Genetic programs of epithelial cell plasticity directed by transforming growth factor-β

Epithelial–mesenchymal transitions (EMTs) are an essential manifestation of epithelial cell plasticity during morphogenesis, wound healing, and tumor progression. Transforming growth factor-β (TGF-β) modulates epithelial plasticity in these physiological contexts by inducing EMT. Here we report a transcriptome screen of genetic programs of TGF-β-induced EMT in human keratinocytes and propose functional roles for extracellular response kinase (ERK) mitogen-activated protein kinase signaling in cell motility and disruption of adherens junctions. We used DNA arrays of 16,580 human cDNAs to identify 728 known genes regulated by TGF-β within 4 hours after treatment. TGF-β-stimulated ERK signaling mediated regulation of 80 target genes not previously associated with this pathway. This subset is enriched for genes with defined roles in cell–matrix interactions, cell motility, and endocytosis. ERK-independent genetic programs underlying the onset of EMT involve key pathways and regulators of epithelial dedifferentiation, undifferentiated transitional and mesenchymal progenitor phenotypes, and mediators of cytoskeletal reorganization. The gene expression profiling approach delineates complex context-dependent signaling pathways and transcriptional events that determine epithelial cell plasticity controlled by TGF-β. Investigation of the identified pathways and genes will advance the understanding of molecular mechanisms that underlie tumor invasiveness and metastasis.

TGF-β3-induced Palatogenesis Requires Matrix Metalloproteinases

Cleft lip and palate syndromes are among the most common congenital malformations in humans. Mammalian palatogenesis is a complex process involving highly regulated interactions between epithelial and mesenchymal cells of the palate to permit correct positioning of the palatal shelves, the remodeling of the extracellular matrix (ECM), and subsequent fusion of the palatal shelves. Here we show that several matrix metalloproteinases (MMPs), including a cell membrane-associated MMP (MT1-MMP) and tissue inhibitor of metalloproteinase-2 (TIMP-2) were highly expressed by the medial edge epithelium (MEE). MMP-13 was expressed both in MEE and in adjacent mesenchyme, whereas gelatinase A (MMP-2) was expressed by mesenchymal cells neighboring the MEE. Transforming growth factor (TGF)-β3-deficient mice, which suffer from clefting of the secondary palate, showed complete absence of TIMP-2 in the midline and expressed significantly lower levels of MMP-13 and slightly reduced levels of MMP-2. In concordance with these findings, MMP-13 expression was strongly induced by TGF-β3 in palatal fibroblasts. Finally, palatal shelves from prefusion wild-type mouse embryos cultured in the presence of a synthetic inhibitor of MMPs or excess of TIMP-2 failed to fuse and MEE cells did not transdifferentiate, phenocopying the defect of the TGF-β3-deficient mice. Our observations indicate for the first time that the proteolytic degradation of the ECM by MMPs is a necessary step for palatal fusion.

Transcription factor EGR-1 suppresses the growth and transformation of human HT-1080 fibrosarcoma cells by induction of transforming growth factor beta 1.

The early growth response 1 (EGR-1) gene product is a transcription factor with role in differentiation and growth. We have previously shown that expression of exogenous EGR-1 in various human tumor cells unexpectedly and markedly reduces growth and tumorigenicity and, conversely, that suppression of endogenous Egr-1 expression by antisense RNA eliminates protein expression, enhances growth, and promotes phenotypic transformation. However, the mechanism of these effects remained unknown. The promoter of human transforming growth factor beta 1 (TGF-beta 1) contains two GC-rich EGR-1 binding sites. We show that expression of EGR-1 in human HT-1080 fibrosarcoma cells uses increased secretion of biologically active TGF-beta 1 in direct proportion (rPearson = 0.96) to the amount of EGR-1 expressed and addition of recombinant human TGF-beta 1 is strongly growth-suppressive for these cells. Addition of monoclonal anti-TGF-beta 1 antibodies to EGR-1-expressing HT-1080 cells completely reverses the growth inhibitory effects of EGR-1. Reporter constructs bearing the EGR-1 binding segment of the TGF-beta 1 promoter was activated 4- to 6-fold relative to a control reporter in either HT-1080 cells that stably expressed or parental cells cotransfected with an EGR-1 expression vector. Expression of delta EGR-1, a mutant that cannot interact with the corepressors, nerve growth factor-activated factor binding proteins NAB1 and NAB2, due to deletion of the repressor domain, exhibited enhanced transactivation of 2- to 3.5-fold over that of wild-type EGR-1 showing that the reporter construct reflected the appropriate in vivo regulatory context. The EGR-1-stimulated transactivation was inhibited by expression of the Wilms tumor suppressor, a known specific DNA-binding competitor. These results indicate that EGR-1 suppresses growth of human HT-1080 fibrosarcoma cells by induction of TGF-beta 1.

