AU570762B2

AU570762B2 - Cloning of cdna for il-3

Info

Publication number: AU570762B2
Application number: AU37854/85A
Authority: AU
Inventors: Hugh Douglas Campbell; Susan Joy Clark; Donna Ruth Cohen; Andrew John Hapel; Rosemary Anna Johnson; Sanie Ymer; Ian Gordon Young
Original assignee: Australian National University; Biotech Australia Pty Ltd
Current assignee: Australian National University; Inhibin Pty Ltd
Priority date: 1983-12-23
Filing date: 1984-12-24
Publication date: 1988-03-24
Anticipated expiration: 2004-12-24
Also published as: AU3785485A

Description

Cloning of cDNA and expression of Murine-Interleukin-3.

The present invention relates to the molecular cloning of cDNA for murine inter leukin-3 , the use thereof , inter alia , as a hybridisation probe for eukaryotic interleukin-3 , and for expression of biologically active murine interleukin-3 .

DISCUSSION OF PRIOR ART

Interleukin-3 (IL-3) is one of a number of colony stimulating factors which are known to regulate haematopoiesis. The colony stimulating factors are hormone-like glycoproteins which are biologically active at very low levels. As they are produced in very minute quantities in vivo, they have proven to be very difficult to characterize by conventional biochemical methods. Also, due to their high level of activity, purification procedures have been hindered and although murine IL-3 derived from myelomonocytic leukaemia cell line WEHI-3 has been recently purified to apparent homogeneity (Ihle, J.N. et al; J.IMMUN. 131,282-287 (1983)), it was not necessarily clear at the time that the major protein in this preparation, from which the N-terminal sequence was derived was, in fact, murine IL-3.

IL-3 is produced by mitogen or antigen-activated T lymphocytes and by a number of continuous cell lines. It may also be produced by bone marrow cells. IL-3 is involved in regulating the growth and differentiation of pluripotent stem cells leading to the production of all major blood cell types. It has a broad range of biological activity and appears to be identical with a number of other factors named on the basis of their biological activities. These include multi-colony stimulating factor (CSF), haematopoietic growth factor, burst promoting activity, P-cell stimulating factor, mast cell growth factor, histamine-producing cell stimulating factor and Thy 1-inducing activity.

It would be desirable to characterise this protein, as well as the DNA coding for the protein, so that large quantities of the protein could be produced for further study, as it is considered that it could play an important role in bone marrow transplantation, inter alia for the treatment of leukaemia. allegedly been purified to apparent homogeneity. However, as indicated above, the very high activity of the protein could lead to possible uncertainty about the N-terminal sequence reported and also synthetic DNA probes based on this sequence would have a high level of degeneracy which would preclude or substantially inhibit the molecular cloning of the DNA for the protein via the N terminal sequence data.

DESCRIPTION OF THE INVENTION

The present invention relates to the molecular cloning of cDNA for murine IL-3 and its use, inter alia, as a hybridisation probe and for expression of biologically active interleukin-3. The cDNA for murine IL-3 was obtained by a methodology as will be hereinafter described.

In one form the present invention provides a method for the molecular cloning of murine interleukin-3 cDNA, said method comprising:

(i) providing a cell line capable of producing murine interleukin-3;

(ii) isolating mRNA therefrom;

(iii) fractionating said mRNA from step (ii) and identifying the fraction having interleukin-3 translation activity;

(iv) preparing a cDNA library from the mRNA of step

(iii); (v) cloning said cDNA library and separating said cloned DNA; (vi) hybridizing said cloned DNA with mRNA isolated from a cell line capable of producing murine interleukin-3; (vii) eluting bound mRNA and identifying the bound mRNA which has interleukin-3 translation activity to thereby identify the clone possessing cDNA for murine interleukin-3; (viii) purifying cDNA from the clone identified in step (vii); The term interleukin-3 translation activity refers to mRNA having the ability to be translated and yield biologically active IL-3.

In another aspect the present invention provides cDNA for murine interleukin-3 in substantially pure rorm, or a DNA sequence complementary thereto.

In another form the present invention provides a polynucleotide having a base sequence corresponding to cDNA for murine interleukin-3, or a part thereof, or a functional equivalent thereof, in substantially pure form. The invention further provides a double stranded DNA sequence, one strand of which is a polynucleotide as described above.

In another form the present invention provides RNA corresponding to the cDNA as described above.

In another form of the present invention there is provided a hybridization probe useful for identifying interleukin-3 in eukaryotic cells, tissue sections or in DNA or in RNA preparations derived from such cells, including particularly human cells, said probe comprising a polynucleotide having at least about ten bases, said polynucleotide having a sequence corresponding to cDNA, DNA complementary thereto, double stranded DNA or RNA of murine interleukln-3, and a label attached to said polynucleotide.

Preferably said probe comprises a polynucleotide having the sequence of c-DNA or part thereof for murine interleukin-3 and the label attached thereto. Preferably said label is a radioactive label or biotin.

Preferably said polynucleotide for said probe comprises at least fifteen bases.

In a further aspect of the present invention there is provided a method for identifying interleukin-3 DNA or RNA in eukaryotic cells or tissue sections or extracts thereof characterised in that the probe of the present invention is used.

