Cloning and characterization of human ubiquitin-processing protease-43 from terminally differentiated human melanoma cells using a rapid subtraction hybridization protocol RaSH

Defects in growth control and differentiation occur frequently in human cancers. In the case of human melanoma cells, treatment with a combination of fibroblast interferon (IFN-beta) and the protein kinase C activator mezerein (MEZ) results in an irreversible loss of proliferative potential and tumorigenic properties with a concomitant induction of terminal differentiation. These changes in cellular properties are associated with an induction and suppression in specific subsets of genes that occur in a temporal manner. To identify the complete repertoire of gene changes occurring during melanoma reversion to a more differentiated state a number of molecular approaches are being used. These include, subtraction hybridization using temporally spaced cDNA libraries, random cDNA isolation and evaluation by reverse Northern blotting and high throughput microarray analysis of subtracted cDNA clones. In the present study we have used a novel approach, rapid subtraction hybridization (RaSH), to identify and clone an additional gene of potential relevance to cancer growth control and terminal cell differentiation. RaSH has identified a human ubiquitin-processing protease gene, HuUBP43, that is differentially expressed in melanoma cells as a function of treatment with IFN-beta or IFN-beta + MEZ. HuUBP43 is a type I interferon inducible gene that is upregulated in a diverse panel of normal and tumor cells when treated with IFN-beta via the JAK/STAT kinase pathway. This gene may contribute to the phenotypic changes induced by IFN-beta during growth arrest and differentiation in human melanoma cells and other cell types as well as the antiviral and growth inhibitory effects of interferon.     

Differentiation induction subtraction hybridization (DISH): a strategy for cloning genes displaying differential expression during growth arrest and terminal differentiation.

Human cancers often display aberrant patterns of differentiation. By appropriate chemical manipulation, specific human cancers, such as human melanoma, leukemia and neuroblastoma, can be induced to lose growth potential irreversibly and terminally differentiate. Treatment of HO-1 human melanoma cells with a combination of recombinant human fibroblast interferon (IFN-beta) and the antileukemic compound mezerein (MEZ) results in irreversible growth arrest, a suppression in tumorigenic properties and terminal cell differentiation. A potential mechanism underlying these profound changes in cancer cell physiology is the activation of genes that can suppress the cancer phenotype and/or the inactivation of genes that promote the cancer state. To define the repertoire of genes modulated as a consequence of induction of growth arrest and terminal differentiation in human melanoma cells, we are using a differentiation induction subtraction hybridization (DISH) approach. A subtracted cDNA library, differentiation inducer treated cDNAs minus uninduced cDNAs, was constructed that uses temporally spaced mRNAs isolated from HO-1 cells treated with IFN-beta+MEZ. Approximately 400 random clones were isolated from the subtracted DISH library and analyzed by reverse Northern and Northern blotting approaches. These strategies resulted in the identification and cloning of both 30 known and 26 novel cDNAs displaying elevated expression in human melanoma cells induced to growth arrest and terminally differentiate by treatment with IFN-beta+MEZ. The DISH scheme and the genes presently identified using this approach should provide a framework for delineating the molecular basis of growth regulation, expression of the transformed phenotype and differentiation in melanoma and other cancers.     

Interferon-beta from melanoma cells suppresses the proliferations of melanoma cells in an autocrine manner

Interferon (IFN)-alpha and IFN-beta have been utilized in the treatment of melanoma as a form of cytokine therapy. While previous studies have demonstrated that melanocytes and melanoma cells produce a number of cytokines, it remains unclear whether or not melanocytes and melanoma cells per se produce IFN-alpha or IFN-beta. In the present study, we investigated the expression of IFN-alpha or IFN-beta in human melanocytes and five melanoma cell lines: G-361, C32TG, MMAc, MEWO and VMRC-MELG at both mRNA and protein levels. Both IFN-alpha and IFN-beta mRNA were detected in normal human melanocytes. Likewise, IFN-alpha mRNA was detected in all five melanoma cell lines. However, IFN-beta mRNA was only detected in one melanoma cell line, VMRC-MELG. When melanocytes and melanoma cells were treated with a potent IFN inducer, polyinosinic:polycytidylic acid (poly I:C), the mRNA expression of both IFN-alpha and IFN-beta was significantly upregulated. Poly I:C was not able to induce melanocytes or melanoma cells to produce detectable amounts of IFN-alpha protein, but able to induce a significant amount of IFN-beta in melanocytes and two of the melanoma cell lines: MMAc and VMRC-MELG. Moreover, similar to exogenous IFN-alpha and IFN-beta, poly I:C significantly inhibited the proliferation of all five melanoma cell lines. This suppressive effect was partially blocked by anti-IFN-beta antibody treatment in the IFN-beta-producing melanoma cell lines: MMAc and VMRC-MELG, but not in the non-IFN-beta-producing cell lines: G-361, C32TG and MEWO. Collectively, these studies have demonstrated for the first time that human melanocytes and melanoma cells produce IFN-beta. Furthermore, melanoma cells are capable of suppressing their own proliferation via secretion of endogenous IFN-beta. This finding may have important implications for melanoma therapy.     

