WO1989002472A1

WO1989002472A1 - Regulation of expression of gm-csf gene

Info

Publication number: WO1989002472A1
Application number: PCT/AU1988/000370
Authority: WO
Inventors: Mary Frances Shannon; Mathew Alexander Vadas
Original assignee: Amrad Corporation Limited
Priority date: 1987-09-21
Filing date: 1988-09-21
Publication date: 1989-03-23
Also published as: EP0391911A1; JPH03502040A; EP0391911A4

Abstract

A method for controlling the expression of granulocyte/macrophage colony stimulating factor (GM-CSF) in cells which express GM-CSF, or for controlling the expression of other haemopoietic cytokines which contain the Ck-1 or Ck-2 DNA sequences in cells where they are expressed, which comprises the step of regulation of the binding of nuclear protein(s) in said cells with Ck-1 and Ck-2 in the promoter region of the GM-CSF gene or the said haemopoietic cytokine genes. The step of regulation of the binding of the nuclear protein(s) with Ck-1 and Ck-2 sequences may consist of the induction or promotion of binding, or alternatively, it may consist of the prevention or inhibition of binding. By way of example, regulation of the binding of the nuclear protein(s) may be effected by the use of DNA of a sequence which will act as a competitive inhibitor in the binding of protein(s).

Description

REGULATION OF EXPRESSION OF GM-CSF GENE
This invention relates to the regulation of the expression of the gene for human granulocyte/macrophage colony stimulating factor (GM-CSF), as well as to diagnosis of leukaemia and other related diseases which might be associated with abnormalities in the cells expressing GM-CSF.

Human granulocyte/macrophage colony stimulating factor (GM-CSF) is a 22-kDa glycoprotein which stimulates the formation of granulocyte, macrophage, granulocyte/macrophage and eosinophil colonies from normal bone marrow progenitor cells in vitro (1). GM-CSF has a direct action also on the function of mature peripheral blood granulocytes (2,3). The gene encoding human GM-CSF is 2.5 -3 kb long (4,5) and maps to the long arm of chromosome 5 (6).

Antigen or mitogen activated T cells and T cell lines produce relatively high levels of GM-CSF (7,8,9).

Other cytokines such as interleukin-l and tumour necrosis factor (TNF) activate the expression of the GM-CSF gene in endothelial cells (10). Primary human stromal cells can also be induced to produce GM-CSF (11).

It appears that there may be multiple forms of control of GM-CSF production, both at a transcriptional and post- transcriptional level. mRNA stability has been shown to be involved in controlling the induction of mouse
GM-CSF in macrophages (12) and a sequence in the 3' untranslated region is responsible for the instability of
GM-CSF mRNA(13). It was recently reported that a 650bp fragment from the promoter region of the hGM-CSF was involved in T cell-specific expression of the gene and was only active in PHA/PMA-stimulated T cells (14).

Computer analysis of the GM-CSF genes from mouse and man has shown that the most highly conserved sequences are in the promoter region and in the 3' non-coding sequence of the mRNA (5), indicating the potential importance of these regions in the regulation of expression of these genes. Homologies with other cytokines can also be found especially in the promoter region of the genes (5,15,16).

It has now been discovered that sequences within the promoter region of the GM-CSF gene, especially those sequences shared with other cytokine genes, are the binding sites for nuclear proteins which confer cellspecificity or inducibility on the GM-CSF gene.

Accordingly, in a first aspect of the present invention, there is provided a method for controlling the expression of granulocyte/macrophage colony stimulating factor (GM-CSF) in cells which express GM-CSF, or for controlling the expression of other haemopoietic cytokines containing CK-1 or CK-2 in cells where they are expressed, which comprises the step of regulation of the binding of nuclear protein(s) in said cells with the promoter region of the GM-CSF gene.

The step of regulation of the binding of the nuclear protein(s) with the promoter region of the GM-CSF gene may consist of induction or promotion of such binding, or alternatively it may consist of prevention or inhibition of such binding. By way of example only, regulation of the binding of the nuclear protein(s) may be effected by inhibition of protein binding using specific inhibitors having affinity for CK-1 or CK-2, or by use of DNA of an appropriate sequence to act as a competitive inhibitor in the binding of protein.

In a particular aspect of this invention, it is the binding of the nuclear protein(s) with the region within the GM-CSF promoter region which is close to the transcription start site, more particularly the region spanning the cytokine-l (CK-1) and/or cytokine-2 (CK-2) specific sequences of the promoter region, which is regulated. By way of example, the method of this aspect of the invention includes induction of the formation of the complex described herein as nuclear factor (NF)-GMb.

NF-GMb is thus formed by interaction of nuclear protein from GM-CSF expressing cells with the promoter region spanning the CK-1 and CK-2 specific sequences. The NF-GMb complex then regulates the transcription of this gene. As described herein, NF-GMb is inducible and accompanies production of GM-CSF mRNA. Furthermore this complex is absent in cell lines not producing GM-CSF.

