Truncation of the c-myb gene by a retroviral integration in an interleukin 3-dependent myeloid leukemia cell line

Regulation of the differentiation of WEHI-3B D+ leukemia cells by granulocyte colony-stimulating factor receptor

To investigate the role of the G-CSF receptor (G-CSFR) in mediating the action of G-CSF, WEHI-3B D+ murine myelomonocytic leukemia cells were transfected with a plasmid containing the murine G-CSFR gene. Overexpression of G-CSFR in transfected clones was demonstrated by northern blotting, binding of [125I]rhG-CSF and cross-linking experiments. A high level of expression of the G-CSFR did not promote or suppress cellular proliferation or initiate differentiation; however, exposure of transfected cells to G-CSF in suspension culture caused a large percentage of the population to enter a differentiation pathway, as determined by two markers of the mature state, the ability of cells to reduce nitroblue tetrazolium (NBT) and to express the differentiation antigen Mac-1 (CD11b) on the cell surface. Thus, upon treatment with 10 ng/ml of G-CSF, 60% or more of transfected cells exhibited NBT positivity; whereas, in contrast, nontransfected cells exhibited only 6% NBT positivity in response to G-CSF. An eightfold increase in Mac-1 expression over that of the parental line was also observed in transfected cells exposed to G-CSF. The growth rate of the transfected clones was decreased by exposure to G-CSF, presumably due to terminal differentiation. The findings suggest that the predominant function of G-CSF and its receptor in WEHI-3B D+ cells is to mediate differentiation and that the level of the G-CSFR portion of the signal transduction mechanism in this malignant cell line is important for a response to the maturation inducing function of the cytokine.

New sites of proviral integration associated with murine promonocytic leukemias and evidence for alternate modes of c-myb activation

Murine promonocytic leukemias involving insertional mutagenesis of the c-myb locus can be induced by replication-competent retroviruses. In previously studied promonocytic leukemic cells induced by Moloney murine leukemia virus (called MML), the provirus has been invariably integrated upstream of exons 3 or 4 and the leukemic cells expressed aberrant RNAs with fused virus-myb sequences. Furthermore, Myb expressed by these cells has been shown to be truncated by 47 or 71 amino acids. The present report examines the mechanisms of myb activation in leukemias induced by two other retroviruses, amphotropic virus 4070A and Friend strain FB29 (the leukemias are called AMPH-ML and FB-ML, respectively). This study revealed two additional c-myb proviral insertion sites in these promonocytic leukemias. One FB-ML had a proviral integration in exon 9, and expressed a C-terminally truncated Myb protein of 47 kDa similar to that previously demonstrated to be expressed in the myelomonocytic cell lines NFS60 and VFL-2. However, a sequence of reverse-transcribed and amplified RNA from this leukemia demonstrated that the truncation involved a loss of 248 amino acids compared with a loss of 240 amino acids in the myelomonocytic cell lines. Another leukemia had a provirus integrated in the 5' end of c-myb upstream of exon 2 (in the first intron) and produced a Myb protein that was indistinguishable on sodium dodecyl sulfate-polyacrylamide gel electrophoresis from normal Myb. This latter leukemia (FB-ML R1-4-10) expressed Myb with the smallest N-terminal truncation observed so far in promonocytic leukemias; translation begins at an ATG within c-myb exon 2, leading to loss of only 20 amino acids from the N terminus. Unlike the proteins produced in Moloney murine leukemia virus-induced promonocytic leukemias (MML) that have larger truncations, this protein has an intact DNA binding region and does not contain N-terminal amino acids encoded by gag. However, this protein is similar to all N-terminally truncated Mybs so far studied, in that the truncation resulted in deletion of a casein kinase II phosphorylation site which has been proposed to be involved in regulation of DNA binding.

Transformation of chicken myelomonocytic cells by a retrovirus expressing the v-myb oncogene from the long terminal repeats of avian myeloblastosis virus but not Rous sarcoma virus

To test the effect of long terminal repeat (LTR) regulatory sequences on the transforming capability of the v-myb oncogene from avian myeloblastosis virus (AMV), we have constructed replication-competent avian retroviral vectors with nearly identical structural genes that express v-myb from either AMV or Rous sarcoma virus (RSV) LTRs. After transfection into chicken embryo fibroblasts, virus-containing cell supernatants were used to infect chicken myelomonocytic target cells from preparations of 16-day-old embryonic spleen cells. Both wild-type AMV and the virus expressing v-myb from AMV LTRs (RCAMV-v-myb) were able to transform the splenocyte cultures into a population of immature myelomonocytic cells. The transformed cells expressed the p48v-Myb oncoprotein and formed compact foci when grown in soft agar. In contrast, the virus expressing v-myb from RSV LTRs (RCAS-v-myb) was repeatedly unable to transform the same splenocyte cells, despite being able to infect fibroblasts with high efficiency. This difference in the transforming activities of v-myb-expressing viruses with different LTRs most likely results from the presence of a factor (or factors) within the appropriate myelomonocytic target cell that promotes specific expression from the AMV but not from the RSV LTR.

Production of multiple growth factors by a newly established human thyroid carcinoma cell line

A multiple growth factor‐producing tumor cell line (NIM‐1) was newly established from a patient with thyroid cancer and remarkable neutrophilia. NIM‐1 cells also caused severe neutrophilia in nude mice bearing tumors. NIM‐1‐conditioned medium (NIM‐1CM) contained activities that supported not only granulocyte, macrophage and eosinophil colony formation of human bone marrow cells but also the growth of colony‐stimulating factor (CSF)‐dependent cell lines, NFS60‐KX and TF‐1. Northern blot hybridization analysis revealed the constitutive expression of granulocyte‐CSF (G‐CSF), granulocyte/macrophage‐CSF (GM‐CSF) and interleukin(IL)‐6 mRNAs in NIM‐1 cells. Enzyme‐linked immunosorbent assays (ELISA) using NIM‐1CM also confirmed the production of IL‐la and a small amount of IL‐1β besides G‐CSF, GM‐CSF and IL‐6 in NIM‐1 cells. In addition, unexpected production of IL‐11 in NIM‐1 cells was detected by northern blot hybridization analysis and by bioassay using an IL‐11‐dependent cell line. Therefore, NIM‐1 cell line is shown to produce multiple cytokines including potentially megakaryopoietic growth factors such as GM‐CSF, IL‐6 and IL‐11.

