The junction-associated protein AF-6 interacts and clusters with specific Eph receptor tyrosine kinases at specialized sites of cell-cell contact in the brain.

The AF-6/afadin protein, which contains a single PDZ domain, forms a peripheral component of cell membranes at specialized sites of cell-cell junctions. To identify potential receptor-binding targets of AF-6 we screened the PDZ domain of AF-6 against a range of COOH-terminal peptides selected from receptors having potential PDZ domain-binding termini. The PDZ domain of AF-6 interacts with a subset of members of the Eph subfamily of RTKs via its COOH terminus both in vitro and in vivo. Cotransfection of a green fluorescent protein-tagged AF-6 fusion protein with full-length Eph receptors into heterologous cells induces a clustering of the Eph receptors and AF-6 at sites of cell-cell contact. Immunohistochemical analysis in the adult rat brain reveals coclustering of AF-6 with Eph receptors at postsynaptic membrane sites of excitatory synapses in the hippocampus. Furthermore, AF-6 is a substrate for a subgroup of Eph receptors and phosphorylation of AF-6 is dependent on a functional kinase domain of the receptor. The physical interaction of endogenous AF-6 with Eph receptors is demonstrated by coimmunoprecipitation from whole rat brain lysates. AF-6 is a candidate for mediating the clustering of Eph receptors at postsynaptic specializations in the adult rat brain.

Phosphorylation and regulation of Raf by Akt (protein kinase B).

Activation of the protein kinase Raf can lead to opposing cellular responses such as proliferation, growth arrest, apoptosis, or differentiation. Akt (protein kinase B), a member of a different signaling pathway that also regulates these responses, interacted with Raf and phosphorylated this protein at a highly conserved serine residue in its regulatory domain in vivo. This phosphorylation of Raf by Akt inhibited activation of the Raf-MEK-ERK signaling pathway and shifted the cellular response in a human breast cancer cell line from cell cycle arrest to proliferation. These observations provide a molecular basis for cross talk between two signaling pathways at the level of Raf and Akt.

Differentiation stage-specific inhibition of the Raf-MEK-ERK pathway by Akt.

Extracellular signals often result in simultaneous activation of both the Raf-MEK-ERK and PI3K-Akt pathways (where ERK is extracellular-regulated kinase, MEK is mitogen-activated protein kinase or ERK kinase, and PI3K is phosphatidylinositol 3-kinase). However, these two signaling pathways were shown to exert opposing effects on muscle cell hypertrophy. Furthermore, the PI3K-Akt pathway was shown to inhibit the Raf-MEK-ERK pathway; this cross-regulation depended on the differentiation state of the cell: Akt activation inhibited the Raf-MEK-ERK pathway in differentiated myotubes, but not in their myoblast precursors. The stage-specific inhibitory action of Akt correlated with its stage-specific ability to form a complex with Raf, suggesting the existence of differentially expressed mediators of an inhibitory Akt-Raf complex.

The PDZ protein erbin modulates beta-catenin-dependent transcription.

Erbin is a member of the leucine-rich repeat and PDZ domain family that can regulate proliferation, differentiation and cell adhesion. As a binding partner of the receptor tyrosine kinase ErbB2, erbin targets this receptor to the basolateral membrane of polarized epithelial cells. In addition, erbin is known to inhibit the Ras-mediated activation of the mitogen-activated protein kinase pathway. Recently we identified the proto-oncoprotein beta-catenin as a ligand of the PDZ domain of erbin. Here we demonstrate that erbin acts as a negative regulator of the beta-catenin/T-cell-factor-dependent gene expression. In contrast, a mutant of erbin with a deletion of the N-terminal leucine-rich repeat allows the PDZ domain of erbin to increase the beta-catenin/T-cell-factor-dependent transcription. This mutant localizes to the nucleus and mimics a putative splice variant found in keratinocytes. Thus, erbin has the potential to act as an inhibitor as well as an activator of the beta-catenin-regulated gene expression.

The Bcr kinase downregulates Ras signaling by phosphorylating AF-6 and binding to its PDZ domain.