In another aspect, the present invention extends to the diagnosis of leukaemia and other related diseases which might be associated with abnormalities in the expression of GM-CSF. In this aspect of the invention detection of abberations or abnormalities in the identity of the nuclear protein(s) or the promoter region to which these nuclear protein(s) bind, or in the nature of the binding between the nuclear protein(s) and the promoter region, provides an indication of altered production of
GM-CSF. Abnormalities which may arise in this regard would include lack of capacity of the protein(s) to bind to DNA. Such an abnormality may, for example, be detected by a DNA-gel retardation assay as described herein, or by competitive ELISA techniques.

Abnormalities associated with lack of capacity to activate transcription may be detected by use of an in vitro transcription system to screen for function. The gel retardation assay as described herein may also be used as a screen for agents which might inhibit the binding of nuclear protein(s) to the promoter region.

In work leading to the present invention, the interaction of nuclear proteins from cells which express
GM-CSF and cells where GM-CSF is not produced with synthetic oligonucleotides spanning conserved regions'of the GM-CSF promoter, has been examined. As a result of this work, it has been shown.that there are cell-specific as well as inducible nuclear proteins which interact with
DNA fragments from the GM-CSF promoters.

Comparison of the promoter regions from a number of cytokine genes has revealed some homologous sequences conserved between different genes and across species (5,14,15). One decanucleotide sequence (cytokine (CK)-1), 5'GRGRTTYCAY3' (R=A or G; Y=C or T) is found in both human and murine IL-2, IL-3, GM-CSF and G-CSF genes (Fig 4). In addition a second sequence (CK-2), 5'TCAGRTA3', lying 3' to the decanucleotide, is conserved in both human and murine GM-CSF and IL-3 genes (Fig 4). This sequence is not found in human or murine G-CSF or IL-2. The CK-1 sequence is also repeated further upstream in the human
GM-CSF (14) and murine IL-3 (31) genes but without the extra flanking conserved sequence.

A 41bp oligonucleotide probe spanning the CK-1 and CK-2 sequences has been designed to elucidate their role in nuclear protein binding. These two sequences are flanked by GM-CSF sequences not conserved-in the other genes.

Two DNA-protein complexes which are specific for the 41bp oligonucleotide spanning these conserved sequences have been identified. It would appear that these two complexes are generated by two or more different proteins since the proteins involved in these complexes (described herein as NF-GMa and NF-GMb) elute from a heparin-sepharose column at 0.1-0.3M KC1 and 0.6M KC1, respectively.

The CK-1 sequence has been postulated to account for the co-ordinate expression of some of these genes in activated T cells (14,15). However, G-CSF is not expressed in activated T cells (13) despite the fact that its promoter contains -a copy of CK-1. It seems unlikely, therefore, that this sequence alone is responsible for T cell expression of GM-CSF or IL-3. However; the extended conserved sequence of CK-1 and CK-2 noted above between
GM-CSF and IL-3 genes may be involved in T cell expression of these genes. It is noteworthy that IL-3 and GM-CSF are often co-ordinately expressed following Con A stimulation of murine T cell clones (35).

Treatment of U5637 cells with PMA results in an increase in GM-CSF mRNA and an increase in the level of the NF-GMb complex but not NF-GMa. The ability to induce the NF-GMb complex in PMA-treated U5637 cells, concomitant with GM-CSF mRNA production suggests that the protein(s) involved in the NF-GMb complex may be responsible for the inducibility of the GM-CSF gene. The situation observed here parallels that found with proteins binding to the octamer motif of the immunoglobulin genes (25,26). Two proteins, one of which is lymphoid specific, bind to this motif. The lymphoid specific protein is inducible with lipopolysaccharide (26) as is NF-GMb with PMA in US 637 cells. Extracts from a human melanoma LiBr cell line contain the protein(s) responsible for NF-GMa, but NF-GMb could not be induced by PMA treatment of these cells.

Nuclear proteins prepared from the mouse SP2 cell line did not bind specifically to the GM-CSF conserved sequence.

It would appear, therefore, that the protein(s) which bind to this GM-CSF sequence are limited in their cellular distribution. The result is consistent with previous results which show that a GM-CSF/CAT fusion gene is expressed in activated T cells but not in a human
B-lymphoblastoid cell line (14).

Further features of the present invention will be apparent from the follow.ing detailed description and drawings. In the drawings:
Figure 1 shows (a) Restriction enzyme map of the promoter region of the human GM-CSF gene. The sequence is numbered from the start of transcription (+1). The sequence between the SacI site at -11 and -96 is expanded to show both the sequence conserved between the cytokine genes and the TATA box sequence. P;PstI,S;SacI, D;Dde I.

(b) The sequence of the 41bp synthetic oligonucleotide which span the conserved GAGATTCCAC and TCAGGTA sequences. The conserved regions are underlined.

Fiaure 2 shows interaction of nuclear proteins from
PMA-treated U5637 cells with synthetic oligonucleotides.

