Establishment of a C3Hf Mammary Tumor Cell Line Expressing Endogenous Mouse Mammary Tumor Virus: Antigenic and Genetic Relationships of This Virus with Highly Oncogenic Mouse Mammary Tumor Viruses

A single-cell clone of C3Hf mammary tumor cells (clone 14) was developed into a continuous cell line expressing high levels of endogenous mouse mammary tumor virus (MMTV) with less than 0.1% murine leukemia virus expression. Comparison of the C3Hf MMTV protein profile on sodium dodecyl sulfatepolyacrylamide gel electrophoresis with that of C3H MMTV revealed that the protein content of the two viruses was quite similar. However, oligonucleotide fingerprints obtained of MMTV 70S RNA revealed that approximately 20% of the large oligonucleotides examined were unique to each virus. The oligonucleotide fingerprint indicated that although the viruses were similar, they differed in their genetic content. The differences in the two viruses extended to immunological differences in the major envelope glycoprotein, gp52. C3Hf MMTV competed only partially in a homologous radioimmunoassay for gp52 of C3H MMTV, whereas C3H MMTV gave complete competition, indicating that gp52 of C3H MMTV contained type-specific determinants not present on gp52 of C3Hf MMTV. Comparison of C3Hf MMTV with highly oncogenic C3H, GR, and RIII MMTVs in a homologous C3H MMTV gp52 assay gave two patterns of reactivity: complete competition by GR and C3H MMTV and incomplete competition by C3Hf and RIII MMTV. Absorption of anti-C3H MMTV serum by either C3Hf MMTV or RIII MMTV removed all antibodies against both viruses but not against GR and C3H MMTVs. These results indicate that C3H and GR MMTVs are more closely related to each other than to RIII and C3Hf MMTVs.     

Integration of New Endogenous Mouse Mammary Tumor Virus Proviral DNA at Common Sites in the DNA of Mammary Tumors of C3Hf Mice and Hypomethylation of the Endogenous Mouse Mammary Tumor Virus Proviral DNA in C3Hf Mammary Tumors and Spleens

To understand the molecular mechanisms by which the endogenous murine mammary tumor virus (MuMTV) proviruses are expressed and produce late-occurring mammary tumors in C3Hf mice, we analyzed, by the use of restriction enzymes and the Southern transfer procedure, genomic DNA from normal organs of mammary tumor-bearing and tumor-free mice and from 12 late-occurring C3Hf mammary tumors. We found, by using the restriction enzymes EcoRI and HindIII, that in addition to the preexisting endogenous MuMTV proviruses, new MuMTV-specific proviral DNA was integrated into new sites in the host genome in all 12 of the tumors that we examined. PstI digests of C3Hf tumor DNA revealed that the new proviral DNA found in C3Hf tumors was of endogenous origin. Moreover, the respective sizes of at least one of the new DNA fragments generated by EcoRI or HindIII digestion were the same in at least 50% of the C3Hf tumors analyzed, suggesting that the integration site of this new proviral DNA could be at the same location in the host genome of these tumors. Our results may imply that mammary tumorigenesis in C3Hf mice results from activation of cellular oncogenes by an MuMTV proviral DNA promoter. Specific hypomethylation of MuMTV proviral DNA was detected in the mammary tumors and spleens of C3Hf tumor-bearing mice. Our results indicated that most, if not all, of the hypomethylated MuMTV proviral DNA sequences were derived from the endogenous MuMTV provirus located at the MTV-1 locus, a locus responsible for the production of MuMTV antigens and increased incidence of mammary carcinoma in C3Hf mice. In spleens of non-tumor-bearing mice of ages 3, 6, 9, and 12 months, there was progressive hypomethylation of proviral DNA with increasing age, suggesting a possible correlation between demethylation of MuMTV proviral DNA in the spleens of C3Hf mice and the expression of endogenous MuMTV.     