A further aspect of the present invention is that the cDNA for murine IL-3 can be cloned into a cloning vector and biologically active IL-3 can be expressed by prokaryotic cells as either a fusion protein or as a non-fusion protein corresponding to essentially mature IL-3. Hence, large quantities of biologically active murine IL-3 can be produced for biological research or as a growth factor reagent to add to haematopoietic cell lines. Thus in a further aspect, the present invention provides a method for the expression of murine interleukin-3, particularly in prokaryotes, as well as recombinant clones useful for such expression, and murine IL-3 produced thereby.

The method of expression comprises:

(i) providing a DNA substantially corresponding to c-DNA of murine-interleukin-3 in substantially pure form;

(ii) inserting said DNA into an appropriate cloning vector such that either the coding region for murine IL-3 is in phase with a prokaryotic protein, (e.g. B-galactosidase) or that the coding region for murine IL-3 is under the control of a prokaryotic promoter (e.g. the tac promoter) within the vector;

(iii) transforming prokaryotic cells with said plasmid constructions described in (ii) and cultivating said cells to produce murine IL-3; and

(iv) separating murine interleukin-3 from the cells of step (iii).

The present invention provides recombinant plasmids useful to express biologically active murine IL-3 in prokaryotes, said plasmids referred to as pBTA 611; pBTA 610 and pBTA 640.

The present invention also provides murine IL-3 in substantially pure form as a fusion protein or as a non-fusion protein. These proteins have N-terminal amino acids as depicted in Figure 8 of the accompanying drawings. Although these proteins are not glycosylated they are still biologically active. DESCRIPTION OF PREFERRED EMBODIBENTS

The invention will be further described by reference to the following drawings wherein:

Figure 1 depicts the interleukin-3 transiation activity of mRNA from WEH1-3 cells, fractionated on a ucrqse density gradient;

Figure 2 shows the titration curves for IL-3 activity in oocyte translates;

Figure 3 depicts the cDNA insert for pILM3 showing relevant restriction sites as well as the sequencing strategy;

Figure 4 depicts the nucleotide sequence for cDNA insert of pILM-3 and amino acid sequence of IL-3;

Figure 5 is a restriction map of vector pBTA 611; Figure 6 is a restriction map of vector pBTA biu;

Figure 7 is a restriction map of vector pBTA 640.

Figure 8 depicts the N-terminal sequence of the proteins expressed by using vectors of Figs 5-7;

Figure 9 depicts schematically the murine IL-3 gene and

Figure 10 depicts the nucleotide sequence of the murine IL-3 gene and flanking regions.

The invention will be further described also, with reference to the following methodology.

Translation of IL-3 mRNA in oocytes

Two different murine cell lines were compared as sources of IL-3 mRNA. The myelomonocytic line WEHI-3 is known to produce IL-3 constitutively (Lee, J.C., Hapel, A.J. & Ihle, J.N., J. Immun. 128, 2393-2398 (1982)) whereas the lymphoma line EL-4 produces IL-3 after stimulation with phorbol myristate acetate (PMA) (Ihle, J.N., Rebar, L., Keller, J., Lee, J.C. & Hapel, A.J.; Immun. Rev 63, 5-32 (1982)). Messenger RNA was prepared from both cell lines using the guanidine thiocyanate method (McCandliss, R., Sloma. A & Pestka, S.; Meth. Enzym. 79. 51-59 (1981); Chirgwin, J.M., Przybyla, A.E., MacDonald, R.J. & Rutter, W.J. Biochemistry 18. 5294-5299 (1979)) and attempts were made to translate the mRNAs by microinjection into X.laevis oocytes (Sloma, A., McCandliss, R. & Pestka, S. Meth. Enzym. 79. 68-71 (1981)). The translates were assayed for biologically active IL-3 using the IL-3 dependent cell line 32D cl-23 (Ihle. J.N. et al. J. Immun, 129, 1377-1383 (1982)). It was found that mRNA from ΕL-4 could be translated in the oocytes to give readily detectable levels of IL-3 after 48 h incubation as shown in Table 1. Host of the detectable activity was present ia the incubatlon aedium indicating that the oocytes were secreting mature IL-3. Assays of oocyte extracts showed that they were inhibitory. mRNA from WEHI-3 cells regularly gave higher levels of IL-3 in the translation assays than preparations from EL-4 (Table 1), in keeping with the higher levels of IL-3 found in culture supernatants of the WEHI-3 cell line (data not shown) .