(a) Competition assays to determine the specificity of binding to the synthetic 41bop annealed oligonucleotides 0.2ng radiolabelled oligonucleotides was mixed with 2pg nuclear protein and 2 pg poly (dI:dC). Increasing concentrations of the specific oligonucleotide (GM) (lanes 2-5) or unrelated oligonucleotide (X) (lanes 7-10) were added; lanes 1 & 6, no competitor; lanes 2 & 7, 5ng lanes 3 & 8, lOng; lanes 4 & 9, 25ng; lanes 5 & 10, 50ng; a and b indicate the two specific complexes NF-GMa and NF-GMb, F indicates the unbound oligonucleotide. (b) Separation of
U5637 nuclear extract on heparin sepharose. The KC1 concentration (M) used for elution is indicated under each track.

Lanes 1, 3, 5 and 7 had 50mM KC1 in the binding reactions and lanes 2, 4, 6 and 8 had 200mM KC1.

Figure 3 shows: (a) Interaction of nuclear proteins, from different cell lines after PMA treatment, with the synthetic oligonucleotides. Extracts from PMA-treated
HUT78 cells (2pig) (lane 1), U5637 cells (2pg) (lane 2), SP2 cells (6pg) (lane 3), LiBr cells (6pig) (lane 4) were used with 0.2ng radiolabelled oligonucleotides; lane 5 had not protein. a and b indicate the specific complexes and F indicates the unbound oligonucleotide.

(b) Effect of PMA treatment on the formation of the
NF-GMa and NF-GMb complexes. Lane 1, untreated U5637 cells; lane 2, PMA-treated U5637 cells; lane 4, untreated
LiBr cells, lane 5, PMA-treated LiBr cells.

Figure 4 shows the conserved sequences found in the promoter region of cytokine genes. All the sequences have been compared to the human GM-CSF sequence. The non conserved bases are shown in small letters. The numbering is relative to the transcription start site of each gene (+1).

Figure 5 shows densitometer scans of gel retardation assays in which various oligonucleotides were assessed for the degree of competition of binding with a radio labelled
GM-CSF sequence incorporating CK-1.

Fiaure 6 shows gel retardation assays using various oligonucleotides incorporating CK-1 and HUT78 T-cell extract.

Fiaure 7 shows transfectioh of Jurkat cells with single or multiple copies of CK-1 from either GM-CSF or G-CSF cloned upstream of tkCAT in pBLCAT2. Lane 1, pGCK-1 +
PMA; lane 2, pGCK-l; lane 3, pGMCK-l + PMA; lane 4, pGMCK-l; lane 5, pG(4)CK-I + PMA; lane 6, pG(4)CK-l; lane 7, pGM(5)CK-1 + PMA; lane 8, pGM(S)CK-l; lane 9, pBLCAT2; lane 10, 1 unit of CAT enzyme. (The number in brackets after the pG or pGM symbols is the number of copies of the sequence cloned upstream of the tk promoter.) Fiaure 8 shows identification of the molecular weight of
NF-GMa by photocross linking. Track 1, NF-GMa complex; track 2; + 200 fold excess cold GM oligo; track 3, + 200 fold excess cold x/y oligo. Molecular.weight markers are indicated beside the gel.

Fiaure 9 shows TNF-a induction of NF-GMa. HUVE cells (a) or FLOW cells (b) were grown in the presence of TNF-a for the indicated times. NF-GMa was detected in nuclear extracts using the GM-CSF sequence as a radio labelled probe. a, NF-GMa; f, unbound DNA.

Figure 10 shows that CK-1 responds to TNF-a. FLOW cells were transfected with the following DNA samples: (1) no DNA; (2) and (3) pSV2 CAT; (4) and (5) pBL
CK-1(4); (6) and (7) pBL CAT2. Lanes 2, 4 and 6 were treated for lOhr with 100 units/ml TNF-a before harvesting and assaying for CAT activity.

MATERIALS AND METHODS
DNA probes.

Two complementary 41bp oligonucleotides were synthesiaed with EcoRI ends (Fig.lb). Each oligonucleotide was end-labelled with y-32P-ATP (Bresa) and polynucleotide kinase (Pharmacia). The radiolabelled oligonucleotides were annealed by heating to 1000C for 3 min in 25mM Tris-HCl pH7.6, 150mM NaCl and cooling at room temperature for 15 min. Unlabelled oligonucleotides were also annealed as described above to give a final concentration of lOng/pl and used as specific competitors in the binding reactions.

Cell lines
U5637 is a human bladder carcinoma cell line which constitutively produces GM-CSF and G-CSF (18). HUT78 is a
T-lymphoblastoid cell line derived from a patient with
Sezary syndrome (19). SP2 10-Ag 14 is a mouse myeloma cell (20) and LiBr is a human melanoma cell line (21).

Jurkat cells are a human t-lymphoblastoid cell line; FLOW cells are an embryonic fibroblast primary cell line, and human umbilical vein endothelial cells are also primary cultures. GM-CSF mRNA can be detected in U5637 and HUT78 cells but not in LiBr and SP2 cells.

All cell lines were routinely grown in RPMI medium supplemented with 10% foetal calf serum. Cells were harvested at lox106 cellsSml or at 80% confluence for non-adherent and adherent cell lines respectively, following treatment for 6hr with l0ng/ml of phorbol-12 myristate-13-acetate (PMA). Untreated cells were grown for the same period of time but without PMA.

