This invention was funded in part by research grant Al 28900 from the National Institutes of Health, which provides to the United States Government certain rights in this invention.

BACKGROUND OF THE INVENTION

Field of the Invention The present invention relates to DNA oligonucleo¬ tides that bind specifically to transcriptional regulatory proteins (TRPs) and the use of such oligonucleotide se¬ quences in (a) the detection and/or isolation of TRPs, (b) analysis of TRP induction by cytokines and (c) analysis of TRP action.

Description of the Background Art

Cell surface receptors for extracellular polypeptide ligands are widely expressed on most mammalian cells and serve to transduce proliferative and differentiation sig- nals to the cell's interior. Multistep signaling pathways lead from the binding of growth factor receptors, for example, the epidermal growth factor receptor (EGFR) to subsequent intracellular processes (J. Schlessinger et al . , Neuron 9, 383 (1993)) . Many growth factor receptors, like EGFR, possess intrinsic tyrosine kinase activities, which, after ligand activation, autophosphorylate the receptor cytoplasmic tail and phosphorylate other cellular sub¬ strates (A. Ullrich et al . , Cell 61, 203 (1990)) .

A common response for receptor and non-receptor signaling pathways involves the activation of Ras (L.S.

Mulcahy et al . , Nature 313, 241 (1985)) . Autophosphoryl- ated EGFR serves as a docking site for SH2 domain-contain-

ing proteins (CA. Koch et al . , Science 252, 668 (1991) , which lead to Ras activation (E.J. Lowenstein et al . , Cell 70, 431 (1992); N. Li et al . , Nature 363, 85 (1993; N. . Gale et al . , Nature 363, 88 (1993); M. Rozakis-Adcock et al . , Nature 363, 83 (1993) ; S.E. Egan et al . , Nature 363, 45 (1993); M. Rozakis-Adcock et al . , Nature 360, 689 (1992- ); K. Baltensperger et al . , Science 260, 1950 (1993); J.P. Olivier et al . , Cell 73, 179 (1993)) . Activated Ras then initiates a cascade of serine/threonine phosphorylations through Raf and MAP kinases (S.A. Moodie et al . , Science 260, 1658 (1993); C . Lange-Carter et al . , Science 260, 315 (1993)) leading to phosphorylation of nuclear proteins (C.S. Hill et al . , Cell 73, 395 (1993) ; H. Gille et al . , Nature 358, 414 (1992) ; R.H. Chen et al . , Mol . Cell . Biol . 12, 915 (1992) ) .

It is widely speculated that the biological consequences of receptor tyrosine kinase activation depends on changes in gene expression. However, it has been diffi¬ cult to document a direct connection between the cytoplas- mic phosphorylation events stimulated by receptor tyrosine kinase activation and modulation of nuclear transcription factors. In contrast, recent studies of signals generated by interferons (IFNs) have revealed an apparently simpler path to the nucleus (D.E. Levy et al . , New Biologist 2 , 923

(1990) ) . Though lacking intrinsic catalytic domains (G. Uze et al . , Cell 60, 225 (1990)) , IFNα. receptors stimulate tyrosine phosphorylation of a family of proteins that serve as DNA-binding and transcriptional-activating factors (C Schindler et al . , Science 257, 809 (1992); K. Shuai et al . , Science 258, 1808 (1992) ; M. J. Gutch et al . , Proc . Natl . Acad. Sci . USA 89, 11411 (1992)) . These proteins, generi- cally termed "signal transducers and activators of tran¬ scription (STATs) , " normally sequestered in the cytoplasm (D.E. Levy et al . , Genes Dev. 2, 1362 (1989) ; D.S. Kessler et al . , Genes Dev. 4, 1753 (1990)) , upon tyrosine phosphor¬ ylation, assemble into a multimeric complex, translocate to the nucleus, and bind cis-acting enhancer elements in the appropriate regulatory regions. One example of a STAT is

ISGF3 which binds to regulatory regions of IFN-stimulated genes. IFNγ stimulates phosphorylation of the 91 kDa protein subunit of ISGF3 (termed p91) , leading to its appearance in the nucleus and its binding to a distinct genetic element, activating IFNγ-responsive genes (K. Shuai et al . , supra) .

When mouse 3T3 fibroblasts are stimulated by v- sis (platelet-derived growth factor or PDGF) , a nuclear factor, termed sis-inducible factor (SIF) , binds to a DNA sequence which is a regulatory element in the c-fos promot¬ er (T.E. Hayes et al . , Proc. Natl . Acad. Sci . USA 84, 1272 (1987); B.J. Wagner et al . , EMBO J. 9, 4477 (1990)) . The rapid, transient stimulation of this DNA-binding activity led the present inventor and his colleagues to investigate its possible relationship to ISGF3 activation. This, led the present inventor to the discovery of the subject matter of the present invention, including (1) novel DNA oligonucleotide sequences to which such TRPs bind and (2) uses for these oligonucleotides.

SUMMARY OF THE INVENTION

The present invention is directed to an oligonu¬ cleotide having between about 11 and 150 nucleotides and comprising the nucleotide sequence TTTCCNGGAAA [SEQ ID NO:l] , which oligonucleotide binds directly or indirectly to the transcriptional regulatory protein p91. The above sequence is preferably TTTCCGGGAAA [SEQ ID NO:2] .

The above oligonucleotide may be double stranded. A preferred double stranded oligonucleotide has the se¬ quence: 5'-gatcTAGGGATTTCCGGGAAATGAAGCT-3' SEQ ID NO:3

3' ATCCCTAAAGGCCCTTTACTTCGActag-5' SEQ ID NO:4

The present invention is further directed to a composition useful for measuring the ability of a compound to stimulate transcription of DNA, which composition com- prises a recombinant DNA molecule comprising an oligonucle¬ otide as described above, operably linked to a promoter and to an expressible reporter nucleotide sequence.

Also provided is a host cell transfected with a composition as described above.

The present invention is further directed to a method for detecting the presence of a transcriptional regulatory protein in a sample, which protein is capable of binding to an oligonucleotide as described above, which method comprises :

(a) contacting the sample with an oligonucleotide as described above under conditions wherein the protein binds to the oligonucleotide to form a complex;

(b) detecting the presence of the protein-oligonucle- otide complex, wherein the presence of the complex indicates the presence of the protein in the sample.

Also provided is a method for determining whether a protein is a DNA-binding protein with transcriptional regulatory protein activity, comprising detecting the direct or indirect binding of a protein in a sample con- taining the protein with an oligonucleotide as described above.

The present invention is directed to a method for detecting the presence or measuring the activity of a compound, in a sample, which compound is capable of indue- ing a transcriptional regulatory protein in a cell, wherein the transcriptional regulatory protein binds to DNA and stimulates transcription, which method comprises:

(a) contacting the sample with a host cell as des¬ cribed above under conditions in which the ex- pressible DNA is expressed;

(b) measuring the expression of the DNA;

(c) comparing the expression in step (b) with the expression of expressible DNA not linked to the above oligonucleotide, wherein an increase in the expression of the expressible

DNA indicates the presence in the sample of, or the activi¬ ty of, the compound.

The present invention also provides a DNA-binding protein, other than p91, which protein, after tyrosine phosphorylation, binds to the DNA sequence TTTCCNGGAAA (SEQ ID NO:l) or to a derivative thereof. The DNA binding protein may be inducible in a cell by a cytokine, growth factor or hormone. The DNA binding protein may be tyrosine phosphorylated in response to activation of a cell with a cytokine, growth factor or hormone. The cytokine, growth factor or hormone is preferably interleukin-2, interleukin-3 or erythropoietin.

In another embodiment, the above DNA-binding protein binds directly or indirectly to the double stranded

DNA sequence

5' -gatCTAGGGATTTCCGGGAAATGAAGCT-3' SEQ ID NO:3 3' ATCCCTAAAGGCCCTTTACTTCGActag-5' . SEQ ID NO:4

The above DNA-binding protein is preferably a transcriptional regulatory protein.

