|18 December 2000, Volume 19, Number 55, Pages 6472-6481|
|Table of contents Previous Article Next [PDF]
|Ets and retroviruses-transduction and activation of members of the Ets oncogene family in viral oncogenesis|
|Donald G Blair1 and Meropi Athanasiou2|
1National Cancer Institute, Division of Basic Science, Basic Research Laboratory, Oncogene Mechanisms Section, Frederick, Maryland, MD 21702-1207, USA
2Intramural Research Support Program, SAIC, Frederick, Maryland, MD 21702-1207, USA
Correspondence to: D G Blair, NCI-Frederick, Bld 469, Room 102, Frederick, Maryland, MD 21702-1207, USA
Studies of retroviral-induced oncogenesis in animal systems led to the initial discovery of viral oncogenes and their cellular homologs, and provided critical insights into their role in the neoplastic process. V-ets, the founding member of the ETS oncogene family, was originally identified as part of the fusion oncogene encoded by the avian acute leukemia virus E26 and subsequent analysis of virus induced leukemias led to the initial isolation of two other members of the ETS gene family. PU.1 was identified as a target of insertional activation in the majority of tumors induced by the murine Spleen Focus Forming virus (SFFV), while fli-1 proved to be the target of Friend murine leukemia virus (F-MuLV) in F-MuLV induced erythroleukemia, as well as that of the 10A1 and Graffi viruses. The common features of the erythroid and myeloid diseases induced by these viruses provided the initial demonstration that these and other members of the ETS family play important roles in hematopoietic development as well as disease. This review provides an overview of the role of ETS genes in retrovirally induced neoplasia, their possible mechanisms of action, and how these viral studies relate to current knowledge of the functions of these genes in hematopoiesis. Oncogene (2000) 19, 6472-6481.
ETS; retrovirus; oncogenesis; E26; F-MuLV; SFFV
Studies of genes associated with retroviral-induced neoplasia forms the basis for much of our present knowledge of oncogenes, and have contributed to our understanding of both gene function and the neoplastic process. Like so many of the oncogenes which have become well known and well-studied over the past 20 years, the ETS family of transcription factors was first identified through studies of retrovirally induced disease. The goal of this brief review is to examine the role of those members of the ETS family of oncogenic transcription factors linked directly to retrovirally-induced cancer.
The aspects of the retroviral life cycle that allows them to interact with cellular genes to produce aberrant gene expression and tumorigenesis are well known and have been exhaustively reviewed elsewhere (Coffin et al., 1997). Genomic sequences, including those with oncogenic potential, can be captured by retroviruses (Figure 1a), with the initial step thought to be viral integration adjacent to and in the same transcriptional orientation as a gene with oncogenic potential (proto-oncogene). This is followed by genetic rearrangements and recombination events leading to the generation of a virus carrying all or part of the cDNA sequence coding for the cellular gene product. The captured sequences, with the sole exception of the src gene captured in Rous Sarcoma virus (RSV), replace viral sequences, so that all such viruses but RSV are replication defective and require a replication-competent helper virus to propagate. The recombinant viruses then express truncated and/or mutated forms of the captured cellular sequences under the control of viral promoter elements. Infection of the appropriate target cells by such virus recombinants can induce cell transformation, and the tumors are characteristically poly- or oligo-clonal.
Alternatively, in the process termed insertional activation (Figure 1b), the retrovirus integrates in a site near a cellular gene (although near can be 10's of kilobases away in some cases). This integration can activate expression of the modified or normal cellular locus through one of two basic mechanisms. Activation can occur through the influence of virus enhancer elements on normal or cryptic promoters within the target gene locus (Figure 1b1). Viral sequences are often integrated in the opposite transcriptional orientation in relation to the cell gene, and need not be deleted or rearranged. The site of virus integration can be upstream, as shown in Figure 1b1, or downstream of the cell gene, and can be kilobases from the locus and exert influences over long chromosomal distances. The virally induced aberrant expression of the cell gene, in conjunction with other genetic events such as the inactivation of tumor suppressors, leads to neoplastic transformation.
In the second type of event, the cellular gene is expressed as a fused message under the control of the virus promoter. This can occur if a portion of the virus is deleted, allowing either the 5' or 3' proviral LTR, or splicing between a virus donor site and a recipient site within the cellular locus (see Figure 1b2), to generate a fused virus-cellular message. In some cases it results in the incorporation and translation of regions of the cell locus not normally expressed, generating abnormal forms of the cell protein. In both types of activation the requirement for multiple rare events results in tumors that are almost always monoclonal.
Members of the Ets gene family have been involved in both gene capture and retroviral insertions, although insertional activation of Ets genes has been more frequent. Initially, however, analysis of oncogenes captured by retroviruses was much easier technically, and it was studies of such an oncogene-containing retrovirus that led to the identification of the prototypical Ets gene and provided the name by which this family of related genes is now known.
It is important to note that all retroviruses, including the oncogenic ones, have undergone a variety of selective processes, prior to and during isolation, that contribute, often subtly, to the oncogenic properties of the virus. During multiple cycles of virus infection, virus mutants are produced as a result of the poor fidelity of retroviral reverse transcriptases, and over time the virus that infects and replicates most efficiently in the target cells of the organism will become the major variant in the population. This process often selects for viruses with altered envelope genes (host range variants) or altered promoter/enhancer sequences (LTRs), which can influence the ability of the captured oncogene sequences to be expressed in specific tissues. Mutations are also induced in the captured sequence, altering and in some cases activating its ability to transform the target cells. Perhaps even more crucial is the 'investigator-initiated selection', since most viruses were identified only because a single diseased chicken, mouse or cat was identified, allowing the virus responsible to be recovered, amplified and isolated.
The avian erythroleukemia virus E26 identifies the first ets gene
In 1962 a report from Bulgaria (Ivanov et al., 1962) described a retrovirus isolated from a case of spontaneous avian leukosis that upon in vivo passage could induce massive erythroblastosis in newborn chicks. The isolate was designated E26, and initial studies indicated it had the unusual property of inducing both myeloid and erythroid leukemias. E26 induced an erythroblast (Sotirov, 1981) that exhibited a more differentiated phenotype than those seen in erythroid disease induced by the previously identified Avian Erythroblastosis virus (AEV). However, like another virus, AMV (Avian Myeloblastosis virus), the E26 virus could also transform myeloblasts in vivo and in vitro (Graf et al., 1978; 1979; Moscovici et al., 1981; Radke et al., 1982). In addition, E26 could transform quail fibroblasts in tissue culture (Graf et al., 1979) and this property, together with its ability to induce a novel erythroid disease, suggested the E26 virus represented a unique isolate different from AMV or the erythroblastosis-inducing AEV.