Preparation of nuclear extracts.

Nuclei were prepared as described by Dignam et.al.

(22). Nuclear proteins were extracted by constant agitation for 30 min with 0.5M KC1 at 40C. Following centrifugation at 20,000 rpm for 30 min (Beckman TL100.3) the supernatant was dialysed against three changes of TM buffer containing 100mM KC1 for 12-16 hrs. (TM buffer is 50mM Tris-HCl pH7.6, 12mM MgCl2, lmM EDTA, lmM DTT, 20% glycerol; 23). The protein extracts were stored at -70tC. Protein concentration was determined using the
Bio-Rad assay (Bio-Rad).

Gel retardation assays.

For binding reactions, 0.1-1.0 ng of radiolabelled fragment (5-10,000 cpm) was mixed with 1-3pg of nuclear extract in a final volume of 20p1 containing 25mM
Tris-HCl pH7.6, 6.25mM MgCl2, 0.5mM EDTA, 0.5mM DTT, 10% glycerol and 80-200mM KC1. 0.5-3pg of poly (dI:dC) was used in all reactions as non-specific competitor.

Specific. competitors were added to each reaction as described in individual experiments. The reactions were analysed on 5% polyacrylamide gels in low ionic strength buffer (24). The gels were pre-electrophoresed at 20V/cm for 2 hr and electrophoresed at the same voltage for 1-2 hr. Following electrophoresis the gels were dried and autoradiographed overnight or for 1-2 days.

Binding reactions with synthetic oligonucleotides were as described above. The retardation patterns were analysed on 11% polyacrylamide gels in 0.5 x TBE (1.0 x
TBE is 50mM Tris/borate pH8.3, lmM EDTA).

Heparin-Sepharose column chromatography.

A lml heparin-sepharose (Pharmacia) column was equilibrated with TM buffer containing 100mM KC1 (TM.1).

Six mg of crude nuclear extract from PMA-treated U5637 cells was loaded onto the column in the- same buffer.

Following extensive washing with TM.1 the bound protein was eluted in a stepwise fashion with 3ml each of 200, 300 and 600mM KC1 in TM buffer. The eluates from each salt concentration were collected, dialysed into TM.1 and the protein concentration estimated with the Bio-Rad assay.

The fractions were tested for binding activity to the synthetic oligonucleotides.

Transfection of Jurkat cells.

Jurkat cells were passage several days before transfection and grown to a density of lx106 cells/ml in
RPMI supplemented with 10% foetal calf serum. Cells were harvested and washed twice in electroporation buffer (20mM
HEPES, pH 7.6, 157mm NaCl, 5mM KC1, 0.1% glucose) and resuspended at lox107 cells/ml for electroporation.

Aliquots of 2x106 cells were mixed with 15pg DNA and 10% foetal calf serum and placed on ice for 15 mins.

Electroporation was with a Bio-Rad Gene Pulsar with settings of 270V and 960pF. The cuvettes were placed on ice for a further 10 min and then transferred to a T25 flask with 8ml RPMI with 10% foetal calf serum and incubated for 40-48 hr before harvesting. For stimulation, cells were treated with PMA (20pg/ml) and
PHA (2pg/ml) for 16 hr before harvesting.

Cells were harvested by centrifugation, washed in
PBS and cytoplasmic extracts prepared according to Gorman et.al. (36). Protein concentration was assayed and CAT assays performed on all extracts using 25pg of protein in each assay.

RESULTS
A conserved cytokine-specific sequence binds nuclear proteins.

Two complementary oligonucleotides (each 41 bp long) spanning the sequence 5'GAGATTCCAC3' (Fig.lb), were synthesized in order to investigate the interaction of nuclear proteins with this sequence. Two specific retarded complexes, labelled a and b in Fig.2a, were generated with extracts from PMA-stimulated HUT78 or U5637 cells. These complexes will be referred to as nuclear factor (NF)-GMa and NF-GMb. These two complexes could be competed out completely with 50ng of unlabelled annealed oligonucleotides (Fig.2a, lanes 1-5) but not by the same concentration of an unrelated oligoncleotide (Fig.2a, lanes 6-10) nor with 3pg of poly (dI:dC) (not shown).

Other retarded complexes running higher on the gel or closer to the free DNA are not consistently observed and cannot be competed with increasing concentrations of specific competitor (Fig.2a). Increasing the salt concentration in the binding reactions from 80mM to 200mM greatly reduces this non-specific interaction and enhances the specific interactions by approximately 3-fold (data now shown). Subsequent binding reactions were therefore carried out at 200mM KC1.

The two retarded complexes could result either from interaction with multimers of the same protein or with two distinct proteins. The nuclear extract from U5637 cells was fractionated on a heparin-sepharose column. The proteins eluting from the column with 0.1M, 0.2M, 0.3M and 0.6M KCl were tested in retardation assays (Fig.2b). The protein(s) responsible for the NF-GMa complex eluted from the column with 0.1-0.3M KC1 and that for NF-GMb with 0.6M
KC1 (Fig.2b, lanes 5, 6 and 7, 8), suggesting the involvement of two distinct proteins in these complexes.

