JPH03504796A - Nucleic acids encoding TGF-β3 and uses thereof - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

[Detailed description of the invention] Nucleic acid encoding TGF-β3 and its use in mammalian cells in culture Peptides capable of inducing current transformation include transforming growth The name of factor (TGF) is given (Reference 1.2). α-type TGF is competes with epidermal growth factor (EGF) for binding to the same cell surface receptors Do (3).

The 50 amino acid TGF=α species has been purified and has sequence homology with EGF. is shown (4). TGF-α is synthesized by various transformed cell lines (3.5.6.91). The 50 amino acid TGF-α starts with 160 amino acids. It is synthesized as part of the acid precursor molecule, which leads to N- and C-terminal protein addition. Upon water-splitting processing, the mature peptide is produced (7.8). apparent Detection of TGF-α species with higher molecular weight (l, 2.9) above This may be due to variations in the processing of amino acid precursors (92).

β-type TGF activity is associated with kidney (12), placenta (13) and platelet (14,15) It has been isolated from many normal tissues, including tumor cells (10, 11). T GF-β is present in platelets, which is similar to platelet-derived growth factor (PDGF) and It also contains an EGF-like peptide (16). Bovine TGF-β has been shown to be effective in rats. It promotes wound healing17), induces fetal RMM cells to differentiate, and induces cartilage-specific It has been shown that it is possible to synthesize large macromolecules80). However, the injection of TGF-β Treatment of NRK fibroblasts with EGF results in an increase in the number of membrane receptors for EGF. Become a fruit (20). This observation suggests that EGF and TGF-α in these cells (10.11). difference Furthermore, TGF-β alone induced AKR-2B fibroblasts to colonize soft agar. (21). Increased amounts of TGF-β are present in some transformed cells. secreted by the vesicles (22). In addition to its ability to stimulate cell proliferation, TG F-β has been shown to inhibit anchorage-dependent proliferation of various human cancer cell lines. Tetaku 13). Currently, TGF-β has been developed in African green monkeys (BSC-1). may be identical or very similar to growth inhibitory substances isolated from cells; It is considered (24). TGF-β stimulates the proliferation of certain cell line types. Whether it acts to inhibit or inhibit depends on the physiological conditions of the cell and other components. It appears to depend on a number of variables, including the presence of a long factor.

Bovine TGF-β had been purified to a stage where it could be sequenced. The most mature protein The first 15 amino terminal residues are A1a-Leu-A5p-Thr-A5 n-T yr-CM CP he-3er-3er-T hr-G 1y-L ys-Asn-CM C- (where CMC is cysteine or semi-cystine residue S-carboxymethylcysteine, which represents a group).

Human TGF-β from human placenta and platelets had been purified to the same stage. womb Board TGF-β was reported to have the following amine terminal sequence: A1a-L eu-A 5p-T hr-A 5n-T yr-CM CP he-(S er-S er)-T hr-G 1u-L ys-A 5n-CM C-V alX -G In-L eu-T yr-11e-A 5p-P he-X -(Lys)-A5p-Leu-GIy- (where X is undetermined and , CMC are the same as defined above). Platelet TGF-β has the following amino terminal sequence: The reported 2A 1a-L eu-A 5p-T hr-A 5n-Ty r-X-P he-3er- (where CMC and X are the same as defined above ).

h) TGF-β has two extremely similar molecular weights (Mr = 12.500). It consists of polypeptide chains, which are held together by covalent bonds through disulfide bonds. It was reported that there were. This disulfide bond gives the structure of the TGF-β molecule It was thought that it played an important role in

Several other proteins related to TGF-β by some amino acid sequence homology factors were listed. The β-chains of inhibin A and B are homologous system It is related to TGF-β due to the arrangement of the amino acid residues and the homology of some other amino acid sequences. This suggests that inhibin or more precisely activin (inhibin) Hybin β, or a dimer of β8 chain) is presumed to be structurally related to TGF-β. It had been. Inhibin suppresses the release of FSH from the pituitary gland, while Bottle increases the release of FSH (81.82). TGF-β has this activity. is not known.

Special table Hei 3-504796 (3) Mullerian inhibitors have a C-terminal region homologous to TGF-β. It inhibits Mullerian-induced tumor growth (83.84).

TGF-β prepared by purification from biological material is HTLV-II + or pose a risk of contamination with infectious substances such as hepatitis viruses. Therefore, the book The aim of the invention is to produce TGF-β from a source that does not pose a risk of contamination. Ru.

Another objective is to hybridize with DNA encoding biologically active TGF-β. The objective is to produce nucleic acids that induce When appropriately labeled, this nucleic acid For diagnostic assays of GF-β mRNA and of the DNA encoding TGF-β. Useful for distance.

Yet another object of the present invention is to obtain a vector containing DNA encoding TGF-β. production of host cell transformants expressing biologically active TGF-β It is to be.

According to the invention, the above object comprises: (a) containing a nucleic acid encoding TGF-β; (b) transform a heterologous eukaryotic host cell with this vector; , (c) culturing the transformed cells, and (d) extracting TGF-β from the culture. This is achieved by a method consisting of recovering.

Coding two subtypes of TGF-β (TGF-β and TGF-β3) Nucleic acids are provided which are useful in constructing the vectors described above.

also labels this nucleic acid or a nucleic acid that can hybridize with it, Diagnosis of DNA or mRNA encoding TGF-β or related proteins Used for testing.

The production of TGF-β derivatives by recombinant methods is the code for TGF-β disclosed in the present invention. This is made possible by knowledge of arrays. These derivatives encode TGF-β. Silent and expressing mutants in a given nucleic acid are included.

Silent variants do not arise from the substitution of one synonymous fudon for another. Ru. Here, both codons code for the same amino acid, but this substitution For example, by changing the secondary structure of TGF-β mRNA, TGF- can have a positive effect on the yield of β, and its useful substitutions are transformation Confirm by screening TGF-β yield from the body.

The expressed TGF-β variants fall into three classes: deletions, substitutions or insertions. divided into one or more of the following: Deletions can be made without inserting replacement residues. Characterized by removal of amino acid residues. The deletion mutant of TGF~β is When preparing fragments of TGF-β, for example, it is possible to delete immune epitopes. It is useful when it is desirable to

Substitutional variants are variants in which one amino acid residue is replaced by another . Such mutations, particularly substitutions targeted within the primary amino acid sequence, It is extremely difficult to prepare by methods other than synthesis. These are the biology of TGF-β. useful in altering the activity of Substitution mutants include TGF-β subtypes. and allelic forms of TGF-β.

TGF-β is found as a disulfide-linked dimer.

Another variant includes a predetermined variant of TGF-β with a normal homodimeric disulfide TGF-β chain having a natural amino acid sequence disulfide-bonded by a di-bond and heterodimers of the TGF-β subtype. this place If so, such mutants contain biologically active TGF-β as well as biologically inactive TGF-β. Contains sexual TGF-β.

Insertional variants are those in which one or more residues are present in or in the internal TGF-β sequence. Which is the variant located at the end of. This class of variants includes T.G. Fusion proteins obtained by insertion of F-β into the carboxy or amino/terminal residues Contains strong qualities. TGF-β fusions with bacteria or other immunogenic proteins generates antibodies against TGF-β or a predetermined fragment thereof It is useful for

FIG. 1a is a schematic representation of the TGF-β1 mRNA, showing the coding sequence. It is shown enclosed in a box. The 112 amino acid TGF-βI (hatched) is the code It is encoded at the 3' end of the sequence. The sequenced cDNA insert λβc1. 3.19.3.32.4.10, 4.33.4.37 and 5.7b (see below) The genomic DNA sequences of the three untranslated regions are arranged on the schematic diagram.

Figures 1b (+) to 1b (II) (hereinafter collectively referred to as Figure 1b) are BreTG determined from several overlapping cDNA and genomic 3' regions 1 shows the sequence and deduced amino acid sequence of F-β, cDNA. Stable The 5' terminal region that can fold into an appin loop is underlined. This blur TG F-β, cDNA encodes a protein of 390 amino acids, of which C - The terminal 112 amino acids (boxed) encode mature TGF-β1. Before The hydrophobic domain found at the N-terminus of the precursor is overlined. above The underlined A rg-A rg dipeptide is a protein for the release of TGF-β. It is located in front of the protein hydrolytic cleavage site. Three possible Ns in pre-TGF-β1 - Glycosylation sites are marked with a line above them. After the stop codon, underline followed by a GC-rich sequence and a downstream TATA-like sequence.

Figure 2 shows the genomic fragment encoding the TGF-β1 exon and its The figure shows the deduced amino acid sequence of . Arrow indicates ff1 RNA processing part The position (intron-exon connection point) is shown. Residue numbers correspond to Figure 1b.

Figure 3 shows bovine TGF-β, mature and human TGF-β, and human and pig TGF-β. This is a comparison of the known N-terminal sequences of the amino acid sequences of TGF-β. TGF -β, and non-conservative substitutions in TGF-β1 that constitute the β3 subtype. Indicated by dots above residues. Unless otherwise specified, this drawing will not be referred to in this specification. Use the number of amino acid residues indicated.

Figures 4a to 4c show the cDNA sequences of human and porcine TGF-β. It is. The human h cDNA sequence covers part of the blurred sequence region and all of the mature sequence. are doing. Human and pig sequences are designated hu4 and 10+11.3, respectively. They are lined up in rows. Gear knobs are aligned to maximize nucleotide homology. introduced inside.

Homologous bases are marked with an asterisk.

Figure 5 shows the amino acid sequences encoded by the cDNAs in Figures 4a to 4C. It indicates the column. Homologous residues are indicated with an asterisk. Translation of candidate pig sequences The starting methionyl residue is at position 43.88 or 90 (boxed methionyl residue). The C-terminal residue of both the human and pig sequences is 499 (pig) or 2 This is Seryl 04 (human).

TGF-β can be used in recombinant cell culture while retaining its growth-altering activity. It turned out to be extremely difficult to synthesize.

