Ski acts as a co-repressor with Smad2 and Smad3 to regulate the response to type beta transforming growth factor.

The c-ski protooncogene encodes a transcription factor that binds DNA only in association with other proteins. To identify co-binding proteins, we performed a yeast two-hybrid screen. The results of the screen and subsequent co-immunoprecipitation studies identified Smad2 and Smad3, two transcriptional activators that mediate the type beta transforming growth factor (TGF-beta) response, as Ski-interacting proteins. In Ski-transformed cells, all of the Ski protein was found in Smad3-containing complexes that accumulated in the nucleus in the absence of added TGF-beta. DNA binding assays showed that Ski, Smad2, Smad3, and Smad4 form a complex with the Smad/Ski binding element GTCTAGAC (SBE). Ski repressed TGF-beta-induced expression of 3TP-Lux, the natural plasminogen activator inhibitor 1 promoter and of reporter genes driven by the SBE and the related CAGA element. In addition, Ski repressed a TGF-beta-inducible promoter containing AP-1 (TRE) elements activated by a combination of Smads, Fos, and/or Jun proteins. Ski also repressed synergistic activation of promoters by combinations of Smad proteins but failed to repress in the absence of Smad4. Thus, Ski acts in opposition to TGF-beta-induced transcriptional activation by functioning as a Smad-dependent co-repressor. The biological relevance of this transcriptional repression was established by showing that overexpression of Ski abolished TGF-beta-mediated growth inhibition in a prostate-derived epithelial cell line.

Association of specific DNA binding and transcriptional repression with the transforming and myogenic activities of c-Ski.

The ski oncogene encodes a transcription factor that induces both transformation and muscle differentiation in avian fibroblasts. The first 304 amino acids of chicken Ski, the transformation domain, are both necessary and sufficient to mediate these biological activities. Ski's biological duality is mirrored by its transcriptional activities: it coactivates or corepresses transcription depending on its interactions with other transcription factors. Ski represses transcription through specific binding to GTCTAGAC (GTCT element) but it possesses a transferable repression activity that can function independently of this DNA element. In this study, we locate this repression domain to the NH2-terminal two-thirds and the GTCT binding region to the COOH-terminal one-third of Ski's transformation domain. Mutations in the transformation domain of c-Ski reveal a strong correlation between GTCT-mediated transcriptional repression and the biological activities of transformation and myogenesis. We also show that a dimerization domain located at the COOH terminal end of the Ski protein increases its transforming activity and its binding to GTCTAGAC.

Heterodimers of the SnoN and Ski oncoproteins form preferentially over homodimers and are more potent transforming agents.

sno is a member of the ski oncogene family and shares ski 's ability to transform avian fibroblasts and induce muscle differentiation. Ski and SnoN are transcription factors that form both homodimers and heterodimers. They recognize a specific DNA binding site (GTCTAGAC) through which they repress transcription. Efficient homodimerization of Ski, mediated by a bipartite C-terminal domain consisting of five tandem repeats (TR) and a leucine zipper (LZ), correlates with efficient DNA binding and cellular transformation. The present study assesses the role of SnoN homodimerization and SnoN:Ski heterodimerization in the activities of these proteins. Unlike Ski, efficient homodimerization by SnoN is shown to require an upstream region of the protein in addition to the TR/LZ domain. Deletion of the TR/LZ from SnoN decreases its activity in transcriptional repression and cellular transformation. When co-expressed in vitro, c-Ski and SnoN preferentially form heterodimers. In vivo, they form heterodimers that bind the GTCTAGAC element. Tethered Ski:Sno hetero-dimers that lack TR/LZ domains are more active than either their monomeric counterparts, tethered Ski:Ski homodimers or full-length SnoN and c-Ski. This work demonstrates, for the first time, the differences between dimer formation by Ski and SnoN and underscores the importance of dimerization in their activity.

A domain necessary for the transforming activity of SnoN is required for specific DNA binding, transcriptional repression and interaction with TAF(II)110.

sno is a member of the ski oncogene family and shares ski's ability to transform avian fibroblasts and induce muscle differentiation. Ski and Sno are nuclear proteins that form homodimers and heterodimers. Ski activates transcription of cellular and viral enhancers and we have identified a DNA binding site (GTCTAGAC) through which it represses transcription. In this work, we show that SnoN binds this site and represses transcription of reporters with this binding site as an upstream element. Using fusions with the Gal4-DNA binding domain in a heterologous reporter assay, we identify a tripartite repression domain in SnoN. A 107 amino acid stretch of the SnoN repression domain, that contains two of the subdomains, is closely related to the minimal region of Ski required for transformation. The third subdomain is unique to SnoN. By analysing deletions involving each of the subdomains, we show that subdomains II and III are also required for DNA binding and cellular transformation. We provide evidence for a quenching mechanism of transcriptional repression by which subdomain II binds to TAF(II)110.