Assay for translation in oocytes

EL-4 cells were grown to a cell density of 10 per ml in spinner cultures in DMEM medium supplemented with 10% fetal calf serum. The cells were collected by centrifugation, resuspended in serum-free medium containing

100 ng ml^-1 PMA and subcultured into 150cm² culture flasks at 5 x 10⁸ cells per flask. The cells were incubated for 10 h before collection. WEHI-3 cells were grown to a density of 10⁶ per ml in RPMI-1640 medium supplemented with 10% fetal calf serum and collected by centrifugation. Afterwards, both cell pellets were resuspended in 0.1 volume of serum-free medium and mRNA was prepared hy the guanidine thiocyanate-CsCl method (McCandliss, R., Sloaa, A. & Pestka, S. Meth. Enzym. 79, 51-59 (1981); Chirgwin, J . M. , Przybyla, A.E . , MacDonald, R.J, & Rutter, W.J . Biochemistry 18, 5294-5299 (1979)). The mRNAs were further purified by chromatography on oligo(dT)-cellulose (Gray. P.W. et al Nature 295. 503-508 (1982)) and after ethanol precipitation, were resuspended in buffer (10mM Tris-HCl, 0.1mM EDTA, pH 7.5) at a concentration of lugul^-1. The mRNAs were injected into 15 X. laevis oocytes ( 50nl per oocyte) according to standard methods (Sloma, A., McCandliss, R. & Pestka, S. Meth. Enzym. 79, 68-71 (1981)). After incubation, the medium was withdrawn and an extract of the oocytes prepared (Sloma, A., McCandliss. R. & Pestka, S. Meth. Enzym. 79. 68-71 (1981)).

The incubation medium, the oocyte extract and a mixture of the two were assayed for IL-3 activity using the IL-3 dependent cell line 32D cl-23 (Ihle, J.N. et al J. Immun.

129, 1377-1383 (1982)) according to the methods described previously, measuring incorporation of ³H-thymidine as an index of cell growth.

Hybrid plasmids containing IL-3 cDNA IL-3 mRNA from WEHI-3 was fractionated on a sucrose density gradient and the fractions assayed by oocyte translation (Fig. 1). The major peak of activity sedimented at 12S. This procedure resulted in 10-20-fold purification of the IL-3 mRNA. RNA from the peak fractions was used as a template to synthesize double-stsanded cDNA using the method of Land et al (Land, H., Grez, M., Hauser, H., Lindenmaier, W. & Schutz. G. Nucleic Acids Res. 9, 2251-2266 (1981)). The resultant cDNA was inserted into pAT153 by G-C tailing (Michelson. A.M. & Orkin, S.H. J. biol. Chem. 257, 14773-14782 (1982)) and the cDNA library transformed into strain HB101. Approximately 5,000 individual clones were isolated. Of a small sample tested, 92% were ampicillln sensitive indicating the presence of cDNA inserts in the majority of the clones. Since the cDNA was not size fractionated prior to cloning, the sizes of inserts in the cDNA library varied from about 200 to 1,000 base pairs (bp). It was expected that even the short inserts would give satisfactory results in the mRNA hybridization-translation assay (see below).

Clones were grown in pools of 10 and DNA prepared by the method of Holmes and Quigley (Holmes, D.S. & Quigley, M. Analyt. Biochem. 114, 193-197 (1981)). The plasmids were converted to a linear form with BamHI restriction endonuclease, contaminating RNA was destroyed by alkaline hydrolysis and the DNA was bound to nitrocellulose filters. The filter-bound DNA was hybridized with WEHI-3 mRNA and bound mRNA eluted and translated in oocytes. One positive pool of 10 clones was detected. The activity observed in translates derived from the positive pool was quite low but was consistently above the background and the values of the other clone pools. DNA was therefore prepared from the individual clones of the pool and the mRNA hybridization translation assay repeated. One of the clones (pILMl) was clearly positive (Table 2). The IL-3 activity detected showed the same titration curve as oocyte translates of WEHI-3 mRNA (Fig. 2).

Nucleotlde sequence analysis The cDNA insert in pILMl was quite short (139 bp excluding the G-C tails). It was subclotied into the Smal site of M13mp8 after shortening the G-C tails with Bal 31 exonuclease. The DNA sequence was determined on both strands using the chain termination method (Sanger, F., Nicklen, S. & Coulson, A.R. Proc. natn. Acad Sci. U.S.A. 74, 5463-5467 (1977); Sanger, F., Coulson, A.R., Hong, G.F., Hill, D.F. & Petersen, G.B. J. molec. Biol. 162, 729-773 (1982)). An oligonucleotide primer corresponding to nucleotides 92-112 of this cDNA insert was synthesized (Tanaka, T. & Letsinger, R.L. Nucleic Acids Res. 10, 3249-3260 (1982); McBride. L.J. & Caruthers. M.H. Tetrahedron Lett. 24, 245-248 (1983)) and labelled using

[ -³²P]dATP and polynucleotide kinase (Goodman, H.M. & MacDonald. R.J. Meth. Enzym. 68, 75-90 (1979)). The labelled probe was used to examine the remaining clones in the cDNA library (Grunstein, M. & Hogness, D.S. Proc. natn. Acad. Sci. U.S.A. 72, 3961-3965 (1975); Wallace, R.B. et al Nucleic Acids Res. 6, 3543-3557 (1979)). No other clones containing this sequence were detected.

