Spontaneous germ line virus infection and retroviral insertional mutagenesis in eighteen transgenic Srev lines of mice.

SWR/J-RF/J hybrid mice spontaneously acquire new germ line ecotropic proviruses at high frequency. In the studies described here, we used these hybrids to produce 18 transgenic mouse lines, each carrying a single newly acquired Srev locus (SWR/J-RF/J ecotropic proviral locus). All of the newly acquired proviruses identified in mosaic founder SWR/J-RF/J mice that could be transmitted through the germ line were also present in somatic tissues, demonstrating that viral integration occurred before the germ line was set aside from the somatic lineages. Quantitative analysis of proviral DNA copy numbers in somatic and germinal tissues of mosaic founder parents combined with structural analysis of Srev loci indicated that these proviruses are acquired after multiple rounds of somatic viral reinfection and that most of these viral integration events occurred after DNA replication in the zygote and before DNA replication in the four-cell embryo. The frequency of provirus acquisition in Srev lines that expressed the infectious ecotropic virus was similar to that in SWR.RF mice carrying Emv-16 and Emv-17, suggesting that the chromosomal integration site of the parental locus is not an important determinant for high-frequency provirus acquisition. The frequency of recessive lethal mutations induced by spontaneous viral integration was 5%, which was similar to that induced by preimplantation embryo infection. This approach represents a simple and viable strategy for inducing and studying mutations that affect mammalian development.     

The endogenous ecotropic murine retroviruses Emv-16 and Emv-17 are both capable of producing new proviral insertions in the mouse genome.

New germ line proviral insertions are acquired at a high frequency by the progeny of SWR/J-RF/J hybrid female mice that carry the endogenous ecotropic murine leukemia proviruses Emv-16 and Emv-17. The tight linkage of these RF/J strain proviral loci has prevented genetic segregation of the retroviral genomes. Hence, it is not known whether both of these proviruses are capable of giving rise to new proviral insertions. We have molecularly cloned Emv-16 and Emv-17 and have characterized them in vitro and in vivo. Restriction enzyme analysis of the recombinant clones revealed that the proviral genomes are very similar to each other and closely resemble the wild-type AKR virus. A comparison of the flanking cellular DNA suggests that the Emv-16 and Emv-17 loci did not arise by simple duplication of a viral insertion site within the RF/J genome but most likely are independent integration events. Both proviruses produce infectious virus when transfected into NIH 3T3 cells, indicating that they are nondefective retroviruses. Exogenous infection of SWR/J mice with either Emv-16 or Emv-17 leads to viremia in the host animals, and in both cases, progeny of viremic females acquire new proviral insertions. The ability of these retroviruses to generate novel retroviral integration sites in the mouse genome provides a simple method for inducing insertional mutations in mice.     

Acquisition of endogenous ecotropic MuLV can occur before the late one-cell stage in the genital tract of SWR/J-RF/J hybrid females

Endogenous ecotropic MuLV proviral loci are acquired by the progeny of some [SWR/J x (SWR/J x RJ/J)F1] N2 hybrid females obtained by two successive backcrosses of RF/J mice onto the SWR/J background. This results most likely from an infection of early embryos or oocytes by MuLV particles originating from maternal tissues. However, the time and site of infection are not yet known. Using oviductal transfers of embryos at the one-cell stage, we show here that three of 88 N3 embryos from [SWR/J x (SWR/J x RF/J)F1] N2 hybrid females transferred to virus-free foster mothers harbored new proviral integrations, whereas none of 61 SWR/J embryos transferred to [SWR/J x (SWR/J x RF/J)F1] N2 hybrid females had acquired any proviruses. These data support the infection of oocyte and/or early one-cell embryo as the initial event leading to new proviral insertions.     

Chromosomal localization of Emv-16 and Emv-17, two closely linked ecotropic proviruses of RF/J mice.