Transcriptional repression by v-Ski and c-Ski mediated by a specific DNA binding site.

The Ski oncoprotein has been shown to bind DNA and activate transcription in conjunction with other cellular factors. Because tumor cells or myogenic cells were used for those studies, it is not clear that those activities of Ski are related to its transforming ability. In this study, we use a nuclear extract of c-ski-transformed cells to identify a specific DNA binding site for Ski with the consensus sequence GTCTAGAC. We demonstrate that both c-Ski and v-Ski in nuclear extracts are components of complexes that bind specifically to this site. By evaluating the features of the sequence that are critical for binding, we show that binding is cooperative. Although Ski cannot bind to this sequence on its own, we use cross-linking with ultraviolet light to show that Ski binds to this site along with several unidentified cellular proteins. Furthermore, we find that Ski represses transcription either through upstream copies of this element or when brought to the promoter by a heterologous DNA binding domain. This is the first demonstration that Ski acts as a repressor rather than an activator and could provide new insights into regulation of gene expression by Ski.

DNA binding and transcriptional activation by the Ski oncoprotein mediated by interaction with NFI.

The Ski oncoprotein has been found to bind non-specifically to DNA in association with unindentified nuclear factors. In addition, Ski has been shown to activate transcription of muscle-specific and viral promoters/enhancers. The present study was undertaken to identify Ski's DNA binding and transcriptional activation partners by identifying specific DNA binding sites. We used nuclear extracts from a v-Ski-transduced mouse L-cell line and selected Ski-bound sequences from a pool of degenerate oligonucleotides with anti-Ski monoclonal antibodies. Two sequences were identified by this technique. The first (TGGC/ANNNNNT/GCCAA) is the previously identified binding site of the nuclear factor I (NFI) family of transcription factors. The second (TCCCNNGGGA) is the binding site of Olf-1/EBF. By electophoretic mobility shift assays we find that Ski is a component of one or more NFI complexes but we fail to detect Ski in Olf-1/EBF complexes. We show that Ski binds NFI proteins and activates transcription of NFI reporters, but only in the presence of NFI. We also find that homodimerization of Ski is essential for co-activation with NFI. However, the C-terminal dimerization domain of c-Ski, which is missing in v-Ski, can be substituted by the leucine zipper domain of GCN4.

Identification of a core functional and structural domain of the v-Ski oncoprotein responsible for both transformation and myogenesis.

The v-ski oncogene promotes cellular transformation and myogenic differentiation. In quail embryo fibroblasts the two properties are displayed simultaneously and terminal muscle differentiation occurs only among cells already transformed by v-ski. To understand how the two phenotypes are derived from a single gene, we have undertaken to identify functionally important regions in v-ski and to test whether these regions can promote one phenotype without the other. We have generated both random and targeted mutations in v-ski and evaluated the effects of these mutations on expression, intracellular location, transformation, and myogenesis. Among a total of 26 mutants analysed, we have not found complete separation of the myogenic and transforming properties. Mutations in the region of v-Ski encoded by exon 1 of c-ski frequently abolish both its transformation and muscle differentiation activities, whereas mutations outside of this region are always tolerated. When expressed in cells from a minigene containing only the exon 1 sequence, the protein displays the transforming and myogenic activities similar to v-Ski. These results argue that the amino acid sequence encoded by exon 1 contains the core functional domain of the oncoprotein. To determine whether this functional domain has a structural counterpart, we have fragmented the v-Ski protein by limited proteolysis and found a single proteolytically stable domain spanning the entire exon 1-encoded region. Physical studies of the polypeptide encoded by exon 1 confirms that it folds into a compact, globular protein. The finding that both the transforming and myogenic properties of v-Ski are inseparable by mutation and are contained in a single domain suggests that they are derived from the same function.