The advent of molecular cloning provided an explanation for the novel properties of E26. Analysis of cloned E26 revealed that it possessed a unique tripartite structure (Figure 2), and that, in addition to retrovirus-derived sequences, E26 contained sequences derived from two cellular oncogenes. One was myb, which had previously been identified in AMV (Radke et al., 1982), but the second was a novel sequence, and was designated ets, for e26 transformation-specific sequence (Leprince et al., 1983; Nunn et al., 1983; 1984).
Using the viral ets sequence as a probe, it was possible to identify c-ets-1, the cellular homolog of v-ets, first in chickens (Leprince et al., 1983) and subsequently in multiple species (de Taisne et al., 1984; Watson et al., 1985). These genes, v-ets and c-ets-1, proved to be the first of what, with the subsequent isolation of the related ets-2 (Watson et al., 1985; 1986), became the ETS family of transcription factors. This family, now defined by the presence of a conserved DNA binding domain recognizing a canonical ... GGAA/T ... sequence, contains nearly 30 different members, represented by over 70 isolates from multiple species. The different members of the ETS family have been reviewed in detail elsewhere (Ghysdael and Boureux, 1997; Papas et al., 1997; Watson et al., 1990), as well as in other articles in this volume.
The E26 virus is one of three known retroviruses that have acquired multiple cellular oncogenes (Coffin et al., 1997). The 5.6 kD E26 genomic RNA (Figure 2) contains v-mybE and v-ets sequences flanked by the characteristic retroviral unique 5' and 3' regulatory sequences and deleted portions of viral gag and env (Leprince et al., 1983; Nunn et al., 1983). The virus-derived sequences are consistent with E26 having arisen through recombination between avian leukosis virus and cellular myb and ets-1 genes.
Like most oncogene-containing retroviruses, the v-mybE and v-ets are truncated versions of their cellular homologs and contain multiple mutations. The 0.8 kD E26 v-mybE contains an internal portion the c-myb gene, which is also a sub-fragment of the 1.2 kb v-mybA sequences found in AMV. While the AMV v-Myb protein contains 9 mutations, the protein encoded by v-mybE differs from c-Myb by only a single amino acid. In contrast, v-ets contains three amino acid mutations and at the RNA level differs from chicken c-ets-1 at both its 5' and 3' ends (Gegonne et al., 1987; Leprince et al., 1990; Watson et al., 1988). The alterations at the 3' end of v-ets are functionally critical to the transforming properties of the virus, since the residues encoded by the 3' region of c-ets were shown to be capable of repressing the DNA binding potential of c-ets (Lim et al., 1992). The novel 85 nucleotide 3' end of v-ets was apparently generated by an inversion of 3' c-ets sequences (Lautenberger and Papas, 1993), suggesting that the virus structure was generated by a complex combination of recombination and biological selection for a transforming virus, consistent with the rare occurrence of such tripartite viruses.
The E26 gene fusion encodes a 135 kD oncoprotein (p135gag-myb-ets), which in transformed cells is localized to the nucleus (Boyle et al., 1984). The myb and ets portions of the fusion protein have retained their respective DNA binding and transactivation domains (DBD and T.D. Figure 2), giving the protein the potential for complex interactions with multiple promoter elements.
The contributions of the oncogene-derived segments of the virus have been studied both in vivo and in vitro, and the results indicate that both v-ets and v-mybE can contribute to the transformation of different lineages and cell types in ways that demonstrate the subtleties involved in the neoplastic process. Viruses engineered to independently express v-mybE- and v-ets-encoded proteins demonstrated that the unique myb-ets gene fusion was critical to the properties of E26, but also showed interesting differences between in vitro and in vivo behavior. In vitro v-mybE, but not v-ets, could transform myeloid cells, while both were able to generate poorly replicating, erythroblast-like cells with a mature erythroid phenotype which were described as weakly transformed (Metz and Graf, 1991b). Co-expression of the two gene products as independent proteins resulted in increased proliferation and a more transformed phenotype, while the fused oncogenes generated transformed cells with a unique, more immature phenotype (Metz and Graf, 1991b). These immature cells could differentiate into both myoblasts and eosinophils in response to TPA or ras oncogene expression (Graf et al., 1992).
While the individual v-mybE and v-ets genes were able to transform specific target cells in tissue culture, they failed to induce either myeloid or erythroid disease in vivo (Metz and Graf, 1991a). However, co-injection of viruses separately expressing the two genes resulted in erythroid leukemias with only a slight delay relative to the normal E26. Analysis of transformed cells derived from these leukemic birds revealed that they contained recombinant viruses in which v-mybE and v-ets sequences were once again fused to each other in a myb-ets configuration. These new recombinant viruses expressed myb-ets fusion proteins, indicating the fusion, probably in that order, was required for leukemogenicity (Metz and Graf, 1991a), and demonstrating that the viral myb-ets fusion possessed unique activities not found in the individual components.
These results were consistent with and helped to explain properties of viral mutants that had been studied several years earlier. Analysis of E26 mutants had identified several that were temperature sensitive for myeloblast transformation (Beug et al., 1984), and one of these had been mapped to the v-mybE DNA binding domain (Frykberg et al., 1988). Mutation or deletion of the v-ets region by recombinant DNA techniques produced viruses that would transform myeloid cells in vitro but not cause leukemia in vivo (Nunn and Hunter, 1989). However, a mutant that mapped to the DNA binding domain of v-ets was temperature sensitive for erythroid cell transformation in vitro and could also transform myeloid cells, but with an altered phenotype. This observation suggested a role for Ets in myeloid as well as erythroid cell differentiation (Golay et al., 1988). Consistent with this, analysis of a series of deletion mutants showed that while the v-mybE DNA binding or transactivating domains are not required to transform erythroid cells, the DNA binding domain of v-ets is required for myeloid transformation (Domenget et al., 1992).
The role of v-ets has been further clarified by analysing cells transformed in vitro with an E26 virus in which the v-ets sequences were flanked by FLP recombinase target sequences (Rossi et al., 1996). Excision of v-ets by the induced expression of the FLP recombinase generated cells which could differentiate along the erythroid pathway, but failed to differentiate into eosinophils or myelocytes (Rossi et al., 1996). These results supported data indicating that v-ets was required to inhibit erythroid differentiation, but it further suggested that ets might play a role in eosinophil and myeloid differentiation.
Interestingly, the transforming potential of the E26 fusion oncogene is also conserved across multiple species. The ME26 virus, a replication-defective murine retroviral vector expressing the fusion protein, induced both erythroid and myeloid leukemia in newborn mice (Yuan et al., 1989). Cell lines derived from these mice were dependent on the erythroid hormone erythropoietin (EPO) for growth and, similar to what was seen in avian species, resembled early erythroid precursors (Ruscetti et al., 1992). As was seen in the case of E26, viruses that separately expressed either v-mybE or v-ets could not induce leukemias in vivo (Aurigemma et al., 1994) and the erythroleukemic properties of ME26 in mice depend on the presence of a myb-ets gene fusion.
