Search

In this study, we show that c-fgr proto-oncogene expression is limited to normal peripheral blood granulocytes, monocytes, and alveolar macrophages, all of which contain 50 to 100 copies of c-fgr mRNA per cell. The c-fgr RNA molecules in these cells consisted of partially spliced transcripts containing intron 7 and completely spliced molecules capable of encoding the predicted p55 c-fgr protein. The splicing of intron 7 appeared to occur after the splicing of most of the other introns; partially spliced molecules containing intron 7 did not appear to be transported into the cytoplasm. Very low levels of fgr transcripts were also present in U937 promonocytic cells and increased in abundance with 12-O-tetradecanoylphorbol-13-acetate (TPA)-induced differentiation. The level of fgr transcripts began to increase 2 to 4 h after TPA addition, peaked at 8 h, and subsequently declined. Since we found that the half-life of fgr mRNA was longer than 8 h, these changes are best explained by transient transcriptional activation of fgr during TPA-induced differentiation, although nuclear runoff experiments were not sensitive enough to detect this event. Cycloheximide also caused accumulation of c-fgr transcripts in U937 cells; no superinduction was observed when TPA and cycloheximide were added at the same time. Induction by either agent was blocked with actinomycin D. These results demonstrate that the c-fgr gene is expressed in a tissue- and development-specific fashion and suggest that constitutive expression of c-fgr in U937 cells is regulated by a labile transcriptional repressor.

A somatic mutation(s), acquired during the evolution of preleukemia in a 75-year-old Caucasian male of North European origin, resulted in a marked decrease in alpha-globin mRNA. The small amount of alpha-globin mRNA present in bone marrow cells was normally processed, had a normal (alpha 1/alpha 2)-globin mRNA ratio, and was translated normally. No detectable zeta-globin mRNA was found. The alpha- and zeta-globin genes were both hypomethylated and restriction endonuclease maps of the alpha- and zeta-globin genes were comparable in the patient's marrow and fibroblast DNA. The data are most consistent with the acquisition of a mutation(s) that resulted in decreased expression of all four alpha-globin genes.

Cathepsin G is a 26,000-Da serine protease that is found in the azurophil granules of neutrophils and monocytes. The cathepsin G gene is expressed at high levels in U937 promonocytic cells, but is down-regulated with phorbol-induced differentiation. To characterize the genomic sequences responsible for the regulated expression of this gene, we screened a human genomic fibroblast library using cathepsin G cDNA, and obtained two lambda clones that contained the cathepsin G locus. The cathepsin G gene spans 2.7 kilobase pairs of genomic DNA and consists of 5 exons and 4 introns. The genomic organization of cathepsin G is similar to that of human neutrophil elastase, rat mast cell protease II, murine adipsin, and murine cytotoxic T-cell serine proteases, with protease catalytic residues located near the borders of exons 2, 3, and 5. Using in situ hybridization techniques, we localized cathepsin G to chromosome 14q11.2, a site that is near the alpha/delta T-cell receptor complex. Cathepsin G transcription is abolished in U937 nuclei with 2 micrograms/ml alpha-amanitin, indicating that this gene is probably transcribed by RNA polymerase II. The 5' end of the cathepsin G gene was defined by primer extension and S1 nuclease protection assays. A TATA box is found at position -29, and a CAAT box is found at -69 with respect to the transcription initiation site. Having defined the genomic structure and chromosomal location of cathepsin G, we are now attempting to identify the DNA elements in or near this gene that mediate its tissue and development-specific pattern of expression.

We measured steady-state levels of the human globin mRNAs in liver samples from several mid-gestational fetuses. RNA from the epsilon, gamma, beta, zeta, theta, and alpha globin genes were present in fetal liver samples isolated from 10-25-wk embryos. The abundance of all human globin mRNAs declined in older fetuses, presumably because of a gradual reduction in the proportion of erythroid precursors in the liver as development proceeds. The gamma:beta globin mRNA ratio in 10-18-wk fetal erythroblasts was 6-7:1, and in adult erythroid bone marrow the ratio was 0.02:1. In fetal liver samples, the relative abundance of epsilon transcripts was less than 1% that of gamma, and zeta transcripts less than 5% that of alpha. Embryonic transcripts declined in abundance during late fetal development and were not detected in newborn liver or adult erythroid bone marrow. theta globin mRNA also represented a minor species (less than 1% that of alpha) in fetal liver samples, but in contrast to the embryonic mRNAs, was most abundant in adult marrow samples obtained from patients with erythroid hyperplasia. These results support the hypothesis that globin protein levels are regulated by the relative amounts of each globin mRNA at various stages of erythropoietic development.

