Advertisement

COT Kinase Proto-oncogene Expression in T Cells

IMPLICATION OF THE JNK/SAPK SIGNAL TRANSDUCTION PATHWAY INCOT PROMOTER ACTIVATION*
      COT/Tpl-2 proto-oncogene encodes a serine/threonine kinase implicated in cellular activation. In this study we have identified the human COT gene promoter region and three different human COT transcripts. These transcripts, with the same initiation site, display heterogeneity in their 5′ untranslated regions and in their subcellular localization. Activation of Jurkat T cells with either calcium ionophore A23187 or αCD3 and a phorbol ester increases the levels of the different COT transcripts. Analysis of the 5′ flanking region of the human COT gene reveals a unique transcription initiation site and a TATA element 20 nucleotides upstream. Transient expression of COT promoter constructs containing a reporter gene indicates that the transcriptional activity of the 5′ flanking region of the COT gene is regulated by T cell-activating signals. Cotransfection of a dominant negative version of SEK-2 abolishes the inducible transcriptional activity ofCOT promoter, indicating that the inducible expression of the COT gene by T cell activating signals is mediated by the JNK/SAPK signal transduction pathway. All these data indicate stringent regulation of COT kinase proto-oncogene expression.
      MEK-1
      mitogen-activated protein kinase kinase-1
      SEK-1
      mitogen-activated protein kinase kinase-4
      ERK
      extracellular signal-regulated kinase
      JNK
      c-Jun NH2-terminal kinase
      SAPK
      stress-activated protein kinase
      UTR
      untranslated region
      nt
      nucleotide(s)
      kb
      kilobase(s)
      PCR
      polymerase chain reaction
      PDBu
      phorbol 12,13-dibutyrate
      8-Br-cAMP
      8-bromoadenosine 3′:5′-cyclic monophosphate
      RT
      reverse transcription
      bp
      base pair(s)
      COT/Tpl-2 is homologous to members of the mitogen-activated protein kinase kinase kinase family (MAP3K) and has been implicated in cellular activation (
      • Salmerón A.
      • Ahmad T.B.
      • Carlille G.W.
      • Pappin D.
      • Narsimham R.P.
      • Ley S.C.
      ,
      • Fanger G.R.
      • Gerwins P.
      • Widmann C.
      • Jarpe M.B.
      • Johnson G.L.
      , ,
      • Patriotis C.
      • Makris A.
      • Bear S.E.
      • Tsichlis P.N.
      ,
      • Robinson M.J.
      • Cobb M.H.
      ). Overexpression of COT/Tpl-2 induces activation of MEK-11and SEK-1 kinases, activating the ERK and JNK/SAPK signal transduction pathways, respectively (
      • Salmerón A.
      • Ahmad T.B.
      • Carlille G.W.
      • Pappin D.
      • Narsimham R.P.
      • Ley S.C.
      ,
      • Patriotis C.
      • Makris A.M.
      • Chernoff J.
      • Tsichlis P.N.
      ,
      • Troppmair J.
      • Bruder J.T.
      • Munoz H.
      • Lloyd P.A.
      • Kyriakis J.
      • Banerjee P.
      • Avruch J.
      • Rapp U.R.
      ). COT/Tpl-2 kinase regulates the transcription of tumor necrosis factor α and interleukin 2 genes during T cell activation (
      • Ballester A.
      • Calvo V.
      • Tobeña R.
      • Lisbona C.
      • Alemany S.
      ,
      • Ballester A.
      • Velasco A.
      • Tobeña R.
      • Alemany S.
      ) by activating at least AP-1 and NF-κB response elements in the gene promoters (
      • Ballester A.
      • Velasco A.
      • Tobeña R.
      • Alemany S.
      ,
      • Tsatsanis C.
      • Patriotis C.
      • Tsichlis P.N.
      ,
      • Tsatsanis C.
      • Patriotis C.
      • Bear S.E.
      • Tsichlis P.N.
      ,
      • Belich M.P.
      • Salmerón A.
      • Johnston L.H.
      • Ley S.C.
      ,
      • Lin X.
      • Cunningham Jr., E.T.
      • Mu Y.
      • Geleziunas R.
      • Greene W.C.
      ).
      Several COT/Tpl-2 cDNAs comprising the complete coding sequence and 3′ UTR have been reported: two different human cDNAs (GenBankTM accession numbers Z14138 and D14497) (
      • Chan A.M-L.
      • Chedid M.
      • McGovern E.S.
      • Popescu N.C.
      • Miki T.
      • Aaronson S.A.
      ,
      • Aoki M.
      • Hamada F.
      • Sugimoto T.
      • Sumida S.
      • Akiyama T.
      • Toyoshima K.
      ), two rat cDNAs (GenBankTM accession numbers M94454and L15358) (
      • Makris A.
      • Patriotis C.
      • Bear S.E
      • Tsichlis P.N.
      ), and one murine COT cDNA (GenBankTM accession number D13759) (
      • Ohara R.
      • Miyoshi J.
      • Aoki M.
      • Toyoshima K.
      ). Identities between human, rat, and murine COT cDNAs in their coding sequences and 3′ UTRs are 85 and 75%, respectively. The 5′ UTR of the human, rat, and murine COT cDNAs did not reveal any homology, with the exception of the 23 nt upstream from the first COT/Tpl-2 ATG codon.
      The human COT gene is a single copy locus (
      • Chan A.M-L.
      • Chedid M.
      • McGovern E.S.
      • Popescu N.C.
      • Miki T.
      • Aaronson S.A.
      ,
      • Miyoshi J.
      • Higashi T.
      • Mukai H.
      • Ohuchi T.
      • Kakunaga T.
      ) localized on the short arm of chromosome 10 at band p11.2 (
      • Chan A.M-L.
      • Chedid M.
      • McGovern E.S.
      • Popescu N.C.
      • Miki T.
      • Aaronson S.A.
      ). Ohara et al. (
      • Ohara R.
      • Miyoshi J.
      • Aoki M.
      • Toyoshima K.
      ) proposed that the human COT gene contains nine exons, of which the last seven are coding exons. COT kinase was first identified in a truncated form in transformed foci induced in SHOK cells by transfection of the genomic DNA of a human thyroid carcinoma cell line (
      • Miyoshi J.
      • Higashi T.
      • Mukai H.
      • Ohuchi T.
      • Kakunaga T.
      ,
      • Aoki M.
      • Akiyama T.
      • Miyoshi J.
      • Toyoshima K.
      ). This rearrangement occurs in the penultimate coding exon and provides transformation capacity (
      • Aoki M.
      • Hamada F.
      • Sugimoto T.
      • Sumida S.
      • Akiyama T.
      • Toyoshima K.
      ). The rat homologue of the human COT gene (Tpl-2) was identified as an oncogene associated with the progression of Moloney murine leukemia virus-induced T cell lymphomas in rats (
      • Patriotis C.
      • Makris A.
      • Bear S.E.
      • Tsichlis P.N.
      ,
      • Makris A.
      • Patriotis C.
      • Bear S.E
      • Tsichlis P.N.
      ). As with the human and murine homologues, the disruption of the last coding exon of Tpl-2 by insertion of the Moloney murine leukemia virus enhances mRNA levels and appears to unmask the oncogenic potential of the protein (
      • Patriotis C.
      • Makris A.
      • Bear S.E.
      • Tsichlis P.N.
      ,
      • Aoki M.
      • Hamada F.
      • Sugimoto T.
      • Sumida S.
      • Akiyama T.
      • Toyoshima K.
      ,
      • Erny K.M.
      • Peli J.
      • Lambert J-F.
      • Muller V.
      • Diggelmann H.
      ,
      • Ceci J.D.
      • Patriotis C.P.
      • Tsatsanis C.
      • Makris A.M.
      • Kovatch R.
      • Swing D.A.
      • Jenkins N.A.
      • Tsichlis P.N.
      • Copeland N.G.
      ). It has been suggested recently that an amplification of the genomic locus of the COT gene plays a role in human breast cancer (
      • Sourvinos G.
      • Tsatsanis D.A.
      • Spandidos D.A.
      ).
      In this paper we have studied the expression of the humanCOT gene in T cells. We have identified three different human COT mRNAs and the 5′ region flanking the transcription initiation site of the COT gene. We also provide evidence that T cell activation up-regulates the levels of the three COT kinase mRNAs and increases the transcriptional activity of the human COT gene promoter through the JNK/SAPK signal transduction pathway.