In vitro, infection of established IL-3-dependent murine hematopoietic cell lines with ME26 induced these cells to express a more erythroid phenotype, increased levels of GATA-1, -globin, and EPO receptor mRNA, and to acquire the ability to proliferate in response to EPO (Athanasiou et al., 1996b; Ruscetti et al., 1992). Expression of the gag-myb-ets oncogene in these cells, in the presence of Epo, interfered with the apoptotic response normally induced by the removal of IL-3. Since ME26-infected animals demonstrated increased levels of Epo, this result suggested that the apoptotic pathways could be another possible target for the fusion oncoprotein to promote the development of malignant cells (Athanasiou et al., 1996b). Viruses expressing either v-mybE or v-ets alone failed to alter the growth properties of these cells (Athanasiou et al., 1996b).
ME26 could also transform NIH3T3 murine fibroblasts to proliferate in media containing levels of serum insufficient for the growth of non-infected cells (Yuan et al., 1989), a property similar to the effect of E26 on chicken fibroblasts (Jurdic et al., 1987; Ravel-Chapuis et al., 1991). V-ets alone did not transform murine fibroblasts in tissue culture (Y Yuan, M Athanasiou and DG Blair, unpublished observation), but c-ets-1, the cellular homologue of v-ets, has been reported to stimulate proliferation and induce a transformed phenotype (Seth and Papas, 1990; Topol et al., 1992). However, how c-ets-1 induces these changes has not been determined.
Viral studies clearly indicated that regions encoded by both v-ets and v-mybE play a role in the transformation induced by the gag-myb-ets fusion oncogene, but the precise molecular targets and mechanisms remain elusive. Two genes have been identified as targets for the v-mybE (mim-1, (Ness et al., 1989)) and v-ets (rem-1, (Kraut et al., 1995)) portions of the fusion protein. Neither appears to be critical to transformation, however, although rem-1 can promote increased growth of E26 ts transformed cells (Kraut et al., 1995). Interestingly, the myb-ets fusion protein has been shown to inhibit c-ErbA, the thyroid hormone receptor, and the retinoic acid receptor (Rascle et al., 1996). This is the same mechanism thought to be involved in erythroid transformation by the v-erbA-containing AEV, and suggests that erythroid transformation by E26 and AEV could involve a common pathway.
Ets and Myb are both well characterized transcription factors, and the presence of binding sites in a wide variety of genes expressed in hematopoietic cells raises numerous possibilities for aberrant gene activation in E26 infected cells. While both genes are normally thought of as transcriptional activators, nevertheless their action in inducing hematopoietic disease appears to occur through the inhibition of normal differentiation, indicating that the gene fusion can act, either directly or indirectly, as an inhibitor of the signaling pathways involved in those processes. Ets-1 is known to interact with the MafB protein and repress erythroid differentiation (Sieweke et al., 1996, 1997). However, the analysis of the E26 virus described above suggests that the fusion gag-myb-ets oncogene possesses unique properties not found in its individual component parts. The fusion protein has been reported to upregulate the bcl-2 promoter in avian myeloid cells (Frampton et al., 1996), and increased levels of bcl-2 expression have also been seen in ME26-infected murine hematopoietic cell lines (M Athanasiou, unpublished observations). As was proposed for the affects seen in IL-3-dependent murine cell lines (Athanasiou et al., 1996b), the observations in avian cells suggest that a gag-myb-ets-induced apoptotic block could contribute to increased cell survival and tumor development.
Retroviral insertion activates members of the ETS gene family
While oncogene-carrying retroviruses have led to the identification of a number of genes critical to the process of tumorigenesis, the majority of oncogenic retroviruses do not contain viral oncogene (v-onc) sequences (Coffin et al., 1997). These replication competent, v-onc-free viruses generally do not transform cells in culture, but transform cells in vivo through a multi-step process involving insertional activation (see Figure 1b) of cellular genes. Unlike the polyclonal tumors induced by the v-onc-containing viruses, these clonal tumors arise from a single rare cell in which the input virus has been inserted adjacent to, and has activated the expression of, a specific gene. Two members of the ETS family were initially found through site-specific integrations seen in clonal erythroid tumors induced by different viruses, the replication competent Friend strain of MuLV (F-MuLV) and the replication-defective virus SFFV (Spleen Focus-Forming Virus), which encodes a novel env-derived gp55 protein.
Spleen focus-forming virus (SFFV) insertionally activates the ETS genespi-1 PU.1
The Friend leukemia virus complex (Friend, 1957) consists of a mixture of the replication competent friend murine leukemia virus (F-MuLV) and a novel replication-defective virus, spleen focus forming virus (SFFV), which is responsible for the induction of an acute erythroleukemia in adult and newborn mice. The disease progresses from an initial proliferation of proerythroblast cells in the spleen and liver to the clonal expansion of malignant proerythroblastic cells within 3-8 weeks, depending on the strain of SFFV involved. The initial proliferative stage of the disease involves activation of the erythropoietin receptor by the 55 kD, envelope derived, SFFV-encoded protein, and details of the structure and biology of SFFV can be found in several recent reviews (Ben-David and Bernstein, 1991; Ruscetti, 1999).
Despite the fact that it lacks an oncogene and induces a clonal disease, SFFV induces an acute disease even in adult mice. A search for a unique viral integration site to explain the clonal nature of the disease led to the identification of a common integration locus in erythroid tumors (Moreau-Gachelin et al., 1988; Spiro et al., 1988). The gene, termed spi-1 (SSFV proviral integration-1, (Moreau-Gachelin et al., 1988)), was highly expressed as a 1.4 kB message in tumors but was only weakly expressed in spleen cells isolated from the early hyperproliferative stage of the disease (Moreau-Gachelin et al., 1989). Curiously, all proviral integrations (20.20) were localized 5' of the origin of spi-1 transcription, with viral transcription oriented in the opposite direction to that of spi-1, consistent with an enhancer-mediated viral activation mechanism (Figure 1b1). Subsequently, during a screen of a macrophage cDNA library for proteins which could bind to MHC class II promoter sequences, Klemsz et al., (1990) identified a protein capable of binding to the purine-rich sequence 5'-GAGGAA-3' (the PU box) which they named PU.1. PU.1 could act as a transcription factor, and a comparison of the amino acid sequence of the PU.1 DNA-binding domain to that of other members of the ETS gene family, together with its ability to bind to the canonical ... GGAA/T... ETS binding sequence, identified it as a new ETS family member (Klemsz et al., 1990). A similar comparison to the spi-1 sequence confirmed that PU.1 and spi-1 represented independent cloning of the same gene (Goebl, 1990).
