Expression of the human protooncogene c-myc is necessary for replication, and may be inhibited in a sequence-specific, dose-dependent manner by an antisense oligodeoxynucleotide specific for the first five codons of c-myc mRNA. Antisense inhibition of c-myc inhibits the proliferation and enhances the differentiation of the HL-60 human promyelocytic leukemia cell line. In order to raise the efficacy of antisense oligomers, HL-60 cells were grown in a serum-free medium so as to minimize nuclease activity in the culture medium. Daily addition of anti-c-myc oligomer was then found to induce terminal granulocytic differentiation of 80% or more of HL-60 cells, and inhibit colony formation by greater than 50%, comparable to 1% Me2SO.
A series of antisense pentadecamers complementary to a variety of target sequences between the cap and AUG initiation codon regions of c-myc mRNA was synthesized and used to treat human promyelocytic leukemia HL-60 cells. The sensitivity of the cap-region sequences to antisense inhibition of c-myc p65 expression was two to three times that of the original initiation codon antisense sequence. The other target sequences downstream of the cap and up to the AUG initiation codon were comparable to the initiation codon sequence, except that the first splice junction was slightly more sensitive. At the primary initiation codon target, a dodecamer was about half as effective as the original pentadecamer, whereas an octadecamer was about twice as effective. The observation of variation in antisense efficacy as a function of target location in c-myc mRNA may represent a combination of the effects of hybrid arrest, RNase H attack, and interference in RNA processing. Alternatively, the most sensitive targets might be those that are the most exposed in the secondary and tertiary structure of c-myc mRNA.
We have studied the binding of proteins to a synthetic DNA fragment corresponding to a 21 base pair sequence in the upstream region of the human c-myc gene. This sequence has previously been reported to bind the c-myc protein, thus representing a putative site for autoregulation of the c-myc gene (Ariga et al., 1989). Our electrophoretic mobility shift assays indicated that this DNA fragment binds to a factor present in nuclear extracts from HeLa cells. However, analysis of extracts from other cells indicated that this binding does not correlate with the presence of the c-myc protein. No binding was observed using c-myc protein translated in vitro. Also, anti-c-myc antibodies could not inhibit binding of this factor. These experiments indicate, that nucleotides -2193 to -2173 from the P1 promoter of the c-myc gene are able to bind a nuclear factor, and that this factor most likely does not include the c-myc protein.
Malignant transformation of cells in vitro requires the action of cooperating oncogenes. However, the action of a single activated oncogene, if placed under a strong constitutive promoter, is sufficient to transform in vitro established cells. We have investigated the role of cooperating oncogenes in the transformation of normal rat bladder epithelial cells. Effects of polyoma middle T, v-Ha-ras or activated rat c-Ha-ras, independently or in combination with v-myc on in vitro and in vivo behavior of rat bladder epithelial cells were studied. Introduction of polyoma middle T alone, c-Ha-ras or in some cases v-myc into "normal" rat bladder epithelial cells was sufficient for full malignant transformation of these cells. In addition, introduction of polyoma middle T, activated c-Ha-ras or v-myc transformed cells into nude mice, resulted in development of highly invasive adenocarcinomas. These results indicate that full malignant transformation of normal rat bladder epithelial cells did not require the action of introduced cooperating oncogenes.
The myeloid interleukin-3 (IL-3) dependent cell line, FDC-P1, enters the G0 stage of the cell cycle after IL-3 deprivation for 24 hr. The expression of 13 protooncogenes and immediate-early genes was compared with 4 "control" genes after the addition of either IL-3 or phorbol myristate acetate (PMA) to IL-3-deprived cells. mRNA transcripts encoding c-myc and the T-cell receptor c-gamma gene were induced to high levels only after IL-3 addition, whereas c-fos, fos-B, c-jun, jun-B, Krox-20, and Krox-24 were induced transiently only after PMA addition. The remaining genes (fra-1, p53, jun-D, c-Ha-ras, c-Ki-ras, c-raf, beta-actin, ornithine decarboxylase, and histone 2B) were detected after culture with either IL-3 or PMA. When cells were serum- and IL-3-deprived, c-fos, fos-B, c-jun, jun-B, Krox-20, and Krox-24 were detected after exposure to either serum or PMA. Moreover, culture with cycloheximide and PMA resulted in superinduction of these genes, whereas cycloheximide and IL-3 addition did not. mRNAs encoding these genes had half-lives of 10-20 min after PMA treatment, whereas that of beta-actin was longer (greater than 2 hr), suggesting that short mRNA half-lives contributed to the transient nature of these genes. Although c-fos, fos-B, c-jun, jun-B, Krox-20, and Krox-24 expression can be detected in IL-3-dependent cells after exposure to either PMA or serum, these genes were not detected after IL-3 addition, which allows cell-cycle progression. These results document the existence of IL-3 and PMA-responsive genes and demonstrate that IL-3 and protein kinase C agonists can induce distinct patterns of gene expression.