Also provided is a method for isolating or puri¬ fying a DNA binding protein in a sample, which protein binds directly or indirectly to an oligonucleotide as described above, which method comprises:

(a) contacting the sample with an oligonucleotide as above under conditions wherein said protein binds to the oligonucleotide to form a complex, prefer- ably, wherein the oligonucleotide is immobilized on a solid phase carrier;

(b) eluting the protein from the protein-oligonucleo- tide complex, thereby isolating or purifying the protein.

BRIEF DESCRIPTION OF THE DRAWINGS

Figures 1A-1E shows stimulation of DNA-binding activity in response to growth factors. Cells were un¬ treated or treated with the indicated growth factors and cytokines prior to detergent lysis and analysis of extracts by gel band shift using a labeled oligonucleotide contain¬ ing an SIF-recognition sequence. Figure 1A: Swiss 3T3 mouse fibroblasts were treated for 15 min with 40 ng/ml

human recombinant PDGF-BB for 15 min and fractionated into cytoplasmic and nuclear extracts. 5 μg of cytoplasmic and 10 μg of nuclear extracts were added to DNA binding reac¬ tions. Figure IB: Human FS2 fibroblasts were treated with human recombinant EGF (250 ng/ml) , PDGF (40 ng/ml) , IFNo. (500 U/ml) , or IFN (1 ng/ml) for 15 min, as indicated. Figure 1C Human FS2 fibroblasts were treated for the indicated times with PDGF, IFNα, and IFNγ. Cells treated with cycloheximide (lanes 8, 16, and 24) were pretreated for 15 min and throughout stimulation with 50 μg/ml.

Figure ID: P13 cells overexpressing the human CSF1 recep¬ tor (lanes 1-5) , WT7 cells overexpressing the fig FGF receptor (lanes 6-10) , and 3006 cells expressing the human insulin receptor (lanes 11-13) were treated for 15 min, as indicated, with the following ligands: murine IFNα., 5000 u/ml; murine IFN7, 5 ng/ml; human PDGF-BB, 50 ng/ml; human CSF1, 10000 u/ml; acidic FGF, basic FGF, and k-FGF, 200 ng/ml; bovine insulin, 1 /xg/ml. Figure IE: Extracts from human FS2 fibroblasts treated for 15 min with PDGF were analyzed by band shift using a modified SIF binding se¬ quence recognizing SIF with higher affinity. Competitions were carried out using 100-fold molar excess of unlabeled oligonucleotide, as indicated. Antibodies against human pll3, p91 carboxyl-terminal peptide or p48 were added at a final concentration of 1:200 for lanes 5-7, respectively. Figures 2A-2D shows tyrosine phosphorylation of p91 by PDGF and interferons. Figure 2A: Immunoprecipit- ation of ³² P-orthophosphate-labeled (2 hr) extracts of human FS2 fibroblasts stimulated with the indicated factors for 10 min. Immunoprecipitates with anti-p91 antibody were analyzed by SDS-PAGE. Figure 2B: Phosphoamino acid analy¬ sis of labeled p91 excised from SDS-PAGE, hydrolyzed, and analyzed by thin-layer chromatography. Migration of phos¬ phoamino acid standards is indicated at right. Figure 2C: Tryptic phosphopeptide mapping of p91. Immunoprecipitated p91 from orthophosphate-labeled cells untreated or treated with IFNs or PDGF as indicated were digested with trypsin and analyzed by two-dimensional thin-layer electrophoresis

and chromatography. Upon longer exposure, a fourth phos- phopeptide fragment was detected in both unstimulated and stimulated samples. Figure 2D: FGF fails to induce tyro¬ sine phosphorylation of p91. Immunoprecipitated p91 was analyzed by SDS-PAGE and immunoblotting using anti-phospho- tyrosine (lanes 1-3) or anti-p91 (lanes 4-6) antibodies. Figures 3A-3D shows that activation of p91 re¬ quires receptor autophosphorylation but is independent of Ras. Figure 3A: Transfected mouse cells stably expressing wild-type human EGF receptor (HER14, lanes 1-3), no human receptor (3T3, lanes 4 and 5), mutant receptors having the five major phosphotyrosine sites converted to phenylalanine (Y5F, lanes 6 and 7) , or mutant receptor truncated in the cytoplasmic tail at amino acid 991 (D991, lanes 8 and 9) were treated with mouse IFNo. (2500 U/ml) or human EGF (250 ng/ml) for 15 min, as indicated, and nuclear extracts were analyzed by band shift. Figure 3B: Detergent lysates of the same cell lines were analyzed by immunoblotting for the presence of the human EGF receptor using rabbit anti-EGFR antiserum. Figure 3C: Rat L6 myoblasts (lanes 1 and 2) or L6 cells stably transfected with dominant-negative Asn-17 Ras under a hormone-responsive promoter and induced with 1.5 μM dexamethasone overnight (lanes 3 and 4) were treated with PDGF, as indicated. Extracts were analyzed for p91 activation by band shift (top) or for Ras expression by immunoblotting (bottom) . Figure 3D: Extracts from (C) were analyzed for MAP kinase activation by immunoblotting with anti-phosphotyrosine serum.

Figure 4 presents a model for rapid signal trans- duction in response to EGF. Autophosphorylation of the cytoplasmic tail of EGFR following EGF stimulation could create a docking site either for p91 directly or for a cytoplasmic kinase, such as a member of the Jak/Tyk family, leading to tyrosine phosphorylation of p91. Activated p91 directly translocates to the nucleus and binds DNA, in contrast to the Ras-dependent pathway where multiple steps are required to induce Ser/Thr phosphorylation of diverse nuclear proteins.

Figure 5 shows that proteins bind the DSE se¬ quence with higher affinity than to the SIE. Nuclear extracts (5 μg) from Swiss 3T3 cells treated for 15 min with 2500 u/ml murine IFNo., were mixed with labeled DSE in a 5 μl gel band shift reaction. Each reaction contained 1.8 μg of non-specific competitor DNA plus the indicated molar excess of competitor oligonucleotide. Gel band shift reactions were resolved on 5% acrylamide gels. Tenfold molar excess of DSE competed equally to 50-fold excess of SIE. "ns" , the SRF binding site from the human gamma actin promoter; "SIE", mutant sequence from the c-fos promoter defined by Wagner et al . (1990) EMBO J. 9:4477-4484 (1990); "DSE", the DSE standard oligonucleotide.

Figure 6 shows that DSE-binding proteins are induced by erythropoietin, IL2 and IL3. Nuclear extracts (5 μg) from BaF3/ER cells (D'Andrea et al . (1991) Molec. Cell . Biol . 11:1980-1987) or from LyD9 cells grown for 4 hr in the absence of IL3 or from CTLL cells (ATCC TIB214) grown for 4 hr in the absence of IL2, treated as indicated, were analyzed by band shift using a DSE probe.

DESCRIPTION OF THE PREFERRED EMBODIMENTS

The present inventor and his coworkers, studying the relationship between growth factor receptor tyrosine kinases and nuclear regulatory events, found that several polypeptide growth factors stimulate (1) tyrosine phosphor¬ ylation of p91, a transcriptional regulatory protein (TRP) , (2) nuclear accumulation of p91 and (3) binding of p91 to DNA. The term "transcriptional regulatory protein" as used herein, is intended to include proteins that bind DNA either directly or indirectly through adaptor proteins, to modulate the transcriptional activity of the associated gene or genes. Understanding of, and manipulation of, these protein-DNA interactions require the ability to measure the binding of known and yet undiscovered TRPs to specific DNA sequences. Thus, the term "DNA-binding pro¬ tein" as used herein refers either to a protein, which binds directly to DNA (or to an oligonucleotide) or a protein

which binds indirectly to DNA by binding to another protein which binds to or is bound to DNA (Veals, S.A. et al . , Molec . Cell . Biol . 13:196-206 (1993)) .