On Oncogenes and Tumor Suppressor Genes in the Mammary Gland

Well-known oncogenes (e.g., MYC) and tumor suppressor genes (e.g., TP53) are involved in breast cancer. But the roles of newly identified genes, microRNAs, and other noncoding RNAs in the disease have yet to be defined.Breast cancer in humans is associated with genetic alterations of a number of oncogenes (ErbB2, MYC, PIK3CA) and tumor suppressors (TP53, BRCA1/2, RB1, PTEN), as outlined by Lee and Muller. The use of genetically engineered mouse models harboring deletions or mutations in these genes has provided insight into how such alterations drive tumor initiation, progression, and metastasis, and how they influence responses to anticancer agents. Beyond these well-characterized alterations, there has been a recent explosion in new information regarding the molecular pathogenesis of breast cancer, and therefore a need to define the functional roles of newly described potential breast cancer genes. For example, whole-genome sequencing has identified a large number of genes with recurrent sequence alterations in human breast cancer specimens (Wood et al. 2007). Moreover, gene copy number analyses have identified multiple regions of chromosomal gain or loss (Chin et al. 2006). Finally, microRNAs and other types of noncoding RNA have recently emerged as important modulators of cancer phenotypes (Ventura and Jacks 2009). Despite, this new information about cancer-associated molecular alterations, the full characterization of their impact on breast cancer biology in vivo remains incomplete. Therefore, a major current challenge is to functionally annotate these emerging groups of candidate breast cancer tumor suppressors/oncogenes.The generation of additional mouse models with engineered mutations or transgenic expression of new candidate genes will provide important information validating their role in cancer and elucidating their specific biological activities. However, it is important to point out that the existing mouse models do not recapitulate the estrogen receptor (ER)-positive histological subtype of breast cancer, which is an important subset in humans. Thus, current transgenic approaches do not provide a comprehensive platform for modeling the full spectrum of the human disease. An equally important issue is that the considerable time and expense of generating new engineered alleles and breeding to obtain compound mutant strains creates a bottleneck in fully annotating the breast cancer genome. Gain- and loss-of-function experiments in cancer cell lines and in immortalized breast epithelial cells (Weaver et al. 1995) can in many cases determine the roles of novel regulators of cellular transformation. However, factors contributing to cancer pathogenesis through such processes as control of cellular senescence, reprogramming of the cellular differentiation state, and interactions with the microenvironment may not be readily studied in such systems. Although xenografts of human cancer cell lines may be helpful in some cases, the functions of other cancer-relevant factors may be best uncovered by the genetic manipulation of primary cells and their subsequent growth as orthotopic implants (Heyer et al. 2010). By using isogenic cells, it is possible to the study tumor formation of such implants in the context of wild-type mice with a fully intact immune system, thereby fully recapitulating the microenvironment of spontaneous tumors. Breast progenitor/stem cells provide a valuable system in this regard, although the selection of other types of breast epithelial cells would likely influence the ensuing tumor phenotype and have utility in studying histological subtypes of breast cancer.An additional important question in annotating the breast cancer genome is whether a genetic lesion contributes to the maintenance of established tumors—rather than just to tumor initiation—and therefore points to a pathway whose deregulation represents a potential drug target. Lee and Muller (2011) refer to work with doxycline-inducible transgenics indicating that tumors induced by ErbB2 remain dependent on this factor for tumor maintenance. The generation of additional inducible overexpression systems and the use of inducible short hairpin RNA (shRNA) in the transplantable models discussed above will be important for defining additional factors required for tumor maintenance. The molecular and cellular alterations resulting from switching off a candidate oncogene in vivo also provides important benchmarks with which to evaluate new drugs that are designed to target such pathways.Noncoding RNAs represent an emerging class of cancer regulators in need of further study in vivo. Micro RNAs (miRNAs) are the most studied group of noncoding RNAs, and individual miRNAs have been shown to function as oncogenes and tumor suppressors in mouse models (Ventura and Jacks 2009). Deregulated miRNAs in breast cancer include miR-210 (Camps et al. 2008), a hypoxia-regulated miRNA that contributes to cancer growth by regulating energy metabolism. miR-335, -126, -10b, and miR-31 have all been shown to regulate invasion and metastasis (Ma et al. 2007; Tavazoie et al. 2008; Valastyan et al. 2009), and let7b and miR-200 negatively regulate breast cancer stem cell self-renewal (Iliopoulos et al. 2009; Shimono et al. 2009). The functional validation of these and other classes of noncoding RNAs in vivo is likely to yield novel observations about the molecular circuitry of breast cancer cells.The molecular pathogenesis of metastasis is a rapidly evolving area of research, and recent efforts have helped define activators and repressors of breast cancer metastasis as well as factors that influence the site of metastatic growth. Notably, it is now apparent that metastatic lesions have marked differences in mutational profiles compared to the primary tumors. Whole-genome sequencing of paired primary and metastatic basal-cell-like breast cancer, and xenografts of the same primary tumor tissue, identified striking similarities and some interesting differences in the prevalence of genetic mutations between the samples (Ding et al. 2010). Mutations detected in the metastatic tumor were highly concordant with those detected in the tumor xenograft. However, some of these mutations were not present in the primary tumor sample. This suggests that the evolutionary pressures driving metastasis and establishment of primary xenografts may be similar, and highlight the potential usefulness of functionally assessing mutations selectively enriched in both lesions. Whether these genetic changes influence tumor cell adaptability and secondary organ site selection are key questions in ongoing efforts to understand the mechanisms governing metastatic spread. In vivo studies have also established an important causative link between metastasis and the expression of specific micro RNAs (miRNA) and identified pro- and antimetastasis regulators (Ma et al. 2007; Tavazoie et al. 2008). The mechanisms by which the expression of these miRNAs are altered and key molecular targets and cellular functions of these miRNAs require further investigation.As new technologies provide an increasingly detailed view of mutational, gene-expression, and gene copy number profiles that define breast cancer, the potential of personalized medicine in cancer therapy becomes more attainable. The full realization of this goal will first require a broader characterization of novel candidate cancer regulators. It is clear that a multipronged approach using multiple model systems will be needed to translate genetic findings into a comprehensive picture of the underlying circuitry of distinct histological and genetic subsets of breast cancer.     

