GB2210620A - Transforming growth factor beta -2 - Google Patents

Description

CLONING AND EXPRESSION OF TRANSFORMING GROWTH FACTOR 32 22 1 O 6 20

1 INTRODUCTION

The present invention relates to the cloning and expression of human transforming growth factor-Beta 2.

2 BACKGROUND OF THE INVENTION

Transforming growth factor-Beta (TGF-p) is a member of a recently described family of polypeptides that regulate cellular differentiation and proliferation Other members of this family include Mullerian inhibitory substance (Cate et al, 1986, Cell 45:685-698), the inhibins (Mason et al, 1985, Nature 318:659-663) and a protein predicted from a transcript of the decapentaplegic gene complex of Drosophila (Padgett et al, 1987, Nature 325:81-84).

Transforming growth factor-Beta (TGF-0) consists of two identical disulfide linked subunits having molecular weights of 13,000 (Assoian et al, 1983, J Biol Chem.

258:7155-7160; Frolik et al, 1983, Proc Natl Acad Sci.

USA 80:3676-3680; Frolik et al, 1984, J Biol Chem.

260:10995-11000) It has been purified from several tissue sources including placenta (Frolik et al, 1983, Nature 325:81-84), blood platelets (Childs et al, 1982, Proc.

Natl Acad Sci USA 79:5312-5316; Assoian et al, 1983, J.

Biol Chem 258:7155-7160) kidney (Roberts et al, 1983, Biochemistry 22:5692-5698), and demineralized bone (Seyedin et al, 1985, Proc Natl Acad Sci USA 82:119-123) In the presence of 10 % serum and epidermal growth factor, TGF-3 promotes the anchorage independent growth of normal rat kidney fibroblasts (Roberts et al, 1981, Proc Natl Acad.

Sci USA 78:5339-5343; Roberts et al, 1982, Nature 295:417-419; Twardzik et al, 1985, J Cell Biochem.

28:289-297); in the presence of 10 % serum alone, it is able to induce colony formation of AKR-2 B fibroblasts (Tucker et al., 1983, Cancer Res 43:1518-1586) TGF-p has also been shown to cause fetal rat muscle mesenchymal cells to differentiate and produce cartilage specific macromolecules (Seyedin et al, 1986, J Biol Chem 261:5693-5695).

In contrast to its effect on cell proliferation, TGF-f purified from human platelets as well as a functionally related protein isolated from African green monkey cells (BSC-1) has been shown to inhibit the growth of certain cells in culture (Tucker et al, 1984, Science 226:705-707) TGF-Beta has also been shown to inhibit the growth of several human cancer cell lines (Roberts et al, 1985, Proc Natl Acad Sci USA 82:119-123) This inhibitory/stimulatory effect of TGF-f may depend on several factors including cell type and the physiological state of the cells (for review see Sporn et al, 1986, Science 233:532-534).

c DNA clones coding for human (Derynck et al, 1985, Nature 316:701-705), mouse (Derynck et al, 1986, J Biol.

Chem 261:4377-4379) and simian (Sharples et al, 1987, DNA 6:239-244) TGF-$ have been isolated DNA sequence analysis of these clones indicates that TGF-f is synthesized as a large precursor polypeptide, the carboxy terminus of which is cleaved to yield the mature TGF-f monomer Strong sequence homology has been found throughout the TGF-f precursor protein from all of the above sources.

Very recently a protein isolated from bovine demineralized bone has been identified as being related to TGF-f (Seyedin et al, 1987, J Biol Chem 262: 1946-1949).

The protein has also been isolated from porcine platelets (Cheifetz et al, 1987, Cell 48:409-415), a human prostatic adenocarcinoma cell line, PC-3 (Ikeda et al, 1987, Biochemistry 26:2406-2410), and a human glioblastoma cell line (Wrann et al, 1987, EMBO 6:1633-1636) Partial amino acid sequence of this protein indicated that it was homologous to TGF-O and has been termed TGF-82 The human (Derynck et al, 1985, Nature 316:701-705), mouse (Derynck et al, 1986, J Biol Chem 261:4377-4379) and simian 3 (Sharples et al, 1987, DNA 6:239-244) TGF-R described previously has been termed TGF-Al.

3 SUMMARY OF THE INVENTION

The present invention relates to the production of large quantities of TGF-62 by eukaryotic host cells transfected with recombinant DNA vectors containing a TGF-A 2 coding sequence controlled by expression regulatory elements In a specific embodiment, c DNA clones coding for human TGF-A 2 precursor were obtained from a c DNA library made from a tamoxifen treated human prostatic adenocarcinoma cell line, PC-3 The c DNA sequence of one such clone predicts that TGF-62 is synthesized as a 442 amino acid polypeptide precursor from which the mature 112 amino acid TGF-A 2 subunit is derived by proteolytic cleavage This TGF-A 2 precursor, termed TGF-A 2-442, shares a 41 % homology with the precursor of TGF-R 1 In another embodiment, c DNA clones coding for simian TGF-A 2 precursor were obtained from a c DNA library made from an African green monkey kidney cell line, BCS-40 The c DNA sequence of one such clone predicts that TGF-f 2 is also synthesized as a 414 amino acid polypeptide precursor from which the mature 112 amino acid TGF-R 2 subunit is derived by proteolytic cleavage This TGF-02 precursor, termed TGF-$ 2-414, has an amino acid sequence of 414 amino acid residues and is identical to the amino acid sequence of TGF-82-442, except that it contains a single Asparagine residue instead of the 29 amino acid sequence from residue numbers 116 to 135 of the human TGF-62-442 sequence.

Clones from the BSC-40 c DNA library which encode a simian TGF-A 2-442 precursor as well as clones from the human PC-3 c DNA library which encode a human TGF-A 2-414 precursor have also been identified The human and simian TGF-A 2-442 precursors appear to be perfectly homologous at the amino 4 - acid level, as do the human and simian TGF-fi 2-414 precursors.

The mature 112 amino acid monomers of TGF-Al and TGF-A 62 show 71 % homology.

Expression vectors containing the TGF-fi 2 mature coding sequence joined in-phase to the TGF-fl signal and precursor sequences (Co-owned/pending United States Patent Application No 189,984) were constructed and used to transfect Chinese Hamster Ovary cells (CHO cells) The resulting CHO transfectants produce and secrete mature, biologically active TGF-A 2.

3.1 DEFINITIONS The following terms as used herein whether in the t singular or plural, shall have the meanings designated.

TGF-02:

TGF- 2 precursor:

A transforming growth factor-Beta 2 of human or simian origin comprising the amino acid sequence substantially as depicted in FIG la from about amino acid residue number 331 to about amino acid residue number 442.

A family of transforming growth factor-Beta 2 molecules of human or simian origin comprising an amino acid sequence substantially as depicted in FIG la from about amino acid residue number 1 to about amino acid residue number 442, or from about amino acid residue number 1 to about amino acid residue number 442 where the amino acid sequence from amino acid residue number 116 to amino acid residue number 144 is deleted and replaced by a single Asparagine residue The term shall mean a TGF-beta 2 precursor designated TGF-B 2-442 or TGF-A 82-414, whether of human or simian origin.

- Hybrid TGF-i 1/ TGF-fi 2 precursor:

TGF-R 1 precursor:

A novel transforming growth factor-beta precursor molecule comprising the amino acid sequence substantially as depicted in FIG lb from about amino acid residue number 1 to about amino acid number 390.

Simian-transforming growth factor-Betal precursor and signal sequences substantially as depicted in FIG lb from about amino acid residue number 1 to about amino acid residue number 278.