As shown in Figures 1b, 3 and 5, the amino acid sequence of mature TGF-β The sequence contains a large number of cysteine residues (9), at least some of which are derived from the natural source. It is involved in chain opening cross-linking when forming homodimeric TGF-β recovered from the source. It is clear that Furthermore, TGF-β appears to some extent in the extracellular medium. However, TGF-β has a recognition characteristic typical of most secreted proteins. It has a large amino-terminal region that does not include a possible NH,-terminal signal peptide sequence. expressed as a precursor molecule. However, the appropriateness of the primary translation product during recombinant culture Despite the expected difficulty in processing eukaryotic cells, The cells were transformed to express heterologous TGF-β.

The present invention relates to recombinant synthesis of TGF-β. This TGF-β has the sequence shown in Figure 1b. biologically active BreTGF-β, mature TGF-β, their polypeptides with fragment, as well as such pre-TGF-β (shown in Figure 1b). (including TGF-β subtypes or alleles other than TGF-β1 subtype), Insertions, substitutions and/or deletions of mature TGF-β or polypeptide fragments Defined as including deletion mutants.

Biologically active TGF-β means that EGF-enrichment of target cell lines is independent of immobilization. induce sexual proliferation (81) and/or growth inhibition of neoplastic cell lines (23). Defined as something that can be done. Fixation-independent growth refers to growth in soft agar. The ability of TGF-β and EGF treatment to form non-neoplastic target cells meaning (property that is said to be for the transformation of cells, hence the name tranformation) transforming growth factors). This suggests that TGF-β “causes” cancer. (This is because many normal cells now express TGF-β. (This is because it is known to do so.) Conversely, TGF-β has some positive It is known to inhibit the growth of normal cell types and various neoplastic cells such as A349. It is.

Biological activity for purposes of the present invention also typically includes biological activity relative to natural TGF-β. This includes the ability to cross-react with antisera generated by

Natural TGF-β is obtained from platelets or other natural sources. be. Immunological cross-reactivity is a measure of one active epitope, but not necessarily active TGF-β domain involved in the induction of anchorage-independent proliferation of target cells. It is not a measure. However, immunologically cross-reactive proteins themselves are defined in the present invention unless they also exhibit an activity that affects proliferation. It is not biologically active. That is, biologically active TGF-β All of the mutants promote growth in defined assays but are immunologically cross-reactive. Not all TGF-β is biologically active. Induces anchorage-independent proliferation As a corollary of maintaining proper conformation, TGF-β, which is capable of It goes without saying that antisera generated against molecules often show immunological cross-reactivity. It is a theory.

The nucleotide sequence in Figure 1b consists of several overlapping cDNA and gene sequences. Analyze the fragments and determine the contiguous sequence corresponding to TGF-β, precursor mRNA. I got it by doing that. According to Figure 1b, the inishek-ATG starts from the 5' end. The coding sequence of the 390-residue polypeptide, located at 841 nucleotides, has been revealed. I'm sure.

Some regions in the cDNA sequence have unusually high G-C content. Inishi The eta-ATG is surrounded by two G-C rich regions of approximately 200 bp each. It is. Furthermore, some regions of the cDNA, especially the 5' end, contain more than 80% It has a region of G-C content.

The location of these G-C rich regions has been determined by numerous cDNA cloning artifacts. The region and partial length cDNA correspond to the region obtained.

The 5′ untranslated region of TGF-β mRNA is 841 nucleotides long. (assuming that ATG is located at nucleotide 842). It contains a long sequence consisting of This unusually long 5° nontranslation of high G-C content Although the biological relevance of this region is unknown, it is similar to the structure of c-myc mRNA. It is similar to the structure of

However, there is no notable sequence homology between these two sequences. c- The long 5' untranslated region of myc is hypothesized to have functional importance. There is (37). Rich in GC in the 5' untranslated sequence of human c-tnyc mRNA A 5 - Some proximal parts can form stable hairpin loops. It has an area of Similarly, the first 12 untranslated BreTGF-β, cDNA 0bp is theoretically equivalent to a hairpin loop with a calculated stability of -91 kcal. can be folded into a folded structure. This long 5” untranslated sequence and possible stable The apin loop structure plays a role in mRNA stability or in the regulation of transcription. You can also Therefore, improving the yield of TGF-β1 from recombinant cell culture In order to identify structures that can be (e.g., from a tylstambax substance).

Immediately after the stop codon, which precedes base 2015, there is a prominent G- A C-rich sequence follows (underlined in Figure 1b). This array is multiple of CCGCC It consists of repetitions of The unique nature of this sequence is probably due to the 3” untranslated nature of the mRNA. The cause of the inability to clone translated ends as cDNA sequences (probably, colon EII DNA polymerase I converts this sequence into a second (This is due to the fact that it cannot be used as a template for CDNA synthesis of the strand of similar gender High quality repeat sequences are present in human dihydrofolate reductase (38) and human transferrin receptor. condition, human adenosine deaminase (39), and herpesvirus thymidine It is found in the promoter region of the gene for kinase (40). In the latter case , McKnight et al. (40) suggested that these structural elements are of primary importance for transcription efficiency. It was shown that it has.

In addition, the promoter-specific transcription factor Spl appears in the SV40 early promoter region. was shown to bind to sequences in the region and related monkey promoters (4L42 ). In all these cases, the G-C rich repeats have G oldberg-Hog The TATA sequence of ness follows closely. However, in the case of BureTGF-β1, In some cases, these sequences are located in the 3′ untranslated region of the gene, but Interestingly, a TATA-like sequence also follows. This region is the promoter There is no evidence that it can function as such. BreTGF-β, gene sequence stopped Contains the hexanucleotide AATAAA approximately 500 nucleotides downstream from the codon. I'm here. This sequence (32 ), probably BreTGF-β, which functions as a polyadenylation signal for mRNA. (This is the rough TG calculated from Northern hybridization.) Since it matches the size of F-β mRNA, it is also possible that the 3' untranslated region is the intervening sequence. ). B enoist etc. (43) is AA The consensus sequence TTCA follows closely TAAA and immediately precedes the polyA tail. We proposed CTGC. Similar sequence T T CA G G CCb<BreTGF- Following the AATAAA sequence in the 3° untranslated region of βImRN　A, position 2530 provided further support for the assignment of the polyadenylation site.

BreTGF-β1 is a 390 amino acid polypeptide. This array was previously determined. TGF-β1 was determined by comparing the determined NH,-terminus of mature TGF-β1 was found to constitute the C-terminal 112 amino acids of BreTGF-β. . The mature TGF-β1 monomer is derived from the precursor immediately at the NH,-terminus of mature TGF-β. It is cleaved at the preceding A rg-A rg dipeptide. similar dibasic The cleavage site for mature TGF-β is located immediately upstream from the amino terminus. This way The hydrolytic cleavage site of eel tongue/sj is preproenkephalin (44,4 5), calcitonin precursor (46), and flutecotropin-β-ribotropy It has been found in several other polypeptide precursor sequences, including the polypeptide precursor (47). ing. Hydrophobic profiling by the method of Kyte and Doolittle (48) Measurement of the hydrophilic region that makes it accessible to trypsin-like peptidases It is predicted that the A rg-A rg sequence is located within. After translation of the precursor Cleavage produces mature TGF-β monomer. The nature of the blur array is unknown. However, it may yield other biologically active peptides. TGF-β1 and TGF-β.

The precursor contains several pairs of basic residues (Figures 1b and 5), It can also undergo post-translational cleavage, producing another polypeptide substance. mature TGF-β1 apparently contains two A rg-L ys dipeps that cannot be cleaved. Contains chido. As shown in Figure 1b, the pre-TGF-β1 precursor is Two possible N-glycosylation sites, A5n-X-Set or Thr (Figure 1b).

None of these are located within mature TGF-β1. Therefore, fixed Adsorb glycoproteins to lectin and elute TGF-β in the unadsorbed fraction. provides a method for purifying mature TGF-β free from glycoproteins. be done.

The sequence of human TGF-β was determined by direct amino acid sequence analysis and Determined by extraction from β cDNA. Obtained by clostripain digestion The sequence of different TGF-β peptides was determined by incorrect amino acids in the sequencing. The results match the cDNA sequence, except for a few residues that are presumed to be due to the assignment of Ru. The 112 amino acid TGF-β sequence contains nine cysteines, while , the remainder of the precursors contains only two (Figures 1b and 5). So far Studies have shown that by reducing the 25 kd TGF-βI dimer, 12.5 It has been shown that two polypeptide chains of kd are produced (15). TGF- TGF-β obtained after amino-terminal sequencing of β and clostribin digestion Sequencing of one peptide indicates that the TGF-β2R body consists of two identical polypeptides. This strongly suggests that In addition, the properties of this homo 2j1 body are similar to human genomic DNA. Hybridization during southern hybridization of TGF-β exon probe It is also supported by the presence of only one DNA fragment that soybeans.

Chou-F asman analysis of the secondary structure (5o) of the TGF-β polypeptide has extensive β-sheet properties with little, if any, α-helical Indicates that there is no The region immediately in front of the basic dipeptide cleavage site is the α-helical region. This is thought to be due to the steric configuration.

For purposes of the present invention, pre-TGF-β is defined as shown in FIGS. 1b and 5. The normal TGF-β precursor as well as the pre-sequence normally associated with TGF-β. is defined as another precursor form of TGF-β. These latter forms should be considered to be an insertion mutant of the DNA encoding mature TGF-β. It is possible. These mutants are usually in the form of fusions with mature TGF-β. Sequences may be associated with other secreted proteins, e.g. pre-growth hormone, pre-proinsulin. , viral envelope proteins, interferons and mammalian host cells Preferably, it is obtained from a yeast or bacterial presequence recognized by. these The sequence of the secretory leader can be used when it is not desired to synthesize the DNA in vitro. is known to be a suitable source of DNA encoding it. this control of DNA containing the desired signal and pre-TGF-β DNA. It is ligated to the DNA encoding mature TGF-β by restriction enzyme digestion. Only restriction site (linker) and, if necessary, any residue removed during restriction enzyme digestion. necessary to complete the presequence and coding sequence of mature TGF-β. A synthetic oligonucleotide is prepared to introduce the NA fragment. then , this synthesized linker and/or fragment is linked with a substitution signal and T Ligate and clone the restriction enzyme digested fragment containing the GF-β coding region. and the vector is used to transform a bacterial host. child The mutant presequence is then cloned into an expression vector and used to infect host cells. Transform. For illustration using the viral envelope protein presequence Examples will be described later.