The term "activated" or "activation" as used herein, for example, "activation" of a cell by a cytokine, growth factor or hormone, means that one or more proteins in the cell are modified post-translationally such that they can now bind to DNA in a sequence-specific manner, preferably to the oligonucleotide sequence of the present invention. Such modification preferably comprises tyro¬ sine phosphorylation of the protein. "Induction" of a protein, for example by a cytokine, growth factor or hor¬ mone, indicates that the abundance of the protein in the cell increases. In performing the above studies, the present inventor, using synthetic oligonucleotides, discovered DNA sequences that serve as a binding site for cellular pro¬ teins that are induced or activated in response to a vari¬ ety of biological stimuli. In particular these DNA se- quences recognize (or are recognized by) a family of transcriptional regulatory proteins (TRPs) that are acti¬ vated in response to any of a variety of cytokines and growth factors. Such DNA sequences may be native sequenc¬ es, or preferably, are non-native oligonucleotide sequences which bind more strongly to TRPs than do the native se¬ quences.

A preferred DNA sequence according to the present invention is a dyadic symmetry element (DSE) , having the sequence TTTCCNGGAAA (SEQ ID NO:l), or preferably, TTTCCGGGAAA (SEQ ID NO:2) , that has higher binding affinity for TRPs than any other DNA sequence known. Therefore, this nucleotide sequence, preferably in the form of an oligonucleotide of between about 11 and 150 nucleotides, or a derivative thereof, is useful for detecting such TRPs in various biological samples, for purifying such TRPs, and for identifying new TRPs which bind to this sequence motif. While a GENBANK search revealed the presence of TTTCCGGGAAA (SEQ ID NO:2) in a number of published genome sequences,

none of these sequences were as short as 11 to 150 nucleo- tides. Preferably, the oligonucleotides of the present invention are prepared synthetically using methods well- known in the art. Furthermore, the novel oligonucleotides of the present invention are useful in the construction of expres¬ sion vectors, such as plasmids, and host cells transfected therewith, which are useful for testing the biological activity of TRPs and for evaluating the ability of a com- pound, such as a potential drug, for its ability to stimu¬ late or inhibit (a) induction of a TRP, (b) binding of a TRP to DNA or (c) transcriptional regulatory activity of the TRP.

Table I, below, lists a series of DNA sequences which have been shown by others, or discovered by the present inventors, to bind TRPs.

TABLE I

DNA SEQUENCES WHICH BIND TRANSCRIPTIONAL REGULATORY PROTEINS

Name SEOUENCE SEO ID NO:

FOS GTTCCCGTCAA 5

SIE TTTACGGGAA 6

GBP TTTAGAGTAAT 7

MIG TTTATAGTAAG 8

FCγ TTTCTGGGAAA 9

K17 ATTTGGAGAAA 10

IFN GAAANN x 2*

IFN (ICSBP) AGTGATTTCTCGGAAAGAGAG 11

(IRF- -1) CATCATTTCGGGGAAATCAGG 12

IFNγ TT(C/A)CNNNAA IL-6 TT(C/A)CNG(G/T)AA

DSE TTTCCGGGGAA 2

* Two GAAANN motifs in inverted fashion, wherein N may be any nucleotide.

The present inventors discovered that the pattern of known or potential p91-binding DNA sequences predicted a trend towards a perfectly symmetric sequence such as the

DSE, which is symmetric around an "odd" nucleotide (shown in brackets) :

TTTCC[G]GGAAA (SEQ ID NO:2) AAAGG[G] CCCTT (SEQ ID NO:13) .

A preferred double stranded DNA sequence useful according to the present invention, which includes the DSE, has the sequence:

5' -gatcTAGGGATTTCCGGGAAATGAAGCT-3' SEQ ID NO:3 3' ATCCCTAAAGGCCCTTTACTTCGActaσ-5' SEQ ID NO:4 (The DSE sequence, or its complement, is underlined) .

A preferred derivative of the oligonucleotide of the present invention is a multimer, comprising two or more "units" of the basic oligonucleotide. There is no limit to the number of such oligonucleotide units in a multimer although, for practical reasons, 2-10 units are preferred. The multimeric embodiment is useful both as a probe for detecting a TRF, or for purifying a TRF. In addition, the recombinant DNA molecules disclosed herein (see below) comprising the oligonucleotide sequence of the present invention operably linked to a promoter and reporter se¬ quence, may have two or more units or the oligonucleotide sequence present as an enhancer, either contiguously or with unrelated nucleotides intervening between units of the oligonucleotide. A number of cytokines and/or growth factors activate proteins which bind the DSE or derivative DNA sequence of the present invention. These cytokines, growth factors and hormones include, but are not limited to, the interleukins IL2, IL3, IL4, IL5, IL6, and IL11, granulo- cyte-macrophage colony stimulating factor (GM-CSF) , colony stimulating factor-1 (CSF-1) , IFNo;, IFN/5 ^* , IFNγ, epidermal growth factor (EGF) , platelet derived growth factor (PDGF) , leukemia inhibitory factor (LIF) , oncostatin, nerve growth factor (NGF) , ciliary neurotrophic factor (CNTF) and brain- derived neurotrophic factor (BDNF) , erythropoietin, growth hormone, and prolactin. Thus, an oligonucleotide of the present invention is useful for detecting the presence of,

or isolating, a TRP induced by any of these cytokines or growth factors.

As different cytokines may induce different TRPs with different DNA binding specificity, the oligonucleotide of the present invention is expected to detect new TRPs with binding specificity to DNA which are not identical to the TRPs characterized thus far. Furthermore, the oligonu¬ cleotide of the present invention is useful for the purifi¬ cation and characterization of such new TRPs. A nuclear factor, termed sis-inducible factor

(SIF) , binds to a regulatory element in the c-fos promoter in response to v-sis (=PDGF) in mouse 3T3 fibroblasts (Hayes et al . , supra; Wagner et al . , supra This DNA binding activity is rapidly and transiently stimulated by PDGF. SIF binds a ³² P-labeled, double-stranded oligonucleotide probe with the sequence

5 ' -GATCAGCTTCATTTCCCGTAAATCCCTA-3 ' (SEQ ID NO:14], which contains (in the complementary strand to the underlined region) the recognition sequence TTTACGGGAA (SEQ ID NO:6) designated SIE (sis-inducible element) in Table I, above.

To detect the binding of a protein to this SIE sequence, or to a novel DNA sequence of the present inven¬ tion, the following approach may be used. Cells, such as 3T3 cells, are treated with an appropriate cytokine or growth factor, for example, with PDGF for 15 minutes to induce one or more TRPs. Extracts of the nucleus and cytoplasm are prepared using conventional methods and tested for binding to the oligonucleotide sequence by an electrophoretic band shift assay, such as that described by D.E. Levy et al . , Genes Dev. 2 , 1362 (1989) and D.S. Kessler et al . , Genes Dev. 4, 1753 (1990)) .

Little or no specific DNA binding activity is expected in extracts from untreated cells, whereas stimu¬ lated cells should possess significant activity. As de- scribed in the Examples, below, DNA binding activity may be found in both cytoplasmic and nuclear fractions. Further¬ more, latent DNA binding activity may be stimulated in vi tro by treating a cytoplasmic extract, supplemented with

membranes, with the activating agent, for example a cyto¬ kine.

If an antibody specific for a TRP is available, it can be used to specifically interfere with DNA binding of the TRP, thereby assisting in the identification of the TRP. Furthermore, an unknown TRP identified or purified using the oligonucleotide of the present invention can be used to immunize animals to prepare an antibody specific for the TRP using methods well-known in the art. See, for example, Hartlow, E. et al . , Antibodies: A Laboratory

Manual, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY, 1988)

The DNA sequence of the present invention thus can serve as a "probe sequence" similar to those used in a variety of nucleic acid detection systems well-known in the art. Here, however, such DNA probes are used to detect a protein, rather than a nucleic acid, which specifically binds to the DNA sequence.