6 - 4 BRIEF DESCRIPTION OF THE FIGURES

Fi G la Nucleotide sequence of human TGF-$ 2-442 c DNA and deduced amino acid sequence The 2597 bp insert of PC- 21 was subcloned into p EMBL (Dante et al, 1983, Nucleic Acids Res 11:1645-1654) and sequenced on both strands using the dideoxy chain-termination method (Sanger et al, 1977, Proc Natl Acad Sci USA 74:5463-5467) The coding sequence is shown and the deduced amino acid sequence is presented directly above The mature TGF-P 2 sequence is boxed and the signal peptide is overlined Potential glycosylation sites are indicated by asterisks The arrow indicates the putative signal sequence cleavage site The nucleotide sequence of simian TGF-p 2-414 c DNA is identical to the human TGF-p 2-442 c DNA sequence except that nucleotides 346 through 432 (bracketed) are deleted and replaced by the sequence AAT, and except that several silent nucleotide changes occur elsewhere in the structure (indicated by single letters directly below the changed nucleotide) The deduced amino acid sequence for simian TGF-$ 2-414 precursor is identical to the human TGF- 2-442 precursor amino acid sequence except that Asparagine replaces amino acid residues 116 through 144 in the human TGF-A 2-442 structure The nucleotide sequence of a human TGF-82-414 c DNA has been sequenced through the region indicated by broken underlining and was found to be perfectly homologous to the human TGF-A 2-442 c DNA sequence except that nucleotides 346 through 432 are deleted and replaced by the sequence AAT.

FIG lb Nucleotide sequence c_ hybrid TG'-i/i GF-k 2 precursor DNA and deduced amino acid sequence The coding sequence is shown and the deduced amino acid sequence is presented directly above The mature TGF-P 2 sequence is boxed and the precursor signal peptide is overlined.

Glycosylation sites are indicated by asterisks The arrow indicates the putative signal sequence cleavage site The 7 - TGF-P 2 mature coding sequence depicted is of human origin.

The simian TGF-fi 2 mature coding sequence is nearly identical to the human sequence: only 3 silent base changes occur and are indicated by single letters directly below the changed nucleotide Details of the c DNA cloning of TGF-A 2 and the construction of the hybrid TGF-A 1/TGF-f 12 gene are given in the text.

FIG lc Schematic diagram of hybrid TGF- 11/TGF-A 2 precursor gene.

FIG ld Restriction endonuclease maps of p PC-14 ( 2.2 kb) and p PC-21 ( 2 3 kb) The boxed regions indicate coding sequences for TGF-R 2 monomer The ATG denotes the initiating methionine codon The distance between the ATG and Kpn I site in p PC-21 ( 2 34 kb) is approximately 420 bp.

The darkened area indicates the position of the 84-bp insertion in p PC-21 ( 2 3 kb).

FIG le Partial DNA sequence analysis of p PC-14 ( 2 2 kb) A synthetic oligonucleotide 5 '-AGGAGCGACGAAGAGTACTA-3 ' which hybridized approximately 140 bp upstream from the K Dn I site within the insert in p PC-21 ( 2 3 kb) was used to prime DNA sequencing reactions In this region, the sequence of p PC-14 ( 2 2 kb) (upper line) is identical to p PC-21 ( 2 3 kb) up to nucleotides coding for Asn-116 The 84-bp insertion within the Asn-116 codon of p PC-14 ( 2 2 kb) which was found in p PC-21 ( 2 3 kb) is shown The Kpn I site within the insert is denoted.

FIG 2 Homologies of human TGF-A 1 and TGF-$ 2-442 precursor sequences a) Primary sequence homology:

identical residues are boxed Asterisks refer to potential glycosylation sites in TGF-A 2 The potential signal sequence cleavage site and the cleavage site of the mature polypeptide are indicated b) Dot matrix comparison using Gene Pro software Each dot locates a point where 5 out of amino acids are identical Diagonal lines indicate regions of homology.

8 - FIG 3 Northern blot analysis of BSC-40 and PC-3 polyadenylated RNA Polyadenylated RNA was isolated from BSC-40 and PC-3 cells, fractionated on an agarose- formaldehyde gel, transferred to Hybond-N filters and hybridized to l 32 Pl-labelled TGF-g 2 specific probe, p PC-21 (Panel A) or a mixture of l 32 Pl-labelled TGF- 2 an 4 TGF-fl (Sharples et al,1987) specific probes (Panel B) as described in Materials and Methods Lane 1, BSC-40 polyadenylated RNA ( 5 micrograms); lane 2, PC-3 polyadenylated RNA ( 5 micrograms).

FIG 4 Northern blot analysis of polyadenylated RNA from different sources Polyadenylated RNA was isolated from MCF-7 (human mammary carcinoma), SK-MEL 28 (human melanoma), KB (nasopharangeal carcinoma) and HBL-100 (human mammary epithelial) cells and analyzed by Northern blot hybridization to a TGF-f 2 specific probe (p PC-21) as described in Materials and Methods Each lane contains 5 micrograms of polyadenylated RNA from SK-MEL 28 (lane 1), MCF-7 (lane 2), HBL-100 (lane 3) or KB (lane 4) cells.

FIG 5 Bioactivity Assay of Recombinant TGF-A 2 1 f 9 12.5, clone 36 cells were grown to confluency in 100 mm tissue culture dishes Cells were washed 3 X with serum-free media and incubated for 24 hours in 5 ml of serum-free media Media was collected, dialyzed against O 2 M acetic acid, and assayed for inhibition of DNA synthesis in CCL 64 cells as described (Gentry et al, 1987, Mol Cell Biol.

7:3418) In this assay 3 3 pg of purified natural TGF-#1 standard gave 50 % inhibition; the specific activity of purified natural TGF 2 was calculated to be about half that of TGF-i 1.

FIG 6 Western Blot Analysis of Recombinant Proteins Secreted by 109, 12 5, clone 36 Acid dialyzed serum-free conditioned media from 1 f 9, 12 5, clone 36 cells was fractionated by SDS-polyacrylamide gel electrophoresis and analyzed by Western blotting with anti serum made against 9 - the synthetic peptide NH 2-YNTINPEASASPC-COOH as described (Gentry et al, 1987, Mol Cell Biol 7:3418).

DESCRIPTION OF THE INVENTION

The present invention relates to the production of a biologically active, mature form of TGF-A 2 from a TGF-A precursor gene coding sequence and its product The mature biologically active TGF-g 2 may be produced by the cloning and expression of the full-length nucleotide coding sequence of the TGF-A 2 precursor or its functional equivalent in a host cell which processes the precursor correctly so that a mature TGF-A 2 is produced having a biological activity that is virtually indistinguishable from that of authentic natural TGF-82 Functional equivalents of the full length nucleotide coding sequence of the TGF-A 82 precursor include any DNA sequence which, when expressed inside an appropriate host cell, is capable of directing the synthesis, processing and export of mature TGF-A 2 In this regard, hybrid precursor coding sequences including, for example, the TGF- Al precursor sequence joined in-frame to the TGF-,i 2 mature sequence, may be constructed and used to produce biologically active TGF-A 2.

The method of the invention may be divided into the following stages solely for the purposes of description: (a) isolation or generation of the coding sequence for a precursor form of TGF-A 2; (b) construction of an expression vector which will direct the expression of a TGF-A 2 coding sequence; (c) transfection of appropriate host cells which are capable of replicating and expressing the gene and processing the gene product to produce the mature, biologically active form of TGF-A 2; and (d) identification and purification of the mature, biologically active TGF-,52.

Once a transfectant is identified that expresses high levels of bioactive, mature TGF-A 2, the practice of the invention - involves the expansion of that clone and isolation of the gene product expressed.

The method of the invention is demonstrated herein, by way of examples in which c DN As of the TGF-A 2 precursor coding region were prepared, cloned, sequenced, and utilized to construct expression vectors which direct the high-level expression of TGF-A 2 in CHO cells In a specific embodiment, the complete amino acid sequence of the mature form of human TGF-A 2 has been determined and shows an overall homology of 71 % with TGF-Al Using synthetic oligonucleotide probes we have identified clones from a PC-3 c DNA library coding for TGF-A 2 DNA sequence analysis of one of these clones revealed that TGF-A 2, like TGF-i 1, is synthesized as a larger precursor protein, the carboxy terminus of which is cleaved to yield the mature TGF-fi 2 monomer While there is a 71 % homology between TGF-A 8 l and TGF-A 2 throughout the mature portions of these molecules, only a maximum of 31 % homology exists within the rest of the precursor, suggesting that the amino terminal regions of TGF-A 8 l and TGF-P 2 may be functionally distinct.

In a specific embodiment of the invention, expression of a novel TGF-fil/TGF-A 2 hybrid gene in CHO cells is used to produce large amounts of biologically active TGF-A 2.