The 5' ends of most normal human beta globin mRNA molecules correspond to a single transcription initiation site, often referred to as the "CAP" site (1-4). Using S1 nuclease mapping and primer extension techniques, we have determined that a minority of beta globin gene transcripts are longer at the 5' ends. These longer molecules comprise about 10% of total beta globin RNA molecules in normal human bone marrow cells and in peripheral blood reticulocytes. The long molecules are transcribed only from the sense strand of DNA and are probably spliced correctly. A DNA segment that includes imperfect "CCAAT" and "TATA" promoter-like sequences begins approximately 150 base pairs (bp) upstream from the normal beta globin gene promoter; this "pseudo-promoter" may function in the initiation of the long globin RNA molecules.

We have developed a method that permits rapid identification of the consequences of mutations that alter beta-globin RNA processing in erythroid cells. S1 nuclease mapping techniques were used to analyze total bone marrow RNA obtained fron 15 patients who are clinically homozygous for beta-thalassemia and from 5 patients with erythroid hyperplasia from other causes. This analysis was facilitated by the use of single-stranded uniformly labeled DNA probes of high specific activity that were prepared by using recombinant phage M13-beta-globin DNA templates. Two abnormalities of RNA processing were found to occur with high frequency in these patients. Nine thalassemic patients were found to have increased levels of an RNA species that retains all sequences transcribed from intervening sequence 1, implying the presence of mutations that decrease the correct splicing of this intron. Seven of 15 thalassemic patients were found to have an abnormally processed RNA species that retains 19 nucleotides transcribed from the 3' end of intron 1; this abnormality is caused by the G leads to A substitution in intron 1 that is known to create an alternative splice acceptor site.

5-Azacytidine, a cytidine analog, stimulated fetal hemoglobin synthesis in five patients who had either severe beta-thalassemia or sickle cell anemia. After treatment, a reduction in the frequency of methylated cytosine residues was observed at all Hpa II sites examined. Despite causing "global" hypomethylation, 5-azacytidine augmented the synthesis of gamma-globin only. Although gamma-gene hypomethylation and increased gamma-gene expression seem to be linked, hypomethylation near other genes was not sufficient to activate transcription. These data suggest that the gamma genes lie in a unique "preactivational" state responsive to hypomethylation, and that other genes are repressed in bone marrow cells by different mechanisms. DNA hypomethylation and an increased concentration of gamma-mRNA were observed in bone marrow cells 2 days after initiation of treatment, indicating that 5-azacytidine may act directly on differentiated erythroid precursors. This compound probably affects early erythroid progenitors as well, since an increased level of gamma-globin synthesis persists for 1-2 weeks after the drug is stopped. A direct effect on erythroid progenitors was also suggested by in vitro assays: Erythroid colonies derived from progenitor cells obtained on day 2 of treatment produced more Hb F than colonies derived from progenitors obtained before 5-azacytidine was given.

We have surveyed 22 kb of DNA from the region surrounding the human fetal (gamma) globin genes and have identified one fragment that meets all of the criteria for a non-tissue specific enhancer element. The enhancer-containing fragment starts approximately 400 bp 3' to the polyadenylation signal of the A gamma gene and is less than 750 bp in length. Addition of this fragment to plasmids containing a 'gamma-CAT' hybrid gene [consisting of the gamma globin gene promoter fused to the chloramphenicol acetyl transferase (CAT) gene] increases CAT expression 6-23-fold in K562 erythroleukemia cells, depending upon the method of transfection. The increase in expression is essentially independent of the orientation or position of the fragment with respect to the gamma-CAT hybrid gene. The 3' gamma enhancer activates heterologous promoters in erythroleukemia cells, and is also active in non-erythroid cell lines. The enhancer acts by increasing the number of transcripts initiated from the normal gamma globin gene transcription initiation site. The enhancer region contains two DNase I hypersensitive sites in erythroleukemia cells but none in nonerythroid human leukemia cell lines. The 3' gamma globin gene enhancer contains a unique element that is similar to sequences found in an enhancer 3' to the chicken beta globin gene, suggesting that this conserved element may have a role in enhancer function.