The sensitivity of nucleic acid detection assays are increased by altering the manner in which detection is reported or signaled to the observer. Thus, for example, assay sensitivity can be increased through the use of detectably labeled oligonucleotide reagents. A wide vari¬ ety of such labels have been used for this purpose includ- ing enzyme labels (Kourilsky et al . (U.S. Patent

4,581,333) , radioisotopic labels (Falkow et al . (U.S. Patent 4,358,535; Beminger (U.S. Patent 4,446,237) , fluo¬ rescent labels (Albarella et al . , EP 144914), chemical labels (Sheldon III et al . , U.S. Patent 4,582,789; Albarella et al . , U.S. Patent 4,563,417), modified bases

(Miyoshi et al . , EP 119448) , etc. Thus, DNA probes accord¬ ing to the present invention include the oligonucleotide sequence alone or as part of a longer nucleic acid mole¬ cule, labeled with a detectable label, preferably a radio- isotope, an enzyme, a fluorescent label, a chemical label, or a modified base.

Examples of modified or substituted nucleotides useful in constructing a labeled oligonucleotide probe of

the present invention, as an alternative to a radiolabel ( e . g. , ³² P) , include the following. A base analog such as bromodeoxyuridine (BrdUrd) may be incorporated into an oligonucleotide; however, BrdUrd has the disadvantage of being highly light-sensitive. The base 5-methyl cytosine (5MC) may be used in place of cytosine in the synthesis of the probe, preferably outside the binding sequence, for example, in the form of 5MC "tail" added to the probe. An antibody specific for the modified base, such as an anti- body specific for 5MC or specific for a 5MC oligomer, can be used to detect the DNA sequence. Other known modified bases may be used, provided that an antibody specific for the modified base can be prepared. After preparing a gel or blot, the target protein is detected by binding to the probe DNA of the invention, followed by a detectably la¬ beled antibody to the substituted or modified base. Any direct or indirect antibody detection system can be used ( e . g. , rabbit polyclonal antiserum to 5MC plus enzyme- labeled protein A or second antibody, or a biotin-avidin sandwich method well-known in the art, etc.

Biotin can also be linked directly to DNA or an oligonucleotide fragment and its presence detected using any of a number of avidin-based detection systems well- known in the art. Examples of the above approaches are well-known in the art of DNA detection, but can easily be adapted to use according to the present invention. For example, the method of Stavrianopoulos et al . , US Patent 4,987,065

(1/22/91) can be used for in vivo labeling of polynucleo- tide sequences, such as by incorporating a labeled base into a polynucleotide sequence, for making non-radioactive- ly labeled probes at less cost, for use in various detec¬ tion systems. Useful "label"-carrying compounds include glucosylated nucleotides, glycosylated nucleotides, 5-hydroxymethyluracil, BrdUrd and 5-methyl cytosine.

Signaling moieties, which bind covalently or non-covalently to the label of the polynucleotide sequence include anti¬ bodies specific for the modified base. The method of

Gratzner et al . , US Patent 4,529,700 (7/16/85), utilizes monoclonal antibodies (mAbs) specific for BrdUrd or specif¬ ic iododeoxyuridine (IdUrd) for detection of DNA sequences by hybridization with a BrdUrd-containing probe. Sawicki et al . , (1977), Science 174 :70. describes antibodies spe¬ cific for BrdUrd, IdUrd and 6-methyladenosine which detect¬ ed these bases in denatured DNA. Thus, any form of "anti- genic DNA", for example, DNA including a rare and antigenic nucleotide such as BrDUrd and IdUrd incorporated into a nucleotide sequence biosynthetically, can serve as a pro¬ tein binding sequence detectable by an antibody to the rare nucleotides (Paau et al . , US Patent 4,556,643 (12/3/85)) . BrDUrd containing DNA complexed to a protein can be detect¬ ed by any of a variety of immunological techniques, such as immunoenzymatic techniques (Avrameas et al . , European Patent Publication 143,059 (5/29/85))

The oligonucleotide probe of the present inven¬ tion can be chemically modified, for example with at least one base which is covalently modified by at least one 7- iodo-N-2-acetylamino-fluorene (AAIF) (Tchen, US Patent 4,963,477 (10/16/90)) . The presence of this probe is detected after binding to the protein by antibodies specif¬ ic for the modified probe. Antibodies to a related struc¬ ture, acetylaminofluorene, lacking the iodine (anti-.AAF antibodies) made either against N-2- (guanosine-8-yl) - acetylaminofluorene or against the probe bearing acetylaminofluorene residues, also recognize the modified probe.

Derivatives of the oligonucleotide of the present invention may include the oligonucleotide sequence as part of a longer nucleic acid molecule which is of natural or synthetic origin. Also included are derivatives in which the additional sequence is a mutant or variant of a natu¬ rally occurring DNA sequence. These variants or mutants ordinarily are prepared by site-directed mutagenesis

(Adelman et al. , DNA 2:183 (1983)) of DNA. Oligonucleotide or polynucleotide derivatives of oligonucleotide of the present invention typically exhibit the same qualitative

activity, binding to a TRP, as the non-modified DSE oligo¬ nucleotide.

Thus, the present invention provides a method for detecting the presence of a "target" protein molecule, a TRP, in a sample. Such samples are preferably biological samples, including, but not limited to cells, cell culture supernatant, cell or tissue extracts, or particular frac¬ tions thereof, and other biological fluids such as blood, sera, urine, saliva, etc. Binding of the DNA probe to a protein in the sample may be detected by any appropriate means known in the art. Direct or indirect, or competitive binding assays may be used. Association of the label with proteinaceous material of the sample is then detected. In a preferred embodiment, the oligonucleotide sequence is modified by the addition of one or more biotin groups.

Avidin, which binds to biotin with high affinity is used to detect the oligonucleotide-protein complex. The avidin can be attached to a solid support and/or labeled using any of a number of labels well-known in the art, including a radiolabel, an enzymatic label, etc.

The DNA molecules of the present invention may be included in a recombinant DNA construct which contains three operably linked sequences of interest: the protein binding sequence or "probe sequence" that acts as an "enhancer" element or part of a promoter, a "promoter sequence," and a "reporter sequence."

The "probe sequence" is a nucleic acid molecule containing an oligonucleotide sequence of the present invention or a derivative thereof capable of directly or indirectly binding to, and identifying a TRP. The TRP will then enhance the transcription of DNA operably linked to the probe sequence. An "enhancer" as used herein is a regulatory DNA sequence that (a) activates transcription from a promoter operably linked to it, with synthesis beginning at the normal RNA start site, (b) operates in both orientations, and (c) functions even when moved more than 1000 nucleotides from the promoter and from either an upstream or a downstream position. Both enhancers and

promoters (see below) bind sequence-specific DNA binding proteins, such as TRPs, that mediate their effects.

The probe sequence may be a multimer of the oligonucleotide sequence comprising at least two contiguous units, preferably about 2-10 units. The multimer may also have intervening nucleotides between the individual oligo¬ nucleotide units. In fact, the oligonucleotide sequence units may also occur at multiple non-contiguous sites, preferably about 2-10 sites, in the recombinant DNA mole- cule, providing that the oligonucleotide units are operably linked to the promoter and reporter sequences.

A promoter is a region of a DNA (or RNA) molecule which is capable of binding RNA polymerase and promoting the transcription of an "operably linked" nucleic acid sequence. The selection of which strand is transcribed is determined by the orientation of the promoter sequence. This selection determines the direction of transcription since RNA is only polymerized enzymatically by the addition of a nucleotide 5' phosphate to a 3' hydroxyl terminus. Two sequences of a nucleic acid molecule are said to be "operably linked" when they are linked to each other in a manner such that the presence of one sequence, e . g. , the promoter, influences transcription from the linked sequence, e . g. , a protein coding sequence, or a reporter sequence. Operably linked sequences may also include two segments of protein coding sequence which are transcribed onto the same RNA transcript. Thus, two sequences, such as a promoter sequence and a "reporter sequence" are operably linked if transcription commencing in the promoter sequence will produce an RNA transcript of the reporter sequence. In order to be "operably linked" it is not necessary that two sequences be immediately adjacent to one another.

