Cdc2 and Cdk2 kinase activated by transforming growth factor-beta1 trigger apoptosis through the phosphorylation of retinoblastoma protein in FaO hepatoma cells.

The signaling pathway leading to TGF-beta1-induced apoptosis was investigated using a TGF-beta1-sensitive hepatoma cell line, FaO. Cell cycle analysis demonstrated that the accumulation of apoptotic cells was preceded by a progressive decrease of the cell population in the G(1) phase concomitant with a slight increase of the cell population in the G(2)/M phase in response to TGF-beta1. TGF-beta1 induced a transient increase in the expression of Cdc2, cyclin A, cyclin B, and cyclin D1 at an early phase of apoptosis. During TGF-beta1-induced apoptosis, the transient increase in cyclin-dependent kinase (Cdk) activities coincides with a dramatic increase in the hyperphosphorylated forms of RB. Treatment with roscovitine or olomoucine, inhibitors of Cdc2 and Cdk2, blocked TGF-beta1-induced apoptosis by inhibiting RB phosphorylation. Overexpression of Bcl-2 or adenovirus E1B 19K suppressed TGF-beta1-induced apoptosis by blocking the induction of Cdc2 mRNA and the subsequent activation of Cdc2 kinase, whereas activation of Cdk2 was not affected, suggesting that Cdc2 plays a more critical role in TGF-beta1-induced apoptosis. In conclusion, we present the evidence that Cdc2 and Cdk2 kinase activity transiently induced by TGF-beta1 phosphorylates RB as a physiological target in FaO cells and that RB hyperphosphorylation may trigger abrupt cell cycle progression, leading to irreversible cell death.     

Apoptosis of mink lung epithelial cells by co-treatment of low-dose staurosporine and transforming growth factor-beta1 depends on the enhanced TGF-beta signaling and requires the decreased phosphorylation of PKB/Akt

We demonstrate how co-treatment of low-dose staurosporine (STS) and TGF-beta1, which alone have little effect on cell death, markedly induces apoptosis in Mv1Lu mink lung epithelial cells, but not in its clonal variant R1B cells lacking functional TGF-beta signaling. This process was associated with mitochondria-dependent apoptosis and the enhanced TGF-beta/Smad signaling in Mv1Lu cells. When R1B cells were infected with adenovirus carrying wild-type ALK5, a functional TGF-beta type I receptor gene, the apoptotic cell death was significantly restored in these cells following co-treatment of low-dose STS and TGF-beta1. Treatment of Mv1Lu cells with both low-dose STS and TGF-beta1 decreased the activity of phospho-Akt, which is involved in cell survival signal. In addition, pre-treatments of PI3 kinase inhibitors, LY294002 and wortmannin, further increased the apoptosis of MvlLu cells induced by co-treatment of low-dose STS and TGF-beta1. And overexpression of constitutively active Akt (myr-Akt) using adenoviral expression system inhibited the apoptotic cell death of Mv1Lu cells by about 50% upon co-treatment of low-dose STS and TGF-beta1. These results suggest that co-treatment of low-dose STS and TGF-beta1 induces apoptosis of mink lung epithelial cells by enhancing TGF-beta signaling and in part suppressing cytoprotective signaling.     

Antagonistic effects of TNF-alpha on TGF-beta signaling through down-regulation of TGF-beta receptor type II in human dermal fibroblasts