The DNA encoding the complete heterologous presequence was added to the first sequence of TGF-β DNA. It is best to combine the Alternatively, encoding the mature TGF-β coding sequence DNA encoding the complete heterologous presequence and the mature heterologous protein. bind to a short portion (e.g., 21 to 45 base pairs) encoding the protein; As a result, it is useful as an immunogen or by cleavage to produce mature TGF-β A fusion capable of Inserting a collagenase cleavage site between the end and the C-terminus of the heterologous protein fragment By). The purpose of these constructions was to synthesize the natural TGF-β secretory leader. The alternative is to replace a highly efficient secretion system. However, this means that TGF- Secretion of TGF-β is required to produce β in recombinant culture. Not at all.

Other deletion-insertion mutants include viral proteins that are abundantly expressed within the cell. proteins, such as retrovirus core proteins and the large T antigen derived from 5v40. or immunogenic bacterial proteins or polypeptides, e.g. chemotactic to polypeptides, especially the potent chemotactic tripeptide Met-L eu-P he- including the binding of mature TGF-β species.

Mutations in expressed pre-TGF-β, mature TGF-β or fragments thereof As the number and extent of insertions, deletions and substitutions increases It will show an amino acid sequence that deviates little by little from the sequence shown in FIG. This deviation It is measured as the decrease in homology between reTGF-β and the mutant. Immobilization of TGF-β Any protein or polypeptide that exhibits biological activity that promotes independent growth proteins of this invention regardless of the degree of homology they show to the proteins in Figure 1. Included within the scope of light. The reason for this is that some regions of pre-TGF-β (e.g. sequence) is easily mutated, or completely intact as in the case of mature TGF-β. It may even be deleted, thus preserving biological activity. other On the other hand, the nine cysteine residues (and associated disulfonyl residues) in the mature TGF-β molecule Deletion of the peptide bond) has a substantially negative effect on this biological activity, and You may even lose it completely. Furthermore, substitution mutants are functional When residues containing amino acid side chains similar to are substituted, the complete TGF- Although it exhibits growth promoting activity of β, the homology may be small. Functionally similar means that the side the dominant nature of the chain, e.g. hydrophobic, basic, neutral or acidic, or steric Denotes the presence or absence of bulk. Therefore, if a polypeptide is preT The degree of homology it has with GF-β is the main reason for its uniqueness as TGF-β. It's not a perfect scale. However, as a general indicator, mature TGF- which shares at least some biological activity with β and is substantially the same as the sequence of Figure 1b. Within the scope of the term TGF-β are proteins or polypeptides homologous to (e.g. approximately 40% to 100% homologous to pre-TGF-β) a TGF-β variant or any fragment thereof of about 20 residues or more; , mature TGF~β, a variant having an amino acid sequence that is about 75% or more homologous to the sequence including). Regarding its activity in inhibiting the proliferation of neoplastic cells, TGF-β has been Polypeptides known to exert growth-inhibiting activity such as interferon, tumor necrosis factor, and lymphotoxin), but other is not necessarily required to have a region homologous to the sequence in Figure 1b. stomach.

Narrower specific factors in confirming the uniqueness of the polypeptide as TGF-β (a) of mature TGF-β that also substantially neutralizes the activity of the polypeptide in question; Antibiotics that can substantially neutralize growth-inhibiting or anchorage-independent growth-promoting activity. (b) the ability of the serum to compete with TGF-β for TGF-β cell surface receptors; ability of the co-polypeptide. However, immunological and proliferation-promoting uniqueness It is understood that gender is not necessarily the same. Figure 1b Neutralization of mature TGF-β This neutralizing antibody is essential for the growth-promoting activity of TGF-β. It may not be directed to the site, so it may not bind to the candidate protein. . Instead, the antibody binds to a harmless region and acts to neutralize it through steric hindrance. It may be useful. Therefore, candidates mutated in this innocuous region The protein no longer binds to this neutralizing antibody, but still has substantial homology. It may be TGF-β in terms of sex and biological activity.

Mutants thought to be biologically active antagonists to TGF-β The TGF-β residues subjected to site-directed mutagenesis for the preparation of cysteine Residues, Arg18, Lys19, Leu20, Tyr21, I1e22, Ph e24, Leu28, G1y29, Trp30, Trp32, I] e33, P ro36, G1y38, Tyr39, Asn42, G1y46, Pro49, L eu62, Tyr65, Pro70, Va179, Pro80, Leu83, L eu86, Ile89, Va19Q, Tyr91, Tyr92, Leu102 , Asn105, Met106.11e107 and Val108. usually , the substitutions T made at these residues are non-conservative: i.e., the substituted residues are , (a) substantially different in hydrophobicity, e.g. hydrophobic residues ff al, I le, L, eu, Phe or Met) are hydrophilic residues (Arg, Lys %Trp or Asn) or a hydrophilic residue. Groups (Tbr, Ser, His, Gin, Asn5Lys, Asp, G (b) whether lu or Trp) is substituted for a hydrophobic residue; Their effects on the orientation of the peptide backbone are substantially different, e.g. or Gly is substituted by or in place of another residue; ( C) have substantially different charges, e.g. negatively charged residues (Glu or Asp, etc.) are positively charged residues (L　yss　Hr　s or Arg, etc.) ) (and vice versa); or (d) steric bulk. are substantially different, for example bulky residues (His, Trp, Phe or Tyr) has a small side chain (Ala, Gly or S etc.) in place of (and vice versa). In addition, the above target residues each (preferably in pairs) or non-conserved residues (also preferred). (preferably in pairs) are inserted adjacent to the target residue. Usually only one residue at a time Introduce sequence mutations. The region for site-directed mutation studies is from residues 105 to 112.77-95, and 20-49.

Identification of antagonists is routine. EGF-enhanced anchorage-independent target cells candidate with equimolar amounts of TGF-β that can be detected by other methods in proliferation assays (81). Incubate the cultures for fixation-independent growth without inducing fixation-independent growth. Observe.

Antagonists that remain immunologically cross-reactive with natural TGF-β Reagents as standards in assays or, when labeled, in competitive assays It is useful as Antagonists may also be used in the treatment of, e.g., TGF-β-dependent tumors. It is also useful in treating tumors.

Additionally, the same residues targeted for site-directed mutagenesis during the preparation of candidate antagonists is a target for the preparation of agonists. However, here there is no replacement or insertion. The mutated residues that are introduced are conservative; i.e., hydrophobic, electronegative, as described above, etc. Which respective group members are substituted for each other or adjacent to members of the same group inserted (again preferably in pairs).

Also, TGs that are a fusion of one TGF-β allele or subtype with another F-β variants, e.g. TGF-β, with a substituted homologous N-terminal domain. the C-terminal domain from TGF~β1, or some allele or subtype. domain derived from a subtype is replaced with a homologous region derived from another allele or subtype. Substitutional variants that are within the scope of this invention.

Natural or wild type mature TGF-β of Figure 1b obtained from placenta or platelets. The illustrated mutants have characteristics (such as molecular weight) of the natural TGF-β species. However, the properties of TGF-β can be changed considerably, and will be explained in more detail later. In fact, this may actually be the purpose of mutagenesis. Defined herein TGF-β includes natural TGF-β1, but other related biological Polypeptides active in this definition fall within this definition. Insertion mutants and deletion mutants mentioned above. TGF-β species, such as natural mutants or fusion proteins, are derived from natural TGF-β. The molecular weight will probably deviate from the molecular weight confirmed by the experiment. For example, a fusion protein with mature TGF-β The protein or protein TGF-β itself is larger than native mature TGF-β. deletion mutants of mature TGF-β have lower molecular weights. It will have a quantity. Similarly, glycosylation can be achieved by introducing glycosylation sites. in order to obtain TGF-β, or at sites not essential for biological activity. TGF-β is engineered to substitute serine for serine. Finally, non-primates Post-translational processing of human brain TGF-β in mammalian-derived cell lines generates slight heterogeneity in the amino-terminal region of mature TGF-β, and no longer contains alanine. It will not be the amino terminal amino acid.

In the definition of biological activity, it is defined as being able to induce anchorage-independent growth. The term BreTGF-β or its fragments can be obtained by Polypeptide fragments that can be converted into polypeptide fragments that exhibit the desired biological activity. It should be noted that lipeptides are meant to be included. General Usually, the inactive precursor is the mature TGF-β that binds its carboxylic acid via peptide bonds. A fusion protein that is attached at its terminus to an insoluble or gelatin-like protein. It must be of good quality. The sequence at or within this peptide bond region is protein selected to be sensitive to hydrolysis, whereby TGF-β is as generated in situ or as part of an in vitro manufacturing protocol released.

Usually, TGF-β means TGF-β, but the biological activity is not explained above. TG from sources such as pigs, pigs, horses or cows, as long as they meet the established criteria. F-β is included within the definition of TGF-β. TGF-β is not disseminated.

For example, Fuzumi and human TGF-β are both anchorage-independent in the same cell line. It is effective against germs that induce proliferation. Therefore, TGF-β from one species can be used to treat other species. It can be used for medical treatment. The DNA encoding TGF-β of other species is Human pre-TGF labeled genomic library or cDNA from species such as -obtained by probing with β cDNA.

