LU87442A1 - AMPHIREGULIN: A NEW BIFUNCTIONAL GLYCOPROTEIN MODULATING GROWTH - Google Patents

Description

89 / B 52 799

J s | λ GRAND-DUCHY OF LUXEMBOURG

...... L .......... S .........

Minister of ........ 25 ....... January ........ 19.8.9. of the Economy and the Middle Classes, Intellectual Property Service

TltredellVré ......................— I LUXEMBOURG

Invention Patent Application —............-.............................. ............ (1) I. Application

The so-called company: ONCOGEN, 3005 First Avenue, SEATTLE / __ ^ 2) ......... WASHIngîIIQM 98 ± 2J- ^ United States_.dlAmérigue) L_x.épresentée- paii --------- -------- ......... Sir ...... JacgaM_dgj ^^ ae, E_ aglssrnt ....... ^^ ---------- ---------------------------------------------------------- ---------------------------------------------------------- ---------------------------------------------------------- -----------------------------------------: -------- ---------------------- (3) remove (s) this .................. twenty- five january 1900 eighty nine ........................ (4) to ........ J..5 ... ................. hours, at the Ministry of the Economy and the Middle Classes, in Luxembourg: 1. this request for obtaining a patent for invention : "Amphiregulin: a ..... new glycoprotein __________________________________________________________ (5) bifunctional modulator of ...... growth.1 * __________________________________________________________________________________ 2. description in French......................................... ................... of the invention in triplicate; 3. ................ 34 ............................... ............. drawing boards, in triplicate; 4. the receipt of taxes paid to the Luxembourg Registration Office on ..... 2.5 ....... January: „1.9..8.9 .....; 5. the delegation of power, dated ......................................... ............................ on .................. January .. .... 1.9.89,; 6. the successor document (authorization); declares (s) assuming responsibility for this declaration, that the inventor (s) is (are): (6) ......... "does not ...... not .. ....mention!'........................................... .................................................. .................................................. ..........-....................................... .....................................— claim (s) for the above patent application priority one (s) request (s) from (7) ......... orevets .......................... .................................................. ...................................... filed in (8) .... ..in the United States of America ________________ on (9) ..... 25..01..1988 ....... (ISO, ....... 14.8 .., 3.2.7.) ...; ...... on ...... lSJläJSBMMüu ....... 13..1 ... 38.4) ...... and ......... .................................................. ............

msm & ys) ......... on ..... 11..01 ... 1.989 ......................... .................................................. .................................................. .................................................. .................................................. ............

on behalf of (11) s ............... inventors ........................... .................................................. .................................................. .................................................. .................................................. ...................

elect (elect) domicile for him / her and, if designated, for his / her representative, in Luxembourg ............................. .................................................. ........

35, Boulevard Royal .............................................. .................................. .............. (12 ) request (s) the grant of a patent for the invention for the subject described and represented in the abovementioned appendices, ^^ averajeumement of this grant to ................ § .................................................. .................................................. .................................................. ..................................month. (13) r The deposit W7 recipient: ........................................... .................................................. .................................................. .................................................. .................................................. ........................... (14) (IU Π. Statement of Deposit

The above-mentioned Wendering patent patent filed with the Ministry of the Economy and the Middle Classes, Intellectual Property Service in Luxembourg, dated: 25 j Vier 1989 „f *" <jï * J "· #» / * · Λ ".

/? ........ * T, y 4 '. ·· * î £' Pr. The Minister of Economy and the Middle Classes, to ............. .. ™ ..................... hours' · Ί i p. d.

; -V * f / .7 · 'The head of the intellectual property department, f' y- »/ '- V <v--. · * _ J, A68007__ (Ο ψ K EXPLANATIONS RELATING TO FORM DElfëfiÿr“ " w- * J f \ (1) if applicable "Request for certificate of addition to the main patent, to main patent application No ............ of .... ........ ”- (2) enter the surname, first name, profession, address of the applicant, when the latter is an individual or company name, legal form, address of the head office, when the applicant is a legal person - (3) enter pj *>} the surname, first name, address of the authorized representative, industrial property attorney, with special powers, if applicable: "represented by ...... ...... acting as agent "(* 7 / r - (4) date of filing in full - (5) title of the invention - (6) enter the names, first names, addresses of the inventors or the indication "(see) separate designation (will follow)", when the designation is made or will be made in a separate document, o u again the indication ”not to mention”, when the inventor signs or signs a document of non-mention to be added to a designation ηηίοαη, Αΐΐιι fnrm> e_l'7 \ hpAU ^ f iflPiif H'nHHitînn mnriÂl ”ri'ui- MîtyS hn> v /> r /> iimnpi »n ίΡΒΡΛ nrnrf» rrinn int * mnrinnnli * f Pf'T) - tJD Τ-ΐπΓ rinn «: tarnishes *! nremîer Hénnf a AtA pffprfiit ”89 / B 52 799

CLAIM OF PRIORITY

of the patent application / utility model KK IN THE UNITED STATES OF AMERICA From '25.01.1988 (No. 148.327);

From 15.04.1988 (No. 181.884) and From 17.01.1989 Brief Description filed in support of a request for

PATENT

at

Luxembourg

in the name of: otlcoGEN

SEATTLE, WASHINGTON 98121 (United States of America) ^ OUr * "Amphiregulin: a new growth-modulating bifunctional glycoprotein." L1amphiréquline: _a new bifunctional glycoprotein modulating growth.

r- 1. Introduction,

The present invention relates to the production and uses of 1 ’amphiregulin, which is a new growth-regulating protein. The proteins according to the invention strongly inhibit the growth of several cell lines of tumor origin, but promote the growth of several other cell lines. A wide variety of therapeutic applications of this bifunctional growth regulator are described, in particular the treatment of wounds, as well as the diagnosis and treatment of cancers, but without limitation to these applications, 1. Background of the invention .

Cell growth and differentiation appear to be initiated, promoted, nurtured and regulated by a large number of factors and hormones with stimulating, inhibiting and synergistic effects. Alteration and / or collapse of the cell homeostasis mechanism appears to be a fundamental cause of growth-related conditions, including neoplasms. Growth modulation factors are involved in a wide variety of pathological and physiological processes, including signal transduction, cell communication, growth and development, embryogenesis, immune response, hematopoiesis, survival and cell differentiation, inflammation, tissue reconstruction and remodeling, atherosclerosis and cancer. There is therefore a great interest in isolating, characterizing and defining the functional mechanisms of growth modulating factors 2 due to their potential application for the diagnosis, prognosis and treatment of cancer. In addition, gaining knowledge of these factors will promote understanding of the fundamental mechanisms underlying the regulation of normal growth and its disappearance in cancer cells.

Epidermal growth factor (EGF), transforming growth factor oC (TGF £> C), platelet-based growth factor (PDGF), fibroblast growth factor (FGF), nerve growth factor (NGF) , transforming growth factor ß (TGFyS), insulin growth factors I and II (IGF I, IGF II), hematopoietic growth factors such as erythropoietin, colony stimulating factors (CSF 1 and 2) , interleukins (IL-1 to 6), interferons (IFN, ß, ^), tumor necrosis factors o (and ß (TNF o (, ß), leukoregulin, oncostatin M and various other factors less well defined are growth and differentiation modulating proteins produced by various cell types under normal physiological conditions or in response to exogenous stimuli. These factors appear to act mostly in autocrine and paracrine fashion. (For articles review, see: Goustin et al., 1986, Cancer r. Res. 46: 1015-1019? Rozengurt, 1986, Science 234; 161 to 166? Pardee, 1987, Cancer Res. 47: 1488 to 1491? Sachs, 1986, Sci., Amer. 254: 40 to 47? Marshall, 1987, Cell 50: 5 to 6? Melcher and Anderson, 1987, Cell 30: 715 to 720? Clemens and McNurlan, 1985, Biochem. J. 226: 345 to 360? Nathan, 1987, J. Clin. Invest. 79: 319 to 326? Sporn and Roberts, 1986, J. Clin. Invest. 78: 329 to 332? Old, 1987, Nature, 326: 330-331? Beutler and 3

Cerami, 1987, New Eng. J. Med. 316: 379-385; Weinstein, J. Cell. Biochem., 33: 213-224? Zarling et al., 1987, Proc. Natl. Acad. Sei. U.S.A. 83: 9739-9744; Sporn and Todaro, 1985; N. Eng. J. Med. 303; 878 r to 880; Sporn and Roberts, 1985, Nature 313: 745-747).

Biologically active phorbol esters, such as 12-0-tetradecanoylphobol 13-acetate (TPA) are potent promoters of tumors in vivo and elicit and modulate a wide variety of biological and biochemical responses in vivo and also in vitro (Blumberg, 1981, Crit. Rev. Toxicol. 9: 153-197; Slaga, 1983, Cancer Surv. 2: 595-612). It has been known for some time that TPA inhibits the growth of the human mammary adenocarcinoma cell line MCF67. In addition, TPA also modifies the morphology of MCF-7 cells since cells treated with TPA manifest the morphological characteristics of secretory cells (Osborne et al., 1981, J. Clin. Invest. 67: 943 to 951; Valette and al., 1987, Cancer Res. 47: 1615-1620).

3. Overview of the invention.

The present invention relates to amphiregulin which is a new growth regulator which manifests bifunctional and unusual activity regulating cell growth. Amphiregulin inhibits the growth of neoplastic cells, but promotes the growth of certain normal cells. Amphiregulin is an extremely hydrophilic glycoprotein with a median molecular weight of 22,500 daltons, which exists in two distinct but functionally equivalent forms, a truncated form and a larger form. Except for the six additional N-terminal residues existing in the larger form, the two forms are perfectly homologous at the level of the 4 amino acids (Fig. 12). The invention is based in part on the discovery that MCF-7 cells treated with TPA release a glycoprotein which inhibits the growth of the cell line of epidermoid carcinoma A431 and other tumor cell lines, but increases the growth of normal human fibroblasts and certain other cell lines.

The invention is described by means of examples in which the amphiregulin is identified, purified to homogeneity and characterized in detail in terms of structure and functions. In other examples, the isolation and sequencing of cDNA and genomic clones encoding the amphiregulin precursor are described. These clones were used to prepare expression vectors capable of directing the synthesis at a high level of the biologically active amphiregulin in transformed bacterial host cells and transfected eukaryotes.

A wide variety of applications of this remarkable factor fall within the scope of the invention.

4. Brief description of the figures.

Fig. 1. Preparative phase inversion HPLC μ of the overhead and washing fractions.

Fig. 2. HPLC semi-preparative phase inversion of fractions 47 to 62 pooled in FIG. 1.

Fig. 3A. Analytical phase inversion HPLC of pool I of the previous test.

Fig. 3B. Analytical phase inversion HPLC of pool II of the previous test.

Fig. 4. Fraction gel permeation chromatogram of FIGS. 3A and 3B on Bio-Sil 5 TSK 250 columns. Chromatography was carried out as described in section 6.3.2. below. (A) HPLC of the concentrated fraction (Fig. 3A); (B) concentrated fraction HPLC 36 (Fig. 3A); (C) concentrated fraction HPLC 37 (Fig. 3A); (D) HPLC of the combined fractions 41 and 42 combined (FIG. 3B).

Fig. 5. Analysis of purified AR and S-pyridylethylated AR (SPE-AR) by HPLC by gel permeation on Bio-Sil TSK 250 columns. Chromatography was carried out as described in section 6.3.2. below. The molecular weight markers used were 43 kD ovalbumin, 25 kD chymotrypsinogen A, 14 kD ribonuclease and 5 kD insulin, (A) AR, (B) SPE-AR.

Fig. 6. Analysis of the AR and of the SPE-AR on an ultraporous RPSC C3 column with phase inversion (4.6 × 75 mm). The gradient was established between the primary solvent, 0.1% TFA and the secondary solvent, acetonitrile with 0.1% TFA, at a flow rate of 1 ml per minute at room temperature. 1 ml fractions were collected (A) AR, (B) SPE-AR.

Fig. 7. SDS-PAGE analysis of AR proteins. (A) 15% SDS-PAGE gel (0.75 mm x 18 cm x 15 cm); Bio-Rad) with a discontinuous buffer system was examined at a constant current of 30 milliamps. The molecular weight markers used were phosphorylase B 92.5 kD, BSA 66.2 kD, ovalbumin 43 kD, carbonic anhydrase 31 kD, chymotrypsinogen A 25.7 kD, l 21.5 kD soybean trypsin inhibitor, 18.4 kD lactoglobulin, 14.4 kD lysozyme, 6.2 kD aprotonin and the 3 kD insulin subunit. Column 1: AR; column 2: NG-AR; column 3: SPE-AR; column 4: NG-SPE-AR. (B) A 20% SDS-PAGE minigel 6 (0.75 mm x 10 cm 7 cm) was examined under constant voltage of 200 volts. The same molecular weight markers as above were used. Column 1: AR; column 2: NG-AR; column 3: NG-AR treated with phospholipase, separated on phase inversion HPLC.

Fig. 8. IEF analysis of the AR marked with - ^ 5 ^ t The following standards having Pj values between 4.65 and 9.6 were used: phycocyanin, 4.65; - lactoglobulin B, 5.1; bovine carbonic anhydrase, 6.1; human carbonic anhydrase, 6.5; horse myoglobin, 7.0; whale myoglobin, 8.05; ^ -chymotrypsin, 8.8 and cytochrome C, 9.6.

Fig. 9. (A) AR dose-response curve for inhibition of the incorporation of 125d deoxyuridine into the DNA of A431 cells; (B) AR stimulating effect on the incorporation of 125χ_ deoxyuridine into the DNA of human foreskin fibroblasts (Sadamoto).

Fig. 10. Competition for 125χ_ EGF binding on fixed A431 cells or A431 plasma membranes by mouse EGF and AR. The fixing assays were carried out as described in section 6.1.5. below. Samples of 100 JUl and 50 JL_L per well were used for the assays with fixed cells or membranes, respectively.

Symbol Origin of receiver Competitor

• EGF fixed cells

• * AR fixed cells

O EGF membranes

^ AR membranes

Fig. 11. Analysis of hydropathy of AR and human EGF (Kyte and Doolittle). (A) AR (remains 1-7 84); (B) AR (remains 44 to 84); (C) EGF (remains 1 to 53); (D) AR precursor (residues 1 to 252); (E) comparison of human AR (residues 1 to 84) and EGF (residues 1 to 53) (the alignment is in accordance with Fig. 13, so that the cysterine residue 46 of the AR at maturity corresponds to the rest of cysteine 6 from EGF at maturity).

Fig. 12. Amino acid sequence of mature AR and truncated AR. The usual single letter code for amino acids is used: Alanine (A); Arginine (R); Asparagine (N); Aspartic Acid (D); Cysteine (C); Glutamine (Q); Glutamic acid (E); Glycine (G); Histidine (H); Isoleucine (I); Leucine (L); Lysine (K); Methionine (M); Phenylalanine (F); proline (P); Serine (S); Threonine (T); Tryptophan (W); Tyrosine (Y) and Valine (V).

Fig. 13. Comparison of the amino acid sequences of amphiregulin (AR) and members of the EGF superfamily.

Fig. 14. (A) Fixation of 125I-AR on cellulose (o-o), denatured DNA-cellulose (D-O) and native DNA-cellulose (û-û). (B) Fixation of 125I_ AR on cellulose (o -—- o) and DNA-cellulose (0 — a) compared to the fixation of ^ 2 ^ I-EGF on cellulose (û-ù) and l 'DNA-cellulose (V-V).

Fig. 15. Synthetic amphiregulin peptides produced by a solid phase technique. Five peptides corresponding to the residues 166 to 184 (no 259), 108 to 130 (no 264), 31 to 50 (no 279), 71 to 90 (no 280) and 221 to 240 (no 281) have were produced by solid phase synthesis, then purified by HPLC with phase inversion.

Fig. 16. Nucleotide sequence and deduced amino acid sequence from the pAR1 cDNA clone, coding for human amphiregulin.

8

Fig. 17. Genomic sequence of human amphiregulin.

Fig. 18. Schematic diagram of the structure of exon and the protein domains of the AR molecule, in relation to the genomic sequence.

Fig. 19. Nucleotide sequence of the expression vector pDCHBARl.

Fig. 20. Nucleotide sequence of pTacAPARl.

Fig. 21. Nucleotide sequence of pTacAPHILE.

5. Detailed description of the invention.

The present invention relates to amphiregulin (AR), the nucleotic sequences coding for AR and the precursor of AR, in addition to the production of AR according to conventional methods and of recombinant DNA.

AR, which is a new factor in regulating cell growth, is expressed and secreted by MCF-7 cells treated with TPA, in two distinct but functionally equivalent forms. The two forms are identical in structure, except for an additional 6 amino-terminal remains in the largest form. AR has a powerful inhibitory activity on DNA synthesis in neoplastic cells and is therefore of interest as a powerful anti-tumor agent. Of particular interest is the ability of AR to promote the growth of fibroblasts and various other normal cells, which suggests that AR may be useful in treating conditions requiring accelerated cell growth, namely burns and sores.

RA, or its fragments and derivatives, have broad potential therapeutic utility for the treatment and monitoring of neoplastic conditions, as well as for the treatment of wounds, RA can also find its utility in the modulation of resorption bone and immune response, as well as in stimulating the cascade of arachidonic acid. The amphiregulin gene and the products of this gene, methods for producing them and their use are described in detail in the subsections and examples below.

5.1. Production and purification of 11 amphiregulin.

Amphiregulin is secreted by the MCF-7 human breast carcinoma cell line when these cells have been treated with 12-0-tetradecanoylphorbol 13-acetate (TPA) which is a tumor promoting agent. Amphiregulin can be isolated from the conditioned medium of these TPA-treated MCF-7 cells and cultured and can then be purified to high specific activity. Amphiregulin can also be isolated from other cell lines with or without inductive therapy with TPA or other tumor promoting agents. Alternatively, amphiregulin can be produced by combining DNA techniques, or by solid phase peptide synthesis.

The amphiregulin can be purified according to various known operating methods and techniques, in particular, but not limited to, chromatography (for example liquid chromatography with phase inversion, by gel permeation, by liquid exchange, by ion exchange, by dimensional and affinity exclusion), centrifugation, electrophoresis, difference in solubility, and according to various other common techniques for protein purification.

In a specific embodiment of the invention, MCF-7 cells are treated with TPA for 48 hours and cultured in fresh serum-free medium for 4 days. The resulting conditioned medium is collected and used to prepare the crude RA as described in section 6.2. below. A combination of phase inversion liquid chromatography and gel permeation chromatography can be applied to purify the RA to apparent homogeneity, as described in section 6.3. below, and gives purified AR from 1842 to 2270 times compared to the initial crude product. The process is reproducible and gives purified AR preparations with specific activities between 2.7 and 3.4 x 106 units per mg protein.

5.2. The amphiregulin gene.

5.2.1. Isolation and cloning of the amphiregulin gene.

For the application of the method of the invention, the nucleotide sequence encoding human amphiregulin or its functional equivalent can be used to generate recombinant molecules which will direct expression and the human amphiregulin which is the product. The nucleotide sequence encoding AR can be isolated from cellular sources producing AR-like activity. For example, in a specific embodiment, the human mammary carcinoma cell line MCF-7 is used as the source of the nucleotide sequence coding for AR. The coding sequence can be obtained by cloning the cDNA from RNA isolated and purified from these cellular sources, or by genomic cloning. Genomic or cDNA libraries of clones can be prepared starting from the DNA fragments generated according to known techniques, in particular, but not limited to, the use of restriction enzymes.

The fragments which contain the gene for the AR can be identified in various known ways. For example, a fraction of the 11 amino acid sequence of AR can be used to deduce the 11 DNA sequence, which DNA sequence can then be chemically synthesized, radioactively labeled and used as a hybridization probe. .