Synergistic interactions of interleukin 1, interferon-beta, and tumor necrosis factor in terminally differentiating a mouse myeloid leukemic cell line (M1). Evidence that interferon-beta is an autocrine differentiating factor

The effect was investigated of combinations of cytokines known to be cytostatic for some tumor cells, namely interleukin 1 alpha (IL-1 alpha), interferon-beta (IFN-beta), and tumor necrosis factor (TNF), on the growth and differentiation of the mouse myeloid leukemic cell line, M1, cells. IL-1 alpha, IFN-beta, and TNF by themselves are antiproliferative for M1 cells. Treatment of cells with a mixture of any two of the three cytokines resulted in at least additive growth inhibition. None of these cytokines by themselves induced differentiation of M1 cells as assessed by increased expression of Fc receptors (FcR), stimulation of phagocytic activity and by morphologic criteria. However, as little as 1 U/ml IL-1 alpha in conjunction with IFN-beta or TNF increased FcR expression, phagocytic activity and morphologic changes in addition to inhibiting the growth of M1 cells. The combination of IFN-beta and TNF did not induce differentiation, although the growth of the cells was markedly inhibited. Both TNF and lipopolysaccharide (LPS) induced the in vitro production of IFN activity by M1 cells. Furthermore, the induction of differentiation of M1 cells by a combination of IL-1 alpha with either IFN-beta, TNF, or LPS was inhibited by antibody against mouse IFN-beta. Therefore, it appears that IFN-beta provides one of the two required signals for differentiation of M1 cells by these combinations of stimulants, the other being IL-1. Furthermore, the cytostatic effect of TNF by itself on M1 cells was also partly blocked by anti-IFN-beta antibody, suggesting that IFN-beta is also involved in the growth inhibitory effect of TNF for M1 cells. In contrast, the cytostatic effect of IL-1 on M1 cells was not blocked by anti-IFN-beta antibody. In conclusion, both the cytostatic and differentiative effect of TNF appear to be mediated by IFN-beta. Thus, the combination of IL-1 and IFN-beta or inducers of IFN-beta resulted in terminal differentiation of M1 cells. Northern blot analysis using cDNAs for murine IFN-beta1 or human IFN-beta2 showed an increased expression of mRNA for IFN-beta1 but not for IFN-beta2 by stimulation with TNF or LPS, strongly suggesting that IFN-beta 1 rather than IFN-beta 2 is responsible for TNF or LPS effects.     

Enhancement of human melanoma antigen expression by IFN-beta

Although many immunotherapeutic investigations have focused on improving the effector limb of the antitumor response, few studies have addressed preventing the loss of tumor-associated Ag (TAA) expression, associated with immune escape by tumors. We found that TAA loss from human melanomas usually results from reversible gene down-regulation, rather than gene deletion or mutation. Previously, we showed that inhibitors of MAPK-signaling pathways up-regulate TAA expression in melanoma cell lines. We have now identified IFN-beta as an additional stimulus to TAA expression, including Melan-A/MART-1, gp100, and MAGE-A1. IFN-beta (but neither IFN-alpha nor IFN-gamma) augmented both protein and mRNA expression of melanocytic TAA in 15 melanoma lines (irrespective of initial Ag-expression levels). Treatment of low Ag melanoma lines with IFN-beta increased expression of melanocyte-lineage Ags, inducing susceptibility to lysis by specific CTLs. Treatment with IFN-beta also enhances expression of class I HLA molecules, thereby inducing both nominal TAA and the presenting HLA molecule. Data from fluorescent cellular reporter systems demonstrated that IFN-beta triggers promoter activation, resulting in augmentation of Ag expression. In addition to enhancing TAA expression in melanomas, IFN-beta also stimulated expression of the melanocytic Ag gp100 in cells of other neural crest-derived tumor lines (gliomas) and certain unrelated tumors. Because IFN-beta is already approved for human clinical use in other contexts, it may prove useful as a cotreatment for augmenting tumor Ag expression during immunotherapy.     