Cell specific interactions with the conserved cytokine sequence.

We have compared the retardation band patterns obtained with extracts prepared from U5637, HUT78, LiBr and SP2 cell lines. The extracts were all prepared from cells treated for 6hr with 10ng/ml PMA. Both of the specific retarded bands, NF-GMa and NF-GMb were obtained using the radio labelled oligonucleotides and extracts from
PMA-treated U5637 and HUT78 cells (Fig.3a, lanes 1 and 2), although HUT78 cell extracts always yielded a 3-4 fold concentration of the proteins involved in both these complexes. Extracts from the SP2 cell line did not result in either of the specific complexes but gave a diffuse retarded band migrating above NF-GMa (Fig.3a, lane 3).

This interaction is competed with increasing concentrations of poly (dI:dC). Extracts from PMA-treated
LiBr cells bound to the GM-CSF specific oligonucleotide giving rise to NF-GMa but not NF-GMb. The amount of
NF-GMa formed with extracts from LiBr cells was always approximately 3-fold lower than that seen in HUT78 or
U5637 cells (Fig.3a, lane 4).

Effect of PMA stimulation on the formtaion of NF-GMa and
NF-GMb DNA-protein complexes.

In order to determine if PMA treatment induced the production of the proteins involved in the two cytokine specific complexes, extracts were prepared from the U5637 and LiBr cell lines before and after PMA treatment. The retarded patterns obtained with these extracts indicate that in U5637 cells the protein involved in the NF-GMb complex was induced 5-10 fold (Fig.3b, compare lanes 1 and 2). In LiBr cells this induction was not observed and no change in the level of NF-GMa was seen (Fig.3b, lanes 3 and 4).

Definition of binding sites of nuclear proteins.

As previously described, the CK-1 sequence is conserved across a large number of cytokines, notably
IL-3, G-CSF and IL-5 (see Table 1). CK-2 on the other hand is present only in GM-CSF and IL-3.

TABLE 1 CK-1 SEOUENCES hGM (1) G A G A T T C C A C hGM (2) G g G A T T a C Ag 7/10 mGM G A G A T T C C A C 10/10 hIL-3 G A G g T T C C A t 8/10 mIL-3 (1) G A G g T T C C A t 8/10 mIL-3 (2) G A G A T T C C A C 10/10 hIL-2 G g G A T T t C A C 8/10 mIL-2 G g G A T T t C A C 8/10 h-G G A G A T T C C A C 10/10 m-G G A G A T T C C c C 9/10 hIL-5 a A G A T T C t t C 7/10 hIL-6 a g G t T T C C A a 6/10 h-IFN (R) G A G t T T C C t t 7/10 hIL-4 (R) G A a A T T a C A C 8/10 consensus G R G N T T N C N N
In order to define the precise binding sites of the nuclear proteins, double stranded oligonucleotides spanning the CK-1 sequence and flanking DNA from a number of cytokine gene regions have been constructed (see Table 2)r and used in competition assays.

It should be noted that these oligonucleotides of between 35-45 base pairs were totally different except for the ten base pairs of the CK-1 sequence. The exceptions are GM-CSF and IL-3, where the CK-2 sequence is also present
TABLE 2
GM
CK1 CK2 5' AATTCTGATAAGGGCCAGGAGATTCCACAGTTCAGGTAGTTG3'
3 'GACTATTCCCGGTCCTCTAAGGTGTCAAGTCCATCAACTTAA5'
X 5 'TCGAGAGCTCCCGGGTCGACTGCAGAAGCTTC3'
3 'CTCGAGGGCCCAGCTGACGTCTTCGAAGAGCT5'
G 5' AATTCAAAGGAGGATCAGAGATTCCACAATTTCACAAAACG3'
3 'GTTTCCTCCTAGTCTCTAAGGTGTTAAAGTGTTTTGCTTAA5'
IL3 5'CTGTGGTTTTCTATGGAGGTTCCATGTCAGATAAAGATCC3' 3' GACACCAAAAGATACCTCCAAGGTACACTCTATTTCTAGGS'
ILS 5' GATGTATTAACCCAAAGATTCTTCGTAATAGAAAAT3' 3'CTACATAATTGGGTTTCTAAGAAGCATTATCTTTTA5'
CK2 5'TCGAGAGCTCCTCAGGTAACTGCAGAAGCTTC3' 3 'AGCTCTCGAGGAGTCCATTGACGTCTTCGAAG5'
Gel

retardation assays in which an oligonucleotide based on the GM-CSF sequence incorporating CK-1 was radio labelled with 32P and mixed with nuclear proteins from cells that only produced NF-GMa (HUT 78 T-cell extract) resulted in a single specific retarded band (see
Fig.6). The oligonucleotides in Table 2 were assessed for the degree of competition of binding. Gels were densitometer scanned and the percent competition calculated from controls with no competitor added. Fig.5 plots the densitometer scans and shows that there was efficient competition for binding not only by "cold"
GM-CSF oligonucleotide but also those derived from IL-3,
G-CSF and IL-5 incorporating the CK-1 region In contrast, oligonucleotides not having a CK-1 or CK-1 like sequence (X) or only having CK-2 did not compete.

