Neutrophilic differentiation induced by human upper airway fibroblast-derived granulocyte/macrophage colony-stimulating factor (GM-CSF)

We have established primary lines of fibroblasts from nasal polyp (NP) tissues as well as from normal nasal (NN) mucosa and have examined the ability of these cells to release hormone-like peptide messenger molecules (cytokines). Our results show that human upper airway fibroblasts release granulocyte/macrophage colony-stimulating factor (GM-CSF), granulocyte-CSF (G-CSF), and macrophage-CSF (M-CSF) in vitro. We also show that fibroblasts derived from NP tissue express the gene for GM-CSF at a higher level, and release the GM-CSF product in greater amounts, than NN fibroblasts. In addition, we have examined the ability of these fibroblasts and their conditioned medium (CM) to induce differentiation of human hemopoietic progenitor cells. After 7 d, cultures of these cells in RPMI-10% fetal bovine serum contained 5 +/- 2.5% (mean +/- SD) neutrophils. In contrast, culture of progenitor cells with fibroblasts resulted in significantly greater neutrophilic differentiation (18 +/- 4%). Culture in fibroblast-CM induced a similar degree of differentiation, and fibroblast-CM from NP fibroblasts elicited greater differentiation compared to CM from NN fibroblasts (17.5 +/- 3 versus 12 +/- 3%). The neutrophilic differentiation induced by fibroblast-CM can be fully inhibited by preincubating this CM with a monoclonal neutralizing antibody to human GM-CSF. Thus, our results demonstrate: (1) the ability of human upper airway fibroblasts to release GM-, G-, and M-CSF in vitro; (2) that fibroblasts derived from NP tissues express the gene and release the product GM-CSF at greater levels compared to NN fibroblasts; and (3) that fibroblast-derived GM-CSF causes neutrophilic differentiation of human hemopoietic progenitors.

Activation of the granulocyte-monocyte colony stimulating factor gene in acute myeloid leukaemia cells is not related to gene rearrangement

Several reports have documented that leukaemic blasts produce a number of cytokines among them the granulocyte-monocyte colony stimulating factor (GM-CSF). We analysed the structure of the gene that codes for GM-CSF in 44 acute myeloid leukaemia (AML) cases in an attempt to establish whether the autocrine production of GM-CSF was due to a structural gene alteration. No structural alteration was detected in the GM-CSF gene in any of the 44 cases studied. We, therefore, conclude that the autocrine production of GM-CSF by leukaemia blasts is not dependent on gene rearrangement.

Influence of medium exchange schedules on metabolic, growth, and GM-CSF secretion rates of genetically engineered NIH-3T3 cells

The metabolic and secretory characteristics of NIH-3T3 fibroblasts transfected with a cDNA encoding human granulocyte-macrophage colony stimulating factor (GM-CSF) were examined as a function of the culture medium exchange schedule. The rates of glucose and glutamine consumption and of lactate and ammonia production were measured over exchange schedules ranging from complete medium replacement weekly (1/week) to complete medium replacement daily (7/week). All measured metabolic rates increased with increased medium exchange rates and accelerated sharply between exchange rates of 3.5/week and 7/week. The lactate/glucose and ammonia/glutamine yield coefficients, however, remained invariant at about 1.9 and 1.0 mol/mol, respectively, under all medium perfusion conditions. A shift-up in medium perfusion rates from 3.5/week to 7/week resulted in increased metabolic rates that resembled those observed in the cultures that were exchanged at the 7/week rate throughout, showing that the metabolic rates could be directly controlled by the perfusion rate. Differential regulation of medium versus serum perfusion demonstrated that increased NIH-3T3 cell metabolism was directly proportional to the serum flux to which the cells were exposed. Thus a limiting serum component is responsible for the altered metabolic and growth rates. The GM-CSF production by the transfected 3T3 cells was stable but exhibited substantial transient increases during periods of cell proliferation, demonstrating that the secretion of transfected gene products can be highly modulated even when the cDNA is driven from a constitutive promoter. These studies show that the metabolic and secretory behavior of genetically engineered cells is influenced by the medium exchange schedule.