Activating mutations within the various RAS protooncogenes have been detected in many human and murine hematopoietic neoplasms. Their causal significance has been difficult to assess, partly due to the lack of an animal model directly linking these mutations to hematopoietic neoplasms. A high-titer, helper-free recombinant retrovirus was used to introduce an activated Harvey RAS gene under the transcriptional control of the Moloney leukemia virus LTR into murine bone marrow cells. Eleven of fifteen mice reconstituted with these bone marrow cells developed fatal thymic lymphomas 10-12 week post-transplant. Analysis of DNA and RNA from tumor cells revealed the integrated proviral genome and provirally encoded RAS mRNA respectively. Immunophenotyping and T-cell receptor rearrangement analysis of fresh tumor cells and of cell lines derived from these tumors showed them to be T-cells arrested midway through thymic development. Despite evidence of proviral integration in marrow cells of mice with thymic tumors, no other hematologic abnormalities were detected. The short latency and reproducibility of thymic lymphoma development in mice transplanted with marrow transduced with this retrovirus suggests a direct causal effect of expression of an activated RAS gene in the transformation of a bone-marrow-derived lymphoid progenitor.
Three chemically-induced precancerous mammary hyperplasias, independently isolated in BALB/c mice, all contained mouse mammary tumor virus (MMTV) proviral DNA integrated into a common region in chromosomal DNA, designated Int-5 (Formerly Int-H, Gray et al., 1986). This site was cloned from a hyperplastic outgrowth (D2) into lambda phage. A 1.7 Kb Hind III DNA fragment, which flanks the 5' end of the MMTV insert, was generated from the cloned Int-5 region. This fragment was used as probe (IH-2) to localize Int-5 to mouse chromosome 9. The IH-2 sequence was highly conserved in DNA of several mammalian species including man, in three other widely divergent vertebrate phyla, and in C. elegans. The Int-5 region, containing 5.6 Kb 5' and 12.8 Kb 3' to the MMTV integration site in D2, was cloned in EMBL-3 from a BALB/c genomic DNA library. cDNA complementary to poly A+ lactating mammary gland RNA, annealed with Sst-1 fragments spanning most of the BALB/c Int-5 clone. The highest level of Int-5 specific poly A+ mRNA was detected in D2 tumor. Lactating mammary gland and D2 hyperplastic alveolar nodule contained 5-fold less Int-5 RNA while liver contained 8-fold less Int-5 RNA. Int-5 cDNA (IH-3) annealed with two RNA species of approximately 3.3 and 4.0 Kb. These data are consistent with the hypothesis that Int-5 contains an oncogene, different from any other previously described, involved in early events in some models of chemical carcinogenesis.
Mutations of c-Ha-ras at codon 12 were detected from 11 out of 27 fresh tissues and cell-lines of human gastric cancer patients using PCR-restriction analysis. Further statistical investigation showed that the presence of point mutations was related to the distal metastases and the survival time of gastric cancer patients after surgical operations.
Much controversy regarding the effect of interferons in control of the animal cell cycle can be reconciled by acknowledging that, until recently, few laboratories enjoyed access to both pure interferons and pure growth-factor preparations. Using such reagents, we show data that suggest that antimitogenic state in Balb/c-3T3 cells by beta-interferon resembles the antiviral state in two important regards: dosage and kinetics. Picomolar concentrations of homogeneous beta-interferon inhibit the mitogenic response to recombinant platelet-derived growth-factor B chain homodimer. The antimitogenic response is seen only when cells are exposed to pure beta-interferon for an extended time prior to platelet-derived growth-factor treatment. The interferon-mediated antimitogenic state is exerted at some point in the cell cycle subsequent to c-myc gene induction.