Transactivation and transformation by Myb are negatively regulated by a leucine-zipper structure

The negative regulatory domain of the c-myb protooncogene product (c-Myb) normally represses transcriptional activation by c-Myb. We show here that a leucine-zipper structure is a component of the negative regulatory domain, because its disruption markedly increases both the transactivating and transforming capacities of c-Myb. We also demonstrate that this leucine-zipper structure can interact with cellular proteins. Our results suggest that an inhibitor that suppresses transactivation binds to c-Myb through the leucine zipper and that c-Myb can be oncogenically activated by missense mutation.

Myb and Ets proteins cooperate in transcriptional activation of the mim-1 promoter

In the generation of the acutely transforming avian retrovirus E26, both myb and ets genes have been transduced, leading to the production of a Gag-Myb-Ets fusion protein. This co-occurrence of v-myb and v-ets oncogenes suggests that the two might have a functional relationship. To look for such a relationship, we tested the transcriptional activation activity of Myb alone or with coexpressed Ets-1 or Ets-2. Using the promoter of the v-Myb-inducible mim-1 gene as a target, we found that full-length c-Myb gene products were poor activators of transcription, while an oncogenic (truncated) form of this protein was a strong trans-activator. However, coexpression of Ets-2 with full-length or truncated forms of Myb greatly increased trans-activation. Coexpression of Ets-1, Fos, Jun, or Myc with Myb did not increase trans-activation of the mim-1 promoter. The ability of Myb and Ets-2 to transactivate was cooperative, since Ets-2 alone gave little or no activation. Bacterially synthesized Ets-2 protein was found to bind specifically to the mim-1 promoter, suggesting that it may be a target for both Myb and Ets proteins. Thus, Myb and Ets proteins can cooperate in transcriptional activation, and their co-occurrence in the E26 virus may reflect a functional relationship between these two oncoproteins. Truncated forms of Myb may have a reduced need for cooperating factors such as Ets-2, and this might constitute an important mechanism associated with oncogenic activation.

Expression of the Evi-1 zinc finger gene in 32Dc13 myeloid cells blocks granulocytic differentiation in response to granulocyte colony-stimulating factor

Expression of the Evi-1 gene is frequently activated in murine myeloid leukemias by retroviral insertions immediately 5' or 90 kb 5' of the gene. The Evi-1 gene product is a nuclear, DNA-binding zinc finger protein of 145 kDa. On the basis of the properties of the myeloid cell lines in which the Evi-1 gene is activated, it has been hypothesized that its expression blocks normal differentiation. To explore this proposed role, we have constructed a retrovirus vector containing the gene and examined its effects on an interleukin-3-dependent myeloid cell line that differentiates in response to granulocyte colony-stimulating factor (G-CSF). Expression of the Evi-1 gene in these cells did not alter the normal growth factor requirements of the cells. However, expression of the Evi-1 gene blocked the ability of the cells to express myeloperoxidase and to terminally differentiate to granulocytes in response to G-CSF. This effect was not due to altered expression of the G-CSF receptor or to changes in the initial responses of the cells to G-CSF. These results support the hypothesis that the inappropriate expression of the Evi-1 gene in myeloid cells interferes with the ability of the cells to terminally differentiate.

Expression of the Evi-1 zinc finger gene in 32Dc13 myeloid cells blocks granulocytic differentiation in response to granulocyte colony-stimulating factor

Expression of the Evi-1 gene is frequently activated in murine myeloid leukemias by retroviral insertions immediately 5' or 90 kb 5' of the gene. The Evi-1 gene product is a nuclear, DNA-binding zinc finger protein of 145 kDa. On the basis of the properties of the myeloid cell lines in which the Evi-1 gene is activated, it has been hypothesized that its expression blocks normal differentiation. To explore this proposed role, we have constructed a retrovirus vector containing the gene and examined its effects on an interleukin-3-dependent myeloid cell line that differentiates in response to granulocyte colony-stimulating factor (G-CSF). Expression of the Evi-1 gene in these cells did not alter the normal growth factor requirements of the cells. However, expression of the Evi-1 gene blocked the ability of the cells to express myeloperoxidase and to terminally differentiate to granulocytes in response to G-CSF. This effect was not due to altered expression of the G-CSF receptor or to changes in the initial responses of the cells to G-CSF. These results support the hypothesis that the inappropriate expression of the Evi-1 gene in myeloid cells interferes with the ability of the cells to terminally differentiate.

Continuous activation of primitive hematopoietic cells in long-term human marrow cultures containing irradiated tumor cells

Human hematopoietic cells can be maintained in vitro for many weeks in the absence of exogenously provided hematopoietic growth factors if an adequate stromal cell containing adherent layer is present. We have now extended the use of this type of long-term culture (LTC) system to create a model of perturbed hematopoiesis in which human tumor cells that constitutively produce a variety of factors are co-cultured together with normal human marrow cells. In the present study, we used the human bladder carcinoma cell line (5637) because these cells were known to produce not only a variety of factors active directly on hematopoietic cells but also factors that can stimulate hematopoietic growth factor production by human marrow stromal cells. Analysis of mRNA extracted from the adherent layer and measurement of growth factor bioactivity in the medium of established LTC of human marrow containing irradiated 5637 cells, showed increased levels of interleukin-1 and -6, as well as granulocyte and granulocyte-macrophage colony-stimulating factor production by comparison to control cultures. As in normal cultures, high proliferative potential clonogenic hematopoietic cells were found almost exclusively in the adherent layer of these co-cultures, but these primitive cells were maintained in a state of continuous turnover, in contrast to control cultures where the same cell types showed the expected oscillation between a quiescent and a proliferating state following each weekly change of the medium. A similar perturbation of primitive progenitor cycling was achieved by adding medium conditioned by 5637 cells twice a week to otherwise normal LTC. The presence of irradiated 5637 cells in the LTC or the addition of 5637 conditioned medium also resulted in modest (2- to 3-fold) but sustained increases in the total hematopoietic progenitor population, as well as in the final output of terminally differentiated granulocytes and macrophages. These findings indicate that primitive hematopoietic cells in LTC can be kept in a state of continuous activation for many weeks by appropriate endogenous or exogenous hematopoietic growth factor provision and that this does not necessarily lead either to their rapid exhaustion or to a large amplification in output of mature progeny.