Emv-16 and Emv-17, the two closely linked ecotropic proviral loci of RF/J mice, have been mapped to chromosome 1 between leaden, ln, and the mouse engrailed homeo-box locus, En-1, by using recombinant inbred strains and conventional backcross analysis.     

Emv-13 (Akv-3): a noninducible endogenous ecotropic provirus of AKR/J mice

All AKR/J mice carry at least three endogenous ecotropic viral loci which have been designated Emv-11 (Akv-1), Emv-13 (Akv-3), and Emv-14 (Akv-4) (Jenkins et al., J. Virol. 43:26-36, 1982.) Using two independent AKR/J-derived sets of recombinant inbred mouse strains, AKXL (AKR/J x C57L/J) and AKXD (AKR/J x DBA/2J), as well as the HP/EiTy strain (an Emv-13-carrying inbred strain partially related to AKR/J mice) (Taylor et al., J. Virol. 23:106-109, 1977), we have examined the association of these endogenous viral loci with virus expression. Strains which transmit Emv-11 or Emv-14 or both were found to produce virus spontaneously, whereas strains that transmit Emv-13 alone were negative for virus expression. Restriction endonuclease digestion and hybridization with an ecotropic virus-specific hybridization probe of DNAs from strains which transmit only Emv-13 yielded enzyme cleavage patterns identical to those observed with DNAs from strains transmitting Emv-11 or Emv-14 or both. These findings indicate the absence of any gross rearrangement of Emv-13 proviral sequences. Cell cultures derived from recombinant inbred strains that carry only Emv-13 failed to express detectable infectious virus, viral proteins, or cytoplasmic ecotropic virus-specific RNA even after treatment with 5-iodo-2-deoxyuridine or 5-azacytidine, an inhibitor of DNA methylation. Our results indicate that a mechanism(s) other than methylation of Emv-13 proviral DNA is responsible for inhibition of Emv-13 expression.     

Role of recombinant ecotropic and polytropic viruses in the development of spontaneous thymic lymphomas in HRS/J mice.

The biological and genetic characteristics of murine leukemia viruses (MuLV) derived from leukemic and normal HRS/J mice were studied. T1-oligonucleotide fingerprinting and mapping of viral RNAs from unpassaged isolates revealed the presence of complex mixtures of viral genomes. MuLV that were purified by endpoint dilution were genetically heterogeneous. Thus, endogenous retroviral sequences expressed in the tissues of HRS/J mice readily recombined with one another. Furthermore, the regular recovery of recombinant ecotropic MuLV suggested reciprocal in vivo complementation of a genetic defect(s) in each of the endogenous ecotropic proviruses Emv-1 and Emv-3. Some recombinant ecotropic viruses contained sequences in the p15E-U3 region that were not derived from Emv-1 and Emv-3 but were found in recombinant polytropic HRS/J viruses. Finally, comparison of the genetic structures of leukemogenic and nonleukemogenic MuLV of this strain implied that the oncogenic phenotype of these MuLV is encoded within env or the U3 region of the genome or both. Our results are consistent with a stepwise convergent evolution of recombinant MuLV in vivo in individual HRS/J mice. Ultimately, this process of selection results in formation of leukemogenic polytropic viruses.     

Prenatal transmission and pathogenicity of endogenous ecotropic murine leukemia virus Akv.