Other methods which can be applied to isolate the AR gene include, but are not limited to, the chemical synthesis of the sequence itself of the gene from a known sequence which may, for example, be derived from the sequence of amino acids of AR. Alternatively, in vitro translation of the selected mRNA, followed by functional or immunological assays of the translation products is applicable. The identified and isolated gene can then be inserted into an appropriate cloning vector. A large number of known vector-host systems can be used. Possible vectors include, but are not limited to, plasmids and modified viruses in the event that the vector system is compatible with the host cell. Such vectors are notably, but not limited to, bacteriophages, such as derivatives of bacteriophage lambda, or plasmids, such as derivatives of plasmids PBR322 and pUC. Recombinant molecules can be introduced into host cells by transformation, transfection, infection, electroporation and so on.

In a particular embodiment, the AR gene expressed by MCF-7 cells is cloned by selecting the mRNA produced in response to the treatment of MCF-7 cells with TPA and by constituting a library of cDNA in bacteriophage 7 ^ 9 ^ 10 . Since the untreated MCF-7 cells do not seem to synthesize and secrete the AR protein, it was assumed that the induction of AR by TPA could be observed at the level of transcription and that such a cDNA library could be considerably enriched for sequences containing the AR. Survey of the cDNA library using degenerate and presumed to be the best oligonucleotides based on the use of human codons (Lathe, 1985, J. Mol. Biol. 183: 1 to 12 and as described in section 8.1. below) has led to the isolation of AR cDNA clones. These cDNA clones were then used to characterize the AR mRNA and the AR gene. In addition, the nucleotide sequence of the AR cDNA (section 8.2. Below and Pig. 16) can be used to deduce the primary amino acid sequence of the AR.

Due to the inherent degeneration of the coding nucleotide sequences, other DNA sequences which code for substantially the same amino acid sequence or a functionally equivalent amino acid sequence may be used to apply the methods of the invention. These modifications to the nucleotide sequences for AR include deletions, additions or substitutions of different nucleotides leading to a sequence which codes for an identical or functionally equivalent gene product. The gene product may contain deletions, additions or substitutions of amino acid residues in the sequence leading to silent modifications which thus give a bioactive product. These amino acid substitutions can be made on the basis of the similarity of polarity, charge, solubility, hydrophobic character, hydrophilic character and / or amphipatic character of the residues in question. For example, negatively charged amino acids include aspartic acid and glutamic acid; positively charged amino acids include lysine and arginine; 13 amino acids comprising uncharged polar head groups and non-polar head groups having similar hydrophilicity indices include leucine, isoleucine, valine; glycine, alanine; asparagine, glutamine; serine, threonine; phenylalanine and tyrosine.

The AR cDNA can be used as a probe to detect the AR raRNA in TPA-induced MCF-7 cells. The AR mRNA is approximately 1400 base pairs in length.

As described in section 9 below, the AR cDNA can be used to characterize the AR gene. The assignment of the human AR gene to the chromosone was determined by in situ hybridization of the cDNA of the AR labeled with normal chromosones in metaphase, which reveals that the human AR gene resides in the 4ql3-21 region of chromosone 4 , which is a region believed to be involved in the differentiation of lymphocytes (see section 9.2. below). CDNA was also used to isolate genomic DNA spanning the entire AR gene, including the 5 'regulatory region. The AR gene has been shown to be around 10 kb in length and is subdivided into 6 exons. The 5 'regulatory region was incorporated into an expression vector containing a chloramphenicol acetyltransferase (CAT) gene without promoter. This constructed vector was able to stimulate transcription of the CAT gene when it is introduced transiently into MCF-7 cells and the activity was stimulated 6 to 7 times by an addition of TPA (see section 9.4. below). In specific embodiments of the invention, this 5 'regulatory region can be used in expression vectors to regulate transcription of the AR gene or other structural genes. The AR-CAT hybrid construct can also be used to test for factors that regulate AR expression, as described in section 9A.

5.2.2. Construction of expression vectors containing the sequence coding for amphiregulin.

To express a biologically active and mature form of AR, it was necessary to choose an expression vector / host system which not only ensures high levels of transcription and translation, but also the adequate processing of the gene product. This is especially important when using the complete coding sequence of the amphiregulin precursor in constructed expression products, since the mature form of amphiregulin appears to be derived from the precursor by cell processing events. . For example, a mammalian host cell can be chosen because of its ability to properly process and secrete amphiregulin into the extracellular medium.

Specifically, it appears that two forms of mature amphiregulin are synthesized as the middle part of a common precursor of 252 amino acids from which the two mature forms are released by alternating proteolytic processing events. In addition, amphiregulin is glycosylated and may undergo tyrosine sulfation, which further emphasizes the importance of choosing an expression system that is capable of performing these post-translational changes, if it is desired, in the final product.

Various animal host / expression vector systems (i.e., vectors which contain the elements necessary to direct replication, transcription and translation of the coding sequence for RNA in an appropriate host cell) can be used with equal success by those skilled in the art. These are in particular, but not limited to, the viral expression vector / mammalian host cell systems (for example cytomegalovirus, vaccinia virus, adenovirus, etc.); insect vector / insect cell expression systems (eg baculovirus)? or non-viral promoter expression systems from the genomes of mammalian cells (eg, the mouse metallothionin promoter).

The expression elements of these vectors vary by their power and their specificities. Depending on the host / vector system that is used, any one of many transcription and translation elements can be used. For example, when cloning into mammalian cell systems, promoters isolated from the genome of mammalian cells (e.g. mouse metallothionine promoter) or isolated from viruses that grow in these cells (e.g. 7.5K promoter vaccinia virus or long terminal repeat of Moloney mouse sarcoma virus) can be used. Promoters produced using recombinant DNA techniques or synthetic techniques can also be used to ensure transcription of the inserted sequences.

Specific initiation signals are also required for sufficient translation of the inserted sequences encoding the protein. These signals are in particular the AGT initiation codon and the adjacent sequences. In cases where the complete AR gene, including its own start codon and adjacent sequences, is inserted into the appropriate expression vectors, additional translation control signals may be unnecessary.

16

However, in cases where only part of the coding sequence is inserted, exogenous translation control signals, in particular the initiation codon ATG, must be provided. In addition, the initiation codon must be in phase with the reading frame of the sequences coding for the RNA in order to ensure the translation of the entire inserted part. These exogenous translation control signals and initiation codons can have different origins, both natural and synthetic. Expression efficiency can be enhanced by including transcription attenuation sequences, enhancement elements, etc.

Any of the methods already described for inserting DNA fragments into a vector can be applied to construct the expression vectors containing the AR gene and the appropriate transcription / translation control signals. These methods may include in vitro recombinant DNA techniques, synthetic techniques and in vivo recombination techniques (genetic recombination).

For example, in cases where an adenovirus is used as an expression vector, the sequence coding for 11AR can be ligated to a transcription / translation control complex of the adenovirus, for example the late promoter and the signal sequence. tripartite. This hybrid gene can then be inserted into the genome of the adenovirus by recombination in vitro or in vivo. Insertion into a nonessential region of the viral genome (for example the E1 or E3 region) leads to a recombinant virus which is viable and capable of expressing AR in infected hosts. Likewise, the 7.5 K promoter of vaccinia can be used.

Another expression system that could be used to express AR is an insect system. In such a system, the Autographa californica nuclear polyhedrosis virus (AcNPV) is used as a vector to express foreign genes. This virus multiplies in the cells of Spodoptera frugiperda. The sequence coding for the AR can be cloned into nonessential regions (for example the polyhedrin gene) of the virus and placed under the control of an AcNPV promoter (for example the polyhedrine promoter). The successful insertion of the sequence coding for the AR leads to the inactivation of the polyhedrin gene and to the production of the non-occluded recombinant virus (that is to say the virus free of the protein envelope for which the polyhedrin gene). These recombinant viruses are then used to infect Spodoptera frugiperda cells in which the inserted gene is expressed.

Retroviral vectors prepared in amphotropic packaging cell lines allow highly efficient expression in many types of cells. This method makes it possible to evaluate the treatment, the regulation or the function of the inserted sequence coding for the protein which are specific for the type of cell.

In addition, it is possible to choose a host cell strain which modulates the expression of the inserted sequences or which modifies and processes the gene product in the specific manner desired. Expression by certain promoters can be increased in the presence of certain inducers (for example zinc and cadmium ions for metallothionine promoters). For this reason, the expression of the AR produced by genetic engineering can be controlled. This is important if the protein produced by the cloned foreign gene is lethal to the host cells. In addition, the modifications (eg glycosylation) and processing (eg cleavage) of the proteins produced are important for the function of the protein. Different host cells have characteristic and specific mechanisms for post-translational processing and for protein modification. It is possible to choose suitable cell lines or host systems to ensure appropriate modification and processing of the expressed foreign protein.

Expression vectors which can be used in accordance with the present invention include, but are not limited to, the following.

The plasmid pSV2Neo. The plasmid pSV2Neo is described by Southern et al. (1982), J. Mol. Applied Genetics 1, 327-341.

The plasmid pSV2dhfr. The plasmid pSV2dhfr (Subramani et al., 1981, Mol. Cell. Biol. 1, 854 to 864) contains the mouse dihydrofolate reductase (dhfr) gene under the control of the SV40 promoter.

The plasmid pH3M. The plasmid pH3M (Aruffo et al., 1987, Proc. Natl. Acad. Sci. USA, 84, 3365 to 3369), contains a hybrid promoter composed of the immediate early enhancer AD169 of the human cytomegalovirus (CMV) fused at positions - 67 to +80 LTR of HIV. The LTR is followed by a polylinker with 5 ′ to 3 ′ sites: HindIII, Xbal, Xhol, BstXl, Xhol, PstI and Xbal. The small splicing signals of the t antigen / polyadenylation of pSV2 and an origin of replication of SV40 are downstream of the polylinker. The latter allows amplification in cells containing the large T antigen of SV40, such as COS and WOP cells. A suppressor tRNA gene maintains growth in pVX-based vectors and bacteria carrying the P3 plasmid, such as MC1061 / P3.

The plasmid pH3M / b0ncM. The plasmid pH3M / b0ncM

19 was acquired from Najma Malik, Oncogen. It contains the untranslated monkey TGF / 15 signal sequence fused to the sequence coding for Oncostatin M inserted into pH3M at Hindlll / Xhol sites (completed). The sequence TGFfi> consists of the untranslated region 51 of 85 base pairs starting at a HindiII site (starting from the polylinker pSP65) and at a PstI site adjacent to the TGFfi cAND, followed by 87 base pairs coding for the signal sequence of amino acid 29, which is normally cut in a glycine-leucine dipeptide sequence.

The plasmid pH3ARP. The plasmid pH3ARP is a vector based on pH3M designed for the transient expression of the precursor of AR in COS cells. It contains 13 base pairs of the 5 'untranslated region of the AR, the complete coding region and the 3' untranslated sequences. This plasmid is formed by ligation of the 4 kb fragment of the vector pH3M digested with HindIII and Xbal, of the oligonucleotides EVSAL3 and EVSAL4 and of the fragment of cDNA HgAl / Xbal of 1.1 kb purified on gel from PARI. EVSAL3 and EVSAL4 are complementary with 5 ′ HindIII, EcoRV, SalI and Hgal sites. The product also contains the EcoRI to Xbal polylinker sites from pEMBL18 immediately following the 3 ′ untranslated region.

EcoRV Hgal HindlII Sali

EVSAL3 5 'AGCTTGATATCGTCGA

CGCAL4 3 'ACTATAGCAGCTGACGC

The plasmid pSVDR / bOM. The plasmid psVDR / bOM was acquired from Jeff Kallestad, Oncogen. It is a fusion product between pSV2dhfr and pH3M / bOM and it constitutes a vector with a cDNA actuated by the CMV / HIV promoter of pH3M flanked by a signal sequence of TGF ß and polyadenylation signals of SV40, in addition of the mouse dhfr gene activated by the SV40 early promoter. It was produced by ligation of pH3M / bOM digested with BamHI / NruI (completed) with pSV2dhfr digested with BamHI.

The plasmid pHras. The plasmid pHras is a mammalian expression vector developed by Stan McKnight, Univ. of Washington, Dept of Pharmacology. It is a pML vector which contains a fragment of 754 EcoRI base pairs up to Tthllll from Harvey Ras LTR, a polylinker with Smal, BamHI, Sali, PstI, HindiII and Clal followed by the mouse DHFR cDNA and signal 3 untranslated / polyadenylation of hepatitis B virus. The distance from BamHI in the polylinker to the ATG of DHFR is 54 base pairs.

The plasmid pHLARGE. The plasmid pHLARGE is a mammalian expression construct which contains the 10 kb AR genomic sequences (Smal to HinlII) activated by Harvey Ras LTR and followed by a promoter-free and non-functional DHFR gene. The following fragments were formed by ternary ligation: a Smal / HindIII fragment of pHras of 4.6 kb, a Smal / HindIII genomic fragment of 6.0 kb originating from pARH12, and a HindiII genomic fragment of HindiII of 6.4 kb from pARH6.

The plasmid pHLARGlpcD. The plasmid pHLARGlpcD is a product of polycistronic mammalian expression. It contains the AR precursor cDNA sequence followed by the mouse dhfr gene activated by a unique Harvey Ras LTR. The primary transcription product is a polycistronic message comprising 64 base pairs between the AR stop codon and the ATG initiation codon of dhfr, which allows the ribosome to effect a reinitiation and to continue the translation. . The pH3ARP was digested with EcoRV and the 700 base pair fragment was isolated. This fragment contained the 5 'untranslated region of 13 base pairs of AR, the region encoding the complete precursor of AR and a 5' untranslated sequence of 21 base pairs. This fragment was ligated with the pHras phosphatase vector, completed and cut with BamHI.

The plasmid pLOSNL. The plasmid pLOSNL was acquired from Dusty Miller, Fred Hutchison Cancer Research Center, Seattle, WA. This retroviral expression vector was designed to generate viral stocks free of high titer amphotropic helpers capable of infecting a large variety of hosts but not of replicating therein. The vector contains an LTR mutated Moloney mouse leukemia virus with deleted packaging signals; psi + sequences; the ornithine transcarbamylase (OTC) gene containing two Xhol sites for the insertion of a cDNA and the subsequent inactivation of the OTC gene, SV2Neo, which is a selectable marker, the 3 ′ LTR and a backbone of pBR322 resistant to Amp.

The plasmid pLARSNL. The plasmid pLARNSL is an amphotropic retroviral expression construct containing the coding region of the cDNA of the AR precursor, free of poly (A) tail, activated by the viral LTR. SV2Neo allows selection based on the expression of transfectants. PLARSNL was generated by ligation of the 800 base pair SalI fragment from pHLARGlpcD with the vector fragment pLOSNL digested with XHol and phosphate of 7.7 kb.

22 5.2.3. Identification of transfectants or transformants which express 11 amphiregulin as a gene product.

The host cells which contain the recombinant sequence coding for the AR and which express the biologically active product at maturity, can be identified according to at least four general routes: (a) DNA-DNA hybridization, DNA-RNA or antisense RNA-RNA, · (B) the presence or absence of "marker" functions of the gene; (c) assessing the degree of transcription as measuring the expression of the transcribed products of the AR mRNA in the host cell and (d) detecting the mature gene product revealed by the immunoassay and, lastly remedy, its biological activity.

According to the first approach, the presence of the sequence encoding human AR inserted into the expression vector can be detected by DNA-DNA hybridization using probes which include nucleotic sequences which are homologous to the sequence encoding the Human AR.

According to the second approach, the recombinant expression vector / host system can be identified and selected on the basis of the presence or absence of certain “marker” gene functions (for example thymidine kinase activity, resistance resistance to methotrexate, the phenotype of transformation, the formation of an occlusion body in a baculovirus, etc.). For example, if the sequence coding for AR is inserted into a sequence of a marker gene of the vector, the recombinants containing the sequence coding for AR can be identified by the absence of the function of the marker gene. Alternatively, a marker gene can be mounted in tandem with the AR sequence under the control of a promoter which is identical to that used to control expression of the AR coding sequence or which is different from it. Expression of the marker in response to induction or selection indicates expression of the AR coding sequence.

According to the third approach, the transcription activity for the region coding for the AR can be evaluated by hybridization tests. For example, polyadenylated RNA can be isolated and analyzed by Northern blot using a probe homologous with the sequence coding for AR or with particular parts of it. Alternatively, the total nucleic acids of the host cell can be extracted and subjected to hybridization assays with such probes.

According to the fourth approach, the expression of the protein product at maturity can be evaluated by immunological route, for example by a Western blot, immunoassays such as radioimmunoprecipitation, the immunoenzymatic test in solid phase (ELISA), etc. The final test for the success of the expression system, however, is to detect the biologically active AR produced by the gene. When the host cell secretes the gene product, the cell-free medium obtained from the cultured transfecting host cells can be assayed for AR activity. When the gene product is not secreted, cell lysis products can be subject to such an activity assay. In either case, the biological assays, for example the growth inhibition or stimulation assays described herein and analogous assays, can be performed.

24 5.3. Structure of amphirequline.

The amino acid sequencing of the purified AR reveals that the preparation is formed by an uneven mixture of two almost identical forms of AR (see Fig. 12). One of the forms, which is the larger of the two, constitutes approximately 16% of the preparation. The other form, which is truncated AR, constitutes the rest and most of the preparation and differs from the longer homologue only by the absence of the aminoterminal hexapeptide, SVRVEQ. The two forms are, moreover, perfectly homologous in terms of amino acids.

AR is an extremely hydrophilic glycoprotein with a median relative molecular weight of 22,500 daltons. Treatment with N-glycanase, which removes N-linked carbohydrates, migrates AR to the single band state with a relative molecular weight of 14,000, in good agreement with the molecular weight calculations based on AR gel permeation chromatography data. In non-reducing conditions, AR migrates in the same way. Therefore, AR is a single chain glycoprotein.

The primary structure of AR has been compared to that of many protein sequences. These comparisons show that the AR is a remarkable protein having a sensitive structure homology with other of the proteins compared. The primary structure of AR, however, is related to that of various other growth factors, most of which belong to the EGF superfamily. It seems rational to consider AR as a member of this group of proteins, because various of its structural features are closely parallel to the highly conserved structural features observed in the family of EGF proteins. For example, the N-terminal sequence of AR resembles the N-terminal sequences of TGFy3, VGF and MGF, in that this region is rich in residues of proline, serine and threonine. The AR also includes sites for N-linked glycosylation as contained in the N-terminal precursor regions of TGF and VGF. AR also presents structural homology with other EGF-like proteins, with the conservation of 6 cysteine residues intervening in 3 disulfide bridges which define the secondary structure of the mature forms of these growth factors. Nevertheless, the hydropathy analysis reveals significant differences between the homologous sequences of AR and EGF. For more detailed comparisons between the AR and other members of the EGF superfamily, reference should be made to Table I, Fig. 13 and sections 9.7 and 9.8.

26

TABLE I

Protein Species Region # 1 Region # 2

Human EGF GLHDGVCMYIE CNCWGYIGERC

TGF 0 (Human CFH-GTCRFLV CVCHSGYVGARC

VGF GLH-GDCIHAR CRCSHGYTGIRC Vaccine

SGF Shope GLNNGTCFTIA CVCRINYEGSRC

Human Factor IV GLNXGSCKDDI CWCPFGFEGKNC

Human X Factor CQNXGKCKDGL CTCLEGFEGKNC

Human Factor XII GLHXGRCLEVE CHCPVGYTGPFC

Human Protein C CCGXGTCIDGI CDCRSGWEGRFC

AR Human CIH-GECKYIE CKCQQEYFGERC

Combined homology ratings for regions # 1 and # 2.