Transforming growth factor-beta stimulates the production of the extracellular matrix, whereas TNF-alpha has antifibrotic activity. Understanding the molecular mechanism underlying the antagonistic activities of TNF-alpha against TGF-beta is critical in the context of tissue repair and maintenance of tissue homeostasis. In the present study, we demonstrated a novel mechanism by which TNF-alpha blocks TGF-beta-induced gene and signaling pathways in human dermal fibroblasts. We showed that TNF-alpha prevents TGF-beta-induced gene trans activation, such as alpha2(I) collagen or tissue inhibitor of metalloproteinases 1, and TGF-beta signaling pathways, such as Smad3, c-Jun N-terminal kinase, and p38 mitogen-activated protein kinases, without inducing levels of inhibitory Smad7 in human dermal fibroblasts. TNF-alpha down-regulates the expression of type II TGF-beta receptor (TbetaRII) proteins, but not type I TGF-beta receptor (TbetaRI), in human dermal fibroblasts. However, neither TbetaRII mRNA nor TbetaRII promoter activity was decreased by TNF-alpha. TNF-alpha-mediated decrease of TbetaRII protein expression was not inhibited by the treatment of fibroblasts with either a selective inhibitor of I-kappaB-alpha phosphorylation, BAY 11-7082, or a mitogen-activated protein kinase/extracellular signal-regulated kinase inhibitor, PD98059. Calpain inhibitor I (ALLN), a protease inhibitor, inhibits TNF-alpha-mediated down-regulation of TbetaRII. We found that TNF-alpha triggered down-regulation of TbetaRII, leading to desensitization of human dermal fibroblasts toward TGF-beta. Furthermore, these events seemed to cause a dramatic down-regulation of alpha2(I) collagen and tissue inhibitor of metalloproteinases 1 in systemic sclerosis fibroblasts. These results indicated that TNF-alpha impaired the response of the cells to TGF-beta by regulating the turnover of TbetaRII.     

Functions of the two adenovirus early E1A proteins and their conserved domains in cell cycle alteration, actin reorganization, and gene activation in rat cells.

Rat embryo cells were infected with adenovirus type 5 mutants that code for only one of the two early E1A proteins, mutants with defects in one of the two conserved regions common to the two proteins, or mutants with defects in the 46-amino-acid region unique to the 289-amino-acid E1A protein. Cells were scored for altered cell cycle progression, disruption of actin stress fibers, and activation of E2A expression. Mutants lacking either E1A protein were able to cause all of these effects; but mutants lacking a 243-amino-acid protein had less effect, and mutants lacking a 289-amino-acid protein much less effect, than wild-type virus. A mutation in any of the three conserved regions caused a defect in each E1A effect. To investigate the reported function of conserved domain 2 in mitosis, we monitored by fluorescence-activated cell sorter the reduction in Hoechst 33342 fluorescence that occurs when cells divide after undergoing a round of DNA replication in 5-bromodeoxyuridine. A smaller percentage of adenovirus-infected cells than mock-infected cells divided within a given period after completing a round of DNA replication. Viruses with mutations in conserved domain 2 were defective for initiation of cellular DNA replication, as were all other E1A mutants we have examined, but had no specific defect in cell division compared with wild-type virus. Thus, although there may be some specialization of function between the two E1A proteins and between their conserved domains, it was not apparent in the aspects of E1A function and the mutants that we examined.     

Restoration of transforming growth factor-beta type II receptor reduces tumorigenicity in the human adrenocortical carcinoma SW-13 cell line.

Transforming growth factor-beta (TGF-beta) is a potent growth suppressor. Acquisition of TGF-beta resistance has been reported in many tumors, and has been associated with reduced TGF-beta receptor expression. In this study, we examined TGF-beta 1, TGF-beta type I receptor (TbetaRI) and TGF-beta type II receptor (TbetaRII) expression in SW-13 adrenocortical carcinoma cells by Northern and Western blot analysis. SW-13 cells did not express TbetaRII mRNA or protein. We have investigated the role of TbetaRII in modulating tumorigenic potential using stably transfected SW-13 cells with TbetaRII expression plasmid. TbetaRII-positive SW-13 cell growth was inhibited by exogenous human TGF-beta1 (hTGF-beta1) in a dose-dependent manner. In contrast, SW-13 cells and control clones transfected with empty vector remained hTGF-beta1-insensitive. Xenograft examination in athymic nude mice demonstrated that TbetaRII-positive SW-13 cells reduced tumor-forming activity. Reconstructing the TbetaRII can lead to reversion of the malignant phenotype of TbetaRII-negative human adrenocortical carcinoma, which contains SW-13 cells. Reduced TbetaRII expression may play a critical role in determining the malignant phenotype of human adrenocortical carcinoma.     