OBJECTIVE: Mouse strains carrying endogenous ecotropic murine leukemia viruses (MuLV) are capable of expressing infective virus throughout life. Risk of transplacental transmission of MuLV raises concerns of embryo infection and induction of pathogenic effects, and postnatal MuLV infection may lead to tumorigenesis. METHODS: Endogenous ecotropic MuLV-negative SWR/J embryos were implanted into Akv-infected viremic SWR/J mice, into spontaneously provirus-expressing AKR/J mice, and into noninfected SWR/J control mice; virus integration and virus expression were investigated at 14 days' gestation. Tumor development was monitored over 18 months. RESULTS: Of 111 embryos, 20 (18%) recovered from Akv-infected SWR/J mice, which had developed normally, were infected. New proviruses were detected in 10 of 111 (9%) embryos from Akv-infected SWR/J mice, and in 2 of 60 (3%) embryos from AKR/J mice; none expressed viral protein. Of 127 embryos recovered from Akv-infected SWR/J mice, 16 (13%) were dead; 4 of 5 (80%) were infected and expressed viral protein. Of 71 embryos from AKR/J mice, 11 (15%) were dead, and 2 of 2 had virus integration; virus expression was not detected. Numbers of dead embryos recovered from experimentally infected, viremic SWR/J mice and from spontaneously endogenous MuLV-expressing AKR/J mice were significantly higher, compared with numbers from nonviremic SWR/J control mice, and embryo lethality was significantly associated with prenatal provirus expression. Postnatal inoculation of Akv induced lymphoblastic lymphomas in 15 of 24 (61%) SWR/J mice within mean +/- SD latency of 14 +/- 2.4 months. Only 3 of 39 (8%) control mice developed lymphomas (P < 0.005). CONCLUSION: Embryos in MuLV-viremic dams are readily infected, and inappropriate prenatal expression of leukemogenic endogenous retroviruses may play a critical role in embryo lethality and decreased breeding performance in ecotropic provirus-positive mouse strains.     

Chromosomal assignment of two endogenous ecotropic murine leukemia virus proviruses of the AKR/J mouse strain.

The AKR/J mouse strain is genetically fixed for three different ecotropic murine leukemia virus genomes, designated Akv-1, Akv-3, and Akv-4 (Emv-11, Emv-13, and Emv-14). With recombinant inbred strains and crosses with linkage-testing stocks, Akv-3 and Akv-4 were placed on the mouse chromosome map. Akv-3, which encodes a replication-defective provirus, maps near the agouti coat color locus, a, on chromosome 2. Akv-4, which is replication competent, maps near the neurological mutant gene locus trembler, Tr, on chromosome 11. Akv-1 and Akv-2 (Emv-12), an ecotropic provirus carried by AKR/N but not AKR/J, have previously been mapped to chromosome 7 and 16, respectively. Thus, the four Akv proviruses mapped to date are on four different chromosomes. Akv-3 is the second ecotropic murine leukemia virus provirus to be mapped near the agouti locus. The results are discussed in relation to possible nonrandomness of viral integration.     

Molecular cloning and mapping of the ecotropic leukemia provirus Emv-23 provides molecular access to the albino-deletion complex in mouse chromosome 7

Genetic analysis of radiation-induced deletion mutations involving the chromosome 7 albino (c) locus has expanded the functional map of this 6 to 11-cM region of the mouse genome. To generate one of many points of molecular access necessary for intensifying the analysis of the genes and phenotypes associated with this particular complex of deletions, we have cloned an endogenous ecotropic leukemia provirus (Emv-23), known to be closely linked to c, along with its flanking chromosome 7 sequences. A unique-sequence probe (23.3), derived from a region immediately 5' to the proviral integration site, was found to map less than 0.5 cM from c in a standard backcross analysis. Southern blot analysis of DNAs from animals carrying homozygous or overlapping albino deletions demonstrated that the 23.3 probe was deleted in several relatively small c-region deletions. The deletion mapping of the 23.3 probe places the Emv-23 locus between c and Mod-2, just proximal to a region important for male fertility and juvenile fitness. Mapping of this locus also provides a refinement of the genetic/deletion map for several mutations within this deletion complex.     

Pattern of expression of ecotropic murine leukemia virus in gonads of inoculated SWR/J mice.