Derivatives of TGF-β are included within the scope of this invention. Derivatives include other TGF-β Covalent or collective conjugation with molecules, 2ffi entities or unrelated chemical moieties Includes glycosylated and glycosylated forms. Covalent derivatives are known in the art. Depending on the method, at the N- or C-terminus, or at the amino acid of TGF-β They are prepared by attaching functional groups to groups found in the chain. these Derivatives may include, for example, carboxy termini, alkyl amines or carboxy side chains. aliphatic or acyl esters or amides of residues containing (e.g. asparagine) Funschgate of acid residues to alkyl amines); 0-acyl derivatives and the N- of amino terminal amino acids or amino group-containing residues. Acyl derivatives (e.g. fMet-Leu-Phe or immunogenic proteins) Conjugates with macrobiotic substances) are included. Derivatives of acyl group include n of 03 to C10. - alkyl groups containing alkyl (thereby forming an alkanoyl species), and and carbocyclic or heterocyclic compounds (thereby forming aroyl species) selected from the group consisting of. Reactive groups bind proteins via reactive side groups. It is a known bifunctional compound used for crosslinking insoluble matrices. It is preferable to

Covalent or aggregated derivatives can be used as reagents in immunoassays or for affinity Useful in purification methods. For example, purification of anti-TGF-β antibodies or cell surface receptors Alternatively, for use in the assay, TGF-β can be prepared using cyanobromide by a method known per se. be insolubilized by covalently binding to activated sepharose, or Adsorbed onto polyolefin surfaces (with or without glutaraldehyde crosslinking) without). TGF-β can also be labeled with detectable groups: e.g., in diagnostic assays. For use in biological samples, TGF-β can be determined in biological samples, especially by competitive immunoassays. To diagnose levels, radioiodized or diluted Covalently linked to earth chelates or conjugated to other fluorescent moieties .

Usually, TGF-β mutants are produced in a predetermined manner, i.e. in a site-specific manner. It is prepared by The purpose of site-directed mutagenesis is to induce TGF-β, as described above. That is, the aim is to construct DNA encoding TGF-β that exhibits biological activity. Ru.

Although the mutation site is predetermined, the mutation itself does not need to be predetermined. . For example, to optimize the effect of a mutation at a certain site, it is necessary to Metainduction is performed at the target codon and the expressed TGF-β mutant is optimized for activity. Screening will be carried out accordingly. At a predetermined site in DNA with a known sequence Methods for making substitution mutations are well known (eg, M13 primer mutagenesis).

(Hereafter, margin) TGF-β mutagenesis usually involves amino acids on the order of about 1-5 amino acid residues. or by deleting about 1 to 10 residues. replace, delete, Insertions or any combination thereof are performed to obtain the final construct. As above, insert For example, amino or carboxy termini with hydrophobic or immunogenic proteins. Contains fusions. Mutations in the DNA that harbor such mutants ultimately does not place this sequence out of reading frame in the expression vector. The resulting protein should contain biologically active T. It does not become GF-β. This mutation also affects the secondary mRNA structure that suppresses translation. Preferably, it does not create regions of complementarity that could generate structures.

A heterozygous 2j1 form of TGF-β is included in the present invention. These usually include one T GF-β chains are from one subclass (e.g., TGF-β,) and the other is from another subclass. 2M bodies derived from subclasses (eg, TGF-βS) are included. Similarly, T GF-β amino acid sequence variants may be combined with other amino acid sequence variants or with natural T It is prepared as a homodimer or heterodimer with the GF-β sequence. 2 amounts of hetero The body has a first TGF- a dimer containing a β sequence and a second TGF-β sequence that is not biologically active. Ru. Such heterodimers would be active as TGF-β antagonists.

The heterodimer synchronizes host cells with DNA encoding the selected TGF-β chain. It is easily prepared by time-transformation. TGF produced in this way- A portion of β is expressed as the desired heterodimer, producing the highest proportion of heterodimers. Screen the producing transformants.

The DNA encoding TGF-β is produced by chemical synthesis in the placenta or other places. or by screening reverse transcripts of mRNA from cells of It is obtained by screening a genomic library derived from nuclear cells.

This DNA can be used as long as the host cell recognizes the codons used in Figure 1b or does not require the use of the codons listed in Figures 4a-C. This kind of DNA is , similar to the DNA of Figures 1b or 4a-c, are easily produced in vitro. . Also useful in the present invention is that TGF-β as defined herein encodes R, but still capable of hybridizing with such NA or RNA. A nucleic acid that is either NA or DNA. Not coded but hybrid Nucleic acids that generate TGGF-β are not used for recombinant synthesis of TGF-β, but TGGF-β in test cells is Preparing labeled probes for F-β mRNA or genomic DNA diagnostic assays It is useful as an intermediate for manufacturing.

Diagnostic nucleic acids contain substances detectable by methods known per se, e.g. fluorescent groups, radioactive Covalently labeled with an active atom or chemiluminescent group. Next, this used in conventional Southern or Northern hybridization assays. this Such assays can be performed using TGF-β vectors and transformation as described in the Examples below. TGF-βa+RN in tissues during body identification or as a measure of mitogenic activity It is used for in vitro diagnosis such as detecting A.

In the present invention, TGF-β is a vector containing DNA encoding TGF-β. synthesized in host cells transformed with vector. A vector is a replicable nucleic acid structure. It is a construction. Amplify and amplify the DNA encoding TGF-β using a vector. /or DNA or RNA expressing or encoding TGF-β Amplify the DNA that hybridizes with A. The expression vector encodes TGF-β. The coded DNA sequence is a suitable vector capable of expressing TGF-β in a suitable host. A replicable NA construct operably linked to a troll sequence. like this The funtrol sequence contains a transcription promoter, a place to control transcription. an optional operator sequence encoding an appropriate mRNA liposome binding site; sequences that control transcription and translation termination. Ma In addition, in order to express TGF-β in eukaryotic cells, the vector carries the selected gene. It should contain the coded DNA. However, this selection gene It can be supplied by the unbound plasmid in transformation.

Vectors include plasmids, viruses (including phages), and integrative vectors. NA fragment sites, i.e., fragments that can be integrated into the host genome by recombination. Contains As used herein, a vector is defined as a vector that is cloned in a bacterial cell. Although it is a plasmid in the sense that it can be transformed into be incorporated into.

However, all other methods that perform equivalent functions and are or become known in the art vectors are suitable for use in the present invention. A suitable vector is Contains replicon and control sequences derived from species compatible with the expression host I'm sure you're doing it.

DNA regions are functionally linked when they are functionally related to each other. Ru. For example, the presequence or secretory leader DNA is If expressed as a preprotein involved in secretion, the DNA of the polypeptide operably linked; a promoter controls the transcription of a sequence If so, it is operably linked to the coding sequence; the liposome binding site is operably coupled to a cryptographic sequence if it is configured to cause usually , functionally linked means adjacent, and in the case of secretory leaders means adjacent and in the leading phase.

Cell cultures derived from multicellular organisms are preferred host cells in the present invention. principle For example, whether the source is derived from a vertebrate culture or a non-vertebrate culture, Any higher eukaryotic cell culture can be used. However, vertebrate cells are the most important The growth of vertebrate cells in culture itself has become a routine method in recent years [ Ti5sue Culture, Academic Press, Kruse and Patterson ed. (1973)]. Useful host cell lines Examples include VERO and HeLa cells, Chinese hamster ovary (CHO ) cell line, as well as W138, BHK.

COS-7,293 and MDCK cell lines.

A number of eukaryotic cells are known to synthesize endogenous TGF-β. i.e. many possible Certain host cells synthesize TGF-β of the host species.

Therefore, this TGF-β is the TGF-β produced by transcription and translation of the transforming DNA. Present in GF-β. For example, Nomster TGF-β is transformed into CHO cells. cell extracts containing human TGF-β present in the cells and therefore collected from such cells. Exist in the product. TGF-β in animals and humans is active across species lines; Although the above is not necessarily inconvenient, it is necessary to minimize the amount of endogenous animal TGF-β. It is desirable to choose a host/vector system that secretes only This is due to (a) low level (b) select a host animal cell line that synthesizes animal TGF-β; Transforming an animal cell line with the vector (described above) for secretion of TGF-β at (c) recover human TGF-β from the culture medium; By transforming human cell lines, raw hTGF-β is not contaminated. is achieved.

It is desirable to use host cell lines that differentiate to synthesize endogenous TGF-β. Sometimes it's good. Examples include megakaryoblasts, prokaryocytes or basophil megakaryocytes. Includes in. When suitable cell lines are not available, these can be used in mammalian EBV perpetuation of megakaryoblasts, prokaryocytes or basophil megakaryocytes recovered from the bone marrow of animals It can be prepared by Immunization using antibodies specific for host TGF-β Desired species are isolated from transformed cell cultures by affinity chromatography. of TGF-β is recovered.

Typically, expression vectors for such cells include an origin of replication (for extrachromosomal amplification), a TG A promoter placed upstream of the F-β coding sequence (and an enhancer if desired) ), RNA splice site (intron-containing TGF-β encoding when using genomic DNA), and the polya located 3′ of the TGF-β sequence. Contains a transcription termination sequence including a denylation site.

Transcription and translation controls in expression vectors for transformation in vertebrate cells The virus sequence is preferably obtained from a viral source, e.g. The promoters used are polyoma, adenovirus 2, and most preferably monkeys. Virus 40 (SV40) Power 1 et al. 1 question. SV40 early and late promos This is particularly useful because both of them are linked to the SV40 virus origin of replication. This is because it is easily obtained from this virus as a fragment containing 4).

In addition, from the Hindl11 site to the BglI site located in the virus replication origin. If it contains a sequence of about 250 bp extending with 0, then tJsrgLX force), Alternatively, larger SV40 fragments can be used.

TGF-β appears to be toxic to mammalian cell transformants and therefore Inducible promoters, e.g. metallothionein promoter, Drosophila fever promoter, Alternatively, the yield may be improved by using the mouse mammary tumor virus promoter. stomach. Furthermore, the TGF-β genomic promoter, a control normally associated with TGF-β, Roles and/or signal sequences may also be used (such controls provided that the sequence is compatible with and recognized by the host cell). TGF-β When an untranslated region is included in the expression vector, immediately 5' to the start codon. Substituting an A or T base in place of the G or C base on the side, resulting in 3' untranslation of the cDNA. Yields can be improved by deleting G-C rich domains in the sequence. Wear.

Origins of replication can be created by constructing the vector to include a foreign origin, e.g. SV40 or other viruses (e.g., polyoma, adenovirus, VSV , or BPV) and can also be obtained by deriving from the host It can be obtained by the cell's chromosome replication mechanism. vector into host cell chromosome When incorporated, the latter is often sufficient.

Instead of using a vector containing a viral origin of replication, a selection marker and T Transforming mammalian cells by cotransformation with GF-β DNA I can do it. Examples of suitable selectable markers are dihydrofolate reductase (DHFR) or is thymidine kinase.