Human c-fps/fes when expressed at sufficiently high levels in NIH 3T3 cells can induce cellular transformation. To probe the mechanism of action of the c-fps/fes product NCP92, we have examined the biological and biochemical consequences of stabilizing phosphotyrosine in cells that overexpress NCP92. In this study we report that when cells expressing c-fps/fes are incubated with low concentrations of sodium vanadate, a tyrosine phosphatase inhibitor, the transforming activity of c-fps/fes can be potentiated by nearly two orders of magnitude. This effect was associated with a threefold increase in the level of phosphotyrosine in NCP92 and its major cellular substrates. Unlike c-src, NCP92 had a single tyrosine phosphorylation site, and vanadate treatment induced its phosphorylation in vivo. This was found to have a positive effect on NCP92 kinase specific activity, but at the low concentrations of vanadate that were used, this effect was very small. The results are consistent with the hypothesis that potentiation was a consequence of the stabilization of phosphotyrosine in critical targets of transformation of NCP92, rather than from a direct effect of vanadate on NCP92 kinase. The potentiating effect of vanadate was relatively specific for c-fps/fes; this reagent did not affect the high transforming activity of gag-v-fps/fes, nor did it enhance the less active c-ras, c-src, or EGF receptor genes, although the latter two genes encode tyrosine kinases. The specificity of the biological response of c-fps/fes to the stabilization of phosphotyrosine suggests that this molecule has a distinct mode of regulation and mechanism of action.
A bacterial expression vector containing a segment of the v-abl gene from Abelson murine leukemia virus (A-MuLV) was constructed such that the gag region of v-abl was replaced by a sequence encoding the IgG-binding domain of the S. aureus protein A. pabl HP, a fusion protein encoded by this vector was rapidly purified to near homogeneity by affinity chromatography on IgG-Affigel and Mono Q FPLC. The Km of the pabl HP kinase for ATP varied with [Val5]-angiotensin II concentration and was 21.2 microM at saturating concentrations of [Val5]-angiotensin II. The Km for [Val5]-angiotensin II at saturating concentrations of ATP was 3.8 mM. The turnover number, at 20 degrees C, was 62 mumol min-1 mumol-1. Initial rate studies support a ternary complex kinetic mechanism for phosphoryl transferase. The substrate specificity of the pabl HP kinase was further characterized using synthetic peptides. This expression system, which enables the rapid purification of recombinant v-abl kinase is suitable for the comparative enzymological study of mutant v-abl enzymes generated by site-directed mutagenesis.
Several lines of evidence suggest that phosphorylation of tyrosine residue 416 (Tyr416) has a positive regulatory influence on the functional activity of the protein tyrosine kinase pp60c-src. To further examine the functional importance of phosphorylation at Tyr416, we have eliminated this phosphoacceptor site in four functionally unique mutant forms of the c-src gene product that are phosphorylated on Tyr416 in vivo. Substitution of phenylalanine for Tyr416 suppressed the biological activity of all of the mutant proteins as assayed by colony formation in soft agar and the induction of morphological alterations. However, the extent of this effect, and the degree to which the substitution affected the phosphorylation of substrates in vivo and in vitro, varied considerably in each of the mutants. These results support the notion that phosphorylation of Tyr416 has a positive regulatory effect on the biological activity of c-src; however, this effect does not directly correlate with a general effect on the total level of tyrosine kinase activity in vitro or the level of tyrosine phosphorylation of cellular proteins in vivo.
Using in situ chromosomal hybridization, we have mapped the gene encoding the basic fibroblast growth factor to chromosome 4 at bands q26-q27. This location is on the same chromosome as that of a related gene, FGF.5, but on a distinct region.
The human teratocarcinoma NTERA-2 cl. D1 (NT2/D1) cell is a cloned multipotential embryonal cancer cell line that differentiates into a neuronal phenotype and other cellular lineages with retinoic acid (RA) treatment. Here we report that mRNA for the transforming growth factor-alpha is expressed in these RA-untreated cells and that RA-treatment results in a reduction of mRNA expression within 24 hr of treatment. In total cellular RNA, TGF-alpha mRNA is not detectable by Northern analysis at 6 days when there is increased expression of the human homeotic genes Hu-1 (Hox 2.1) and Hu-2 (Hox 2.2), known markers of RA response in NT2/D1 cells. RA treatment also causes a marked reduction in cloning efficiency and tumorigenicity of these cells. The addition of TGF-alpha or EGF (epidermal growth factor) protein to RA-untreated NT2/D1 cells augments soft agar cloning under limited fetal calf serum conditions. Blocking monoclonal antibodies directed against the EGF receptor (EGFr) can prevent this augmentation. We conclude that TGF-alpha expression inversely correlates with the state of RA-induced differentiation of this human teratocarcinoma cell and that TGF-alpha and EGF proteins are stimulatory growth factors in NT2/D1 cells under these culture conditions.
We have isolated the human homolog (hltk) of the murine tyrosine kinase gene ltk from a K-562 human leukemia cDNA library. The deduced protein sequence of hltk is 17 amino acids longer in the juxtamembrane domain and 28 amino acids shorter in the carboxy terminus than that of murine ltk. The partially identical splicing points of hltk to those of c-ros showed a close genetic linkage between the two. In Northern blot analysis of 35 human malignancies, hltk is preferentially expressed in leukemias (10 out of 18 cases) with no cell lineage specificity, but none of 17 nonleukemic neoplasms expressed hltk gene.