An ecotropic murine leukemia virus (MuLV) isolate has recently been shown to be able to infect the germ line or the early embryo or both when inoculated at birth to SWR/J females (J. J. Panthier, H. Condamine, and F. Jacob, Proc. Natl. Acad. Sci. USA 85:1156-1160, 1988). We have used this isolate to further study this phenomenon. By using the techniques of RNA-RNA in situ hybridization, immunocytochemistry, and transmission electron microscopy, the identities of two important cell types that are infected by ecotropic MuLV in the gonads of inoculated mice were determined. These cells are the thecal cells surrounding the follicles in the ovary and the Leydig cells in the testis. Both types actively synthesize viral RNA and express a viral antigen. Furthermore, we documented the production of viral particles by the thecal cells. The expression of ecotropic MuLV by nonlymphoid cells in vivo may play a key role in the vertical transmission of these viruses by females as well as in their horizontal transmission.     

Mobility of endogenous ecotropic murine leukemia viral genomes within mouse chromosomal DNA and integration of a mink cell focus-forming virus-type recombinant provirus in the germ line

Characterization of endogenous ecotropic Akv proviruses in DNA of low and high leukemic mouse strains revealed the presence of one to six copies of the Akv genome per haploid genome equivalent integrated in the germ line. Low leukemic strains analyzed so far contained only one complete copy of the Akv proviral DNA. The site of integration varied among strains, although genetically related strains often carried the Akv proviral gene in the same chromosomal site. The different substrains of the AKR mouse displayed the presence of variable numbers (two to six) of Akv genomes. In all substrains one Akv genome was present in an identical chromosomal site; this locus probably comprised the progenitor genome. Closely related substrains had several Akv proviral DNAs integrated in common sites. The accumulation of Akv genomes in the germ line of the AKR/FuRdA strain is likely the result of independent integration events, since backcross studies with the Akv-negative 129 strain showed random segregation of all six proviral loci. The AKR/Cnb strain carried a recombinant provirus in the germ line. This provirus resembled in structure the AKR mink cell focus-forming viruses, which are generated by somatic recombination during leukemogenesis. Therefore, the germ-line amplification of Akv proviral DNAs occurs most likely through infection of embryonic cells by circulating virus.     

A nucleotide substitution in the gag N terminus of the endogenous ecotropic DBA/2 virus prevents Pr65gag myristylation and virus replication.

The endogenous ecotropic provirus Emv-3 present in DBA/2 mice is poorly expressed in the animal, as well as in cell cultures. Transfection of proviral DNA into NIH 3T3 cells localized the expression defect to the 5' region of the viral genome, spanning the untranslated region and the N-terminal part of the gag gene. Comparison of the nucleotide sequence of the Emv-3 provirus with the sequence of the highly infectious Akv murine leukemia virus revealed three nucleotide differences within the gag coding region. One of these differences was found in codon 3 of the gag polyprotein, where a Gln codon is seen in Akv and a Pro codon is differences was found in codon 3 of the gag polyprotein, where a Gln codon is seen in Akv and a Pro codon is seen in Emv-3. By site-directed mutagenesis, we showed that the defect of Emv-3 expression indeed is localized to codon 3 of the gag gene. The gag polyprotein of mammalian type C retrovirus contains myristic acid covalently linked to the N-terminal glycine. This myristylation is not seen in the Emv-3-coded gag polyprotein. We showed that the in vitro-mutagenized Emv-3 genome containing a Gln codon at position 3 of the gag gene yields a myristylated gag polyprotein. Thus, it seems most likely that the defect of expression of the Emv-3 provirus is due to the presence of a proline is position 3 of the gag polyprotein, preventing the myristylation.     

Fv-1 regulation of lymphoma development and of thymic ecotropic and xenotropic MuLV expression in mice of the AKR/J x RF/J cross.

