different murine cell types, and that the sequences present in the LTR of SL3-3 exhibit significantly enhanced transcriptional activity in T cells compared with the corresponding region of the Akv LTR. The results suggest that transcriptional elements are primary determinants of cell tropism and of leukaemogenicity of these viruses.

Mapping the viral sequences conferring leukemogenicity and disease specificity in Moloney and amphotropic murine leukemia viruses

The Moloney murine leukemia virus (MuLV) is a highly leukemogenic virus. To map the leukemogenic potential of Moloney MuLV, we constructed chimeric viral DNA genomes in vitro between parental cloned infectious viral DNA from Moloney and amphotropic 4070-A MuLVs. Infectious chimeric MuLVs were recovered by microinjection of recombinant DNA into NIH/3T3 cells and tested for their leukemogenic potential by inoculation into NIH/Swiss newborn mice. Parental Moloney MuLV and amphotropic 4070-A MuLV induced thymic and nonthymic leukemia, respectively, when inoculated intrathymically. With chimeric MuLVs, we found that the primary determinant of leukemogenicity of Moloney and amphotropic MuLVs lies within the 1.5-kilobase-pair ClaI-PvuI long terminal repeat (LTR)-containing fragment. The presence of additional Moloney env-pol sequences with the Moloney LTR enhanced the leukemogenic potential of a chimeric MuLV significantly, indicating that these sequences were also involved in tumor development. Since parental viruses induced different forms of leukemia, we could also map the viral sequences conferring this disease specificity. We found that the 1.5-kilobase-pair ClaI-PvuI LTR-containing fragment of Moloney MuLV was necessary and sufficient for a chimeric MuLV to induce thymic leukemia. Similarly, the same LTR-containing fragment of amphotropic MuLV was necessary and sufficient for a chimeric MuLV to induce nonthymic leukemia. Therefore, our results suggest that specific sequences within this short LTR-containing fragment determine two important viral functions: the ability to transform cells in vivo (leukemic transformation) and the selection of a specific population of cells to be transformed (disease specificity).

Physical mapping of the paralysis-inducing determinant of a wild mouse ecotropic neurotropic retrovirus

We have recently shown that a molecularly cloned ecotropic retrovirus, initially isolated from the brain of a paralyzed wild mouse, retained the ability to induce hind limb paralysis when inoculated into susceptible mice (Jolicoeur et al., J. Virol. 45:1159-1163, 1983). To map the viral DNA sequences encoding the determinant of paralysis, we constructed chimeric viral DNA genomes in vitro between parental cloned infectious viral DNA genomes from this neurotropic murine leukemia virus (MuLV) and from nonneurotropic amphotropic 4070-A MuLV. Infectious chimeric MuLVs, recovered after microinjection of NIH 3T3 cells with these recombinant DNAs, were inoculated into newborn SIM.S and SWR/J mice to test the paralysis-inducing potential. We found that the 3.9-kilobase-pair SalI-ClaI fragment of the neurotropic MuLV comprising the 3' end of pol and all env sequences was sufficient to confer the paralysis-inducing potential to chimeric viruses. Therefore, this region of the neurotropic MuLV genome most likely harbors the primary determinant of paralysis.

Evidence that a major class of mouse endogenous long terminal repeats (LTRs) resulted from recombination between exogenous retroviral LTRs and similar LTR-like elements (LTR-IS)

Two endogenous retroviral long terminal repeats (LTRs) were sequenced and compared to LTR-IS (a family of insertion-element-like sequences with structural features of solitary retroviral LTRs) and to Moloney murine leukemia virus DNA. The sequence comparisons revealed that the major difference between these two endogenous LTRs is a 190-base-pair segment which is also present in LTR-IS elements. Hybridization analysis of DNAs from several mouse species using specific probes shows linkage of the 190-base-pair segment to a LTR-IS specific fragment. It is concluded that the major class of endogenous LTRs has been generated by recombination between exogenous retroviral LTRs and LTR-IS sequences.