Differential effects of microenvironmentally presented interleukin 3 versus soluble growth factor on primitive human hematopoietic cells

The effect of IL-3 on hematopoiesis in long-term culture (LTC) was studied by cocultivating normal human marrow cells with human marrow fibroblast feeders engineered to constitutively produce IL-3 and by adding soluble IL-3 to LTC according to a variety of dose-time schedules. Feeders stably producing 7 ng/ml IL-3, or LTC to which 10 ng/ml IL-3 was added daily for 5 wk, but not once or twice weekly for the same time period, increased the output of mature nonadherent cells and progenitors from LTC as compared to control cultures. At the time of the weekly half-medium change, when primitive clonogenic progenitors in the adherent layer of standard LTC are quiescent, such cells were actively cycling in cultures containing a continuous source of an adequate dose of IL-3. In LTC, where the proportion of IL-3-producing cells in the feeder layer was diluted to 10% and no IL-3 was detectable in culture medium, primitive adherent layer progenitors were, nevertheless, maintained as a population of continuously proliferating cells. Thus, the presence of IL-3 in LTC can enhance the proliferation and differentiation of very early human hematopoietic cells, but the concentration, duration of exposure, and method of IL-3 presentation are important determinants of the ultimate effects observed.

Nucleotide preferences in sequence-specific recognition of DNA by c-myb protein

Using a binding site selection procedure, we have found that sequence-specific DNA-binding by the mouse c-myb protein involves recognition of nucleotides outside of the previously identified hexanucleotide motif. Oligonucleotides containing a random nucleotide core were immunoprecipitated in association with c-Myb, amplified by the Polymerase Chain Reaction and cloned in plasmids prior to sequencing. By alignment of sequences it was apparent that additional preferences existed at each of three bases immediately 5' of the hexanucleotide consensus, allowing an extension of the preferred binding site to YGRCVGTTR. The contributions of these 5' nucleotides to binding affinity was established in bandshift analyses with oligonucleotides containing single base substitutions; in particular, it was found that replacement of the preferred guanine at position -2 with any other base greatly reduced c-Myb binding. We found that the protein encoded by the related B-myb gene bound the preferred c-Myb site with similar affinity; however, B-Myb and c-Myb showed distinct preferences for the identity of the nucleotide at position -1 relative to the hexanucleotide consensus. This study demonstrates that the c-Myb DNA-binding site is more extensive than recognised hitherto and points to similar but distinct nucleotide preferences in recognition of DNA by related Myb proteins.

Acute myeloid leukemia induction by amphotropic murine retrovirus (4070A): clonal integrations involve c-myb in some but not all leukemias

Amphotropic murine retrovirus 4070A was demonstrated to be highly leukemogenic when inoculated intravenously into adult DBA/2 mice that were undergoing an intense chronic inflammatory response, but was nonleukemogenic in the absence of inflammation. The virus-induced promoonocytic leukemias, designated AMPH-ML, are similar morphologically and in cell surface marker expression to monocytic leukemias, called MML and MF-ML, previously shown to be induced by Moloney murine leukemia virus and MF-3 virus (a recombinant between Friend murine leukemia virus and Moloney murine leukemia virus) and resemble certain mature acute monocytic leukemias in humans (AML subtype M5). Approximately two-thirds of the AMPH-MLs (subgroup I) were demonstrated to have alterations in the 5' end of the c-myb locus, an event which occurs in 100% of MML and MF-ML. Data indicate that proviral insertions in AMPH-ML subgroup I resulted in aberrant c-myb mRNA expression and truncation of its translation product at the amino terminus. Approximately one-third of the AMPH-MLs (subgroup II) had not undergone any DNA rearrangements at the c-myb locus. In addition, their transcripts and protein products were of normal size. These latter leukemias also had not undergone DNA rearrangements in c-myc, although retroviruses expressing myc have previously been shown to induce monocyte-macrophage tumors in mice undergoing a chronic inflammation. That subgroup II leukemias had at least one clonal viral insertion suggests that there may be other sites in the cellular genome that can be activated by insertional mutagenesis in these murine acute monocytic leukemias.

Identification of a common viral integration region in Cas-Br-E murine leukemia virus-induced non-T-, non-B-cell lymphomas

The Cas-Br-E murine leukemia virus is a nondefective retrovirus that induces non-T-, non-B-cell lymphomas in susceptible NIH/Swiss mice. By using a DNA probe derived from Cas-Br-E provirus-flanking sequences, we identified a DNA region, originally called Sic-1, rearranged in 16 of 24 tumors analyzed (67%). All proviruses were integrated in a DNA segment smaller than 100 bp and were in the same 5'-to-3' orientation. Ecotropic as well as mink cell focus-forming virus types were found integrated in that specific DNA region. On the basis of Southern blot analysis of somatic cell hybrids and progeny of an interspecies backcross, the Sic-1 region was localized on mouse chromosome 9 near the previously described proto-oncogenes or common viral integration sites: Ets-1, Cbl-2, Tpl-1, and Fli-1. Restriction map analysis shows that this region is identical to the Fli-1 locus identified in Friend murine leukemia virus-induced erythroleukemia cell lines and thus may contain sequences also responsible for the development of mouse non-T-, non-B-cell lymphomas.

Constitutive production of granulocyte colony-stimulating factor by hybrids of a SV40-transformed mouse macrophage and a renal adenocarcinoma cell line

Mouse macrophage BAM3 cells produced colony-stimulating factors (CSFs) after stimulation with bacterial lipopolysaccharide (LPS). By assaying the CSF using various interleukin 3-dependent cell lines, it was shown that most of the CSFs produced by BAM3 cells were granulocyte CSF (G-CSF). The granulocyte-macrophage CSF (GM-CSF) gene was also expressed in BAM3 cells after stimulation with LPS. When BAM3 cells were fused with the mouse renal adenocarcinoma cell line RAG which does not produce G-CSF, two of four hybrid cell lines constitutively produced large quantities of G-CSF. About 300 bp of the promoter region of mouse G-CSF chromosomal gene was inserted upstream of the Escherichia coli chloramphenicol acetyltransferase gene, and introduced into BAM3, RAG and hybrid cells. The G-CSF promoter was activated by stimulation with LPS, in BAM3 cells, but was inert in RAG cells. On the other hand, there was significant constitutive CAT activity in the hybrid cells.