High affinity dimerization by Ski involves parallel pairing of a novel bipartite alpha-helical domain.

c-Ski protein possesses a C-terminal dimerization domain that was deleted during the generation of v-ski, and has been implicated in the increased potency of c-ski in cellular transformation compared with the viral gene. The domain is predicted to consist of an extended alpha-helical segment made up of two motifs: a tandem repeat (TR) consisting of five imperfect repeats of 25 residues each and a leucine zipper (LZ) consisting of six heptad repeats. We have examined the structure and dimerization of TR or LZ individually or the entire TR-LZ domain. Using a quenched chemical cross-linking method, we show that the TR dimerizes with moderate efficiency (Kd = 4 x 10(-6) M), whereas LZ dimerizes poorly (Kd > 2 x 10(-5) M). However, the entire TR-LZ domain dimerizes efficiently (Kd = 2 x 10(-8) M), showing a cooperative effect of the two motifs. CD analyses indicate that all three proteins contain predominantly alpha-helices. Limited proteolysis of the TR-LZ dimer indicates that the two helical motifs are linked by a small loop. Interchain disulfide bond formation indicates that both the LZ and TR helices are oriented in parallel. We propose a model for the dimer interface in the TR region consisting of discontinuous clusters of hydrophobic residues forming "leucine buttons."

Production, characterization and functional activities of v-Ski in cultured cells.

The v-ski oncogene was introduced into mammalian cells in order to study its biochemical and biological properties. v-Ski, produced at relatively high levels by mouse L cells stably transfected with this DNA, was localized to the cell nucleus, was of correct apparent molecular mass, and was capable of complexing with DNA. Transient transfection of reporter plasmids into control or Ski producing mouse L cells revealed that Ski acts as a transcriptional activator of various transcriptional regulatory elements, including CMVie, RSV LTR and SV40. These results indicate that mouse L cells contain the nuclear cofactor(s) required for the ability of v-Ski to bind to DNA and also suggest that the v-Ski present within the cells is functional.

Enhanced expression of mouse c-ski accompanies terminal skeletal muscle differentiation in vivo and in vitro.

Overexpression of either v-ski, or the proto-oncogene, c-ski, in quail embryo fibroblasts induces the expression of myoD and myogenin, converting the cells to myoblasts capable of differentiating into skeletal myotubes. In transgenic mice, overexpression of ski also influences muscle development, but in this case it effects fully formed muscle, causing hypertrophy of fast skeletal muscle fibers. In attempts to determine whether endogenous mouse c-ski plays a role in either early muscle cell determination or late muscle cell differentiation, we analyzed mRNA expression during muscle development in mouse embryos and during in vitro terminal differentiation of skeletal myoblasts. To generate probes for these studies we cloned coding and 3' non-coding regions of mouse c-ski. In situ hybridization revealed low c-ski expression in somites, and only detected elevated levels of mRNA in skeletal muscle beginning at about 12.5 days of gestation. Northern analysis revealed a two-fold increase in c-ski mRNA during terminal differentiation of skeletal muscle cell lines in vitro. Our results suggest that c-ski plays a role in terminal differentiation of skeletal muscle cells not in the determination of cells to the myogenic lineage.

Protooncogene c-ski is expressed in both proliferating and postmitotic neuronal populations.

The cellular protooncogene, c-ski, is expressed in all cells of the developing mouse at low but detectable levels. In situ hybridization and Northern blot analyses reveal that some cells and tissues express this gene at higher levels at certain stages of embryonic and postnatal development. RT-PCR results indicate that alternative splicing of exon 2, known to occur in chickens (Sutrave and Hughes [1989] Mol. Cell. Biol. 9:4046-4051; Grimes et al. [1993] Oncogene 8:2863-2868) does not occur in adult mouse tissues. In the embryo, neural crest cells express the c-ski gene during migration at 8.5 to 9.5 days post coitum (p.c.). Neural crest derivatives such as dorsal root ganglia and melanocytes stain positively with an antibody to the ski protein. At 9 days p.c., the entire neural tube has high levels of c-ski gene expression. By 12-13.5 days only the ependymal layer expresses c-ski above background levels. At 14-16 days p.c., c-ski mRNAs are detected at high levels in the cortical layers of the brain and in the olfactory bulb. In 2 week and 6 week postnatal brains, c-ski gene transcripts are also detected in the hippocampus and in the granule cell layer of the cerebellum. The allantois and placenta exhibit high levels of c-ski mRNAs. Neonatal lung tissue increases c-ski gene expression approximately two-fold compared to prenatal levels. These results suggest that ski plays a role in both the proliferation and differentiation of specific cell populations of the central and peripheral nervous systems and of other tissues.