      EXPERIMENTAL PROCEDURES

       Primers

      Primers located 3′ from the transcription initiation site of the COT gene (+1 nt) have been designated according to their location in COT-1 cDNA: −8D (−1082/−1060 nt), 5′-CTGCA AAAATAAGTGAAAGTGAC; −7D (−778/−758 nt), 5′-GCTTTCATCAGGTTGACTGATGTCA; −6D (−723/−704 nt), 5′-AGAAGGATTCAGAGGTCAGA; −5D (−650/−631 nt), 5′-TTGGGGAGTTTTTCTAACTC; −4D (−66/−448 nt), 5′-GGGATGGAGAGGTAAGCAT; −3D (−340/−318 nt), 5′-ATTGTGCAGACAT TGATTCATTT; −2D (−229/−210 nt), 5′-CTCCACCATATGATTCTAAT; −1D (−102/−83 nt), 5′-AGCT TGGCAAAACTTCTTCA; 1D (122/141 nt), 5′-CCAGCATCGCACCGAAACCTT; 2D (589/609 nt), 5′-CCCGATCCTCCCAAATGCTGG; 3D (107/1090 nt), 5′-CCTGAGTCGTTTGTGCCAAG; 4D (1731/1752 nt), 5′-CATTTTCATTCACACTTGCCAG; 5D (2237/2256 nt), 5′-TCAGCGTCAGACACTCCTCC; 6D (2989/3008 nt), 5′-ATCTTCTTACCGCGAAGAAG; 7D (3984/4003 nt), 5′-CAAGATGTTTGCTTTGCACTA; 8D (4643/4662 nt), 5′-CCAATCCTTGTATGTCAGTT; 9D (4749/4769 nt), 5′-ATGGAGTACATGAGCACTGGA; UpTATA (−49/−30 nt), 5′-CTTCTTGTCACATAGCCCAG; PR (98/115 nt), 5′-GCCTGTGGGAGCCGAGCA; PE (123/140 nt), 5′-AGGTTCGGTGCGATGCTG; 1R (588/607 nt), 5′-AGCATTTGGGAGGATCGGGT; 2R (1070/1090 nt), 5′-CTTGGCACAAACGACTCAGGC; 3R (1731/1752 nt), 5′-CTGGCAAGTGTGAATGAAAATG; 4R (2762/2784 nt), 5′-ATTTGACTTCGGTTTTACTGAGC; 5R (3006/3033 nt), 5′-AAGATGTGGCTACCTATTCCCCTGGCTT; 6R (3767/3790 nt), 5′-GAGTAACAACGTCAGTTTTTACCC; 7R (4932/4951 nt), 5′-TTGGCCACTAAGCAGCAGAG; 8R (5133/5152 nt), 5′-GAGAACATCGGAATCTATT; 9R (6206/6229 nt), 5′-GGCCCCTGTGTAGAGGCAGCAGAA; 5′β-actin, 5′AGCACAATGAAGATCAAGAT; and 3′β-actin, 5′TGTAACGCAACTAAGTCATA.

       DNA Isolation and COT Promoter-Luciferase Reporter Vector Constructs

      Genomic DNA was obtained as described previously (
      • Blin N.
      • Stafford D.V.
      ). A 6.1-kb DNA fragment containing the 5′ region flanking the translation initiation site of COT kinase was obtained with the human genomic DNA PromoterFinderTM DNA walking kit (CLONTECH), using two specific reverse primers deduced from the sequence of the first COT coding exon. Different DNA fragments of the 5′ flanking region of the COTgene were generated by PCR with different direct primers (−8D, −7D, −6D, −5D, −4D, −3D, −2D, and −1D) and PR as reverse primer. The −778Δ PCR product was performed with −7D and upTATA primers. Purified PCR products were cloned in pMOS Blue T-vector (Amersham Pharmacia Biotech). Single clones were selected, and their sequences were compared with the original template. From these constructs, the KpnI/HindIII fragments were cloned in the pGL3-Basic Luc-reporter vector (Promega). Sequencing, using specific oligonucleotides and Sequenase (U. S. Biochemical Corp.), was performed by the Sanger method (
      • Sanger F.
      • Niksen S.
      • Coulson A.R.
      ). Gene Jockey II, DNAstrider 1.1, MacPattern Folder, and Blast programs were used to analyze the DNA sequences.