Granulocyte colony-stimulating factor and its receptor

Granulocyte colony-stimulating factor (G-CSF) is a glycoprotein of Mr of about 20,000, which stimulates proliferation and differentiation of progenitor cells of neutrophils. Recent clinical application of G-CSF has proven that this hormone is effective in treatment of patients suffering from neutropenia. In the last few years, the biochemical and molecular nature of the G-CSF receptor has been characterized. The G-CSF receptor is a glycoprotein of Mr 100-130,000, and is expressed on the cell surface of various myeloid cells. A homodimer of this polypeptide can bind G-CSF with a high affinity, and transduce G-CSF-triggered growth signals into cells. Its extracellular domain contains a sequence of about 200 amino acids which can be found in various cytokine receptors. In addition, it contains an immunoglobulin-like domain and three fibronectin type III domains. The overall structure of the beta-chain (gp130) of the interleukin 6 receptor was found to be very similar to that of the G-CSF receptor.

Insertional mutagenesis: neoplasia arising from retroviral integration

The integration of retroviral proviruses near cellular genes can profoundly affect their expression. Painstaking analysis of insertion sites from a large number of tumors has revealed a number of previously unknown proto-oncogenes, and has elucidated new mechanisms whereby known proto-oncogenes can be activated. A number of these genes have been implicated in tumors of clinical relevance. At the time of writing a great deal remains to be learned of the normal function of these genes in the cell. While it has yet to be demonstrated that retroviral insertion mechanisms play some role in naturally occurring human neoplasms, they must be considered in the context of retroviral gene therapy protocols now being contemplated.

Role of tryptophan repeats and flanking amino acids in Myb-DNA interactions

The c-myb protooncogene codes for a sequence-specific DNA-binding protein that appears to act as a transcriptional regulator and is highly conserved through evolution. The DNA-binding domain of Myb has been shown to contain three imperfectly conserved repeats of 52 amino acids that constitute the amino-terminal end. Within each repeat, there are three tryptophans that are separated by 18 or 19 amino acids and are flanked by basic amino acids. To determine the role of tryptophans and the flanking basic amino acids in the DNA-binding activity of Myb proteins, we have selectively mutagenized individual tryptophans as well as some of the amino acid residues that flank these tryptophans. Replacement of these tryptophans with glycine, proline, or arginine abolished the DNA-binding activity whereas replacement with other aromatic amino acids or leucine or alanine did not appreciably affect this activity. On the other hand the replacement of two amino acids, asparagine and lysine, that flank the last tryptophan with acidic amino acids completely abolished their DNA-binding activity. These results are consistent with a model we present in which the tryptophans form a hydrophobic scaffold that plays a crucial role in maintaining the helix-turn-helix structure of the DNA binding domain. Basic and polar amino acids adjacent to these tryptophans seem to participate directly in DNA binding.

Evi-2, a common integration site involved in murine myeloid leukemogenesis

BXH-2 mice have the highest incidence of spontaneous retrovirally induced myeloid leukemia of any known inbred strain and, as such, represent a valuable model system for identifying cellular proto-oncogenes involved in myeloid disease. Chronic murine leukemia viruses often induce disease by insertional activation or mutation of cellular proto-oncogenes. These loci are identified as common viral integration sites in tumor DNAs. Here we report on the characterization of a novel common viral integration site in BXH-2 myeloid leukemias, designated Evi-2. Within the cluster of viral integration sites that define Evi-2, we identified a gene that has the potential for encoding a novel protein of 223 amino acids. This putative proto-oncogene possesses all of the structural features of a transmembrane protein. Within the transmembrane domain is a "leucine zipper," suggesting that Evi-2 is involved in either homopolymer or heteropolymer formation, which may play an important role in the normal functioning of Evi-2. Interestingly, the human homolog of Evi-2 has recently been shown to be tightly linked to the von Recklinghausen neurofibromatosis locus, suggesting a role for Evi-2 in human disease as well.

Evi-2, a common integration site involved in murine myeloid leukemogenesis

BXH-2 mice have the highest incidence of spontaneous retrovirally induced myeloid leukemia of any known inbred strain and, as such, represent a valuable model system for identifying cellular proto-oncogenes involved in myeloid disease. Chronic murine leukemia viruses often induce disease by insertional activation or mutation of cellular proto-oncogenes. These loci are identified as common viral integration sites in tumor DNAs. Here we report on the characterization of a novel common viral integration site in BXH-2 myeloid leukemias, designated Evi-2. Within the cluster of viral integration sites that define Evi-2, we identified a gene that has the potential for encoding a novel protein of 223 amino acids. This putative proto-oncogene possesses all of the structural features of a transmembrane protein. Within the transmembrane domain is a "leucine zipper," suggesting that Evi-2 is involved in either homopolymer or heteropolymer formation, which may play an important role in the normal functioning of Evi-2. Interestingly, the human homolog of Evi-2 has recently been shown to be tightly linked to the von Recklinghausen neurofibromatosis locus, suggesting a role for Evi-2 in human disease as well.

Transformation of murine bone marrow cells with combined v-raf-v-myc oncogenes yields clonally related mature B cells and macrophages

Murine bone marrow cells infected with replication-defective retroviruses containing v-raf alone or v-myc alone yielded transformed pre-B cell lines, while a retroviral construct containing both v-raf and v-myc oncogenes produced clonally related populations of mature B cells and mature macrophages. The genealogy of these transformants demonstrates that mature myeloid cells were derived from cells with apparent B-lineage commitment and functional immunoglobulin rearrangements. This system should facilitate studies of developmental relationships in hematopoietic differentiation and analysis of lineage determination.