Such markers are proteins, usually transformed cells, i.e. foreign DNA It is an enzyme that allows the identification of cells that were competent for the uptake of . General Identification is usually done in culture media that are toxic or do not incorporate marker proteins. Survival of transformants in culture media where cells cannot obtain essential nutrients I am by.

A vector containing DNA sequences encoding both TGF-β and DHFR When selecting a preferred host mammalian cell for transfection by It is appropriate to choose the host according to the type of DHFR protein used. field When native DHFR protein is used, it is DHFR deficient and therefore Successful testing in selective media lacking hyboxanthin, glycine, and thymidine Enables use of DHFR coding sequence as marker for transfection Preferably, a host cell is selected that is compatible with the host cell. A suitable host cell in this case is Url DHFR, prepared and grown as described by aub and Chasin (55). A Chinese Hamster Ovary (CHO) cell line that lacks activity.

On the other hand, the DHFR protein, which has a low binding affinity for methotrexate (MTX), When using DNA encoding a protein as a control sequence, D It is not necessary to use HFR resistant cells. Mutant DHFR has no effect on MTX. resistant to MTX, provided that the host cell itself is sensitive to MTX. In this case, MTX-containing media can be used as a selection means. Absorbing MTX The majority of eukaryotic cells that can be methotrexate appear to be sensitive to methotrexate. This way One such useful cell line is the CHO line, CHO-K 1 (ATCC No. , CCL61). According to another method, DHFR” host cells are This cell is then infected with DNA encoding the tumor resistance genes, DHFR and TGF-β. Used by co-transforming. The first transfection is neoma Screen for icin resistance and resistant transformants are then amplified on MTX. do.

Other methods suitable for applying recombination to the synthesis of TGF-β in vertebrate cell culture The method is described by G et al. (56), Mantei et al. (57), and L e Vinson et al. (58, 59).

TGF-β was recovered from lysed transformed cells, and insoluble cell debris was separated by centrifugation. Let go. According to an alternative method, culture supernatants derived from transformed cells secreting TGF-β are Separate from cells simply by centrifugation.

It is then treated in the presence of an acid, usually by methods known in the art (15, 16, 17). using gel filtration followed by HPLC and elution with an acetonitrile gradient. Purify GF-β. However, such methods are not always necessary for the preparation of therapeutic products. That doesn't mean it's necessary.

An alternative purification step involves denaturing and precipitating impure proteins. Temperatures and periods that are sufficient for cell lysis but do not denature and precipitate TGF-β Heat the lysate or supernatant; TGF-β probably has extensive disulfide bonds. As a result of the formation of bonds, it is a protein with remarkable thermal stability. Therefore, heating Should be performed in a medium containing small amounts of disulfide reagents such as threitol .

In addition, since TGF-β is known to be stable in 1M acetic acid, heating should be done in an acidic environment. Combine with .

Natural mature TGF-β is not glycosylated. Therefore, this lectin Glycoproteins are adsorbed onto a column, such as Sepharose bound to lentil lectin. from any remaining impure heat- and acid-stable glycoproteins. Separated. Although less preferred, this step may be performed before heat and acid treatment. Can be done.

TGF-β will elute with the unadsorbed fraction. Glycosylated by host By transforming host cells with the 'I' G F-β mutant, endogenous assembled from host cells expressing TGF-β (or undesirable homodimers). Collect recombinant TGF-β. Sugar r tagJ has lectin affinity chromatography, elution of glycosylated TGF-β, and if desired Recombinant TGF-β is transformed into endogenous TGF-β by removal of sugar residues by enzymatic digestion of This makes it possible for the materials to be collected without containing them.

When a highly pure product is desired, the crude or partially purified mixture is then The compound can be used for chromatographic focusing.

TGF-β of the desired purity in a physiologically acceptable carrier, i.e. the dosage and amount used. TGF-β can be administered by mixing it with a carrier that is non-toxic to the recipient at moderate concentrations. Prepare for use. Typically, this includes buffering TGF-β, antioxidants (ascorbic acids, etc.), low molecular weight (approximately 1° residue or less) polypeptides, proteins, amino acids, Carbohydrates (including glucose or dextrins), chelating agents (EDTA ), and other excipients. therapeutic administration The TGF-β used for this must be sterile. This is a sterile filtered membrane. This is easily achieved by filtering through a run (0.2 micron). Usually, T.G. F-β is highly stable against heat and oxidative modification and is therefore stored in aqueous solution.

TGF-79i;! , usually concurrently pending (7) U, S, S, N, 500. As described in more detail in 833, TGF-α or EGF The compound is mixed with an activated drug, such as a seed, and administered as described in that application.

Various therapeutic targets for TGF-β compositions are known.

A first and preferred subject is exposed to promote cut or wound healing. Topical application to tissues. Types of wounds or other trauma that can be treated There are no restrictions on these, including (but not limited to) the following: burns (not 1st degree, 2nd degree) and 3rd degree burns (especially 2nd and 3rd degree burns) ); epidermal and internal surgical cuts (including orthopedic); lacerations, cuts and wounds including pressure ulcers (bedsores), diabetes, dental, hemophilia, and varicose veins; Epidermal ulcer. Doses previously described for wound treatment (17) were It would be suitable as a starting dose in elephants.

The TGF-β composition is in the form of a sterile MFC solution, preferably a physiological saline solution. Mixed or in the form of an ointment or suspension, preferably with purified collagen. Also applied to burns. The composition is also preferably liquid or semi-liquid. It can also be impregnated into transdermal pansoukola, salves and bandages in the form of . Containing self-microbial agents such as silver sulfadiazine in such articles or compositions should be allowed. In addition, necrosis silk removal agents such as proteolytic enzymes are TGF-β mutations that do not hydrolyze TGF-β or are resistant to hydrolysis It can be included when the body is in use.

Additionally, TGF-β is administered systemically for the treatment of wounds and similar trauma. whole body Administration may have undesirable side effects, such as stimulating neoplastic cell proliferation in cancer patients. Useful when resources are absent or limited. TGF-β composition for systemic administration Preferably, the product is formulated as a sterile isotonic parenteral injection or infusion solution.

Activated drugs administered with TGF-β (TGF-α, EGF or other growth factors) of TGF-β present in the activated composition administered to the recipient. is directly dependent on the amount, growth factor selected, and clinical status of the patient.

The initial dose of TGF-β is administered to the treatment area at a concentration of approximately 0.1 to 150 n9/d. and then adjusted according to clinical trials.

Since the TGF-β composition stimulates and promotes cell regeneration, the continuation of this composition application or periodic reapplication. Dosage determined by clinician according to clinical trials will change.

To simplify the examples, some frequently occurring methods are described in abbreviated terms.

Plasmids are identified by a lowercase p followed by and/or after an uppercase letter and/or is shown in numbers. The starting plasmids of the invention are commercially available or is available to anyone from a source other than the It can be constructed from lasmid according to known methods. In addition, other equivalent Although sumids are known in the art, these will be apparent to those skilled in the art.

DNA digestion is the process by which DNA is digested by enzymes that act only at certain locations in the DNA. Refers to the catalytic cleavage of Such enzymes are called restriction enzymes, and each The site for which is specific is called a restriction site. Various restriction enzymes used in the present invention are commercially available. Reaction conditions for these enzymes that are commercially available and established by the enzyme supplier , cofactors and other necessary things were used. Usually, restriction enzymes are Capital letter followed by another letter representing the microorganism from which the restriction enzyme was first obtained. It is indicated by an abbreviation consisting of numbers representing the specific enzyme. Usually about 1μ9 plasmid DNA fragments are combined with about 2 units of enzyme in about 20 pQ of buffer solution. used for Appropriate buffers and substrate amounts for specific restriction enzymes are determined by the manufacturer. Directed. Typically, incubation times of approximately 1 hour at 37°C were used; It may be changed according to the supplier's instructions. After incubation, filter the protein. Digested nucleic acids are removed by ethanol and chloroform extraction. It is recovered from the aqueous fraction by precipitation. Although it is not always done, digestion with restriction enzymes is This is followed by bacterial alkaline phosphatase hydrolysis of the terminal 5' phosphate. When the two restriction ends of a DNA fragment "circularize," or close, forming a circle (preventing insertion of another DNA fragment at the restriction site) prevent. Unless otherwise noted, digestion of the plasmid is followed by removal of the 5” end. No phosphorylation. Dephosphorylation methods and reagents are conventional (85).

"Recovery" or "isolation" of a DNA fragment from a restriction digest refers to Separation of digests on polyacrylamide or agarose gels by electrophoresis and their transfer. Determination of the desired fragment by comparing the mobility of marker DNA fragments with known molecular weights. identification of fragments, cutting out gel sections containing the desired fragments, and Refers to gel separation from DNA. This method is widely known (86.87 ).

"Southern analysis" refers to the determination of the presence of DNA sequences in a digest or DNA-containing composition. , hybridization to known labeled oligonucleotides or DNA fragments. This method is confirmed by dization. Unless otherwise specified, the present invention For Southern analysis, separation of digested product in 1% agarose, E, S outhe Transfer and modification to nitrocellulose by the method of rn (68) and T, Man hybridization as described by Iatis et al. (88).

"Transformation" means that DNA can be replicated as an extrachromosomal element or chromosomal integrant. It means introducing DNA into an organism, as in Unless otherwise noted The method used for the transformation of E. coli in the present invention is described by Mandel et al. 89) CaCQ, law.

“Ligation” refers to the phosphorylation between two double-stranded nucleic acid fragments. Refers to the process of forming diester bonds (T. Maniatis et al., ibid. 146 pages). Unless otherwise noted, ligations were performed using known buffers and conditions. , T for approximately equimolar amounts of the DNA fragments to be ligated l(), 5 μ9 It can be performed using 10 units of 4 DNA ligase ("ligase").

"Preparation" of DNA from transformants refers to the preparation of plasmid DNA from a microbial culture. means to isolate. Unless otherwise noted, Maniatis et al. , p. 90) may be used.

An “oligonucleotide” is a short single-stranded or double-stranded polydeokine nucleotide. which is chemically synthesized by known methods and then polyacrylic Purify on midgel.