The protein kinase Bcr is a negative regulator of cell proliferation and oncogenic transformation. We identified Bcr as a ligand for the PDZ domain of the cell junction and Ras-interacting protein AF-6. The Bcr kinase phosphorylates AF-6, which subsequently allows efficient binding of Bcr to AF-6, showing that the Bcr kinase is a regulator of the PDZ domain-ligand interaction. Bcr and AF-6 colocalize in epithelial cells at the plasma membrane. In addition, Bcr, AF-6, and Ras form a trimeric complex. Bcr increases the affinity of AF-6 to Ras, and a mutant of AF-6 that lacks a specific phosphorylation site for Bcr shows a reduced binding to Ras. Wild-type Bcr, but not Bcr mutants defective in binding to AF-6, interferes with the Ras-dependent stimulation of the Raf/MEK/ERK pathway. Since AF-6 binds to Bcr via its PDZ domain and to Ras via its Ras-binding domain, we propose that AF-6 functions as a scaffold-like protein that links Bcr and Ras to cellular junctions. We suggest that this trimeric complex is involved in downregulation of Ras-mediated signaling at sites of cell-cell contact to maintain cells in a nonproliferating state.

Mutagenesis and selection of PDZ domains that bind new protein targets.

PDZ domains are a recently characterized protein-recognition module. In most cases, PDZ domains bind to the C-terminal end of target proteins and are thought thereby to link these target proteins into functional signaling networks. We report the isolation of artificial PDZ domains selected via a mutagenesis screen in vivo, each recognizing a different C-terminal peptide. We demonstrate that the PDZ domains isolated can bind selectively to their target peptides in vitro and in vivo. Two of the target peptides chosen are the C-terminal ends of two cellular transmembrane proteins with which no known PDZ domains have been reported to interact. By targeting these artificial PDZ domains to the nucleus, interacting target peptides were efficiently transported to the same subcellular localization. One of the isolated PDZ domains was tested and shown to be efficiently directed to the plasma membrane when cotransfected with the full-length transmembrane protein in mammalian cells. Thus, artificial PDZ domains can be engineered and used to target intracellular proteins to different subcellular compartments.

Human c-myb is expressed in cervical carcinomas and transactivates the HPV-16 promoter.

Human c-myb is normally involved in the regulation of proliferation and differentiation of hematopoietic cells. Until now, only a few reports have described elevated c-myb gene expression in epithelial tissue, suggesting that under certain circumstances, c-Myb protein might play a role during the process of malignant transformation of epithelial cells. To investigate a possible role of c-myb during papillomavirus-associated carcinogenesis, we investigated the c-myb mRNA expression in human papillomavirus (HPV)-associated tumors and tumor cell lines. Seven of nine cervical carcinomas and two of three carcinoma cell lines exhibited elevated c-myb transcriptional activity. In contrast to malignant cervical neoplasias, only 3 of 15 condylomata acuminata expressed a sparse signal for c-myb mRNA. Since the c-Myb protein has been described as a potent transcriptional regulator, we investigated the transactivating properties of c-Myb on the HPV-16 promoter/enhancer. Cotransfection of a chloramphenicol acetyltransferase-reporter plasmid containing the HPV-16 enhancer/promoter element with a full-length c-Myb-expressing plasmid resulted in a significant induction (4.3-fold) of the HPV-16 promoter, whereas expression of a carboxy-terminally deleted c-Myb protein led to no effects. Gel shift experiments showed a specific binding of recombinant c-Myb protein on the HPV-16 P97 enhancer. These data indicate that elevated c-myb expression occurs with HPV-associated cell transformation. Since c-Myb has been shown to stimulate the HPV-derived oncoprotein expression via transcriptional activation, it may play a role in the process of HPV-associated cervical carcinogenesis.

Epidermal growth factor receptor expression in human lung cancer cell lines.