A 3' end fragment encompassing the transcriptional enhancers of nondefective Friend virus confers erythroleukemogenicity on Moloney leukemia virus

Nondefective Friend helper murine leukemia virus (Fr-MuLV) induces primarily erythroleukemias in NFS mice, whereas Moloney murine leukemia virus (Mo-MuLV) induces T cell lymphomas. Using molecular clones of these two viruses, we constructed a recombinant in which a 0.62-kilobase fragment encompassing the U3 region at the 3' end of the Fr-MuLV genome replaced the corresponding region of Mo-MuLV. The recombinant virus obtained by transfection of this clone, whose genome is derived primarily from Mo-MuLV, induces almost exclusively erythroleukemias in NFS mice. This and the previous result of Chatis et al. (Proc. Natl. Acad. Sci. U.S.A. 80:4408-4411), showing that the reciprocal recombinant whose genome is primarily derived from Fr-MuLV induces almost exclusively lymphomas, argue that a strong determinant of the distinct disease specificities of Fr-MuLV and Mo-MuLV lies in this 3' end 0.62-kilobase fragment which contains the putative virus enhancers. To more precisely define this determinant, we have begun to construct recombinants in which smaller 3' end fragments of the Fr-MuLV and Mo-MuLV genomes are exchanged. Analysis of the first such recombinant showed that Fr-MuLV can be converted to a lymphoma-inducing virus in NFS mice by substitution of a 0.38-kilobase fragment encompassing the virus enhancers in U3 with the corresponding region of the Mo-MuLV genome.

Spontaneous and induced leukemias of myeloid origin in recombinant inbred BXH mice

BXH-2 recombinant inbred (RI) mice produce high titers of B-ecotropic murine leukemia virus beginning early in life and have a high incidence of non-T-cell leukemias that occur before 1 year of age. The leukemias that develop are in some cases associated with hind limb paralysis. In addition, a dualtropic mink cell focus-forming virus has been isolated from leukemic cells of BXH-2 mice. Immunological and cytochemical characterization of the BXH-2 leukemias showed that they are of the myeloid lineage. To assess the oncogenicity of the BXH-2 viruses, newborn mice of several BXH RI strains were inoculated at birth with biologically cloned B-ecotropic or mink cell focus-forming murine leukemia virus. These studies demonstrated that the B-ecotropic virus can induce myeloid leukemias in other BXH RI strains, whereas the dualtropic mink cell focus-forming isolates were nononcogenic in the strains tested. DNA-DNA reassociation analysis indicated that the organotropism of the B-ecotropic murine leukemia virus is confined to lymphoid tissues. Southern analysis of tumor DNAs showed that there was amplification of ecotropic virus-specific sequences in BXH-2 myeloid tumors and in all leukemias induced in other BXH RI strains by inoculation of the BXH-2 B-ecotropic virus. Although B-ecotropic virus is expressed in central nervous tissues of paralyzed BXH-2 mice, we were unable to induce the disorder in several BXH RI strains inoculated intracranially at birth with either the B-ecotropic or dualtropic virus. These results suggest that the paralysis that occurs in BXH-2 mice is due to the infiltration of leukemic cells into the central nervous system.

Genetic variation among lentiviruses: homology between visna virus and caprine arthritis-encephalitis virus is confined to the 5' gag-pol region and a small portion of the env gene

Visna virus of sheep and arthritis-encephalitis virus of goats are serologically related but genetically distinct retroviruses which cause slowly progressive diseases in their natural hosts. To localize homologous regions of the DNAs of these two viruses, we constructed a physical map of caprine arthritis-encephalitis virus DNA and aligned it with the viral RNA. Cloned probes of visna virus DNA were then used to localize regions of homology with the caprine arthritis-encephalitis virus DNA. These studies showed homology in the 5' region of the genome encompassing U5 and the gag and pol genes and also in a small region in the env gene. These findings correlate with biological data suggesting that the regions of the DNA which are homologous may be responsible for virus group characteristics such as the closely related virus core antigens. Regions which did not show homology such as large sections in the env gene may represent unique sequences which control highly strain-specific characteristics such as the neutralization antigen and specific cell tropisms.