SFFV transformed cells express the virally encoded gp55 envelope-derived protein together with elevated levels of PU.1, and in many cases have also undergone p53 inactivation (Ben David et al., 1988; Mowat et al., 1985). P53-null mice develop SFFV-induced leukemias at an accelerated rate but spi-1/PU.1 is still insertionally activated (Lavigueur and Bernstein, 1991), suggesting that p53 mutations most likely contribute to tumor progression. PU.1 mRNA in erythroleukemic cells is identical to that found in normal cells, indicating that overexpression of the normal gene product is sufficient to provide the necessary transforming functions. The precise mechanism through which PU.1 contributes to the erythroleukemic transformation of cells infected by SFFV is still not defined, but it is likely that it acts to block erythroid differentiation in these cells and permits their continued proliferation. Spi-1/PU.1 transgenic mice develop erythroleukemia, and leukemic erythroblasts from these mice are unable to differentiate and are Epo-dependent for growth (Moreau-Gachelin et al., 1996). Interestingly other lineages do not seem to be affected.
Consistent with an inhibitory role for PU.1 in the differentiation process, when PU.1 is overexpressed in erythroblasts in long-term bone marrow cultures, a polyclonal population of non-differentiating, immortal cells is generated (Schuetze et al., 1993). Proliferation of MEL (Murine Erythroid Leukemia) cells, generated by SFFV transformation, can be inhibited by spi-1/PU.1 antisense oligonucleotides (Delgado et al., 1994). MEL cells also exhibit a reduction in spi-1/PU.1 RNA levels when erythroid differentiation is induced by DMSO, and the introduction of vectors constitutively expressing spi-1/PU.1 during this process results in growth inhibition and increased apoptotic death (Yamada et al., 1997). Spi-1/PU.1 overexpression in the pluripotent human erythroleukemia line K562 leads to reduced expression of erythroid markers and an increase in monocytic markers (Delgado et al., 1998), also consistent with an inhibitory effect on erythroid differentiation. Spi-1/PU.1 expression has also been shown to interfere with the expression of the thyroid and retinoic acid receptors (Gauthier et al., 1993), which can act as modulators of erythroid differentiation, and which are thought to be the target of v-erbA, a part of the erythroleukemia-inducing AEV. Since the E26 myb-ets fusion oncoprotein has also been reported to inhibit these receptors (Rascle et al., 1996), it is tempting to speculate that both ETS family members could act through similar targets to induce erythroid disease.
Multiple insertional activations of fli-1 by different viruses
Friend MuLV insertion identified the ETS family member fli-1: While the Friend virus complex causes disease through the combined action of the replication defective SFFV-encoded gp55 and the insertional activation of spi-1 (with a contribution from the inactivation of p53), the replication-competent F-MuLV component is also capable of inducing a range of erythroid, lymphoid and myeloid disease in susceptible strains (Silver and Kozak, 1986). These diseases, like many of those induced by viruses that lack oncogenes, exhibits a longer latency (3-6 weeks compared to 1-3 weeks for the SFFV-induced disease). In addition, only newborn mice are susceptible to F-MuLV-induced disease, while both adult and newborns are vulnerable to SFFV infection.
The properties of the virus and the disease it generated suggested an insertional activation mechanism, and analysis of F-MuLV-induced erythroleukemias identified a proviral integration site common to 75% (9/12) of those tested (Ben-David et al., 1990). The site was termed fli-1 (Friend leukemia integration 1), and that designation was subsequently given to the gene encoding a cDNA clone that was highly expressed only in the erythroleukemic cell lines containing rearrangements at the fli-1 locus. This RNA contained a 1467 nucleotide open reading frame encoding a 51 kD, 452aa protein. Fli-1 was identified as an ETS family member on the basis of its sequence homology to human erg-2 (100%) and ets-1 (68%) within the 85aa canonical ETS DNA-binding domain (Ben-David et al., 1991). The structural and functional properties of the Fli-1 protein have been recently reviewed (Ghysdael and Boureux, 1997) and are discussed in great detail elsewhere in this volume by Truong and Ben-David.
Fli-1 activation by Cas-Br-E, 10A1 and Graffi MuLV: The insertional activation of the same gene by multiple retroviruses is a common feature of retrovirally induced disease, as is the presence of the same insertionally activated gene in multiple tumor types (Coffin et al., 1997). Although fli-1 was initially identified as an F-MuLV integration site in Friend MuLV-induced erythroleukemia, other viruses also activate it in non-erythroid disease.
Analysis of non-B, non-T cell leukemias induced in NIH/Swiss mice by the ecotropic, nondefective Cas-Br-E strain of MuLV revealed common virus insertions within a 100 bp segment immediately 5' of fli-1 in 71% of the tumors analysed (Bergeron et al., 1991, 1992). The virus was always found to be inserted in the sense (5' to 3') orientation, with the viruses intact and unrearranged in about half of the tumors (Bergeron et al., 1992). Activation of fli-1 expression occurred through the generation of a fused virus-fli-1 message, a mechanism that contrasted to that seen in F-MuLV-induced erythroleukemias. However, as in F-MuLV erythroleukemias, Cas-Br-E tumors were associated with loss of p53 (Bergeron et al., 1993), suggesting that the lack of p53 activity plays a role in tumors induced by both viruses. The Cas-Br-E virus induces a wide variety of tumors, as well as hind limb paralysis, in susceptible mice (Gardner et al., 1978), and like many such viruses, generates host-range recombinant MCF (Mink Cell Focus-forming) viruses during the course of infection. Both ecotropic and MCF-type virus were found to insertionally activate fli-1 with no apparent preference, while insertion at a second locus, Evi-1, which occurs in 22% of the tumors, always involves the ecotropic variant (Bergeron et al., 1992, 1993).
A second virus, the 10A1 host range variant isolate of MuLV (Ott et al., 1990, 1992), also appears to induce hematopoietic tumors through insertional activation of fli-1. Remarkably, 10A1 integrates at the same site as Cas-Br-E, inserting just upstream of the fli-1 protein initiation site and, like Cas-Br-E, activates fli-1 expression by promoter insertion. In both cases the insertion disconnects the sequences encoding Fli-1 protein from the fli-1 promoter and other regulatory regions, placing it instead under the constitutive control of the retroviral LTR. Consistent with these properties, the diseases induced by the two viruses are similar, and analysis of the pattern of markers expressed by 10A1 tumor cells suggests that the cells resemble early hematopoietic precursors (Ott et al., 1994).