       Cells, Transient Transfection Analysis, and Polysome Gradients

      Human leukemia T Jurkat cells were electroporated as previosuly described (
      • Ballester A.
      • Velasco A.
      • Tobeña R.
      • Alemany S.
      ) with 20 μg of the different pGL3-Luc constructs. After a 2-h incubation, cells were stimulated for 12 h with different stimuli: soluble αCD3 (10 μg/ml, obtained as indicated in Ref.
      • Ballester A.
      • Velasco A.
      • Tobeña R.
      • Alemany S.
      ), calcium ionophore A23187 (0.25 μm, Sigma), and/or PDBu (50 ng/ml, Roche Molecular Biochemicals). Cyclosporin A (100 ng/ml), 8-Br-cAMP (0.5 mm, Roche Molecular Biochemicals), PD 98059 (MEK inhibitor) (20 μg/ml, Calbiochem), or SB-20580 (HOG/p38 mitogen-activated protein kinase inhibitor) (20 μg/ml, SmithKline Beecham), and/or okadaic acid (100 ng/ml, Calbiochem) were added 30 min before stimulation. Luciferase activity and protein concentration were measured as described previously (
      • Ballester A.
      • Velasco A.
      • Tobeña R.
      • Alemany S.
      ).
      To perform the polysome fractionation, 8 × 107 cells were lysed as described previously (
      • Nakiely S.
      • Dreyfuss G.
      ) and centrifuged at 10,000 × g for 10 min at 4 °C. The supernatant was spun in a 10–50% linear sucrose gradient buffered with 20 mm HEPES (pH 7.3), 250 mm KCl, 20 mm MgCl2, 2 mm dithiothreitol, and 500 μg/ml heparin at 36,000 rpm for 2 h at 4 °C in a SW 41 rotor (Beckman). Fractions of 1 ml were collected, and ethanol was precipitated. RNA was extracted from the pellets by using the SV total RNA isolation system kit (Promega).

       Northern Blot and Dot Blot Analysis

      By using Ultraspec (Biotecx Laboratories), 20 μg of total RNA was extracted from Jurkat cells stimulated with 50 ng/ml PDBu, 0.25 μmcalcium ionophore, and 100 nm okadaic acid for different times. RNA was electrophoresed and blotted onto a nylon membrane (Nytran, NY 13N, Renner GmbH). Filters were hybridized with a random primer-labeled cDNA (>109 cpm/μg) comprising the entire human COT open reading frame (
      • Chan A.M-L.
      • Chedid M.
      • McGovern E.S.
      • Popescu N.C.
      • Miki T.
      • Aaronson S.A.
      ). Membranes were exposed to x-ray film for 10 days at −70 °C.
      Multiple tissue human poly(A)+ RNA Northern blot (CLONTECH) was hybridized with different probes as described above. Stripping was performed by boiling the membrane for 30 min in 0.1 × SSC and 0.5% SDS. Control autoradiographs were performed after each stripping.
      Dot blots performed with total RNA and RNA isolated from different subcellular fractions (
      • Nakiely S.
      • Dreyfuss G.
      ) were hybridized with the COTcoding probe, with probe C (3837–4646 nt), and with the COTpromoter probe (−778 to −30 nt). The autoradiograph was exposed for 7 days at −70 °C.

       RT-PCR and Primer Extension Assays

      Two μg of total RNA treated with 2.5 units of DNase I (Life Technologies, Inc.) as in Ref.
      • Ballester A.
      • Velasco A.
      • Tobeña R.
      • Alemany S.
      were subjected to the RT reaction (
      • Ballester A.
      • Velasco A.
      • Tobeña R.
      • Alemany S.
      ). One μl of this reaction was used for PCR analysis. Control samples treated with DNase I and not exposed to RT were also subjected to PCR. To perform the RT-PCRs from leukocyte mRNA, 1 μl of human leukocyte Marathon-ready cDNA library (CLONTECH) was used. Controls without a template, with genomic DNA as a template, and with one single primer were performed.
      Seven pmol of the 5′-end-labeled oligonucleotide PE, complementary to nt 123–140 of the three COT transcripts, was annealed to 7 μg of poly(A)+ RNA from PDBu- and calcium ionophore-stimulated Jurkat cells, and a RT reaction was carried out. After synthesis, nucleic acids were precipitated, and reaction products were analyzed on 6% polyacrylamide gels containing 8 murea.

       RNase Protection and S1 Nuclease Protection Assays

      A 434-nt labeled RNA probe (50–80 × 104 cpm/fmol), of which 409 nt were complementary to nt 5820–6229 of COT-1that corresponds to a region of the coding sequence, was obtained using the Maxiscript kit (Ambion). RNase A protection assay was performed with the radioimmune precipitation buffer II kit (Ambion). Dried gels were exposed to x-ray films at −70 °C for 2 days. For the S1 nuclease analysis a radiolabeled DNA probe (probe D (589–2784 nt), 10–30 × 105 cpm/fmol) was obtained using the Prime-A-Probe kit (Ambion) according to the manufacturer's instructions, except that 0.9% alkaline agarose gels were used to purify the labeled probe. Nuclease S1 protection assay was performed with the Multi-NPA kit (Ambion). Agarose gels (0.9%) were exposed to x-ray films at −70 °C for 4 days.