Pericentriolar Targeting of the Mouse Mammary Tumor Virus GAG Protein

The Gag protein of the mouse mammary tumor virus (MMTV) is the chief determinant of subcellular targeting. Electron microscopy studies show that MMTV Gag forms capsids within the cytoplasm and assembles as immature particles with MMTV RNA and the Y box binding protein-1, required for centrosome maturation. Other betaretroviruses, such as Mason-Pfizer monkey retrovirus (M-PMV), assemble adjacent to the pericentriolar region because of a cytoplasmic targeting and retention signal in the Matrix protein. Previous studies suggest that the MMTV Matrix protein may also harbor a similar cytoplasmic targeting and retention signal. Herein, we show that a substantial fraction of MMTV Gag localizes to the pericentriolar region. This was observed in HEK293T, HeLa human cell lines and the mouse derived NMuMG mammary gland cells. Moreover, MMTV capsids were observed adjacent to centrioles when expressed from plasmids encoding either MMTV Gag alone, Gag-Pro-Pol or full-length virus. We found that the cytoplasmic targeting and retention signal in the MMTV Matrix protein was sufficient for pericentriolar targeting, whereas mutation of the glutamine to alanine at position 56 (D56/A) resulted in plasma membrane localization, similar to previous observations from mutational studies of M-PMV Gag. Furthermore, transmission electron microscopy studies showed that MMTV capsids accumulate around centrioles suggesting that, similar to M-PMV, the pericentriolar region may be a site for MMTV assembly. Together, the data imply that MMTV Gag targets the pericentriolar region as a result of the MMTV cytoplasmic targeting and retention signal, possibly aided by the Y box protein-1 required for the assembly of centrosomal microtubules.     