Non-small cell lung cancer (NSCLC) and small cell lung cancer (SCLC) cell lines were studied for epidermal growth factor (EGF) receptor expression. All NSCLC cell lines tested (eight of eight) had specific EGF binding sites, whereas only five of 11 SCLC cell lines bound EGF. NSCLC and SCLC cell lines expressed the same type of high affinity EGF binding sites with a Kd of 0.5 to 4.5 nM; however, NSCLC cells bound significantly more EGF than SCLC cell lines. The amount of binding sites in NSCLC cells ranged between 71 and 1,000 fmol/mg of protein and in SCLC cells, between 26 and 143 fmol/mg of protein. The two SCLC cell lines with EGF binding values within the range of NSCLC belonged to the variant subtype of SCLC. By means of an anti-erbB serum and indirect radioimmunoprecipitation, a strong Mr approximately 170,000 protein band could be detected in the NSCLC cell lines. This protein corresponds to the EGF receptor molecule. Its identity was proven by competition with excess erbB antigen for the antibody during the radioimmunoprecipitation. Furthermore, this Mr 170,000 protein exhibited protein kinase activity as evidenced by in vitro autophosphorylation. The radioactivity incorporated into the Mr 170,000 band in radioimmunoprecipitation and protein kinase assays was 10 to 100 times lower in these SCLC cell lines which were positive in the EGF binding assay compared to the NSCLC cell lines. We conclude that NSCLC in contrast to SCLC expresses high levels of EGF receptors which may be used to facilitate the differential diagnosis in some cases of lung cancer. These data suggest that EGF may play a role in growth and differentiation of NSCLC.

Activation of Mil/Raf protein kinases in mitotic cells.

The c-Raf-1 protein kinase is a major element of several signal transduction pathways and thought to be involved in entry into the S phase of the cell cycle. Here we show that c-Raf-1 as well as the transforming viral fusion protein Gag-Mil, in which most of the amino terminal regulatory region of the avian Raf homologue Mil is deleted, are activated five- to sixfold in mitotic cells. Mitotic activation of Mil/Raf kinase activity correlates with reduced electrophoretic mobility caused by hyperphosphorylation at serine/threonine residues located in the carboxy terminal part of c-Raf-1. Mitotic hyperphosphorylation occurs in various cell-lines indicating that it is ubiquitous. Our data suggest a novel function for Mil/Raf kinases in late stages of the cell cycle.

Direct interaction of CREB protein with 21 bp Tax-response elements of HTLV-ILTR.

The three 21-bp repeats (Tax-responsive elements) of the long terminal repeat (LTR) of the human T-cell leukemia virus (HTLV-I) mediates the response of the Tax protein. All three Tax-responsive elements (TREs) contain a TGACG motif, reminiscent of the CREB/ATF-binding site TGACGTCA. DNA-affinity chromatography with the 5'-TRE resulted in a previous study in proteins of about 32, 36 to 42, 50 and 110 kDa. Here we demonstrate that the 42 kDa protein is the cAMP-response element-binding (CREB) protein. This is shown by phosphorylation of the proteins eluted from the DNA-affinity column with protein kinase A (PKA) in vitro and subsequent indirect immunoprecipitation with a CREB-specific antiserum raised against an internal CREB-specific peptide. This method allows detection of phosphorylated proteins by autoradiography with high sensitivity and is superior to metabolic labeling. One of the phosphorylated proteins co-migrates with immuno-affinity-purified CREB protein--also phosphorylated in vitro--and competes with the peptide antigen, which proves the specificity of the reaction. The purified CREB protein leads to specific DNA-protein complexes in DNA mobility-shift analyses with all three TREs. Comparison of these TRE-CREB complexes with those formed by nuclear extracts from the HTLV-I-transformed T-cell line C81-66-45 indicates that additional cellular factors contribute to the complexes, especially to the middle TRE. This is also shown by using CREB-depleted instead of complete nuclear extracts for DNA mobility-shift assays. Antibodies against CREB but not Tax affect the mobility of the DNA-protein complex.

Tax-independent binding of multiple cellular factors to Tax-response element DNA of HTLV-I.