The envelope gene and long terminal repeat sequences contribute to the pathogenic phenotype of helper-independent Friend viruses

Friend murine leukemia virus (F-MuLV) and Friend mink cell focus-inducing virus (Fr-MCF) are helper-independent murine retroviruses which induce a rapidly fatal erytholeukemia in NIH Swiss mice. Amphotropic clone 4070 (Ampho) is a murine retrovirus which does not cause leukemia in these animals. Mice inoculated with Ampho, an Fr-MCF/Ampho pseudotype, or F-MuLV developed leukemia in 0, 50, and 100% of animals, respectively. To identify the F-MuLV and Fr-MCF sequences responsible for leukemia, we constructed hybrid viral genomes between these viruses and Ampho, using subgenomic fragments of molecularly cloned viral DNA. Transfection of these hybrid viral DNAs into fibroblasts produces recombinant retroviruses. These new viruses are assayed in vivo for their ability to cause leukemia. Recombinant viruses constructed between the Ampho genome and the Fr-MCF envelope gene do not cause leukemia. Similarly, viruses constructed by using either the Fr-MCF long terminal repeat U3 region or the F-MuLV long terminal repeat U3 region and the remainder of the Ampho genome do not cause leukemia. However, if the Fr-MCF envelope gene plus the Fr-MCF U3 region are joined to Ampho, the resulting virus causes erythroleukemia in 14% of mice. Recombinant viruses made between the Fr-MCF envelope gene, the F-MuLV U3 region, and the remainder of the Ampho genome cause erythroleukemia in 38% of mice. This study demonstrates that both the envelope gene of Fr-MCF and the U3 regions of Fr-MCF and F-MuLV contain sequences which contribute to the leukemic phenotype of helper-independent Friend viruses.

Multiple enhancer domains in the 3' terminus of the Prague strain of Rous sarcoma virus

The 3' terminal region of the Prague strain of Rous sarcoma virus (PrRSV) contains at least three distinct domains that comprise two functional enhancer elements. Two of these domains (designated B and C) are found in the U3 region of the 3' long terminal repeat (LTR) while the third (designated A) is located in the sequences immediately preceding the LTR termed XSR sequences. Combinations of adjacent domains [e.g., (A + B or B + C)] are capable of activating the expression of the SV40 early promoter (21 bp repeats and TATA box) coupled to coding sequences from the prokaryotic gene chloramphenicol acetyltransferase (CAT) while a single domain is inactive. Furthermore, duplication or triplication of the central domain B restores activity. The related, Schmidt-Ruppin, strain of RSV, contains an almost identical 3' LTR element, but differs in the enhancer sequences immediately preceding the 3' LTR. A model is presented in which the sequence differences may contribute to the difference in disease spectrum of transformation defective (td) variants of these viruses.

Bovine leukemia virus: unique structural features of its long terminal repeats and its evolutionary relationship to human T-cell leukemia virus

The nucleotide sequence of the long terminal repeat (LTR) of bovine leukemia virus, a unique oncogenic retrovirus of cattle, was determined. The LTR consisted of 530 base pairs (bp) with an inverted repeat of 6 bp at its 5' and 3' ends, flanked by a direct repeat of 6 bp of host cell origin. A tRNAPro binding site for minus-strand DNA synthesis followed the 5' LTR. The U3 region contained putative transcriptional promoters, "CAT" box and "TATA" box, but they had peculiar sequences (C-C-A-A-C-T and G-A-T-A-A-A-T). The U3 region also contained a potential enhancer element, whose sequence partially resembled those of other viral and cellular, especially of immunoglobulin, enhancers. The most striking structural feature of the LTR was an exceptionally long R region (228 bp), which separated a poly(A) addition signal (A-A-T-A-A-A) from a poly(A) site as far apart as 260 bp. The long R region was suggested to form a large stable hairpin structure on a nascent RNA chain, making the two transcription termination signals close together and thus ensuring normal termination of the chain. This structural feature of the bovine leukemia virus LTR was analogous to that of human T-cell leukemia virus LTR and, in fact, slight sequence homology (at most 50%) was observed between the R regions of these two retroviruses, indicating their evolutionary relationship. The unique structural feature of bovine leukemia virus and human T-cell leukemia virus LTRs may thus bear some relation to the biological features commonly shared by these retroviruses.