Tumor necrosis factor receptor associated factor 2 is a mediator of NF-kappa B activation by latent infection membrane protein 1, the Epstein-Barr virus transforming protein.

Latent infection membrane protein 1 (LMP1), the Epstein-Barr virus transforming protein, associates with tumor necrosis factor receptor (TNFR) associated factor 1 (TRAF1) and TRAF3. Since TRAF2 has been implicated in TNFR-mediated NF-kappa B activation, we have evaluated the role of TRAF2 in LMP1-mediated NF-kappa B activation. TRAF2 binds in vitro to the LMP1 carboxyl-terminal cytoplasmic domain (CT), coprecipitates with LMP1 in B lymphoblasts, and relocalizes to LMP1 plasma membrane patches. A dominant negative TRAF2 deletion mutant that lacks amino acids 6-86 (TRAF/ delta 6-86) inhibits NF-kappa B activation from the LMP1 CT and competes with TRAF2 for LMP1 binding. TRAF2 delta 6-86 inhibits NF-kappa B activation mediated by the first 45 amino acids of the LMP1 CT by more than 75% but inhibits NF-kappa B activation through the last 55 amino acids of the CT by less than 40%. A TRAF interacting protein, TANK, inhibits NF-kappa B activation by more than 70% from both LMP1 CT domains. These data implicate TRAF2 aggregation in NF-kappa B activation by the first 45 amino acids of the LMP1 CT and suggest that a different TRAF-related pathway may be involved in NF-kappa B activation by the last 55 amino acids of the LMP1 CT.

Mammalian dwarfins are phosphorylated in response to transforming growth factor beta and are implicated in control of cell growth.

The dwarfin protein family has been genetically implicated in transforming growth factor beta (TGF-beta)-like signaling pathways in Drosophila and Caenorhabditis elegans. To investigate the role of these proteins in mammalian signaling pathways, we have isolated and studied two murine dwarfins, dwarfin-A and dwarfin-C. Using antibodies against dwarfin-A and dwarfin-C, we show that these two dwarfins and an immunogenically related protein, presumably also a dwarfin, are phosphorylated in a time- and dose-dependent manner in response to TGF-beta. Bone morphogenetic protein 2, a TGF-beta superfamily ligand, induces phosphorylation of only the related dwarfin protein. Thus, TGF-beta superfamily members may use overlapping yet distinct dwarfins to mediate their intracellular signals. Furthermore, transient overexpression of either dwarfin-A or dwarfin-C causes growth arrest, implicating the dwarfins in growth regulation. This work provides strong biochemical and preliminary functional evidence that dwarfin-A and dwarfin-C represent prototypic members of a family of mammalian proteins that may serve as mediators of signaling pathways for TGF-beta superfamily members.

Isolation and characterization of mutant cell lines defective in transforming growth factor beta signaling.