EFG rating Coag rating. AR rating

Family of growth factors> 20 <20 <40

Coagulation factors <25> 25 <40

Amphiregulin subfamily <20 <20> 40

The amino acid sequences of the amphiregulin illustrated in FIG. 12, as well as their functional equivalents, fall within the scope of the invention. For example, the amphiregulin produced may contain deletions, additions or substitutions of amino acid residues in the sequence, which result in silent modifications thus giving a bioactive product. These amino acid substitutions can be carried out on the basis of the similarity of polarity, charge, solubility, hydrophobic character, hydrophilic character and / or amphypatic character of the residues in question. For example, the negatively charged amino acids are in particular aspartic acid and glutamic acid; positively charged amino acids include lysine and arginine, and uncharged polar head amino acids of similar hydrophilicity are leucine, isoleucine, valine, glycine, alanine, l , glutamine, serine, threonine, phenylalanine and tyrosine.

5.4. Properties of amphiregulin.

AR is a single chain glycoprotein which has a median molecular weight of approximately 22,500 daltons and which exhibits bifunctional growth modulating activities on various cells in culture. From a structural point of view, AR is related to the EGF family of growth factors and may, moreover, have functional similarities with other members of this family, as indicated by the suitability of the AR to compete effectively with EGF for binding to receptors.

AR is a monomeric protein which requires intracatenary disulfide bridges for its activity, which does not require glycosylation for its biological activities, which is stable in a moderately acid or alkaline medium and which is stable during a heat treatment at 56 °. C for 30 minutes. RA loses its full biological activity when reduced, when heated at 100 ° C for 5 minutes, when subjected to digestion with trypsin, endopeptidase Lys-C, endopeptidase ArG and endopeptidase Glu.

Proteolytic cleavage using phospholipase D appears to convert AR into one or more active fragments with relative molecular weights of about 5,600, as indicated by the SDS-PAGE analysis. The active fragments of the AR fall within the scope of the invention and are considered in more detail in section 5.5. below.

RA has potent antiproliferative effects in vitro on various human cancer cell lines of epithelial origin. For example, AR effectively inhibits DNA synthesis by cells of epidermal carcinoma of the vulva A431, breast adenocarcinoma HTB 132, cervical squamous cell carcinoma CRL 1550, papillary ovarian adenoma HTB 75, ovarian teratocarcinoma HTB 1572 and mammary adenocarcinoma HTB 26. A 50% inhibition of DNA synthesis is observed on A431 cells treated with 0.13nM AR.

Normal NRK-SA6 cells in the rat kidney usually do not form colonies on soft agar. However, the combination of exogenous EGF and TGF ß added to the culture medium of these cells induces growth independent of anchoring, which indicates a transformed phenotype. A combination of exogenous AR and TGF ß also induces growth independent of anchoring of NRK-SA6 cells, but to a lesser degree (see Table VI), which provides direct evidence that AR and EGF are functionally related.

'The synergistic action of AR with TGF $ strongly involves AR as an important factor in regulating cell growth. Therefore, the present invention provides previously inaccessible research tools for investigating the complex and recent past of normal regulation of cell growth and the characteristic loss of growth regulation in neoplastic cells.

RA also induces the proliferation of human foreskin fibroblasts (Table IV). A doubling of DNA synthesis by these cells has been observed at an AR concentration of approximately 50 μM. Maximum stimulation up to approximately six times appears for approximately 5nM.

The mechanism by which AR induces proliferation or inhibition of cell growth is unknown. Nevertheless, preliminary studies aimed at characterizing the binding of AR to receptors suggest that AR may have a specific receptor. AR competes with EGF for binding to EGF receptors and may also compete with EGF for other common receptors, including possibly the specific AR receptor itself -even. It is known that the ligand forming the binding to the EGF receptor initiates a growth stimulation signal, in part, by activating the activity of thyrosine kinase. Binding of AR to EGF receptors can similarly initiate a proliferative response or, conversely, can block the initiation of a proliferative signal by limiting the number of EGF receptors available to bind EGF or other ligands generating the signal.

Alternatively, another mechanism by which AR inhibits cell growth is possible and is suggested by the data collated in Table I. Four clones of the A431 cell line, which have variable number of EGF binding sites per cell, were subjected to a test of the sensitivity to the AR inhibitory activity. The observed lack of correlation between AR sensitivity and the number of EGF binding sites per cell suggests that the growth inhibiting signal generated by AR is initiated by a binding event on a receptor. which does not involve the EGF receptor. Such an event can antagonize the growth proliferation signals generated by other growth factors such as, for example, TGF. Thus, the AR can exert its antiproliferative effect by specifically blocking a mitogenic response of the cell to TGF or to other autocrine growth factors.

AR is an extremely hydrophilic protein, the amino acid sequence of which generates a unique hydropathy profile that has little in common with the profiles of other proteins in the EGF family (Fig. 11). AR is a basic protein with a PI of 7.6 to 8.0.

5.4.1. Characterization of AR induction by TP A.

TPA induces a diverse response in cells, including changes in cell morphology, cell proliferation and differentiation, membrane transport, receptor-ligand interactions, protein phosphorylation and phospholipid metabolism. Protein kinase C (PKC) is a protein kinase dependent on Ca2 + and phospholipids, which functions as a receptor for TPA. The binding of TPA to PKC leads to the phosphorylation of many substrates, in particular growth factor receptors, cytoskeleton proteins and transactivating regulatory proteins.

CAMP is another regulatory protein that exerts various effects on cells, primarily in the intervention of cAMP-dependent protein kinase A. Intracellular levels of cAMP are regulated by a combination of receptor-mediated activation and inhibition of adenylate cyclase. Various studies show that the expression of a gene induced by TPA and cAMP can converge in a common path. To study the regulation of the expression of the AR gene, the Applicant has chosen to observe the effect of various activators or inhibitors of these two regulatory pathways on the production of AR-specific RNA in MCF-7 cells. .

Forsholine is an agent that activates adenylate cyclase and leads to an increase in intracellular cAMP. Administered to MCF-7 cells at 25 ddD for 4 hours, it induces marked stimulation of the AR mRNA, which suggests that the cAMP pathways also play a role in regulating the expression of the AR.

5.5. Amphiregulin related derivatives, analogs and peptides.

The production and use of derivatives, analogs and peptides related to AR are also envisaged and fall within the scope of the invention. These derivatives, analogs and peptides, which manifest growth-modulating activity, can find applications in the diagnosis, prognosis and treatment of a wide variety of neoplastic conditions. These derivatives, analogs or peptides can have increased or weakened biological activities compared to native AR and / or can increase or limit the variety of cells sensitive to the effect of AR as a growth regulator while remaining in the part of the invention. Likewise, the production and use of derivatives, analogs and peptides related to AR which manifest an increased or decreased activity of stimulating growth 32 and / or which increase or limit the variety of cells sensitive to stimulatory activity of growth by RA, can find useful applications, in particular, but not limited to, in the treatment of wounds and burns.

The AR-related derivatives, analogs and peptides which are the subject of the invention can be produced by various known means. The procedures and manipulations at the genetic level and at the protein level are within the scope of the present invention.

At the protein level, numerous chemical modifications could be exploited to produce derivatives, analogs or peptides analogous to RA according to known techniques, in particular, but not limited to, specific chemical cleavage by endopeptidases (for example cyanogen bromide , trypsin, chemotrypsin, protease V8 and the like) or exopeptidases, acetylation, formylation, oxidation, etc.

5.6. Production of anti-phiregulin antibodies.

It is also within the scope of the invention to produce polyclonal and monoclonal antibodies which recognize anphiregulin and related proteins.

Various techniques are known which can be applied to the production of polyclonal antibodies for AR epitopes. To produce these antibodies, various animal hosts can be immunized by injection of the AR protein or of a synthetic AR, in particular, but not limited to, rabbits, mice, rats, etc. Different adjuvants can be used to increase the immunity response depending on the species to which the host belongs, in particular, but not limited to, the adjuvant (complete or incomplete) by 33

Freund, mineral gels such as aluminum hydroxide, surfactants such as lysolecithin and Pluronic polyols, polyanionic agents, protids, oily emulsions, arapedan hemocyanin, dinitrophenol and adjuvants potentially useful for humans, such as BCG (Bacille Calmette-Guerin) and Corynebacterium parvum.

A monoclonal antibody for an AR epitope can be prepared by applying any technique which ensures production of the antibody molecules by cultured immortal cell lines. These are, without limitation, the hybridoma technique originally described by Köhler and Milstein (1975, Nature 256, 49 to 497), the more recent technique of human B cell hybridoma (Kosbor et al., 1983 , Immunology Today 4:72) and the EBV hybridoma technique (Cole et al., 1985, Monoclonal Antibodies and Cancer Therapy, Alan R. Liss, Inc., pages 77-96).

Fragments of the antibody, which contain the iodiotype of the molecule, could be produced by known techniques. For example, such fragments are, without limitation, the F (ab ') 2 fragment which can be produced by digestion of the antibody molecule in pepsin; the Fab 'fragments which can be produced by reducing the disulfide bridges of the F (ab') 2 fragment and the two Fab or Fab fragments which can be produced by reacting the antibody molecule with papain and a reducing agent.

Anticoprs against AR can find their applications in the qualitative and quantitative detection of mature AR as well as its precursors and its subcomponents, in the purification of AR proteins by affinity and also in the elucidation of biosynthesis, metabolism and function of AR. Antibodies to RA can also be useful as diagnostic and treatment agents.

5.7. Uses of amphirequline.

The bifunctional nature of amphiregulin opens up a wide variety of in vitro and in vivo applications. Any compound that contains AR or fragments and derivatives thereof which exhibit growth inhibiting and / or growth stimulating activity, both in isolation and in conjunction with other growth inhibiting or immunomodulating agents which are biologically active, can be used for the application of the process of the invention.

The location of the AR gene in a region involved in the differentiation of lymphocytes suggests that AR may play a role in the development of hematopoietic cells, in activation and in immunosuppression. This function is also confirmed by the homology between the untranslated AR3 region of the AR and similar regions of various other cytokines.

The compounds of the invention can be used for modulation of angiogenesis, bone resorption, immune response and neuronal synaptic and effector functions. AR can also be used to modulate the cascade of arachidonic acid. The enzymatic oxidation of arachidonic acid leads to a multitude of important products such as prostaglandins, thromboxanes, prostacyclins and leukotrienes. These products are extremely potent ubiquitous agents with many physiological effects including, for example, muscle contraction, platelet aggregation, leukocyte migration and gastric secretion. AR, AR-related molecules and their compositions may be especially useful for wound treatment, as well as for the diagnosis and treatment of cancers.

5.7.1. Wound treatment.

RA can be used in a method to treat wounds, such as skin wounds, corneal wounds and various other ruptures of the epithelium or stroma, such as chronic ulcers, burns, surgical incisions, wounds traumatic injuries and lesions to hollow organs lined with an epithelium, such as the esophagus, stomach, small intestine and colon, mouth and genitourinary tract. The method is based on the topical application of a treatment composition containing AR in a physiologically acceptable excipient.

The compositions of the present invention can be used to treat a wide variety of wounds, particularly in substance all skin wounds, corneal wounds and lesions to hollow organs furnished with an epithelium. Wounds suitable for such treatment include those resulting from trauma such as burns, abrasions, nicks and the like, as well as surgical operations, such as surgical incisions and skin grafts. Other conditions susceptible to treatment using the compositions of the present invention include chronic conditions, such as chronic ulcers, diabetic ulcers and other conditions that escape healing (trophic).

AR can be incorporated into physiologically acceptable excipients for application at 36 The affected location. The nature of the excipients can vary "a lot and depends on the intended application location. For application to the skin, a cream or ointment base is usually preferred; suitable bases are in particular lanolin, Silvadene ( Marion) (particularly for the treatment of burns), Aquaphor (Duke Laboratories, South Norwalk, Connecticut) and the like. If desired, AR compositions can be incorporated into bandages and other dressings to ensure continued exposure of the wound to the peptide. Aerosol applications may also be useful.

The concentration of RA in the treatment composition is not critical, but should be sufficient to induce the proliferation of epithelial cells. The compositions can be applied topically to the affected place, for example as eye drops in the eye or else in the form of creams, ointments or lotions on the skin. In the case of the eye, frequent treatment is desirable and is usually done at intervals of 4 hours or less. On the skin, it is desirable to continuously maintain the treatment composition in contact with the affected region during healing, with applications of the treatment composition two to four times a day or even more frequently,

The amount used of the polypeptide of the invention varies with the mode of administration, the use of other active compounds and other similar factors and is generally in the range of about 1 to 100 micrograms. The polypeptide of the invention can be used in a physiologically acceptable excipient, such as saline, phosphate buffered saline, etc. The amount of compound which is used is determined empirically based on the response of cells in vitro and the response of experimental animals to the polypeptides of the invention or to the compositions containing them.

AR derivatives can be used alone or in conjunction with other growth factors, inhibitors or immunomodulators.

5.7.2. Diagnosis and treatment of neoplastic conditions.

The compositions of the present invention can be useful for the diagnosis, prognosis and treatment of a wide variety of neoplastic conditions.

A large number of diagnostic applications of AR and related molecules is envisaged. In the practice of the invention, the polypeptides can be provided with a marker, such as a radioisotope, an enzyme, a fluorescent agent, a chemiluminescent agent, an enzyme fragment, a particle, etc. These compounds can be used to titrate the number of AR receptors on a cell. The identification of AR receptors is an indication of the cell's potential sensitivity to the biological effects of AR and related molecules. AR, AR-related molecules and / or antibodies thereto can be used in competitive assays to detect AR in media, particularly physiological media. Many known diagnostic tests are applicable.

The presence of AR and AR levels in body fluids and tissues can be directly or inversely related to the presence of certain cancers and their spread. Tests 38 making it possible to detect and / or quantitatively measure RA can find their application in the diagnosis and prognosis of cancer. (See section 5.6. Above).

The present invention also relates to the detection of AR mRNA. Assays that involve nucleic acid probes to detect sequences comprising all or part of a known gene sequence are well known and applicable. ARN mRNA levels may reveal emerging or existing neoplastic conditions, and therefore assays, which allow quantitative measurement of AR mRNA levels, are a valuable diagnostic tool.

In addition, malignant cells that express the AR receptor can be detected by using AR or AR-related molecules that are labeled, in a receptor binding test, or using antibodies against the AR receptor itself. Cells can be discerned based on the presence and density of AR receptors, which provides a means of predicting the sensitivity of these cells to the biological activities of AR.

AR can be used as an antineoplastic compound. For an in vivo application, the compositions of the invention can be administered by various routes, in particular, but not limited to, injection, infusion, topical route, parenteral route, etc. The administration can be carried out in any physiologically acceptable excipient, in particular physiological saline solution buffered with phosphate, physiological saline solution, sterilized water, etc. The AR and related molecules can also be encapsulated in liposomes and 39 can be conjugated to antibodies which identify and bind to the tumor or specific antigens of the cell so as to provide a means of "targeting" the compositions of the invention.

RA can be useful in vivo to induce terminal differentiation of tumor cells. These cells have deviated from the ordinary course of cell differentiation characteristic of normal cells and are susceptible to uninterrupted proliferation. On the contrary, normal cells differentiate into cells which are incapable, in most circumstances, of new cell division. Therefore, the ability of RA to reactivate the differentiation of normal cells in tumors and, as a last resort, to stop further tumor growth, may find valuable application in the treatment of tumors.

AR and related derivatives, analogs and their peptides can be used alone or with at least one other antiproliferative compound, including, for example, an interferon, TGF ß1, TGF ß2, tumor necrosis ß factor, o factor (tumor necrosis, etc. for the treatment of carcinomas sensitive to hormones which can attack a wide variety of organs, such as the lungs, breast, prostate, colon, etc. Carcinomas sensitive to hormones can also be treated by inducing the production of AR in carcinoma cells These inducers include, but are not limited to, estrogens, such as 17-ß estradiol, compounds having estrogenic activity, and compounds having antiestrogenic activity, such as Tamoxifen. Any combination of these compounds can also be used as an inducing agent.

40

The compounds of the invention can be used in vitro to inhibit the growth of cells or cell lines sensitive to RA rather than cells which are not sensitive. In this way, heterogeneous mixtures or lines; Cellular cells can get rid of unwanted cells in case the unwanted cells are sensitive to the growth inhibitory activity manifested by RA. For example, the compounds of the invention can be used in vitro to suppress malignant cells from the marrow for autologous marrow transplants and to suppress or inhibit the proliferation of malignant cells in the blood before retransfusion.

The most effective concentration of AR to inhibit proliferation of cells of a given variety can be determined by adding different concentrations of AR to the tumor cells in question and monitoring the degree to which cell proliferation is inhibited. The most effective concentration of inducers taken in isolation and / or combinations of inducers can be assessed by monitoring the production of RA in carcinomatous cells.

6. Example; Production, purification and characterization of 11 human amphiregulin.

The following subsections describe the purification and characterization of amphiregulin produced by MCF-7 cells treated with TPA.

6.1. Materials and processes.

The following methods were applied to induce AR synthesis in MCF-7 cells to prepare and characterize substantially pure AR.

41 6.1.1. SDS-polyacrylamide gel electrophoresis.

The proteins were analyzed on SDS-PAGE gel plates (normal or mini of the Bio-Rad system) as described (Laemmli, 1970, Nature 227: 680 to 685), and were detected by silver staining (Merril et al., 1981, Science 211: 1487-1439).

6.1.2. Isoelectric focusing.

Isoelectric focusing was carried out essentially as described in the Isolab technical manual for Resolve isoelectric focusing gels. In summary, the samples were applied to pre-molded agarose gels (85 mm x 100 mm, pH 3 to 10) and were focused using a Resolve model FR-2500 isoelectric focusing apparatus using a Bio-Rad model 3000 Xi computer-controlled electrophoresis supply. IEF Bio-Rad standards were focused parallel to the samples and stained and the dry gel was exposed to Kodak X-Omat AR film with Cronex DuPont lighting and an intensifying screen.

6.1.3. Protein iodization.

Proteins were labeled using 125i using the chloramine T method (Barridge, 1978, Methods Enzymol. 50: 54-65).

6.1.4. Protein determination.

Protein concentrations were determined by absorption at various wavelengths (210 to 280 nm). The methods of Lowry (Lowry et al., 1951, J. Biol. Chem. 193: 265 to 275) and Bradford (1976, Anal. Biochem. 72: 248 to 252) were used with a serum albumin standard beef (BSA).

6.1.5. Determination of the 125i-EGF binding.

The fixation assays were performed either in 48-well tissue culture plates when freshly cultured A431 and / or 42 formalin-fixed cells were used as described (Carpenter et al. 1979, J. Biol. Chem. 254, 4484-4891; Shoyab et al., 1979, Nature 279: 387-391;

DeLarco et al., 1980, J. Biol. Chem. 255: 3685 to 3690), or by immobilizing plasma membranes on poly (vinyl chloride) plates with 98 wells as described (Kimball and Warren, 1984,

Biochem. Biophys. Acta 771: 82-88).

6.2. Production of 11 amphiregulin.

6.2.1. Collection of conditioned medium from MCF-7 cells treated with TPA.

MCF-7 cells were grown in Corning T 150 tissue culture flasks in a total volume of 25 ml of 50% IMDM (modified Dulbecco medium from Iscove) + 50% DMEM containing 0.6yjLg / ml of insulin and 15% heat-inactivated fetal bovine serum (complete medium for MCF-7). About 1 x 10 ^ cells were added to each flask and incubated at 37 ° C with 5% CO 2. On day six, all of the medium was collected and 20 ml of complete fresh MCF-7 medium containing 100 ng / ml of TPA was introduced into each flask. The medium was collected 48 hours later and rinsed each vial with 15 ml of 50% IDMM + 50% DMEM (serum-free medium) and 25 ml of fresh serum-free medium was introduced into each vial before conducting incubation at 37 ° C with 5% CO 2 · Ob collected the serum-free medium conditioned four days later, centrifuged to separate the debris and stored at -20 ° C. 25 ml of serum-free medium was again introduced into the vials and the conditioned serum-free medium was collected every three or four days. Growth inhibitory activity (GIA) was assayed on an aliquot of conditioned medium from each harvest on A431 cells of human epidermoid carcinoma 43. Usually, from each batch of TPA-treated MCF-7 cells, three to four crops of conditioned medium were collected.