Recent evidence suggests that fli-1 can also be insertionally activated by a third virus, resulting in the generation of yet another type of transformed hematopoietic cell. A small number of tumors induced by virus derived from two molecular clones of Graffi MuLV (Ru et al., 1993) contained common insertion sites in the c-myc (6/30), as well as the fli-1 locus (3/30), with the latter specifically exhibiting elevated levels of fli-1 expression. The Graffi insertions in c-fli-1 were 1 to 5 Kb upstream of the first initiation site and in an antisense orientation as were seen in F-MuLV-induced tumors. Both Graffi and F-MuLV activate fli-1 via their enhancers, while Cas-Br-E and 10A utilize promoter insertion, and it has been suggested that the enhancers of these latter viruses may be weaker than those of Graffi and F-MuLV, requiring a direct linkage to the virus promoter to get sufficient fli-1 expression (Ru et al., 1993). Graffi-induced tumors were reported to be granulocytic on the basis of histochemical characteristics, suggesting fli-1 could also play a role in the development of this class of leukemias (Denicourt et al., 1999).
Integration of Moloney MuLV in the ets-1 locus
A third insertional activation of an Ets gene was found during a search for common integration sites in tumors and tumor-derived cell lines isolated from Mo-MuLV induced rat tumors. A site, originally designated tpl-1 (for tumor progression locus-1), was identified in one primary thymic tumor (but not in a secondary isolated from the spleen), as well as in three cell lines, two of which were independently selected from the same primary tumor explant (Bear et al., 1989). Subsequent analysis indicated that the site of insertion was immediately 5' of the first ets-1 exon, and that both tumor and normal cells encoded the same c-ets-1 mRNAs (Bellacosa et al., 1994), suggesting activation occurred at the level of the viral enhancer. Despite the apparent growth advantage of cells carrying this insertion, there were only subtle differences in ets-1 RNA levels and splicing patterns, raising the possibility that minor regulatory changes, or perhaps activation of additional genes by this proviral insertion (Bellacosa et al., 1991).
The role of fli-1 activation in multiple leukemias
The range of virally induced tumors found to involve insertional activation of fli-1 suggests that abnormal expression of this gene can play a role in a variety of hematopoietic lineages. Fli-1 is normally expressed at a high level in hematopoietic cells and tissues, particularly thymus and spleen (Klemsz et al., 1993), and it is likely to play a functional role in those tissues. The induction of overexpression of Fli-1 in pluripotent human hematopoietic cells through infection with a Fli-1-expressing retroviral vector induces a megakaryocytic phenotype and increases the level of expression of megakaryocytic markers (Athanasiou et al., 1996a). In addition, the elevated levels of fli-1 in these cells also inhibits their ability to respond to inducers of erythroid differentiation (Athanasiou et al., 2000). Fli-1 over-expression has also been shown to block Epo-induced differentiation in HB60, an established human erythroblastic cell line (Tamir et al., 1999), and a similar effect was seen in Friend cells treated with the erythroid inducer HMBA (hexamentylenebisacetamide) (Starck et al., 1999). In avian primary erythroblasts overexpression of fli-1 can also block erythroid differentiation in response to Epo and inhibit the apoptotic response normally seen when Epo is removed, in part through elevation of bcl-2 levels (Pereira et al., 1999). Taken together, these results suggest retroviral induction of fli-1 acts to induce the megakaryocytic and block the erythroid differentiation pathway, perhaps by increasing cell resistance to the induction of an apoptotic response.
Both roles for fli-1 (megakaryocytic activation and erythroid inhibition) have been supported by recent studies involving targeted deletions of fli-1. Mice harboring such deletions die in early embryogenesis and exhibit defects in hematopoiesis and a tendency to hemorrhage (Hart et al., 2000; Spyropoulos et al., 2000). The hematopoietic defects included abnormal megakaryocytic and erythroid differentiation (Hart et al., 2000; Kawada et al., submitted). Significantly, the defective megakaryopoiesis seen in Fli-1 null embryos resembles that seen in patients suffering from Jacobsen or Paris-Trousseau Syndrome, who exhibit a characteristic 11q chromosomal deletion that includes the human FLI-1 locus (Hart et al., 2000). The properties of mice harboring targeted deletions of fli-1 and other members of the ETS gene family are discussed extensively elsewhere in this volume by Bartel and Spyropoulos.
The properties of transgenic mice have also been suggestive of other Fli-1 targets that could contribute to its wide role in neoplasia. Overexpression of fli-1 under control of the H-2Kk promoter leads to elevated expression of the gene in multiple tissues, particularly thymus and spleen (Zhang et al., 1995). While the mice die of an immunologically based renal disease they exhibit numerous hematopoietic abnormalities, including splenomegaly, B cell hyperplasia, and the accumulation of abnormal T and B cells. The B cells exhibit increased proliferation and were refractory to mitogen-induced apoptosis, suggesting elevated levels of fli-1 could affect lymphoid function and programmed cell death, functions likely to be involved in the neoplasms observed following insertional activation of fli-1.
Fli-1 binds and transactivates promoters containing the canonical ETS binding sequence, although its DNA binding activities are distinguishable from other Ets proteins (Bosselut et al., 1993; Zhang et al., 1993). Fli-1 is capable of transactivating the promoters of genes associated with vasculogenesis and cell adhesion (de Launoit et al., 1998; Dube et al., 1999; Gory et al., 1998), and can regulate megakaryocytic gene expression (Bastian et al., 1999; Deveaux et al., 1996; Lemarchandel et al., 1993; Schwachtgen et al., 1997; Zhang et al., 1993).
While Fli-1 is normally thought to act as a transcriptional activator, it has been shown to inhibit transcription of the rb gene during erythroid differentiation (Tamir et al., 1999), apparently by binding to a cryptic Ets binding site within the rb promoter. This could be significant since loss of Rb in mice leads to delayed or impaired erythropoiesis (Clarke et al., 1992; Hu et al., 1997; Jacks et al., 1992; Lee et al., 1992). Fli-1 also interferes with the autoregulatory transcriptional activation of GATA-1 (Athanasiou et al., 2000), a major effector of hematopoietic differentiation, as well as tel, an ETS family member whose loss is associated with leukemic transformation (Kwiatkowski et al., 1998). Interestingly, fli-1 has been reported to be positively regulated by spi-1/PU.1, the gene activated by SFFV-induced erythroleukemia (Starck et al., 1999), suggesting a possible common mechanism of disease or a common pathway of normal gene action.
Fli-1's involvement in erythroid, myeloid and granulocytic leukemias suggests that the gene could play a role in multiple hematopoietic lineages, but the viruses themselves are a determining factor in the disease that develops following infection. The diseases induced by different viruses are clearly due in part to the ability of the virus to infect specific target tissues or cells at varying stages of hematopoietic development. This ability is due both to differences in viral host range (determined by the virus encoded envelope gene), and to the presence of different virus receptors found on the surface of particular cells. While the latter is probably evolutionarily and developmentally fixed, viral envelope genes undergo mutation or recombination during virus replication, leading to changes which can contribute to the disease inducing capacity of the virus.