Some decrease in binding was observed with the controls but this was less and more inconsistent than competition where the oligonucleotide incorporated the CK-1 region. These results demonstrate that NF-GMa is a distinct protein binding only to the CK-1 motif.

Transcriptional properties of CK-1.

To determine the functional significance of the CK-1 sequence, transient transfection experiments were carried out in Jurkat cells. Plasmids were constructed with either single or multiple copies of the CK-1 sequence from
GM-CSF or G-CSF upstream of the thymidine kinase (tk) gene in pBLCAT (37) (Fig.7). The GM-CSF sequence also contains the CK-2 sequence which is necessary for binding of
NF-GMb These plasmids were transfected into Jurkat cells by electroporation. Cell extracts were assayed for CAT activity (36) 48 hr after transfection.

A single copy of either the GM-CSF or G-CSF CK-1 sequence cloned in either orientation upstream of the tk promoter enhanced transcription by 2-3 fold (Fig.7).

Multiple copies of either sequence appeared to have an additive effect enhancing transcription from the tk promoter by 5-10 fold (Fig.7). Treatment of the transfected cells with phorbol myristate acetate (PMA) and phyto-haemagluttinin (PHA) for 12 hr prior to harvest resulted in approx. 30% increase in transcription. The vector pBLCAT2 did not respond to PMA/PHA stimulation.

These results show that CK-1, and its binding protein
NF-GMa, acts as a positive regulator of transcription. It has the properties of a transcriptional enhancer sequence, being functional in either orientation relative to the direction of transcription and multiple copies function in an additive manner.

Identification of the proteins involved in the NF-GMa complex.

To identify the protein(s) responsible for forming the NF-GMa complex a modification of the uv-cross linking method of Wu et.al. (38) was employed. A 40 bp fragment of DNA containing the G-CSF CK-1 sequence was labelled with 32P-dATP and bromodeoxyuridine (BUdR) by chain elongation. The probe was incubated with crude nuclear extract from the HUT78 cells in a scaied up binding reaction (see Materials and Methods). The substitution of
T residues with BUdR did not interfere with NF-GMa complex formation. The NF-GMa complex was excised from a standard polyacrylamide retardation gel following exposure of the gel to uv light and autoradiography. The gel slice was incubated in SDS load buffer and electrophoresed on a 10%
SDS Laemmli gel.

A single protein of molecular weight 43 kD was identified on the protein gel (Fig.8). When 50 fold molar excess CK-1 competitor was added to the binding reaction this protein band was eliminated but in a reaction containing the same quantity of an unrelated DNA fragment as competitor, the 43 kD protein was present (Fig.8).

NF-GMa and NF-GMb in leukaemia.

Nuclear proteins were extracted by normal procedures from peripheral blood mononuclear cells prepared from patients with acute myeloid (AML) or other forms of leukaemia. All samples checked had > 80% leukaemic blasts. The presence of NF-GMa and NF-GMb were determined by the standard gel retardation procedure (described in
Materials and Methods) using the GM-CSF 40bp fragment of
DNA as a radiolabelled probe. A number of normal samples including purified lymphocytes, neutrophils and monocytes were also tested. The results are summarised in the attached Tables 3 and 4.

The NF-GMa protein is present in most leukaemic or normal cells tested. This finding is in agreement with the previous finding of NF-GMa in most cell lines in culture. On the other hand, NF-GMb is absent in all normal cells so far tested with one exception. One normal lymphocyte sample contained low levels of NF-GMb. Of the 24 AML samples tested 8 contained NF-GMb, i.e. 33% of samples. Of the 11 samples from other leukaemias tested, only one T cell acute lymphocytic leukaemia contained
NF-GMb.

If, as proposed, the induction of NF-GMb is required for GM-CSF expression then the presence of NF-GMb in 30% of acute myeloid leukaemias may lead to the expression of
GM-CSF in these cells. Recently, it has been shown that of 22 AML samples analysed for GM-CSF expression by
Northern blotting of mRNA, half contained mRNA for
GM-CSF. On the other hand, analysis of some chronic myeloid leukaemia (CML) or acute lymphocytic leukaemia (ALL) samples did not reveal GM-CSF mRNA (39). This finding of NF-GMb in some cases of AML may be correlated with the ability of these cells to synthesize GM-CSF.

Therefore it may be feasible to diagnose a subset of AML patients, (i.e. those expressing GM-CSF) by screening for the presence of NF-GMb.

TABLE 3 NF-GMa and NF-GMb in leukaemic cells.

Disease No.of samples Presence of
GMa GMb
AML 24 23 8
ALL 3 2
CML 2 1
CLL 2
Lymphoma 1 1
Eosinophilic
leukaemia 1 1
CGL 1 1
NHL 1 1
ALL, acute lymphocytic leukaemia;
CML, chronic myeloid leukaemia;
CLL, chronic lymphocyte leukaemia;
CGL, chronic granulocyte leukaemia;
NHL, non-Hodgkin's lymphoma.