6.2.2. Isolation of raw RA.

About 4500 ml of conditioned medium was thawed and centrifuged at 4 ° C for 15 minutes at 3500 rpm. The supernatant was concentrated in a 2 liter Amicon concentrator using a YM10 membrane (Amicon) at 4 ° C. When the volume retained reached approximately 200 ml, 1000 ml of cold Mili-E-Q water was added and the mixture was reconcentrated to 200 ml.

The concentrate was removed and transferred to a 250 ml pre-cooled Corning centrifuge flask. With stirring, concentrated acetic acid was added slowly to a final IM concentration of acetic acid. The mixture was allowed to stand for 1 hour at 4 ° C and centrifuged for 20 minutes at 40,000g at 4 ° C in a Sorval RC-5B centrifuge. The supernatant was collected and stored at 4 ° C. The pellet was suspended in 30 ml IM acetic acid and recentrifuged as described above. The supernatant was again carefully collected and pooled with the first supernatant, then dialyzed against 17 liters of 0.1M acetic acid in Spectrapore No. 3 dialysis tube (cut molecular weight about 3000). The dialysis pad was changed three times over the course of two days. The retained product was freeze-dried. The dry product was collected, pooled, then weighed and stored at -20 ° C until further use. This material is here called raw powder.

44 6.3. Purification of amphirequline.

6.3.1. Liquid chromatography with phase inversion.

950 mg of crude powder (from about 9 liters of conditioned serum-free medium) was suspended in 300 ml of 0.1% TFA (trifluoroacetic acid) and centrifuged for 20 minutes at 7000 g . The supernatant was carefully collected and placed on a preparative C18 column (2.54 cm x 27 cm; 125-105 micrometers; Waters) brought to equilibrium with 0.1% TFA in 1 'water. The chromatographic support was suspended in acetonitrile (MeCN) with 0.1% TFA and the dispersion was poured into a column with a diameter of 2.54 cm, then the column was washed with 400 ml of MeCN with 0.1% TFA, after which it was brought to equilibrium with 600 ml of 0.1% TFA in water. With a flow rate of 4 ml per minute, the chromatography was carried out at room temperature. The column was washed with 650 ml of 0.1% TFA in water. The top fraction and the washing fractions were collected together. A stepwise elution was then carried out as follows: (1) 650 ml of 20% MeCN / H2O with 0.1% TFA, (2) 650 ml of 40% MeCNC / H2O with 0.1% of TFA, (3) 650 ml of MeCN 60% / H2O with 0.1% of TFA and (4) 650 ml of MeCN 100% with 0.1% of TFA. An aliquot was taken from each fraction and checked for GIA. About 77% of the total GIA activity was observed in the overhead and wash fractions.

Head and wash fractions were injected isocratically into a preparative Partisil 10 ODS-3 column (10 micrometers, 2.2 cm x 50 cm, Whatman) connected to a high performance liquid chromatography system (HPLC) ( Waters). The flow rate was adjusted to 4 ml per minute. After admitting the sample to the column, the column was washed with 250 ml of 0.1% TFA in water. A linear gradient was maintained between the primary solvent, 0.1% TFA in water, and the secondary solvent, acetonitrile containing 0.1% TFA, the gradient conditions being: 0 to 15 % in 10 minutes, 15 to 15% in 30 minutes, 15 to 25% in 150 minutes, 25 to 65% in 5 minutes and 65 to 100% in 10 minutes. All solvents were HPLC grade. 14 ml fractions were collected and the GIA was assayed in aliquots of each fraction. Two large peaks of activity were observed (Fig. 1). The peak eluted first was further purified and characterized (eluted between 20 and 23% acetonitrile).

Fractions 47 to 62 were pooled. 224 ml of 0.1% TFA in water was added to the fraction thus formed. The mixture was injected isocratically on a semi-preparative jiA.-Bondapack-C18 column (7.8 mm x 300 mm, Wauters) at a flow rate of 2 ml per minute at room temperature. The conditions of the linear gradient were: 0 to 17% in 10 minutes, 17 to 17% in 30 minutes, 17 to 25% in 320 minutes and 25 to 100% in 40 minutes. The flow rate was 1 ml per minute during the gradient, the fractions collected being 4 ml. Aliquots were taken for the GIA assay. Two major peaks of activity were observed eluted at acetonitrile concentrations of approximately 20% and 21%, respectively (Fig. 2).

Fractions 49 to = 53 were pooled. 20 ml of 0.1% TFA was added to the fraction thus formed. The mixture was applied isocratically on a ^ u.-Bondapack-C18 column (3.9 mm x 300 mm, Wauters) at a flow rate of 1 ml per minute at room temperature. The gradient conditions were 0 to 18% in 10 minutes, 18 to 18% in 30 minutes, 18 to 25% in 280 minutes and 25 to 100% in 20 minutes, the flow rate being 0.4 ml per minute and the fractions collected being 2 ml. Most of the activity emerged from the column at an acetonitrile concentration of approximately 21.5% (Fig. 3A). Fractions 54 to 59 were pooled (Fig. 2) and chromatographed exactly as described above with respect to Fig. 3A. Most of the activity eluted from the column at an acetonitrile concentration of about 22.2% (Fig. 3B).

6.3.2. Gel permeation chromatography.

Fractions 35 to 38 (Fig. 3A) were individually concentrated to about 70 µl by means of a Speed-Vac concentrator (Savant), into which an equal volume of acetonitrile containing 0.1 was introduced. % of TFA. This sample of 140 μl was injected into two Bio-sil TSK-250 columns (each 7.5 mm × 300 mm, Bio-rad) mounted in tandem. Elution was carried out isocratically with a mobile phase of 50% acetonitrile / water with 0.1% TFA at room temperature. The flow rate was 0.4 ml per minute with a recording speed of 0.25 cm per minute and 0.4 ml fractions were collected on aliquots from which the GIA was assayed. The chromatographic profiles of fractions 35, 36 and 37 (Fig. 3A) constitute Figs. 4A, 4B and 4C, respectively.

Fractions 41 and 42 were pooled (Fig. 3B) and concentrated to 70 microliters, then subjected to gel permeation chromatography as described above. The chromatographic profile is indicated in FIG. 4D.

i 47 6.4. Cell growth measurements.

6.4.1. Measurement of the modulation of cell growth according to the incorporation of the ^ - ^^ I-deoxy-uridine in DNA.

The assays were performed in 96-well dish plates (Falcon 3072). Human epidermoid carcinoma cells of the vulva (A431) were used as cells indicative of growth inhibiting activity (GIA) and human foreskin fibroblasts (Sadamoto) as cells indicative of growth stimulating activity ( GSA). 3.5 x 10 ^ cells in 50 ^ liters of DMEM supplemented with 5% heat-inactivated fetal bovine serum (FBS), penicillin / streptomycin (3) were introduced into all wells except the peripheral wells. PS) and glutamine (test medium). 50 microliters of PBS were introduced into the peripheral wells. 50 microliters of test sample were introduced into the test medium 3 hours later in the test medium, the control cups receiving only 50 microliters of test medium. Three wells were used for each concentration of the test sample. Plates were incubated at 37 ° C for 2 to 3 days. Then, 100 microliters of a solution of ^ - ^^ I-iodo-2 deoxyuridine (4 Ci / mg - 0.5 mCi / ml (2 microliters per ml in test medium)) are introduced into each well. incubated the plates at 37 ° C. After 4 to 6 hours, the medium was aspirated out of the wells which were washed once with 200 microliters of PBS. 200 microliters of methanol were then introduced into each well and the plates were incubated for 10 minutes at room temperature, then the methanol was removed by suction. 200 microliters 48 of IM sodium hydroxide were added to each well 48 and the plates were incubated for 30 minutes at 37 ° C. The sodium hydroxide was removed with Titerteh buffers (Flow Labs). Buffers were transferred to 12mm x 75mm plastic tubes to determine quantitatively the incorporation of 125I-IUdR with a gamma counter.

6.4.2. Colonies measurement on soft agar.

A 0.5 ml base layer of 0.5% agar (Agar Noble; Difco Laboratories, Detroit, Michigan) was added in DMEM containing 10% heat inactivated FBS in Costar tissue culture plates to 24 cups. 0.5 ml of 0.3% agar containing the same mixture of medium and FBS, 5-10 x 102 test cells, was applied over the agar base coat, as well as the factors to be examined in various ways. concentrations. The plates were incubated at 37 ° C. in a humid atmosphere at 5% CO 2 in air and they were re-fed after 7 days by the addition of 0.5 ml of 0.3% agar containing the same medium and the same concentrations of substances to be examined. Unbound and unstained colonies were counted and the number of colonies over 20 cells was determined between days 7 and 14.

6.4.3. Measurement of cell growth inhibition.

A431 cells were spread at 2.1 x 10 ^ cells per well in 0.3 ml of medium (DMEM with 10% FBS, P / S and glutamine), in Costar plates with 24 wells (approximately 2 cm2 per well) and incubated for 2 hours at 37 ° C. Then, in each well, test samples were introduced in various concentrations, in. two copies, in 0.3 ml of medium. The sample-free medium was introduced into the control wells. The plates were incubated at 37 ° C for 72 hours. The medium was then evacuated by suction and the cells were washed with 1 ml of PBS per well, then the cells were released with 0.5 ml of trypsin (0.5%) - EDTA (0.5 mM) and they were counted using a Coulter counter.

6.5. Analysis of the primary structure of AR.

6.5.1. Reduction and S-pyridylethylation.

RA (20 to 30 micrograms) was dried in polypropylene microcentrifuge tubes of

1.5 ml, then it was suspended in 100 microliters of 3M urea in TRIS.HCl, 0.05 M of pH 7.5. 4 microliters of 2-mercaptoethanol was added to the mixture which was stirred and purged with nitrogen, then incubated at 25 ° C. After 2.5 hours, the mixture of 4.5 microliters of freshly distilled 4-vinylpyridine was added and each tube was purged again with nitrogen before 2 hours of incubation at 25 ° C. The reaction mixture was acidified to pH 2.0 with 10% TFA. The S-pyridylethylated AR was purified by reverse phase HPLC on a yu.-Bondapak C18 column (3.9 mm × 300 mm, Waters). The acetonitrile concentration was increased linearly (1% per minute) for 55 minutes, the flow rate being 1 ml per minute and the primary solvent being 0.1% TFA. AR

S-pyridylethylated (SPE-AR) eluted at approximately 26.5% acetonitrile, i.e. an acetonitrile concentration approximately 4% higher than that for untreated AR.

6.5.2. Treatment with N-glycanase.

AR or S-pyridylethylated AR (10 to 20 micrograms) was dried in 1.5 ml polypropylene centrifuge tubes, and then suspended in 80 microliters of phosphate buffer 0, 1 M pH

5.5 and added 10 to 20 microliters of N-glycanase (0.25 units per microliter, Genzyme), after which the mixture was incubated at 37 ° C before adding 50 0.9 ml of 0.2% TFA to the reaction mixture and to inject the sample at a flow rate of 1 ml per minute on an ultrapore RPSC phase inversion HPLC column (4.6 mm × 75 mm, Altex) previously brought to equilibrium with the primary solvent which is 0.1% TFA. Acetonitrile with 0.1% TFA was the secondary solvent. The acetonitrile concentration was increased linearly (0.4% per minute) for 85 minutes at room temperature. N-deglycosylated AR and S-pyridylethylated N-deglycosylated AR eluted at acetonitrile concentrations of approximately 16.5% and 20.5%, respectively.

6.5.3. Enzymatic cleavage of S-pyridylethylated AR treated with N-glycanase.

The cleavage was carried out with the endopeptidase Lys-C in 60 microliters of acetic acid-TRÏS 0.1M buffer of pH 8.0, at 25 ° C for 16 hours, the enzyme / substrate ratio being from 1 to 5 (in weight). Digestions were made with 11endopeptidase Arg and protease V8 of S. aureus in 80 microliters of 0.05M TRIS.HCl, pH 8.0 and 0.1M ammonium bicarbonate, at 37 ° C. for 16 hours, again with an enzyme / substrate ratio of 1 to 5.

6.5.4. Chemical break.

For the cleavage to the methionine residue by CNBr, 5 micrograms of S-pyridylethylated AR treated with N-glycanase were dried in 1.5 ml polypropylene microcentrifuge tubes, then the compound was suspended in 75 microliters of 70% formic acid before adding 25 microliters of CNBr solution (25 mg per ml in 70% HCOOH), mixing and purging the tube with nitrogen. The reaction was carried out for 22 hours at 25 ° C in the dark. The reaction mixture was diluted to 1.1 ml with water and brought to dryness in a Speed-Vac concentrator. The dried sample was raised in suspension in 20 microliters of 2-aminoethanol and allowed to stand for 10 minutes at 22 ° C before drying in vacuo. The mixture was suspended in 0.9 ml at 0.2% TFA.

6.5.5. Isolation of peptides.

The peptides were separated on a C8 phase inversion HPLC column (4.6 mm x 100 mm, Whatman) connected to an HPLC system (Waters). The acidified sample (pH 2.0) was deposited on a column brought to equilibrium with 0.1% TFA (primary solvent) at the flow rate of 1 ml per minute and the column was continued to wash with about 15 ml of 0.1% TFA. Linear gradients were used between the primary solvent and the secondary solvent, namely acetonitrile with 0.1% TFA, the gradient conditions being: 0 to 50% in 125 minutes at the flow rate of 0.5 ml per minute.

6.5.6. Amino acid analysis.

The dried samples were subjected to hydrolysis with HCl at constant boiling point (5.7M, Pierce) containing 1% (by weight) of phenol, under reduced pressure, in a glass hydrolysis flask capped with Teflon (Pierce), at 105 ° C for 16 hours. The hydrolysis products were dried in a Speed Vac concentrator (Savant Instruments) and the derivatives were formed with phenyl isothiocyanate for 20 minutes at room temperature. Phenylthiocarbamylic derivatives of amino acids were studied by phase inversion HPLC on a C18 column (4.5 mm x 250 mm, IBM). The linear gradient between the primary solvent, namely 0.15M sodium acetate pH, was maintained

6.4 with 0.05% triethylamine brought to pH 6.4 with acetic acid, and the secondary solvent, namely 52% acetonitrile, at a flow rate of 1 ml per minute at 38 ° C, 6.5.7. Determination of the amino acid sequence.

Peptide sequences were determined using an Applied Biosystem model 475 gas phase sequencer as described (Hewick et al., 1981, J. Biol. Chem. 256: 7990-7997). Three pre-cycles of Edman degradation were performed before applying the sample in each test. 25% TFA was used to convert the triazolinic derivatives to phenylthiohydantoine amino acids. The identification of these latter acids was carried out continuously on a model 12 OA PTH analyzer (Applied Biosystem), as described (Hunkapiller and Hood, 1983, Science 219: 650 to 659).

6.5.8. Various treatments.

For the present experiments (Table I), 50 to 100 units of AR were used, either semi pure (stage C18 with phase inversion, semi preparative) or apparently pure. N-glycanase and 1O-glycanase were acquired from Genzyme Corp., Boston and the other enzymes from Boehringer Mannheim Biochemicals or Worthington Biochemicals. The treatments were carried out in 50 to 200 microliters of a suitable buffer. Enzymes were used in excess, usually 5-20% (by weight) of the substrate concentration. After the treatments, the GIA was assayed in the samples by eliminating the troublesome substances by various analytical means. In most cases, the pH of the sample was adjusted to about 2 with 10% TFA. The samples were deposited on an ultraporous phase inversion RPSC C3 column (4.6 mm χ-75 mm, Altex) brought to equilibrium with 0.1% TFA. Washing of the column was continued with primary solvent, namely 53 of 0.1% TFA. Elution was then started with acetonitrile containing 0.1% TFA as a secondary solvent. The gradient conditions were: 0 to 50% for 0.2 minutes and 50 to 50% for 19.8 minutes, the flow rate being 0.5 ml per minute and the fractions collected being 2 ml. Aliquots were taken for the GIA assay. AR activity usually eluted in fraction 4.

6.6. DNA binding studies.

As described (Herick and Alberts, 1971, Methods in Enzymology, 21: 198), AR binding was carried out on cellulose-native calf thymus DNA, denatured calf thymus DNA and cellulose. The celluloses (10 mg) were rehydrated in 500 microliters of loading buffer (20 mM Tris, pH 7.4, 10% glycerol, 1 mM EDTA, 1 mM β-mercaptoethanol, BSA at 100 micrograms per ml and NaCl 50 mM). The reactions were performed in duplicate in 1.5 ml Eppendorf tubes with 300 microliters of hydrated cellulose (packed volume) washed five times with loading buffer. Next, 0.2 ng I-AR (specific activity of 425 microcuries per microgram) or other protein labeled in 250 microliters of loading buffer was added to each tube. The samples were mixed and incubated at 4 ° C for 20 minutes with periodic shaking, then washed five times with 200 microliters of loading buffer for each wash. We carried out the elution step by step with

0.1M, 0.15M, 0.25M, 0.6M, IM and 2M NaCl in loading buffer (five times with 200 microliters for each concentration). The specifically fixed material has been defined as being that eluted at a NaCl concentration of 0.25M or more, the material eluting at an NaCl concentration of 0.15M if not less 54 being considered to be non-specifically fixed.

6.7. Results.

6.7.1. Production of AR by MCF-7 cells treated with TPA.

MCF-7 cells grown on a plastic substrate exhibit two typical epithelioid features; cells are small and polygonal. TPA induces profound morphological changes in proportion to the dose applied. After treatment with TPA, MCF-7 cells lose their well-defined morphology, become round and spread out. Depending on the dose applied, the TPA causes a reduction in the number of cells and an increase in the volume of the cells, in comparison with the untreated cells.

The serum-free conditioned medium from MCF-7 cells did not contain any detectable growth inhibiting activity for A431 cells. However, serum-free supernatants separated from MCF-7 cells treated with TPA for two to three days have been shown to inhibit proliferation of A431 squamous cell carcinoma cells. The optimal induction of inhibitory activity was observed between 5 and 100 ng per ml of TPA. Even after elimination of TPA, MCF-7 cells treated with TPA secreted this activity in serum-free medium for at least eight days. Although a reduction in growth inhibitory activity was observed over time after the suppression of TPA, these results indicate an amphiregulin, the expression of which is induced or increased in the MCF-7 cells which have been treated by TPA.

55 6.7.2. Initial characterization.

Table I summarizes the effects of various physical, chemical and enzymatic treatments on the antiproliferative activity of amphiregulin. GIA (growth inhibiting activity) of amphiregulin appeared resistant to treatment with IM acetic acid, IM ammonium hydroxide, 6M urea, 0.01M sodium metaperiodate, heating to 56 ° C for 30 minutes and on treatment with neuraminidase, N-glycanase, O-glycanase, lipase and phospholipase A2, C or D. However, the activity is sensitive to heating for 10 minutes at 100 ° C., to reduction, reduction and reaction with 4-vinylpyridine, as well as digestion with proteinases such as trypsin, endopeptidase Lys-C, 11endopeptidase Arg and endopeptidase Glu (V-8). These results suggest that amphiregulin is a protein containing cysteine units in one or more disulfide bonds which are essential for its biological activity. This protein may contain units of oligosaccharides and / or lipids which are not compulsory for biological activity.