Example 1 Purification and sequence analysis of human TGF-β A Known purification methods can be scaled up and modified to provide sufficient capacity for amino acid sequencing. Homogeneous purified human TGF-β1 was obtained. Warin human platelets (250 units) Extracted with acid-ethanol (IQ) in a g blender. Add ether (4Q) A precipitate is formed, which is vacuum filtered using Whatman No. 1 filter paper. collected by. This precipitate was dissolved in 1M acetic acid (50fff2) overnight and 1M Gage of Biogel P-60 column (10xlOOcx) equilibrated with acetic acid. Purified by filtration.

Fractions containing TGF-β1 were subjected to analytical 5DS-polyacrylamide gel electrophoresis. Identified by dynamic and bioassays (15). Peak fractions were collected, lyophilized, and diluted with 1M vinegar. acid, 8M urea (redissolved in 20x □, followed by Bio in 1M acetic acid, 8M urea) By gel filtration with a gel P-60 column (5X 90 cm), approximately TGF-β with 0% purity was obtained. These peak fractions were then diluted with 1 volume of water. Dilute and semi-preparative RF P C18 (S ynchropak) HP L C column, 20-50% acetonitrile Elution was performed with a Ryl gradient.

When the TGF-β thus obtained was quantified by amino acid analysis, it was found that The yield was found to be approximately 0.5 Rg. Modified 5DS-polyacrylamide gel Electrophoresis was performed as described (60).

Consistent with previous studies, non-reduced TGF-β1 was induced by 5DS-polyacrylamide β-mercaptoethanol migrates as a 25 kD protein in the It was converted to a 12.5 kD species upon reduction with fluorine. This means that TGF-β two 12.5 kD polypeptide chains linked by intramolecular disulfide bridges. suggested that it consists of

In order to obtain information on the protein sequence, purified TGF-β1 was reduced and alkyl and subjected to amino-terminal sequence analysis. TGF-β (1,2 nmol) at 8M Dialyzed into urea, Q, 1M) Lis-HCQ (pH 8,5), 10mM dithiotre Itol was reduced by incubation in 8M urea. next archi The conversion was carried out in the dark and at room temperature in the presence of 50mM iodoacetate. excessive The reaction was stopped after 30 min by addition of β-mercaptoethanol and dialysis. Ta. This TGF-β, (0.7 nmol) was used for direct NH,-terminal sequence analysis. . The reduced and alkylated TGF-β1 (1.2 nmol) was added to 0.75 M Urea, 50mM NH, HCO, 1% clost in 5mM dithiothreitol Digested with levine for 24 hours (15). Add 1% crosstrivine at 12 o'clock. Added after the intermediate reaction. The reaction product was dissolved in 0.1% trifluoroacetic acid. Using ~70% acetonitrile gradient, Synchromapak RP Separation was performed on a Cl8 reverse phase column (4,6x250xx). Sequencing was performed using Rai Amino acid induction by reverse phase HPLC on ninM 1crosorb C-8 column Conductor identification allows the gas phase sequencer (A) described by Hevick et al. (62) to iosystems, model 470A) or significantly improved Beckman 890C spinning power.

This was done using either a 3D sequencer (61). Amino of some peptides The acid sequence was determined. One of these fragments is NH.

terminal segment, while another large peptide gives a sequence of 37 amino acids. , which overlapped the N H terminal sequence and presented a contiguous sequence of 60 residues.

In addition, unmodified TGF-β1 was treated with CNBr. For cleavage at methionine residues This results in a complete loss of biological activity, which indicates that this C-terminal showed that at least a portion of the cuta peptide is required for biological activity (data (data not shown).

Example 27 Isolation of GF-β exon The approach used for the initial identification of the nucleotide sequence encoding TGF-β For this purpose, we used the “long probe” method previously used for TGF-α (7). Adopted. Long oligonucleotides designed based on partial protein sequences 7 for the identification of TGF-β1 exons in human genomic DNA libraries. 1 was used as an imbridization probe. Next, this TGF-β1 extract Son was used as a probe for the isolation of TGF-β+ cDNA.

complementary to sequences encoding amino acids 3-17 and 30-44, respectively. Two 44-base long deoxyoligonucleotides βLPI and βLP2 were synthesized. was chemically synthesized (63.64). The selection of the nucleotide sequence is based on human mRNA (26). This is relatively true in animal DNA. Rare CpG dinucleotides were avoided if possible. Furthermore, amino acids 13 to 17 Sixteen 14-mars were synthesized that were complementary to all possible codons. these The deoxyoligonucleotide and corresponding amino acid sequences are shown below.

NHt-A l aLeuAspTh rAs nTy rCys PheS erSerTh rG l uLysAsnCys Cys@Va l Arg G I uLeuTyrLeuG 1yTrpLysTrp l 1eHisG 1uProLy sG 1yTyrHi sA 1aAs nPheCy s　keuG　1yProcys　ProTyrThe nucleotide indicated by the dot is the code This is a base with no ambiguity in the base.

Using 3'P-labeled βLP-1 as a probe, a low stringency high Screening a human genomic DNA library (28) under hybridization conditions I did a ning. Using low stringency conditions (65), human genome fetal liver Approximately 7°5 After replica plate on the base filter (66), 3! P label 44- It was hybridized with merβLP-1. From 58 hybridized phages Prepare DNA and use BamHI? Southern hybridization of the t'lization mixture °2P labeling using the dot-plot analysis method (68) was hybridized with βLP-1 and βLP-2 oligonucleotides. both The two phage DNAs hybridized with the oligonucleotides are digested and with a pool of 32p-labeled 14-mers by hybridization. Blobbed again. Hybridization with 14-mer is performed using 5xSS C, Q, 5% NP40.6mM EDTA, lxDenhardt solution and 50 uti/x (1) in salmon sperm DNA at 37°C. Autoradiograph Several washes were performed at room temperature in 6xSSC before roughy. Phage βλ58 The original DNA contains oligonucleotides βLP-1, βLP-2, and 14-m hybridized with ar pool. Hybrid with βLP-2 and 14-mar The sequences to be synthesized are located within the same 4 and 2 kbp BamHI fragment. On the other hand, probe βLP-1 was conjugated to a 20 kbp BamHI fragment. Breeded. The hybridizing BamHl fragment was transferred to pBR322. subcloned inside. The nucleo of the smaller hybridizing fragment After subcloning the dideoxynucleotide sequence into M13 derivatives (70), It was determined by the otide chain growth termination method (69).

By screening a genomic DNA library, starting from mature residue 10 An exon encoding part of the coding sequence of TGF-β was isolated. perfect T To obtain the coding sequence of GF-β, we used this exon as a probe and Screening of λgtlO-based cDNA library derived from parturition placenta mRNA I clicked.

Example 3 Isolation of TGF-β cDNA Total RNA was extracted from different cell sources (71) and polyadenylated fflRN A fraction was isolated by oligo(dT)-cellulose chromatography (72) did. dT+t-+s or deoxyoligonucleotide ACACGGGTTC cDNA was prepared by priming with AGGTAC 73). This 2 The heavy chain cDNA was digested with nuclease S1 (Miles Laboratories). ), followed by E. coli DNA polymerase I Klenow fragment (Boehri Mannheiw) and asymmetric EcoRI linker (75) (74) except that the need for EcoRI methylase treatment is avoided using Subcloned into EcoRI-cut λgtlo as follows.

Recombinant phages were plated on E. coli C600Hf1 (74), and replicas were plated on Nitrogen. plated on cellulose filters (66). These were mixed with 50% formamide. 5xSSC, 50mM sodium phosphate (pH 6,8), 0.1% pylori sodium phosphate, 5 x D enhardt solution, 50 μ9/min Q salmon sperm D Restriction flag 3ap labeled (76) of the exon of Example 2 at 42°C in NA. Sodize with ment and 0.2x S'S C10, 1% SDS. Washed at the same temperature. When using 3-labeled deoxyoligonucleotides used low stringency hybridization conditions (65). M1 After subcloning into 3 phage derivatives (70), TGF-β, cDNA restriction The nucleotide sequence of the fragment was synthesized using the dideoxyoligonucleotide chain termination method (6). 9). The resulting cDNA is shown in Figure 1a. Ge as a probe λβC1 was extracted from the human placenta cDNA library using the nom exon (Figure 3). isolated. Approximately 750,000 oligo-dT primed placental cDNA clones As a result of screening, one TGF-β cDNA (λβ C1) was isolated. The previously determined partial TGF-β1 sequence is the leading The frame is shown and the sequence encoding the complete TGF-β1 polypeptide is revealed. Or it becomes. This sequence begins with the N-terminal alanine residue and extends to 112 cods. This is followed by a stop codon, which occurs 20 base pairs from the 3° end. be. Next, A172 glioblasts were isolated using the λβCI EcoRI cDNA insert. A tumor cDNA library was screened and λβC3,19 was isolated. λβ P-labeled Kpnl-Kpnl of C3,19 cDNA insert and above Specifically primed H By screening the T1080 fibrosarcoma cDNA library, λβC4, 10, 4, 33 and 4.37 were obtained. Another similar library - λ with a synthetic 40-mar corresponding to the βC4,33 insert and nucleotides 1 to 40. By screening, λβC5°7b was isolated.

Over 70 different oligo (dT) primed cDNA libraries isolated from All TGF-β cDNAs contain at least a few nucleotides in the 3' untranslated region. Since the cloned genomic DNA was not available, we determined the 3" untranslated sequence using the cloned genomic DNA. Established. Hybridization analysis revealed that the 3′ of the λβC1 cDNA insert The terminus was found to be present in genomic DNA phage βλ58. DNA Sequence analysis revealed that the exo encoding the carboxy-terminal portion of TGF-β1 revealed the presence of a stop codon followed by a 3′ untranslated end (first Figure 1b). The AATAAA hexanucleotide sequence (32) is 50 minutes from the stop codon. 0 bp downstream, allowing assignment of the putative polyadenylation site. This is Assuming that it is indeed a polyadenylation signal, TGF-β, +aRNA The calculated size was determined by Northern hybridization experiments (Example 4). This agrees well with the determined length of 2°3 to 2.5 kd. Genome for 3' untranslated ends Oligo(dT) primed placenta and HTI080 using mu DNA probe By another screening of the cDNA library, one hybrid No cDNA phages were identified.