The promoter sequences of the present invention may be any promoter capable of driving transcription of the reporter sequence and compatible with the host cell being studied. Thus, useful promoters may be prokaryotic, eukaryotic or viral. Examples of suitable prokaryotic promoters include promoters recognized by the T4 (Malik, S.

et al . , J. Biol. Chem. 263:1174-1181 (1984) ; Rosenberg, A.H. et al., Gene 59:191-200 (1987); Shinedling, S. et al . , J. Molec. Biol. 195:471-480 (1987); Hu, M. et al . , Gene 42:21-30 (1986)) , T3, Sp6, and T7 (Chamberlin, M. et al . , Nature 228:227-231 (1970); Bailey, J.N. et al . , Proc. Natl. Acad. Sci. (U.S.A.) .80:2814-2818 (1983) ; Davanloo, P. et al . , Proc. Natl. Acad. Sci. (U.S.A.) 81:2035-2039 (1984)) polymerases; the P _R and P _L promoters of bacteriophage λ (The Bacteriophage Lambda, Hershey, A.D., Ed., Cold Spring Harbor Press, Cold Spring Harbor, NY (1973) ; LAMBDA II,

Hendrix, R.W. , Ed., Cold Spring Harbor Press, Cold Spring Harbor, NY (1980) ) ; the trp, recA, heat shock, and lacZ promoters of E. coli ; the α.-amylase (Ulmanen, I., et al . , J. Bacteriol. 162:176-182 (1985)) and the σ-28-specific promoters of B . subtilis (Gilman, M.Z., et al., Gene 32 :11- 20 (1984)) ; the promoters of the bacteriophages of Bacillus (Gryczan, T.J., In: THE MOLECULAR BIOLOGY OF THE BACILLI. Academic Press, Inc., NY (1982)) ; Streptomyces promoters (Ward, J.M., et al., Mol. Gen. Genet. 203:468-478 (1986)) ; the int promoter of bacteriophage λ; the J la promoter of the β-lactamase gene of pBR322, and the CAT promoter of the chloramphenicol acetyl transferase gene of pPR325, etc. Prokaryotic promoters are reviewed by Glick, B.R., (J. Ind. Microbiol. 1:277-282 (1987)) ; Cenatiempo, Y. (Biochimie H:505-516 (1986)) ; Watson, J.D. et al . (In: Molecular Biology of the Gene, Fourth Edition, Benjamin Cummins, Menlo Park, CA (1987)) ; and Gottesman, S. (Ann. Rev. Genet. ϋ:415-442 (1984)) . Preferred eukaryotic promoters include yeast cycl promoter (Guarante et al . , Proc . Natl . Acad. Sci . USA 78:2199-2203 (1981)), the promoter of the mouse metallothionein I gene (Hamer, D., et al . , J. Mol . Appl . Gen. 1:273-288 (1982)) ; the TK promoter of Herpes virus (McKnight, S., Cell 31:355-365 (1982)) ; the SV40 early promoter (Benoist, C, et al . , Nature (London) 290:304-310 (1981)) ; and the yeast gal4 gene promoter (Johnston, S.A., et al., Proc. Natl. Acad. Sci. (USA) 79:6971-6975 (1982) ; Silver, P. ., et al . , Proc. Natl. Acad. Sci. (USA) 81:5951-

5955 (1984)). All of the above listed references are incorporated by reference herein.

In a system in which the reporter is chloram- phenicol acetyl transferase (CAT) , a preferred promoter is the CAT promoter of the bacterial chloramphenicol acetyl transferase gene.

The third sequence of the recombinant DNA or RNA molecules of the present invention is a "reporter sequence" which may be composed of any set of nucleotides regardless of sequence. A reporter is any gene whose product is used to assess regulation of transcription via a promoter or enhancer. The expression of the "reporter sequence" re¬ sults in the formation of the "reporter molecule" which is readily detectable. Preferably, the reporter sequence will have a length of between about 5 and 3,000 bases and, more preferably, 10-100 bases. The reporter sequence will preferably be selected such that the reporter molecule will have physical and chemical characteristics which will permit or facilitate its identification or detection by means well known in the art. In one embodiment, the pres¬ ence of the reporter molecule will be detected through the use of an antibody or an antibody fragment, capable of specific binding to the reporter molecule. A reporter such as β-galactosidase or luciferase can be assayed enzymati- cally or immunologically.

A preferred reporter molecule is CAT, well-known in the art. See, for example, Lewin, B., GENES IV, Oxford University Press, 1990, chap. 23 ) . Because this is a bacterial gene, it is absent from eukaryotic cells and the enzyme product can be directly assayed in a cell extract. The level of enzyme activity corresponds to the amount of enzyme that was made, which in turn reveals the level of expression. CAT mRNA may also be measured directly.

Typically, a plasmid containing the recombinant DNA molecule of the present invention, including the CAT gene, is introduced into eukaryotic cells (which may be designated "reporter cells"), which are then grown for several days. The reporter cells are treated with the test

material suspected of containing a material capable of inducing or activating a TRP, for example, an extract of other cytokine-treated cells. The CAT-producing reporter cells are extracted, ¹⁴ C-chloramphenicol is added, the reaction incubated about 60 minutes, and the extract is run on thin layer chromatography to separate acetylated and nonacetylated forms and assayed by autoradiography. Total acetylated chloramphenicol corresponds to the units of CAT enzyme activity. When transcription of CAT is activated by the TRP being assayed, CAT synthesis is increased relative to a control lacking the TRP. Thus, the amount of CAT enzyme produced is an indirect measure of transcription induced by the TRP binding to the oligonucleotide sequence of the present invention which is operably linked to the CAT gene. Cytokines or growth factors may induce any of a number of yet undiscovered TRPs which bind to the DNA sequence of the present invention. Thus, the present invention is further directed to a protein, other than p91, inducible in a cell by a cytokine, growth factor or hor¬ mone, which protein binds to the DNA sequence TTTCCGGGAAA (SEQ ID NO:2) or to the double stranded sequence

5' -gatcTAGGGATTTCCGGGAAATGAAGCT-3' SEQ ID NO:3

3' ATCCCTAAAGGCCCTTTACTTCGActag-5' . SEQ ID NO:4 Such a protein may be induced by any cytokine or growth factor, such as those described above. Preferably, the protein is induced by IL2, IL3 or erythropoietin. This protein is preferably a TRP, either stimulatory or inhibi¬ tory.

PROTEIN ISOLATION AND PURIFICATION

The invention provides a naturally occurring TRP protein other than p91 substantially free of other proteins or glycoproteins with which it is natively associated.

"Substantially free" indicates that the protein has been purified away from at least 90% (on a weight basis) , and from even at least 99% if desired, of other proteins and

glycoproteins with which it is natively associated, and is therefore substantially free of them.

Such purification can be achieved by subjecting the cells, tissue or fluids containing the TRP to standard protein purification techniques. In particular, the pres¬ ent invention provides novel methods for the isolation and purification of TRPs. These methods are based on affinity chromatography, which relies on the interaction of the protein to be purified with an immobilized ligand. Here, the ligand for a TRP is an oligonucleotide or derivative of the present invention. This ligand may be able to react with a number of TRPs and thus serves as a means for recov¬ ering a particular class of proteins.