TABLE 4 Presence of NF-GMa and NF-GMb in
normal blood cells.

Cell type Presence of
GMa GMb
Lymphocytes (7)* 6
Neutrophils (10) 10
Monocytes (2) 1
T Blasts** (2) 2 * refers to the number of different samples tested.

** Normal T-cells grown for 1 week in presence of PHA.

Induction of NF-GMa in endothelial cells and fibroblasts.

Treatment of endothelial cells or fibroblasts with
TNF-a leads to an induction of GM-CSF and G-CSF mRNA (10). The mechanism of induction has been shown to be at the transcriptional level (40). Because NF-GMa and NF-GMb may be involved in GM-CSF transcription, the response of these proteins to TNF-a treatment has been analysed.

Both human umbilical vein endothelial (HUVE) cells and embryonic fibroblasts (FLOW) contain low levels of NF-GMa and barely detectable levels of NF-GMb. Treatment of either cell type with 100ng/ml TNF-a induces the level of NF-GMa binding to the GM-CSF 40bp sequence (Fig.9a and b). This increase is time dependent reaching maximum levels at 6hr treatment and decreasing to almost basal levels by 24hr. No consistent changes are seen in the level of NF-GMb (not shown).

The induction of NF-GMa can be correlated with the ability of CK-1 sequence to direct TNF-a dependent transcription when transfected into FLOW cells. A plasmid containing 4 copies of the G-CSF CK-1 sequence upstream of the tk promoter and the CAT reporter gene was transcribed at high levels only after TNF-a treatment of the cells for 10-12hr (Fig.10).

Therefore under certain conditions in some cell types
NF-GMa is an inducible protein which may be involved in
GM-CSF induction by TNF-.

Those skilled in the art will appreciate that the invention described herein is susceptible to variations and modifications other than those specifically described. It is to be understood that the invention includes all such variations and modifications which fall within its spirit and scope.

REFERENCES: 1. Metcalf, D., Begley, C.G., Johnson, G.R., Nicola,
N.A., Vadas, M.A., Lopez, A.F., Williamson, D.J.,
Wong, G.G., Clark, S.C. and Wang, E.A. (1986).

Blood 67:37-45.

2. Vadas, M.A., Nicola, N. and Metcalf, D. (1983).

J.Immunol. 130:795-799.

3. Lopez, A.F., Williamson, D.J., Gamble, J.R., Begley,
C.G., Harland, J.M., Klebanoff, S.J., Waltersdorph,
A., Wong, G., Clark, S.C. and Vadas, M.A. (1986).

J.Clin.Invest. -Ai7:1220-1228.

4. Huebner, K., Isobe, M., Croce, C.M., Golde, D.W.,
Kaufman, S.E. and Gasson, J.C. (1985). Science 2;1Q:
1282-1285.

5. Myatake, S., Otsuka, T., Yokota, T., Lee, R. and
Arai, K. (1985). EMBO J. 4:2561-2568.

6. Le Beau, M.M., Westbrook, C.A., Diaz, M.O., Larson,
R.A., Rowley, J.D., Gasson, J.C., Golde, D.W. and
Sheer, C.J. (1986) Science, 231:984-987.

7. Rimaldi, A., Young, D.C. and Griffin, J.D. (1987),
Blood 69:1409-1413.

8. Wong, g.G., Witek, J.S., Temple, P.A., Wilkens,
K.M., Leary, A.C., Luxenburg, D.P., Jones, S.S.,
Brown, E.L., Kay, R.M., Orr, E.C., Shoemaker, C.,
Golde, D.W., Kaufman, R.J.,-Hewick, R.M., Wang, E.A.

and Clark, S.C. (1985) Science 228:810815.

9. Chen, I.S.Y., Quan, S.G. and Golde, D.W. (1983) Proc.Natl.Acad.Sci.USA. 80:7006-7009.

10. Munker, R., Gasson, J.., Ogawa, M. and Koeffler,
H.P. (1986) Nature, London 323:79-82.

11. Clarke, S.C. and Kamen, R. (1987) Science
236:1229-1237.

12. Thorens, B., Mermod, J-J, and Vassalli, P. (1987),
Cell 48:671-679.

13. Shaw, G. and Kamen, R. (1986) Cell 46:659-667.

14. Chan, J.Y., Slamon, D.J., Nimer, S.D., Golde, D.W.

and Gasson, J.C. (1986) Proc.Natl.Acad.Sci.

A8:8669-8673.

15. Stanley, E., Metcalf, D., Sobieszczuk, P., Gough,
N.M. and Dunn, A.R. (1985), EMBO J. 4:2569-2573.

16. Tsuchiya, M., Raziro, Y. and Nagata, S. (1987),
Eur.J.Biochem. 165, 7-12.

17. Maniatis, T., Fritsch, E.F. and Sambrook, J. (1982),
Molecular Clonina. A Laboratory Manual. Cold
Spring Harbor.

18. Welte, K., Platzer, E., Lu, L., Gabrilove, J.L.,
Levi, E., Mertelsmann, R. and Moore, M.A.S. (1985),
Proc.Natl.Acad.Sci.USA 82:1529-1530.