Transformation of murine bone marrow cells with combined v-raf-v-myc oncogenes yields clonally related mature B cells and macrophages

Murine bone marrow cells infected with replication-defective retroviruses containing v-raf alone or v-myc alone yielded transformed pre-B cell lines, while a retroviral construct containing both v-raf and v-myc oncogenes produced clonally related populations of mature B cells and mature macrophages. The genealogy of these transformants demonstrates that mature myeloid cells were derived from cells with apparent B-lineage commitment and functional immunoglobulin rearrangements. This system should facilitate studies of developmental relationships in hematopoietic differentiation and analysis of lineage determination.

Characterization of the v-myb DNA binding domain

The transforming protein encoded by the v-myb oncogene is a sequence-specific DNA-binding protein that is thought to be involved in the regulation of gene expression. The N-terminal region of the v-myb protein is composed of two highly conserved tandem repeat sequences of unknown function. It has been speculated that the N-terminal v-myb repeats might be crucial for DNA-binding, since N-terminal deletions destroy the DNA-binding activity of the v-myb protein. Here, we have studied the v-myb DNA-binding domain in more detail. Our results show that the N-terminal region of the v-myb protein is sufficient for specific DNA-binding. Dissection of this region suggests that both repeats are required for DNA-binding, but that both repeats play different roles in v-myb protein DNA interaction. We also show that the myb repeats of a drosophila melanogaster homolog of c-myb function as sequence-specific DNA-binding domain. Our results support the view that specific sequence-recognition, mediated by the conserved myb repeats, is a general feature of myb-related proteins.

Effects of 5' and 3' truncations of the myb gene on the transforming ability of avian myeloblastosis virus (AMV)

Proviruses based on the avian myeloblastosis virus (AMV) have been constructed which code for variations of the c-myb and/or v-myb gene product. These proviruses have been used in a soft colony agar assay to assess the contributions of the 5' and 3' deletions of the v-myb oncogene in the cellular transforming activity of the virus. The results indicate that 3' truncations are an integral part of the gene's mechanism of activation and that the truncations on the 5' end of the gene are important either in its mechanism of activation or its expression by viral control elements.

Identification, nuclear localization, and DNA-binding activity of the zinc finger protein encoded by the Evi-1 myeloid transforming gene

Activation of the Evi-1 zinc finger gene is a common event associated with transformation of murine myeloid leukemias. To characterize the gene product, we developed antisera against various protein domains. These antisera primarily detected a 145-kilodalton nuclear protein that bound double-stranded DNA. Binding was inhibited by chelating agents and partially restored by zinc ions.

Identification, nuclear localization, and DNA-binding activity of the zinc finger protein encoded by the Evi-1 myeloid transforming gene

Activation of the Evi-1 zinc finger gene is a common event associated with transformation of murine myeloid leukemias. To characterize the gene product, we developed antisera against various protein domains. These antisera primarily detected a 145-kilodalton nuclear protein that bound double-stranded DNA. Binding was inhibited by chelating agents and partially restored by zinc ions.

Identification and mapping of a common proviral integration site Fli-1 in erythroleukemia cells induced by Friend murine leukemia virus

Friend murine leukemia virus (F-MuLV) induces erythroleukemia when inoculated into newborn BALB/c or NIH/Swiss mice. We have molecularly cloned F-MuLV host cell DNA junction fragments from an erythroleukemia cell line induced by F-MuLV to identify cellular genes involved in the leukemogenic process. One particular proviral integration site, Fli-1, is rearranged in 75% (9/12) of independently isolated erythroleukemia cell lines derived from either BALB/c or NIH/Swiss mice inoculated at birth with F-MuLV. Other hematopoietic neoplasms induced by F-MuLV, including myeloid (granulocytic) and lymphoid tumors, did not show rearrangements of the Fli-1 locus. Similarly, none of 35 erythroleukemia cell lines induced by the Friend virus complexes (FV-A and FV-P) was rearranged at the Fli-1 locus. In contrast, no rearrangements were detected at the Sfpi-1 locus, a preferred site of integration in either FV-P- or FV-A-induced leukemias. Using recombinant inbred mice, the Fli-1 locus was situated on mouse chromosome 9 close to the cellular protooncogene c-ets-1. DNA and RNA analysis suggests, however, that Fli-1 is different from ets-1. Thus, Fli-1 appears to define a distinct locus specifically involved in the induction of erythroid leukemias by F-MuLV.

Mutation of the core or adjacent LVb elements of the Moloney murine leukemia virus enhancer alters disease specificity

Transcriptional enhancers of replication-competent mouse C-type retroviruses are potent determinants of the distinct disease-inducing phenotypes of different viral isolates and can also strongly influence the incidence and latent period of disease induction. To study the contribution of individual protein-binding sites to viral pathogenicity, we introduced mutations into each of the known nuclear factor-binding sites in the enhancer region of the Moloney murine leukemia virus and injected viruses with these mutations into newborn NFS mice. All viruses induced disease. Viruses with mutations in both copies of the leukemia virus factor a (LVa) site, leukemia virus factor c (LVc) site, or in just the promoter proximal copy of the glucocorticoid response element (GRE) had a latent period of disease onset and disease specificity indistinguishable from that of the wild-type Moloney virus. Viruses with mutations in two or three of the GREs, in both copies of the leukemia virus factor b (LVb) site, in two of the four nuclear factor 1 (NF1) consensus motifs, or in both copies of the conserved viral core element showed a significant delay in latent period of disease induction. Strikingly, viruses with mutations in the core element induced primarily erythroleukemias, and mutations in the LVb site also resulted in a significant incidence of erythroleukemias. These and other genetic and biochemical studies suggest models for how subtle alterations in the highly conserved structure of mouse C-type retrovirus enhancers can produce a dramatic effect on disease specificity.