The amino-terminal portion of CD1 of the adenovirus E1A proteins is required to induce susceptibility to tumor necrosis factor cytolysis in adenovirus-infected mouse cells.

Previous work by our laboratory and others has shown that mouse cells normally resistant to tumor necrosis factor can be made sensitive to the cytokine by the expression of adenovirus E1A. The E1A gene can be introduced by either infection or transfection, and either of the two major E1A proteins, 289R or 243R, can induce this sensitivity. The E1A proteins are multifunctional and modular, with specific domains associated with specific functions. Here, we report that the CD1 domain of E1A is required to induce susceptibility to tumor necrosis factor cytolysis in adenovirus-infected mouse C3HA fibroblasts. Amino acids C terminal to residue 60 and N terminal to residue 36 are not necessary for this function. This conclusion is based on 51Cr-release assays for cytolysis in cells infected with adenovirus mutants with deletions in various portions of E1A. These E1A mutants are all in an H5dl309 background and therefore they lack the tumor necrosis factor protection function provided by the 14.7-kilodalton (14.7K) protein encoded by region E3. Western blot (immunoblot) analysis indicated that most of the mutant E1A proteins were stable in infected C3HA cells, although with certain large deletions the E1A proteins were unstable. The region between residues 36 and 60 is included within but does not precisely correlate with domains in E1A that have been implicated in nuclear localization, enhancer repression, cellular immortalization, cell transformation in cooperation with ras, induction of cellular DNA synthesis and proliferation, induction of DNA degradation, and binding to the 300K protein and the 105K retinoblastoma protein.     

Adenovirus type 5 early region 4 is responsible for E1A-induced p53-independent apoptosis.

In the absence of E1B, the 289- and 243-residue E1A products of human adenovirus type 5 induce p53-dependent apoptosis. However, our group has shown recently that the 289-residue E1A protein is also able to induce apoptosis by a p53-independent mechanism (J. G. Teodoro, G. C. Shore, and P. E. Branton, Oncogene 11:467-474, 1995). Preliminary results suggested that p53-independent cell death required expression of one or more additional adenovirus early gene products. Here we show that both the E1B 19-kDa protein and cellular Bcl-2 inhibit or significantly delay p53-independent apoptosis. Neither early region E2 or E3 appeared to be necessary for such cell death. Analysis of a series of E1A mutants indicated that mutations in the transactivation domain and other regions of E1A correlated with E1A-mediated transactivation of E4 gene expression. Furthermore, p53-deficient human SAOS-2 cells infected with a mutant which expresses E1B but none of the E4 gene products remained viable for considerably longer times than those infected with wild-type adenovirus type 5. In addition, an adenovirus vector lacking both E1 and E4 was unable to induce DNA degradation and cell killing in E1A-expressing cell lines. These data showed that an E4 product is essential for E1A-induced p53-independent apoptosis.     

Tumor necrosis factor-alpha induces Bax-Bak interaction and apoptosis, which is inhibited by adenovirus E1B 19K

Tumor necrosis factor (TNF)-alpha-mediated death signaling induces oligomerization of proapoptotic Bcl-2 family member Bax into a high molecular mass protein complex in mitochondrial membranes. Bax complex formation is associated with the release of cytochrome c, which propagates death signaling by acting as a cofactor for caspase-9 activation. The adenovirus Bcl-2 homologue E1B 19K blocks TNF-alpha-mediated apoptosis by preventing cytochrome c release, caspase-9 activation, and apoptosis of virus-infected cells. TNF-alpha induces E1B 19K-Bax interaction and inhibits Bax oligomerization. Oligomerized Bax may form a pore to release mitochondrial proteins, analogous to the homologous pore-forming domains of bacterial toxins. E1B 19K can also bind to proapoptotic Bak, but the functional significance is not known. TNF-alpha signaling induced Bak-Bax interaction and both Bak and Bax oligomerization. E1B 19K was constitutively in a complex with Bak, and blocked the Bak-Bax interaction and oligomerization of both. The TNF-alpha-mediated cytochrome c and Smac/DIABLO release from mitochondria was inhibited by E1B 19K expression in adenovirus-infected cells. Since either Bax or Bak is essential for death signaling by TNF-alpha, the interaction between E1B 19K and both Bak and Bax may be required to inhibit their cooperative or independent oligomerization to release proteins from mitochondria which promote caspase activation and cell death.     