To isolate a long clone encoding IL-3, a further cDNA library was prepared, but in this case the double-stranded cDNA was size fractionated using Sepharose CL-4B to enrich for cDNA fragments larger than 500bp. The resulting cDNA was inserted into pAT153 by G-C tailing and transformed into HB101. The labelled oligonucleotide probe described above was used to identify a longer cDNA clone (pILM3) which was shown by DNA sequencing to carry an insert of 629bp excluding the G-C tails. The DNA sequence of clone pILM3 was determined on both strands using the chain termination method (Sanger, F., Nicklen, S. & Coulson, A.R. Proc. Natn. Acad. Sci. U.S.A. 74, 5463-5467 (1977); Sanger. F., Coulson, A.R., Hong, G.F., Hill. D.F. & Petersen. G.B. J. Molec. Biol. 162, 729-773 (1982)) after subcloning into M13mp8 and 9. The sequencing strategy is in Fig. 3. The subcloning was complicated by the fact that one of the PstI sites flanking the cDNA insert had been lost due to exonucleolytlc removal of one nucleotide during the G tailing of the vector. Clone pILM3 was found to extend beyond pILMl at the 3' end only, both clones possessing the same 5' end. The DNA sequence is shown ln Figure 4 and shows a single large open reading frame. Assuming that the first ATG from the 5' end serves as the initiation codon, the reading frame extends from nucleotides 28-525 inclusive and specifies a polypeptide of 166 amino acids. The 5' untranslated region is therefore quite short (27bp). The method of Land et al (Land. H., Grez, M. , Hauser, H., Lindenmaier, W. & Schutz, G. Nucleic Acids Res. 9, 2251-2266 (1981)), which was used to generate the cDNA clones, is designed to produce a high proportion of cDNA clones with complete 5' terminii. Since both cDNA clones isolated in the course of this work had the same 5' end, it is possible that pILM3 carries a close to full length copy of the 5' untranslated region of IL-3 mRNA. Clone pILM3, however, does not extend to the 3' end of the IL-3 mRNA and is probably missing a significant portion of the 3' untranslated region. The presence of the complete coding sequence for IL-3 is the important feature of the present invention.

The reading frame specifies a protein of MW 18,540 as depicted by the single letter amino acid notation in Figure 4. The strongly hydrophobic stretch of 9 amino acids (residues 13-21) is indicative of the presence of a signal peptide. In the cases studied to date, the last residue of the signal sequence is invariably an amino acid with a small uncharged side chain. Analysis of the signal peptide region suggests a number of potential sites for cleavage by the signal peptidase. Application of the empirical rules proposed by Von Heijne (Von Heijne, G Eur. J. Biochem. 133, 17-21 (1983)) predicts the end of the signal peptide at residue 27 (Ala). This potential cleavage site is shown with an arrow in Fig. 4.

The published N-terminal amino acid sequence of murine IL-3 (Ihle. J.N. Contemp. Topics molec. Immun. (in the press)) is in complete agreement with the reading frame and extends from residues 33 (Asp) to 47 (Val). It seems most unlikely that the signal peptide would end at residue 32 (Arg) and this suggests that the murine IL-3 isolated from the WEHI-3 cell line has been further processed.

The protein sequence of mature processed IL-3 indicates a MW of 15,102 with four potential N-glycosylation sites (Asn-X-Ser or Asn-X-Thr). The four sites are underlined in Fig. 4. One of these is in the region of the published N-terminal sequence and glycosylation at this site would explain the inability to identify the Asn residue during automated Eόjman degradation. ϊfhen the aaino acid composition figures for aarine IL-3 axe normalized to the size predicted by the DMA sequence they give good agreement with the predicted composition (See Table 3 below), These methods enable production of large quantities of IL-3 by prokaryotic cells. The IL-3 produced is not glycosylated but still demionstrates high biological activity. TA

The amino acid sequence of murine IL-3 was compared with the sequences of human interleukln-2 and human immune interferon {Gray, P.W. et al. Nature 295, 503-508 (1982)) hoth of which axe also produced hy antiges-activated T lymphocytes. No significant sequence hoaology was detected. Although hnaan IL-2 is of very similar size to IL-3 it carries no potential sites for N-glycosylation and does not appear closely related.

Use of murine IL-3 cDNA as hybudisatlon probe

General Methods High mol. wt. DNA was prepared from mouse liver or cultured cells by gentle homogenization and treatment with Sarkosyl and proteinase K followed by equilibrium centrifugation in CsCl-ethidium bromide gradients.

DNA fragments were isolated by using electrophoresis on low-melting-point agarose gels (Wieslander, L.1979 Anal. Biochem, 98. 305-309) or by using DEAE-membrane (Schleicher & Schull) in conjunction with normal agarose geis. Bacteriophage lambda DNA minipreparations from plate lysates were carried out as described (Maniatis et al.. 1982 Molecular Cloning: A Laboratory Manual, Cold Spring Harbour Laboratory, NY, USA) except that the plates used were LM (Hanahan, D. (1983) J. Mol. Bio. 166, 557-580) containing agarose instead of agar. The phage DNA pellet was redissolved in 100 ul TE containing 50 ug/ml RNase, incubated at 37° for 20 min. extracted with phenol phenol/CHCl₃(1:1), then CHCl₃, and recovered by ethanol precipitation. The DNA was then redissolved in 50 ul of 10mm Tris/HCl. 0.1mM EDTA, pH 8.0. Plasmid DNA minipreparations were made by the rapid boiling method (Maniatis et al., 1982 supra) except that after RNase treatment, the DNA was further purified as described above for phage DNA minipreparations. Large scale plasmid DNA preparations were made by the method of Clewell & Helinski (1969, Procl. Nat. Acad. Sci. USA, 62. 1159-1166) following amplification with chloramphenicol. Isolated EcoRI fragments from recombinant bacteriophage were ligated to pAT153 which had been cleaved with EcoRI and dephosphorylated with calf intestinal alkaline phosphatase (Maniatis, 1982 supra) Transformation of E. coll strain C600 (r_k-, m_k ⁺) with ligation mixes was accomplished by the method of Hanahan (1983). The synthetic random 10-mer used as a pximer for prohe synthesis was prepared by the phosphoramidite method (Tenaka & Letsinger (1982) Nucleic Acids Res. 10 , 3249-3260) using an equimolar mixture of all four bases at each position.