The human T-cell leukemia virus type I (HTLV-I) promoter contains three copies of imperfect repeats of a 21-base pair sequence designated here as TRE (Tax-response element) that is responsive to the virally encoded transactivator protein Tax. We have identified and separated four nuclear proteins from C81-66-45 cells, an HTLV-I immortalized Tax-expressing human T-lymphocyte line (Salahuddin et al., 1983), that interact with the TRE-DNA, none of which are identical with the Tax-protein. The proteins identified have molecular weights of about 32, 36 to 42, 50 and 110 kD. Four different methods were used to identify the proteins. First, from different cell lines three or all four of the nuclear proteins were specifically cross-linked by UV irradiation to the radioactively labeled TRE-DNA fragment. Second, TRE-DNA binding proteins sedimented through a glycerol density gradient at rates corresponding to proteins of native molecular weights of 35 to 50 kD and 110 kD. Third, only the 50 kD protein was retained on a biotinylated DNA-streptavidin matrix when the DNA fragment contained the TRE-DNA. Fourth, extensive purification by several cycles of TRE-DNA affinity chromatography resulted in the 32, 36 to 42 and 110 kD proteins and to less extent the 50 kD factor. Two abundant proteins of 75 and 80 kD were competed out by poly[d(I-C)] in all reactions. The cAMP-response element CRE, TGACGTCA, present in the 21 base-pair sequence, appears to be essential for specific protein-TRE-DNA interactions because mutation of the two G's destroys this complex. This result suggests that the cAMP response element binding protein, CREB, is involved in the protein-TRE-DNA complex and in mediating the Tax response.

Isolation and characterization of the tax protein of HTLV-I.

The transactivator protein tax of the human T-cell leukemia virus type I, HTLV-I, is responsible for transactivation of gene expression of viral and cellular genes and is involved in the onset of adult T-cell leukemia, ATL. Genetic deletion studies have implicated a region of the HTLV-I LTR designated as tax-acceptor region, TAR, which is the target of the tax protein. Using antibodies against a tax carboxyterminal synthetic decapeptide the tax protein was purified from an HTLV-I immortalized human T-lymphocyte cell line by immunoaffinity chromatography. The tax protein, purified to apparent homogeneity binds to double-stranded DNA irrespective of its origin from either a nuclear or cytoplasmic fraction of the HTLV-I immortalized cell-line - both of which harbor similar quantities of tax protein. The tax protein binds less to single-stranded DNA and not to single-stranded RNA in vitro. It also binds to DNA-cellulose and heparin-Sepharose. Nuclease treatment of isolated nuclei does not release the tax protein under conditions which release known DNA-binding proteins, such as the myb protein. Transactivation by the tax protein presumably involves host-cell factors, since it does not recognize specific DNA sequences.

Specific association of Mil/Raf proteins with a 34 kDa phosphoprotein.

Mil/Raf protein kinases are intermediates in signaling pathways leading to differentiation, mitogenesis and cellular transformation. To gain insight into the activity of Mil/Raf kinases at the molecular level we aimed to identify proteins specifically interacting with Mil/Raf proteins. A phosphoprotein of 34 kDa (pp34) was found to be associated with c-Raf as well as with viral and activated forms of Mil/Raf proteins in exponentially growing interphase cells. pp34 association was not detectable in mitotic cells. Serum stimulation or coexpression of activated Ras led to decreased electrophoretic mobility of pp34 complexed to Mil/Raf proteins while serum starvation rendered pp34 undetectable. Moreover, the association with pp34 became undetectable in parallel with the onset of morphological cellular transformation caused by overexpression of a constitutively activated mutant of c-Raf in an inducible expression system. Thus, the association of Mil/Raf proteins with pp34 is altered in the course of cell cycle progression, serum stimulation and cellular transformation. These events represent hallmarks of cellular Mil/Raf functions, rendering pp34 a candidate protein involved in Mil/Raf function

Mapping of a small phosphopeptide at the carboxyterminus of the viral myb protein by monoclonal antibodies.

Several myb-specific monoclonal antibodies were produced and their antigen recognition sites characterized using a series of bacterially expressed truncated myb proteins. The monoclonal antibodies were used for analysing the in vivo phosphorylation site of the oncogene protein from avian myeloblastosis virus (AMV), p48v-myb. The p48v-myb protein labeled metabolically with [32P]orthophosphate was isolated from the AMV-transformed chicken myeloblast cell line BM-2 by immunoaffinity chromatography. Phosphoamino acid analysis indicated that it was phosphorylated mainly on serine and to a lesser extent (less than 5%) on threonine residues. Indirect immunoprecipitation of phosphopeptides from trypsin-digested [32P]-labeled purified p48v-myb protein by use of the myb-specific monoclonal antibodies allowed the mapping of a small phosphopeptide at the carboxyterminus of p48v-myb.