The various aspects of the method of the invention are described in more detail in the subsections below and in the examples that follow.

5.1 ISOLATION OR GENERATION OF THE TGF-BETA 2 CODING REGION The nucleotide coding sequence for TGF-A 2 is depicted in FIG la In the practice of the method of the invention, the nucleotide sequence depicted therein, or fragments or functional equivalents thereof, may be used to generate the recombinant molecules which will direct the expression of the TGF-A 2 product in appropriate host cells In a specific 11 - embodiment, a TGF-il/TGF-A 2 hybrid gene (FIG lb) was constructed and used to transfect CHO cells Transfectants producing as much as 500 pg mature, biologically active TGF-82 per ml culture medium were isolated.

Due to the degeneracy of the nucleotide coding sequences, other DNA sequences which encode substantially the same amino acid sequences as depicted in FIG la and FIG lb may be used in the practice of the present invention for the cloning and expression of TGF-A 2 Such alterations include deletions, additions or substitutions of different nucleotide residues resulting in a sequence that encodes the same or a functionally equivalent gene product The gene product may contain deletions, additions or substitutions of amino acid residues within the sequence, which result in a silent change thus producing a bioactive product Such amino acid substitions may be made on the basis of similarity in polarity, charge, solubility, hydrophobicity, hydrophilicity and/or the amphipathic nature of the residues involved For example, negatively charged amino acids include aspartic acid and glutamic acid; positively charged amino acids include lysine and arginine; amino acids with uncharged polar head groups or nonpolar head groups having similar hydrophilicity values include the following:

leucine, isoleucine, valine; glycine, alanine; asparagine, glutamine; serine, threonine; phenylalanine, tyrosine.

The nucleotide coding sequence for TGF-A 2 may be obtained from cell sources that produce TGF-f 2 like activity The coding sequence may be obtained by c DNA cloning of RNA isolated and purified from such cellular sources or by genomic cloning Either c DNA or genomic libraries of clones may be prepared from the DNA fragments generated using techniques well known in the art including but not limited to the use of restriction enzymes The fragments which encode TGF-A 2 may be identified by screening such libraries with a nucleotide probe that is substantially 12 complementary to any portion of the sequence depicted in FIG la Full length clones, i e, those containing the entire coding region for the TGF-A 2 precursor may be selected for expression.

In an alternate embodiment of the invention, the coding sequence of FIG la could be synthesized in whole or part, using chemical methods well known in the art See, for example, Caruthers et al, 1980, Nuc Acids Res Symp.

Ser 7:215-223; Crea and Horn, 1980, Nuc Acids Res.

9 ( 10):2331; Matteucci and Carruthers, 1980, Tetrahedron Letters 21:719 and Chow and Kempe, 1981, Nuc Acids Res.

9 ( 12):2807-2817 Alternatively, the protein could be produced using chemical methods to synthesize the amino acid sequence depicted in FIG la in whole or in part For example, peptides can be synthesized by solid phase techniques on a Beckman 990 instrument, and cleaved from the resin as previously described (Gentry, L E, et al, 1983, J Biol Chem 258:11219-11228; Gentry, L E and Lawton, A, 1986, Virology 152:421-431) Purification can be accomplished by preparative high performance liquid chromatography The composition of the peptides was confirmed by amino acid analysis.

In a specific embodiment, described in the examples herein, the TGF-A 2 coding sequence was obtained by c DNA cloning of human TGF- 2 precursor coding sequences derivea from polyadenylated RNA isolated from tamoxifen-treated human prostatic adenocarcinoma cell line, PC-3, previously shown to produce TGF-A 2 The entire coding region of one c DNA clone was sequenced and compared to the published sequence of human TGF-$ 1 (see FIG 2).

DNA sequence analysis of TGF-A 2 c DNA clones indicates that TGF-p 2, like TGF-f 1, is synthesized as a large precursor protein, the carboxy terminus of which is cleaved to yield the mature 112 amino acid TGF-A 2 monomer TGF-02 has been shown to have a molecular weight of 24,000 composed 13 of two disulfide-linked 13,000 dalton subunits (Ikeda et al., 1987, Biochemistry 26:2406-2410; Cheifetz et al, 1987, Cell 48:409-415) Therefore, the production of mature TGF- A 2 requires proper proteolytic cleavage as well as the formation of intra and inter-molecular disulfide bonds An amino terminal hydrophobic leader sequence (residue 3-19) is present in the precursor and may be responsible for directing the protein out of the cell The mature TGF-A 2 may still be associated with the remaining portion of the precursor during this process.

TGF-A 2 shows 71 % homology with TGF-A 1 throughout the mature portion of the precursor, implying a functional similarity which is supported by experimental evidence (Seyedin et al, 1987, J Biol Chem 262:1946-1949; Cheifetz et al, 1987, Cell 48:409-415) The amino portion of the precursor region of TGF-fl from human, rodent and simian sources (Derynck et al, 1985, Nature 316:701-705; Derynck et al, 1986, J Biol Chem 261:4377-4379; Sharples et al, 1987, DNA 6:239-244) is highly conserved and suggests that this part of the molecule may have an important biological function In contrast, there is no more than 31 % homology between the N-terminal precursor regions of TGF-A 1 and TGF-A 2 After cleavage of the putative signal peptide, the TGF-A 2 precursor would also contain more amino acids than TGF-Al precursor The primary structural differences within the amino terminal region of the TGF-A 1 and TGF-A 2 precursor proteins may reflect functional differences However, significant homologous regions within the precursors are found in isolated blocks suggesting the conservation of important functional domains even within the N-terminal precursor region.

Northern blot analysis revealed two major size classes of TGF-f 2-specific m RNA of 4 1 and 6 5 kb in BSC-40 cells.

Tamoxifen-treated PC-3 cells contain three TGF-A 2 transcripts of 4 1 kb, 5 1 kb, and 6 5 kb These different- 14 - sized messages could be the result of differential RNA splicing, polyadenylation, or both as has been described for other genes (Helfman et al, 1986, Mol Cell Biol 6:35823595; Sayre et al, 1987, Proc Natl Acad Sci USA 84:2941-2945) Preliminary analysis of another TGF-A 2 c DNA clone shows that it contains a 3 '-untranslated region approximately lkb larger than that of p PC-21 and p PC-14 and contains a different polyadenylation site suggesting that alternative polyadenylation is one factor responsible for 1 the generation of multiple TGF-A 2 m RN As observed on Northern blots.

BSC-40 cells contain comparable levels of TGF-1 i and TGF-A 2-specific transcripts: tamoxifen-treated PC-3 cells contain more TGF-fil m RNA than TGF-A 2 (FIG 3 B) The latter result is unexpected since these cells produce more TGF-A 2 protein than TGF-A 1 (Ikeda et al, 1987, Biochemistry 26:2406-2410) and suggests a post transcriptional level of regulation regarding the synthesis of this growth modulator.

Experiments aimed at understanding transcriptional and translational control mechanisms which lead to the production of TGF-Al and TGF-A 2 are in progress Production of adequate amounts of TGF-A 2 by recombinant DNA techniques, as has been done for TGF-i 1, should aid further in designing experiments to explore the different effects of this protein.

In another embodiment, the TGF- 2 coding sequence was obtained by c DNA cloning of simian TGF-$ 2 precursor coding sequences derived from polyadenylated RNA isolated from an African green monkey cell line, BSC-40 The entire coding region of one c DNA clone was sequenced The human and simian TGF-82 precursors appear to have identical amino acid sequences, and their nucleotide sequences are nearly identical.

- 5.2 CONSTRUCTION OF EXPRESSION VECTORS CONTAINING THE TGF-BETA 2 CODING SEQUENCE In order to express a biologically active, mature form of TGF-P 2, an expression vector/host system should be chosen which provides not only for high levels of transcription and translation but for the correct processing of the gene product This is especially important when employing the entire coding sequence of a TGF-A 2 precursor in the expression constructs because the mature form of TGF-A 2 appears to be derived from the precursor product via cellular processing events In addition, an expression/host cell system which provides for secretion of the product may be selected.