Example 4 Diagnostic method using TGF-β cDNA probe RNA was extracted from liver cancer HEP-G2, Wilms tumor TuWi, glioblastoma A172, bladder Cancer T24, squamous epidermoid carcinoma A431, breast cancer MCF-7, nasopharyngeal carcinoma KB, fibrosarcoma H T1080, Burkitt's lymphoma B-lymphoblast Daudi and RajiST - Recovered from lymphoblast M o l t-4. Peripherally as described (53) Blood lymphocytes (PBL) were prepared and treated with staphylococcal enterotoxin B. and mitogenic induction using phorbol myristate. In this case, after 24 hours RNA was collected. Polyadenylated mRNA A (4 μg) was dissolved in formaldehyde. - Electrophoresis on 1.2% agarose gel29) and nitrocellulose filter (30). Under the high stringency conditions used above, λβC1 A 3′P-labeled (76) EcoRI cDNA insert was used as a probe. . By comparing the positions of 28S and 18S rRNA on the gel, TGF-β A length of 2.3-2.5 kb is suggested for mRNA. Partial mRNA Although degradation cannot be excluded, in some cases relatively small mRNA species may exist. There may be.

TGF-β mRNA is expressed in tumor cells of neuroectodermal origin, such as TuWi (Wil+ ns tumor) and A172 (glioblastoma), as well as cancer cell line T24 bladder cancer , A431 (squamous epidermoid carcinoma), MCF-7 (breast cancer) and KB (nasopharyngeal carcinoma). It was detectable in all human tumor cell lines. c D N A claw The fibrosarcoma-derived cell line selected as the source of mRNA for T1080 contained relatively high levels of TGF-β mRNA. TGF- βmRNA is derived from cell lysates derived from solid tumors of mesodermal, endodermal and ectodermal origin. tumor cell lines of hematopoietic origin, such as Daudi (Burkitt's lymphoma B-lymphoblast), Raji (Burkitt's lymphoma B-lymphoblast) lymphoblasts) and Molt-4 (T-cell leukemia). Ta. TGF-β mRNA is clearly associated with placental and peripheral blood lymphocyte (PBL) mRNA. Since it is also detectable in NA, the presence of TGF-β mRNA is limited to tumor cells. Not determined. The amount of TGF-β mRNA was significantly and significantly increased after mitogenic stimulation of PBL. increased significantly. TGF-β mRNA could not be detected in human liver; However, it was present in the HEP-G2 hepatoma cell line. In all cases, TGF- β mRNA migrated as a seed with an apparent length of 2.3-2.5 kb. Some smell Therefore, relatively small tpRNA species of approximately 1.8 to 1.9 kb may exist. However, this may be due to partial degradation of mRNA.

(Hereafter, margin) Example 5 Recombinant synthesis of TGF-β The plasmid used for recombinant synthesis of TGF-β1 was pM BTE 5. . The following method for preparing this plasmid is the one actually used for its construction. Preferable to complex methods.

p342E (79) was digested with EcoRI and extracted with Escherichia coli DNA polymerase I (Frame). (now fragment) and four dNTPs, digested with Sal+, and 1 (containing the Amp' gene of pB R322) is recovered.

At the same time, p342E was digested with SalI and Hindll, and HBsAg The fragment encoding this is recovered as Fragment 2.

Finally, the SV40 genome was simultaneously digested with Hindll and Hincll, Flag the 596bp fragment containing the SV40 origin and early promoter Collect as ment 3.

Fragments 1.2 and 3 were ligated with a three-component ligation and the ligation mixture was transformed into an E. coli strain. 294 (ATCC31446) for transformation. Untransformed cultures Plate on picillin medium plates and select resistant colonies. Is it a transformant colony? p342E-blunt was recovered.

p342E blunt was simultaneously digested with Hindll and EcoRI, Collect large vector fragments. Replace this fragment with the following sequence: Polylinker to: and this ligation mixture was used to form E. coli ATCC31446 as described above. It was transformed. Recover pCVSV-HBs from ampicillin-resistant transformants .

pCVSV-HBs was simultaneously digested with Hindll and EcoRI, and the vector -Isolate the 18bp H1ndlll-E coRI flag ment will not appear in the gel due to its small size).

p, a plasmid containing DNA encoding the herpes simplex gD protein. gD-DHF R-T runc (European patent application 84.305909.8) was simultaneously digested with 5tul and Hindlll to obtain the signal peptide of herpes simplex. DNA encoding and the N-terminal portion of the mature H3V-1gD protein. A fragment of approximately 760 bp containing the coding region is recovered. European patent application 84 ．． Plasmid pJ2.9 of 305909.8 can be used similarly.

pi3c1 (Figure 1a) was digested with SmaI and BamHI, resulting in a 480bp Collect the fragments. This fragment is the N-terminus of mature TGF-β1. contains the sequence encoding residues 314 through 314, and encodes pre-TGF-β. contains most of the sequences in

p/3C1 was fragmented with BamHI and EcoRI to generate a 270bp fragment. Collect the waste. These two separate digestions of pβC1 were performed using BamHl-EcoRI This was done because the pβC1 fragment contains a SmaI site. This 270 The bp fragment contains sequences encoding the remainder of the TGF-β1 molecule; It extends 20 bp beyond the stop codon.

pCVSV-HBs vector fragments 760, 270 and 48 as described above. Connect with the 0 bp fragment by a four-component ligation. The construct thus obtained ( pCV　VgD) is a hybrid under the control of the SV40 early promoter. coding sequence (pre-TGF-β, g linked in frame to the precursor fragment) Part of D-1 envelope protein and signal peptide of herpes simplex gD-1 ). This hybrid coding sequence then contains the 3′ untranslated sequence and a hepatitis surface antigen polyadenylation signal.

pCVSVgD was digested with EcoRI and blunted with Flenow and 4 dNTPs. and then digested with PstI. In this way, two types of fragments were obtained and a hybrid Code sequence of code and S V 4. Fragment containing Q promoter (fragment Collect component A).

pCVSVgD was digested with BamHI and blunted with Flenow and 4 dNTPs. and then digested with PstI. After these digestions, four types of fragments are obtained. . A fragment containing the pB R322 origin and the Amp' gene (approximately 1900 bp) is recovered as fragment B.

Fragmen) A and B are ligated and this ligation mixture is used to incubate E. coli ATCC31. , 446. Recovering plasmid pMBTE6 from Amp' colony do.

Plasmid pMBTE5 along with plasmid pFD11 (90) was DHFR-deficient. CHO cells (55) were transfected. The latter plasmid carries DHFR and thus make transfected cells resistant to methotrexate. This allows selection of transformants expressing TGF-β. Arayu DHFR- mammalian host cells are suitable for use. Alternatively, any feeding Using mammalian host cells, host cells are immunized with a plasmid encoding neomycin resistance. due to their ability to co-transform cells and grow on neomycin-containing medium. Transformants can be identified.

Transfected CHO cells were cultured in HGT- medium. cells were selected. Cells were grown to full extent in 15 ci diameter plates. Next The cells were cultured in serum-free medium for 48 hours and then collected. Pour the culture medium The presence of TGF-β was determined by soft agar assay as described (78). I checked using the standard method.

Supernatant medium (50i) was lyophilized and supplemented with 4mM HCQ-0.1% bovine serum alcohol. It was dissolved in Bumin (700μQ). This solution and serial 3-fold dilutions (200 μg ) was tested. The number of colonies in soft agar with a diameter of >89 μl was counted. saturation The maximum response (plateau value) obtained in the presence of an amount of TGF-β1 per plate It was about 1500. Less than 5 colonies were obtained in the absence of TGF-β1. It was. A half-maximal response was observed in cells transformed with the negative control plasmid. A 9-fold dilution of the sample obtained from the cell was obtained. By serial dilution of MBTE6 supernatant Calculating from the values obtained, half of the maximum value is obtained when diluted 70 times. I understand.

Cells transformed with MBTE6 and pFDll by serial dilution assay. pFD even before subcloning and selection in MTX-containing medium. 11 alone or pFDll and control plasmid (herpes coding sequence is bacterial 5TII signal peptide. A plasmid similar to pMBTE6 with the exception of Compared to CHO cells, they synthesize approximately 8-10 times more TGF-β per medium IRQ. It is shown that This additional amount of TGF-β is human TGF-β1.

Biologically active TGF-β1 is found in the culture medium and therefore CHO cells cleave BreTGF-β to produce natural mature TG, similar to the way that CHO cells do. It is concluded that it secretes F-β. This conclusion suggests that TGF-β in soft agar, The slope of the concentration dilution curve for both endogenous natural TGF-β and recombinant TGF-β1 therefore, although their affinities for TGF-β receptors are not identical, This is greatly reinforced by showing that they are also similar.

Example 6 Isolation of DNA encoding TGF-β Pig ovary cDNA λ A 1.5x 10” plaque from the library was placed in 5x SSC 120% formamide. Denhardt, 0.1% sodium pyrophosphate, 0, 05 M N a P O4, 0,1% SDS and 50μ9/xQ salmon sperm DN 71 hybridization buffer at pH 6.8 containing A at 42°C overnight for 3. Using a relabeled E. coli insert (1050 bp) of λβC1, Screened under isostatic conditions.

Washing was performed with 2XSSC at 37°C. Positively hybridized plaque The phage were purified from and their DNA inserts sequenced. Approximately 200 positives The cDNA inserts that hybridized to TG were divided into three groups: pig TG; F-β, (2 plaques), to GC that does not encode TGF-β polypeptide Enriched CDNA (6) and encodes porcine TGF-β3 downstream from the mature residue IO. The gene fragment λ11.3゜λ11.3 containing the DNA The labeled EcoRI insert was incubated under high stringency conditions (same as above; only and washed with 50% formamide and Q, 1x ssc at 42 °C. 1xlO' plaques derived from a pig ovary cDNA lambda library were Rescreened. Of the 20 positively hybridizing plaques One (λ10) contained the complete porcine TGF-β3 sequence. total number The nucleotide sequence and the post-assignment amino acid encoded by λ10+11.3 The amino acid sequences are shown in Figures 4a-c.