The incidence of spontaneous thymic lymphoma has been studied in crosses between AKR/J and RF/J mice. AKR mice develop a high incidence of this disease. RF mice transmit a marked resistance to development of the disease to F1 hybrid mice of the AKR x RF cross. This resistance is associated with a reduction of endogenous ecotropic and xenotropic MuLV expression in the prelymphomatous thymus. The RF gene governing the coordinate suppression of these three phenotypes has been mapped to the Fv-1 locus. These results indicate that the particular Fv-1 allele of AKR mice provides a permissive genetic background for endogenous ecotropic and xenotropic MuLV expression and that these viral activities may be etiologically involved in the development of spontaneous thymic lymphoma in the mouse.     

Studies of the mechanism of spontaneous germline ecotropic provirus acquisition in mice.

SWR/J--RF/J hybrid mice spontaneously acquire new germline ecotropic proviruses at high frequency. We have performed ovarian transplantation and in situ hybridization studies to delineate the mechanism and developmental stage of germline provirus acquisition. In addition, we have developed a novel, efficient and simple method to introduce single copy proviruses into the mouse germline. The results reported here have direct implications for understanding how proviruses are acquired in the germline, for using murine leukemia viruses as insertional mutagens, and for using retroviral vectors to introduce foreign genes into the mouse germline.     

Genetics of dietary obesity in AKR/JxSWR/J mice: segregation of the trait and identification of a linked locus on Chromosome 4

We describe a new multiple gene mouse model of differential sensitivity to dietary obesity that provides a tool for dissecting the genetic basis for body composition and obesity. AKR/J and SWR/J male mice, as well as male progeny of intercrosses between these strains, were fed a high-fat diet for 12 weeks beginning at 5 weeks of age. Body weight and energy intake were assessed weekly. At the conclusion of the dietary manipulation, an adiposity index was calculated by dividing the weight of seven dissected adipose depots by the carcass weight. AKR/J mice had approximately sixfold greater adiposity than SWR/J mice. Examination of the segregation of the adiposity trait in the progeny of crosses between these strains indicates that the trait is determined by a minimum of one to four genetic loci and that there is significant dominance of the AKR/J genotype. A preliminary analysis with markers linked to the known mouse obesity genes ob, db, tub, and fat showed no linkage with these loci. However, a quantitative trait locus was found that maps distal to the db gene on Chromosome (Chr) 4. This locus has been designated dietary obese 1 or Do1.     

Linkage of the Fv-2 gene to a newly reinserted ecotropic retrovirus in Fv-2 congenic mice

Restriction enzyme and Southern gel analyses were used to determine the number and location of endogenous ecotropic retroviruses in the germ line of several mouse strains congenic at the Fv-2 gene locus. A new endogenous ecotropic provirus was observed in the germ line of B6.S (Fv-2ss) mice, in addition to the resident provirus found in its congenic partner C57BL/6 (Fv-2rr). This new provirus was similar in structure to the C57BL provirus. The SIM strain of mice, the donors of the Fv-2s allele in B6.S mice, does not contain ecotropic proviruses, suggesting that the new provirus in the B6.S mouse strain arose by germ-line reintegration during the construction of this strain. Mendelian segregation analysis indicated that this new provirus was linked to the Fv-2 gene locus on chromosome 9. In three other Fv-2s congenic mouse strains--B10.C (47N), B6.C (H-7b), and C57BL/6J Trfa, Bgsd--no additional ecotropic endogenous viruses were detected, suggesting that the reinsertion event that occurred during the construction of B6.S is not essential for the acquisition of the Fv-2s phenotype in the C57BL genetic background. Although numerous reports of germ-line reinsertions of ecotropic virus in high-virus mouse strains have been received, the present results provide definitive evidence that similar germ-line amplifications of endogenous ecotropic virus can occur in a low-virus mouse strain.     

Identification of ecotropic proviral sequences in high- and low-ecotropic-virus-producing mouse strains.