56

TABLE II

Effects of various treatments on GIA for amphiregulin

Resistance Treatment%

Acetic acid IM (2 h at 4 ° C) 102 NH4OH IM (2 h at 4 ° C) 97 56 ° C for 30 minutes 96 100 ° for 10 minutes 5

Urea 6M (2 h at 25 ° C) 82 2-Mercaptoethanol 0.2M (2.5 h at 25 ° C) 6 2-Mercaptoethanol + 4-vinylpyridine (2.5 h at 25 ° C) + (2 h at 25 ° C) 1 Sodium metaperiodate (0.01M) (6h at 4 ° C, dark) 86 TPCK - trypsin (6h at 37 ° C) 3 TPCK-trypsin + soybean trypsin inhibitor (6h at 37 ° C ) 82

Soybean trypsin inhibitor (6 h at 37 ° C) 109

Endopeptidase Lys-C (20 h at 25 ° C) 5

Endopepdidase Arg (16h37 ° C) 4

Endopepdidase Glu (4 p.m. to 37oC) 6

Neuraminidase (4 p.m. at 37 ° C) 103 N-glycanase (4 p.m. at 37 ° C) 90 O-glycanase (4 p.m. at 37 ° C) 102

Neuraminidase + N-glycanase (4 p.m. at 37 ° C) 94

Neuraminidase + O-glycanase (16 h at 37 ° C) 105

Phospholidase A2 (6h at 37 ° C) 82

Phospholidase C (6 h at 37 ° C) 95 57

Phospholidase D (6h at 37 ° C) 80

Lipase (6h at 37eC) 111 6.7.3. Purification of amphiregulin.

A summary of the purification of amphiregulin is presented in Table III. A purification of 1842 times with a yield of 5.1% was obtained for the fraction TSK 250-1 and a purification of 2270 times with a yield of 1.7% was obtained for the fraction TSK 250-11. This process is reproducible. Amphiregulin has been purified with similar results on four different occasions. The specific activity of purified amphiregulin is between 2.6 and 3.4 x 106 units per mg of protein. Purified amphiregulin from the TSK 250-1 column with an activity of approximately 3.0 x 106 has been used for structural determinations and other biochemical studies.

Note for Table III.

Details are given in section 6.1.3. and its subsections, above. One unit of GIA is the amount of factor required to inhibit the incorporation of 125I-deoxyuridine by 50% into A431 cells.

58 l n ntoh ο rö rn _, θΓ "- r ~~ .........

U kοο ™ * ”* · * (N m μ co ο μ.ηπ 1-1 '_η cn cn coco“ · ί \ 0> Ρ Π, λ ri ^ σινο oot'- co mmC' ν 3 ^ 3. 00 Η CN ΗΗ .....

0 <Η ι — Ι ι — I Η CN

cl Q) Ë Ο ™ ^ Η Γ ~ »f 'Η Η ^ C0 Γ" · Φ tfP Ο w, * · * "*" *> ·> * "·" "> 0 Η οο οο in.-ι cn --Η η • rj 1 ^ φ <d pi Η α N · »*

CO

0) ω ι .3 'Φ GX ο οο Ν' co in Π -P Ο1 51 oo n * σ> h öc Hr-f-com π Ή · Η u 6 Cl 00 Ν 'H Ν' CO 00> · * * 'n * σι ·. ^ '- ·. in cn Ν 'oo cn, ¾ -h -h .. u * "" in cd h co mn · r · "σι co co £ +) OS) IÖ Η H CN 00 σι O co t" · CN CO σ 00 ι O '(1) dd D-! H CN 00 CN CN ·. ··· p * £ ÇuZj CN CN (N ON 00 i 3

H

H H

H 00

(1) I

DO O

rij H

S O * X 00 CN, 0 S. j rtî co in „h m ο Ό omrocNoo

®UrHm 2 S “^ ^ ^ ^ ^ iH 00 CN CN

H in 0 "il) ^ 1-1

U -P

* ri eH

Ή G

• M 3 U 3 a co

(0 CD O O

-tG '-' OO.-l in ^.

• H CTl 10 O 00 OC U0 CN 00 ’Ό" LO CO

CD 'CD S · · CN 00 00 Ν' Ι '"' ιη-_Γ * 'Φ +) CO CN · 00 CN CN Ή ü, co' J r- vo θ'— · H 00 CN CN Ή Ό) P σ m S Cu 3 te ^ S Λ

S J Ο Ο m ~ 0 CO CO CN 00 CN

Hg o m ο ^ co vo .....

Q ^. 00 OC CN CN CN Γ0ΗΗΟΉ> • ·

Qj H

.-P 'CD · ft) Nj 0Q U H

G n, ® P ι <. | »Ss ι III ι — ι

Ο η) Cl) Ctj 1.1. (-. '0) j. j. G HMH I

h ®'® <dd "° v t & ^ v v S O

g £ ®iSÎ 8 “· S | 'g 3 * g * · fc Γ, n)>' Ηω> 2 co u & G co G ^ Eh

J -r-t -H

59 6.7.4. Analysis of amphiregulin (AR) and S-pyridylethylated amphi-regulin (SPE-AR) by HPLC.

Gel permeation chromatograms of AR and SPE-AR on a TSK 250 column (7.5 mm x 500 mm, Biorad) constitute Figs. 5A and 5B, respectively. The activity eluted at the same time as the protein peak (5A). The molecular weights of AR and SPE-AR, evaluated from the data in Fig. 5, are about 14,000 and 17,000, respectively. The molecular weights calculated for the truncated AR and Ar structures shown in Fig. 12 are around 8,500 and 9,100 daltons, respectively.

The AR and the SPE-AR elute in the form of symmetrical single peaks of an ultra porous RPSC C3 phase inversion HPLC column (4.5 mm x 75 mm, Alt ex) (Fig. 6A and B). The GIA elutes with the protein peak (Fig. 6A). SPE-AR elutes at an acetonitrile concentration of approximately 21%, which is approximately 4% more acetonitrile than for the untreated protein. These results suggest that the AR is rich in cysteine residues which are modified by 4-vinylpyridine, which makes the AR more hydrophobic and causes it to elute at a higher concentration of acetonitrile.

6.7.5. SDS-PAGE analysis of the AR.

Fig. 7A illustrates an analysis of AR, N-glycanase-treated AR (NG-AR), SPE-AR and N-glycanase-treated SPE-AR (NG-SPE-AR) in 15% acrylamide under reducing conditions (Fig. 7A). AR and SPE-AR migrate into the gel as a single broad diffuse band where the median relative molecular weight is approximately 22,500 with an interval of approximately 20,000 to 25,000 (columns 1 and 3) . Treatment of AR and SPE-AR with N-60 glycanase accelerates the migration of these proteins. NG-AR and NG-SPE-AR migrate to the single band state with median molecular weights of 14,000 and 14,500 daltons, respectively. Similar results were observed when the proteins were passed through the 15% gel under non-reducing conditions (data not shown). The treatment of RA with neuraminidase, O-glycanase or neuraminidase and O-glycanase together does not modify its electrophoretic mobility in the gel under reducing or non-reducing conditions. Therefore, AR is a single chain glycoprotein containing one or more N-linked oligosaccharide chains which are not necessary for growth inhibiting activity of AR in vitro.

Treatment of NG-AR or AR with phophoslipase D (PLD) (cabbage) leads to a reduction in GIA up to 80% of the control. The phospholipase D-treated NG-AR was subjected to a phase inversion HPLC on a C3 column and the purified active peaks were analyzed on 20% SDS-PAGE (FIG. 7B). A single band of Mr 5600 has been observed. These results show that phospholipase D or one or more contaminating proteases of the preparation of PLD convert the AR into one or more active fragments.

6.7.6. Isoelectric focusing (IEF) analysis of AR.

The IEF analysis of the 125i-labeled AR performed as described in section 6.1.1.2., Above. The AR focuses in a single wide band with a PI between 7.6 and 8 (Fig. 8). The NG-AR focuses in a single band with a PI of around 8.05. AR is therefore a basic single chain N-linked glycoprotein.

61 6.7.7. Biological properties.

The inhibition of the incorporation of 125d deoxyuridine into the DNA of A431 cells by various concentrations of the purified AR is illustrated in FIG. 9A. At about 0.2 ng per well, 50% inhibition of DNA synthesis was observed. Therefore, a 50% inhibition of DNA synthesis in A431 human squamous cell carcinoma cells was observed at a concentration of approximately 0.13 nM of the pure protein. It should however be observed that the AR GIA depends on the experimental conditions, for example the number of cells per well (cell density), the time of application of the factor, the duration of the treatment, the serum concentration and various other variables.

When A431 cells were cultured in the presence and absence of various AR concentrations, cell growth being monitored by direct cell count, it was observed that AR inhibits proliferation of A431 cells in proportion to the dose applied (data not shown). Therefore, the degree of incorporation of ^ i-deoxyuridine into DNA is a good measure of cell growth.

1 ΛΡ

The stimulation of the incorporation of deoxyuridine into the DNA of human foreskin fibroblasts (Sadamoto) by various concentrations of purified AR is illustrated in FIG. 9B. Twice stimulation of the incorporation of 125i-dexyxyidine was observed at approximately 0.7 ng per ml, ie an AR concentration approximately 0.05 nM. The maximum response was approximately six-fold stimulation of the incorporation of 125I-deoxyuridine into these fibroblasts. Therefore, AR acts as a growth factor on human fibroblasts, even in the presence of 5% FBS.

The effect of RA on the incorporation of 125j-deoxyuridine into the DNA of various human tumor and non-tumor cell lines and certain non-human cell lines has been studied. These data are collated in Table IV. RA inhibits the growth of various clones of A431 cells, human breast carcinone HTB 132 cells, human ovarian teratocarcinoma cells HTB 1575, human cervical squamous cell carcinoma cells CRL 155, human ovarian papillary adenoma cells HTB 75 and human breast adenocarcinoma cells HTB 26. It shows no significant effect on various other cells (Table III). AR simulates the incorporation of ^ 25j_ deoxyuridine into various lines of human fibroblastic cells, human pituitary tumor cells CRL 7396, human ovarian adenocarcinoma cells HTB 77, African monkey kidney cells BSC 1 and kidney cells from SA6 rats. RA does not significantly affect the proliferation of T, B or endothelial cells. RA does not suppress the human T or cytotoxic T cell responses in mixed leukocyte cultures (MLC).

AR is a monomeric protein which requires disulfide bonds between chains for its activity, which does not require glycosylation for its activity, which is stable in a moderately acidic or basic medium and which is stable during heating for 30 minutes at 56 ° C. RA loses complete biological activity when reduced, when heated for 5 minutes at 100 ° C or when subjected to digestion with trypsin, endopeptidase Lys-c, endopeptidase ARG or the endopeptidase GLU.

63 ® s 0 eu ό> · ρ 3 ΰ S a ..........

> (0 .3'S QflÔOODDOOOwÏÏOÛQiflin • H H λ j ........ · · 03 3 ?, Γ '· · -P 3 £ g ™' Z'z S5 “5“ i 10 in <D (1)

H φ U

3 'S-dsu

S 5-h.hH £ 3 “2 îî> q * q * ° * q * q * q * Q ° Q Q o O

φ 3âSc ** ’Ü ïi ü 3 ^ g ^ G -P Φ Ο Ό

S V

S υ ". 3, _i 03 1 3 φ 0 ° -P 3 X H Lf> Φ

> n ro in (N O -H

H w ^ H M-) p C ft Ή

D 5 a a) - φ "J

rtj Φ a in ÎUO '-' O- W,. > U r »Μ O 2 -P d Ό h3q hcs r» oo ω> x o-H ^

'Rn DooHCQinvO CO) Ufflß g £ 3 -P

<> -h (0 E-ι cn m> ^ UW>) t "<0 Ό Φ ^ E-irt, U ffi t ^ linft '-' ΡΓ'ΌΟΟΦ 2 m (Om-rlCQCQ fc rtj ^ P <0 ^ <0OO> .3 $ h H> CBEH Pi U ® IdHO) fflCOÜ 1 Π ffi P Φ td NS ® 3 ai ft in 3 Φ ~ b φ Φ · Η P σο- Φ 3 (H ® Ό tj! D Π · Ϋ Φ OP d Φ Φφ-Η-Η'φΗΜΌ ΦΦ c Π P (0 Φ P μ ^ "4) Φ30 Jfiüüffl -ri Z Φ · ΗΦ> · Η · Η · Η · γΗ β ιίΠΗ'Φ (BÄ Φ 3 3 (1) 0) J π * Ö (0 Ό Ο P (0 (0 (0 Ο-PttJH g 3 U 'P & ΟιΗ U3 d .b IH ËH 10g ggg ai Φ ρ σ * Μ'φ'φ (0 Φ rtj • S, dog ο φ> g gHHoeoo-HH (0M ^ -H oco g- d E i0 e μ o ic (0 (03po - ihE (0> djq.h

5.b P g P -P g g Ë ft OirH 03 Μ U Ο 'Φ CC

_ro * Φ Φ (0 Φ <0 φ G Φ -H 3 3 JG -H -H

. rn Ό Φ Ό H g φ φφφφϋ, 0 (0, θ> (D00 + I ΦΦ

M S 'Hg ^ HOHuiggÊE H h-rlM g O PP

m bj & O Οι-P d0r "000030-PÛ <® O 01 ai Ό $ d Ό fi'J) QrH droddÜdO-P 'Φ 0 d Φ Φ d 3Φ u dd -ri (0 U -P ri! · γΗ -H -h -H 03 m -H + J + J Φ 'Ü' Ü m Zj Φ Ο Φ 0 <POUUU Ο Φ (0 ΦΦϋ0303 "} n, g P g (0ΡΦΡΡΡΡεΗΟ3εεΡ (0ι003θΟΟ3Ο1 in rb Ο <0Ο <1 > Ο <0Η <0 <0ί0 <0Ο, ΰΦΟΟ <0Ηι-) ΦΕΗφΦ

J: u cudgououuoocOHccuaaHroHH

ïï -H Ο · Ρ O -P Ο d OOOO -P, d 3 -P -P 000 3 \ 3 3 m ucodfOdiocacôuc ^ HuocppHii ^ ii-t w P 'Φ P' Φ P 'Φ H' φ 'φ' Φ 'Φ PEHP ίΙ'ΦίίΙΗΗΗΗ ίΟ'ϋιΟΌ'ΦΌ'ΦΌΌ'ΰΌΦ> ι Φ <0 <0 Ό ·>) - · Η Φ (0ΦΦ

UitÎUrtlEHidS ^ rfîrtîidUJUUUrtifefoUmUO

03 ΦΦΦΦΦΦΦΦΦΦΦΦΦΦΦΦΦΦΦΜ-Ι-ΗΦ

IllililllüilllllllSsc ^ ΟΟΟΟΟΟΟΟΟΟΟΟΟΟΟΟΟΟΟΟΟ · Η (0 iumtdtctßüiffiCötdüimwtatütßWtdwmcQwcoPi> f- 64

CO W I ^ in U OJ

© in 3 Φ 3 C

+1 CH © H © © +> 0 0 +> +> Ό • HUU ω 0! —I · U ld Ö © n) Ο ^ Ή g »wg HW fd trt · η O 0) H fl) <ÜT ( • H +> Ό Ό 3 S © 2 Vj φ Ό -φ R Ό) 3

Ο Ό dd £ + d C

ιη φ -h 'Ô -hx (N 3 o + j * ri +> o fi · η g 3 R co co H te id P <d "a) d H +) 3 3 crt (d -h (d σ <£ σ | Ό gh 9 Λ 8 "{" f * 3 • rl g dd I +> en o "h en d © '^ ^ Φ Φ> +> ™ id

o. "D -H C H H

en> en ^ a, 3 m -ri o S © W m en + J m h O $ Φ u +> S w υ Ö Ό © Φ / η

Φ 0) Ό Φ S

© H C S Ό U 0 ai h h H d ‘Ί <Η'Φϋ R fi

- 'Φ in S O in R

Ό in O in ”in ^

Ό 'Φ 3 -H -H Q

^ ü +> ‘5 <d O ^

H Φ (d H gH M H

in C H g U O

"Ψ in d 3 u B * B

<d Φ O M H id 5

'Φ in R® h -H

in 3 · Φ M 9 “ΦΗΦΗ„ .Ç ΦΗ H -H H O 'H Td

3 Ό 3 '© ^ »O

H HH „H Φ O

H (d H 3 R U O

Φ Φ U o <ÿP HH

fi en u h τ 'o u Φ ed "m +> d) U H φ u -H id h

3 3 ’H fl) H

“D 3 <+> -5> o 3U

O © © Λ §fi <! , C 'Sl, H j,

HtnUen fi M +> t!

ü H d · Η o in S

- 3 H o m 5 S

Di 3 H '© of end, in o +> -Pc +>? 'Φ'. ft o -HR h 2 +> +> (ü> H> ® • H in ^ -hh, d Φ dh +> H +> h 3 +> O fi 5 fi 3 h in <d O ed 9 · Φ m © +) © - a - £ + in h <N Ό 3 h Ό »Ö Φ fi HH fl) fl) H id

3 H H '© U' © U 3 H

in © Ό 3 +> - nen +) · ηη φ hh in H td Φ h ed h o

Ehvo denn den © »© h © 3 en 3 3enü Ό © E en t> φη φ

Hd © o h © u in β d © h © d © © d o © o Ο Ό O, in D d U D d H * zj •

O

© fi U! 3 65 The effect of AR and EGF on the formation of a colony of NRK-SA6 cells in soft agar in the absence and in the presence of TGFyJ was the subject; of an observation, the results of which are collated in Table V. EGF induces growth of SA6 cells independent of anchoring as a function of dose in the presence of TGF ß, but AR has been shown to be a very inductive weak formation of SA6 colonies on soft agar (Table V).

TABLE y

Effect of AR and EGF on the formation of NRK-SA6 cell colonies on soft agar.

Addition Number of (per ml) colonies per 6 fields None 0 TGFfè (1 ng) 0 AR (2 ng) 0 AR (4 ng) 0 AR (2 ng) + TGFyâ (1 ng) 7 AR (4 ng) + TGFyâ (1 ng) 8 EGF (2 ng) 0 EGF (4 ng) 6 EGF (2 ng) + TGFß (1 ng) 30 EGF (4 ng) + TGFß (1 ng) 114

The assays are performed as described in the section. A colony is a cluster of at least 6 cells.

66 6.7.8. Effect of AR on the binding of EGF to its specific receptors.

RA has been shown to inhibit 125j-egf binding on A431 cells, as well as on A431 plasma membranes (Fig. 10). A 50% inhibition of 125j_egp binding to fixed cells and membranes was observed at approximately 1.1 nM and 1.8 nM EGF, respectively, while a 50% reduction in l binding EGF on cells and membranes was observed at approximately 1.8 nM and 5.7 nM AR, respectively. The unlabeled EGF completely inhibits the interaction between the 125j_ggp receptors and higher concentrations in both systems. However, the maximum competition with RA was 75% and 50% for binding to cells and membranes, respectively. The competition curves for the AR are also not parallel to those recorded with the EGF. These results suggest that AR manifests a lower affinity than EGF for EGF receptors. Likewise, the AR may have its own specific receptor closely related to the EGF receptor.

Different clones of the A431 cell line having various sensitivities to inhibition by RA were selected and an absence of correlation between the sensitivity to RA and the number of EGF receptors per cell or KD was observed. (Table VI).

67

TABLE VI

Lack of correlation between the sensitivities to amphiregulin of various A431 clones and the number of KDs or EGF receptors

Parameters for fixing the EGF Sensitivity relative to the AT (GIA)

KD Clone Sites

A431 fixation by (NM) cells A-3 4.6 x 105 1.80 34 F-8 9.0 x 105 3.13 20 A-2 5.1 x 105 1.38 11 A-8 5.3 x 105 3.18 1

Various clones of A431 were selected by culture in soft agar. The fixation of the EGF was carried out as described in section 6.6.1.5. above. Kg and the numbers of binding sites were calculated from Scatchard diagrams (Scatchard et al., 1949, Ann., N.Y., Acad. Sci. 51: 660 to 672).