Expression of Mouse Mammary Tumor Virus Superantigen mRNA in the Thymus Correlates with Kinetics of Self-Reactive T-Cell Loss

Mouse mammary tumor virus (MMTV) encodes a superantigen (Sag) that is expressed at the surface of antigen-presenting cells in conjunction with major histocompatibility complex (MHC) type II molecules. The Sag-MHC complex is recognized by entire subsets of T cells, leading to cytokine release and amplification of infected B and T cells that carry milk-borne MMTV to the mammary gland. Expression of Sag proteins from endogenous MMTV proviruses carried in the mouse germ line usually results in the deletion of self-reactive T cells during negative selection in the thymus and the elimination of T cells required for infection by specific milk-borne MMTVs. However, other endogenous MMTVs are unable to eliminate Sag-reactive T cells in newborn mice and cause partial loss of reactive T cells in adults. To investigate the kinetics of Sag-reactive T-cell deletion, backcross mice that contain single or multiple MMTVs were screened by a novel PCR assay designed to distinguish among highly related MMTV strains. Mice that contained Mtv-17 alone showed slow kinetics of reactive T-cell loss that involved the CD4(+), but not the CD8(+), subset. Deletion of CD4(+) or CD8(+) T cells reactive with Mtv-17 Sag was not detected in thymocytes. Slow kinetics of peripheral T-cell deletion by Mtv-17 Sag also was accompanied by failure to detect Mtv-17 sag-specific mRNA in the thymus, despite detectable expression in other tissues, such as spleen. Together, these data suggest that Mtv-17 Sag causes peripheral, rather than intrathymic, deletion of T cells. Interestingly, the Mtv-8 provirus caused partial deletion of CD4(+)Vbeta12(+) cells in the thymus, but other T-cell subsets appeared to be deleted only in the periphery. Our data have important implications for the level of antigen expression required for elimination of self-reactive T cells. Moreover, these experiments suggest that mice expressing endogenous MMTVs that lead to slow kinetics of T-cell deletion will be susceptible to infection by milk-borne MMTVs with the same Sag specificity.     

Mouse Mammary Tumor Virus Suppresses Apoptosis of Mammary Epithelial Cells through ITAM-Mediated Signaling

Many receptors in hematopoietic cells use a common signaling pathway that relies on a highly conserved immunoreceptor tyrosine-based activation motif (ITAM), which signals through Src family tyrosine kinases. ITAM-bearing proteins are also found in many oncogenic viruses, including the mouse mammary tumor virus (MMTV) envelope (Env). We previously showed that MMTV Env expression transformed normal mammary epithelial cells and that Src kinases were important mediators in this transformation. To study how ITAM signaling affects mammary cell transformation, we utilized mammary cell lines expressing two different ITAM-containing proteins, one encoding a MMTV provirus and the other a B cell receptor fusion protein. ITAM-expressing cells were resistant to both serum starvation- and chemotherapeutic drug-induced apoptosis, whereas cells transduced with these molecules bearing ITAM mutations were indistinguishable from untransduced cells in their sensitivity to these treatments. We also found that Src kinase was activated in the MMTV-expressing cells and that MMTV-induced apoptosis resistance was completely restored by the Src inhibitor PP2. In vivo, MMTV infection delayed involution-induced apoptosis in the mouse mammary gland. Our results show that MMTV suppresses apoptosis through ITAM-mediated Src tyrosine kinase signaling. These studies could lead to the development of effective treatment of nonhematopoietic cell cancers in which ITAM-mediated signaling plays a role.     

Dexamethasone Induction of the Intracellular RNAs of Mouse Mammary Tumor Virus

We have studied the kinetics of dexamethasone induction of mouse mammary tumor virus (MMTV) RNAs and proteins in virus-infected rat XC cells and GR mouse mammary tumor cells. A detectable increase in viral RNA in infected XC cells was present within 10 min after hormone addition, and half-maximal induction was achieved in less than 2 h. The increase in viral RNA concentration was apparent first in nuclear RNA and later in the cytoplasm. Within the first 15 min of induction, only genome-sized RNA (35S, 7.8 kilobases) was present in augmented amounts, whereas the major subgenomic RNA (24S, 3.8 kilobases) did not appear until at least 30 to 60 min postinduction. The sequential appearance of these RNAs, the probable mRNA's for the gag and env proteins, paralleled the order of appearance of the gag and env proteins, respectively, after hormone treatment. An additional species of viral RNA (20S, 2.5 kilobases) was detected during these induction experiments, but the role of this RNA is not known. Both subgenomic RNAs contain sequences derived from both the 5′ and 3′ termini of genomic RNA and are presumably spliced. After dexamethasone induction of infected XC cells, we detected two smaller env-related proteins which were not found in full hormone induction. The functional role of these smaller proteins is not known. A previously reported smaller species of RNA (13S, 1.0 kilobase) did not appear to be induced and was shown to be cellular rather than viral in origin. In the fully induced infected XC and GR mammary tumor cells, the only viral RNAs present were the 35S and 24S RNAs. In addition, mammary tumors contained only these two viral RNAs. Thus, tumor cells appear to contain only the viral RNAs which direct the synthesis of the gag, pol, and env proteins of the virion.     