      RESULTS

       Identification of Three Human COT Transcripts

      Hybridization of a leukocyte poly(A)+ RNA Northern blot with a codingCOT probe yielded two hybridization signals at 3.0 and 7.3 kb (Fig. 1 A). To investigate whether the occurrence of these transcripts was due to the different length of the 5′ UTR, we cloned and sequenced a 6.1-kb DNA genomic fragment containing the 5′ flanking region of the COTtranslation initiation site. Several probes from this region were generated by PCR (Fig. 1 A). Analysis of the poly(A)+ RNA Northern blot with these probes revealed a hybridization signal only at 7.3 kb (Fig. 1 A), indicating that this COT mRNA species has a large 5′ UTR.
      Figure thumbnail gr1
      Figure 1Identification of three COTtranscripts. A, human leukocyte poly(A)+ RNA Northern blot was hybridized with the coding sequence of COT kinase and with three different probes from the 5′ flanking region of the COT translation initiation site. A schematic representation of the localization of these probes in the genomic DNA fragment is also shown. According to the numbering given to the sequence of the COT-1 transcript, probe A corresponds to nt 290–608, probe B to nt 2234–2784, and probe C to nt 3837–4646. These probes were generated by PCR. The coding COT probe was generated by PCR with primers 9D and 9R and reverse transcribed mRNA from Jurkat cells as a template. B, PCRs from cDNA of Jurkat cells were performed with the indicated primers. Diagrams show the positions of the primers in the different humanCOT cDNAs. M, molecular weight markers. C, S1 nuclease protection analysis ofCOT-1 mRNA in Jurkat cells with probe D (nt 589–2784). RNAs (30 μg) from Jurkat T lymphocytes (T) or yeast (Y) were used. Undigested probe (U.P.) was also electrophoresed. D, relative levels ofCOT-1/COT-2 and COT-2/COT-3transcripts were measured by RT-PCR in different human tissues.Lanes: 1, liver; 2, pancreas;3, muscle; 4, peripheral blood leukocytes; 5, lung; and 6, kidney. Relative levels of COT-1/COT-2 were determined using the primers 6D (0.5 μm), 8D (0.5 μm), and 8R (1 μm). The 630- and 565-bp PCR products correspond to COT-2 and COT-1 transcripts, respectively. Relative levels of COT-2/COT-3transcripts were detected with primers 1D (0.5 μm), 6D (0.5 μm), and 7R (1 μm). The 527- and 299-bp PCR products correspond to COT-2 andCOT-3, respectively.
      To determine the sequence of the 5′ UTR of the different COTmRNAs, RT-PCR analysis was performed. The direct and reverse primers used were deduced from the sequence of the 6.1-kb genomic fragment containing the 5′ flanking region of the COTtranslation initiation site. PCR of overlapping fragments was performed using as a template reverse transcribed mRNA of Jurkat cells. Different PCRs were performed with the combinations of each direct primer and all the different 3′-located reverse primers. Control samples treated with DNase I and not exposed to RT were also subjected to PCR. Controls without a template, with genomic DNA as a template, and with one single primer were also performed. The different PCR products obtained (Fig. 1 B) were analyzed by restriction mapping and by sequencing (data not shown). The same overlapping PCR products were obtained when human leukocyte cDNA was used as a template (data not shown). This analysis revealed that theCOT gene is transcribed with three different 5′ UTRs. The transcription start site of these three 5′ UTRs was delimited to the same 30-nt region by PCR analysis, using as a template cDNA from Jurkat cells as well as human leukocyte cDNA (data not shown). (See “Determination of the Transcription Start Site of the Human COT Gene” for the location of the exact start transcription site.) To establish the 5′ UTR of COT-1 by a method other than RT-PCR, S1 nuclease analysis was performed. A DNA probe (probe D) complementary to the 589–2784-nt sequence of the 5′ UTR ofCOT-1 was hybridized with RNA obtained from Jurkat cells and incubated with nuclease S1 (Fig. 1 C). This probe contains the entire sequence of probe B and is extended to the DNA sequence of probe A. The intron/exon boundaries of the three different 5′ UTRs of COT transcripts are shown in TableI.
      Table IIntron-exon organization of the 5′ region of the human COT gene
      mRNAExonSizentSequence at intron-exon junction
      Splice acceptorSplice donor
      nt
      COT-1150841–5084ATTAAACATGgttagtttct
      COT-21a1921–192AAGGCCGCAGgtaatccagg
      1b2292984–3212ttgttttagATGCAATCTTCACCTCATGAGgtaggtgctg
      1c3594726–5084tctttcctagACTCTCCAGATATTAAACATGgttagtttct
      COT-31a1921–192AAGGCCGCAGgtaatccagg
      1c3594726–5084tctttcctagACTCTCCAGATATTAAACATGgttagtttct
      Positions have been designated according to their location in the COT-1 transcript. Exon sequences are shown in uppercase letters; intron sequences are shown in lowercase letters.
      We also investigated the possibility of alternative splicing in the coding sequence and 3′ UTR of COT transcripts. RT-PCR analysis revealed no alternative splicing in these regions. The coding sequence and 3′ UTR region of COT transcripts have a size of 2.5 kb (data not shown). Considering the length of the 5′ UTR of the three COT transcripts, the 7.3-kb COTmRNA species detected in the leukocyte poly(A)+ RNA Northern blot (Fig. 1 A) should correspond toCOT-1, and the 3.0-kb signal should correspond toCOT-2 and COT-3. The three COTtranscripts were detected by RT-PCR in all human tissues tested (Fig.1 D), indicating that none of the different COTtranscripts is tissue-specific, although the relative amounts seem to vary between the different tissues. When the PCRs were performed as described in the legend to Fig. 1 D, the ratio ofCOT-2 to COT-1 oscillated between 0.5 for liver or pancreas and 1.6 for muscle. The ratio of COT-3 toCOT-2 varied from 9.8 for lung to 1.2 for pancreas.

       Determination of the Transcription Start Site of the Human COT Gene

      The transcription start site of the COT gene was delimited by RT-PCR analysis to a 30-nt region (data not shown). The exact transcription start site of the COT gene was determined by primer extension on poly(A)+ RNA from stimulated Jurkat cells with a primer complementary to primer 1D (PE primer). This sequence is complementary to the three COTtranscripts. As shown in Fig. 2, a single product corresponding to a 140-base extended fragment was detected. The first transcribed base has been designated +1, to facilitate numbering of the different COT transcripts. Sequence analysis revealed a putative TATA box located at position −20 nt (Figs. 2 and 3 A), which is in agreement with the preferred position occupied by this element in a typical eukaryotic promoter (
      • Bucher P.
      ).
      Figure thumbnail gr2
      Figure 2Determination of the transcription start site of the human COT gene. A, a32P end-labeled primer complementary to nt 123–140 of human COT cDNAs was annealed to poly(A)+ RNA of Jurkat cells stimulated for 6 h with PDBu (50 ng/ml) and calcium ionophore A23187 (0.25 μm) (lane 2) or to poly(A)+ RNA from Dictyostelium discoideum (lane 1) and extended with reverse transcriptase. The products were analyzed in parallel with sequencing reactions (ACGT) carried out on a genomic PCR product using the same primer. The arrow shows the position of the extended product. The sequences of the sense strand near the band and the TATA box are shown below.
      Figure thumbnail gr3
      Figure 3Nucleotide sequence and putative regulatory elements of the 5′ flanking region of the human COTgene and determination of its transcriptional activity. A, the sequence is numbered relative to the transcription start site, which is referred to as +1. The putative regulatory elements are indicated in bold letters belowunderlined sequences. B, Jurkat cells were transfected with the pGL3, pGL3–778 (5′-3′), pGL3–778Δ, and pGL3–778 (3′-5′) constructs (20 μg/0.8 ml). The graph shows the value of LRU/mg of protein (prot.) performed three times in duplicate. A schematic representation of the different constructs is also shown. The nomenclature of the different constructs is denoted relative to the transcription initiation site (position +1).
      To confirm that the DNA region 5′ flanking the defined transcription start site of the human COT gene has promoter activity, transient expression experiments with the pGL3-Luc basic vector linked to different fragments of this DNA region were carried out. Jurkat cells were transfected with pGL3, pGL3–778 (5′-3′), and pGL3–778Δ as well as with the pGL3–778 (3′-5′) construct, and luciferase activity was measured (Fig. 3 B). The pGL3–778 (5′-3′) construct, containing nt −778 to +115 of the COTgene, exhibited a transcriptional activity 20-fold higher than that of the empty pGL3 vector. The −30 to +115-nt region is essential in maintaining this increase, because deletion of the −30 to +115-nt fragment, in the pGL3–778Δ construct, resulted in transcriptional activity similar to pGL3. No promoter activity over background levels was detected with the pGL3–778 (3′-5′) construct, indicating that this region contains the transcription start site of the COT gene and not only cis-response elements.