Expression of the c-fes proto-oncogene in granulocyte-macrophage colony-stimulating factor-dependent acute myelogenous leukemia cells grown autonomously

In the present article, we show that 6 of 69 acute myelogenous leukemia (AML) samples exhibited autonomous in vitro growth that was dependent on endogenous granulocyte-macrophage colony-stimulating factor (GM-CSF). Cytoplasmic RNA harvested from all 6 leukemia specimens contained GM-CSF transcripts easily detectable by mRNA hybridization. All 6 GM-CSF-expressing leukemia samples simultaneously displayed high levels of transcripts for the c-fes proto-oncogene previously shown to be expressed in GM-CSF sensitive myeloid cells, whereas only 2 of the 48 AML samples not expressing GM-CSF accumulated c-fes mRNA. Seven of additional 14 GM-CSF-expressing specimens showed specific signals upon RNA hybridization with the c-fes probe, but failed to grow autonomously. These results suggest that c-fes and GM-CSF genes may be coordinately regulated in AML blasts and that GM-CSF-mediated growth autonomy may be linked to c-fes expression.

Differentiation of T cell lymphokine gene expression: the in vitro acquisition of T cell memory

A simple in vitro experimental system was devised to reflect the in vivo generation of a T cell anamnestic response so that T cell differentiation could be examined at the level of lymphokine gene expression. Comparison of neonatal and adult T cells revealed that both populations expressed the genes for interleukin 2 (IL-2) and its receptor, but only adult T cells were capable of transcribing mRNAs for IL-3, IL-4, IL-5, IL-6, interferon gamma, and granulocyte/macrophage colony-stimulating factor. However, neonatal T cells could be induced to undergo functional differentiation in vitro, thereby acquiring the capacity to express the lymphokine gene repertoire characteristic for adult T cells. These data suggest that the T cells generated from neonatal blood by a primary stimulation in vitro are functionally indistinguishable from the T cells in adult blood that presumably have undergone primary stimulation in vivo. Therefore, we propose that the term "memory cell" be applied to those T cells that can be identified by their differentiated state of inducible effector-lymphokine gene expression.

Hematopoietic colony-stimulating factors

In summary, hematopoietic growth factors have been discovered, biochemically characterized, cloned, produced by recombinant DNA technology, and put into clinical use in a period of 25 years. We are approaching a greater understanding of the cellular anatomy and molecular mechanisms that regulate production of the CSFs, the ways in which the CSFs interact with their cell surface receptors and trigger their biological effects, the nature of these receptors themselves and their mechanisms of signal transduction, and the effects of the CSFs in vitro and in vivo on hematopoietic progenitor cells and mature leukocytes. However, many questions remain. What is the mechanism that couples growth-factor binding to the triggering of cellular proliferation? How do multi-CSF and GM-CSF cross-compete at the level of the cell-surface receptor, and yet show no primary amino acid sequence homology? What are the mechanisms that regulate the tissue expression profile of multi-CSF compared to the genetically similar growth factor GM-CSF? And, what are the optimal dosages, schedules of administration, and combinations of CSFs optimal for each of several conditions of marrow failure? These are but a few of the questions that continue to occupy much current research interest.

The in situ expression of granulocyte-macrophage colony-stimulating factor (GM-CSF) mRNA at the maternal-fetal interface

Granulocyte-macrophage colony-stimulating factor (GM-CSF) is produced by cells in the placenta, is known to be a growth factor for trophoblast cells in vitro and when injected into pregnant mice at risk for mid-gestation fetal resorption, dramatically lowers the fetal death rate while stimulating placental and fetal growth. We describe here the localization of GM-CSF mRNA expression in murine placenta by in situ hybridization. It is found in small round cells (lymphoid-like) and endothelial cells in the maternal decidua. In addition, GM-CSF transcripts are located in cells of the spongiotrophoblast zone (trophoblast-like cells), but not in the labyrinthine zone. These results indicate that GM-CSF may be influencing the growth and function of the fetal placenta in a paracrine-autocrine manner. These results support earlier observations that link GM-CSF production during pregnancy to decidual T-lymphocytes and further suggest a placental source within the invasive trophoblast.

Site-specific mutagenesis using asymmetric polymerase chain reaction and a single mutant primer

A method is described for preparing site-specific mutants using a polymerase chain reaction (PCR) based protocol. The protocol requires a single mutant primer, and has been used to introduce mutations into DNA fragments ranging in size from 200 bp to 1569 bp in length in the GM-CSF, beta-actin, human growth hormone and erythropoietin genes. Sequence analysis of PCR derived mutant fragments shows an error rate of less than one bp change per 1500 bp incorporated. Single base pair mutations have been introduced into these genes which create unique restriction sites. We demonstrate that these mutant templates may be used for competitive PCR to quantitate mRNA and DNA. This method thus offers a rapid means for producing competitive templates for use in quantitative PCR.