Selective DNA binding of the human cellular myb protein isolated by immunoaffinity chromatography using a monoclonal antibody.

The bacterially expressed v-myb protein served as antigen for the isolation of several monoclonal antibodies, one of which recognized the human cellular myb protein (p75hu-c-myb) indicating a conserved epitope. The epitope was mapped to amino acid positions 208-232 by the use of several bacterially expressed v-myb proteins with various deletions. Furthermore, a synthetic oligopeptide which had been selected on the basis of its hydrophilicity by computer analysis of the v-myb oncogene (amino acids 213-231) blocked the action of this monoclonal antibody, indicating the immunological significance of this region. The monoclonal antibody allowed efficient purification of the p75hu-c-myb protein by immunoaffinity chromatography. The purified protein binds to double-stranded DNA in vitro in a filter-binding assay. Since the monoclonal antibody does not interfere with DNA binding it allowed analysis of DNA-protein interaction in a modified McKay assay using the purified p75hu-c-myb protein. Specific binding was observed predominantly to one of 12 lambda DNA fragments in vitro in the presence of high molar excess of competing co-polymer poly [d(I:C)]. Enhancer/promoter-like sequences of SV40 were not preferentially recognized.

Transcriptional activation by human c-myb and v-myb genes.

The human c-myb proto-oncogene is the cellular progenitor of the viral v-myb oncogene and codes for a 75 kD protein involved in growth regulation and differentiation in a number of cells. Fusion proteins in which human c-myb sequences are linked to the DNA binding domain of the yeast transcriptional activator GAL4 can activate transcription from a reporter gene which carries the chloramphenicol acetyl transferase (CAT) gene linked in cis to a repeat of the GAL4 binding site. Deletions of carboxyterminal sequences allowed the identification of the domain responsible for transcriptional activation, which is located between amino acid residues 275 to 327. Deletion of this activator domain results in abrogation of the transcriptional activation. The GAL4-v-myb fusion protein can also activate transcription whereas no transactivation by GAL4-c-myb is observed, indicating that a carboxyterminal domain of c-myb which is absent from v-myb apparently negatively regulates transcriptional activation. Dimer formation which is required for transactivation by GAL4 fusion proteins can, when GAL4 is truncated, be mediated by a region of the c-myb protein upstream of the transactivator domain possibly including the transactivator domain itself but not a putative leucine zipper located downstream of this region.

Analysis of DNA sequences present in complexes of v-Myc and cellular DNA.

We have isolated cellular DNA sequences from the myelocytomatosis virus-transformed quail fibroblastic cell-line (MC29-Q8) by indirect immunoprecipitation of v-Myc-DNA complexes and subsequent cloning of the DNA. The v-Myc-DNA complexes were obtained from isolated nuclei pretreated with 150 mM salt and indirect immunoprecipitation of the p110 Gag-Myc protein with the IgG of a Gag-specific monoclonal antibody. A non-specific monoclonal IgG was used as control to account for non-specific interactions. The DNA from the precipitates was isolated, cloned and characterized. 21 positive and 13 control clones with inserts ranging in size from 25 to 330 nucleotides were sequenced and analyzed for sequence homologies to known DNA-motives. Some of the sequences over-represented in the specific DNA fragments corresponded to elements which have been previously described to promote DNA amplification. Six of these DNA fragments were tested for their ability to promote DNA amplification by insertion into a plasmid containing the HSV-1 tk gene with a truncated promoter. These constructs were transfected into mouse Ltk- cells which grow under HAT selection only upon amplification of the tk-carrying plasmids. Two of the six DNA fragments showed the capability to amplify their plasmids in cis and create stable cellular clones. Copy numbers of the amplified plasmids in these Ltk+ clones ranged from 470 to 680 whereby the amplified sequences were integrated as large clusters of head-to-tail tandems. The data presented here suggest that the Myc protein may be involved in DNA amplification and therefore may play a role in DNA replication.

Nuclear and cytoplasmic distribution of cellular myb protein in human haematopoietic cells evidenced by monoclonal antibody.