IFN increases class I MHC antigen expression on adenovirus-infected human cells without inducing resistance to natural killer cell killing.

It has been previously shown that unstimulated NK cells cannot preferentially lyse adenovirus serotypes 2 and 5-infected human cells. In this study, the ability of IFN to promote the selective NK cell-mediated lysis of adenovirus-infected human cells was determined. The relationship between target cell susceptibility to NK cell-mediated killing and class I Ag expression was also analyzed through the use of adenovirus serotype 2 and 5 mutants that do not make the adenovirus early region 3 19-kDa class I binding protein. IFN induced the selective lysis of adenovirus serotype 2 and 5-infected human cells by activating NK cells (IFN-alpha) and protecting uninfected, but not adenovirus-infected cells, from NK cell-mediated lysis (IFN-gamma). IFN-gamma increased the expression of class I Ag on the surface of cells infected with the adenovirus early region 3 deletion mutants, dl327 or dl801, to a level equal to or greater than that expressed on uninfected cells. Despite the increased expression of class I Ag, IFN-gamma could not protect these adenovirus-infected cells from NK cell-mediated lysis. Thus, dl327 or dl801 infection prevented IFN-gamma's induction of cytolytic resistance to NK cell-mediated killing but left IFN-gamma's induction of class I Ag intact. Surface class I Ag levels were substantially higher on IFN-gamma-treated, dl327-, and dl801-infected cells in comparison to cells infected with wild type adenovirus serotype 5. Again, higher target cell levels of class I Ag did not correlate with increased resistance to NK cell-mediated lysis because there was equivalent NK cell-mediated killing of IFN-gamma-treated adenovirus serotype 5-, dl327-, or dl801-infected cells. Thus, IFN-gamma only protects uninfected cells from NK cell-mediated killing, irrespective of target class I Ag levels, and thereby concentrates NK lytic activity on just adenovirus-infected cells. These data demonstrate that IFN-gamma's ability to protect target cells from NK cell-mediated cytolysis is unrelated to IFN-gamma's induction of surface class I MHC Ag.     

Albumin endocytosis in endothelial cells induces TGF-beta receptor II signaling

Vascular endothelial cells undergo albumin endocytosis using a set of albumin binding proteins. This process is important for maintaining cellular homeostasis. We showed by several criteria that the previously described 73-kDa endothelial cell surface albumin binding protein is the 75-kDa transforming growth factor (TGF)-beta receptor type II (TbetaRII). Albumin coimmunoprecipitated with TbetaRII from a membrane fraction from rat lung microvascular endothelial cells. Albumin endocytosis-negative COS-7 cells became albumin endocytosis competent when transfected with wild-type TbetaRII but not when transfected with a domain-negative kinase mutant of TbetaRII. An antibody specific for TbetaRII inhibited albumin endocytosis. A mink lung epithelial cell line, which expresses both the TGF-beta receptor type I (TbetaRI) and the TbetaRII receptor, exhibited albumin binding to the cell surface and endocytosis. In contrast, mutant L-17 and DR-26 cells lacking TbetaRI or TbetaRII, respectively, each showed a dramatic reduction in binding and endocytosis. Albumin endocytosis induced Smad2 phosphorylation and Smad4 translocation as well as increased protein expression of the inhibitory Smad, Smad7. We identified regions of significant homology between amino acid sequences of albumin and TGF-beta, suggesting a structural basis for the interaction of albumin with the TGF-beta receptors and subsequent activation of TbetaRII signaling. The observed albumin-induced internalization of TbetaRII signaling may be an important mechanism in the vessel wall for controlling TGF-beta responses in endothelial cells.     