The ability of a particular virus strain to infect and spread is also influenced by subtle differences in viral LTR sequences as well. These subtle sequence differences influence both the level of viral expression and the level of insertionally activated expression driven by the LTR promoter or enhancer sequences. How the virus, cell and fli-1 gene itself each contribute to this process, however, remains to be determined.
Conclusions and future implications
The analysis of avian and murine retrovirus-induced disease originally led to the identification of the ETS family of oncogenes, and the properties of virally induced diseases that involve ETS family members still pose questions that are basic to understanding oncogenesis in man. Rearrangements and translocations seen in human cancers that lead to increased expression of abnormal gene products are mimicked in the insertional activation and viral oncogene recombinants seen in retroviral diseases. The myb-ets fusion in the E26 virus and the fusions, involving fli-1 and other members of the ETS family, seen in Ewing's Sarcomas (Delattre et al., 1992), and could suggest a propensity of these genes to generated oncogenic fusions.
The viral-induced disease provide in vivo models in which cellular and immunological factors interact, as they do in human cancer, to generate the response of the entire organism to disease. The development of animal models through transgenic and conditional, targeted gene disruption technologies should allow a more thorough dissection of the roles of specific genes in specific tissues. The recent generation of fli-1 knockout mice (Hart et al., 2000; Spyropoulos et al., 2000), for example, confirms the critical role of the gene in hematopoiesis and particularly megakaryopoiesis. The generation of mice in which gene disruption is specifically targeted to tissues and developmental stages, and the crossing of mice deficient in ets-1, spi-1/PU.1, and fli-1 will help to separate the specific and redundant functions of these genes in hematopoiesis. An understanding of the normal functions of these genes, together with knowledge of the effects of their virus-induced aberrant expression, may eventually answer the questions first raised by those studying the retrovirally-induced neoplasias in birds and rodents.
We want to express our gratitude for the opportunity to work with Takis Papas, who first introduced us to the ETS family of oncogenes, and whose enthusiasm, support, encouragement and friendship we were fortunate to experience through our many years of association. The authors also wish to thank Dennis Watson for helpful suggestions and permission to cite unpublished data, and Karen Cannon for help in preparation of the manuscript. This work was supported in part with Federal Funds from the National Cancer Institute, National Institutes of Health, under contract No. N01-CO-56000. The content of this publication does not necessarily reflect the views or policies of the Department of Health and Human Services, nor does mention of trade names, commercial products, or organizations imply endorsement by the US Government.
Athanasiou M, Clausen PA, Mavrothalassitis GJ, Zhang XK, Watson DK and Blair DG. (1996a). Cell Growth Differ. 7, 1525-1584. MEDLINE
Athanasiou M, Mavrothalassitis GJ, Sun-Hoffman L and Blair DG. (2000). Leukemia 14, 439-445. MEDLINE
Athanasiou M, Mavrothalassitis GJ, Yuan CC and Blair DG. (1996b). Oncogene 12, 337-344. MEDLINE
Aurigemma RE, Shen-Ong GL, Athanasiou M, Hanson CA, Yuan CC, Blair DG and Ruscetti SK. (1994). Molecular Biology of Haematopoiesis, Vol. 3. Nader GA, Shadduck RK, Levine AS and Takaku F (eds). Intercept, Ltd.: Andover, Hampshire,, pp. 669-679.
Bastian LS, Kwiatkowski BA, Breininger J, Danner S and Roth G. (1999). Blood 93, 2637-2644. MEDLINE
Bear SE, Bellacosa A, Lazo PA, Jenkins NA, Copeland NG, Hanson C, Levan G and Tsichlis PN. (1989). Proc. Natl. Acad. Sci. USA 86, 7495-7499. MEDLINE
Bellacosa A, Datta K, Bear SE, Patriotis C, Lazo PA, Copeland NG, Jenkins NA and Tsichlis PN. (1994). J. Virol 68, 2320-2330. MEDLINE
Bellacosa A, Lazo PA, Bear SE and Tsichlis PN. (1991). Mol. Cell Biol. 11, 2864-2872. MEDLINE
Ben-David Y and Bernstein A. (1991). Cell 66, 831-834. MEDLINE
Ben-David Y, Giddens EB and Bernstein A. (1990). Proc. Natl. Acad. Sci. USA 87, 1332-1336. MEDLINE
Ben-David Y, Giddens EB, Letwin K and Bernstein A. (1991). Genes Dev. 5, 908-918. MEDLINE
Ben-David Y, Prideaux VR, Chow V, Benchimol S and Bernstein A. (1988). Oncogene 3, 179-185. MEDLINE
Bergeron D, Houde J, Poliquin L, Barbeau B and Rassart E. (1993). Leukemia 7, 954-962. MEDLINE
Bergeron D, Poliquin L, Houde J, Barbeau B and Rassart E. (1992). Virology 191, 661-669. MEDLINE
Bergeron D, Poliquin L, Kozak CA and Rassart E. (1991). J. Virol 65, 7-15. MEDLINE
Beug H, Leutz A, Kahn P and Graf T. (1984). Cell 39, 579-588. MEDLINE
Bosselut R, Levin J, Adjadj E and Ghysdael J. (1993). Nucleic Acids Res. 21, 5184-5191. MEDLINE
Boyle WJ, Lampert MA, Lipsick JS and Baluda MA. (1984). Proc. Natl. Acad. Sci. USA 81, 4265-4269. MEDLINE
Clarke AR, Maandag ER, van Roon M, van der Lugt NM, van der Valk M, Hooper ML, Berns A and te Riele H. (1992). Nature 359, 328-330. MEDLINE
Coffin JM, Hughes SH and Varmus HE. (1997). Retroviruses. CSHL Press, Cold Spring Harbor, NY.,
Delattre O, Zucman J, Plougastel B, Desmaze C, Melot T, Peter M, Kovar H, Joubert I, de Jong P, Rouleau G, Aurias A and Thomas G. (1992). Nature 359, 162-165. MEDLINE
de Launoit Y, Audette M, Pelczar H, Plaza S and Baert JL. (1998). Oncogene 16, 2065-2073. MEDLINE
Delgado MD, Gutierrez P, Richard C, Cuadrado MA, Moreau-Gachelin F and Leon J. (1998). Biochem. Biophys. Res. Commun. 252, 383-391. MEDLINE
Delgado MD, Hallier M, Meneceur P, Tavitian A and Moreau-Gachelin F. (1994). Oncogene 9, 1723-1727. MEDLINE
Denicourt C, Edouard E and Rassart E. (1999). J. Virol. 73, 4439-4442. MEDLINE
de Taisne C, Gegonne A, Stehelin D, Bernheim A and Berger R. (1984). Nature 310, 581-587. MEDLINE
Deveaux S, Filipe A, Lemarchandel V, Ghysdael J, Romeo PH and Mignotte V. (1996). Blood 87, 4678-4685. MEDLINE
Domenget C, Leprince D, Pain B, Peyrol S, Li RP, Stehelin D, Samarut J and Jurdic P. (1992). Oncogene 7, 2231-2241. MEDLINE
Dube A, Akbarali Y, Sato TN, Libermann TA and Oettgen P. (1999). Circ. Res. 84, 1177-1185. MEDLINE
Frampton J, Ramqvist T and Graf T. (1996). Genes Dev. 10, 2720-2731. MEDLINE
Friend C. (1957). J. of Exper. Med. 105, 307-318.