To isolate and characterize effector molecules of the transforming growth factor beta (TGFbeta) signaling pathway we have used a genetic approach involving the generation of stable recessive mutants, defective in their TGFbeta signaling, which can subsequently be functionally complemented to clone the affected genes. We have generated a cell line derived from a hypoxanthine-guanine phosphoribosyltransferase negative (HPRT-) HT1080 clone that contains the selectable marker Escherichia coli guanine phosphoribosyltransferase (gpt) linked to a TGFbeta-responsive promoter. This cell line proliferates or dies in the appropriate selection medium in response to TGFbeta. We have isolated three distinct TGFbeta-unresponsive mutants following chemical mutagenesis. Somatic cell hybrids between pairs of individual TGFbeta-unresponsive clones reveal that each is in a distinct complementation group. Each mutant clone retains all three TGFbeta receptors yet fails to induce a TGFbeta-inducible luciferase reporter construct or TGFbeta-mediated plasminogen activator inhibitor-1 (PAI-1) expression. Two of the three have an attenuated TGFbeta-induced fibronectin response, whereas in the other mutant the fibronectin response is intact. These TGFbeta-unresponsive cells should allow selection and identification of signaling molecules through functional complementation.

The recombinant proregion of transforming growth factor beta1 (latency-associated peptide) inhibits active transforming growth factor beta1 in transgenic mice.

All three isoforms of transforming growth factors beta (TGF-betal, TGF-beta2, and TGF-beta3) are secreted as latent complexes and activated extracellularly, leading to the release of the mature cytokines from their noncovalently associated proregions, also known as latency-associated peptides (LAPs). The LAP region of TGF-beta1 was expressed in a baculovirus expression system and purified to homogeneity. In vitro assays of growth inhibition and gene induction mediated by TGF-beta3 demonstrate that recombinant TGF-beta1 LAP is a potent inhibitor of the activities of TGF-betal, -beta2, and -beta3. Effective dosages of LAP for 50% neutralization of TGF-beta activities range from 4.7- to 80-fold molar excess depending on the TGF-beta isoform and activity examined. Using 125I-labeled LAP, we show that the intraperitoneal application route is effective for systemic administration of LAP. Comparison of concentrations of LAP in tissues shows a homogenous pattern in most organs with the exception of heart and muscle, in which levels of LAP are 4- to 8-fold lower. In transgenic mice with elevated hepatic levels of bioactive TGF-betal, treatment with recombinant LAP completely reverses suppression of the early proliferative response induced by TGF-beta1 in remnant livers after partial hepatectomy. The results suggest that recombinant LAP is a potent inhibitor of bioactive TGF-beta both in vitro and in vivo, after intraperitoneal administration. Recombinant LAP should be a useful tool for novel approaches to study and therapeutically modulate pathophysiological processes mediated by TGF-beta3.

Human immunodeficiency virus tat gene transfer to the murine central nervous system using a replication-defective herpes simplex virus vector stimulates transforming growth factor beta 1 gene expression.

The high incidence of neurological disorders in patients afflicted with acquired immunodeficiency syndrome (AIDS) may result from human immunodeficiency virus type 1 (HIV-1) induction of chemotactic signals and cytokines within the brain by virus-encoded gene products. Transforming growth factor beta1 (TGF-beta1) is an immunomodulator and potent chemotactic molecule present at elevated levels in HIV-1-infected patients, and its expression may thus be induced by viral trans-activating proteins such as Tat. In this report, a replication-defective herpes simplex virus (HSV)-1 tat gene transfer vector, dSTat, was used to transiently express HIV-1 Tat in glial cells in culture and following intracerebral inoculation in mouse brain in order to directly determine whether Tat can increase TGF-beta1 mRNA expression. dSTat infection of Vero cells transiently transfected by a panel of HIV-1 long terminal repeat deletion mutants linked to the bacterial chloramphenicol acetyltransferase reporter gene demonstrated that vector-expressed Tat activated the long terminal repeat in a trans-activation response element-dependent fashion independent of the HSV-mediated induction of the HIV-1 enhancer, or NF-kappaB domain. Northern blot analysis of human astrocytic glial U87-MG cells transfected by dSTat vector DNA resulted in a substantial increase in steady-state levels of TGF-beta1 mRNA. Furthermore, intracerebral inoculation of dSTat followed by Northern blot analysis of whole mouse brain RNA revealed an increase in levels of TGF-beta1 mRNA similar to that observed in cultured glial cells transfected by dSTat DNA. These results provided direct in vivo evidence for the involvement of HIV-1 Tat in activation of TGF-beta1 gene expression in brain. Tat-mediated stimulation of TGF-beta1 expression suggests a novel pathway by which HIV-1 may alter the expression of cytokines in the central nervous system, potentially contributing to the development of AIDS-associated neurological disease.