TGF-beta1 induces rearrangement of FLK-1-VE-cadherin-beta-catenin complex at the adherens junction through VEGF-mediated signaling

VEGF and TGF-beta1 induce angiogenesis but have opposing effects on vascular endothelial cells: VEGF promotes survival; TGF-beta1 induces apoptosis. We have previously shown that TGF-beta1 induces endothelial cell apoptosis via up-regulation of VEGF expression and activation of signaling through VEGF receptor-2 (flk-1). In context with TGF-beta1, VEGF signaling is transiently converted from a survival into an apoptotic one. VEGF promotes cell survival in part via activation of PI3K/Akt by a mechanism dependent on the formation of a multi-protein complex that includes flk-1 and the adherens junction proteins VE-cadherin and beta-catenin. Here we report that TGF-beta1 induces rearrangement of the adherens junction complex by separating flk-1 from VE-cadherin and increasing beta-catenin association with both flk-1 and VE-cadherin. This rearrangement is caused neither by changes in adherens junction mRNA or protein expression nor by post-translational modification, and requires VEGF signaling through flk-1. These results show that the adherens junction is an important regulatory component of TGF-beta1-VEGF interaction in endothelial cells.     

A kunitz-type protease inhibitor bikunin disrupts ligand-induced oligomerization of receptors for transforming growth factor (TGF)-beta and subsequently suppresses TGF-beta signalings

We previously found that bikunin (bik), a Kunitz-type protease inhibitor, suppresses transforming growth factor-beta1 (TGF-beta1)-stimulated expression of urokinase-type plasminogen activator (uPA) in human ovarian cancer cells that lack endogenous bik. In the present study, we tried to elucidate the mechanism by which bik also inhibits plasminogen activator inhibitor type-1 (PAI-1) and collagen synthesis using human ovarian cancer cells. Here, we show that (a) there was an enhanced production of both uPA and PAI-1 in HRA cells in response to TGF-beta1; (b) the overexpression of bik in the cells or exogenous bik results in the inhibition of TGF-beta1 signaling as measured by phosphorylation of the downstream signaling effector Smad2, nuclear translocation of Smad3, and production of PAI-1 and collagen; (c) bik neither decreased expression of TGF-beta receptors (TbetaRI and TbetaRII) in either cell types nor altered the specific binding of 125I TGF-beta1 to the cells, indicating that the effects of bik in these cells are not mediated by ligand sequestration; (d) TbetaRI and TbetaRII present on the same cells exclusively form aggregates in TGF-beta1-stimulated cells; (e) co-treatment of TGF-beta1-stimulated cells with bik suppresses TGF-beta1-induced complex formation of TbetaRI and TbetaRII; and (f) a chondroitin-4-sulfate side chain-deleted bik (deglycosylated bik) does not inhibit TGF-beta1 signaling or association of type I/type II receptor. We conclude that glycosylated bik attenuates TGF-beta1-elicited signaling cascades in cells possibly by abrogating the coupling between TbetaRI and TbetaRII and that this probably provides the mechanism for the suppression of uPA and PAI-1 expression.     

Differential distribution of the adenovirus E1A proteins and colocalization of E1A with the 70-kilodalton cellular heat shock protein in infected cells.