Pig c labeled with 1×1O” plaques derived from human ovarian eDNA library Screened using DNA. One positive plaque (λhu4) identified did. λhu4 is the nucleotide sequence shown in Figure 4a=c and the amino acid after assignment. It has an array. Figure 5 shows the attribution of pig and human TGF-β3 cDNAs. This is to compare the amino acid sequences of The start codon of the pig precursor candidate is boxed. Surrounded.

TGF-β3 was expressed in recombinant cell culture, and from this nucleotide and amino acid of TGF-β, allowing for deviations in the amino acid sequence (which will be apparent to those skilled in the art). Collected in substantially the same manner as the case. The complete precursor of human TGF-β3 has not been disclosed. To express human TGF-β3, the remaining N-terminal precursor must be Genome or cDNA library for DNA encoding the body sequence DNA encoding available human sequences (residues) can be reprobed or starting from codon 3) to porcine TGF-β, residue 297 of the precursor (shown in Figure 5). (counting as shown) or join the DNA encoding up to DNA containing mature human TGF-β3 and xenobiotic mammalian or viral strains. It would be desirable to prepare DNA encoding a fusion of Gnar.

〆 4005 (197g).

3498 (1980).

5343 (1981).

11, Roberts, A, B, et al, Nature 295: 417-419 (1982).

12. Roberts, A. B. et al., Biochemistr. y22:5692-5698 (1983).

+3. Frolik, C.A. et al., Proc. Natl., Acad. , Sci. USA 80: 3676-3680 (1983).

14, Children, C, B, et al, Proc, Nat l, Acad, Sci, USA 79: 5312|5316 (1982).

15, As5oian, R, et al, J, Biol, C hem, 258: 7155-7160 (1983).

16, As5oian, R, et al, Nature 309 : 804406 (1984).

17, 5porn, M, B, et al, 5science 219 : 1329-1331 (1983).

1g, Frolik, C, A, et al, J, Biol, Ch em, 259: 10995-11000 (1984) B 19.Tucker, R.F., et al., Proc., Natl., Acad. , Sci. LISA 81: 6757-6761 (1984).

20, ^5soian, R, et al, Ce1l 36: 35-41 (1984).

21, Tucker, R, F, et al, Cancer Re s, 43: 1581-1586 (1983).

22, nzano, M, ^, et al, Mo1ec, Ce 11. Biology 5: 242-247 (198T).

23, Roberts, A, B, et al, Proc, Natl, A cad, Sci, USA 82: 119-12328, Lawn, R,M , et al., Ce1l 15: 1157-1174 (1978).

29, Dobner, P. R., et al., Proc. Natl., Aca. d, Sci, USA 78: 2230-2234 (1981).

30, Thoa+as+P, S, + Proc, Natl, Acad, Sci , USA 77: 5201-5205 (1980).

33, Levy, W.P., et al., Proc, Natl, Acad , Sci. USA 78: 6186-6190 (1981).

34, Renderknecht, E, et al, J, Biol, C hem, 259:6790-6797 (1984).

35. No citation.

36, no citation.

37. Batt6y, J, et al, Cedan 34: 779-78 7 (1983).

38, McGrogan, M. et al., J. Biol. Chem. 280(4): 2307-2314 (1985).

39, Valerio, D, et al, EMBOJ, 5(1): 113-119 (1986).

40, McKnight, S, L, et al, Ce1l 37: 2 53-262<1984).

41, Dynan, W, S, et al, Ce1l 35: 79-8 7 (1983).

42, Gidoni, D., Nature 312: 40 9-4I3 (1'#4).

43, Benoist, C, et al, Nucl, Ac1ds Res , 127-142 (1980).

44, Noda, et al, Nature 295: 202-2 06 (1982).

45, Gubler, U. et al., Nature 295: 20 6-208 (1982).

46, Amara+S, G, et al, Nature 298: 2 40-244 (19g2).

47, Nakanishi, S, et al, + Nature 278: 423-427 (1979).

4B, Kyte, J, et al, J, Mo1. Biol, 157: 105105-132 (19°49, Winzler, R, J,, Ho rmonal Proteins and Peptides 1″.

(Li, C,!,, ed,, New York, Academic Pre ss), p, 1-15 (1973).

50, Garnier, J. et al., J. Mo1. Biol, 1 20:97-120 (1978).

51, no citation.

52, no citation.

53, no citation.

'58. A. Levinson et al, EP 117. D60A.

59, A. Levinson et al, EP 117,058Aa6 0, Laemmli, υ, ni,, Nature 227: 680-68 5 (1970).

6], Rodriguez,) 1. et al, + Anal, Bioch Em, 140:53g-547 (1984).

64, Beaucage, S, L, et al, Tetrahedro Lett, 22: 1859 (1981).

66, Benton, W, D, et al, 5science 1 96:180-182 (1977).

67, Kafatos, F, C, et al, Nucl, ^aids R es, 7-: 1541-1552 (1979).

68, 5outhern, E, M, , J, Mo1. Biol, 98:50 3-517 (1975).

69, Sanger, F. et al, Proc, Natl. Acad, Sci, LISA 74 2 5463-54U7 (1977).

70, Messing, J, at al, Nucl, Ac1ds , Res, 9: 309-321 (1981).

71, Ullrich, A, et ah, 5science 196 : 1313-1317 (1977).

72, Avrv, H, et al, Proc, Natl, Acad, Sci , USA 89: 140g-1412 (1972).

73, Vickens, M, P, et al, J, Biol, C hem, 253: 2483-2495 <1978>.

74, Huynh, T, V, et al, “DNA Cloni ng Technical, A Practical Approach” (GIover, D, ed,, rRL, 0xford).

75, Norris, E, et al, Gene 7: 355-362 (1979).

7L Taylor, J, M, et al, Biochim, Bioph ys, Acta 442: 324'-330 (1976).

77, Crowley et al, Molec, Ce11. B iol, 3: 44-55 (1983).

78, Anzano et al, Molec, Ce11. Biol, 5-:242-247 (1985).

79, EP 73,658A0 80, 5eyedin et al, J, Biol, Chem, 262 ( 5): 1946-1949 (1987).

81, Ying et al, Biochem, Biophys, R es, Commun, 135: 950-956 (1986).

82, Mason et al, Nature 318: 659-66 3 (1985).

83, Fuller et al, J, C11n, Endocrin, M etab, 54: 1051-1055 (1982).

84, Fuller et al, Gynecol, Onco1.　 17: 124-132゜85, T. Maniatis et al. Molecular Cloning, pp, 133-134 (198 2).

86, R, Lawn et al, NucIeic Ac1ds Res , 9:6103-6114 (1981).

8’1. D, Goeddel et al, Nucleic Ac1ds Res, 8: 4057 (1980).

90, Simonsen and Levinson, Proc, Natl. , Acad, Sci. 80: 2495-2499 (1983).

91, Deryck et al, Cancer Res, 47: 707-712 (1987).

Q? Rrine++, +n pt *1. , CeIt AR: A') q-A40 (19B'l).

hJ-T to F: ALDTNYCFS5TEKNCC VRQLY n Immediately LGn-W engineering 0PXGYh-ρ-T cFs: HANFCTAP CPY Engineering WSLDT----QY5KVLAL-Y)JQ--HNPGAhJ- TGF, : YANFC5GPCPYLR5ADτ---TH5TV LGL-YNT--LNPnh, +9-TCF-S entry PCCVPQAL engineering P LPmVYYV-GRXPKVEQLSN'M Engineering VR5CXC580901oo 　　　　　110 10φ11.1 E) ILL AKPOlo 20 30 40 S.

10411 3 55G5REAAkfFSSLLLHVCWGLLLTR PR5PRASLPGSIQ;tLORALWLAhu4 　　　　　　　　　　　　　　　　　　　　　　　　　　　 CQ%4LL international search report □22; international search report US 8801945 S^　23028

Claims (22)

[Claims] 1. (a) Construct a vector containing a nucleic acid encoding TGF-β3, and ( b) transforming eukaryotic host cells with this vector; and (c) culturing the transformed cells. and (d) recovering TGF-β3 from the culture. 2. 2. The eukaryotic cell according to claim 1, wherein the eukaryotic cell is a Chinese hamster ovary cell line. Method. 3. 2. The method of claim 1, wherein TGF-β3 is pre-TGF-β3. 4. Pre-TGF-β3 has a viral secretion signal sequence at its N-terminus. , which is a fusion having a sequence of mature TGF-β3 at its C-terminus. Method. 5. Nucleic acid encoding pre-TGF-β3 is functionally linked to the viral promoter 5. The method of claim 4, wherein the method is coupled to 6. Nucleic acid encoding pre-TGF-β3 is functionally linked to an inducible promoter. 5. The method of claim 4, wherein the method is coupled to 7. 2. The method of claim 1, wherein mature TGF-β3 is recovered from the culture medium. 8. TGF that does not contain one or more introns present in genomic DNA - DNA encoding β3. 9. Extrachromosomal DNA encoding TGF-β3. 10. isolated from the cells of the desired species and encodes other proteins of that species mRNA encoding TGF-β3 of the desired species without mRNA. 11. Hybridization with DNA encoding mature human TGF-β3 in Figure 5 Nucleic acids released from cells that can be 12. The nucleic acid according to claim 11, which is a DNA containing at least 44 bases. 13. Figure 5 DNA encoding mature TGF-β3 and low stringency 12. The nucleic acid according to claim 11, which is capable of hybridizing under conditions of 14. 12. The nucleic acid of claim 11, labeled with a detectable moiety. 15. The DNA according to claim 8 or 9, which is labeled with a detectable moiety. 16. 11. The mRNA of claim 10, which is labeled with a detectable moiety. 17. A replicable vector containing DNA encoding TGF-β3. 18. 18. The vector according to claim 17, wherein the DNA does not contain introns. 19. A replicable vector containing the nucleic acid according to claim 11. 20. A host cell containing the vector of claim 17. 21. 21. The cell according to claim 20, which is a eukaryotic cell. 22. 21. The cell according to claim 20, which is a bacterial cell.