In particular, it appears that the mature TGF-,82, a disulfide-linked homodimer of 112 amino acids per subunit may be formed by cellular processing involving proteolytic cleavage between the Arg-Ala amino acids of the precursor (residue numbers 330 and 331 in FIG la) In addition, the TGF-,2 precursor contains three potential N-glycosylation sites not found in the mature form; the proper glycosylation of the precursor may be important to the cellular synthesis and release or secretion of the mature molecule Moreover, the mature form of TGF-P 2 comprises a disulfide-linked dimer involving nine cysteine residues per subunit Some of these are involved in interchain and others in intrachain disulfide bonds which affect the tertiary structure and configuration of the mature molecule, and, as a result, its biological activity Thus, the ability of a host cell used in the expression system to correctly express and process the TGF-62 gene product is important to the production of a biologically active, mature TGF-p 2.

A variety of animal host/expression vector systems (i.e, vectors which contain the necessary elements for directing the replication, transcription and translation of the TGF-i 2 coding sequence in an appropriate host cell) may 16 - be utilized equally well by the skilled artisan These include, but are not limited to, virus expression vector/mammalian host cell systems (e g, cytomegalovirus, vaccinia virus, adenovirus, and the like); insect virus expression vector/insect cell systems (e g, baculovirus); or nonviral promoter expression systems derived from the genomes of mammalian cells (e g, the mouse metallothionine promoter).

The expression elements of these vectors vary in their strength and specificities Depending on the host/vector system utilized, any one of a number of suitable transcription and translation elements may be used For instance, when cloning in mammalian cell systems, promoters isolated from the genome of mammalian cells, (e g mouse metallothionien promoter) or from viruses that grow in these cells, (e g vaccinia virus 7 5 K promoter) may be used.

Promoters produced by recombinant DNA or synthetic techniques may also be used to provide for transcription of the inserted sequences.

Specific initiation signals are also required for sufficient translation of inserted protein coding sequences.

These signals include the ATG initiation codon and adjacent sequences In cases where the entire TGF-A 2 gene including its own -i-iiaation codon cod adjacent sequences are inserted into the appropriate expression vectors, no additional translational control signals may be needed However, in cases where only a portion of the coding sequence is inserted, exogenous translational control signals, including the ATG initiation codon, must be provided Furthermore, the initiation codon must be in phase with the reading frame of the TGF-A 82 coding sequences to ensure translation of the entire insert These exogenous translational control signals and initiation codons can be of a variety of origins, both natural and synthetic The efficiency of 17 - expression may be enhanced by theinclusion of transcription attenuation sequences, enhancer elements, etc.

Any of the methods previously described for the insertion of DNA fragments into a vector may be used to construct expression vectors containing the TGF-R gene and appropriate transcriptional/translational control signals.

These methods may include in vitro recombinant DNA techniques, synthetic techniques and in vivo recombinations (genetic recombination).

In cases where an adenovirus is used as an expression vector, the TGF-,q 2 coding sequence may be ligated to an adenovirus transcription/translation control complex, e g, the late promoter and tripartite leader sequence This chimeric gene may then be inserted in the adenovirus genome by in vitro or in vivo recombination Insertion in a non- essential region of the viral genome (e g, region El or E 3) will result in a recombinant virus that is viable and capable of expressing TGF-A 2 in infected hosts Similarly, the vaccinia 7 5 K promoter may be used.

An alternative expression system which could be used to express TGF-82 is an insect system In one such system, Autographa californica nuclear polyhedrosis virus (Ac NPV) is used as a vector to express foreign genes The virus grows in Spodo Dtera frugiperda cells The TGF-A 2 coding sequence may be cloned into non-essential regions (for example the polyhedrin gene) of the virus and placed under control of an Ac NPV promoter (for example the polyhedrin promoter).

Successful insertion of the TGF- 2 coding sequence will result in inactivation of the polyhedrin gene and production of non-occluded recombinant virus (i e, virus lacking the proteinaceous coat coded for by the polyhedrin gene) These recombinant viruses are then used to infect Spodoptera frugiperda cells in which the inserted gene is expressed.

In addition, a host cell strain may be chosen which modulates the expression of the inserted sequences, or 18 - modifies and processes the gene product in the specific fashion desired Expression from certain promoters can be elevated in the presence of certain inducers, (e g zinc and cadmium ions for metallothionein promoters) Therefore, expression of the genetically engineered TGF-i 2 may be controlled This is important if the protein product of the cloned foreign gene is lethal to host cells Furthermore, modifications (e g glycosylation) and processing (e g, cleavage) of protein products are important for the function of the protein Different host cells have characteristic and specific mechanisms for the post-translational processing and modificatin of proteins Appropriate cell lines or host systems can be chosen to ensure the correct modification and processing of the foreign protein expressed.

5.3 IDENTIFICATION OF TRANSFECTANTS OR TRANSFOR- MANTS EXPRESSING THE 2 TGF-BETA GENE PRODUCT The host cells which contain the recombinant-TGF-02 coding sequence and which express the biologically active, mature product may be identified by at least four general approaches: (a) DNA-DNA hybridization; (b) the presence or absence of "marker" gene functions; (c) assessing the level of transcription as measured by the expression of TGF-A 2 m RNA transcripts in the host cell; and (d) detection of the mature gene product as measured by immunoassay and, ultimately, by its biological activity.

In the first approach, the presence of the TGF-i 2 coding set:ne :td in the ax-ression vectzr czn be detected by DNA-DNA hybridization using probes comprising nucleotide sequences that are homologous to the TGF-p 2 coding sequence substantially as shown-in FIG la, or portions or derivatives thereof.

In the second approach, the recombinant expression vector/host system can be identified and selected based upon 19 - the presence or absence of certain "marker" gene functions (e.g, thymidine kinase activity, resistance to antibiotics, resistance to methotrexate, transformation phenotype, occlusion body formation in baculovirus, etc) For example, if the TGF-42 coding sequence is inserted within a marker gene sequence of the vector, recombinants containing the TGF-j 2 coding sequence can be identified by the absence of the marker gene function Alternatively, a marker gene can be placed in tandem with the TGF-82 sequence under the control of the same or different promoter used to control the expression of the TGF-02 coding sequence Expression of the marker in response to induction or selection indicates expression of the TGF-A 2 coding sequence.

In the third approach, transcriptional activity for the TGF-beta 2 coding region can be assessed by hybridization assays For example, polyadenylated RNA can be isolated and analyzed by Northern blot using a probe homologous to the TGF-beta 2 coding sequence or particular portions thereof Alternatively, total nucleic acids of the host cell may be extracted and assayed for hybridization to such probes.

In the fourth approach, the expression of the mature protein product can be assessed immunologically, for example by Western blots, immunoassays such as radioimmuno- precipitation, enzyme-linked immunoassays and the like The ultimate test of the success of the expression system, however, involves the detection of the biologically active TGF-A 2 gene product Where the host cell secretes the gene product the cell free media obtained from the cultured transfectant host cell may be assayed for TGF-62 activity.

Where the gene product is not secreted, cell lysates may be assayed for such activity In either case, biological assays such as the growth inhibition assay described herein or the stimulation of anchorage independent growth in target cells (Twardzik and Sherwin, 1985, J Cell Biochem.

- 28:289-297; Delarco and Todaro, 1978, Proc Natl Acad Sci.

U.S A 75:4001-4005) or the like may be used.

Once a clone that produces high levels of biologically active, mature TGF-A 2 is identified, the clone may be expanded and the TGF-A 2 may be purified using techniques well known in the art Such methods include immunoaffinity purification, chromatographic methods including high performance liquid chromatography, and the like.

6 EXAMPLE: c DNA CLONING OF TGF-BETA-2 PRECURSOR FROM PC-3 CELLS The following examples describe the c DNA cloning of TGF-f 2 precursor coding sequences from the human prostatic adenocarcinoma cell line, PC-3, from which TGF-beta-2 was previously isolated.

6.1 MATERIALS AND METHODS The following procedures were used to clone c DN As encoding human TGF-beta 2 precursor.

6 1 1 GROWTH OF CELLS AND RNA EXTRACTION.