Five distinct localization patterns were observed for the adenovirus E1A proteins in the nuclei of infected HeLa cells: diffuse, reticular, nucleolar, punctate, and peripheral. The variable distribution of E1A was correlated with the time postinfection and the cell cycle stage of the host cell at the time of infection. All staining patterns, with the exception of peripheral E1A localization, were associated with the early phase of infection since only the diffuse, reticular, nucleolar, and punctate staining patterns were observed in the presence of hydroxyurea. Because the E1A proteins (12S and 13S) stimulate the expression of the cellular heat shock 70-kilodalton protein (hsp70), we examined the intracellular distribution of hsp70 in the adenovirus-infected cells. Whereas hsp70 was predominantly cytoplasmic in the cells before infection, after adenovirus infection most of the protein was now found within the nucleus. Specifically, hsp70 was found within the nucleoli as well as exhibiting reticular, diffuse, and punctate nuclear staining patterns, analogous to those observed for the E1A proteins. Double-label indirect immunofluorescence of E1A and hsp70 in infected cells demonstrated a colocalization of these proteins in the nucleus. Translocation of hsp70 to the nucleus was dependent upon both adenovirus infection and expression of the E1A proteins. The localization of hsp70 was unaltered by infection with an E1A 9S cDNA virus which does not synthesize a functional E1A gene product. Moreover, the discrete nuclear localization patterns of E1A and the colocalization of E1A with hsp70 were not observed in adenovirus-transformed 293 cells which constitutively express E1A and E1B. E1A displayed exclusively diffuse nuclear staining in 293 cells; however, localization of E1A into the discrete nuclear patterns occurred after adenovirus infection of 293 cells. Immunoprecipitation of labeled infected-cell extracts with a monoclonal antibody directed against the E1A proteins resulted in precipitation of small amounts of hsp70 along with E1A. These data indicate that the adenovirus E1A proteins colocalize with, and possibly form a physical complex with, cellular hsp70 in infected cells. The relevance of this association, with respect to the function of these proteins during infection and the association of other oncoproteins with hsp70, is discussed.     

Isolation and analysis of adenovirus type 5 mutants containing deletions in the gene encoding the DNA-binding protein.

A genetic system is described which allows the isolation and propagation of adenovirus mutants containing lesions in early region 2A (E2A), the gene encoding the multifunctional adenovirus DNA-binding protein (DBP). A cloned E2A gene was first mutagenized in vitro and then was introduced into the viral genome by in vivo recombination. The E2A mutants were propagated by growth in human cell lines which express an integrated copy of the DBP gene under the control of a dexamethasone-inducible promoter (D. F. Klessig, D. E. Brough, and V. Cleghon, Mol. Cell. Biol. 4:1354-1362, 1984). The protocol was used to construct five adenovirus mutants, Ad5d1801 through Ad5d1805, which contained deletions in E2A. One of the mutants, Ad5d1802, made no detectable DBP and thus represents the first DBP-negative adenovirus mutant, while the four other mutants made truncated DBP-related polypeptides. All five mutants were completely defective for growth and plaque formation on HeLa cell monolayers. Furthermore, the two mutants which were tested, Ad5d1801 and Ad5d1802, did not replicate their DNA in HeLa cells. The mutant Ad5d1804 encoded a truncated DBP-related protein which contained an entire amino-terminal domain derived from the host range mutant Ad5hr404, a variant of Ad5 which multiplies efficiently in monkey cells. While results of a previous study suggest that the amino-terminal domain of DBP could act independently of the carboxyl-terminal domain to enhance late gene expression in monkey cells, the Ad5d1804 polypeptide failed to relieve the block to late viral protein synthesis in monkey cells. The mutant Ad5d1802 was used to study the role of DBP in the regulation of early adenovirus gene expression in infected HeLa cells. These experiments show that E2A mRNA levels are consistently reduced approximately fivefold in Ad5d1802-infected cells, suggesting either a role for DBP in the expression of its own gene or a cis-acting defect caused by the E2A deletion. DBP does not appear to play a significant role in the regulation of adenovirus early regions 1A, 1B, 3, or 4 mRNA levels in infected HeLa cell monolayers since wild-type Ad5- and Ad5d1802-infected cells showed very little difference in the patterns of expression of these genes.