In a preferred embodiment, the oligonucleotide of the present invention is immobilized on a solid support or carrier. By "solid phase carrier or support" is intended any support capable of binding the oligonucleotide or DNA molecule of the present invention. Well-known supports, or carriers, include glass, polystyrene, polypropylene, poly- ethylene, dextran, nylon," amyloses, natural and modified celluloses, polyacrylamides, agaroses, and magnetite. Methods for coupling nucleic acids to the solid phase, the solid phase substances useful in these methods, and means for elution, are well known to those of skill in the art. The protein bound to the DNA directly or indirectly may be eluted from the solid phase using any of a number of meth¬ ods well-known in the art.

In addition to the specific methods described above, purification steps prior to affinity purification may include one or more standard method, such as ammonium sulfate precipitation, size exclusion chromatography (gel filtration) , ion exchange chromatography, differential precipitation and the like, all well-known in the art. Also useful is the method known as hydrophobic interaction chromatography (HIC) which is based on the interaction between the solute and the gel that is hydrophobic. Hydro¬ phobic interactions are strongest at high ionic strength, therefore, this form of separation is conveniently per-

formed following salt precipitations or ion exchange proce¬ dures. Elution from HIC supports can be effected by alter¬ ations in solvent, pH, ionic strength, or by the addition of chaotropic agents or organic modifiers, such as ethylene glycol. General principles of HIC are described in U. S. Patents 3,917,527 and 4,000,098. Purification of specific proteins using HIC is disclosed in the following U.S. Patents: U.S. 4,332,717; U.S. 4,771,128; U.S. 4,743,680; U.S. 4,894,439; U.S. 4,908,434; and U.S. 4,920,196. The DSE oligonucleotide or a functional deriva¬ tive thereof, according to the present invention, can be used to assay for TRPs at any stage of a purification process as described above.

Having now generally described the invention, the same will be more readily understood through reference to the following examples which are provided by way of illus¬ tration, and are not intended to be limiting of the present invention, unless specified.

EXAMPLE I Stimulated 3T3 Cells or Fibroblasts

Produce DNA-Binding Proteins

A. MATERIALS AND METHODS

3T3 cells were treated with PDGF for 15 min, and nuclear and cytoplasmic extracts. Protein extracts from nuclei or cytoplasm of cells treated with various cytokines were prepared and tested for binding to a SIF recognition DNA sequence by electrophoretic gel band shift in an assay previously described (D.E. Levy et al . , Genes Dev. 2 , 1362 (1989) ; D.S. Kessler et al . , Genes Dev. 4, 1753 (1990)) . The DNA probe used ³² P-labeled, double-stranded oligonucleo¬ tide with the sequence 5' -GATCAGCTTCATTTCCCGTAAATCCCTA-3' (T.E. Hayes et al . , Proc . Natl . Acad. Sci . USA 84, 1272 (1987) ; B.J. Wagner et al . , EMBO J. 9 , 4477 (1990)) . Also tested was a modified oligonucleotide sequence (DSE in Table I, above) , which binds the same proteins with higher apparent affinity.

Cells were incubated in medium containing 0.5% serum for 24-48 h prior to treatment with cytokines. B. RESULTS

1. PDGF-Treated 3T3 Cells Little or no specific DNA binding activity was present in extracts from untreated cells while abundant activity was readily detected in extracts from PDGF-stimu¬ lated cells (Figure 1A) . DNA binding activity was detected in both cytoplasmic (lane 2) and nuclear fractions (lane 4) , suggesting cytoplasmic activation followed by nuclear translocation similar to ISGF3 (Kessler et al . , supra; Dale et al . , supra) . Similar cytoplasmic activation of SIF was reported (H.B. Sadowski et al . , Nature 362, 79 (1993)) ; furthermore, latent SIF DNA binding activity could be stimulated in vi tro in cytoplasmic extracts supplemented with membranes, reinforcing the conclusion that factor activation occurs in the cytoplasm.

2. EGF-Treated Human Fibroblasts

SIF DNA-binding activity was also detected in extracts from human fibroblasts treated with EGF (Figure

IB, lane 2) , but not in untreated cells (lanes 1 and 5) or in cells treated with FGF (lane 3) . Surprisingly, SIF band shift activity was also detected in extracts of cells treated with either IFNo; or IFNγ (lanes 6 and 7) . Activa- tion by PDGF, IFNα, and IFNγ was rapid, occurring within 5 min, transient, and resistant to inhibition of protein synthesis by cycloheximide (Figure 1C) . Duration of the responses differed: SIF was undetectable after 1 h of PDGF treatment (lane 5) while it was still detectable for up to 6 h in response to IFNγ (lane 23) . Interestingly, SIF activation occurred in cells treated with PDGF at 4°C while IFN activation required a 37°C incubation (0. Silvennoinen, D.E. Levy, unpublished data.) .

The specificity of SIF activation was tested in cell lines overexpressing growth factor receptors (Figure ID) . Mouse fibroblasts expressing the mouse colony stimu¬ lating factor 1 (CSF1) receptor (M. Mohammadi et al . , Nature 358, 681 (1992)) responded to CSF1 treatment by

stimulating SIF DNA-binding activity (lane 5) . However, acidic, basic, and k-FGF failed to activate SIF in FGF- responsive, Swiss 3T3 cells or in WT7 cells (M. Mohammadi et al . , Nature 358, 681 (1992)) overexpressing the FGF receptor 1 ( fig) gene (lanes 8-10) . Likewise, insulin treatment (lane 13) did not lead to SIF activation in cells overexpressing the human insulin receptor (C.K. Chou et al . , J. Biol . Chem. 262, 1842 (1987)) . These cell lines could be stimulated to activate SIF binding activity in response to PDGF or IFN (lanes 7 and 12) . Despite absence of SIF stimulation by FGFs and insulin, all these growth factors caused receptor autophosphorylation and an increase in tyrosine phosphorylation of cellular proteins.

Comparison of the SIF DNA sequence with IFNγ response elements (R.N. Pearse et al . , Proc. Natl . Acad.

Sci . USA 90, 4314 (1993)) suggested that it might serve as a binding site for the 91 kDa subunit of ISGF3a (p91) . To test this possibility, band shift reactions were treated with antibodies to the different components of ISGF3 (Fig- ure IE) . The PDGF-stimulated band shift was unaffected by antibodies to the p48 and pll3 components of ISGF3; in contrast, this DNA-binding activity was abrogated by anti¬ bodies to p91, concomitant with the appearance of a more slowly migrating super-shifted complex (lanes 5-7) . Speci- ficity of DNA-binding was demonstrated by competition with SIF-binding sequences (lanes 3 and 4) but absence of compe¬ tition by an irrelevant sequence (the ISRE, lane 8) . Similar involvement of p91 could be demonstrated for the binding activity stimulated by both IFNα and IFNγ.

EXAMPLE II

Tyrosine Phosphorylation in p91 Activation To test the involvement of tyrosine phosphoryla¬ tion in p91 activation, cells were labeled with ³² P-ortho- phosphate and treated with IFNs or growth factors. Human FS2 fibroblasts were labeled with ³² P-orthophosphate (0.8 mCi/ml) for 2 h and either left unstimulated or stimulated with IFNα., IFNγ, or PDGF, as indicated. Lysates were

subjected to immunoprecipitation using anti-p91c sera (C Schindler et al . , Science 257, 809 (1992)) and analyzed by SDS-PAGE, phosphoamino acid analysis, and phosphopeptide analysis, as previously described (W. Li et al . , Mol . Cell . Biol . 12, 5824 (1992)). Immunoprecipitation of labeled extracts with anti-p91 showed a stimulation of phosphoryla¬ tion of this protein in response IFNα, IFNγ, and PDGF (Figure 2A) . p91 precipitated from IFNα-treated cells (lane 2) co-precipitated with the 113 kDa subunit of ISGF3 (Schindler et al . , supra) . However, this protein was not detected in samples from IFNγ-treated or PDGF-treated cells (lanes 3 and 4) . This reflects the exclusive activation of pll3 in response to IFNα, while p91 phosphorylation occurs in response to all three ligands. Phosphoamino acid analy- sis revealed the presence of phosphotyrosine in p91 exclu¬ sively from treated samples, while untreated samples con¬ tained only phosphorylated serine residues (Figure 2B) . Peptide map analysis of tryptic digestion products revealed phosphorylation of the same, single new tryptic phospho- peptide in response to all three agents (Figure 2C) .