Nucleic acid hybridization Agarose gel electrophoresis and transfer of DNA to nitrocellulose were performed by standard methods (Maniatis et al. , 1982 supra). Nitrocellulose replicas of plates containing bacteriophage lambda plaques were prepared for hybridization as described by Benton & Davis (1977) (Science (Wash.), 196, 180-182). Binding of DNA to nitrocellulose for filter hybridization assay was accomplished essentially as described by Kafatos et al. (1979) Nucleic Acids Res. 1, 1541-1552). Conditions for the various hybridizations were as follows. Southern hybridization: prenybridization and hybridization (18h) at 65° in 6xSSc. 5x Denhardt's solution (Denhardt. 1966), 100 ug/ml sonicated, denatured salmon sperm DNA, 0.5% (w/v) SDS and 10mM EDTA; washing, 10 min at room temperature and 60 min at 65° in 2xSSC.

Filter hybridization assays of DNA. samples: prehybridization (4-6h) and hybridization (18h) as for Southern hybridization except that 5xSSC and 50mM sodium phosphate buffer, pH 6.8 replaced the 6xSSC, and the lmM ATP and lmM sodium pyrophosphate were included; washing, 20 min at room temperature in 2 changes of 2xSSC, 0.5% SDS and then 120 min at 65° in 0.1xSSC. 0.5% SDS followed by a further 30 min. under the same conditions in fresh buffer.

Plaque hybridization: prehybridization (5h) and hybridization (16h) at 42° in 50% (v/v) formamide, 5xSSPE, 5x Denhardt's solution, 100 ug/ml sonicated, denatured salmon sperm DNA, 0.1% (w/v) SDS, ImMATP and lmM sodium pyrophosphate; washing, 20-30 min at room temperature in 4 changes of 2xSSC, 0.1% SDS then 3h at 67° in 2 changes of 1xSSC, 0.1% SDS.

Hybridization in situ in dried agarose gels was performed by a modification of published methods. Briefly, agarose gels following electrophoresis of DNA fragments were soaked in lmM MgSO. 0.5 ug/ml ethidium bromide and then dried between sheets of cellophane on a commercial gel dryer at 60º (2h) . The gel/cellophane sandwich was wetted with water, .and the separated gel mas incubated for 20 min in 0.5N NaOH, 1 m HaCl then neutralized for 20 min in 0.5 M Tris/HCl pH 7.4, 1 M NaCl. The gel was then added directly to reserved hybridization liquid from the previous plaque screening and hybridization was allowed to proceed at 42º for 16h. Washing was for 10 min at room temperature in 2 changes of 2xSSC, 0.5% SDS, for 60 min at 60° in 4 changes of 2xSSV, 0.1% SDS. and then for 60 min at 60° in 4 changes of 0.1xSSC. 0.1% SDS.

All hybridization and washing of filters or dried gels at elevated temperatures was performed with continual agitation. Autoradiography was performed using Kodak XAR-5 or Fuji RX film at -70° with intensifying screens. Probe Synthesis Probes were prepared by primed synthesis using the 467 bp HindiII-NcoI fragment of the IL-3 cDNA-containing plasmid pILM-3 as template and a synthetic random 10-mer as primer. A typical small scale reaction contained, in 20 ul, 100ng of template DNA and 1 ug of primer which had been heated together at 100° for 3 min and snap chilled to 0°, 50mM sodium phosphate buffer pH 6.8, 10mm MgCl₂ 10mM DTT, 500 uM DGTP, 500 uM DCTP, 500 uM TTP, 3 uM (variable) [³²P]-dATP (50 uCi). and 2.5 units of E. coll DNA polymerase I large fragment. The reaction was allowed to proceed for 2h at 47°. Incorporation into acid precipitable material (Maniatis et al. Molecular Cloning: A Laboratory Manual: Cold Spring Harbour Laboratory, NY, USA. 1982) was

70%. For some Southern blots, the probe DNA was separated from low-molecular-weight material by chromatography on Sephadex G-50 fine (Comb et al. (1983) DNA. 2 , 213-229). Generally, the probe was used without separation by including ATP (1mM) and sodium pyrophosphate (1 mM) in the prehybridization and hybridization fluid (Ullrich et al. (1984) EMBO J . , 3, 361-364). Probe DNA was denatured before use by heating for 5 min at 100° and snap chilling. For library screening, the probe synthesis was scaled up and the probe was used at a concentration of 1x10⁶ cpm/ml

(specific activity: 4x10⁸ cpm/ug cDNA). For Southern hybridization amd filter hybridization of agarose gel fractions, probes were used at a concentration of 4x10⁶ cpm/ml (specific activity 6x10⁸ cpan/ug cDNA) .