A carboxyl-terminal region of the ski oncoprotein mediates homodimerization as well as heterodimerization with the related protein SnoN.

Ski is a nuclear oncoprotein, and possibly a transcriptional factor, that has been shown to be involved in both transformation and myogenesis. In attempts to understand the molecular mechanisms underlying the function of Ski, the protein-protein interactions of Ski with itself and with its close relative, SnoN, were investigated. It was found that while both v-Ski and c-Ski bound themselves and each other as bacterial fusion proteins, only c-Ski formed homodimers that could be detected by covalent cross-linking of the native in vitro translated protein in solution. The results also showed that c-Ski formed heterodimers with SnoN. Deletion analysis showed that the carboxyl-terminal third of c-Ski, which is deleted in v-Ski, was required for stable dimer formation in solution. This region consists of two predicted structural motifs that constitute the c-Ski dimerization domain. The more amino-terminal motif is predicted to be mostly alpha helical and is comprised of five tandem repeats of 25 amino acids each and was required for c-Ski dimerization. The second motif is a predicted leucine zipper that was not required for dimerization but greatly increased the fraction of Ski protein detected as dimers. This minor c-Ski homodimerization domain appeared to be required for Ski-Sno heterodimer formation.

Induction of the c-ski proto-oncogene by phorbol ester correlates with induction of megakaryocyte differentiation.

Overexpression of v-ski blocks the terminal differentiation of chicken erythroblasts, and in cooperation with v-sea causes transformation of these cells, indicating that c-ski may play a role in regulating either proliferation or differentiation in hematopoietic cells. We examined c-ski expression in four different myeloid cell lines which can be induced to differentiate by exposure to phorbol 12-myristate 13-acetate (PMA). Two of the cell lines are multipotent and have the ability to differentiate into either erythrocytes or megakaryocytes (K562 and HEL cells), one cell line differentiates exclusively into megakaryocytes (CHRF-288-11), and the fourth cell line differentiates into either monocytes or granulocytes (HL-60). Our findings indicate that c-ski mRNA is up regulated by PMA only in those cell lines which respond by differentiating along the megakaryocyte lineage. The extent of differentiation and the observed increase in c-ski mRNA levels are positively correlated with the PMA concentration used to induce differentiation. Experiments in which CHRF-288-11 cells were treated with the protein kinase C (PKC) activator bryostatin 1 indicate that c-ski mRNA induction is not a general effect of PKC activation. The results strongly suggest that c-ski expression is correlated with megakaryocyte maturation.

Sequence and biological activity of chicken snoN cDNA clones.

cDNA clones of the ski-related gene, sno, were isolated from a chicken cDNA library and sequenced. In contrast to the human system, from which two forms of sno cDNAs have been isolated, we obtained only one type of chicken sno cDNA, that encoding snoN. The coding region for chicken snoN was inserted into the retroviral vectors RCAS(A) and RCASBP(A) and introduced into chicken embryo fibroblasts (CEFs) or quail embryo cells (QECs). Like the various forms of ski, snoN appears to be localized in the nucleus, and high levels of snoN expression cause transformation of CEFs and muscle differentiation of QECs. In contrast to ski however, low-level expression of snoN cannot induce transformation, and is only weakly myogenic.

Activation of the c-ski oncogene by overexpression.

The v-ski oncogene is a truncated version of the cellular proto-oncogene, c-ski, and lacks sequences coding for both the N- and C-terminal ends of the c-ski protein. In the region of overlap, v-ski and c-ski differ by only one amino acid. To determine whether these differences underlie v-ski's oncogenic activation, we have cloned cDNAs for several alternatively spliced c-ski mRNAs and introduced these cDNAs into replication-competent retroviral vectors. The biological activities of these c-ski constructs have been compared with those of v-ski. We found that all c-ski gene products, when expressed at high levels from the promoter in the retroviral long terminal repeat, can induce morphological transformation, anchorage independence, and muscle differentiation in avian cells. Cells that are susceptible to ski-induced transformation and myogenesis normally express endogenous c-ski at low levels. Thus, it appears that overexpression of ski is sufficient for oncogenic and myogenic activation.

Transformation-defective v-ski induces MyoD and myogenin expression but not myotube formation.