Systematic analysis of the ability of stromal cell lines derived from different murine adult tissues to support maintenance of hematopoietic stem cells in vitro

Hematopoietic stem cells interact with a complex microenvironment both in vivo and in vitro. In association with this microenvironment, murine stem cells are maintained in vitro for several months. Fibroblast-like stromal cells appear to be important components of the microenvironment, since several laboratories have demonstrated that cloned stromal cell lines support hematopoiesis in vitro. The importance of the tissue of origin of such cell lines remains unknown, since systematic generation of stromal cell lines from adult tissues has never been accomplished. In addition, the capacity of stromal cell lines to support reconstituting stem cell has not been examined. We have previously described an efficient and rapid method for the immortalization of primary bone marrow stromal cell lines (Williams et al., Mol. Cell. Biol. 8:3864-3871, 1988) which can be used to systematically derive cell lines from multiple tissues of the adult mouse. Here we report the immortalization of primary murine lung, kidney, skin, and bone marrow stromal cells using a recombinant retrovirus vector (U19-5) containing the simian virus large T antigen (SV40 LT) and the neophosphotransferase gene. The interaction of these stromal cells with factor-dependent cells Patterson-Mix (FDCP-Mix), colony forming units-spleen (CFU-S), and reconstituting hematopoietic stem cells was studied in order to analyze the ability of such lines to support multipotent stem cells in vitro. These studies revealed that stromal cell lines from these diverse tissues were morphologically and phenotypically similar and that they quantitatively bound CFU-S and FDCP-Mix cells equally well. However, only those cell lines derived from bone marrow-supported maintenance of day 12 CFU-S in vitro. One lung-derived stromal cell line, ULU-3, supported the survival of day 8 CFU-S, but not the more primitive CFU-S12. A bone marrow-derived stromal cell line, U2, supported the survival of long-term reconstituting stem cells for up to 3 weeks in vitro as assayed by reconstitution 1 year post-transplant. These studies suggest that adherence of HSC to stromal cells is necessary but not sufficient for maintenance of these stem cell populations and that bone marrow provides specific signals relating to hematopoietic stem cell survival and proliferation.

Primary combined immunodeficiency resulting from defective transcription of multiple T-cell lymphokine genes

The circulating T lymphocytes of a female child with recurrent opportunistic infections were normal in number and phenotype but exhibited poor proliferation and decreased synthesis of the T-cell growth factor interleukin (IL) 2 in response to mitogens. Recombinant IL-2 fully restored the proliferative responses of her T cells, suggesting that her poor immune function was related to IL-2 deficiency. Northern blot analysis of total cellular RNA from the patient's T cells revealed markedly decreased levels of IL-2 mRNA of normal size. In addition, mRNA levels of other lymphokines selectively expressed by T cells, which include IL-3, IL-4, and IL-5, were either severely depressed or absent. The levels of interferon gamma mRNA were moderately decreased, while those of granulocyte-macrophage colony stimulating factor, a lymphokine the production of which is not restricted to T cells, were unaffected. The decreased level of lymphokine mRNA in the patient's T lymphocytes was not from enhanced catabolism but resulted from a diminution in the transcription rate of the affected lymphokine genes. Normal transduction via the T-cell receptor/CD3 complex of biochemical signals necessary for the initiation of lymphokine gene transcription indicated that the defect was distal to the membrane signal-transducing apparatus. The defect is hypothesized to involve a T-cell-specific trans-acting regulatory factor required for transcription of the affected lymphokine genes.