Here we report a survey of c-rel proto-oncogene transcription in murine tissues, cell lines and lymphoid cells. In addition to the previously described 7.5-kb mRNA, we have identified a mRNA of 2.5-kb. As DNA hybridization indicates that there is only one gene with significant homology to c-rel in the mouse genome, it appears that multiple mRNAs are transcribed from c-rel. The nucleotide sequence of a cDNA clone derived from the 2.5-kb c-rel mRNA demonstrates that the 7.5- and 2.5-kb mRNAs encode identical proteins. The different size of the two mRNAs is due to variation in the length of the 3' untranslated region, which arises from the use of alternate polyadenylation signals. These mRNAs are present at low levels in organs tested, and in cell lines representing a wide variety of lineages. Fibroblasts are the only cells in which expression was not detectable. In B-cell lines representing different stages of differentiation, the highest levels of mRNA are seen in B-lymphomas, and this level drops markedly in plasmacytomas. There is a transient increase of 10- to 20-fold in the level of c-rel mRNAs in T-cells treated with concanavalin A, while lipopolysaccharide-stimulated B-cells exhibit a transient 5-fold elevation of c-rel expression. This study indicates that the control of c-rel expression can vary between and within different cell lineages, and the widespread expression of this gene points to a fundamental cellular function, rather than one restricted to hematopoietic cells as previously suggested.
The relationship of antisense oligodeoxynucleotide inhibition to the predicted secondary structure of human c-Ha-ras oncogene mRNA was examined. Three antisense pentadecade-oxynucleotides complementary to the 5' cap region, a predicted loop in the 5' untranslated region, and the initiation codon region of c-Ha-ras mRNA were synthesized. T24-transformed NIH3T3 cells were treated for 24 hr with each anti-ras oligomer or control sequence. The levels of c-Ha-ras p21 antigen were then analyzed by radioimmunoprecipitation. Inhibition of p21 expression was sequence-specific and dose-dependent. The efficacy of the cap sequence in reducing p21 expression was greater than that of the initiation codon target, which in turn was more effective than the target upstream of the initiation codon. Antisense inhibition of p21 expression correlated with inhibition of focus formation.
The chicken cellular c-ets-1 locus encodes for two related proteins generated by alternative splicing: the widely expressed p54c-ets-1 protein and the cellular homolog of the v-ets-encoded domain of the E26-transforming protein P135gag-myb-ets, p68c-ets-1 which has been found so far only in the spleen. We have prepared a new site specific antiserum directed against the amino-terminus of p68c-ets-1, which is highly hydrophobic by contrast to the hydrophilic NH2 terminus of p54c-ets-1. This antiserum specifically immunoprecipitated p68c-ets-1 and P135gag-myb-ets only in denaturing and reducing conditions. Despite these biochemical differences at their amino-terminal parts, p68c-ets-1, as p54c-ets-1, is a nuclear protein able to bind to DNA in vitro. Unlike p54c-ets-1 which is expressed at high levels in T- and B-lymphoid cells, p68c-ets-1 is not expressed in lymphoid cells of the spleen. Thus, the two c-ets-1 encoded proteins, although both exhibiting DNA-binding properties in vitro, display differences both in the nature of their specific NH2 termini, and in their level and pattern of expression.
The protooncogene c-ets-1 is preferentially expressed in lymphoid cells. The protein product of this gene has been found to be a phosphorylated nuclear protein. When lymphocytes are stimulated with calcium ionophore, hyperphosphorylation of c-ets-1 occurs. In order to study the biological and biochemical functions of the c-ets-1 protein in detail, it is important to prepare adequate quantity of the c-ets-1 protein. To this end, we have cloned, sequenced, and expressed mouse c-ets-1 cDNA in baculovirus expression vector. Sequence analysis indicated that mouse c-ets-1 cDNA codes for a 50/51-kd protein. Since the mouse c-ets-1 protein in mouse lymphocytes is a 60/62-kd protein, the result obtained indicated that the c-ets-1 protein undergoes posttranslational modification by phosphorylation. When the c-ets-1 cDNA was expressed in the baculovirus expression vector, insect cells infected with a recombinant virus synthesizes a protein of the same size but with 50-100 times more of the c-ets-1 protein than that of mouse lymphocytes. The Staphylococcus aureus V8 protease mapping analysis of mouse c-ets-1 proteins synthesized in mouse and insect cells showed that they are identical. Thus, the c-ets-1 protein synthesized in insect cells will allow us to purify and study the functions of the c-ets-1 protein in detail.