Repetitive structure in the long-terminal-repeat element of a type II human T-cell leukemia virus

The majority of human T-cell leukemia virus isolates (HTLV-I) are associated with clinically aggressive adult T-cell leukemia/lymphomas. By contrast, HTLV-II has been isolated from a patient with a relatively benign hairy T-cell leukemia. To characterize differences in the viral genomes that might contribute to these different pathologies, we determined the nucleotide sequence of the long terminal repeat (LTR) of a HTLV-II provirus. Comparison with the type I HTLV LTR reveals that, whereas the overall structural features are similar, the two sequences differ markedly throughout most of the length of the LTR. Despite the overall differences, the sequences of several functional regions of the two LTRs are conserved. These include the 5' boundary of U3, the RNA cap site, and the tRNAPro-binding site immediately 3' to the LTR. Another point of similarity is a 21-base sequence that is repeated four times in the U3 region of HTLV-II and three times in the U3 region of HTLV-I. This sequence has a formal analogy to, but no common sequence with, viral transcriptional enhancers. The U3 region of HTLV-II possesses a series of imperfect tandem direct repeats, 42 bases long, 21 bases long, 19 bases long, and 7 bases long. These structures differ from those of HTLV-I except for the 21-base repeat sequence. Thus, the structure of HTLV-II differs substantially from that of HTLV-I in the region that governs transcriptional initiation and tissue specificity. Such differences may account for some of the differences in clinical presentation of HTLV-associated adult T-cell leukemia/lymphomas and hairy T-cell leukemia.

Trans-acting transcriptional activation of the long terminal repeat of human T lymphotropic viruses in infected cells

The transcription initiation signals for retroviruses lie within the long terminal repeat (LTR) sequences that flank the integrated provirus. Two subtypes of human T lymphotropic virus (HTLV) are associated with different disease phenotypes. In this article it is shown that marked differences exist in the ability of LTR sequences of these subtypes to function as transcriptional elements in differentiated cell types. It is also shown that trans-acting regulatory factors present in HTLV-infected cells stimulate gene expression directed by these LTR sequences in a type-specific manner. These results have implications for understanding the diverse biological effects of HTLV infection.

Leukemia induction by a new strain of Friend mink cell focus-inducing virus: synergistic effect of Friend ecotropic murine leukemia virus

A new strain of Friend recombinant mink cell focus-inducing retrovirus, FMCF -1-E, was found to induce leukemias in NFS and IRW mice. Although the isolate was obtained from a stock of FMCF -1 ( Troxler et al., J. Exp. Med. 148:639-653, 1978), FMCF -1-E was distinguishable from FMCF -1 by oligonucleotide fingerprinting and antigenic analysis, using monoclonal antibodies. These analyses suggested that FMCF -1-E is a distinct FMCF isolate rather than a simple variant of FMCF -1. After neonatal inoculation, the latency for leukemia induction was 3 to 8 months. A similar long latency was also seen when Friend murine leukemia virus 57 was inoculated into adult (6-week-old) IRW mice. However, sequential inoculation of FMCF -1-E at birth followed by Friend murine leukemia 57 at 6 weeks of age led to a shortened latency period (2.5 to 4 months). Only neonatal inoculation of Friend murine leukemia virus 57 was able to induce a more rapid appearance of leukemia. The leukemia cell type in the majority of cases, regardless of virus inoculation protocol, was erythroid, but occasional myeloid, lymphoid, and mixed leukemias were also observed. In contrast to NFS and IRW mice, BALB/c mice were resistant to leukemia induction by FMCF -1-E and also showed some transient resistance to leukemia induction by Friend murine leukemia virus 57.