A monoclonal antibody was used for analysing the expression of the cellular myb (c-myb) protein in a variety of established human tumor cell lines and its decrease after induction of differentiation. Differentiated resting human T-cells and B-cells do not express detectable amounts of c-myb protein. However, upon mitogenic stimulation in vitro T-cells exhibit strong expression of the c-myb protein as demonstrated by immunocytochemical staining and indirect immunoprecipitation. In contrast to the transformed T-lymphoblastic cell line Molt-4, where c-myb protein is a nuclear antigen, it was found in proliferating normal T-cells almost exclusively distributed in the cytoplasm. Screening of a total of 70 fresh human malignant lymphomas by immunohistochemical staining indicates the presence of the c-myb protein primarily in non-Hodgkin's lymphomas with a large growth fraction, i.e. precursor cell-derived lymphoblastic lymphomas of B-cell type and T-cell type (9/10, 3/4, respectively) and anaplastic large cell Ki-1 lymphomas (5/9), which originate from activated lymphoid cells. The c-myb protein was located predominantly in the nucleus and in some cases additionally in the cytoplasm. The different subcellular locations suggest a dual functional role. While nuclear localisation is exhibited by transformed haematopoietic cells, cytoplasmic localisation appears to be characteristic for proliferating normal T-cells and points to a second property of the c-myb protein other than interaction with DNA.

Expression of p64c-myc and neuroendocrine properties define three subclasses of small cell lung cancer.

Twelve human small cell lung cancer (SCLC) cell lines and 6 non-SCLC cell lines were analysed with respect to expression of the c-myc, c-myb, and c-raf1 protooncogenes at the protein level. Analysis of p64c-myc protein expression in 12 SCLC cell lines resulted in the observation that it is present at high levels not only in cells with low, but also in those with moderate neuroendocrine differentiation. Neuroendocrine differentiation was based on parameters such as growth rate, colony formation, L-Dopa decarboxylase (DDC) activity, bombesin, and neurotensin described before. Surprisingly, in two cell lines with low neuroendocrine differentiation but without c-myc protein expression (SCLC-86M1 and NCI-H526) p75c-myb expression was observed which may therefore be able to substitute for the p64c-myc protein. Analysis of p74c-raf1 expression did not result in correlation with any growth or differentiation parameter since it was expressed at low levels in 11 out of 12 cases. We conclude that SCLC in vitro can be classified in three rather than two previously defined subclasses. In addition to the classic subclass with slow growth, high neuroendocrine differentiation, and absent or very low p64c-myc expression and the variant subclass with fast growth, absent to very low neuroendocrine differentiation, and high p64c-myc expression, we suggest a third subclass designated as transitional with moderate growth, moderate neuroendocrine differentiation, and high p64c-myc expression. Data on a small number of non-SCLC cell lines tested showed that high levels of p64c-myc correlate with high in vitro growth rates. This indicates that high p64c-myc levels may be associated with high proliferative activity, and lack of differentiation in lung cancer in general. The p74c-raf1 protein was found in all non-SCLC cell lines. Whether this classification of SCLC cell lines is applicable to SCLC in vivo remains to be determined.

Interaction of the v-and c-Myb proteins with regulatory sequences of the human c-myc gene.

Eight c-Myb-binding sites have been identified in the regulatory region of the human c-myc gene using gel retardation and DNAase I footprint assays with purified bacterially expressed full-length and carboxy-terminally truncated c-Myb proteins. These binding sites exhibit different affinities whereby strong binding correlates better with conservation of the palindromic sequences, AACXGTT or AACGTT, than the previously described consensus sequence. Flanking AT-rich sequences further increase the binding affinity. The c-Myb-binding sites are arranged in pairs consisting of one high- and one low-affinity binding site. Binding of the Myb proteins to these sites is non-cooperative. The v-Myb protein protects two nucleotides fewer than the c-Myb protein. Co-transfection of reporter CAT genes, containing upstream human c-myc sequences including exon 1, with c-Myb-expressing constructs resulted in positive transactivation, which was eightfold with full-length Myb and 14-fold with the truncated Myb. This result suggests that the Myb protein could participate in regulation of human c-myc gene expression.