Further demonstration of tyrosine phosphorylation of p91 in response to growth factors involved immunoblot¬ ting with anti-phosphotyrosine antibodies (Figure 2D) . Proteins from cells treated with various factors were immunoprecipitated with anti-p91, fractionated by SDS-PAGE, and transferred to nitrocellulose. Cells were lysed in 50 mM Tris, pH 7.5, 150 mM NaCl, 1% Triton X-100, 500 μM sodium orthovanadate, 1 mM EDTA, 1 mM PMSF, 3 μg/ l apro- tinin, 0.5 μg/ml leupeptin and immunoprecipitated with anti-p91c sera. Immune complexes were separated on SDS- PAGE, transferred to nitrocellulose, analyzed using anti- phosphotyrosine monoclonal antibodies (Upstate Biotechnolo¬ gy, ICN, and Oncogene Science) or anti-p91c and HRP-labeled second antibody, and detected by enhanced chemiluminescence (Amersham) .

A 91-kDa band could be detected with anti-pTyr antibodies in extracts from IFN- or EGF-treated cells (lanes 1 and 3) , but not from cells treated with FGF (lane

2) . Thus, the 91 kDa subunit of ISGF3, or a protein close¬ ly related to it, is tyrosine phosphorylated in response to a number, but not all, growth factors, leading to its nuclear accumulation and its ability to bind specifically to DNA regulatory elements. The absence of SIF binding activity in FGF-treated cells is therefore due to the inability of this growth factor to stimulate tyrosine phosphorylation of p91.

EXAMPLE III Involvement of EGF Receptor in p91 Activation

The mitogenic action of EGFR and other receptor tyrosine kinases appears to signal through Ras and may require intervention by intermediary proteins specifically recognizing the autophosphorylated cytoplasmic tail of the receptor. To test the involvement of the EGFR in tyrosine phosphorylation of p91, cells expressing different mutant human EGFRs (N. Li et al. , unpublished data) were activated with EGF or with IFNα (Figure 3A) . Only cells expressing wild-type receptor (lane 3) showed full stimulation of p91 in response to EGF. Truncation of the EGFR cytoplasmic tail, leaving only the kinase catalytic domain and removing the major autophosphorylated tyrosines, greatly diminished the ability of EGF to stimulate p91 (lane 9) . Likewise, mutation of the five autophosphorylation sites to the phenylalanine residue also prevented the normal induction of p91 DNA-binding activity in response to EGF (lane 7) , in spite of similar levels of receptor expression (Figure 3B) .

In contrast to their responses EGF stimulation, all cells responded normally to IFNα (lanes 2, 4, 6, and 8) . This result suggests that tyrosine autophosphorylation of the receptor is required for p91 activation. Possibly, p91, an SH2-domain containing protein (C Schindler et al . , Proc . Natl . Acad. Sci . USA 89:7836 (1992) ; X.-Y. Fu, Cell 70, 323 (1992)) , is recruited directly to a tyrosine kinase activated by EGF stimulation by recognition of phosphory¬ lated tyrosine residues. Indeed, Sadowski et al . (supra) found that an SH2 domain-containing protein and antibodies

to phosphotyrosine prevented SIF activation in vi tro (Sadowski et al . , supra) .

Mutations eliminating tyrosine autophosphoryl¬ ation sites on EGFR have little effect on the mitogenic properties of EGF (S.J. Decker, J. Biol . Chem. 268, 9176 (1993)); hence, p91 activation, which requires receptor autophosphorylation, may occur independently of mitogene¬ sis.

To further assess the connection between p91 activation and mitogenesis, we investigated the involvement of Ras in p91 action in cells expressing dominant-negative mutations of H-Ras. Mutation of serine to asparagine at position 17 of H-Ras (N ¹⁷ -Ras) prevents guanine nucleotide exchange, inhibiting Ras activation in a dominant fashion and suppressing mitogenic responses to growth factors (D.W. Stacey et al . , Oncogene 6, 2297 (1991); D.W. Stacey et al . , Mol . Cell . Biol . 11, 4053 (1991)). In both parental type cells and cells expressing N ¹⁷ -Ras (E.Y. Skolnik et al . , Science 260, 1953 (1993)), PDGF stimulated p91 DNA- binding activity (Figure 3C, top) , despite abundant levels of mutant Ras proteins (bottom) . However, the N ¹⁷ -Ras mutation prevented the Ras-mediated activation of MAP kinase by PDGF (Figure 3D) , an essential step in PDGF- induced mitogenesis. GENERAL DISCUSSION OF EXAMPLES I-III

The present findings point to a distinct mecha¬ nism for growth factor signaling, directly linking receptor tyrosine kinase activation to modulation of nuclear events. This is schematically illustrated in Figure 4. EGF, PDGF, and CSF1 stimulated the tyrosine phosphorylation of a latent, cytoplasmic transcription factor, causing it to translocate to the nucleus and ac¬ quire the ability to bind DNA; this effect is similar to that observed by stimulation of IFN receptors which are not tyrosine kinases. This pathway is independent of Ras and the action of MAP kinase, which are thought to lead to activation of the fos promoter through ser/thr phosphoryl-

ations (S.A. Moodie et al . , supra; Lange-Carter et al. , supra) .

Recent studies showed that the tyrosine kinase Tyk2 is necessary for IFNα signaling (L. Velazquez et al . , Cell 70, 313 (1992)), and the related enzyme Jak2 appears to mediate IFNγ action. Cytoplasmic tyrosine kinases may also be necessary for growth factor stimulation of p91. The biologic significance of activation of a common transcription regulatory factor shared by mitogenic growth factors and by anti-proliferative, differentiation- stimulating cytokines is still unclear. Receptor tyrosine kinases can stimulate cells to proliferate or to differen¬ tiate and withdraw from the cell cycle depending upon cellular context and specific culture conditions (J. den Hertog et al . , Cell Growth Diff . 2 , 155 (1991)) . Indeed,

CSFl stimulates mitogenesis of fibroblasts transfected with the CSFl receptor (Roussel et al . , supra) but fulfills distinct signaling functions during macrophage differentia¬ tion and osteoclast activation. Likewise, EGF, FGF, and NGF can stimulate either cell division or neurite outgrowth in pheochromocytoma cells (J. Schlessinger et al . , Neuron 9, 383 (1993)) . Differential activation of distinct sets of genes following activation of ligand-specific sets of latent transcription factors may therefore modulate the outcome of a simultaneous, Ras-mediated common signal.

One might expect a lack of specificity to result from multiple signaling pathways impinging on a single transcription factor, ISGF3 p91. A potential solution is suggested by studies of IFN induction, wherein IFNα and IFNγ, both of which stimulate tyrosine phosphorylation of p91, nevertheless lead to the activation of distinct sets of genes (K. Shuai et al . , supra; R.N. Pearse et al . , supra; Y. Kanno et al . , supra) . However, only IFNα leads to the additional tyrosine phosphorylation of the 113 kDa transcription factor subunit (C Schindler et al . , Science 257, 809 (1992)) , while IFNγ and PDGF activate only p91 (see Figure 2A) . Assembly of activated p91 and pll3 into the ISGF3a complex following IFNα treatment allows associa-

tion with the 48 kDa ISGF3g subunit (D.E. Levy et al . , Genes Dev. 2 , 1362 (1989) ; D.S. Kessler et al . , Genes Dev. 4, 1753 (1990)) , a protein with a distinct DNA-binding specificity that directs the complex to a regulatory ele- ment unique to IFNα-stimulated genes (S.A. Veals et al . , Mol . Cell . Biol . 12, 3315 (1992)) .