Expression of novel cytokine transcripts in the murine placenta

It is increasingly apparent that lymphohematopoietic cytokines play a unique role during gestation. For example, placentally derived cells, including trophoblast and choriocarcinomas, respond to granulocyte-macrophage (GM)/colony-stimulating factor (CSF) and to colony-stimulating factor 1 (CSF-1), both formerly considered to be hematopoietic cytokines. It has been shown that CSF-1 is produced by the uterine epithelium and GM-CSF by the decidua. However, evidence is emerging that placentally derived cytokines may also influence reproductive function, and the question arises whether anything unique about their expression allows them to function in this particular environment. We have therefore analyzed the expression of cytokine genes in the murine placenta using a panel of cDNA probes that detect GM-CSF, CSF-1, tumor necrosis factor (TNF), interleukin (IL)-1, IL-2, IL-3, and IL-5. We report here the detection of mRNA encoding IL-1, TNF, and CSF-1 that are identical in size to those found in macrophage and fibroblast cell lines. In contrast, five distinct GM-CSF transcripts, four of which are larger than T-cell GM-CSF transcripts, were present. These novel transcripts ranged in size from 5.2, 3.9, 2.4, and 2.1 to 1 kb. Restriction analysis did not reveal any major structural alterations in the placental GM-CSF gene. Thus, the unique placental GM-CSF mRNAs detected most likely result from modified transcription of the GM-CSF gene in the placenta rather than transcription from a modified placental GM-CSF gene. Transcription of enkaryotic genes involves a number of regulatory mechanisms that can generate functionally and structurally diverse polypeptides from a single gene. Modified transcription of the GM-CSF gene in the placenta may serve to generate functionally diverse cytokines which provide the growth and differentiation signals that help to sustain pregnancy. This observation may clarify the unique role played by GM-CSF in reproductive function.

Human thymic epithelial cells produce IL-6, granulocyte-monocyte-CSF, and leukemia inhibitory factor

The development of conditions for culturing normal human thymic epithelial (TE) cells free from contaminating stromal cells has allowed us to characterize a number of cytokines produced by TE cells. Using cDNA probes for human IL-6, granulocyte-monocyte-CSF, and leukemia inhibitory factor (LIF), we identified mRNA for these cytokines by RNA blot analysis of total RNA preparations derived from TE cells. We demonstrated that TE cells produced IL-6 transcripts and that TE cell culture supernatants contained IL-6 biologic activity, as determined by the ability to support proliferation of the T1165 plasmacytoma line. The 1.0-kilobase (kb) transcript of granulocyte-monocyte-CSF was also detected in TE cell-derived total RNA. TE cell culture supernatants contained LIF activity, as determined by proliferation of the murine cell line DA-1a, and a 4.0-kb LIF transcript was detected in TE cell-derived total RNA preparations. The 4.0-kb LIF transcript from TE cell-derived total RNA corresponded in size to the LIF transcripts in PMA-activated T lymphocytes. Thus, using biologic assays and RNA blot analysis, we demonstrated that cultured normal human TE cells produced both immunoregulatory cytokines and cytokines that drive various differentiation stages of human hematopoiesis. Our findings support the hypothesis that TE cells may play a role in providing cytokines that are important for the proliferation and differentiation of hematopoietic precursor cells that migrate to the thymus during fetal and postnatal human thymic development.

Interleukin-6 (IL-6) is an intermediate in IL-1-induced proliferation of leukemic human megakaryoblasts

We have examined the in vitro effects of recombinant human (rh) interleukin-1 (IL-1) on the growth of purified megakaryoblasts obtained from patients with acute megakaryoblastic leukemia. We demonstrate that both IL-1 alpha and IL-1 beta treatment of these cells led to stimulation of DNA synthesis (as shown by increase of 3H-thymidine incorporation up to 35-fold) and also resulted in colony formation of leukemic megakaryoblasts. However, the stimulatory effect of IL-1 was dependent on endogenous production of IL-6, because addition of neutralizing monoclonal antibody (MoAb) to IL-6 abrogated the stimulatory activity of IL-1. In contrast, neutralizing MoAbs to granulocyte (G)-colony stimulating factor (CSF), granulocyte-macrophage (GM)-CSF, and macrophage (M)-CSF failed to counteract the growth-enhancing effects of IL-1. Leukemic megakaryoblasts accumulated IL-6 mRNA and released IL-6 protein into their culture supernatant when exposed to rh IL-1 but failed to disclose transcripts for G-, GM-, and M-CSF under these conditions. Analysis of IL-6 receptor (IL-6R) transcript levels demonstrated that megakaryoblasts constitutively expressed IL-6R mRNA and that these transcripts are down-regulated to undetectable levels upon exposure to IL-1 and IL-6. Increase of 3H-thymidine incorporation by megakaryoblasts could be duplicated by exogenous IL-6 that could be blocked by neutralizing MoAb to IL-6. In conclusion, our results suggest that leukemic megakaryoblasts could produce and secrete IL-6, and express IL-6R, and that the growth-enhancing effect of IL-1 on these cells is indirect, via production of IL-6 by leukemic cells.