The human prostatic adenocarcinoma cell line, PC-3, was grown in tamoxifen-supplemented medium as described (Ikeda et al, 1987, Biochemistry 26:2406-2410) MCF-7 cells were grown in Dulbecco's modified Eagle's medium containing 10 % fetal calf serum and 6 units/ml insulin All other cell lines were grown in the same medium without insulin Polyadenylated RNA was isolated by oligold Tl- cellulose chromatography as described (Purchio and Fareed, 1979, J Virol 29:76 Z-769).

6.1 2 c DNA LIBRARY CONSTRUCTION AND SCREENING Double-stranded c DNA was synthesized from polyadenylated RNA isolated from PC-3 cells treated with tamoxifen for 24 hours as described (Maniatis et al, 1982, 21 - in Molecular Cloning: A Laboratory Manual, Cold Spring Harbor Press, Cold Spring Harbor, New York) c DNA fractions greater than 1000 base pairs were cloned into lambda gt O l O as described (Webb et al, 1987, DNA 6:71-79) The library was first screened in duplicate with a l 32 Pl-labeled 24-fold degenerate probe complimentary to DNA encoding amino acids WKWIHEP (probe 1) which are conserved between TGF-R 1 and TGF-fi 2:

l 5 '-GGTTCGTGTATCCATTTCCA-3 'l C A SG A Positive clones were then screened with a second 128-fold degenerate probe complimentary to DNA encoding amino acids CFRNVQD (probe 2); five out of these seven amino acids are specific for TGF-Beta 2:

l 5 'TCTTGAACGTTTCTGAAGCA-3 'l C C A C A A G T Hybridization was performed at 42 'C in 6 XSSC, 5 X Denhart's solution, 0 15 m M pyrophosphate, 100 micrograms/ml denatured calf thymus DNA, 100 micrograms/ml yeast t RNA and 1 m M EDTA (Maniatis et al, 1982, in Molecular Cloning: A Laboratory Manual, Cold Spring Harbor Press, Cold Spring Harbor, New York) Filters were washed at 42 'C in 2 XSSC, 0 1 % Na Dod 504, 4 four times for 30 min Several c DNA clones were isolated which hybridized to both probes and were subcloned into p EMBL (Dante et al, 1983, Nucleic Acids Res 11:1645-1654).

One clone (p PC-21) containing a 2 6 kb insert was sequenced on both strands by the dideoxy chain-termination method (Sanger et al, 1977, Proc Natl Acad Sci USA 74:5463- 5467) using various restriction and exonuclease III deletion fragments combined with specific oligonucleotide priming (Henikoff, 1984, Gene 28:351-359) Another clone (p PC-14) containing a 2 2 kb insert was partially sequenced Dot 22 - matrix analysis was performed on an IBM ATPC using Gene Pro software from Riverside Scientific Enterprises (Seattle, WA).

6 1 3 NORTHERN BLOT ANALYSIS Polyadenylated RNA was fractionated on a 1 % agaroseformaldehyde gel (Lehrach et al, 1977, Biochemistry 16:4743-4751), transferred to a nylon membrane (Hybond, 32 Amersham), and hybridized to l 32 Pl-labeled probe.

Hybridization was carried out at 42 'C in 50 % formamide containing 0 9 M Na Cl, 50 m M sodium phosphate (p H 7 0), 5 m M EDTA, 0 1 % Na Dod 504, 4 X Denhardt's solution, 0 4 mg/ml of yeast t RNA, and 0 25 mg/ml of denatured calf thymus DNA.

Filters were washed at 65 C in 0 25 X SSC, 0 1 % Na Dod 504, 1 dried, and exposed to Cronex-4 X-ray film (Du Pont) with the aid of Lightening Plus intensifier screens (Du Pont).

6.2 RESULTS A c DNA library was constructed using polyadenylated RNA isolated from tamoxifen treated PC-3 cells Earlier observations indicated that tamoxifen treatment resulted in a 2 to 5-fold increase in the secretion of TGF-A 2 (Ikeda et al., 1987, Biochemistry 26:2406-2410) The library was screened with probes 1 and 2 as described above Five clones were obtained which hybridized to both probes: one clone, p PC-21, containing a 2 6 kb insert, was chosen for sequencing Another clone, p PC-14, containing a 2 2 kb insert, was partially sequenced The DNA and deduced amino acid sae-snces are shown figure 1.

p PC-21 contains a single open reading frame coding for a deduced polypeptide of 442 amino acids; the 112 carboxy terminal amino acids comprise the mature TGF-P 2 monomer (boxed in FIG la) The first methionine encoded by the open reading frame is immediately followed by a stretch of hydrophobic and uncharged amino acids (overlined in FIG la) 23 characteristic of a signal peptide Neither the nucleotide sequence coding for this methionine nor those coding for the next two methionines present in the open reading frame are homologous to the consensus sequence for the initiating methionine sequence (Kozak, 1986, Cell 44:283-292) Because translation usually initiates with the first methionine in an open reading frame and because regions homologous to TGF-A 1, as discussed below, occur upstream of the second methionine, the first methionine has been tentatively assigned as the site of translation initiation It appears then, that TGF-A 2, like TGF-A 1, is expressed as part of a much larger secreted precursor The p PC-21 clone contains 467 bp upstream of the putative initiating methionine and a 3 ' untranslated region of approximately 800 bp including a poly (Al track, fifteen bases upstream of which is located a polyadenylation signal sequence (Proudfoot and Brownlee, 1976, Nature 263:211-214).

The nucleotide sequence homology within the coding regions of the TGF-A 1 and TGF-A 2 p PC-21 c DNA clone was determined to be 53 % The regions coding for the mature proteins have 57 % homology while the upstream precursor regions have 48 % homology After optimal alignment of the two sequences, several nucleotide insertions were noted in the TGF-A 2 precursor region, one of which extended for 75 nuclectides Whether these insertions are due to the presence of extra exons in TGF-A 2 is unknown No significant homology was detected between the DNA sequences in the non-coding regions of the two clones In fact, while TGF-fil had extended G-o orich non-coding regions, TGF-B 2 had extensive A-T rich non-coding regions Both c DNA clones contain repeating structural motifs in the 3 ' noncoding region with the repeats in TGF-Al consisting of (purine) CCCC (Sharples et al, 1987, DNA 6:239-244) and in TGF-p 2 of ATG or A(pyrimidine)(purine).

24 - Restriction mapping of many clones revealed that one clone, p PC-14, lacked a Kpn I restriction site located in the amino portion of TGF-82 coding sequence Restriction maps of p PC-14 and p PC-21 are shown in FIG ld p PC-14 was sequenced over a strech of about 100 nucleotides corresponding to this region of the molecule by specifically priming with a 20-mer oligonucleotide complementary to nucleotides 277 to 296 in FIG la The results show that the p PC-14 clone contains an 87 nucleotide deletion (nucleotide positions 346 to 432 in FIG la; see also FIG.

le) that accounts for the missing K Dn I site and which is replaced by the sequence AAT, the codon for Asparagine The results suggest that the p PC-14 clone encodes a shorter TGF-A 2 precursor of 414 amino acids differing from the sequence encoded by p PC-21 only in that amino acid residues 116 through 144 are deleted and replaced by a single Asparagine residue.

Although the entire coding region of p PC-14 was not determined, it is probably in perfect agreement with the p PC-21 coding sequence since, except for the Kpn I site, restriction maps of the two clones overlap perfectly (FIG.

ld) Furthermore, a simian clone encoding a 414 amino acid TGF-A precursor containing the same 29 amino acid deletion and replacement has been identified, as described in example Section 7, infra This simian clone has a coding sequence which is nearly identical to that of the human p PC-21 clone in the regions 5 ' and 3 ' to the deletion.

Figure 2 A shows the deduced protein sequence of human TGF-l (Derynck et al, 1985, Nature 316:701-705) compared to that of human TGF-A 2-442 It was determined that TGF-A 2 is 7,_ homolcgous with human TGF-4 ' throughout the mature portion of the molecule as reported previously (Marquardt et al., 1987, J Biol Chem, in press) The amino portion of the precursor upstream of the mature molecule shows a 31 % homology between TGF-f 11 and TGF-f 12-442 The dot matrix - homology comparison shown in figure 2 B reveals that significant homology exists in several specific areas of the proteins Comparison of the N-terminal amino acid sequences in the putative signal peptide region reveals no significant homology.