6.7.9 Chemical structure of 11 amphiregulin.

The amino acid sequence of amphiregulin was deduced from the microsequencing analysis of the S-pyridylethylated protein (SPE-AR) and of the fragments produced by the endoproteinase Lys-C, the proteinase V8 from Staphylococcus aureus and 1 ' endopeptidase Arg. The sequences of the truncated protein and of the protein containing the six additional amino acids are as follows: 68 AR truncated 1 10 20

VVKPPQNKTE SENTSDKPKR KKKGGK

30 40 50

N G K N RRNRKKKNPC NAEFQNFCIH GE

60 70 78

CKYIEHLE AVTCKCQQEY FGERCGEK

Larger AR 1 10 20

SVRVEQVVKP PQNKTESENT SDKPKR

30 40 50

KKKG GKNGKNRRNR KKKNPCNAEF Q

60 70

NFCIHGECK YIEHLEAVTC KCQQEYF

80 84

GER C G E K

The above sequences are formed using the usual letter abbreviations for each amino acid. Remains of amino acids whose identity is questionable are indicated in parentheses. The letter "X" indicates an amino acid of unknown identity. The amount of the larger AR was found to be approximately 16% of that of the truncated AR.

The sequence of the AR was compared with that of all the proteins from the National Biomedical Research Foundation (PIR 13) database, from the genetic sequence database (Boit Beranek and Newman, Inc., Los Alamos National laboratory (release no 50) and the DNA sequence library of the European Molecular Biology Laboratory (release # 12). This research revealed that AR is a new protein and is a member of the EGF superfamily. growth factors. This 69 family includes EGF (mouse, man, rat, beef, etc.) "TGFO (, viral growth factors (vaccinia, myxomatosis and Shope's fibroma), the activator of human plasminogen , bovine factor IX, human factor X, LDL receptor, bovine protein C, human proteoglycan core protein, the product of the Drosophila Notch gene, the product of C. eleqans lin 12 gene and cell line gene product of sea urchin S. purpuratus. t of the structure of the AR with the structures of the proteins of the EGF superfamily reveals that the AR, like the other members of the EGF superfamily, contains the six essential cysteine residues for identification, respects the conservation of the spacing cysteine residues and contains some of the characteristic and highly conserved amino acids. It is apparent that RA lies between members of the family of growth factors that resemble EGF and TGF and those that resemble MGF, SFGF, especially in terms of the use of asparagine. For example, at position 2, (first cysteine = 1), all TGF and EGF have a proline, VGF has glycine, and MGF, SFGF and AR have asparagine. In the same loop, at position 7, EGF and VGF have glycine, TGF have glutamine and MGF, SFGF and AR have asparagine. However, in the third loop, the AR does not have the beta turn stored at position 34 (an asparagine in MGF and SFGF or a glycine in all other members), but includes a glutamic acid (low beta turn potential ). The AR has a different structure for the highly conserved third loop, which can affect binding to the receiver. AR does not have the asparagine residues that may be used in MGF and SFGF as sites of glycosylation in the proper structure of growth factor.

The N-terminal sequence of AR has a certain homology with the N-terminal sequences of TGF, VGF and MGF in that it is rich in prolines, serines and threonines and includes, like TGF and VGF, potential N-linked glycosylation sites (in fact there is such a site) with the possibility of o-linked glycosylation in this region which is rich in serines, threonines and prolines. This region is highly conserved between the N-terminus of the TGF precursors and the N-terminus of VGF and appears to be a highly glycosylated region.

AR is an extremely hydrophilic protein, especially in the region containing amino acids 8 to 45 (Fig. 2). The RA hydropathy profile shows little similarity to that of other members of the EGF family. The hydropathy profile of the C-terminal part of the RA, although structurally related to those of other members of the EGF family, is different from the profiles of other members of this family (Fig. 11 B and C).

6.7.10. AR specifically fixes DNA.

125I-AR was tested for its ability to bind to cellulose, denatured DNA-cellulose and native DNA-cellulose, according to the procedure described in section 6.6. above. The results presented in Fig. 14A indicate that the AR binds specifically to DNA-cellulose whose DNA is denatured or native, but does not bind to cellulose. Binding to native DNA is three or four times greater than binding to denatured DNA.

The ability of the AR to bind specifically to DNA makes the difference between the AR and the other 71 members of the EGF superfamily. For example, EGF is unable to bind to DNA-cellulose (Fig. 14B). These results strongly suggest that the hydrophilic N-terminal domain of 45 amino acids of AR, which does not exist in EGF or in the other members of the EGF family, may be a nuclear localization sequence involved in targeting. AR on the nucleus, where the factor interacts with DNA.

7. Example; antibodies against 11 amphirequline.

The production of antibodies against AR and synthetic AR, in addition to the AR precursor peptides, is described in the subsections below.

7.1. Materials and processes.

7.1.1. Synthesis of solid phase peptides.

Amphiregulin was produced and purified as described in section 6 above. Five peptides corresponding to residues 31 to 50 (n ° 279), 71 to 90

(n ° 280), 108 to 130 (n ° 264), 166 to 184 (n ° 259) and 221 to 240 (n ° 281) of the primary structure of the AR

(Fig. 15) were synthesized according to solid phase techniques in an automatic peptide synthesizer Applied Biosystems Inc., substantially as described (Merrifield, 1973, J. Amer. Chem. Soc. 85: 2149). The synthesized peptides were detached from the support resins using an HF cleavage technique (Tarn et al., 1988, J. Amer. Chem. Soc. 105: 6442). The synthetic peptides were purified by HPLC with phase inversion.

7.1.2. Antibody production.

Polyclonal antisera against mature AR and the syntethic peptides of AR, chemically conjugated to arapedan hemocyanin, were prepared using young adult white rabbits from New Zealand. Synthetic peptides 72 were conjugated to arapedan heraocyanin as described (Ishikawa et al., 1983, Immunoassay 4; 235) and RA was not conjugated at maturity.

AR or conjugated peptides were emulsified with Ribi's adjuvant system (Ribi Immunochem. Research, Inc., Hamilton, MT). A total dose of 1 ml was administered to each rabbit, namely 0.8 ml intramuscularly at two sites on each hind leg and 0.2 ml subcutaneously at one site. To produce antisera against mature RA, 30 micrograms of mature AR were used for primary inoculation, followed by 15 micrograms for subsequent booster injections. To produce antisera against synthetic AR peptides, 100 micrograms of conjugated peptide was used for primary inoculation and 50 micrograms for subsequent booster injections. Booster injections were given every 3 to 4 weeks and rabbits were bled 10 to 14 days after the booster injections.

7.1.3. Immunoprecipitation.

The RA was radioactively labeled with 125 d using the chloramine T method (Barridge, 1978, Methods Enzymol. 50: 54-65). 10 microliters of polyclonal serum (or dilution with v PBS) were mixed with 10 microliters of 125I-AR (50 mg per ml, specific activity of 425 microcuries per microgram) and 30 microliters of PBS and the assay was carried out essentially as described (LeBier et al., 1982, J. Immunol. 129: 2287-2292).

7.1.4. Radio immunoassay.

The polyclonal sera were diluted to 1.50 in PBS. The 125i_ar (1 microgram per ml, specific activity of 425 microcuries per microgram) was diluted to 1: 200 with TNEN / 0.1% BSA (20 mM Tris, pH 7.4, 73 EDTA 5 niM, NaCl 150 mM, 0.05% NP-40, 01% BSA). 10 microliters of unlabeled AR (1 mg per ml), 10 microliters of diluted polyclonal serum and 30 microliters of diluted 125 I-AR were mixed and incubated for 45 minutes at room temperature. The protein A- solution was prepared

Sepharose as follows: 200 mg of protein A-Pharmacia dry Sepharose was rehydrated with 1.4 ml of PBS and 80 microliters of the rehydrated product were washed and diluted to 400 microliters with a solution containing Tris 100 mM pH 8.1, 150 mM NaCl, 0.5% Triton X-100, 2 mM PMSF, 1% Aprotinin and 0.5 micrograms per ml of Levpeptine. Then 20 microliters of protein A solution was added to the mixture which was incubated for a further 30 minutes. The Sepharose beads were then washed twice with 500 microliters of TNEN / 0.1% BSA, resuspended in 50 microliters of SB (80 mM Tris pH 6.8, 3% SDS, 15 % glycol, 0.01% bromophenol blue, 5% β-mercaptoethanol) and they were heated for 5 minutes at 100 ° C. The samples were centrifuged and the radioactivity of the supernatants was assayed in a gamma counter. This procedure makes it possible to detect barely 100 pg of AR.

7.1.5. Solid phase immunoenzymatic test (ELISA).

The micro-ELISA solid phase procedure described in the E.U.A. patent application was modified. n ° 740 124 of May 20, 1985. TerasaXi microplates were prepared by introducing, into each well, 5 microliters of synthetic AR peptides which were diluted in PBS to various concentrations before leaving the solution. evaporate to dryness at room temperature overnight. 5% Blotto in PBS was added to the plates which were allowed to stand at room temperature 74 for one hour. Next, 10 microliters of dilutions of the polyclonal antibody solution (diluted in PBS / 1% Blotto / 0.05% Triton X-100) were introduced into each well before incubation at room temperature for one hour, then four washes with PBS. 5 microliters of goat anti-rabbit IgG conjugated with peroxidase (Cappel) were diluted 1: 1000 in PBS / 1M Blotto / 0.05% Triton X-100 before each procedure. incubation at room temperature for 1 hour, then six washes with PBS. 10 microliters of chromogenic Micro-EIA solution (Genetic Systems Corp.) was added to the 1:20 dilution, then the absorbance was read at DO492 after 20 minutes and 40 minutes, according to Genetic Systems Micro-EIA protocol .

7.1.6. Western blot.

Samples were subjected to electrophoresis on a 15% polyacrylamide gel or 16% tricine-polyacrylamide gel with a BioRad minigel. The gel was placed adjacent to a sheet of nitrocellulose and sandwiched in the cassette so that the nitrocellulose was on the cathode side. The transfer of the protein was carried out with a current at 400 mA at constant power for 30 minutes. The nitrocellulose was then removed and soaked overnight at 4 ° C in 2.5% Blotto / PBS / 0.2% NP-40.

Next, the nitrocellulose filter was exposed to antiserum diluted in 2.5% Blotto / PBS / 0.2% NP-40 as reaction buffer for 90 minutes at room temperature on a platform tilting, before four washes (5 minutes for each wash) in 2.5% Blotto / PBS / 0.2% NP-40.

75

Protein A conjugated with alkaline phosphatase (Cappel) diluted 1: 500 in 2.5% Blotto / PBS / 0.25% NP-40 was then applied to the nitrocellulose sheet. incubation at room temperature for 60 minutes in a tilting bowl. The filter was then washed four times (3 minutes for each wash) with PBS / 0.2% NP-40, then for 5 minutes in "APS" buffer (TRIS 100 nM pH 9.5, NaCl 100 mM, MgCl2 5 niM). The filter was developed in chromogenic reagent (40 ml of APS buffer, 12 mg of 5-bromo-4-chloro-3-indolyl phosphate (Sigma), 6.8 mg of p-nitro-tetrazolium blue (Sigma ), prepared just before use and filtered through a 0.2 micron filter), then the reaction was stopped with water and the filter was allowed to air dry.

7.2. Characterization of antibodies against AR.

Polyclonal sera directed against mature RA and synthetic peptides of RA have been characterized by radioimmunoprecipitation, radioimmunoassay, ELISA and Western blotting. The results are given in Table VII.

76

TABLE VII

Characterization of the antibodies against 11 amphiregulin and the synthetic peptides of 11AR

Polyclonal dosage serum -

against RIP RIA ELISA WB

AR at maturity l: 256a 1: 250 ND 1: 250 (0.100 bp) b (1 ng)

Peptide 259 1: 256 ND ND ND

Peptide 264 1: 256 1: 250 1: 100 1: 100 (0.100 bp) (5 ng) (0.1 ng) 1: 500 (1 ng)

Peptide 279 ND ND 1: 100 ND

(5 ng)

Peptide 280 ND ND 1: 100 ND

(5 ng)

Peptide 281 ND ND 1: 100 ND

(5 ng) a = a dilution of the antiserum is necessary; b = the values in brackets indicate the sensitivity of the assay; RIP = radioimmunoprecipitation; RIA = radioimmunoassay; ELISA = solid phase immunoenzymatic test; WB = Western blot; ND = Not Determined.

8. Example: cloning of the cDNA of the 11 amphiregulin precursor.

The following example describes the cloning and analysis of the cDNAs coding for the amphiregulin precursor originating from cells of human mammary carcinoma MCF-7 treated with TPA.

77 8.1. Materials and processes.

8.1.1. RNA preparation.

RNA was isolated from subconfluent cells cultured in T150 tissue culture flasks or from fresh frozen tissue samples by applying the guanidinium method described by Chirgwin (1979, Biochemistry, 18, 5294). The cells of 4 T150 vials were washed in PBS, treated with trypsin, washed again in PBS and the pellet was resuspended in 8 ml of guanidinium isothiocyanate solution 4M. The DNA was collected by passing the solution through a No. 18 needle. The sample was deposited on 2.4 ml of 5.7M CsCl in a Beckman SW41Ti tube and centrifuged at 32,000 rpm. minute for 20 hours at 20 ° C. The pellets were resuspended in 300 microliters of 2.5M CsCl and precipitated with 2 volumes of ethanol at room temperature. The pellets were resuspended in 300 microliters of water, then 35 microliters of 3M sodium acetate (pH 5.2) and 800 microliters of ethanol were added to precipitate the RNA at -70 ° C. The precipitation was repeated and then the RNA was washed briefly before being resuspended in 100 microliters of water, quantitatively assayed at DO250 and stored at -70 ° C.

8.1.2. Statistical labeling with 32p-.TTp.

The specific fragments of the region coding for the AR were excised out of a low temperature agarose gel (BioRad) and they were labeled with 32p_ TTP (NEN, 3000 Curies per millimole) according to the labeling process. statistic developed by Feinberg (1983, Biochem., 137, 266). By chromatography on a 1.5 ml Sephadex column, the unincorporated deoxyribonucleosides were separated. Specific activities were typically 0.5 to 2.5 x 10 ^ counts per minute per microgram.

8.1.3. RNA gels and Northern blot analysis.

10 or 20 micrograms of total RNA was subjected to electrophoresis on 1.2% agarose / formaldehyde gels for Northern blot analysis. Agarose / formaldehyde gels were prepared in 40 mM N-morpholinopropanesulfonic acid (MOPS), pH 7.0 / 1 mM EDTA / 10 mM sodium acetate / 2.2 M deionized formaldehyde (pH 5.6 ) in an IBI VCV gel machine, with frosted plates. The RNA was denatured in the same buffer with 50% formamide at 65 eC and passed through 1 X MOPS between 20 and 30 mA for 3 to 5 hours. The gel was rinsed in 10x SSC for 30 minutes, then transferred to Hybond-N membranes (Amersham), cross-linked in UV (1200 microjoules), then proceeded to prehybridization and to hybridization in 5 × SSPE (1 × SSPE is 0.18 M NaCl / 10 mM sodium phosphate, pH 7.7, 0.1 mM EDTA), 5 × Denhardt medium, SDS at 0, 5% and 20 micrograms per ml of DNA from denatured salmon sperm, as carrier. Hybridizations were performed at 42 ° C for 16 hours with 2 x 10 ^ counts per minute per ml of P-ARBP1. Qn iav © fingerprints several times in 2x SSC with 0.1% SDS at room temperature, then in 1x SSC with 0.1% SDS at 65 ° C, before exposure to Kodak X-OMAT film with two Dupont Lightning Plus intensifier screens at -70oC. The bands were rated as (+) if they are clearly visible after exposure overnight, (+/-) if they are visible after 3 days of exposure and (++) if they are detectable after 6 hours.

79 8.1.4. Cloning of cDNA.

MCF-7 cells were harvested to obtain RNA after treatment with 100 micrograms per ml of 12-0-tetradecanoylphorbol 13-acetate (TPA) for 24, 40 and 72 hours. The poly (A) + RNA was isolated from the pooled aliquots of these samples and used as a template for the synthesis of double-stranded cDNA substantially as described (Gubler and Hoffman, 1983, Gene 25: 263). L-tailed cDNA was ligated into the EcoRI site of ^ gt10 using the BRI oligonucleotide adapters (Rose et al., 1986, Proc. Natl. Acad. Sci. U.S.A. 83: 1261-1265).

Specifically, 5 micrograms of poly (A) + MCF-7 RNA treated with TPA were heat denatured and a primer of 5 micrograms of oligo (dT) was fixed using 100 units of reverse transcriptase in 45 microliters of reaction volume. Second strand synthesis was performed with 4 units of RNase H and 115 units of pol DNA. I of E. coli. Next, 10 micrograms of T4 DNA polymerase was used to remove the 3 'overhangs and obtain non-sticky ends. The 6.8 micrograms of double-stranded cDNA were fractionated according to dimensions on a column of Sephadex G50 to select the cDNAs of more than 500 base pairs, then a dG tail was fixed on 150 ng of cDNA with terminal deoxynucleotidyl transferase. The dG-tailed cDNA was ligated into the EcoRI site of Λ gt10 using BRI adapters (AATTCCCCCCCCCCCCCC). The bound DNA was packaged in vitro (Grosveld et al., 1981, Gene 13: 227-237) and deposited on C600 rK "mK hfl of E. coli with an efficiency of 106 recombinants per microgram of cDNA. From 2.5 × 10 5 recombinants, two i 80 fingerprints were taken on nitrocellulose, then the filters were examined with degenerate and presumed to be better oligonucleotide probes, marked with 32 p (table VIII below) deduced. of the primary AR sequence as determined by automated Edman degradation of the reduced S-pyridylethylated AR and treated with N-glycanase (section 6, above).

31

"(-NOT

<D

ü ^ ^ G <D φ

u 0> H H

a ο a a S ® 3 s g ® _ J. I m r "on

Oi fc S.'J0 ^ 1/1 Ό un - '(J) cm \ û cm en' η ö ^ "- 3 * ® M tyi 'Φ Ό a ο u U U U U W Eh

<<- O

a

O U ~ U

W <"b Wb Wü <d Eh '-'U <

O O Ü U

Z Eh U U OR W fn

Eh <U U

U ü <Eh

H OEh Eh U U U Eh O

w U E- "ΐ Eh> O U U U ~

W <> <KU «E KO

3 - - <U O Eh ^ Eh x * Φ a one

mü - - U <U O O

3 G * Eh X Eh WEH <ü W Eh ÛH X Eh 3 <s> - <'-ο ο υ v u Eh «a * D un

2 O1 Eh U Eh U «C ~ U Ü U U

'© U Eh -WEh <ÜWE- <WEhüUC> EhPhO

TO U U U '-' EhUUOU

EhU <<O O * £ U O

OEh H <Z Eh OU J <W <CU Ο X Eh O Eh Eh <0 <ÜEh

Eh U <0 O U <O O U

0 »Eh Xh OR WEh SB> * Eh CU O û Eh

O <<U O * C O U

<OEhO EhOOUUC

or WEh had O OR WEh w Eh W Eh wo <Eh OUOUEh <

Eh U <U O <<U U O

WEh OU Z Eh X Eh H <O £ h> <Eh O

Eh <Eh (J) Eh O O Eh

<U Eh ο U O> H ü U O O

OR WEh X Eh h <- <OEh> <2 Eh <U Eh Eh O U Eh ro m en m on

Φ -, H 00 rH

Ό .r. tj · a Q) u <η · <τ Eh G £ Q Q * φ 3 W X 2 O J * OS “g φ OS“ “C0'® <<3 ^ 3 3 <* φ ® - en '£

* φ t S

The restriction analysis revealed a scarcity of sites in the inserted cDNA segment of pAR1, with unique sites for Bsml, EcoRV, Hgal, Nael, PvuII, Smal and SstI and two sites for Sspl. No digestion was done in the segment inserted by BamHI, Clal, EcoRI, HindlII, Kpnl, PstI, Pvul, SphI, StuI and Xbal. A 170 base pair Bsml to PvuII fragment was isolated and used to probe a second cDNA library of 100,000 recombinants which had been obtained substantially as described above. Thirteen positive clones were identified comprising inserted segments ranging from 300 base pairs to 1.3 kb, six being greater than 1 kb, and five containing a single SsTI site known to be located within 100 base pairs from the 5 'end of pARl. Four of the last five inserted segments (pAR3, pAR5, pAR9, pAR13) were subcloned for further restriction with sequence analysis. All of these clones had identical restriction maps based on Bsml, EcoRV, PvuII, SstI and Smal, except pAR9 which had a 3 'truncation of 100 base pairs and which later appeared to come from an A5 track, probably sufficient for priming with oligo (dT).