       Regulation of Human COT Promoter Activity by T Cell-Regulating Signals

      To further study the transcriptional activity of the 5′ flanking region of theCOT gene, transient expression experiments were conducted in Jurkat cells with pGL3 plasmids containing 1082 bp of theCOT gene 5′ flanking region (Figs. 3 and4 A) as well as 5′ deletion fragments (Fig. 4 A).
      Figure thumbnail gr4
      Figure 4Regulation of the human COTpromoter-luciferase construct activity by T cell stimulatory signals. A, a schematic representation of theCOT promoter-based reported gene constructs. B, Jurkat cells were transfected with the different constructs and 2 h later were stimulated or not with PDBu (50 ng/ml) and calcium ionophore A23187 (0.25 μm) or with PDBu (50 ng/ml) and soluble αCD3 (10 μg/ml). The graph shows the mean of fold induction of four different experiments performed in duplicate. A value of 1 was given to that obtained with the basic COT promoter (pGL3–102) under no stimulation conditions. C, Jurkat cells were transiently transfected with the pGL3–340 (15 μg) or pGL3–778 (15 μg) constructs together with pscαDN-SEK-2 (10 μg) or with empty vector (10 μg), and cells were stimulated with PDBu (50 ng/ml) and calcium ionophore A23187 (0.25 μm) or with PDBu (50 ng/ml) and soluble αCD3 (10 μg/ml). The graph shows the mean of the LRU/mg of protein (prot.) values of three experiments performed in triplicate. D, cells were transfected with pGL3–340 COT promoter construct and activated with calcium ionophore A23187 and/or PDBu. Cyclosporin A (100 ng/ml), MEK inhibitor (MEK I, 20 μg/ml), HOG inhibitor (HOG I, 20 μg/ml), okadaic acid (Oka, 100 ng/ml), and 8-Br-cAMP (cAMP, 0.5 mm) were added to the incubation media 30 min prior to activation with calcium ionophore A23187 (0.25 μm) and PDBu (50 ng/ml). The graph shows the mean fold induction of four experiments performed in duplicate.
      Comparison of the relative promoter activities of the different constructs indicated that the progressive removal of 5′ sequences up to −650 did not significantly affect the COT promoter activity in unstimulated cells (Fig. 4 B). Deletion of the −466 to −650 DNA fragment significantly decreased the transcriptional activity. Further deletion of nt −340 to −229 further decreased the activity of the COT promoter (Fig. 4 B). However, the pGL3–229 and pGL3–102 constructs still exhibited a transcriptional activity 7-fold higher than vector pGL3 (data not shown).
      To determine whether COT promoter activity is regulated by T cell-activating signals, Jurkat cells were transfected with the different COT promoter-derived constructs and stimulated with PDBu (50 ng/ml) and calcium ionophore (0.25 μm) or with αCD3 (10 μg/ml) and PDBu (50 ng/ml). Comparison of the luciferase activities produced by each different plasmid in unstimulated and stimulated cells showed that deletion of the sequences located between positions −1082 to −340 did not significantly affect the 3-fold induction relative to the unstimulated activity of each construct. Further removal of the sequences from position −340 to −229 abolished the 3-fold induction by these signals (Fig.4 B).
      One AP-1 binding site (
      • Angel P.
      • Imagawa M.
      • Chiu R.
      • Stein B.
      • Imbra R.
      • Rahmsdorf H.J.
      • Jonat C.
      • Herrlich P.
      • Karin M.
      ,
      • Lee W.
      • Mitchell P.
      • Tjian R.
      ) is found at position −327 nt of the 5′ flanking region of the COT gene (Fig. 3 A). The AP-1 transcription factor is up-regulated in T lymphocytes activated with PDBu and calcium ionophore or with αCD3 and PDBu, and the JNK/SAPK signal transduction pathway mediates its activation. To determine whether this signal transduction pathway regulates, at least in part, activation of the COT promoter, Jurkat cells were cotransfected with pGL3–340 or pGL3–788 together with the dominant negative form of JNK kinase, DN-SEK-2 (MKK7-KL), that inhibits the activation of c-Jun. Cells were stimulated or not with PDBu (50 ng/ml) and calcium ionophore (0.25 μm) or with αCD3 (10 μg/ml) and PDBu (50 ng/ml). As shown in Fig. 4 C, the expression of the DN-SEK-2 abolished the increase of the promoter-driven transcription of the pGL3–340 and pGL3–778 constructs by T-cell activating signals.
      To further analyze the signal transduction mechanism by which theCOT promoter is activated, transient transfection experiments with the pGL3–340 construct were performed. Addition of PDBu or calcium ionophore by itself did not increase the luciferase activity, indicating that an integration of both signals has to occur to activate COT promoter-driven transcription (Fig.4 D). Transfected cells were also incubated with different inhibitors or activators of protein kinases or protein phosphatases. The transfected cells were incubated with PDBu (50 ng/ml) and calcium ionophore (0.25 μm) in the presence or absence of cyclosporin A (100 ng/ml), MEK inhibitor (20 μm), HOG inhibitor (20 μm), okadaic acid (100 ng/ml), or 8-Br-cAMP (0.5 mm) at doses that have already been reported to regulate the activation of Jurkat T cells (
      • Ballester A.
      • Velasco A.
      • Tobeña R.
      • Alemany S.
      ). Because addition of PDBu and calcium ionophore to Jurkat T cells increases the phosphorylation state of many proteins involved in the signal transduction mechanism, the addition of okadaic acid, an inhibitor of protein phosphatases 1 and 2A, to these activated Jurkat cells could induce a further increase in the phosphorylation state of the proteins. According to the results obtained in Fig. 4 D, the addition of okadaic acid did not increase the luciferase activity. Addition of cyclosporin A prior to activation of the cells reduced the luciferase activity, indicating that calcineurin (protein phosphatase 2B) is at least partially involved in the activation of the COT promoter. MEK inhibitor, which blocks the ERK signal pathway, and HOG/p38 mitogen-activated protein kinase inhibitor hardly diminished the stimulatory signal of PDBu and calcium ionophore. Activation of cAMP-dependent protein kinase by the addition of 8-Br-cAMP increased the luciferase activity by about 1.7-fold (Fig.4 D).