Granulocyte-macrophage colony-stimulating factor can stimulate macrophage proliferation via persistent activation of Na+/H+ antiport. Evidence for two distinct roles for Na+/H+ antiport activation

Macrophages respond to a variety of extracellular stimuli which can modulate the proliferation, development, activation and functional activity of these cells. The effects of two such agents, granulocytemacrophage colony-stimulating factor (GM-CSF, which stimulates proliferation) and platelet-activating factor (PAF, which stimulates chemotaxis and bactericidal activity), on cellular signal transduction mechanisms were compared. PAF can stimulate inositol lipid hydrolysis leading to Ca2+ mobilization. GM-CSF on the other hand has no effect on these events. Both agonists do, however, share an ability to activate an amiloride-sensitive Na+/H+ antiport and, furthermore, amiloride analogues are shown to inhibit the proliferative effects of GM-CSF on these cells. Long-term incubations with either PAF or GM-CSF demonstrate that it is only those cells pretreated with the latter which show a persistent activation of the antiport together with a sustained increase in intracellular pH. PAF-treated cells exhibit only a transitory increase in antiport activity, their intracellular pH levels returning to resting levels in spite of the continuous presence of the agonist in the medium. These effects of GM-CSF and PAF on Na+/H+ exchange are observed in both bicarbonate-free and bicarbonate-containing medium. These results lead us to suggest that the Na+/H+ antiport has a role in macrophage proliferation and in the regulation of intracellular pH during the oxidative burst stimulated by PAF and other agonists, and that differential mechanisms whereby this antiport is regulated exist in macrophages.

Characterization of the sequence-specific interaction of mouse c-myb protein with DNA

We have examined parameters that affect sequence-specific interactions of the mouse c-myb protein with DNA oligomers containing the Myb-binding motif (CA/CGTTPu). Complexes formed between these oligomers and in vitro translated c-myb proteins were analysed by electrophoresis on non-denaturing polyacrylamide gels using the mobility-shift assay. By progressive truncation of c-myb coding sequences it was demonstrated that amino acids downstream of a region of three imperfect 51-52 residue repeats (designated R1, R2 and R3), which are located close to the amino terminus of the protein, had no qualitative or quantitative effect on the ability to interact specifically with this DNA motif. However, removal of only five amino acids of the R3 repeat completely abolished this activity. The contribution of individual DNA-binding domain repeats to this interaction was investigated by precisely deleting each individually: it was demonstrated that a combination of R2 and R3 was absolutely required for complex formation while the R1 repeat was completely dispensible. c-myb proteins showed quantitatively greater interaction with oligomers containing duplicated rather than single Myb-binding motif, in particular where these were arranged in tandem. Moreover, it was observed that c-myb protein interacted with these tandem motifs as a monomer. These findings imply that a single protein subunit straddles adjacent binding sites and the implications for c-myb activity are discussed.

Characterization of the sequence-specific interaction of mouse c-myb protein with DNA

We have examined parameters that affect sequence-specific interactions of the mouse c-myb protein with DNA oligomers containing the Myb-binding motif (CA/CGTTPu). Complexes formed between these oligomers and in vitro translated c-myb proteins were analysed by electrophoresis on non-denaturing polyacrylamide gels using the mobility-shift assay. By progressive truncation of c-myb coding sequences it was demonstrated that amino acids downstream of a region of three imperfect 51-52 residue repeats (designated R1, R2 and R3), which are located close to the amino terminus of the protein, had no qualitative or quantitative effect on the ability to interact specifically with this DNA motif. However, removal of only five amino acids of the R3 repeat completely abolished this activity. The contribution of individual DNA-binding domain repeats to this interaction was investigated by precisely deleting each individually: it was demonstrated that a combination of R2 and R3 was absolutely required for complex formation while the R1 repeat was completely dispensible. c-myb proteins showed quantitatively greater interaction with oligomers containing duplicated rather than single Myb-binding motif, in particular where these were arranged in tandem. Moreover, it was observed that c-myb protein interacted with these tandem motifs as a monomer. These findings imply that a single protein subunit straddles adjacent binding sites and the implications for c-myb activity are discussed.

Phenotypes and mechanisms in the transformation of hematopoietic cells

Interleukin 3 (IL-3) is a growth factor that supports the proliferation of early hematopoietic stem cells, as well as cells that are committed to a variety of the myeloid lineages. The mechanisms by which IL-3 functions have been studied through the use of a series of IL-3-dependent cell lines isolated from myeloid leukemias or long-term bone marrow cultures. A variety of studies have implicated tyrosine phosphorylation in IL-3 signal transduction. One of the substrates of phosphorylation is a 140 kDa, IL-3-binding protein that is speculated to be the biologically relevant IL-3 receptor. IL-3, through tyrosine phosphorylation, supports viability and growth through the regulation of transcription of a series of genes including c-myc and c-pim-1. The c-myc gene contributes to viability, in part, by regulating the transcription of the ornithine decarboxylase gene. The role of growth factors in differentiation is less clear. By studying IL-3-dependent myeloid leukemia cell lines, two genes have been identified whose altered expression is associated with blocking the ability of the cells to differentiate. The c-myb gene is a nuclear DNA binding protein that has been implicated in myeloid transformation in a number of systems. The Evi-1 gene is a novel gene of the zinc finger family of transcriptional activators. Possible mechanisms by which these genes interfere with normal differentiation are discussed.

Delineation of three functional domains of the transcriptional activator encoded by the c-myb protooncogene

The c-myb protooncogene encodes a sequence-specific DNA-binding protein (c-Myb) that induces transcriptional activation or repression. We have identified three functional domains of the mouse c-Myb protein that are responsible for DNA binding, transcriptional activation, and negative regulation, respectively. In addition to the DNA-binding domain, which is located near the N terminus, an adjacent region (the transcriptional activation domain) containing about 80 amino acids was found to be essential for transcriptional activation. Deletion of a region spanning about 175 amino acids of the C-proximal portion increased transcriptional activation markedly, revealing that this domain normally represses activation. Differences between the transcriptional activation and repression functions of c-Myb and v-Myb are discussed in the light of these functional domains. Our results suggest that transcriptional activation may be involved in transformation by myb gene products.