Differentiation of nonbeating embryonic stem cells into beating cardiomyocytes is dependent on downregulation of PKC beta and zeta in concert with upregulation of PKC epsilon

Cardiomyocyte differentiation overall has been analyzed in vivo and in vitro at the molecular level by homologous recombination, gene mutation studies, and by transgenics; however, the roles of many signal transduction mechanisms that drive this differentiation process are still not fully understood. One set of signal transduction components that has been studied in detail in mature, differentiated cardiomyocytes is the PKC isotype superfamily. However, while the function of each isotype is slowly being uncovered in adult cardiomyocytes, limited information persists concerning their function in the differentiation process of cardiomyocytes. To begin analyzing the function of specific PKC isotypes in the differentiation process, we employed an established model for differentiating ES cells into cardiomyocyte-positive embryoid bodies (EBs) in vitro. RT-PCR, Western analyses, and confocal microscopy all showed that the expression of specific PKC isotypes was significantly changed as ES cells differentiated into cardiomyocytes. More importantly, by using antagonists specific for each isotype we found that this change was a final step in the differentiation process. PKC beta and zeta downregulation served to promote differentiation (beating), while upregulation of PKC epsilon appeared to amplify differentiation (beating). Finally, melding classical tools (i.e., ionic exchange glass beads) with recently developed methods for differentiating ES cells creates a possible novel technique for investigating differentiation of ES cells into cardiomyocytes as well as other cell types.     

Interferon-gamma induces altered oncogene expression and terminal differentiation in A431 cells

In tumor cell lines in which oncogene expression is abnormal, modulation of the expression of the oncogene (myc, src, or ras) by interferons (IFNs) has been observed concurrently with cell growth inhibition or phenotypic reversion. Oncogene expression has also been reported to vary during the differentiation of several neoplastic cell lines. Treatment of monolayer cultures of A431, a human epidermoid carcinoma cell line, with IFN-gamma resulted in rapid morphological alterations and cell death not seen with either IFN-alpha or IFN-beta. These changes were accompanied by elevated expression of mRNA's for p21 (the c-ras gene product) and the epidermal growth factor receptor as well as increases in the biosynthetic rate of their respective proteins. These effects likewise appeared to be specific for IFN-gamma. Growth inhibition by IFN-gamma was also observed when A431 cells were grown in a three dimensional in vitro culture system. Immunohistochemical staining of these "tumoroids" with a differentiation specific, anti-keratin antibody indicated that IFN-gamma enhanced expression of this keratin. This observation suggests that the killing by IFN-gamma of A431 cells may result from an acceleration of terminal differentiation.     

Regulation of the terminal event in cellular differentiation: biological mechanisms of the loss of proliferative potential

Human plasma has been demonstrated to contain factors that induce the sequential expression of nonterminal and terminal adipocyte differentiation in 3T3 T mesenchymal stem cells. We now report the development of methods for the isolation of purified populations of nonterminally differentiated cells and terminally differentiated cells, and we show that it is possible to experimentally induce transition from the nonterminal to the terminal state of differentiation. With this model system it is therefore now possible to examine the biological and molecular processes associated with the terminal event in differentiation, i.e., the irreversible loss of proliferative potential. In this regard, we demonstrate that transition from the nonterminal to terminal state of differentiation is a complex metabolic process that consists of at least two steps and that this process can be triggered by pulse exposure to an inducer for approximately 12 h but that approximately 24-48 h is required for the process to be completed. The data also establish that induction of the terminal event in differentiation requires protein synthesis but not RNA and DNA synthesis. These and additional results suggest that loss of proliferative potential associated with the terminal event in cellular differentiation is a distinct regulatory process, and we suggest that defects in this regulatory process may be of etiological significance in the pathogenesis of specific human diseases, especially cancer.     

Induction of beta interferon by human immunodeficiency virus type 1 and its gp120 protein in human monocytes-macrophages: role of beta interferon in restriction of virus replication.