       Up-regulation of COT mRNA Levels

      An RNase protection assay was performed to determine whether the increase in the transcriptional activation of the 5′ flanking region of the humanCOT gene by T cell activating signals correlates with an increase in COT mRNA levels after T lymphocyte stimulation. A riboprobe from the COT coding sequence was hybridized with RNA isolated from Jurkat cells stimulated with soluble αCD3 (10 μg/ml) and PDBu (50 ng/ml) for different times. As shown in Fig. 5 A, αCD3 and PDBu stimulation transiently increased COT mRNA levels (∼4-fold). A similar increase in the level of COTtranscripts was detected by RT-PCR analysis of mRNA of Jurkat cells stimulated with PDBu and calcium ionophore for different times, using primers that amplified a COT coding sequence fragment. As a control, a 187-bp fragment of β-actin was also amplified in each reaction (Fig. 5 B).
      Figure thumbnail gr5
      Figure 5Induction of total COTmRNA levels after Jurkat T cell activation. A, RNase A protection analysis of COT mRNA levels in Jurkat cells stimulated with PDBu (50 ng/ml) and αCD3 (10 μg/ml). RNAs from Jurkat cells stimulated for different times fromSaccharomyces cerevisiae (Y) andD. discoideum (D. d.) were used for the RNase protection assay, using COT and β-actin fragments as probes. Undigested probes (U.P1 and U.P2) were also electrophoresed. The histogram represents the relative value of dividing the absorbance of the COT Rnase-protected product by the absorbance of the β-actin (β-Act) Rnase-protected product at the indicated time. B, quantitative RT-PCR analysis of the total COT mRNA of Jurkat cells stimulated for different times with calcium ionophore A 23187 (0.25 μm) and PDBu (50 ng/ml). PCRs were performed with primers 9D (1 μm) and 8R (1 μm) and β-actin primers (0.1 μm) (β-actin product, 187 bp). The figure shows one of the three experiments performed. The histogram represents the relative value of dividing the absorbance of the COT PCR product by absorbance of the β-actin PCR product at the indicated time, expressed as the mean ± S.D. from three different experiments.
      To distinguish COT-1 from COT-2 andCOT-3, a Northern blot with total RNA from Jurkat cells stimulated for different times with PDBu (50 ng/ml), calcium ionophore (0.25 μm), and okadaic acid (100 nm) was hybridized with the coding COT probe (Fig. 6 A). Whereas a hybridization signal of 3.0 kb, corresponding to COT-2 andCOT-3, was detected 3 h after stimulation, theCOT-1 message was first detected 6 h after stimulation. In agreement with the Northern blot analysis, COT-1transcript levels determined by RT-PCR were not increased at 4 h after stimulation of Jurkat cells with the stimuli described above (Fig. 6 B). At this time of stimulation, COT-2 andCOT-3 levels were increased to an equal extent (Fig.6 B).
      Figure thumbnail gr6
      Figure 6Induction of COT-1,COT-2, and COT-3 mRNA levels in activated Jurkat cells. A, Northern blot analysis of the kinetics of induction of COT mRNAs isolated from Jurkat cells stimulated with PDBu (50 ng/ml), calcium ionophore (0.25 μm), and 100 nm okadaic acid, using the coding sequence of COT kinase as a probe. Quantification of the RNA loading was performed by methylene blue staining of the membranes after transfer for detection of the 28 S and 18 S RNAs. B, RT-PCRs from RNA of Jurkat cells stimulated for 4 h with the stimuli (St) described above or not (C). The PCRs were performed with the different COT primers indicated in the diagrams, at a concentration of 1 μm. In each reaction tube a β−actin fragment was also amplified using primers at a final concentration of 0.1 μm. The figures show one representative experiment of the four performed. The histogram represents the relative value of dividing the absorbance of the COT-1, COT-2, and COT-3 PCR products by the absorbance of the β-actin (β-Act) PCR product, expressed as the mean ± S.D. of the four different experiments.

       Subcellular Distribution of COT Transcripts

      Next, we decided to investigate the subcellular distribution of COT-1,COT-2, and COT-3. RNA was isolated from the cytoplasmic fraction and the nuclear fraction of intact Jurkat cells. A dot blot performed with these RNAs was hybridized with a probe specific for the 5′ UTR of COT-1 (probe C), theCOT coding probe, and a probe containing the −778 to −30 nt sequence of the COT promoter (Fig.7 A). Comparison of the hybridization signals obtained with the different probes and RNA fractions indicated that the COT-1 transcript is mainly located in the nuclear fraction.
      Figure thumbnail gr7
      Figure 7Distribution of COTtranscripts. A, a dot blot performed with total RNA, RNA isolated from the cytoplasmic fraction, and RNA isolated from the nuclear fraction was hybridized with probe C (nt 3837–4646), theCOT coding probe, and a probe generated from theCOT promoter region (nt −778 to −30). B, distribution of COT transcripts in polysomes. The figure shows the relative absorbance corresponding to the different sucrose fractions. The 28 S and 18 S RNAs from 6 μl of each fraction were separated on a 2.2 m formaldehyde-agarose denaturing gel and photographed after ethidium bromide staining. RT-PCR analysis of the different fractions was performed. COT-1 levels were determined with primers 8D (1 μm) and 8R (1 μm), COT-2 levels were determined with primers 6D (1 μm) and 7R (1 μm), and COT-3 levels were determined with primers 1D (1 μm) and 7R (1 μm). Primers for β-actin detection were used at a concentration of 0.1 μm.
      We next decided to study the distribution of the differentCOT messengers to polysomes. The postmitochondrial supernatant of Jurkat cells was subjected to sucrose gradient fractionation, and RNA was isolated from the different fractions. The levels of the different COT transcripts were measured by RT-PCR analysis. As shown in Fig. 7 B, the fraction ofCOT-1 located in the cytoplasmic fraction is not associated with polysomes. In addition, only a small fraction of COT-2was not loaded with ribosomes. The fact that both COT-2 and COT-3 were detected in fractions corresponding to small polysomes indicates a low translation efficiency of these transcripts. As a control of polyribosome-associated mRNA, we assayed the same fractions for β-actin messenger (Fig.7 B). Stimulation of Jurkat cells with PDBu and calcium ionophore did not change the distribution of these transcripts in the different fractions (data not shown).