In TGF-A 2, the signal sequence cleavage site is predicted to be after amino acid 20 (serine) and after amino acid 29 (glycine) in TGF-61 (Von Heijne, 1983, Eur J.

Biochem 133:17-21) This cleavage site directly precedes the first block of homology between TGF-Al and TGF-A 2 which extends for 34 amino acids downstream After removal of the signal sequences, the TGF-fil and TGF-A 2 precursors would share identical N-termini over the first four amino acids, including the cysteine at position 4 Fourteen amino acids downstream of this putative N-terminus, 19 out of the next 21 amino acids are conserved between TGF-Al and TGF-fi 2, a homology block larger than any seen even in the C-terminal region containing the mature TGF-R protein Several more blocks of strong homology exist within the region upstream of the mature protein as seen in figure 2 A and 2 B These domains are separated by long stretches of non homologous amino acids.

The TGF-A 2 precursor has three potential N- glycosylation sites (located at residues 72, 168, and 269 in FIG la) Only the first site is conserved in TGF-A 1, and lies within a larger block of conserved residues, suggesting that this potential glycosylation site has important structural and/or functional characteristics.

After removal of the signal sequence, the TGF-A 2 precursor would contain either 31 or 59 amino acids more than its TGF-A 1 counterpart An additional cysteine residue in TGF-A 2 is located just upstream of a large region of non homologous amino acids that precedes the mature sequence.

As with TGF-A 1, the cleavage site of the mature TGF-A 82 protein occurs just after a region of 4-5 basic amino acids as shown in FIG 2 A The mature region contains nine cysteines Conservation of 7 of the 9 cysteines is characteristic for the different members of the TGF-A family Hydropathy analyses of TGF-fi 11 and TGF-A 2 reveal similar patterns in both the precursor and mature regions with both proteins being generally hydrophilic in nature (data not shown).

Figure 3 A shows a Northern blot analysis using p PC-21 to probe polyadenylated RNA from BSC-40 (an African green monkey kidney cell line) and tamoxifen-treated PC-3 cells.

PC-3 cells contain three major TGF-f 12-specific m RNA species of 4 1, 5 1 and 6 5 kb in size (FIG 3 A, lane 2); BSC-40 cells contain predominantly the 4 1 and 6 5 kb transcripts and lesser amounts of the 5 1 kb RNA (FIG 3 A, lane 1).

Note that the p PC-21 probe does not detect the 2 5 kb TGF- f 1-specific m RNA species present in this cell line under the hybridization conditions used here These results and previous observations (Sharples et al, 1987, DNA 6:239-244) suggest that BSC-40 cells contain both TGF-f 1 and TGF-f 2specific m RNA's In order to demonstrate this more clearly, Northern blots were hybridized to a mixture containing equal amounts of TGF-f 1 and TGF-f 12 probes radiolabelled to the same specific activity Lane 1 of FIG 3 B shows that BSC-40 cells contain the 2 5 kb TGF-fl-specific m RNA as well as the 4 1 and 6 5 kb TGF-Beta 2 m RNA species: lane 2 of figure 3 B shows that tamoxifen treated PC-3 cells also contain the 2.5 kb TGF-A 1-specific m RNA FIG 3 B also demonstrates that tamoxifen-treated PC-3 cells contain more TGF-1 pl-specific than TGF-f 12-specific message.

The identification of TGF-fi 2-specific c DNA clones has enabled us to screen for TGF-fi 2 m RNA in various cell lines.

The northern blot shown in FIG 4 shows that TGF-P 2-specific transcripts could be detected in HBL 100 (a normal epithelial cell line derived from human milk, obtained from Dr Greg Schultz), MCF-7 (a human mammary carcinoma cell liney, SK- 27 - MEL 28 (a melanoma cell line), and KB cells (a nasopharyngeal carcinoma cell line) contain very low levels of TGF-$ 2 m RNA.

7 EXAMPLE: c DNA CLONING OF TGF-BETA-2 PRECURSOR FROM BSC-40 CELLS The following examples describe the c DNA cloning of TGF-P 2 coding sequences from the African green monkey kidney cell line, BSC-40, shown to contain TGF-p 2 specific m RN As (Section 6, supra) The results indicate that simian TGF- 2, like human TGF-A 2, is synthesized as one of at least two longer precursors from which the mature TGF-f 2 molecule is derived by proteolytic cleavage.

7.1 MATERIALS AND METHODS The following procedures were used to clone c DN As encoding simian TGF-A 2 precursor.

7.1 1 GROWTH OF CELLS AND RNA EXTRACTION BSC-40 cells were grown in Dulbecco's modified Eagle medium containing 10 % fetal calf serum Polyadenylated RNA was isolated by oligold Tl-cellulose chromatography as described (Purchio and Fareed, 1979, J Virol 29:763-769).

7.1 2 c DNA LIBRARY CONSTRUCTION AND SCREENING Double-stranded c DNA was synthesized from BSC-40 polyadenylated RNA as described (Maniatis et al, 1982, Molecular Cloning: A Laboratory Manual, Cold Spring Harbor Laboratory, Cold Spring Harbor, NY, 371-372) and after treatment with Eco RI methylase was ligated to oligonucleotide linkers containing an Eco RI restriction enzyme recognition site (Eco RI linkers) The c DNA was digested with Eco RI and fractionated by chromatography on Sephacryl S-1000 c DNA fractions of greater than 750 base pairs were pooled and ligated into lambda gtl O which had been cut with Eco RI (Davis et al, 1980, A Manual for Genetic Engineering: Advanced Bacterial Genetics; Cold Spring Harbor Laboratory, Cold Spring Harbor, NY), packaged (Grosveld et al, 1981, Gene 13:227-237) and plated on E.

coli C 600 r K-m Khfl The library was screened by plaque hybridization (Bentonet et al, 1977, Science 196:180-182) to l 32 Pl-labeled p PC-21 and p PC-14 probes Clone p BSC-40- 16, which hybridized the p PC-21 probe, and clone p BSC-40-1, which hybridized the p PC-14 probe, were isolated and subcloned into p EMBL The TGF-R 2 coding sequence of p BSC- 40-1 was determined by sequencing both strands using the dideoxy chain-termination method (Sanger et al, 1977, Proc.

Natl Acad Sci USA 74:5463-5467) p BSC-40-16 was partially sequenced.

7.2 RESULTS Two clones were obtained from a BSC-40 c DNA library which hybridized alternatively to probes constructed from the human TGF-R 2-442 and TGF-R 2-414 precursor coding sequences.

Clone p BSC-40-16, which hybridized to the TGF-42-442 probe, was sequenced over a 150 nucleotide strech (nucleotides 300 to 450 in FIG la) expected to contain the coding sequence for the 29 amino acid segment from positions 346 to 432 in FIG la Thie results show that, over this region, p BSC-40-16 encodes an amino acid sequence which is identical to the corresponding sequence in the human TGF- p 2-442 c DNA clone, p PC-21, and suggest that p BSC-40-16 encodes a 442 amino acid TGF-R 2 precursor.

Clone p BSC-40-1, which hybridized to the TGF-beta 2- 414 probe, was sequenced over the entire coding region The results show that this clone encodes a 414 amino acid TGF-R 2 precursor which is identical to the human TGF-R 2-442 precursor except that amino acid residues 116 through 144 human TGF-fi 2-442 are deleted and replaced by a single Asparagine residue At the nucleotide level, p BSC-40-1 differs from human TGF-62-442 in the deletion region:

nucleotides 346 through 432 in FIG la are deleted and replaced by the codon for Asparagine, AAT Except for 13 silent base changes, the two structures are otherwise perfectly homologous over the rest of the coding sequence.

8 EXAMPLE: EXPRESSION OF TGF-82 The following examples describe the expression of mature, biologically active TGF-P 2 in Chinese Hamster Ovary cells (CHO cells) transfected with a recombinant plasmid containing the coding sequence for mature human TGF-fi 2 ligated down-stream and in-frame with the coding sequence for the simian TGF-f 1 precursor, under the regulatory control of the SV 40 promoter sequences The construct directed the synthesis and secretion of mature, biologically active TGF-$ 2 at a level of about 0 5 mg/L.