Interleukin-1 and interleukin-2 control granulocyte- and granulocyte-macrophage colony-stimulating factor gene expression and cell proliferation in cultured acute myeloblastic leukemia

In vitro proliferation of leukemic cells purified from 10 cases of acute myeloblastic leukemia (AML) was analyzed in basal conditions or in the presence of exogenous recombinant (r) Interleukin (IL) 1. In parallel, blasts from 5 of these patients were studied for granulocyte-macrophage colony-stimulating factor (GM-CSF) or granulocyte-CSF (G-CSF) mRNA. IL-1 augmented the spontaneous AML cell proliferation in all cases and induced de novo expression or increased amounts of GM-CSF and/or G-CSF transcripts in 4 of the 5 cases evaluated. IL-1-induced AML cell proliferation was modulated by neutralizing anti-GM-CSF or anti-G-CSF antibodies in those cases in which CSF mRNAs were induced or increased by exogenous cytokine. In the same cases, biosynthetic labelling and immunoprecipitation studies using monospecific anti-GM-CSF antibodies showed that IL-1 also increased the levels of GM-CSF protein synthesis. Addition of neutralizing anti-IL-1 antibodies to AML cell cultures completely abolished ongoing GM-CSF synthesis, suggesting that endogenous IL-1 is needed to maintain autocrine production of CSFs. The effects of rIL-2 were investigated in a larger series of 21 patients. The cytokine reduced spontaneous AML cell proliferation in 8 cases. It caused complete disappearance of GM-CSF mRNA in 1 case, and marked reduction of G-CSF mRNA in 2 cases. Increased AML cell proliferation was observed in 2 of 21 cases. These findings suggest that expression of CSF genes and cell proliferation in AML are under the control of different cytokines acting in autocrine or paracrine fashion.

Both granulocyte-macrophage colony-stimulating factor and monocytic colony-stimulating factor are produced by the human T-cell line, HUT 102

We present the first report of a T-helper cell line, HUT 102, constitutively producing both granulocyte-macrophage colony-stimulating factor (GM-CSF) and monocytic colony-stimulating factor (M-CSF), as detected by sensitive enzyme-linked immunosorbent assays and Northern blot analysis. However, neither amplification nor structural change of the GM-CSF and M-CSF genes was detected by Southern blot analysis. In the case of HUT 102, in which human T-lymphotropic leukemia virus type I (HTLV-I) is integrated, the viral protein, which acts as a trans-acting transcriptional activator, may induce the production of both GM-CSF and M-CSF.

The repeated sequence CATT(A/T) is required for granulocyte-macrophage colony-stimulating factor promoter activity

The hematopoietic growth factor GM-CSF (granulocyte-macrophage colony-stimulating factor) is expressed by activated but not resting T lymphocytes. Previously, we localized GM-CSF-inducible promoter activity to a 90-bp region of GM-CSF 5'-flanking sequences extending from bp -53 to +37. To more precisely identify the GM-CSF DNA sequences required for inducible promoter activity in T lymphocytes, we have performed mutagenesis within a region of GM-CSF 5'-flanking sequences (bp -57 to -24) that contains the repeated sequence CATT(A/T). Mutations that do not alter the repeated CATT(A/T) sequence do not eliminate inducible promoter activity, whereas mutation or deletion of either of the CATT(A/T) repeats eliminates all inducible promoter activity in T-cell lines and in primary human T lymphocytes. Thus, both copies of the direct repeat CATT(A/T) are required for mitogen-inducible expression of GM-CSF in T cells.