      DISCUSSION

      In this paper we have identified the promoter region of the humanCOT gene and demonstrated that its activity is inducible by T cell-activating signals. We have also identified three different human COT mRNAs, COT-1, COT-2, andCOT-3, with different lengths in the 5′UTR but a common transcription initiation site. This site is located 4748 nt upstream of the translation initiation site of COT kinase. The first exon of COT-1 (denominated exon 1) comprises these 4748 nt and the first 336 nt of the coding sequence of COT kinase (see Table Iand Fig. 1). The lack of splicing in the 5′UTR of COT-1mRNA species results in a predominantly nuclear distribution. The physiological significance of this finding remains to be established. Nevertheless, the possibility that COT-1 mRNA is stored in the cell nucleus and a later processing of its 5′ UTR triggers the transport of the generated transcript to the cytoplasm should not be excluded. A similar situation has been described for other mRNAs (
      • Weil D.
      • Brosset S.
      • Dautry F.
      ). The 5′ UTR of the COT-2 transcript is generated when the 5′ region flanking the translation initiation site of human COT kinase undergoes splicing, and this UTR comprised of exons 1a, 1b, and 1c. This 5′ UTR contains 444 nt upstream of the coding sequence of COT kinase and exhibits a putative open reading frame located in exon 1b. COT-3 mRNA, with a 215-nt 5′ UTR, is comprised of exons 1a and 1c and does not have any open reading frame upstream of the translation initiation site of COT kinase. The occurrence of upstream open reading frames has only been detected in about 10% of vertebrate mRNAs, and their physiological role is still unclear. Interestingly, the majority of these mRNA species code for proteins involved in signal transduction (30, 31 and references therein).
      Toyoshimo and co-workers (
      • Aoki M.
      • Hamada F.
      • Sugimoto T.
      • Sumida S.
      • Akiyama T.
      • Toyoshima K.
      ,
      • Ohara R.
      • Miyoshi J.
      • Aoki M.
      • Toyoshima K.
      ) reported a genomic structure of the human COT gene with 9 exons, from which the 2 upstream exons are noncoding, and defined a 5′ UTR of the COT transcript (GenBankTM accession number D14497) with 366 nt; this sequence corresponds to part of the 5′ UTR of the COT-2transcript defined here. Chan et al. (
      • Chan A.M-L.
      • Chedid M.
      • McGovern E.S.
      • Popescu N.C.
      • Miki T.
      • Aaronson S.A.
      ) reported aCOT transcript (GenBankTM accession numberZ14138) (denominated est in their study) with a 159-nt long 5′ UTR; this sequence corresponds to part of the 5′ UTR of COT-1 defined here.
      It has recently been reported that an increase in COTmRNA levels, determined by RT-PCR analysis, of a 136-nt fragment of the COT coding sequence plays a role in human breast cancer (
      • Sourvinos G.
      • Tsatsanis D.A.
      • Spandidos D.A.
      ). Based on the data obtained here, demonstration of this role in tumorigenesis would require analysis of the expression of the differentCOT transcripts in tumoral versus normal tissues, because the study of the subcellular and polysome distribution of the different COT transcripts indicates that COT-1 is not attached to polysomes but is mainly located in the nucleus and is, therefore, probably not translated.
      We did not find any evidence of additional splicing variants in the protein coding exons of the described human COT gene (data not shown). This is probably because COT kinase does not have any of the known signaling modules, and consequently changes in the coding sequence (with the exception of the last exon) would result in the modification of at least one of the XI regions necessary for a functional protein kinase (
      • Hanks S.K.
      • Quinn A.M.
      • Hunter T.
      ). Deletion of the last exon provides transformation capacity to COT kinase (
      • Patriotis C.
      • Makris A.
      • Bear S.E.
      • Tsichlis P.N.
      ,
      • Aoki M.
      • Hamada F.
      • Sugimoto T.
      • Sumida S.
      • Akiyama T.
      • Toyoshima K.
      ,
      • Erny K.M.
      • Peli J.
      • Lambert J-F.
      • Muller V.
      • Diggelmann H.
      ,
      • Ceci J.D.
      • Patriotis C.P.
      • Tsatsanis C.
      • Makris A.M.
      • Kovatch R.
      • Swing D.A.
      • Jenkins N.A.
      • Tsichlis P.N.
      • Copeland N.G.
      ).
      The COT gene is induced during T cell activation. Both combinations of stimuli (PDBu and calcium ionophore or PDBu and αCD3) induce an increase in COT mRNA levels as well as activation of the COT promoter-driven transcription. A number of consensus sequences reported to bind specific trans-acting factors regulated by activating T signals are present in the 5′ flanking region of the COT gene (Fig. 3). Thus, three AP-1 binding sites (
      • Angel P.
      • Imagawa M.
      • Chiu R.
      • Stein B.
      • Imbra R.
      • Rahmsdorf H.J.
      • Jonat C.
      • Herrlich P.
      • Karin M.
      ,
      • Lee W.
      • Mitchell P.
      • Tjian R.
      ) were found at positions −79, −327, and −637 nt. One PEA-3 motif (
      • Martin M.E.
      • Piette J.
      • Yaniv M.
      • Tang W.J.
      • Folk W.R.
      ) was located at −267 nt. The analysis also revealed one consensus recognition motif for ets (
      • Macleod K.
      • Leprince D.
      • Stehelin D.
      ) at position −808 nt and four consensus binding sequences for OCT-1 (
      • Ho C.
      • Bhat N.K.
      • Gottschalk L.R.
      • Lindsten R.
      • Thompson C.B.
      • Papas T.S.
      • Leiden J.M.
      ) at positions −293, −501, −708, and −1037 nt. The ets and OCT-1 response elements have been reported to be regulated by phorbol esters (
      • Faisst S.
      • Meyer S.
      ). Two cAMP-response element (CRE)-like sequences (
      • Faisst S.
      • Meyer S.
      ) are found at positions −41 and −164 nt, which could account for the increased COT promoter activity observed upon 8-Br-cAMP treatment. This finding also suggests that different signal transduction pathways can mediate COT regulation. The relative activities of the different COT promoter constructs in the presence and absence of T cell activating signals indicated that a PDBu and calcium ionophore response element is located in the −340 to −229-nt region of the COT promoter. Activation of the AP-1 response elements in T cells requires an integration of both PDBu and calcium ionophore signals, is independent of ERK pathway activation, is sensitive to cyclosporin A, and is up-regulated by the JNK/SAPK signal transduction pathway (
      • Ballester A.
      • Velasco A.
      • Tobeña R.
      • Alemany S.
      ,
      • Su B.
      • Jacinto E.
      • Hibi M.
      • Karin M.
      • Ben-Neriah Y.
      ). The same requirements are needed for activation of the −340COT promoter-driven transcription, indicating that at least the AP-1 binding site present at −327 nt could play a role in the PDBu and calcium ionophore-triggered COT promoter activation.
      COT kinase activity is crucial for the transduction mechanism of activating signals in T cells during G0/G1transition (
      • Salmerón A.
      • Ahmad T.B.
      • Carlille G.W.
      • Pappin D.
      • Narsimham R.P.
      • Ley S.C.
      ,
      • Ballester A.
      • Calvo V.
      • Tobeña R.
      • Lisbona C.
      • Alemany S.
      ,
      • Ballester A.
      • Velasco A.
      • Tobeña R.
      • Alemany S.
      ,
      • Tsatsanis C.
      • Patriotis C.
      • Tsichlis P.N.
      ,
      • Tsatsanis C.
      • Patriotis C.
      • Bear S.E.
      • Tsichlis P.N.
      ,
      • Belich M.P.
      • Salmerón A.
      • Johnston L.H.
      • Ley S.C.
      ). The data presented here indicate that the expression of the COT gene is regulated by these same signals.