In vitro cultivated human monocytes show a time-dependent differentiation into macrophages, characterized by an increased expression of macrophage-specific antigens. Monocytes-macrophages were infected with human immunodeficiency virus type 1 strain Ba-L (HIV-1Ba-L) at different stages of differentiation. When 7-day cultured macrophages were infected in the presence of antibodies to beta interferon (IFN-beta), a significant increase in HIV-1 p24 release was detected. This effect was not observed in 1-day monocytes. This finding suggests that IFN-beta secreted by the infected macrophages inhibits p24 release. Treatment of cultured macrophages with recombinant gp120 (rgp120) protein resulted in the induction of IFN-beta mRNA and in an antiviral state to vesicular stomatitis virus. This rgp120-induced antiviral state was largely neutralized by antibodies to IFN-beta, whereas anti-IFN-alpha antibodies were ineffective. In cultured macrophages, 0.1 IU of IFN-beta per ml was sufficient to induce a marked inhibition of vesicular stomatitis virus yield, whereas this dose was ineffective in 1-day monocytes. These results indicate that (i) HIV-1 (possibly in part through its gp120 protein) induces low levels of IFN-beta in macrophages and (ii) this IFN-beta is very effective in inducing an antiviral state in differentiated macrophages.     

Cycloheximide induces expression of the human interferon beta 1 gene in mouse cells transformed by bovine papillomavirus-interferon beta 1 recombinants.

Mouse cells transformed by a bovine papillomavirus recombinant vector containing the human interferon (IFN) beta 1 (IFN-beta 1) gene could be induced to produce human as well as mouse IFNs. The optimal conditions for induction of human IFN and of its mRNA in these transformants resembled those needed for mouse IFN: high concentrations of DEAE-dextran and low concentrations of polyriboinosinic acid-polyribocytidylic acid. Superinduction by inhibitors of protein synthesis which strongly stimulate IFN-beta 1 induction in human cells had only a small effect on human IFN induction in bovine papillomavirus IFN-beta 1-transformed mouse cells. In contrast, cycloheximide without double-stranded RNA could induce significant levels of human IFN in the bovine papillomavirus IFN-beta 1 mouse transformants. After cycloheximide treatment, these cells contained IFN-beta 1 mRNA whose 5' ends originated in the authentic start site of the human IFN-beta 1 gene, as shown by S1 nuclease mapping. The transferred human gene, propagated extrachromosomally in the mouse cells, was, therefore, inducible under conditions different from those in human cells. The results also confirmed that the inhibitor of protein synthesis, cycloheximide, can induce expression of a human IFN gene.     

Timing of IFN-beta exposure during human dendritic cell maturation and naive Th cell stimulation has contrasting effects on Th1 subset generation: a role for IFN-beta-mediated regulation of IL-12 family cytokines and IL-18 in naive Th cell differentiation

Type I IFNs, IFN-alpha and IFN-beta, are early effectors of innate immune responses against microbes that can also regulate subsequent adaptive immunity by promoting antimicrobial Th1-type responses. In contrast, the ability of IFN-beta to inhibit autoimmune Th1 responses is thought to account for some of the beneficial effects of IFN-beta therapy in the treatment of relapsing remitting multiple sclerosis. To understand the basis of the paradoxical effects of IFN-beta on the expression of Th1-type immune responses, we developed an in vitro model of monocyte-derived dendritic cell (DC)-dependent, human naive Th cell differentiation, in which one can observe both positive and negative effects of IFN-beta on the generation of Th1 cells. In this model we found that the timing of IFN-beta exposure determines whether IFN-beta will have a positive or a negative effect on naive Th cell differentiation into Th1 cells. Specifically, the presence of IFN-beta during TNF-alpha-induced DC maturation strongly augments the capacity of DC to promote the generation of IFN-gamma-secreting Th1 cells. In contrast, exposure to IFN-beta during mature DC-mediated primary stimulation of naive Th cells has the opposite effect, in that it inhibits Th1 cell polarization and promotes the generation of an IL-10-secreting T cell subset. Studies with blocking mAbs and recombinant cytokines indicate that the mechanism by which IFN-beta mediates these contrasting effects on Th1 cell generation is at least in part by differentially regulating DC expression of IL-12 family cytokines (IL-12 and/or IL-23, and IL-27) and IL-18.     