Selection of lineage-restricted cell lines immortalized at different stages of hematopoietic differentiation from the murine cell line 32D

Erythropoietin (Epo), granulocyte-macrophage colony-stimulating factor (GM-CSF) and granulocyte colony-stimulating factor- (G-CSF) dependent cell lines have been derived from the murine hematopoietic cell line 32D with a selection strategy involving the culture of the cells in FBS-deprived medium supplemented only with pure recombinant Epo, GM-CSF, or G-CSF. The cells retain the diploid karyotype of the original 32D clone, do not grow in the absence of exogenous growth factor, and do not induce tumors when injected into syngeneic recipients. The morphology of the Epo-dependent cell lines (32D Epo1, -2, and -3) was heterogeneous and evolved with passage. The percent of differentiated cells also was a function of the cell line investigated. Benzidine-positive cells ranged from 1-2% (32D Epo3) to 50-60% (32D Epo1). These erythroid cells expressed carbonic anhydrase I and/or globin mRNA but not carbonic anhydrase II. The GM-CSF- and G-CSF-dependent cell lines had predominantly the morphology of undifferentiated myeloblasts or metamyelocytes, respectively. The GM-CSF-dependent cell lines were sensitive to either GM-CSF or interleukin-3 (IL-3) but did not respond to G-CSF. The G-CSF-dependent cell lines grew to a limited extent in IL-3 but did not respond to GM-CSF. These results indicate that the cell line 32D, originally described as predominantly a basophil/mast cell line, has retained the capacity to give rise to cells which proliferate and differentiate in response to Epo, GM-CSF, and/or G-CSF. These cells represent the first nontransformed cell lines which can be maintained in growth factors other than IL-3 and which differentiate in the presence of physiologic signals. As such, they may represent a model to study the molecular mechanisms underlying the process of hematopoietic differentiation, as well as sensitive targets for bioassays of specific growth factors.

Activation of the Mlvi-1/mis1/pvt-1 locus in Moloney murine leukemia virus-induced T-cell lymphomas

The Mlvi-1/mis-1/pvt-1 locus, located approximately 270 kilobase pairs 3' of the c-myc protooncogene, was originally discovered as a common region of provirus integration in Moloney murine leukemia virus-induced rat T-cell lymphomas. The same locus was shown subsequently to be coamplified with c-myc and to be involved in chromosomal translocations in a variety of human and animal neoplasms. Provirus integration in Mlvi-1 in Moloney murine leukemia virus-induced rat T-cell lymphomas activates the c-myc protooncogene. The studies reported here were aimed to determine whether, in addition to the activation of c-myc, provirus integration affected the expression of other neighboring genes. Provirus integration was shown to occur in three clusters separated by regions of uninterrupted DNA. The proviruses in all three clusters had integrated in a single-transcriptional orientation, and they appeared intact. Systematic hybridization of Mlvi-1 clones to rat, mouse, and human genomic DNA revealed three patches of evolutionarily conserved sequences. Two of them were mapped in regions targeted by the provirus, and the third was mapped immediately 5' to the provirus clusters. A probe derived from the conserved sequences 5' of the integrated proviruses detected a tumor-specific RNA transcript in tumors carrying a provirus in Mlvi-1 or in the neighboring Mlvi-4 and c-myc loci. The highest level of RNA transcript expression, however, was seen in a CD4+ CD8+ tumor cell line that was not carrying a provirus in this region. We conclude that provirus insertion in this region activates both c-myc and another gene that is located in the immediate vicinity of the integrated Mlvi-1 proviruses and may be developmentally regulated in T cells.

Activation of the c-H-ras proto-oncogene by retrovirus insertion and chromosomal rearrangement in a Moloney leukemia virus-induced T-cell leukemia

A rearrangement of the c-H-ras locus was detected in a T-cell line (DA-2) established from a Moloney leukemia virus-induced tumor. This rearrangement was associated with the high-level expression of H-ras RNA and the H-ras gene product, p21. DNA from DA-2 cells transformed fibroblasts in DNA transfection experiments, and the transformed fibroblasts contained the rearranged H-ras locus. The rearrangement involved one allele and was present in tissue from the primary tumor from which the cell line was isolated. Cloning and sequencing of the rearranged allele and comparison with the normal allele demonstrated that the rearrangement was complex and probably resulted from the integration of a retrovirus in the H-ras locus between a 5' noncoding exon and the first coding exon and a subsequent homologous recombination between this provirus and another newly acquired provirus also located on chromosome 7. These events resulted in the translocation of the coding exons of the H-ras locus away from the 5' noncoding exon region to a new genomic site on chromosome 7. Sequencing of the coding regions of the gene failed to detect mutations in the 12th, 13th, 59th, or 61st codons. The possible reasons for the complexity of the rearrangement and the significance of the activation of the H-ras locus to T-cell transformation are discussed.

Murine T cells express a cell surface receptor for multiple extracellular matrix proteins. Identification and characterization with monoclonal antibodies

Cell-cell and cell-extracellular (ECM) protein interactions are mediated through heterodimers termed integrins. We have demonstrated that dendritic epidermal T cell (DETC) lines adhere to the ECM proteins, fibronectin, fibrinogen, and vitronectin but not to collagen, laminin, or control proteins. This adhesion was blocked by the tetrapeptide RGDS, but not the control peptide, RGES. We have derived a hamster mAb H9.2B8, and a rat mAb, 8.18E12, from immunizations with DETC lines. The mAbs in combination, but not individually, specifically inhibited the adhesion of DETC lines to fibronectin, fibrinogen, and vitronectin. Immunoprecipitation analysis revealed that both mAbs reacted with a heterodimer composed of noncovalently linked 140- and 95-kD subunits. The 140-kD subunit can be reduced to 120- and 23-kD fragments. Although the two mAbs did not cross-compete for binding to DETC, sequential immunoprecipitation studies indicated that they react with the same 120-kD fragment. While all DETC cell lines and several T cell clones were reactive with the mAbs, the mAbs were not reactive with normal spleen, lymph node, thymus, or skin. Stimulation of splenic T cells with Con A or allogeneic cells induced mAb reactivity after 1 wk in vitro. These data demonstrate that a single lymphocyte receptor, with biochemical features characteristic of integrins, mediates RGD-dependent binding to the ECM proteins, fibronectin, fibrinogen, and vitronectin. Furthermore, since this integrin is expressed by long-term activated T cells, this receptor may play a physiological role in T cell function.