      Acknowledgements

      We thank Luis Alvarez Querido and Jose Gonzalez Castaño for critical reading of the manuscript, Joaquin Perez for technical assistance, Drs. Manolo Fresno and Tadashi Nishida for the pSCα DN-SEK-2 construct, and Dr. A. Arnero (Sandoz-España) for providing cyclosporin A.

      REFERENCES

        • Salmerón A.
        • Ahmad T.B.
        • Carlille G.W.
        • Pappin D.
        • Narsimham R.P.
        • Ley S.C.
        EMBO J. 1996; 15: 817-826
        • Fanger G.R.
        • Gerwins P.
        • Widmann C.
        • Jarpe M.B.
        • Johnson G.L.
        Curr. Opin. Genet. Dev. 1997; 7: 67-74
        • Hunter T.
        Cell. 1997; 88: 333-346
        • Patriotis C.
        • Makris A.
        • Bear S.E.
        • Tsichlis P.N.
        Proc. Natl. Acad. Sci. U. S. A. 1993; 90: 2251-2255
        • Robinson M.J.
        • Cobb M.H.
        Curr. Opin. Cell Biol. 1997; 9: 180-186
        • Patriotis C.
        • Makris A.M.
        • Chernoff J.
        • Tsichlis P.N.
        Proc. Natl. Acad. Sci. U. S. A. 1994; 91: 9755-9759
        • Troppmair J.
        • Bruder J.T.
        • Munoz H.
        • Lloyd P.A.
        • Kyriakis J.
        • Banerjee P.
        • Avruch J.
        • Rapp U.R.
        J. Biol. Chem. 1994; 269: 7030-7035
        • Ballester A.
        • Calvo V.
        • Tobeña R.
        • Lisbona C.
        • Alemany S.
        J. Immunol. 1997; 159: 1613-1618
        • Ballester A.
        • Velasco A.
        • Tobeña R.
        • Alemany S.
        J. Biol. Chem. 1998; 273: 14099-14106
        • Tsatsanis C.
        • Patriotis C.
        • Tsichlis P.N.
        Oncogene. 1998; 19: 2609-2618
        • Tsatsanis C.
        • Patriotis C.
        • Bear S.E.
        • Tsichlis P.N.
        Proc. Natl. Acad. Sci. U. S. A. 1998; 95: 3827-3832
        • Belich M.P.
        • Salmerón A.
        • Johnston L.H.
        • Ley S.C.
        Nature. 1999; 397: 363-368
        • Lin X.
        • Cunningham Jr., E.T.
        • Mu Y.
        • Geleziunas R.
        • Greene W.C.
        Immunity. 1999; 10: 271-280
        • Chan A.M-L.
        • Chedid M.
        • McGovern E.S.
        • Popescu N.C.
        • Miki T.
        • Aaronson S.A.
        Oncogene. 1993; 8: 1329-1333
        • Aoki M.
        • Hamada F.
        • Sugimoto T.
        • Sumida S.
        • Akiyama T.
        • Toyoshima K.
        J. Biol. Chem. 1993; 268: 22723-22731
        • Makris A.
        • Patriotis C.
        • Bear S.E
        • Tsichlis P.N.
        J. Virol. 1993; 67: 4283-4289
        • Ohara R.
        • Miyoshi J.
        • Aoki M.
        • Toyoshima K.
        Jpn. J. Cancer Res. 1993; 84: 518-525
        • Miyoshi J.
        • Higashi T.
        • Mukai H.
        • Ohuchi T.
        • Kakunaga T.
        Mol. Cell. Biol. 1991; 11: 4088-4096
        • Aoki M.
        • Akiyama T.
        • Miyoshi J.
        • Toyoshima K.
        Oncogene. 1991; 6: 1515-1519
        • Erny K.M.
        • Peli J.
        • Lambert J-F.
        • Muller V.
        • Diggelmann H.
        Oncogene. 1996; 13: 2015-2020
        • Ceci J.D.
        • Patriotis C.P.
        • Tsatsanis C.
        • Makris A.M.
        • Kovatch R.
        • Swing D.A.
        • Jenkins N.A.
        • Tsichlis P.N.
        • Copeland N.G.
        Genes Dev. 1997; 11: 688-700
        • Sourvinos G.
        • Tsatsanis D.A.
        • Spandidos D.A.
        Oncogene. 1999; 18: 4968-4973
        • Blin N.
        • Stafford D.V.
        Nucleic Acids Res. 1976; 3: 2303-2308
        • Sanger F.
        • Niksen S.
        • Coulson A.R.
        Proc. Natl. Acad. Sci. U. S. A. 1977; 74: 5463-5467
        • Nakiely S.
        • Dreyfuss G.
        Curr. Opin. Cell Biol. 1997; 9: 420-429
        • Bucher P.
        J. Mol. Biol. 1990; 212: 563-578
        • Angel P.
        • Imagawa M.
        • Chiu R.
        • Stein B.
        • Imbra R.
        • Rahmsdorf H.J.
        • Jonat C.
        • Herrlich P.
        • Karin M.
        Cell. 1987; 49: 729-739
        • Lee W.
        • Mitchell P.
        • Tjian R.
        Cell. 1987; 49: 741-752
        • Weil D.
        • Brosset S.
        • Dautry F.
        Mol. Cell. Biol. 1990; 10: 5865-5875
        • Hanks S.K.
        • Quinn A.M.
        • Hunter T.
        Science. 1988; 241: 42-52
        • Martin M.E.
        • Piette J.
        • Yaniv M.
        • Tang W.J.
        • Folk W.R.
        Proc. Natl. Acad. Sci. U. S. A. 1988; 85: 5839-5843
        • Macleod K.
        • Leprince D.
        • Stehelin D.
        Trends Biochem. Sci. 1992; 17: 251-256
        • Ho C.
        • Bhat N.K.
        • Gottschalk L.R.
        • Lindsten R.
        • Thompson C.B.
        • Papas T.S.
        • Leiden J.M.
        Science. 1990; 250: 814-818
        • Faisst S.
        • Meyer S.
        Nucleic Acids Res. 1992; 20: 3-26
        • Su B.
        • Jacinto E.
        • Hibi M.
        • Karin M.
        • Ben-Neriah Y.
        Cell. 1994; 77: 726-736