Electron Microscopy of the Nucleic Acid of Mouse Mammary Tumor Virus

Mouse mammary tumor virus (MTV) was isolated from the milk of RIII mice by density-gradient centrifugation in Ficoll. The homogeneity of the preparation was demonstrated in electron micrographs. The nucleic acid was extracted with phenol in the presence of Pronase. Its viral origin was attested by failure of ribo-nuclease and deoxyribonuclease treatment of the virus preparation to destroy the filamentous molecules; after phenol extraction, the molecules were destroyed by ribonuclease but not by deoxyribonuclease. Rotary shadowed preparations were examined in the electron microscope. The length distribution of the RNA filaments showed peaks at 1.2, 2.4, and 3.6 mum. The molecular weight of the longest molecule of MTV-RNA was estimated as 3.6 x 10(6) daltons.     

Structure of the Mouse Mammary Tumor Virus: Polypeptides and Glycoproteins

The polypeptide and glycoprotein compositions of the mouse mammary tumor virus virion from primary monolayer cultures of BALB/cfC3H mouse mammary tumor cells were studied by polyacrylamide gel electrophoresis by using internal and external labeling and Coomassie blue and periodic acid Schiff (PAS) staining. Twelve polypeptides were reproducibly resolved by the combined methods. Five major polypeptides were demonstrable with estimated molecular weights of 52,000, 36,000, 28,000, 14,000, and 10,000. Seven minor polypeptides were also consistently detected and had estimated molecular weights of 70,000, 60,000, 46,000, 38,000, 30,000, 22,000, and 17,000. Carbohydrate was associated with five of these polypeptides as measured by PAS stain or [(3)H] glucosamine labeling, or both. These glycoproteins had estimated molecular weights of 70,000, 60,000, 52,000, 36,000 and 10,000. The majority of the PAS stain and glucosamine was found in the 52,000 and 36,000 dalton peaks.     

Preferential Binding of Mouse Mammary Tumor Virus to B Lymphocytes

Mouse mammary tumor virus (MMTV) has been shown to preferentially infect B lymphocytes in vivo. We have used recombinant envelope-coated fluospheres and highly purified MMTV particles to study the distribution of the viral receptors on fresh mouse lymphocytes. A preferential dose-dependent binding to B lymphocytes was observed which could be competed with neutralizing antibodies. In contrast, T-lymphocyte binding remained at background levels. These results strongly suggest a higher density of viral receptor molecules on B lymphocytes than on T lymphocytes and correlate with the preferential initial infection of B lymphocytes observed in vivo.     

In Vitro System for Production of Mouse Mammary Tumor Virus

An in vitro system for production, purification, and concentration of mouse mammary tumor virus is described. Monolayer cultures of C(3)H mouse mammary tumor cells propagated at 34 C in roller bottles in the presence of dexamethasone, a glucocorticoid hormone, release B-type particles which possess ribonucleic acid and a ribonucleic acid-dependent deoxyribonucleic acid polymerase. One thousandfold concentration by ultracentrifugation with subsequent gradient fractionation yielded > 7 x 10(10) particles per ml in the 1.16- to 1.18-g/ml region. Mouse mammary tumor virus produced in this system was free of detectable C-type virus.     