Southern hybridization of mnrine genomic DNA To assess the number of copies of the IL-3 gene present In the murine genome, DNA from different mouse strains was digested with EcoRI and subjected to agarose gel electrophoresis, transfer to nitrocellulose membrane and hybridization to a [ ³²P]-labelled probe derived from the

467 bp Hlndlll-Ncol fragment covering most of the coding portion of the murine IL-3 cDNA as described above. A single hybridizing fragment of appropriate intensity for a single copy gene was observed in all strains tested. The size of the genomic EcoRI fragment varied from 9kb for

Balb/c and some other strains to 16kb for New Zealand Black. DNA from Balb/c and New Zealand Black gave identical results with BamHI , bands of 2.7kb and 10kb being observed. With Hindlll. a single band of 6.4kb was observed in both cases. These results are consistent with the presence of a single copy of the IL-3 gene in the haploid mouse genome. The presence of two bands in the case of BamHI is due to the presence of a BamHI site known to lie between Hindlll and Ncol sites in the cDNA which delimit the probe fragment used. A single band is obtained with Hindlll. in spite of the presence of a Hindlll site in the cDNA since this Hindlll site was used to excise the HindIII-Ncol fragment used to synthesize the probe. Thus, only DNA to one side of this Hindlll site will hybridize with the probe.

Genomic DNA from several cultured murine cell lines was examined by Southern hybridization. WEHI-3 (D-), WEHI-3 (D⁺) and A4A showed a band at 9kb as for Balb/c and an additional band, of similar intensity, at 4kb. B6, the original cell line from which the WEHI lines were derived, had a single band 9kb.

Cloning of the IL-3 gene from Balb/c mice Genomic DNA from Balb/c liver was digested to completion with EcoRI, size fractionated on gels and a fraction enriched for the IL-3 gene cloned into gtWES. B. The lambda library mas screened by plague hybridization (Benton W.P. & Davis R.-W. (1177) Science (Wash) 196, 180-182) using the IL-3 cDNA-derived probe.

Eight positive plaques were obtained from a total of 70,00 plaques. ; Six of these were iplaque purified and phage DNA mlnipxeparations nexe prepared and suhjecteαs to restriction endonuclease analysis All six lambda clones contained an 8.6 kb EcoRI insert and gave identical patterns asfter double digestion with EcoRI/BamHI and EcoRI/HlndlII. EcoRI excises the inseerts from the phage arms, so that the orientation of the inserts would not affect the pattern in these double digests.

To facilitate further work, the 8.6 kb insert from one of the recombinant phages was recloned into the EcoRI site of pAT153 (Twigg & Sherratt. (1980) Nature. 283. 216-218).

This plasmid was designated pILMll and was further analysed by restriction endonuclease digestion and agarose gel electrophoresis of plasmid DNA minipreparations. In situ hybridization of restriction fragments to the CDNA probe was used to localize the IL-3 gene to a 3.7 kb HlncII-EcoRI fragment (Figure 9). This fragment was isolated by low-melting point agarose gel electrophoresis of a HincII-EcoRI digest of pILMll plasmid DNA.

Nucleotide sequence of the murine IL-3 gene

The nucleotide sequence of a 3490 bp region of the 3.7 kb HincII-EcoRI fragment containing the gene was determined on both strands by the chain-termination method (Sanger et al ., Proc. Natl. Acad. Sci. U.S.A.. 74, 5463-5467 (1977)). Figure 9 shows a summary of some of the gel readings used to assemble the sequence. The sequence of this region is presented in Figure 10.

Organization of the IL-3 gene

The IL-3 gene is made up of five exons interrupted by four introns. This arrangement is shown schematically in Figure 9 and is indicated in Figure 10.

The sequence of the five exons is in complete agreement with our previously determined cDNA sequence. In particular, nucleotide 2302 is A (Figure 10) as is the corresponding nucleotide 463 in the cDNA. Yokota et al. (1984 Procl. Natl. Acad. Sci. U.S.A.. 81, 1070-1074) have reported that this nucleotide is G in their cDNA clone which would change the encoded amino acid from Thr to Ala. Whether this change represents allelic variation or some artefact arising during cDNA. synthesis is not known. The predicted amino acid seguence of IL-3 from the WEHI-3 cell line, the source of mRMA for our cDNA and from normal Balb/c mice, the source of genomic DNA in the present work is identical.

Transcriptional control signals

Transcription of the IL-3 gene almost certainly begins at nucleotide 494 indicated in Figure 10. Several cDNA clones obtained by two different groups by methods likely to preserve the 5' terminus of the mRNA intact terminate at this position. A potential TATA box, is located upstream from the start of transcription (Figure 10). The distance between the TATA sequence and the start of transcription is 27 nucleotides, similar to the mean distance observed eukaryotic promoters.