8.2. Analysis of the nucleotide sequence of AR cDNA clones.

Exact oligonucleotide primers were used to sequence the two strands of the pAR1 clone of 1230 base pairs using the dideoxy chain termination method (Sänger et al., 1977, Proc. Natl. Acad. Sci. USA, 74: 5463 at 5467). The complete nucleotide sequence and the amino acid sequence deduced therefrom for the clone pAR1 are illustrated in FIG. 16. An open reading frame of 965 base pairs begins at nucleotide # 1. The first AUG is at position 210, but 83 does not meet the optimal Kozak consensus for translation initiation sites (Kozak, 1984, Nucleic Acids Research 12: 857 to 872; Kozak, 1986, Cell 44: 282 to 292) due to the absence of a purine at the -3 position. However, these AUG triplets can serve as the initiating codon J, as Kozak observed in approximately 3% of the messages examined, and the presence of adenosine at position -1 in these three "less favored" AUGs and in the mRNA of AR is of possible importance. Although a second methionine is located at nucleotide 378, which does not conform to the consensus sequence of the initiator, the first AUG is believed to be the true translation initiation site because it is followed by an expected series of 19 amino acids with predominantly hydrophobic residues which is interrupted by 3 prolines, which is typical of a signal peptide sequence (Heijne, 1983, J. Biochem. 133: 17-21).

The longest reading frame starting with a methionine encodes a polypeptide of 252 amino acids which includes the signal peptide of 19 residues. The coding sequence is preceded by 209 nucleotides of the 5 'untranslated sequence and followed by a TAA end of translation signal and of 262 nucleotides of 3' untranslated sequence. A potential poly (A) addition signal sequence, AATAA, is located 64 nucleotides upstream of the series of 15 adenylate residues, probably to constitute the poly (A) tail. The 3 'untranslated region of AR contains four copies of the ATTTTA sequence, which is also present in certain lymphokines, cytokines and various protooncogenes. In GM-CSF, this sequence is involved in the destabilization and degradation of RNA (Shaw, 1986, Celle 46: 659 to 667). The conservation of this motif in molecules of similar functionality 84 suggests that RA may also have a relationship with this class of lymphokines.

The cDNA sequence confirms the peptide sequence of the mature AR (section 6 above), except at position 113 of the amino acids (Fig. 15) that the sequencing of the protein revealed to be aspartic acid (D ) and which was presumed to be asparagine (AAT = N) from the cDNA clones. Of the five cDNAs examined, all had their 5 'ends in the first 25 base pairs of the pAR1 sequence.

Comparison of the cDNA sequence with the best presumed probes revealed 74% (ARK41) and 77% (ARK31) of general homology. No probe had a match on more than eight successive base pairs, but overall, 50 of the 67 nucleotides aligned for ARK41 gave a detectable signal under conditions that were not very restrictive. The use of codons by the AR mRNA sequence differs greatly from the frequencies of use indicated by Lathe (Lathe, 1985, J. Mol. Biol. 183: 1 to 12), which explains why degenerate oligonucleotides have given more intense signals in this circumstance. From the cDNA and protein sequences, two proteins with a molecular weight of 9772 and 9173 are provided for the two forms of mature AR on the basis of the sequences of cDNA (FIG. 16) and of the protein (section 6.7.9. above). The RA hydropathy profile is shown in Fig. HD.

The AR precursor includes 3 potential N-glycosylation sites, including one in the N-terminal domain (position 30 in Fig. 16) and 2 in the hydrophilic region of the mature AR (positions 113 and 119) . It is known that glycolysis contributes for 10 to 12 kd to the molecular weight of the AR at maturity and that it is capable of being carried out by the addition of carbohydrate at one or both of these sites in the field. hydrophilic.

The N-terminal domain rich in serine of the precursor of the AR (Fig. 15) contains three potential sites of sulfation of tyrosine in Y ^ l, γ82 and γ87, on the basis of the presence of acid residues, or amino acids inducing torsion, and the absence of residues which could contribute to steric hindrance (Huttner, 1987, TIBS 12: 301 to 303). Most of the proteins sulfated on tyrosine are secretory proteins. Changes by sulfation of tyrosines are presumed to be involved in the activation, transport or proteolytic processing of precursor proteins. The serine-rich domain of the AR may also contain 0-linked carbohydrate chains given the abundance of sine / threonine residues (23 of 81) in this region of the molecule, which are sites for such links. In this connection, O-glycosylated forms of another member of the EGF family, namely TGF û (, have been identified (Ignotz et al., 1986, PNAS, 83, 6307 to 6311).

None of the serine residues in the RA precursor adapts to the consensus sequence for the kinase phosphorylation sites dependent on cAMP (Grima et al., 1985, Proc. Natl. Acad. Sci. Ü .SA 82: 617 to 621), and none of the tyrosine residues shows any similarity in the flanking sequence with known phosphotyrosine residues.

The hydrophilic domain of the AR protein at maturity is composed of numerous positively charged amino acids (16 out of 37 residues are lysine or arginine), in particular two successive series of 4 or 5 basic charged residues. The nuclear targeting signal 86 of the large T antigen of SV40 is similar to this region of the AR and contains a characteristic KKKRK sequence preceded by small amino acids (glycine, alanine, proline) which can allow the formation of a 0 structure. (-helical (Burglin et al., 1987, EMBO 6: 2617-2625; Dingwall et al., 1987, EMBO 6: 69-74). The mutation analysis defined four consecutive basic remains as the predominant feature of the nuclear localization sequence of SV40 (Manford, 1984, Cell, 37, 801 to 813) Other nuclear proteins comprising similar series of basic amino acids which are believed to be involved in nuclear targeting are notably nucleoplasmin , the large T protein of the polyoma virus, histones and c-myc. The fusion of various nuclear signal sequences on the genes of chicken pyruvate kinase, albumin, immunoglobulins, ferritin and / 2 -galactosidase, leads to localization ation of these otherwise cytoplasmic proteins in the nucleus (Moreland, 1987, Mol. Cell. Biol. 7: 4048 to 4057). Whether or not the hydrophilic sequences of AR are involved in targeting this growth modulator in the nucleus has not been verified, but the ability of AR to bind specifically to DNA implies that AR has a functional role within the nucleus. In this respect, the AR can regulate DNA synthesis, the cell cycle and various other events in the nucleus.

The precursor of TGF o (contains multiple cysteines in the C-terminal cytoplasmic domain, some of which undergo covalent attachment of palmitate (Bringman, 1987, Cell, 48, 429 to 440). On the contrary, the precursor of AR does not contain cysteines in its cytoplasmic domain and it is therefore not to be expected that it contains palmitate residues. Although the biological function of palmitate fixation is not known / this constitutes yet another difference between l 'AR and the other members of the EGF superfamily.

8.3. Cellular sources for the synthesis of 11 amphiregulin.

Northern blot analysis (Thomas / 1980, PNAS, 77, 5201) using 32p-labeled ot AR cDNA fragments revealed hybridization on a single 1.4 kb RNA species at from various normal human tissues and from various tumor cell lines. A single band was observed on the Northern blots, using either AREB1 (fragment BstEII to PvuII of 480 base pairs of pAR1 containing the complete coding region of the AR at maturity and part of the precursor and transmembrane N-terminal domains) , or AR170 (Bsml to PuvII fragment of 170 base pairs of pAR1 containing only the c-terminal half of the AR at maturity) as labeled probes. The results collated in Table IX below show that a low level of expression (+) of the RNA of the AR has been observed in the tissue of the normal lung and pancreas of the adult. Weaker but detectable expression (+/-) has been observed in normal kidney, liver and brain tissue.

88

TABLE IX

Normal human tissue AR RNA

Lung +

Liver +/-

Pancreas +

Kidney +/-

Missed

Brain +/-

Placenta

Intestine, fetal

Since the AR was originally isolated from MCF-7 cells treated with TPA, it was interesting to determine the time for the induction of RNA from the AR by TPA in these cells. MCF-7 cells were treated with TPA for 0, 1, 3, 6, 18, 24 and 48 hours and the total RNA was isolated, then passed in quantity of 10 micrograms on each column of a gel 1.2% formaldehyde agarose and then transferred to nylon membranes and examined with the ARBP1 probe complementary to the complete coding region of mature AR. An impressive 1.4 kb RNA increase in AR was observed as early as 1 hour after TPA treatment, with the highest rates being achieved between 18 and 24 hours. Subsequent Northern blots, probed using ^ p-ARBPl, revealed that TPA stimulates RNA synthesis in RA in other breast cancer lines, although the highest rates have never were as high as in TPA induction MCF-7 cells.

Many tumor cell lines have been assayed for RNA levels in RA

89 both before and after 24 hours of induction by TPA. The results are collated in Table X below. With few exceptions, the only detectable sources of RNA from AR were human breast cancer lines treated with TPA. One such exception is the Caki-1 line, or clear kidney human renal cell carcinoma line, which constitutively expresses high levels of AR RNA which are comparable to the amounts observed in induction-induced MCF-7 cells. by the TPA. All of the TPA treated breast cancer cell lines that have been observed have shown specific AR hybridization.

Apparently, RA fulfills a certain functional role in the tissue of the lungs, pancreas, kidneys, liver and brain of adults, perhaps by being involved in a wide variety of processes, including wound healing , tissue regeneration and maintenance of neurons.

90 i 143 ο ß & +,,. .

"5 ° ί + + ΐ + + + -Î + ι ιΊ-ι ι-ί. + + + + + - .5 Φ

Α Μ ÔJ

Α

X I

§ »ϋ Ö + ^ § fl · 2 +11+ 1111+ ιιιιιι 03 β + J + Η β • Η φ

U

β fl β ο “> 1

νιι Φ U

* • 1 χ φ u & -, 3H -ri g fl (0 · Η Φ (0 Φ 5 ß (Ö Μ Η Φ - fl (Ö Η · Η Ο 'Φ Α φ £ ϋ Η (Ο β φ Ο S Φ β Φ g Φ C0 Φ Φ CO Α β 2 2 gmaggg φ -mg · η m ft ^ 5 ο Φ'Φοφο η ρ ο ΦΟ'Φ fi ßflgfigß β Ο β Η g Μ X · Η · Η · Η rl · Η Φ kl β ft Π Ο ΦΦΟΦΟΦΦΗΦΟ ft Φ Ο Ο ^ ggtllllhgg © tllhg Ό> ιβ u Φ ΟΟΦΦΦΟΟΟΦΦΗβ-Hkifl

OrHßßO + JOßß +> Ο · Η fl ft £>

Ο Φ · Η · Η Ο 10 Ο · Η -rH / (0 ID ο S 'Φ g φ fiEüUßidßüü "J β Φ Φ + ι' Φ k Ιη ^ ι'Φ 4J '" μ Φ + Ι'Φ Φ Ε ¢ - W

flO (0 φ fl »Φ fl Φ (0 g * Φ fl fl> igfl (0 φβυθΦ2Φθοο2 <0 ί ο η Ο ß kl Ο ß ß XJ **** ·» «. * · Η · Η ^ 3 C Ή Ο Ο ßßßßß ß β k υ φ φ φ ug k.

• Η -Η · Η · Η · Η 'k · Η Ο k · Η g CO kl β, Ο Φ Φ Φ Φ Φ ΦΦ, βίΟ Ο ß' Φ (0 Ο · Η 03 03 03 03 Ü3 03 03 U Ο ft Εη 2 U Οι Οι + Γ "/ - * Φ ζ- 'CM (Ν Λ p *

kl - + Z- 'CM Z-' z-, fO

• H cm 03 03 O z ^ Φ ζ ~ ή O E-ι H nz-xvo M * vom

H ni cm cMîUîr! 'Λ - + 03 - + co m J

3Ï CM 03 - 03 m Eh H - + H U

h.5 Hm B mtrjHu rk □ 03 ffl Eh r ~ H tu 03 Eh - J J "" φ2 EH '-' ffimco eh ffl S i-3 2 05 œ -ZH 00 ŒJ'-z H U pi U Φ

d '—ο Φ I CM v_z I

Φ + O - + 03 03 | CD H ft ^ CM

'Φ r'HP-ss 03 co i ft r-i i-ι co

ß I I - + I I 03- + 1 -H KOSO I

Cn ft 01 I <<I I O. If 1- + 0) - + (- 3¾

H U 03 Eh Q Q fcc! Eh H Φ "I H I ffl 1D

i-3 2EC322 BfflbU 03 U E ri! 03 K

91 9. Example; Genomic cloning and analysis of the amphiregulin gene.

The example below describes the genomic cloning of the human amphiregulin gene, its structure and its intron / exon organization. The functional and evolutionary implications relating to the EGF superfamily of growth factors are also considered.

9.1. Materials and processes.

9.1.1. Southern blot hybridizations.

Total genomic DNA (Maniatis, 1982, in Molecular Cloning: A Laboratory Manual) was isolated from subconfluent cells in T150 tissue culture flasks. 20 micrograms of DNA were digested with the restriction enzymes indicated, electrophoresed on 0.8% agarose gel and transferred to Hybond-N (Amersham) with background buffer SSC 20x (Southern, 1975, J. Mol. Bio., 98, 503-517), then DNA was fixed by exposure to 1200 micro joules of short wavelength UV. Filters were hybridized at 65 ° C overnight in hybridization buffer (SSC 6x, Denhardt 5x solution, 0.5% SDS and 20 micrograms per ml denatured salmon sperm DNA) containing, per ml , 2 x 106 counts per minute of the specific fragment of AR labeled with The probes were labeled statistically up to a specific activity of 5 to 25 × 10 8 counts per minute per microgram. Filters were washed thoroughly in 1x SSC at 65 ° C, then autoradiographed overnight at -70 ° C.

9.1.2. Construction of the genomic library.

The lambda bacteriophage L41.1 was chosen as the cloning vector and arms treated with phosphatase were prepared after digestion with BamHI, EcoRI and HindIII. We put high weight DNA 92

molecular analysis of MCF-7 cells to be digested with HindIII, then electrophoresis was carried out on an 0.8% agarose gel at low temperature (BioRad), then dimensional fractionation. The agarose fraction, enriched to the maximum in the desired restriction fragment, was melted at 65 ° C., then the DNA was extracted with phenol and NaOAc, after which it was precipitated with ethanol and resuspended. suspension at 25 to 100 micrograms per ml. The constitution of the libraries consisted in an overnight ligation at 14 ° C. with 100 ng of lambda L41.1 arms and 40 nG of DNA.

fractionated according to dimensions and extracted, in a reaction volume of 5 microliters, in the presence of T4 DNA ligase (Biolabs). The recombinant phage was packed in vitro with extracts (Grosveld, 1981, Gene, 13, 227 to 237) prepared with the following E. coli strains: BHB2688 N205 recA “(lambda imm4 ^ 4 cITS be red3 Eam4 Sam7) and BHB2690 N205 recA “(lambda imm434 cits be red3 Daml5 Sam7). Libraries were titled on E. coli LE392. Genomic clones were unraveled with AR-specific DNA probes by in situ plaque hybridization (Benton, 1977, Science 196, 180-182) and DNA was isolated from purified positive plaque clones (Huynh , 1985 in DNA cloning techniques: a practical approach, D. Glover, ed. (Oxford: IEL Press), pages 49 to 78). The HindIII inserted segments were excised and subcloned into pEMBL 18 for restriction analysis and propagation.

9.1.3. Assay by primer extension.

The RNA from MCF-7 cells was isolated as described in section 8, above. Synthetic oligonucleotides complementary to nucleotides 40 to 60 (ARAP) and 76 to 97 (ARCP) in the AR cDNA sequence were labeled using 93 32p with T4 polynucleotide kinase to a specific activity of 2 to 5 x 10® shots per minute per microgranime. 1,000,000 counts per minute of labeled oligonucleotide were used to prime synthesis of the first strand of cDNA on 50 micrograms of MCF-7 RNA, much as in section 8.1.4. above. The products were treated with RNase A, extracted with phenol and chloroform, precipitated with ethanol, heat denatured in 80% formamide at 100 eC for 5 minutes and analyzed by gel electrophoresis ordinary sequencing of 8% polyacrylamide with 7M urea.

9.1.4. Assay by CAT.

MCF-7 cells were grown as described in section 6.2.1. above. For each assay, 1 to 2 x 10 ^ cells were plated in a 100 mm dish in 10 ml medium and 24 hours later, 20 micrograms of plasmid DNA were added to the cells (Southern and Berg, 1982) in superhelix precipitated by calcium phosphate. The cells were rinsed with fresh medium 4 hours after transfection and shocked with 25% glycerol for 90 seconds. The cells were rinsed again and covered with 20 ml of fresh medium containing 0 or 100 ng of TPA per ml. The cells were washed and harvested 40 hours after glycerol shock, then lyzed by ultrasound in 100 microliters of 0.25 M Tris.HCl (pH 7.8). The CAT activity was assayed essentially as described by Gorman et al. (1982). The cell extract (3 to 50 microliters) was added to 2.5 mCi of ^ C-chloramphenicol- (NEN) in a reaction volume of 150 microliters of 0.5 M Tris (pH 7.8) and 0.05 mM acetylCoA. The reaction mixtures were incubated for 2 hours at 37 ° C, then extracted with 1 ml ethyl acetate and developed on TLC silica gel plates with chloroform: 1-butanol (95 : 5). The CCM plates were dried and autoradiographed. Acetylated and non-acetylated ^ c-chloramphenicol was assayed quantitatively by counting the excised spots in Optifluor. The number of CAT enzyme activity units was calculated in micrograms of acetylated chloramphenicol per hour per micrograms of protein in the cell extract.

9.1.5. In situ chromosomal hybridization.

Plasmid pAR9 was statistically labeled with ^ h-ttp which was used for hybridization with metaphase human cells prepared from phytohemagglutinin stimulated peripheral blood lymphocytes at the 500 to 500 band stage. 800. This probe corresponds to an AR cDNA which is greatly lacking in the 3 'untranslated region and inserted into the EcoRI site of pEMBL18.

Hybridization was carried out as described previously (Le Beau, 1985, PNAS, 82, 6692 to 6696) with 2, 4, 20 and 40 ng of probe per ml of hybridization mixture. Autoradiographs were prepared with Kodak NTB-2 nuclear emulsion by exposing the slides for 7 to 60 days. The chromosome bands were visualized by quinicrine-raoutarde staining.

9.2. Location of the AR gene in the chromosome.

The human AR gene has been located in the chromosome by in situ hybridization of the 3 h-labeled AR cDNA with chromosomes in normal metaphase. The silver grains were counted on 50 metaphase spreads or 30% were distributed non-statistically at bands 4ql3-4q21. The result was confirmed by polymerase chain reaction (PCR) 95 on hybrid DNA from somatic hamster / human cells containing only the human chromosome 4. The oligonucleotide primers from exon 3 of AR and the flank intron generated a 220 base pair PCR fragment only in human DNA and hybrid DNA containing chromosome 4, while hamster DNA remained negative,

The 4ql3-4q21 region of the chromosome also contains the genes for gro or activity simulating the growth of melanoma (Anisowicz, A, et al., 1987, Proc. Natl. Acad. Sci. USA, 84, 7188 to 7192? Richmond, A . et al., 1988, EMBO J. 7, 2025 to 2033), the c-kit receptor (Yarden et al., 1987, EMBO J. 6, 3341 to 3351), platelet factor 4 (PF4) (Griffin et al., 1987, Cytogenetic Cell Genet., 45, 43 to 73), the inducible factor of interferon gamma OP-10 (Luster, AD et al., 1987, Proc. Natl. Acad. Sci. USA, 84, 1868 to 1871), vitamin D binding protein (group specific component) (Cooke, NE et al., 1986, Human Genetics 73, 225-229; McCombs, JL et al., 1986, Cytogenet Cell Genet, 42, 62-64) and statherine which is a saliva protein regulating calcium (Sabatini, LM et al., 1987, Am. J. Hum. Genet., 41, 1048 to 1060). The gene for EGF is located distally at 4q25.