Structure of the Mouse Mammary Tumor Virus: Characterization of Bald Particles

The polypeptide, antigenic, and morphological structure of the mouse mammary tumor virus was studied following protease digestion of intact virions. Intact, untreated virions (rho = 1.17 g/ml) had characteristic envelope spikes, five major polypeptides, and were precipitated by antisera against gp52. Two of the major polypeptides, with molecular weights of 52,000 (gp52) and 36,000 (gp36), had carbohydrate moieties. Protease treatment resulted in spikeless, "bald" particles (rho = 1.14 g/ml), which had altered surface antigenicity and which contained neither gp52 nor gp36. These data indicated that gp52 and gp36 were on the viral envelope. Bald particles retained a 28,000 dalton polypeptide (p28) which was proposed as the major internal polypeptide.     

Notch and Wingless Regulate Expression of Cuticle Patterning Genes

The cell surface receptor Notch is required during Drosophila embryogenesis for production of epidermal precursor cells. The secreted factor Wingless is required for specifying different types of cells during differentiation of tissues from these epidermal precursor cells. The results reported here show that the full-length Notch and a form of Notch truncated in the amino terminus associate with Wingless in S2 cells and in embryos. In S2 cells, Wingless and the two different forms of Notch regulate expression of Dfrizzled 2, a receptor of Wg; hairy, a negative regulator of achaete expression; shaggy, a negative regulator of engrailed expression; and patched, a negative regulator of wingless expression. Analyses of expression of the same genes in mutant N embryos indicate that the pattern of gene regulations observed in vitro reflects regulations in vivo. These results suggest that the strong genetic interactions observed between Notch and wingless genes during development of Drosophila is at least partly due to regulation of expression of cuticle patterning genes by Wingless and the two forms of Notch.     

Generation of a Novel MMTV-tTA Transgenic Mouse Strain for the Targeted Expression of Genes in the Embryonic and Postnatal Mammary Gland

We have generated a new and improved transgenic mouse strain that permits a temporally controlled expression of transgenes throughout mammary gland development. High expression of the tetracycline-regulatible transactivator (tTA) under control of the mouse mammary tumor virus long terminal repeat (MMTV-LTR) is restricted to mammary epithelial cells and the salivary gland. The novel MMTV-tTA mouse strain induces a sustained transactivation of responder transgenes, which can be swiftly suppressed through administration of doxycycline (Dox). An important characteristic of this strain is its expression in early progenitor cells of mammary gland anlagen beginning at day 13.5 of embryonic development. We show here that the MMTV-tTA can be used in combination with GFP reporter strains to visualize CK8/CK14-dual positive progenitors in newborn females and their derived basal and luminal epithelial cell lineages in adult females. Our observations suggest that the novel MMTV-tTA can be utilized to express exogenous proteins in multipotent mammary progenitors during the earliest stages of mammary gland development to assess their biological significance throughout mammogenesis. Moreover, we demonstrate that the expression of the MMTV-tTA is sustained during mammary gland tumorigenesis in female mice expressing wildtype ErbB2. This makes this strain particular valuable to target the expression of exogenous proteins into developing mammary tumors to assess their significance in biological processes, such as tumor cell growth and survival, metabolism, and metastasis.     

Mouse Mammary Tumor Virus Superantigen Expression in B Cells Is Regulated by a Central Enhancer within the pol Gene

Expression of mouse mammary tumor virus (MMTV)-encoded superantigens in B lymphocytes is required for viral transmission and pathogenesis. The mechanism of superantigen expression from the viral sag gene in B cells is largely unknown, due to problems with detection and quantification of these low-abundance proteins. We have established a sensitive superantigen-luciferase reporter assay to study the expression and regulation of the MMTV sag gene in B-cell lymphomas. The regulatory elements for retroviral gene expression are generally located in the 5′ long terminal repeat (LTR) of the provirus. However, we found that neither promoters nor enhancers in the MMTV 5′ LTR play a significant role in superantigen expression in these cells. Instead, the essential regulatory regions are located in the pol and env genes of MMTV. We report here that maximal sag expression in B-cell lines depends on an enhancer within the viral pol gene which can be localized to a minimal 183-bp region. Regulation of sag gene expression differs between B-cell lymphomas and pro-B cells, where an enhancer within the viral LTRs is involved. Thus, MMTV sag expression during B-cell development is achieved through the use of two separate enhancer elements.