19. Gazder, A.F., Carney, D.N., Bunn, P.A., Russell,
E.K., Jaffe, E.S., Schechter, G.P. and Guccion, J.G.

(1980), Blood 55:408-417.

20. Shulman, M., Wilde, C.D. and Kohler, G., (1978),
Nature (London) 276:269-270.

21. Asano, S. and Riglar (1981), Cancer Res.

41:1199-1204.

22. Dignam, J.D., Lebowitz, R.M. and Roeder, R.G.

(1983), Nucl.Acids Res. 11:1475-1489.

23. Jones, K.A., Yamamoto, K.R. and Tjian, R. (1985),
Cell 42, 559-572.

24. Sen, R. and Baltimore, D. (1986), Cell 46, 705-716.

25. Landolfi, N., Capra, J.D. and Tucker, P.W. (1986),
Nature (London) 323:548-551.

26. Staudt, L.M., Singh, H., Sen, R., Wirth, T., Sharp,
P.A. and Baltimore, D. (1986) Nature (London)
323:640-643.

27. Royer, H.D. and Reinherz, E.L. (1987),
Proc.Natl.Acad.Sci.USA 84:323-236.

28. Maniatis, T., Goodbourn, S. and Fisher, J.A. (1987)
Science 236:1237-1245.

29. Goodbourn, S., Burstein, H. and Maniatis, T. (1986),
Cell 45:601-610.

30. Zinn, K. and Maniatis, T. (1986), Cell 45:611-618.

31. Campbell, H.D., Ymer, S., Fung, M.C. and Young, I.G.

(1985), Eur.J.Biochem. 150:297-304.

32. Holbrook, N.J., Smith, K.A., Fornace, A.J., Comean,
C.M., Wiskocil, R.L. and Crabtree, G.C. (1984),
Proc.Natl.Acad.Sci.USA 81:1634-1638.

33. Fuse, A., Fujita, T., Yasumitsu, H., Kashima, N.,
Hasegawa, R. and Taniguchi, T. (1984),
Nucl.Acids.Res. 12:9323-9331.

34. Nagata, S., Tsuchiya, M., Asano, S., Yamamoto, O.,
Hirata, Y., Kubota, N., Oheda, M., Nomura, H. and
Yamazaki, T. (1986), EMBOJ. 5:575-581.

35. Kelso, A. and Metcalf, D. (1985), Exo.Hematol.

13:7-15.

36. Gorman, C.M., Moffat, L.F. and Howard, B.H. (1982),
Mol.Cell.Biol. 2:1044-1051.

37. Luckow, B. and Schutz, G. (1987), Nucl.Acids Res.

15:5490.

38. Wu, C., Wilson, S., Walker, B., David, I., Paisley,
T., Zimarino, U. and Ueda, H. (1987), Science
234:1247-1253.

39. Young, D.C., Wagner, K. and Griffin, J.D. (1987).

J.C1in.Invest. A7:100-106.

40. Seelentag, K., Mermod, J-J., Montesano, R. and
Vassalli, P. (1987), EMBO J ±:2261-2265.

Claims

1. A method for controlling the expression of granulocyte/macrophage colony stimulating factor (GM-CSF) in cells which express GM-CSF, or for controlling the expression of other haemopoietic cytokines containing CK-1 or CK-2 in cells where they are expressed, which cqmprises the step of regulation of the binding of nuclear.

protein(s) in said cells with the promoter region of the GM-CSF gene.

2. A method according to claim 1, wherein said regulation step comprises induction or promotion of said binding of nuclear protein(s) with said promoter region.

3. A method according to claim 1, wherein said regulation step comprises prevention or inhibition of said binding of nuclear protein(s) with said promoter region.

4. A method according to claim 1, wherein said regulation step comprises regulation of binding of said nuclear protein(s) to the region spanning the cytokine-l (CK-1) and/or cytokine-2 (CK-2) specific sequences of the promoter region of the GM-CSF gene.

5. A method according to claim 4, wherein said regulation step comprises induction of the formulation of the NF-GMa and/or NF-GMb complexes by interaction of nuclear protein(s) with said region spanning the CK-1 and/or CK-2 specific sequences.

6. A method of diagnosis of diseases associated with abnormalities in the expression of GN-CSF, which comprises the step of detection of abberations or abnormalities in the identities of nuclear protein(s) in GM-CSF expressing cells, in the promoter region of the GM-CSF gene in said cells, or in the nature of binding of said nuclear protein(s) with said promoter region.

7. A method of diagnosis of diseases associated with the expression of GM-CSF which comprises the step of detection of the NF-GMa and/or NF-GMb nuclear protein(s) in cell samples taken from a patient.

8. A method for the determination of an agent which is effective in controlling the binding of nuclear protein(s) with the promoter region of GM-CSF gene in cells which express GM-CSF, which comprises the step of screening the effect of a candidate agent on a nuclear extract in a gel retardation assay.