Characterization of the transforming activity of p80, a hyperphosphorylated protein in a Ki-1 lymphoma cell line with chromosomal translocation t(2;5).

We have molecularly cloned a cDNA encoding a protein uniquely expressed and hyperphosphorylated at tyrosine residues in a Ki-1 lymphoma cell that contained chromosomal translocation t(2;5). The encoded protein p80 was shown to be generated by fusion of a protein-tyrosine kinase and a nucleolar protein B23/nucleophosmin (NPM). The coding sequence of this cDNA turned out to be virtually identical to that of the fusion cDNA for NPM-anaplastic lymphoma kinase (ALK) previously cloned from the transcript of the gene at the breakpoint of the same translocation. Overexpression of p80 in NIH 3T3 cells induced neoplastic transformation, suggesting that the p80 kinase is aberrantly activated. The normal form of p80 was predicted to be a receptor-type tyrosine kinase on the basis of its sequence similarity to the insulin receptor family of kinases. However, an immunofluorescence study using COS cells revealed that p80 was localized to the cytoplasm. Thus, subcellular translocation and activation of the tyrosine kinase presumably by its structural alteration would cause the malignant transformation. We also showed that a mutant p80 lacking the NPM portion was unable to transform NIH 3T3 cells. Thus, the NPM sequence is essential for the transforming activity, suggesting that the chromosomal translocation is responsible for the oncogenesis. Finally, Shc and insulin receptor substrate 1 (IRS-1) were tyrosine-phosphorylated and bound to p80 in p80-transformed cells. However, mutants of p80 that were defective for binding to and phosphorylation of Shc and insulin receptor substrate 1 could transform NIH 3T3 cells. Association of these mutants with GRB2 was still observed, suggesting that interaction of p80 with GRB2 but not with Shc or IRS-1 was relevant for cell transformation.

Cross-modulation by transforming growth factor beta in human tuberculosis: suppression of antigen-driven blastogenesis and interferon gamma production.

In tuberculosis, Mycobacterium tuberculosis (MTB)-stimulated T-cell responses are depressed transiently, whereas antibody levels are increased. Lymphoproliferative responses of peripheral blood mononuclear cells (PBMCs) from Pakistani tuberculosis (TB) patients to both mycobacterial and candidal antigens were suppressed by approximately 50% when compared to healthy purified protein derivative (PPD)-positive household contacts. Production of interferon gamma (IFN-gamma) in response to PPD also was depressed by 78%. Stimulation with PPD and the 30-kDa alpha antigen of MTB (30-kDa antigen) induced greater secretion of transforming growth factor beta (TGF-beta), but not interleukin 10 (IL-10) or tumor necrosis factor alpha (TNF-alpha), by PBMCs from TB patients compared to healthy contacts. The degree of suppression correlated with the duration of treatment; patients treated for <1 month had significantly lower T-cell blastogenesis and IFN-gamma production and higher levels of TGF-beta than did patients treated for >1 month. Neutralizing antibody to TGF-beta normalized lymphocyte proliferation in response to PPD, partially restored blastogenesis to candidal antigen, and significantly increased PPD-stimulated production of IFN-gamma in TB patients but not in contacts. Neutralizing antibody to IL-10 augmented, but did not normalize, T-cell responses to both PPD and candida in TB patients and candidal antigen in contacts. TGF-beta, produced in response to MTB antigens, therefore plays a prominent role in down-regulating potentially protective host effector mechanisms and looms as an important mediator of immunosuppression in TB.

Eradication of established intracranial rat gliomas by transforming growth factor beta antisense gene therapy.