Coordinate secretion of interleukin-1 beta and granulocyte-macrophage colony-stimulating factor by the blast cells of acute myeloblastic leukemia: role of interleukin-1 as an endogenous inducer

Acute myeloblastic leukemia (AML) blasts have been shown to produce a variety of cytokines in culture such as interleukin-1 (IL-1), IL-6, granulocyte-, macrophage-, and granulocyte-macrophage colony-stimulating factor (GM-CSF), and tumor necrosis factor-alpha (TNF alpha). Using two sensitive and specific enzyme-linked immunosorbent assays for IL-1 beta and GM-CSF, we document in the present study that the production of the two cytokines by AML blasts in culture is coordinated. First, we observe a striking correlation between the levels of GM-CSF and IL-1 beta released by the cells. Thus, a high production of IL-1 beta is always concordant with a high production of GM-CSF and, conversely, low production of IL-1 beta is concordant with low levels of GM-CSF. Second, neutralization of intrinsic IL-1 using antibodies that are specific for IL-1 alpha and -1 beta suppresses the release of GM-CSF by the cells. Third, neutralization of the endogenous source of IL-1 also results in an abrogation of GM-CSF mRNA. Fourth, the production of both IL-1 beta and GM-CSF is up-regulated by exposing AML blasts to an exogenous source of IL-1, suggesting a positive regulation of autocrine growth factor production. Taken together, our results indicate that GM-CSF production by AML blasts is mediated by endogenously produced IL-1. Both IL-1 beta and -1 alpha are produced by AML blasts, although IL-1 beta appears to be more abundant. Spontaneous colony formation by AML blasts is abrogated by the addition of neutralizing antibodies against IL-1 beta and GM-CSF, whereas each antibody alone has little effect on blast proliferation. Taken together, our results are consistent with the view that the production of IL-1 beta by AML blasts supports autocrine growth in culture, through induction of CSFs or other cytokines that stimulate blast proliferation.

Examination of the role of CSF-1 independence in myc retrovirus induced monocyte tumorigenesis

These studies were initiated as an attempt to estimate the number and nature of genetic changes that are required in addition to c-myc deregulation during monocyte tumorigenesis, and to determine whether the oncogenic changes that can be created in vitro resemble the actual changes that occur in vivo. We found that superinfecting myc-immortalized monocytes with a colony stimulating factor-1 (CSF-1) expressing retrovirus strongly promoted tumorigenesis, whereas granulocyte/macrophage-CSF (GM-CSF) and v-fms retroviruses, or the spontaneous acquisition of CSF-1 independence did so only moderately. In addition myc-infected monocytes isolated from mice at a stage prior to tumor formation are more tumorigenic than in vitro myc-immortalized monocytes, but they were still largely CSF-1 dependent, and were not as tumorigenic as reinnoculated tumor cells. In the simplest model only two oncogenic activations are required for monocyte/macrophage transformation, immortalization of the cells with c-myc and deregulation of the CSF-1 gene. However, not all mechanisms that result in loss of CSF-1 dependence lead to full tumorigenicity, suggesting that in vivo tumorigenesis may involve multiple secondary events including growth factor independence.

cis-acting determinants of c-myc mRNA stability

Rapid turnover of the c-myc message mediates both the low basal level of mRNA and the rapid response to changes in transcription. The primary RNA instability determinant (RID sequence) resides in the 3' untranslated region (UTR), within an 80 base region that is rich in A and U residues. In contrast to granulocyte-macrophage colony-stimulating factor (GM-CSF) mRNA in which the RID sequence has been mapped to a repeating AUUUA sequence, mutation of the only copy of this sequence in the c-myc 3' UTR has no effect on RNA turnover. Thus the c-myc RID sequence appears to be quite different from that of GM-CSF, which may account for the differential regulation of half-life exhibited by these mRNAs. c-myc mRNA turnover is also tightly coupled to translation since translational inhibitors stabilize this mRNA. Mutation of the initiating AUG to a termination codon stabilizes c-myc RNA, arguing that loading with polysomes (perhaps accompanied by localization on the cytoskeleton) is also required for proper message turnover.

Adenosine-uridine binding factor requires metals for binding to granulocyte-macrophage colony-stimulating factor mRNA

Post-transcriptional gene regulation plays an important role in the expression of granulocyte-macrophage colony-stimulating factor (GM-CSF). Cytokine secretion by activated lymphocytes or mast cells is preceded by dramatic stabilization of the normally labile GM-CSF mRNA. The 3'-untranslated region of GM-CSF and other labile mRNAs contain the destabilizing motif adenosine-uridine-uridine-uridine-adenosine (AUUUA). We recently identified a cytoplasmic protein denoted the adenosine-uridine binding factor (AUBF) which binds with high affinity and specificity to AUUUA elements in synthetic RNA transcripts. We now demonstrate that AUBF binds specifically to GM-CSF mRNA through the destabilizing AUUUA elements. The formation of AUBF-GM-CSF RNA complexes required calcium or magnesium which were sensitive to EDTA or EGTA. A variety of other divalent metals blocked magnesium-dependent AUBF activity. These observations suggest that AUBF may protect GM-CSF mRNA from rapid degradation and play a crucial role in the expression of cytokine genes.