Detection of granulocyte colony-stimulating factor produced by a newly established human hepatoma cell line using a simple bioassay system

A colony-stimulating factor(CSF)-producing tumor cell line (KX-87) was established from a patient with hepatocellular carcinoma and marked granulocytosis. This cell line formed tumors on nude mice in high frequency and the mice revealed marked granulocytosis. In clonogenic assays of human bone marrow cells, KX-87 conditioned medium (CM) supported the formation of colonies mainly consisting of neutrophilic granulocytes but had no burst-promoting activity. The molecular weight of the colony-stimulating activity (CSA) in KX-87CM was estimated about 25,000 daltons by gel filtration and a new bioassay system. In principle, a subline of murine hemopoietic cell line NFS-60 was cloned which was dependent on KX-87CM. Then the growth of this subline was examined by a rapid and sensitive colorimetric tetrazolium assay. From these results, it was concluded that the CSA which KX-87 cell line produced was G-CSF.

Origins and properties of hematopoietic growth factor-dependent cell lines

Studies of the growth regulation, differentiation and transformation of myeloid cells have been greatly facilitated by the availability of a variety of hematopoietic growth factor-dependent cell lines. These cell lines have been isolated from long-term bone marrow cultures and myeloid tumors using interleukin 3 (IL-3) as a growth factor. Using growth factor-dependent cells, it has been shown that growth regulation by IL-3 involves binding to a high-affinity receptor of 140 Kd and activation of tyrosine phosphorylation. IL-3 binding is associated with a number of cellular responses which are required for maintenance of viability, including induction of transcription of the c-myc and ornithine decarboxylase (ODC) genes. In addition, IL-3 regulates the expression of transcription of the gamma T cell receptor locus. The properties of the IL-3-dependent lines are consistent with the hypothesis that they are transformed in their ability to terminally differentiate. In some of the cell lines, this transformation may terminally differentiate. In other of the cell lines, this transformation may be due to the altered expression of the c-myb gene. In other cell lines, transformation is associated with the activation of the expression of a novel gene, termed Evi-1, of the zinc finger family of transcriptional factors. Comparable transformation of erythroid lineage cells is speculated to be due to the activation of the expression of another novel gene termed spi-1. These studies have emphasized the value of well-characterized hematopoietic growth factor-dependent cell lines in advancing our understanding in the basic biology of myeloid cells.

The pathology of murine myelogenous leukemias

Murine myelogenous leukemias can be classified into several distinct subgroups based on morphology, cytochemical staining, and immunoreactivity. The leukemias invariably involve the spleen and the extent of infiltration into other tissues is variable. The myelogenous nature of the leukemia is readily apparent in well-differentiated leukemias on the basis of morphology; with poorly differentiated leukemias, positive staining with chloroacetate esterase, nonspecific esterase, and certain monoclonal antibodies such as Mac-1, is helpful to establish myelogenous differentiation. Subgrouping of myelogenous leukemias depends on the presence or absence of monocytic differentiation, as ascertained by staining with Mac-2, electron microscopy or phagocytosis. Leukemias showing no monocytic differentiation can be classified as myeloblastic, corresponding to the FAB M1 and M2 subtypes in humans. Leukemias exhibiting both monocytic and granulocytic features are myelomonocytic, corresponding to the FAB M4 subtype. Tumors with only monocyte differentiation arise primarily as solid tumors in mice, and a leukemic phase is variable.

Isolation of human cDNA clones of myb-related genes, A-myb and B-myb

cDNA clones of the myb-related genes A-myb and B-myb were obtained by screening human cDNA libraries. The predicted open reading frame of B-myb could encode a protein of 700 amino acid residues. Although the C-terminal end has not been cloned yet, an almost entire coding region of A-myb, which is 745 amino acid long, was determined. The A-myb and B-myb proteins are highly homologous with the myb protein in three regions. Domain I, which is 161 amino acid long, is well conserved in the myb gene family. The homology between human-myb and A-myb in domain I is 90% at the amino acid level. Domain II, which is about 85 amino acid long, is less well conserved. Although it is a short stretch, domain III is found in the C-terminal region. The mRNAs of A-myb and B-myb were 5.0 and 2.6 kb, respectively. The mRNA expression pattern of the myb gene family in various tumors is presented.

Regulation of proliferation in a murine colony-stimulating factor-dependent myeloid cell line: superinduction of c-fos by the growth inhibitor 8-Br-cyclic adenosine 3':5' monophosphate

We have investigated the effect of 8-Br-cyclic adenosine 3':5' monophosphate (cAMP), a pharmacological activator of cAMP-dependent protein kinase, on the proliferation and the nuclear proto-oncogene induction in a murine granulocyte macrophage colony-stimulating factor (GM-CSF)-dependent myeloid cell line. Cells were growth arrested by granulocyte macrophage colony-stimulating factor and serum deprivation and were allowed to proceed in the cell cycle by addition of the lymphokine in the presence or absence of 8-Br-cAMP. 3H-thymidine incorporation assays showed that addition of 8-Br-cAMP inhibited the entry of cells into S phase and the subsequent proliferation. Northern analysis showed that 8-Br-cAMP had opposite effects on c-fos and c-myc mRNA induction. 8-Br-cAMP induced c-fos in the absence of any GM-CSF. In the presence of GM-CSF, c-fos mRNA was superinduced (30-fold induction compared to four- to fivefold by each signal alone). On the contrary, 8-Br-cAMP was not able to induce c-myc in the absence of growth factor and hardly interfered with the induction of c-myc by GM-CSF. Phorbol myristate acetate (PMA), a pharmacological activator of the lipid and CA++-dependent protein kinase C, was shown to induce nuclear proto-oncogene mRNA in the GM-CSF-dependent cell line. We investigated the effect of 8-Br-cAMP on PMA-induced c-fos and c-myc mRNA levels. When both cAMP dependent and lipid-dependent kinase systems were co-stimulated in the absence of GM-CSF, c-fos message was again superinduced (60-fold induction). On the contrary, c-myc message induction by PMA was inhibited by 80% by coactivation of cAMP-dependent protein kinase with 8-Br-cAMP. Our data indicate that an antiproliferative signal induces or even superinduces c-fos message and hardly interferes with c-myc induction, suggesting that the intracellular pathways resulting in c-fos and c-myc induction may be distinct and that two different pathways can lead to c-fos induction, with synergistic effects when both are activated.