Like human gliomas, the rat 9L gliosarcoma secretes the immunosuppressive transforming growth factor beta (TGF-beta). Using the 9L model, we tested our hypothesis that genetic modification of glioma cells to block TGF-beta expression may enhance their immunogenicity and make them more suitable for active tumor immunotherapy. Subcutaneous immunizations of tumor-bearing animals with 9L cells genetically modified to inhibit TGF-beta expression with an antisense plasmid vector resulted in a significantly higher number of animals surviving for 12 weeks (11/11, 100%) compared to immunizations with control vector-modified 9L cells (2/15, 13%) or 9L cells transduced with an interleukin 2 retroviral vector (3/10, 30%) (P < 0.001 for both comparisons). Histologic evaluation of implantation sites 12 weeks after treatment revealed no evidence of residual tumor. In vitro tumor cytotoxicity assays with lymph node effector cells revealed a 3- to 4-fold increase in lytic activity for the animals immunized with TGF-beta antisense-modified tumor cells compared to immunizations with control vector or interleukin 2 gene-modified tumor cells. These results indicate that inhibition of TGF-beta expression significantly enhances tumor-cell immunogenicity and supports future clinical evaluation of TGF-beta antisense gene therapy for TGF-beta-expressing tumors.

Arrest of endotoxin-induced hypotension by transforming growth factor beta1.

Septic shock is a cytokine-mediated process typically caused by a severe underlying infection. Toxins generated by the infecting organism trigger a cascade of events leading to hypotension, to multiple organ system failure, and frequently to death. Beyond supportive care, no effective therapy is available for the treatment of septic shock. Nitric oxide (NO) is a potent vasodilator generated late in the sepsis pathway leading to hypotension; therefore, NO represents a potential target for therapy. We have previously demonstrated that transforming growth factor (TGF) beta1 inhibits inducible NO synthase (iNOS) mRNA and NO production in vascular smooth muscle cells after its induction by cytokines critical in the sepsis cascade. Thus, we hypothesized that TGF-beta1 may inhibit iNOS gene expression in vivo and be beneficial in the treatment of septic shock. In a conscious rat model of septic shock produced by Salmonella typhosa lipopolysaccharide (LPS), TGF-beta1 markedly reduced iNOS mRNA and protein levels in several organs. In contrast, TGF-beta1 did not decrease endothelium-derived constitutive NOS mRNA in organs of rats receiving LPS. We also performed studies in anesthetized rats to evaluate the effect of TGF-beta1 on the hemodynamic compromise of septic shock; after an initial 25% decrease in mean arterial pressure, TGF-beta1 arrested LPS-induced hypotension and decreased mortality. A decrease in iNOS mRNA and protein levels in vascular smooth muscle cells was demonstrated by in situ hybridization and NADPH diaphorase staining in rats treated with TGF-beta1. Thus these studies suggest that TGF-beta1 inhibits iNOS in vivo and that TGF-beta1 may be of future benefit in the therapy of septic shock.

Caenorhabditis elegans genes sma-2, sma-3, and sma-4 define a conserved family of transforming growth factor beta pathway components.

Although transforming growth factor beta (TGF-beta) superfamily ligands play critical roles in diverse developmental processes, how cells transduce signals from these ligands is still poorly understood. Cell surface receptors for these ligands have been identified, but their cytoplasmic targets are unknown. We have identified three Caenorhabditis elegans genes, sma-2, sma-3, and sma-4, that have mutant phenotypes similar to those of the TGF-beta-like receptor gene daf-4, indicating that they are required for daf-4-mediated developmental processes. We show that sma-2 functions in the same cells as daf-4, consistent with a role in transducing signals from the receptor. These three genes define a protein family, the dwarfins, that includes the Mad gene product, which participates in the decapentaplegic TGF-beta-like pathway in Drosophila [Sekelsky, J. J., Newfeld, S. J., Raftery, L. A., Chartoff, E. H. & Gelbart, W. M. (1995) Genetics 139, 1347-1358]. The identification of homologous components of these pathways in distantly related organisms suggests that dwarfins may be universally required for TGF-beta-like signal transduction. In fact, we have isolated highly conserved dwarfins from vertebrates, indicating that these components are not idiosyncratic to invertebrates. These analyses suggest that dwarfins are conserved cytoplasmic signal transducers.