The ski oncogene induces muscle differentiation in otherwise nonmyogenic quail embryo cells (C. Colmenares and E. Stavnezer, Cell 59:293-303, 1989). Here we report that v-ski induces both MyoD and myogenin expression, suggesting that activation of these muscle regulatory genes may be a critical step in ski-induced myogenesis. We also describe a transformation-defective mutant of v-ski (tdM5i) that fails to induce myotube formation, although it induces the expression of many muscle-specific genes, including the MyoD and myogenin genes. Therefore, if activation of MyoD and myogenin expression is a necessary component of the myogenic program triggered by ski, it is clearly insufficient to account for complete muscle differentiation.

Structure and activities of the ski oncogene.

The ski oncogene is the transforming gene, v-ski, of the defective SKV avian carcinoma viruses. V-ski transformation causes increased proliferation of embryo fibroblasts and also induces a number of genes characteristic of the muscle lineage. In natural SKV isolates the 49 kDa v-ski polypeptide is expressed as a fusion protein with N-terminal gag and other viral sequences. The cellular homologue of v-ski, c-ski, is a large gene comprising at least 70 kb and containing at least seven coding exons. V-ski consists of most of the first five coding exons of c-ski. The proteins encoded by both genes are nuclear proteins that bind DNA and contain recognised motifs common to known nuclear regulatory proteins. A c-ski-related gene called sno has also been recently identified and, like c-ski, it is expressed in several human tumour cell lines.

The ski oncogene induces muscle differentiation in quail embryo cells.

Quail embryo cells (QECs) are primary cultures of fibroblastoid cells that become myogenic after infection with avian retroviruses expressing the ski oncogene (SKVs). ski also stimulates proliferation of QECs and induces morphological transformation and anchorage-independent growth. Paradoxically, ski-transformed clones picked from soft agar are capable of muscle differentiation. ski-induced differentiation is essentially indistinguishable from that of uninfected myoblasts in culture with regard to muscle-specific gene expression, commitment, and inhibition by growth factors or other oncogenes. However, ski-induced myoblasts have less stringent requirements for growth and differentiation. Uninfected QECs cannot differentiate and do not express an early marker for the myogenic lineage. Clonal analysis indicates that at least 40% of QECs are converted by ski to differentiating myoblasts. The data suggest that ski induces either the capacity for differentiation in an "incompetent" muscle precursor or the determination of nonmyogenic cells to the myogenic lineage.

The v-ski oncogene encodes a truncated set of c-ski coding exons with limited sequence and structural relatedness to v-myc.

The nucleotide sequence of a biologically active v-ski gene from a cloned proviral segment shows that ski is a 1,312-base sequence embedded in the p19 region of the avian leukosis virus gag gene. The v-ski sequence contains a single open translational reading frame that encodes a polypeptide with a molecular mass of 49,000 daltons. The predicted amino acid sequence includes nuclear localization motifs that have been identified in other nuclear oncoproteins. It also contains a proline-rich region and a set of cysteine and histidine residues that could constitute a metal-binding domain. Two regions of the amino acid sequences of v-ski and v-myc are related, and the two proteins exhibit similar distributions of hydrophobic and hydrophilic amino acids. Cloned segments of the chicken c-ski proto-oncogene totaling 65 kilobases have been analyzed, and regions related to v-ski have been sequenced. The results indicate that v-ski is derived from at least five coding exons of c-ski, that it is correctly spliced, and that it is missing c-ski coding sequences at both its 5' and 3' ends. The c-ski and avian leukosis virus sequences that overlap the 5' virus/v-ski junction in Sloan-Kettering virus contain an 18-of-20-base sequence match that presumably played a role in the transduction of ski by facilitating virus/c-ski recombination.

Polyproteins containing a domain encoded by the V-SKI oncogene are located in the nuclei of SKV-transformed cells.

SKV-transformed nonproducer clones were isolated from infected quail and chicken embryo cells. Analysis of intracellular viral RNAs by the Northern technique revealed that each clone contained a single SKV genome (either 5.7 or 8.9 kb) but no genome of the helper virus. Analysis of intracellular viral proteins containing gag determinants revealed that each clone contained a single species of either 55, 110, or 125 kDa. The intracellular location of these proteins was determined by indirect immunofluorescence employing either monoclonal antibodies (anti-p19gag) or conventional antiserum against gag proteins. All three of the SKV-specific proteins were localized to the nuclei of the transformed cells.