8.1 MATERIALS AND METHODS 8.1 1 CELL CULTURE Dihydrofolate reductase (dhfr)-deficient Chinese Hamster Ovary (CHO) cells (Urlaub and Chasin, 1980 Proc.

Natl Acad Sci U S A 77:4216) were propagated in Ham's F-12 medium (Gibco Laboratories, NY) supplemented with 10 % fetal bovine serum (FBS) and 150 ug/ml of L-proline.

Penicillin and streptomycin were included at 100 U/ml and ug/ml, respectively CHO transfectants were grown in Dulbecco's modified Eagle's medium containing the same supplements as those listed above CHO cells and their derivatives were routinely passaged by trypsinization at a 1:5 splitting ratio.

Methotrexate (Sigma, MO) was prepared at a stock concentration of 10 mg/ml in water Dilute Na OH ( 0 2 M) was added in order to solubilize the drug (final p H of 6) The - stock was filter-sterilized and stored at -20 C Stock solutions of methotrexate in media ( 100 u M) were kept at 4 C for no longer than 1 month.

8 1 2 DNA MANIPULATIONS AND PLASMID CONSTRUCTIONS Restriction enzymes, T 4 DNA ligase, calf intestinal phosphatase, the Klenow fragment of DNA polymerase I and other DNA reagents were purchased from Bethesda Research Laboratories, MD Standard DNA manipulations were performed as outlined in Maniatis, T, et al, 1982, Molecular Cloning: A Laboratory Manual, Cold Spring Harbor Laboratory, New York.

Plasmid p SV 2 (A 1-TGF-dhfr), which contains the simian TGF-f 1 c DNA and the mouse dhfr gene in tandem as well as intervening SV 40 sequences, was constructed as described (Gentry et al, 1987, Mol Cell Biol 7:3418).

Plasmid p SV 2/$I 1-2/dhfr was constructed as outlined in Section 8 2, infra.

8 1 3 DNA TRANSFECTIONS 6 Approximately 24 hours after seeding 10 dhfrdeficient CHO cells onto 100 mm dishes, the cultures were transfected with 20 ug of Nde I linearized p SV 2-(A 1-TGF-dhfr) plasmid as a calcium phosphate precipitate (Wigler, M, et al, 1979, Proc Natl Acad Sci U S A 76:1373-1376).

Briefly, 20 ug of linearized DNA was added to 1 ml of 250 m M sterile Ca C 12 A 1 ml portion of 2 X HEPES solution ( 280 m M Na Cl, 50 m M HEPES, 1 5 m M sodium phosphate, p H 7 1) was then added dropwise, and the mixture was allowed to sit on ice for 30 minutes The precipitate was then dispersed dropwise over the cells containing 10 ml of the F 12 media After incubation at 37 'C for 4 hours, the media was removed and replaced with 10 ml of F 12 media containing 25 % glycerol for seconds at room temperature Cells were rinsed once with 20 ml of F 12 media and incubated in the nonselective F 12 media ( 20 ml) for an additional 48 hours Selection for dhfr-expressing transfectants was accomplished by replacing the media with DMEM supplemented with 10 % dialyzed FBS (Gibco, N Y) and 150 ug/ml L-proline Colonies were observed after culturing the cells 10-14 days in the selection media Ten colonies were aspirated by a pasteur pipet and expanded.

8.1 4 SELECTION OF METHOTREXATE RESISTANT CELLS Dihydrofolate reductase (dhfr)-amplified cells were derived from theprimary transfectants essentially as described (Gasser, C S and Schimke, R T, 1986, J Biol.

Chem 261:6938-6946) After expansion, 105 cells were seeded onto 100 mm dishes and adapted to increasing concentrations of methotrexate The plate containing visible colonies at the highest methotrexate concentration was trypsinized and adapted to that concentration of methotrexate for at least two additional 1:5 cell passages.

Cells ( 105) were then seeded onto 100 mm dishes in 5 times the concentration of methotrexate The dish containing visible colonies was again trypsinized and adapted in the methotrexate containing medium Cells were frozen back at various stages of amplification in media containing 40 % FBS, % dimethyl sulfoxide and 50 % DMEM Methotrexate was not included in the freezing media.

8.1 5 GROWTH INHIBITION ASSAY Mink lung epithelial cells, Mv 1 Lu (Accession Number CCL-64, American Type Culture Collection), which are extremely sensitive to TGF-f 1, were utilized for the growth inhibition assay The assay was performed using the thymidine analog 5 '-l 125 Il-iodo-2 'deoxyuridine ( 125 Id U) to assess DNA synthesis One unit of activity was defined as the amount required to inhibit 50 % incorporation of 125 Id U compared to untreated CCL-64 cells.

32 - To assay transfected cells for secretion of active TGF-62, serum free supernatants were collected from one 24- hour collection on confluent cultures of cells and dialyzed extensively against 0 2 M acetic acid The acetic acid was removed by lyophilization and the sample was re-dissolved in sterile complete culture medium for assays.

8.2 CONSTRUCTION OF TGF-BETA 1/TGF-BETA 2 HYBRID PRECURSOR GENE FOR TGF-BETA 2 EXPRESSION A hybrid TGF-beta precursor gene consisting of simian TGF-Beta 1 precursor coding and 5 ' untranslated sequences joined in-frame with human TGF-Beta 2 mature coding and 3 ' untranslated sequences was constructed as illustrated in FIG lc.

p PC-21 was first digested with Eco RI, filled-in with Klenow enzyme, the 2 3 Kb fragment ligated into Hinc II digested p EMBL, and used to transform E coli Two clones, p PC-21/Hinc II and p PC-21/Hinc II, having inserts in opposite orientations, were used to generate overlapping Exo III digest fragments by digesting both with Sst I and Bam HI followed by Exo III digestion, Klenow repair, religation of the DNA, and transformation of E coli Two clones, Exo 5 9 and Exo 25 C were found to contain different lengths of 5 ' and 3 ' sequences, respectively, and were subcloned into p EMBL to generate p EMBL 5 9 and p EMBL 25 C. p EMBL 5 9 was digested with Hind III, blunt ended with Klenow enzyme,

digested with Kpn I, and the 0 6 Kb fragment (fragment 1) was isolated Exo 25 C was digested with Eco RI and Kpn I and the 1 1 Z frag-men z (fragment 2) was isolated.

p G 562 was digested with Bam HI, filled in with Klenow enzyme, digested with Eco RI and ligated to fragments 1 and 2 (p G 562 was derived from p G 520 (Mackett et al, 1984, J Virol.

49:857) by deletion of a single Eco RI site) The mixture was used to transform E coli and p G 562/CIFB was isolated.

p G 562/CIFB was digested with Pst I and Eco RI and the 1600 bp fragment was isolated and further digested with Xho II The resulting 400 bp Xho II-Eco RI fragment was isolated (fragment 3) p SV 2-beta-TGF (Gentry et al, 1987, Mol Cell Biol 7:3418) was digested with Apa I and Eco RI and the large 3000 bp fragment was isolated (fragment 4).

Two complimentary strands of DNA with the sequences shown below were synthesized, phosphorylated, annealed and ligated to fragments '3 ' and '4 ' described above.

CAA CAT CTG CAA AGC TCC CGG CAC CGC CGA GCT TTG GAT GCG GCC TAT TGC TTT AGA AAT GTG CAG GAT AAT TGC TGC CTA CGT CCA CTT TAC ATT GAT TTC AAG AGG 3 ' ' GATC CCT CTT GAA ATC AAT GTA AAG TGG ACG TAG GCA GCA ATT ATC CTG CAC ATT TCT AAA GCA ATA GGC CGC ATC CAA AGC TCG GCG GTG CCG GGA GCT TTG CAG ATG TTG GGCC 3 ' The ligation mixture was used to transform E coli and p$ 1/$ 2 was isolated.