Frykberg L, Metz T, Brady G, Introna M, Beug H, Vennstrom B and Graf T. (1988). Oncogene Res. 3, 313-322. MEDLINE
Gardner MB, Klement V, Henderson BE, Casagrande J, Bryant ML, Doutherty MF and Estes JD. (1978). Tumors of early life in man and animals. Severi L. (ed.):. Monteluce, Italy, , pp. 343-356.
Gauthier JM, Bourachot B, Doucas V, Yaniv M and Moreau-Gachelin F. (1993). EMBO J. 12, 5089-5096. MEDLINE
Gegonne A, Leprince D, Pognonec P, Dernis D, Raes MB, Stehelin D and Ghysdael J. (1987). Virology 156, 177-180. MEDLINE
Ghysdael J and Boureux A. (1997). Oncogenes as transcriptional regulators. Vol. 1. Yaniv M and Ghysdael J. (eds). Birkhauser Verlag: Basel, Switzerland, pp. 29-88.
Goebl MK. (1990). Cell 61, 1165-1166. MEDLINE
Golay J, Introna M and Graf T. (1988). Cell 55, 1147-1158. MEDLINE
Gory S, Dalmon J, Prandini MH, Kortulewski T, de Launoit Y and Huber P. (1998). J. Biol. Chem. 273, 6750-6755. MEDLINE
Graf T, Beug H, Royer-Pokora B and Meyer-Glauner W. (1978). In: Clarkson B, Marks PA and Till JE (eds). Differentiation of normal and neoplastic hematopoietic cells. Cold Spring Harbor, Cold Spring Harbor Laboratory, 625-639.
Graf T, McNagny K, Brady G and Frampton J. (1992). Cell 70, 201-213. MEDLINE
Graf T, Oker-Blom N, Todorov TG and Beug H. (1979). Virology 99, 431-436. MEDLINE
Hart A, Melet F, Grossfeld P, Chien K, Jones C, Tunnacliffe A, Favier R and Bernstein A. (2000). Immunity 13, 167-177. MEDLINE
Hu N, Gulley ML, Kung JT and Lee EY. (1997). Cancer Res. 57, 4123-4129. MEDLINE
Ivanov X, Mladenov Z, Nedyalk S and Todorov TG. (1962). Bugarian Acad. Sci. Bull. Inst. Pathol. Comp. Animaux Domestiques. 9, 5-36.
Jacks T, Fazeli A, Schmitt EM, Bronson RT, Goodell MA and Weinberg RA. (1992). Nature 359, 295-300. MEDLINE
Jurdic P, Benchaibi M, Gandrillon O and Samarut J. (1987). J. Virol. 61, 3058-5765. MEDLINE
Klemsz MJ, Maki RA, Papayannopoulou T, Moore J and Hromas R. (1993). J. Biol. Chem. 268, 5769-5773. MEDLINE
Klemsz MJ, McKercher SR, Celada A, Van Beveren C and Maki RA. (1990). Cell. 61, 113-124. MEDLINE
Kraut N, Frampton J and Graf T. (1995). Oncogene 10, 1027-1036. MEDLINE
Kwiatkowski BA, Bastian LS, Bauer Jr TR, Tsai S, Zielinska-Kwiatkowska AG and Hickstein DD. (1998). J. Biol. Chem. 273, 17525-17530. MEDLINE
Lautenberger JA and Papas TS. (1993). J. Virol. 67, 610-612. MEDLINE
Lavigueur A and Bernstein A. (1991). Oncogene 6, 2197-2201. MEDLINE
Lee EY, Chang CY, Hu N, Wang YC, Lai CC, Herrup K, Lee WH and Bradley A. (1992). Nature 359, 288-294. MEDLINE
Lemarchandel V, Ghysdael J, Mignotte V, Rahuel C and Romeo PH. (1993). Mol. Cell. Biol. 13, 668-676. MEDLINE
Leprince D, Gegonne A, Coll J, de Taisne C, Schneeberger A, Lagrou C and Stehelin D. (1983). Nature 306, 395-397. MEDLINE
Leprince D, Gesquiere JC and Stehelin D. (1990). Oncogene Res. 5, 255-265. MEDLINE
Lim F, Kraut N, Framptom J and Graf T. (1992). EMBO J. 11, 643-652. MEDLINE
Metz T and Graf T. (1991a). Cell. 66, 95-105.
Metz T and Graf T. (1991b). Genes. Dev. 5, 369-380.