Inducible and non-inducible factors co-operatively activate the GM-CSF promoter by interacting with two adjacent DNA motifs

The upstream region of the mouse granulocyte macrophage colony stimulating factor (GM-CSF) gene between positions -95 and -73 is required for phorbol-12-myristate-13-acetate (PMA)- and calcium ionophore (A23187)-inducible transcriptional activity in vivo. To study the mechanism of GM-CSF gene activation in T cells, nuclear extracts from non-stimulated and PMA/A23187-stimulated Jurkat cells were used in DNA binding assays. DNA mobility shift assays with wild type and mutant oligonucleotides revealed that this region contains at least two DNA binding motifs. One is the binding sequence GGTAGTTCCCC (positions -91 to -81), recognized by NF-GM2 (nuclear factor of GM-CSF 2), and the other is a GC-rich sequence (GC-box). NF-GM2, induced in Jurkat cells by PMA/A23187 stimulation, effectively competed with DNA containing the NF-kappa B binding sequence, suggesting that it has NF-kappa B-like activity. By UV cross-linking analysis, three cross-linked bands, corresponding to Mr 165, 70, and 60 kd for NF-GM2, and 110 and 130 kd for A1 and A2, respectively, were identified. Transfection experiments showed that activation of the GM-CSF gene in response to PMA/A23187 stimulation was abolished by point mutations in either the GC-box or the NF-GM2 site; these mutations also abolished binding of the respective proteins. These results indicate that constitutive (GC-box binding factor) and inducible (NF-GM2) factors regulate the GM-CSF promoter co-operatively in a PMA/A23187-inducible manner in vivo.

Expression of granulocyte and granulocyte-macrophage colony-stimulating factors by human non-hematopoietic tumor cells

The expression of granulocyte colony-stimulating factor (G-CSF) mRNA was studied in human non-hematopoietic tumors, including 18 cases of lung cancers 10 cases of stomach cancers, three cases of glioblastomas, and one case each of breast phyllode sarcoma, thyroid cancer, and hepatocellular carcinoma. Northern blot analysis detected G-CSF mRNA in two of the lung cancer cases, in one of the glioblastoma cases, and in both the breast phyllode sarcoma and hepatocellular carcinoma cases. Since G-CSF receptors were not detected on the tumor cells by 125I-G-CSF binding assay, G-CSF autocrine loop are probably not involved in the growth of these G-CSF-producing tumors. Interestingly, granulocyte-macrophage colony-stimulating factor (GM-CSF) mRNA was concomitantly expressed in most of these G-CSF-producing tumors. No major gene deletions or rearrangements of G-CSF and GM-CSF genes were demonstrated by Southern blot analysis in the tumors expressing G-CSF and GM-CSF mRNAs except for one of the glioblastomas (G3) in which one chromosome 17 allele was deleted. Although the mechanism of the concomitant expression of G-CSF and GM-CSF mRNA is unknown, relatively high frequency of this phenomenon suggests the presence of common transcriptional factors acting on regulatory regions of G-CSF and GM-CSF genomes.

Multiple determinants of eukaryotic mRNA stability

Regulated changes in mRNA stability play an important role in modulating the level of expression of many eukaryotic genes. In several systems, specific sequence determinants that dictate mRNA instability have been identified. Thus, the presence of instability determinants, and not the absence of sequences that dictate stability, appears to be required for regulated mRNA degradation. These instability determinants presumably interact with specific nucleases or other trans-acting factors that regulate the accessibility of the domain to nucleases. Although each RNA destabilization pathway has unique features, in many cases RNA degradation requires ongoing protein synthesis. In some of the systems discussed, the mRNAs are degraded co-translationally, perhaps by a ribosome-associated ribonuclease. For other messages, the mechanistic reasons for the dependence of mRNA degradation on ongoing protein synthesis are still unknown.