We studied the abundance and structure of globin mRNAs present in K562 cells both before and after induction of hemoglobin synthesis by hemin. In vitro translation of poly(A)+ RNA from K562 cells generated protein products corresponding to alpha, A gamma-, G gamma-, epsilon-, and zeta-globin mRNAs. Individual globin mRNAs increased 1.5- to 3-fold after induction. Similar results were obtained by measuring steady-state mRNA concentrations of induced and uninduced cells by using S1 nuclease mapping. Globin gene transcripts were correctly initiated and processed. In addition, S1 nuclease analysis revealed the presence of delta-globin mRNA in both control and induced cells. A small percentage of delta-globin transcripts appeared to be initiated upstream from the normal initiation site. beta-Globin mRNA was not detected in any studies. The results (i) suggest that hemin induction of K562 cells is mediated at a transcriptional level and (ii) reveal the dissociation of delta- and beta-globin gene expression in K562 cells compared with normal erythroid cells.

We transfected human complement receptor 2 (CR2/CD21) cDNA containing eukaryotic expression constructs into CR2-negative mouse L cells and human K562 erythroleukemia cells. We subsequently selected stably transformed cells that expressed human CR2, as assessed by flow microfluorimetry analysis and immunoprecipitation of 125I-labeled surface membranes using the monoclonal anti-CR2 antibody, HB5. Utilizing flow microfluorimetry analysis, epitopes recognized by anti-CR2 mAb HB5, OKB7, B2, and four other anti-CR2 antibodies were detected on CR2 expressing transfectants but not parental cells. In addition, CR2 expressing transfected cells efficiently formed rosettes with sheep erythrocyte intermediates bearing human C3bi and C3d, but not C4b or C3b, consistent with the known ligand specificity of CR2. CR2 containing transfectants were also demonstrated to specifically bind EBV. Infection with EBV of CR2 expressing L cells and K562 cells resulted in mean expression of Epstein-Barr nuclear Ag (EBNA) at 48 h in 0.35% of CR2 expressing L cells and 3.7% of CR2 expressing K562 cells. Parental L cells and K562 cells did not express EBNA after EBV infection. These results indicate that CR2 alone is sufficient to transfer both C and EBV receptor functions to heterologous cells. In addition, expression of EBNA was found to be significantly higher in human K562 than mouse L cells, both expressing the same recombinant receptor. These results suggest that mechanisms other than CR2 binding lead to inefficient EBV infection and/or EBNA synthesis in mouse fibroblasts.

The human beta-globin gene is expressed--but the human fetal (gamma) and embryonic (epsilon) globin genes are not--in an induced mouse erythroleukemia cell line (M11-X) that contains most of human chromosome 11. A 24-hr exposure of M11-X cells to 5-azacytidine before induction causes "global" DNA hypomethylation but selective activation of the human gamma-globin genes. Genomic DNA is remethylated 2-3 days after exposure to 5-azacytidine, but sequences near the human and mouse globin genes remain hypomethylated, suggesting that the remethylation process is inhibited in these regions.

Several acute transforming retroviruses encode tyrosine-specific protein kinases which possess structural and functional relationships to cell-surface receptors for certain growth factors. One such tyrosine kinase is encoded by the onc gene, v-fgr, of Gardner-Rasheed feline sarcoma virus (GR-FeSV). Recently, we have isolated and characterized the human gene, c-fgr, corresponding to the viral onc sequence and have shown that c-fgr is a unique gene located on the short arm of chromosome 1 (ref. 7). Here we report that certain lymphomas (but not sarcomas or carcinomas) express fgr-related messenger RNA. This transcript is detected in Burkitt's lymphoma cell lines naturally infected with Epstein-Barr virus (EBV), but not in EBV-negative Burkitt's lymphoma cells. Normal umbilical cord or peripheral blood lymphocyte lines established in vitro by EBV infection also contain detectable c-fgr mRNA. Moreover, a 50-fold increase of the steady-state c-fgr mRNA concentration is observed when uninfected Burkitt's lymphoma cell lines are deliberately infected with EBV. These findings demonstrate for the first time the induction of a proto-oncogene in response to infection by a DNA tumour virus.