Moreau-Gachelin F, Ray D, Mattei MG, Tambourin P and Tavitian A. (1989). Oncogene 4, 1449-1456. MEDLINE
Moreau-Gachelin F, Tavitian A and Tambourin P. (1988). Nature 331, 277-280. MEDLINE
Moreau-Gachelin F, Wendling F, Molina T, Denis N, Titeux M, Grimber G, Briand P, Vainchenker W and Tavitian A. (1996). Mol. Cell. Biol. 16, 2453-2463. MEDLINE
Moscovici C, Samarut J, Gazzolo L and Moscovici MG. (1981). Virology 113, 765-768. MEDLINE
Mowat M, Cheng A, Kimura N, Bernstein A and Benchimol S. (1985). Nature 314, 633-636. MEDLINE
Ness SA, Marknell A and Graf T. (1989). Cell. 59, 1115-1125. MEDLINE
Nunn M, Weiher H, Bullock P and Duesberg P. (1984). Virology 139, 330-339. MEDLINE
Nunn MF and Hunter T. (1989). J. Virol. 63, 398-402. MEDLINE
Nunn MF, Seeburg PH, Moscovici C and Duesberg PH. (1983). Nature 306, 391-395. MEDLINE
Ott D, Friedrich R and Rein A. (1990). J. Virol. 64, 757-766. MEDLINE
Ott DE, Keller J and Rein A. (1994). Virology 205, 563-568. MEDLINE
Ott DE, Keller J, Sill K and Rein A. (1992). J. Virol. 66, 6107-6116. MEDLINE
Papas TS, Bhat NK, Spyropoulos DD, Miaatvedt AE, Vournakis J, Seth A and Watson DK. (1997). Leukemia 11, (Suppl 3) 557-566. MEDLINE
Pereira R, Quang CT, Lesault I, Dolznig H, Beug H and Ghysdael J. (1999). Oncogene 18, 1597-1608. MEDLINE
Radke K, Beug H, Kornfeld S and Graf T. (1982). Cell 31, 643-653. MEDLINE
Rascle A, Ferrand N, Gandrillon O and Samarut J. (1996). Mol. Cell. Biol. 16, 6338-6351. MEDLINE
Ravel-Chapuis P, Leprince D, Pain B, Li R, Domenget C, Stehelin D, Samarut J and Jurdic P. (1991). J. Virol. 65, 3928-3931. MEDLINE
Rossi F, McNagny KM, Logie C, Stewart AF and Graf T. (1996). Curr. Biol. 6, 866-872. MEDLINE
Ru M, Shustik C and Rassart E. (1993). J. Virol. 67, 4722-4731. MEDLINE
Ruscetti S, Aurigemma R, Yuan CC, Sawyer S and Blair DG. (1992). J. Virol. 66, 20-26. MEDLINE
Ruscetti SK. (1999). Int. J. Biochem. Cell. Biol. 31, 1089-1109. MEDLINE
Schuetze S, Stenberg PE and Kabat D. (1993). Mol. Cell. Biol. 13, 5670-5678. MEDLINE
Schwachtgen JL, Janel N, Barek L, Duterque-Coquillaud M, Ghysdael J, Meyer D and Kerbiriou-Nabias D. (1997). Oncogene 15, 3091-3102. MEDLINE
Seth A and Papas TS. (1990). Oncogene 5, 1761-1767. MEDLINE
Sieweke MH, Tekotte H, Frampton J and Graf T. (1996). Cell 85, 49-60. MEDLINE
Sieweke MH, Tekotte H, Frampton J and Graf T. (1997). Leukemia 11, (Suppl 3) 486-488. MEDLINE
Silver J and Kozak C. (1986). J. Virol. 57, 526-583. MEDLINE
Sotirov N. (1981). J. Natl. Cancer Inst. 66, 1143-1149. MEDLINE
Spiro C, Gliniak B and Kabat D. (1988). J. Virol. 62, 4129-4135. MEDLINE
Spyropoulos DD, Pharr PN, Lavenburg KR, Jackers P, Papas TS, Ogawa M and Watson DK. (2000). Mol. Cell. Biol. 20, 5643-5652. MEDLINE
Starck J, Doubeikovski A, Sarrazin S, Gonnet C, Rao G, Skoultchi A, Godet J, Dusanter-Fourt I and Morle F. (1999). Mol. Cell. Biol. 19, 121-135. MEDLINE
Tamir A, Howard J, Higgins RR, Li YJ, Berger L, Zacksenhaus E, Reis M and Ben-David Y. (1999). Mol. Cell. Biol. 19, 4452-4464. MEDLINE
Topol LZ, Tatosyan AG, Ascione R, Thompson DM, Blair DG, Kola I and Seth A. (1992). Cancer Lett. 67, 71-78. MEDLINE
Watson DK, Ascione R and Papas TS. (1990). Crit. Rev. Oncog. 1, 409-486. MEDLINE
Watson DK, McWilliams-Smith MJ, Flordellis CS and Papas TS. (1988). Advances in gene Technology: Molecular biology of the Endocrine System. Puett D, Ahmad F, Black S, Lopez DM, Melner MH, Scott WA and Whelan WJ. (eds). Cambridge University Press: Cambridge, p., 158.
Watson DK, McWilliams-Smith MJ, Kozak C, Reeves R, Gearhart J, Nunn MF, Nash W, Fowle JRD, Duesberg P, Papas TS and et al. (1986). Proc. Natl. Acad. Sci. USA 83, 1792-1796. MEDLINE
Watson DK, McWilliams-Smith MJ, Nunn MF, Duesberg PH, O'Brien SJ and Papas TS. (1985). Proc. Natl. Acad. Sci. USA 82, 7294-7298. MEDLINE
Yamada T, Kondoh N, Matsumoto M, Yoshida M, Maekawa A and Oikawa T. (1997). Blood 89, 1383-1393. MEDLINE
Yuan CC, Kan N, Dunn KJ, Papas TS and Blair DG. (1989). J. Virol. 63, 205-215. MEDLINE
Zhang L, Eddy A, Teng YT, Fritzler M, Kluppel M, Melet F and Bernstein A. (1995). Mol. Cell. Biol. 15, 6961-6970. MEDLINE
Zhang L, Lemarchandel V, Romeo PH, Ben-David Y, Greer P and Bernstein A. (1993). Oncogene 8, 1621-1630. MEDLINE
Figure 1 Mechanisms of retrovirus-induced oncogene activation. (a) Retroviral capture of a cellular oncogene. The structure of the virus and a hypothetical cellular oncogene locus containing three exons () are shown. In the virus, the open boxes are the unique 3' and 5' sequences, and the positions of the viral gag, pol, and env genes, and the transcriptional orientation () are indicated. In this case, proviral integration is shown as occurring just 5' of the first exon of the cell gene, in the same transcriptional orientation as the gene itself. The hypothetical defective retrovirus formed after multiple genetic steps contains portions of the gag and env genes, and has deleted portions of exons 1 and 3. (b) Retroviral insertional activation of a cellular locus. The cell and viral structures are as in (a), with (1.) and (2.) showing two possible outcomes, depending on the orientation of the virus integration. In (1.) the virus is integrated in the opposite transcriptional orientation to the cell gene, and the viral enhancer is shown activating a cellular promoter to generate the oncogene message. In (2.) the virus has integrated in the same transcriptional orientation as the cell gene. In the first instance, splicing from a viral splice donor to a splice acceptor within the first coding exon occurs, generating a fused message. In the second case, deletion of the 3' end of the virus allows the read-through generation of a fusion message. The messages (¾) and splices (........) are indicated
Figure 2 Structure of the E26 virus and its gag-myb-ets fusion oncogene. The structure of the integrated E26 provirus, together with the resulting virus message and transcribed protein are shown schematically. Different regions of the virus and encoded protein are shown by differences in fills and shading, and by the labels indicating the regions derived from viral gag (gag [left to right hatched box]) and env (env [shaded box]) sequences, as well as myb (mybE [right to left hatched box]) and ets (v-ets [cross hatched box]). In the viral RNA the positions of mutations (relative to the cellular genes) are indicated by ([black box]) and in the protein by (*). DBD and T.D. indicate the positions of the DNA binding and transactivation domains respectively, while ([horizontal line box] [vertical line box]) represents the retroviral LTR and (=) the cell DNA flanking the provirus
|18 December 2000, Volume 19, Number 55, Pages 6472-6481|
|Table of contents Previous Article Next [PDF]