Plasmid p 1 l/$ 2 was digested with Eco RI, filled in with the Klenow fragment of DNA polymerase I, cut with Hind III and the 1600 bp fragment was isolated: p SV 2, 11/$ 2 was constructed by inserting this fragment into p SV 2, neo which had been previously digested with Hind III and Hpa I to eliminate the neo gene.

p SV 2, 11/$ 2 was digested with Pvu I and Eco RI, filled in with Klenow enzyme, digested with Nde I and the 2 6 kb (approx) Nde I-Eco RI fragment was isolated and ligated to p SV 2,dhfr which had been digested with Nde I and Pvu II The ligation mixture was used to transform E coli and p SV 2/fi 1- P 2/dhfr was isolated.

34 - 8.3 EXPRESSION OF TGF-BETA 2 IN CHO CELLS p SV/A 1-g 2/dhfr was used to transfect dhfr-deficient CHO cells and dhfr-amplified cells were derived from the primary transfectants as described in Materials and Methods, supra.

Positive clones were identified by bioassay (inhibition of-mink lung epithelial cells (CCL-64) as described (Gentry et al, 1987, Mol Cell, Biol 7:3418) Recombinant proteins were also detected by Western blotting using an anti-peptide antisera made against the sequence NH 2YNTINPEASASPC-COOH (Gentry et al, 1987, Mol Cell, Biol.

7:3418) which is present in mature TGF-beta 2.

One line, l I 9, 12 5, was found to secrete 240 ng/ml TGF-A 2 (FIG 5) This line was then cloned by limiting dilution in 96 well plates One clone, 159, 12 5, cl 36, produced approximately 500 ng/ml (FIG 5).

Analysis of the protein secreted by this clone by Western blotting using anti-peptide antiserum is shown in FIG 6, revealing the presence of the mature 24 kd TGF-A 2 dimer as well as the larger (approx 90 kd) precursor form.

9 DEPOSIT OF MICROORGANISMS The following microorganisms have been deposited with the Agricultural Research Culture Collection, Northern Regional Research Center (IVRL) and have been assigned the following accession numbers:

Microorganism Plasmid Accession No.

Escherichia ccli HBE 01 p PC-21 B-18256 Escherichia coli HB 101 p PC-14 B-18333 Escherichia coli HB 101 p BSC-40-1 B-18335 Escherichia coli HB 101 p BSC-40-16 B-18334 Chinese Hamster Oary p Sv/Pl- 2/dhfr

Claims (1)

1. WHAT IS CLAIMED IS:

   1 A nucleotide sequence encoding transforming growth factor-fi 2 precursor comprising the nucleotide coding sequence substantially as depicted in FIG la from about nucleotide residue number 1 to about nucleotide residue number 1326.

   2 A nucleotide sequence encoding transforming growth factor-P 2 precursor comprising the nucleotide coding sequence substantially as depicted in FIG la from about nucleotide residue number 1 to about nucleotide residue number 1326 in which the nucleotide sequence from nucleotide residue numbers 346 to 432 is deleted and replaced by the nucleotide sequence "AAT".

   3 A nucleotide sequence encoding mature transforming growth factor-p 2 comprising the nucleotide sequence substantially as depicted in FIG la from about nucleotide residue number 991 to about nucleotide residue number 1326.

   4 A transforming growth factor-42 precursor comprising the amino acid sequence substantially as depicted in FIG la from about amino acid residue number 1 to about amino acid residue number 442.

   A transforming growth factor-A 2 precursor comprising the amino acid sequence substantially as depicted in FIG la from about amino acid residue number 1 to about amino acid residue number 442 in which the amino acid sequence from residue numbers 116 to 144 is deleted and replaced by a single Asparagine residue.

   6 Transforming growth factor-42 precursor comprising the amino acid sequence substantially as depicted in FIG la 36 - from about amino acid residue number 20 to about amino acid residue number 442.

   7 Transforming growth factor-d 2 precursor comprising the amino acid sequence substantially as depicted in FIG la from about amino acid residue number 20 to about amino acid number 442 in which the amino acid sequence from residue numbers 116 to 144 is deleted and replaced by a single Asparagine residue.

   8 Transforming growth factor-42 comprising the amino acid sequence substantially as depicted in FIG la from about amino acid residue number 331 to about amino acid residue number 442.

   9 A nucleotide sequence encoding hybrid transforming growth factor l/transforming growth factor-P 2 precursor comprising the nucleotide coding sequence substantially as depicted in FIG lb from about nucleotide residue number -70 to about nucleotide residue number 1755.

   A hybrid transforming growth factor 1/transforming growth factor-d 2 precursor comprising the amino acid sequence substantially as depicted il FIG lb from about amino acid residue numbe C I to about amino acid residue number 390.

   1 i A hybrid transforming growth factor-P 1/transforming growth factor-p 2 precursor ccmprising the amino acid sequence substantially as depicted in FIG lb from about amino acid residue number 30 to about amino acid residue number 390.

   12 A method for producing transforming growth factor- P 2 comprising:

   (a) culturing a eucaryotic cell containing a nucleotide sequence encoding transforming :- 37 - growth factor- 2 under the control of a second nucleotide sequence that regulates gene expression so that a peptide or protein having transforming growth factor-52 activity is produced by the eucaryotic cell; and (b) recovering the transforming growth factor-,82 from the culture.

   13 The method according to claim 12 in which the nucleotide sequence encoding transforming growth factor-52 comprises the nucleotide sequence substantially as depicted in FIG la from nucleotide number 1 to 1339.

   14 The method according to claim 12 in which the nucleotide sequence encoding transforming growth factor-d 2 comprises the nucleotide sequence substantially as depicted in FIG lb from nucleotide number -70 to 1755.

   The method according to claim 12 in which the eucaryotic cell comprises a Chinese Hamster Ovary cell.

   16 The method according to claim 12 in which the second nucleotide sequence which controls gene expression comprises an SV 40 promoter.

   17 The method according to claim 12 in which the second nucleotide sequence comprises a promoter and a sequence encoding a selectable marker for which the eucaryotic cell is deficient, so that the eucaryotic cell containing the transforming growth factor-42 coding sequence can be identified.

   18 The method according to claim 17 in which the selectable marker comprises dihydrofolate reductase.

   38 - 19 The method according to claim 18 further comprising exposing the eucaryotic cell to methotrexate, so that resistant colonies are selected which contain amplified levels of the coding sequences for dihydrofolate reductase and transforming growth factor-fi 2.

   The method according to claim 19 in which the - eucaryotic cell comprises a dihydrofolate reductase-deficient Chinese Hamster Ovary cell.

   21 A method for producing transforming growth factor- p 2, comprising (a) culturing transfectant 1 P 9, 12 5 clone 36 as deposited with the ATCC; and (b) recovering the transforming growth factor- Beta 2 from the culture.

   22 The method according to claim 21 in which the transfectant is cultured in the presence of methotrexate.

   23 A eucaryotic cell containing a nucleotide sequence encoding transforming growth factor-d 2 under the control of a second nucleotide sequence that regulates gene expression so that the eucaryotic cell produces active transforming growth factor-d 2.

   24 The eucaryotic cell according to claim 23 in which the nucleotide secuence encoding the transforming grcwth factor-42 comprises the nucleotide sequence substantially as depicted in FIG lb from nucleotide number -70 to 1755.

   The eucaryotic cell according to claim 23 which comprises a Chinese Hamster Ovary cell.

   39 - 26 The eucaryotic cell according to claim 23 in which the second nucleotide sequence which controls gene expression comprises an SV 40 promoter.

   27 The eucaryotic cell according to claim 23 in which the second nucleotide sequence comprises a promoter and a sequence encoding a selectable marker for which the eucaryotic cell is deficient, so that eucaryotic cell containing the simian transforming growth factor-p 2 coding sequence can be identified.

   28 The eucaryotic cell according to claim 27 in which the selectable marker comprises dihydrofolate reductase.

   29 The eucaryotic cell according to claim 28 which comprises a dihydrofolate reductase-deficient Chinese Hamster Ovary cell.

   A cell line comprising 19, 12 5, clone 36 as deposited with the ATCC.

   P Rt /-shet 191: -: P:e Of:e A S:: '-s H-: F -:: ' -'":::r Lnn:n V;WC:R 4 T;-ecr:c:: ma be o:e rca e T: e a-Fze Ofn:,