Expression of biologically active murine lnterleukin-3 Interleukin-3 was expressed in E. coll as a functionally active protein by two approaches. Firstly as fusion proteins with the first few amino acids of B-galactosidase and secondly as a non fusion protein corresponding to one of the putative mature protein sequences. The first amino acid of the mature protein is still unclear since aspartate, amino acid 33 (Ihle, J.N. et al., J. Immun. 131. 282-287. 1983) as well as Alanine, 9.9.27) have been reported to be the N-termlnal amino acid. Therefore in the construction of IL-3 a number of start points were chosen.

Construction of IL-3 Fusion Expression Clones (i) PBTA611

This expression clone was designed to express a protein consisting of 10 amino acids of B-galactosidase fused to the IL-3 gene coding from Arginine, amino acid 32, to the end of the protein (see Figure 8).

Approximately 5 ug of PILM3 DNA was digested with Haelll and the DNA fragments separated on a 6% polyacrylamide gel. pBR322 DNA digested with Haelll was run as molecular weight markers. The 450bρ Haelll fragment from pILM3 was cut out of the gel and eluted in 300 ul of 500 mM Ammonium Acetate, 1 mM EDTA pH 8.0, 0.1% SDS at 37ºC for 16 hours. The supernatant was ethanol precipitated, redissolved in 300 ul of 300 mM sodium acetate and ethanol precipitated again. The 450bp Haelll fragment was ligated to Pst I linkers d(pGCTSCAGC) , then digested with Pst I and electrophoresed on a 6% polyacrylamide gel. The Haelll fragment with Pst I cohesive ends was eluted as described above, and then ligated with Pst I, digested and phosphatased pUC9 DNA (Vierira and Messing, Gene 19, 259, 1983). The ligation mix was transformed into the Ion- strain NB42 and plated on L+Ampicillin plates. DNA was prepared from a number of colonies, digested with Pst I and electrophoresed on a 6% polyacrylamide gel.

The majority of clones had 450bp Pst I inserts. The orientation of the inserts was determined by BamHI digestion. In the correct orientation, where the IL-3 gene is in phase with the B-galactosidase coding region, a 290bp BamHI fragment is expected. Clones with this restriction profile, called BTA611 (Figure 5), were picked and assayed for expression of an IL-3 protein. BTA611 colonies were Inoculated into 3 mis of L broth and incubated overnight with shaking at 37ºC. Approximately 300 ul of culture was TCA precipitated and the proteins separated on a 15% polyacrylamide SDS gel. An extra 17 K dalton protein band was observed in the BTA611 track which did not appear in the control track of pUC9 in NB42. (ii) BTA610

This expression clone was designed to express a protein consisting of 9 amino acids of B-galactosidase fused to IL-3 gene coding from Alanine, amino acid 27, to the end of the gene (see Figure 8).

Approximately 5 ug of pILM3 DNA was digested with Hindlll and Ncol and end-filled with Klenow polymerase and the DNA fragments separated on a 6% polyacrylamide gel. The 457bp HindiII/Ncol fragment was eluted from the gel and ligated into the end-filled BamHI site of pUC8 (Vieira and Messing, Gene 19, 259, 1983 ) . The ligation mix was transf ormed into a lon- strain NB42 and plated onto L+Ampicillin plates. DNA was prepared from a number of colonies and digested with Haelll to observe the insert and EcoRI/Bglll to determine orientation. Clones with the insert in the correct orientation, where the IL-3 gene is in phase with the B-galactosidase gene were called BTA610 (Figure 6). Cultures of BTA610 were assayed for the presence of a 17 K dalton protein by polyacrylamide SDS gel electrophoresis as described previously. A predominant protein band was observed in the BTA610 track at 17.7 K daltons and was not present in the control track of pUC8 in NB42.

Biological Activity of the IL-3 Expression Constructs

Cultures of BTA611. BTA610, pUC8 and pUC9 in NB42 were incubated at 37°C for 16 hours in 50 mis L broth and 50 ug/ml Ampicillin. The cell pellets were lysed by sonication and the supernatant from each culture assayed for 32D cl=23. (Ihle. J.N. et al.. J. Immun. 129., 1377-1383, 1982). Both BTA610 and BTA611 culture sonicates demonstrated high levels of biologically active IL-3, (2 x

10⁴ units/ml) and both pUC8 and pUC9 were negative for IL-3 activity.