Modulation of the Antigenic Phenotype of Human Melanoma Cells by Differentiation-inducing and Growth-suppressing Agents

Tumor cells often display alterations in their normal program of cellular differentiation. A promising approach for the treatment of cancer involves the induction of terminal differentiation and a loss of proliferative capacity in cancer cells. In human melanoma cells, the combination of mezerein (MEZ) and fibroblast interferon (IFN-β), results in a rapid and irreversible suppression of cell growth with a concomitant increase in the synthesis of melanin. The induction of terminal differentiation is associated with alterations in the expression of several cellular genes, including fibronectin, ISG-15 and ISG-54, and changes in the expression of specific cell surface antigens, including intercellular adhesion molecule-1 (ICAM-1) and HLA Class I antigens. In the HO-1 human melanoma cell line, induction of terminal differentiation by MEZ plus IFN-β results in an induction and/or increased expression of ICAM-1. HLA Class I antigens and HLA Class II antigens. IFN-β and MEZ alone can modulate expression of these antigens to a lower extent than does the combination of compounds. Induction of terminal differentiation and the irreversible suppression of cell growth is not a prerequisite for antigenic modulation in HO-1 cells. This is indicated by the inability of immune interferon (IFN-γ), a strong inducer of ICAM-1, HLA Class I antigens and HLA Class II antigens synthesis, or the combination of IFN-β plus IFN-γ which synergistically but reversibly suppresses HO-1 growth. to induce melanin synthesis or terminal differentiation in HO-1 cells. The inhibitor of protein kinase C, H-7, only marginally alters 72 hr growth suppression induced by MEZ or the interferons, used alone or in combination. In several experiments, H-7 only partially and variably inhibited the enhanced expression of ICAM-1, HLA Class I antigens and HLA Class II antigens in HO-1 cells treated with MEZ. IFN-β or IFN-γ, used alone or in various combinations. This model system will be useful in defining the biochemical, genomic and antigenic changes associated with the chemical induction of terminal differentiation and the loss of proliferative capacity in human melanoma cells.     

Interleukin-6- and leukemia inhibitory factor-induced terminal differentiation of myeloid leukemia cells is blocked at an intermediate stage by constitutive c-myc.

Interleukin-6 (IL-6) and leukemia inhibitory factor (LIF), two multifunctional cytokines, recently have been identified as physiological inducers of hematopoietic cell differentiation which also induce terminal differentiation and growth arrest of the myeloblastic leukemic M1 cell line. In this work, it is shown that c-myc exhibited a unique pattern of expression upon induction of M1 terminal differentiation by LIF or IL-6, with an early transient increase followed by a decrease to control levels by 12 h and no detectable c-myc mRNA by 1 day; in contrast, c-myb expression was rapidly suppressed, with no detectable c-myb mRNA by 12 h. Vectors containing the c-myc gene under control of the beta-actin gene promoter were transfected into M1 cells to obtain M1myc cell lines which constitutively synthesized c-myc. Deregulated and continued expression of c-myc blocked terminal differentiation induced by IL-6 or LIF at an intermediate stage in the progression from immature blasts to mature macrophages, precisely at the point in time when c-myc is normally suppressed, leading to intermediate-stage myeloid cells which continued to proliferate in the absence of c-myb expression.     

IFN-beta pretreatment sensitizes human melanoma cells to TRAIL/Apo2 ligand-induced apoptosis

All human melanoma cell lines (assessed by annexin V and TUNEL assays) were resistant to apoptosis induction by TRAIL/Apo2L protein. TRAIL/Apo2L activated caspase-8 and caspase-3, but subsequent apoptotic events such as poly(ADP-ribose) polymerase cleavage and DNA fragmentation were not observed. To probe the molecular mechanisms of cellular resistance to apoptosis, melanoma cell lines were analyzed for expression of apoptosis regulators (apoptotic protease-associated factor-1, FLIP, caspase-8, caspase-9, caspase-3, cellular inhibitor of apoptosis, Bcl-2, or Bax); no correlation was observed. TRAIL/Apo2L was induced in melanoma cell lines by IFN-beta and had been correlated with apoptosis induction. Because IFN-beta induced other gene products that have been associated with apoptosis, it was postulated that one or more IFN-stimulated genes might sensitize cells to TRAIL/Apo2L. Melanoma cell lines were treated with IFN-beta for 16-24 h before treatment with TRAIL/Apo2L. Regardless of their sensitivity to either cytokine alone, >30% of cells underwent apoptosis in response to the combined treatment. Induction of apoptosis by IFN-beta and TRAIL/Apo2L in combination correlated with synergistic activation of caspase-9, a decrease in mitochondrial potential, and cleavage of poly(ADP-ribose) polymerase. Cleavage of X-linked inhibitor of apoptosis following IFN-beta and TRAIL/Apo2L treatment was observed in sensitive WM9, A375, or WM3211 cells but not in resistant WM35 or WM164 cells. Thus, in vitro IFN-beta and TRAIL/Apo2L combination treatment had more potent apoptotic and anti-growth effects when compared with either cytokine alone in melanoma cells lines.