The arrangement of endogenous ecotropic retroviruses in selected high- and low-ecotropic-virus-producing mouse strains was examined by Southern blot hybridization analysis, using an ecotropic retrovirus-specific DNA probe. High-ecotropic-virus-producing mouse strains of the AKR family displayed heterogeneity with respect to the number of copies and the sites of insertion of endogenous ecotropic specific DNA. This diversity was seen even among individuals of the same AKR subline. Contrastingly, individuals within the same low-ecotropic-retrovirus-producing mouse strain showed no evidence of variability in their endogenous ecotropic proviral sequences. These results favored the hypothesis that germ line proviral reinsertion was responsible for the proviral sequence heterogeneity observed in high-ecotropic-virus-producing mouse strains.     

Diverse wild mouse origins of xenotropic, mink cell focus-forming, and two types of ecotropic proviral genes

We analyzed wild mouse DNAs for the number and type of proviral genes related to the env sequences of various murine leukemia viruses (MuLVs). Only Mus species closely related to laboratory mice carried these retroviral sequences, and the different subclasses of viral env genes tended to be restricted to specific taxonomic groups. Only Mus musculus molossinus carried proviral genes which cross-reacted with the inbred mouse ecotropic MuLV env gene. The ecotropic viral env sequence associated with the Fv-4 resistance gene was found in the Asian mice M. musculus molossinus and Mus musculus castaneus and in California mice from Lake Casitas (LC). Both M. musculus castaneus and LC mice carried many additional Fv-4 env-related proviruses, two of which are common to both mouse populations, which suggests that these mice share a recent common ancestry. Xenotropic and mink cell focus-forming (MCF) virus env sequences were more widely dispersed in wild mice than the ecotropic viral env genes, which suggests that nonecotropic MuLVs were integrated into the Mus germ line at an earlier date. Xenotropic MuLVs represented the major component of MuLV env-reactive genes in Asian and eastern European mice classified as M. musculus molossinus, M. musculus castaneus, and Mus musculus musculus, whereas Mus musculus domesticus from western Europe, the Mediterranean, and North America contained almost exclusively MCF virus env copies. M. musculus musculus mice from central Europe trapped near the M. musculus domesticus/M. musculus musculus hybrid zone carried multiple copies of both types of env genes. LC mice also carried both xenotropic and MCF viral env genes, which is consistent with the above conclusion that they represent natural hybrids of M. musculus domesticus and M. musculus castaneus.     

Lack of ecotropic virus involvement in induction of lymphomas in DBA/2J mice by 7,12-dimethylbenz(a)anthracene.

DBA/2 mice carry a single endogenous ecotropic murine leukemia provirus, Emv-3, that is replication defective because of a single nucleotide substitution in codon 3 of p15gag. However, when weanling DBA/2 mice are treated percutaneously with 7,12-dimethylbenz(a)anthracene (DMBA), ecotropic virus replication is induced in almost all of the treated mice. Previous studies have shown that this induction results from DMBA-induced reverse mutations in codon 3 that allow efficient virus replication. In addition to ecotropic virus replication, DMBA also induces lymphomas in 100% of the treated mice. These results have raised the possibility that ecotropic virus replication is causally associated with the development of lymphomas in DBA/2 mice, perhaps via the insertional activation or mutation of cellular proto-oncogenes. To test this possibility, we compared lymphoma incidence after percutaneous DMBA treatment in DBA/2J-dv/dv mice, which carry two copies of Emv-3, with lymphoma incidence in DBA/2J-d+18J/d+18J mice, which lost both copies of Emv-3 by homologous recombination involving the long terminal repeat sequences. The results of this study conclusively demonstrated that Emv-3 is not causally associated with the development of DMBA-induced lymphomas in DBA/2J mice. Interestingly, histopathological and molecular analyses of the lymphomas indicated that the majority of the lymphomas in both strains of mice were of the B-cell lineage. This was unanticipated, since the majority of chemically induced lymphomas in other inbred strains are thymic lymphomas, presumably of the T-cell lineage. Thus, DBA/2 mice appear to present a unique model system for the investigation of chemically induced B-cell lymphomas in mice.