Construction of IL-3 Non Fusion Expression Clones (i) PBTA640

This expression clone was designed to express mature IL-3 from Alanine at amino acid 27 to the end of the gene, under the control of the tac promoter (Amann al., Gene 25, 167-178. 1983). ptac12 was digested with PvuII and ligated with a 22mer linker d(pAAATGGCTTCAATCAGTGGCCA) . This linker includes the ATG start codon and amino acids 27-31 of IL-3. The last 6bp of the linker is a Ball site. The ptac12 linker ligation mix was transformed into MC1061 and plated onto L+Ampicillin plates. DNA from a number of colonies were digested with Ball and EcoRI to verify the cloning of the linker molecules and linker orientation. DNA from the correct construction with the ATG 5' to the IL-3 codons, was then digested with Ball and phosphatased and ligated with the 450bp Haelll fragment from pILM-3. The ligation mix was digested with Ball, since a Ball site is only created when the Haelll fragment is in the non-coding orientation. The ligation mix was then transformed into Ion- NB42 F (laclq) and plated onto L+Ampicilllin plates. BNA from a number of colonies was prepared and digested with EcoRI/BamHI to verify cloning of the Haelll fragment and its orientation. The majority of nlones had the Haelll fragment in the correct orientation where the IL-3 coding region was in phase with the linker coding region. These clones were called BTA640 (Figure 7). BTA640 colonies were inoculated into 3 mis of M9 broth and induced at an OD₆₀₀ of 0.6 with 0.2% lactose for 3 hours. Approximately 300 ul of culture was TCA precipitated and the proteins separated on a 15% polyacrylamide gel. An extra protein band at 16.8 K daltons was observed in the BTA640 track which did not appear in the control track of ptac12 In NB42 F .

DNA transfer vectors and expression The cDNA insert encoding IL-3 was excised from the plasmid pILM3, made blunt ended, and inserted into the SV40 shuttle vector pSV2neo³⁷ at the unique Hind III site using standard methods. Insertion of the coding sequence into pSV2neo places the expression of IL-3 under the control of the SV40 early promoter. One of the resultant recombinant plasmids (designated pILM4) was shown to have the insert in the correct orientation by restriction enzyme digestion.

The plasmid DNA was transformed into monkey COS cells.

After 48 hrs growth it was shown that the culture medium contained good levels of IL-3, assayed using IL-3 dependent cell lines.

This method enables the production of IL-3 by mammalian cells.

Claims

1. A method for the molecular cloning of murine interleukin-3 cDNA, said method comprising:

(i) providing a cell line capable of producing murine interleukin-3; (ii) isolating mRNA therefrom; (ill) fractionating said mRNA from step (ii) and identifying the fraction having interleukin-3 translation activity; (iv) preparing a cDNA library from the mRNA of step

(iii); (v) cloning said cDNA library and separating said cloned DNA; (vi) hybridizing said cloned DNA with mRNA isolated from a cell line capable of producing murine interleukin-3; (vii) eluting bound mRNA and identifying the bound mRNA which has interleukin-3 translation activity to thereby identify the clone possessing cDNA for murine interleukin-3; (viii) purifying cDNA from the clone identified in step (vii).

2. cDNA prepared from the method of claim 1.

3. A polynucleotide having a sequence substantially equivalent to cDNA for murine interleukin-3 or part thereof, in substantially pure form.

4. A polynucleotide complementary to the polynucleotide of claim 3.

5. A polynucleotide comprising RNA corresponding to the polynucleotide of claim 3 or claim 4.

6. Murine interleukin-3 in substantially pure form and having a amino acid sequence as depicted in Figure 4 of the accompanying drawings.

7. A hybridization probe useful for identifying interleukin-3 DNA or RNA in eukaryotic cells, said probe comprising a polynucleotide having at least about ten bases, said polynucleotide having a sequence corresponding to c-DNA of murine interleukin-3, or the DNA complementary thereto, or the double stranded DNA thereof, or RNA having a sequence corresponding to the mRNA sequence of murine interleukin-3, and a label attached to said polynucleotide.

8. The probe of claim 7 wherein said label is selected from a radioactive label or blotin.

9. The probe of claim 7 or 8 wherein said polynucleotide comprises at least fifteen bases.

10. The probe of any one of claims 7 to 9 wherein said polynucleotide has a sequence corresponding to a cDNA of murine interleukin-3 or part thereof.

11. A method for identifying interleukin-3 DNA or RNA in eukaryotic cells or tissue sections or extracts thereof characterized in that the probe of any one of claims 7 to 10 is used.

12. cDNA for murine interleukin-3 substantially as hereinbefore described and as depicted in Figure 4.

13. A method for producing biologicallly active Il-3 said method comprising:

(i) providing a DNA substantially corresponding to c-DNA or part thereof of murine 11-33 in substantially pure form;

(ii) inserting said DNA into an appropriate cloning vector such that either the coding region for murine Il-3 is in phase with a prokaryotic protein or that the coding region for murine Il-3 is under the control of a prokaryotic promoter within the vector, to result in a recombinant cloning vector ;

(iii) transferring prokaryotic cells with said recombinant vector of (ii) and culturing said cells to pxoduc e murine Il-3.

14. The method of claim 13 wherein said murine interleukin-3 is expressed as a fusion protein.

15. The method of claim 14 wherein said fusion protein comprises mature murine interleukin-3 and an oligopeptide comprising at least the first 9 amino acids of

-galactosidase.

16. The method of claim 13 wherein said murine Il-3 is expressed as a non-fusion protein.

17. The method of claim 16 wherein the said non-fusion protein begins at any of the first 32-40 amino acids of murine Il-3. pBTA611 and pBTA640.

19. Murine interleukin-3 produced by the method of any one of claims 13 to 17.