The complete ISGF3 complex displays a DNA-binding specificity distinct from that of any of the individual subunit proteins on their own (S.A. Veals et al . , Mol . Cell . Biol . 13, 196 (1993)) . Similar specificity-determin¬ ing subunits analogous to p48 and pll3 may be involved in each signaling pathway from a distinct receptor, combining with the common p91 subunit and directing the complex to appropriate target genes. The absence of p91 involvement in insulin and FGF signaling suggests yet further complexi¬ ty. Possibly, a unique specificity of insulin or FGF receptors for proteins with distinct SH2 domains could direct them to activate distinct cytoplasmic transcription factors. Combinatorial assembly of these different sub- units would result in a high degree of potential signaling specificities.

EXAMPLE IV

Comparison of Protein Binding to the DSE Oligonucleotide Sequence and the SIE Oligonucleotide Sequence

Nuclear extracts (5 μg) from Swiss 3T3 cells were treated for 15 min with 2500 u/ml murine IFNα and the extracts mixed with ³² P-labeled DSE. The binding of pro¬ teins in the extract to the oligonucleotide were measured in a 5 μl gel band shift reaction. Each reaction contained 1.8 μg of non-specific competitor DNA plus an excess of competitor oligonucleotide. Gel band shift reactions were resolved on 5% acrylamide gels.

The results are shown in Figure 5. Tenfold molar excess of DSE competed equally to 50-fold excess of SIE. Thus, proteins bind the DSE sequence with higher affinity than to the SIE oligonucleotide sequence. Thus, the DSE oligonucleotide, or a larger oligonucleotide containing the

DSE sequence is useful as a probe for detecting a TRP-type of DNA binding protein, for isolating such a protein, and for measuring the activity of such a protein.

EXAMPLE V

Induction of DSE-Binding Proteins by Interleukin-2,

Interleukin-3 and Erythropoietin

Cells responsive to particular cytokines or growth factors were used to test the induction of proteins which bind to the DSE-containing oligonucleotide. BaF3/ER cells (D'Andrea et al . (1991) Molec. Cell. Biol. 11:1980- 1987) , responsive to erythropoietin and IL3, were grown in the absence of either erythropoietin or IL-3. LyD9, re¬ sponsive to IL-3, cells were grown for 4 hr in the absence of IL-3. CTLL cells (ATCC TIB214) , responsive to IL-2, were grown for 4 hr in the absence of IL2. The cells were then treated with their appropriate cytokine/growth factor for 15 minutes. Nuclear extracts (5 μg) prepared from each of the treated cell populations, as well as unstimulated control cells, were analyzed by gel band shift assays, as described above, using a ³² P-DSE probe.

The results, shown in Figure 6, indicate that DSE-binding proteins are induced in these cells by the cytokine/growth factor to which they are responsive: erythropoietin, IL-2 or IL-3. It is noteworthy that the IL-2 and IL-3 resulted in a DNA-protein complex of a dif¬ ferent molecular weight than that induced by IFNα. Fur¬ thermore, the present inventors found that an antibody specific for p91 does not bind to the TRPs induced by IL-2, IL-3 or erythropoietin. These results indicate that novel TRPs which bind to the DSE oligonucleotide sequence are induced by IL-2, IL-3 and erythropoietin.

The references cited above are all incorporated by reference herein, whether specifically incorporated or not. Having now fully described this invention, it will be appreciated by those skilled in the art that the

same can be performed within a wide range of equivalent parameters, concentrations, and conditions without depart¬ ing from the spirit and scope of the invention and without undue experimentation. While this invention has been described in con¬ nection with specific embodiments thereof, it will be understood that it is capable of further modifications. This application is intended to cover any variations, uses, or adaptations of the invention following, in general, the principles of the invention and including such departures from the present disclosure as come within known or custom¬ ary practice within the art to which the invention pertains and as may be applied to the essential features hereinbe¬ fore set forth as follows in the scope of the appended claims.

SEQUENCE LISTING

(1) GENERAL INFORMATION:

(i) APPLICANT: LEVY, David E.

(ii) TITLE OF INVENTION: DNA SEQUENCE WHICH BINDS TRANSCRIPTIONAL REGULATORY PROTEINS ACTIVATED IN RESPONSE TO VARIOUS CYTOKINES AND USES THEREOF

(iii) NUMBER OF SEQUENCES: 14

(iv) CORRESPONDENCE ADDRESS:

(A) ADDRESSEE: Morrison & Foerster

(B) STREET: 2000 Pennsylvania Avenue

(C) CITY: Washington

(D) STATE: D.C.

(E) COUNTRY: USA

(F) ZIP: 20006-1812

(v) COMPUTER READABLE FORM:

(A) MEDIUM TYPE: Floppy disk

(B) COMPUTER: IBM PC compatible

(C) OPERATING SYSTEM: PC-DOS/MS-DOS

(D) SOFTWARE: PatentIn Release #1.0, Version #1.25

(vi) CURRENT APPLICATION DATA:

(A) APPLICATION NUMBER:

(B) FILING DATE:

(C) CLASSIFICATION:

(vii) PRIOR APPLICATION DATA:

(A) APPLICATION NUMBER:

(B) FILING DATE:

(viii) ATTORNEY/AGENT INFORMATION:

(A) NAME: LIVNAT, SHMUEL

(B) REGISTRATION NUMBER: 33,949

(C) REFERENCE/DOCKET NUMBER: 15661-20010.00

(ix) TELECOMMUNICATION INFORMATION:

(A) TELEPHONE: (202) 887-1500

(B) TELEFAX: (202) 887-0763

(C) TELEX: 90-4030

(2) INFORMATION FOR SEQ ID NO:l:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 11 base pairs

(B) TYPE: nucleic acid

(C) STRANDEDNESS : single

(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:l: TTTCCNGGAA A 11

(2) INFORMATION FOR SEQ ID NO:2:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 11 base pairs

(B) TYPE: nucleic acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:2: TTTCCGGGAA A 11

(2) INFORMATION FOR SEQ ID NO:3:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 28 base pairs

(B) TYPE: nucleic acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:3: GATCTAGGGA TTTCCGGGAA ATGAAGCT 28

(2) INFORMATION FOR SEQ ID NO:4:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 28 base pairs

(B) TYPE: nucleic acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:4: ATCCCTAAAG GCCCTTTACT TCGACTAG 28

(2) INFORMATION FOR SEQ ID NO:5:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 11 base pairs

(B) TYPE: nucleic acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:5: GTTCCCGTCA A 11

(2) INFORMATION FOR SEQ ID NO:6:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 10 base pairs

(B) TYPE: nucleic acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:6: TTTACGGGAA 10

(2) INFORMATION FOR SEQ ID NO:7:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 11 base pairs

(B) TYPE: nucleic acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:7: TTTAGAGTAA T 11

(2) INFORMATION FOR SEQ ID NO:8:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 11 base pairs

(B) TYPE: nucleic acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:8: TTTATAGTAA G 11

(2) INFORMATION FOR SEQ ID NO:9:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 11 base pairs

(B) TYPE: nucleic acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:9: TTTCTGGGAA A 11

(2) INFORMATION FOR SEQ ID NO:10:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 11 base pairs

(B) TYPE: nucleic acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:10: ATTTGGAGAA A 11

(2) INFORMATION FOR SEQ ID NO:11:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 21 base pairs

(B) TYPE: nucleic acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO-.11: AGTGATTTCT CGGAAAGAGA G 21

(2) INFORMATION FOR SEQ ID NO:12:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 21 base pairs

(B) TYPE: nucleic acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:12: CATCATTTCG GGGAAATCAG G 21

(2) INFORMATION FOR SEQ ID NO:13:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 11 base pairs

(B) TYPE: nucleic acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:13 AAAGGCCCTT T 11

(2) INFORMATION FOR SEQ ID NO:14:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 28 base pairs

(B) TYPE: nucleic acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:14: GATCAGCTTC ATTTCCCGTA AATCCCTA 28