Gro, IP-10 and PF4 belong to a class of peptides related by their structure which can constitute a family of growth factors gathered on chromosome 4. The c-kit codes for a cellular surface receptor which is in relation of structure and function with different growth factor receptors that contain tyrosine-specific kinase activity, including the 96 EGF receptor, the Neu oncogene, the PDGF receptor and the insulin receptor. Generally speaking, the genes for ligands and their receptors are. mapped on separate chromosomes, but some are mapped on common chromosomal locations (Groffin, 1983, Nucl. Acids Res. 11: 6331 to 6339; Pettenati, 1987, Proc. Natl. Acad. Sci. USA, 84: 2970 to 2974 ). The ligand for c-kit has not been identified and the RA should be tested for such activity.

Specific translocations are observed in many malignant conditions. The most frequently observed cytogenetic aberration in congenital acute lymphoblastic leukemia (ALL), t (4; 11) (q21? Q23) involves the region on which the RA was mapped (Heim, S. et al., 1987, Leukemia, 1, 16-23). ALL are now classified by morphology as defined by the Franco-American-British Cooperation Group (FAB), markers of B and T cells and chromosomal analysis. These classifications serve as the basis for the diagnosis, prognosis and treatment of ALL. A third of all ALL cases involve specific translocations and t (4; ll) corresponds to a group with an unfavorable prognosis, very common in children below 16 months with a median survival of less than 1 year ( Kocova et al., 1985, Cancer Genetics and Cytogenetics, 16, 21-32). Trans locations involving the 4q21 region have also been observed in T lymphomas (Levine, EG et al., 1986, Cancer Res., 46, 6481 to 6488) and in a case of acute myeloblastic leukemia (AML) (Selypes, A et al., 1987, Human Genet, 76, 106-108). This region also contains the genes for a syndrome of dental dysgenesis and the "piebald 97! Trait" which is an inherited disorder leading to speckled pigmentation of the skin due to deficient migration and differentiation of melanoblasts, (Hoo, JJ and collaborators , 1986, Human Genet., 73, 230-231).

The studies of factorial links between AR and genetic disorders located on the chromosome at 4ql3-4q21 should make it possible to determine the significance of this co-localization. Current cytogenetic analysis suggests that the 4q21 region contains genes involved in lymphocyte differentiation. As a last resort, these associations may suggest other biological activities or applications for this growth-regulating molecule.

9.3. Genomic cloning.

Southern blot analysis (section 9.1. Above) of MCF-7 DNA digested with HindIII, EcoRI or BamHI revealed unique bands of 12 kb, 8 kb and 20 kb, respectively, when hybridization with a probe of 830 base pairs (pAR1 digested by Pvull / Bsml and labeled with 32P) containing the 5 ′ part of the AR gene, which suggests that the latter is a single copy gene. A 170 base pair probe (AR 170, BsmI-PvuII) from the 3 'end of the mature AR, including the coding regions of the transmembrane and cytoplasmic domain, hybridized to the same HindIII fragments , EcoRI and BamHI and also on an additional 7.5 kb HindIII fragment, which implies that most of the coding region of the AR is divided between two HindIII fragments of 12 and 6.5 kb. Identical bands were observed by Southern analysis of DNA digests of placenta, brain, melanoma (SK-ML 28), breast cancer (HTB36), 98 squamous cell carcinoma (A431) and cancer of the human lung, suggesting that the AR gene has not undergone any significant reorganization or amplification.

We have chosen to clone the two HindiII fragments of MCF-7 DNA because it is likely that it contains most, if not all, of the AR gene and its flank sequences. HindIII-digested MCF-7 DNA was subjected to dimensional fractionation and the appropriate fractions were ligated in L47.1. Among the many positive clones, two clones, namely 7 ^ ARH12 and ARH6, were selected for more detailed characterization. The inserted 12 and 6.4 kb segments were subcloned into pEMBL18 and various restriction sites were mapped. With exact oligonucleotide primers and by direct sequencing of various smaller subclones, the sequence of all exons and their junctions to introns was determined and revealed no disagreement with the cDNA sequence.

9.4. Characterization of the regulatory region 51.

Genomic cloning of the RNA led to the isolation of a 6.5 kb 5 'flank sequence contained in the 12 kb HindIII fragment. The fragment of 688 base pairs from the first EcoRI 5 ′ site of exon 1 to the SmaI site in exon 1 (position 40 in the cDNA) was subcloned and sequenced for the two strands. The data are shown in Fig. 17.

A primer extension was carried out on the MCF-7 RNA by means of two distinct oligonucleotides originating from exon 1. A major transcription starting site, confirmed by the two oligonucleotides, was located at 1 pair of bases at 99 5 'on the longest cDNA clone. At positions +1 and +2, two minor sites were observed in the cDNA sequence, which confirms that many of the cDNA clones are practically full length. To functionally confirm the regulatory region of the AR gene, the Applicant has constructed a hybrid gene (pXARElCAT) containing the 5 'flank region of the EcoRI AR to Smal AR of 688 base pairs (sequence 648 to +40 with + 1 as a major transcription starting site) causing the expression of a chloramphenicol acetyltransferase (CAT) gene without promoter. This constructed product was able to stimulate transcription of the CAT gene when it is introduced transiently into MCF-7 cells and the activity was stimulated 6 to 7 times by the addition of TPA. This confirms both structural and functional points of view that the Applicant has cloned the 5 'end of the AR gene. The Applicant then built a series of mutants by deletion 5 'CAT containing the sequences of the AR -539 to +40 (Ela), -387 to +40 (Elb), -277 to +40 (Elc), -148 at +40 (Eld), -77 to +40 (Ele), -79 to +40 (Elg) or +19 to +40 base pairs (Elh) in the same vector. The basic activity was lost during the deletion beyond the -77 part, i.e. Elh did not show any measurable activity, and all of these constructed products showed an improved expression of 3.5 to 7 times in response to a 40 hour treatment with 100 ng per ml of TPA. These constructed products can also be used for transient assays to evaluate the effectiveness of morphogens, mitogens, growth factors, drugs or any crude fermentation extracts on the regulation of the expression of RA, this which thus makes it possible to identify new therapeutic agents.

100

Nuclear fast passage experiments reveal 3 to 5-fold increased AR transcription in response to TPA, suggesting that increased promoter activity is at least partially responsible for TPA induction of AR . The following Northern analysis of MCF-7 cells treated with TPA in the presence or absence of actinomycin D identifies the AR as being a relatively stable RNA having a half-life greater than 4 hours. The induction of the expression of AR under the effect of TPA therefore has multiple aspects involving an increased transcription of a stable mRNA.

9.5. Intron / exon organization of the AR gene / protein domains of the AR, evolutionary and functional implications.

The complete genomic sequence of human amphiregulin is shown in FIG. 17 and the relationship between the exons and the protein domains of the AR molecule is shown schematically in FIG. 18.

Primer extension analysis of AR mRNA and studies using AR / CAT promoter constructs (CAT = chloramphenicol acetyltransferase which is a marker gene) locates a functional promoter in the 64 base pairs in 5 'before the end of the longest cDNA clone. Therefore, there is a 29 base pair TATA consensus box upstream of the 5 'end of the cDNA sequence. The 3 'end of the gene is preceded by a consensus polyadenylation signal sequence at 64 nucleotides from the poly (A) tail in cDNA. The primary transcribed product of the AR gene is approximately 10.2 kb.

Six exons encode the precursor of human AR and occupy 10.2 Tcb of genomic DNA. The »» 101 AR exons are 112 to 270 base pairs in length and are interrupted by 1.25 kb to 2.1 kb introns. The five introns of the AR interrupt the coding sequence, so that many of the protein domains are products of a single exon. Exon 1, code for the signal and 5 'untranslated peptide domains, exon 2 codes for the N-terminal precursor, exon 5 contains the cytoplasmic region and exon 6 represents the 3' untranslated region . Together, exons 3 and 4 encode the mature AR protein, including hydrophilic and EGF-like sequences, as well as the suspected transmembrane domain. The junction between exons 3 and 4 takes place between the second loop and the third of the EGF-like region.

When the exon junctions are superimposed on an alignment of the amino acid sequences of AR, EGF and TGF, it is obvious that all these proteins are coded by two exons and that the interrupt intron is located at the same position. In addition, the 3 'exon of each also codes for the transmembrane domain of the corresponding precursor proteins. On the contrary, the EGF-like region is contained in a single exon in all the other mammalian EGF homologs for which the exon structure has been determined, in particular the 9 repeating units analogous to EGF in the precursor of EGF (Gray, 1983, Nature, 303, 722-725); all three in the LDL receptor (Yamamoto, 1984, Cel, 39, 27 to 38); and that in rat fibronectin (Patel, 1987,

Embo, J., 6, 2565 to 2572) and human coagulation factor XII (Cool, 1987, J. Biol. Chem., 262, 1366 to 1373). Invertebrate counterparts, such as the Drosophila Notch gene and C. elegans lin-12 also contain multiple EGF-like repeat units 102 (Kiud, 1986, Molec. Cell. Biol., 6, 3094-3108; Greenwald , 1985, Cell, 43, 583-590). Unlike mammalian genes, most of the EGF-like repeating units in Notch are contained in a single coding region in which only 4 of 36 motifs are separated by intervening sequences. It is remarkable that 2 of these 4 motifs show a tighter alignment with the AR than with the repetitive consensus of Notch and that one is even interrupted by an intron with the same CXC sequence as EGF, TGFûf and AR. The nematode lin-12 gene contains at least 11 EGF-like units, 9 of which are located on the first exon and the last two of which are on separate exons, with the intact EGF-like unit flanked by introns.

Differences in the structure of exons encoding EGF-like repeating units from different organisms suggest that they may have different origins. One group contains the intron-delimited EGF-like motif and the other group, of which the AR is a member, has an interrupted motif. The similarity of the sequences between the two groups could be the result of a convergent evolution or else of an insertion of intron after their divergence from a common ancestral gene.

The localization of the intron in the same CXC sequence of EGF, TGFo (and AR and one of the repeating units of Notch suggests a common origin, but a more detailed examination reveals a phase difference of the introns. The intron phase means that the intron interrupts the reading frame after the first (I), second (II) or third (0) nucleotide of a codon. This phase is believed to be of evolutionary importance in since it allows the slow passage of exons containing a functional domain 103 among various proteins. EGF and TGF Cl have a phase II intron in the EGF-like unit, while AR and Notch have phase introns I. A similar displacement of a nucleotide is observed in the penultimate intron inside the protease domain of complement factor B and elastase (Cambell, 1983, PNAs, 80, 4464; Swift, 1984 , J. Biol. Chem. 259, 14271) The non-statistical positions of these introns may imply that an insertion sequence of introns ons specific exists in these genes. Alternatively, there may have been selection pressure to cut the coding region at this cycle. Comparison of sequences at exon-intron junction sites in the EGF-like unit for EGF, TGF, human AR and Notch's first repeating unit does not indicate direct repeating units as they are often observed with transposable elements. However, all four have the sequence 'gtaagt' forming the boundary of the junction. This sequence may play a certain role in the origin or in the splicing of this intron.

Viral homologous EGFs do not contain introns, but alignment between VGF, EGF, TGF and AR reveals a homology extending across the transmembrane domain, while the growth factor of myxomatosis viruses and Shope papilloma ends before this hydrophobic region. Recent studies provide indications that the precursor of TGF û (is synthesized as a transmembrane protein and that the differential proteolytic cleavage leads to the secretion of larger forms which may be specific for the cell type (Teixido, 1987, Nature , 326, 883 to 885).

Other homologous EGFs that contain transmembrane domains include coagulation factors 104, LDL and the homeotic Notch and lin-12 genes, although none are adjacent to an EGF-like repeating unit. The presence of the EGF motif in different precursors linked to the membrane suggests that in certain circumstances, it may be treated as a secreted growth factor or else may remain associated with the membrane and may play a role in intercellular and / or intracellular communication.

Some of the homologous ARs include homeotic invertebrate genes. Homeotic gene products are involved in the regulation of cell development. For example, the Notch gene product is involved in the correct conversion of an ectodermal cell into a neuroblast or a dermoblast. In Notch mutants, all of these cells become nerve cells and the offspring, all of the brain and none of the skin, die. A homeotic gene can function autocrine to establish or maintain a specific state in cells expressing the gene product and may be involved in the regulation of such states in adjacent cells through the intervention of cell-to-cell interactions. The fact that RA also intervenes or does not intervene in the regulation of the state of development of certain types of cells should be investigated.

9.6. Treatment of amphirequline at maturity.

Characterization of the AR cDNA revealed that the 78- and 84-amino acid forms of AR are synthesized as the middle part of a transmembrane precursor of 252 amino acids (section 8.2 above). The sites for the proteolytic cleavage of the AR precursor which could lead to the release of the AR at maturity do not correspond with the cleavage sites by any known protease. At the N-terminal end, the sites are Asp-Asp / Ser-Val and Glu-Gln / Val-Val, while the C-terminal site is Glu-Lys / Ser-MET. The two forms of AR appear to be the result of an alternating proteolytic treatment of a common precursor, since all the cDNA and the genomic clones revealed the same sequence in this region. An interesting fact is that an intron is located between the two N-terminal cleavage sites and it is possible that this "intron slip" can also explain these differences.

The MCF-7 cells produce 80% of their AR in the 84 amino acid form which is the largest, while the cell line of choriocarcinoma HTB-36 produces 80% of its AR in the 78 amino acid form which is the smaller. To determine whether this difference is associated with alterations in DNA, the Applicant has used the polymerase chain reaction (Scharf, 1986, Science, 233, 1076 to 1078) to isolate exon 3 from the AR of MCF-7 cells and HTB 36 cells, which constitute a line which constitutively produces large quantities of AR mRNA. An AR intron-specific "sense" oligonucleotide and an "antisense" oligonucleotide from the exon 3 region were used to specifically amplify the 220 base pair intermediate fragment of the AR gene from both sources. 'DNA. Direct sequence analysis did not reveal any disagreement in the intron / exon junctions, or in the coding sequence of exon 3 between these two sources of human DNA, which suggests more than the two forms of the AR result from an alternating proteolytic cleavage of the precursor.

106 9.7. Comparison of structure of growth factors AR and analogues to EGF.

AR shows clear sequence homology with other EGF-like proteins, with the conservation of the 6 cysteines involved in three disulfide bonds which define the secondary structure of growth factors at maturity. Previous (computer-aided) comparisons of amino acid sequences have led to the division of EGF-like motifs into two distinct groups: growth factors and blood clotting factors (see Fig. 13). The Applicant has selected two sequence motifs for comparison with AR and other known DNA or protein sequences. Selection was based on the apparent ability of these sequences to distinguish between the two groups of proteins and because very rare interruptions were necessary for optimal alignment. The exact sequences are given in Table I. The first region corresponds to the first 11 amino acids of the second loop of EGF and the second corresponds to the third cysteine loop of EGF. Four representative examples of each group of growth factors and blood coagulation factors were selected to generate consensus sequences based on their well-established functional and structural homology. Human sequences were selected whenever possible, since the Applicant wanted to compare them to human AR as a last resort. The growth factors in question are: human EGF, TGF and VGF and the growth factor of Shope papilloma and the coagulation factors are human factors IX, X and XII and protein C. The RA was also subjected to a comparison with the databases for these two blocks of homology. The consensus sequence has been weighted based on the frequency with which a remainder appears in a given position.

The structure-function analysis of various derivatives of TGFC ^, VGF and EGF has recently led to the identification of various residues which are necessary for the biological activity of these growth factors. Recombinant proteins, synthetic peptides, site-specific chemical derivatives and proteolytic degradation have all been useful in producing modified molecules. Some generalizations include (the positions are relative to the alignment in Fig. 15): the 6 cysteines (positions 1, 19, 15, 26, 28 and 39) and their disulfide loops (1-15, 9-26, 28 -37) are necessary for biological activity, N-terminal extensions have little effect on activity, an aromatic residue (F, Y) is necessary at position 8, a non-conservative modification of γ32 # d41 or L ^ 2 leads to a loss of activity and / or a serious decrease in binding to the receptor or in autophosphorylation.

With the exception of the last two, the AR

meets all the criteria which define the residues necessary for binding to the EGF receptor and / or mitogenic activity. The mature AR is cut just before D ^ 1 and is absolutely free of this and of "crucial" leucine 42, but still competes for binding to the EGF receptors and replaces the EGF in certain dosages of mitogens. However, these differences may be responsible for the unsaturable kinetics of binding to receptors and for its marked functional differences with EGF in 108 certain assays, such as synergy with TGFyô, the differential effect on the selected A431 subclones. , cross-links, phosphorylation assays, and the ability to bind to DNA. The extremely hydrophilic region at the N-terminus of the mature AR can also confer some of these differences in receptor binding or biological activity.

9.8. AR as a subcategory of the EGF superfamily.

The Applicant has calculated a matrix of evolutionary distances based on the alignment of FIG. 13 and the computer analysis based on Table I. This matrix was used to deduce an evolution tree for the EGF superfamily. The search criteria requires the presence of the cysteines and adds a point for each remainder that suits the weighted consensus sequence. This tree of evolution predicts that AR is placed in a new subcategory of growth factors similar to EGF, on the basis of both structural and functional homology. Such a model predicts that other members of this family of growth factors may exist and that they can be identified by homology with the above sequence diagrams based on a ternary criterion: a combined score of two regions of growth factors. growth greater than 20, a combined rating of coagulation factor regions less than 20 and a combined rating of AR regions greater than 20. All new molecules meeting these criteria would be expected to have some homology of function with EGF , TGFty, VGF or AR. Molecules with a combined AR region score greater than 40 would be classified as members of the AR family within the EGF superfamily and would fall within the scope of the present invention.

10. Example; bacterial expression of amphirequline.

10.1. Materials and processes.

10.1.1. Plasmid constructs.

The plasmid pARDl. The plasmid pARD1 contains the 1.4 kb EcoRI cDNA fragment (pAR1) in pEMBL18 with a modification of T to C at position 532 of the nucleotides in the cDNA, corresponding to the valine-valine sequence at the amino-terminal end of mature AR. This modification of the base was carried out by site-directed mutagenesis and was confirmed by sequence analysis. The constructed product gives access to the junction between the amino-terminal domain of the precursor of AR and the hydrophilic region by creating a DdeI site (CTAAG).

The plasmid pARSTOP. The plasmid pARSTOP is an intermediate construct used in the preparation of an AR secretion vector. PARD1 was digested with Sspl and Xbal and the fragment of 525 base pairs encoding the 9 carboxy-terminal amino acids of mature AR, the transmembrane and cytoplasmic domains and the 3 'untranslated region was isolated . The 4.7 kb fragment (Sspl-Xbal) containing the remaining amino-terminal part of the AR cDNA was gel-purified and ligated with the complementary kinase-treated and renatured ARSTOP1 and ARSTOP2 oligonucleotides (see below). -after). These oligonucleotides have a non-sticky 5 ′ end compatible with an Sspl site, followed by a sequence coding for the 9 carboxyterminal amino acids of the mature AR sequence, a TAA stop codon and EcoRV and XbAI sites.

110

YFGERCGEK * EcoRV / Xbal ARSTOP1 5 'ATTTCGGTGAACGGTGTGGGGAAAAGTAAGATATCT

ARSTOP2 3 'TAAAGCCACTTGCCACACCCCTTTTCATTCTATAGAGTC

- The plasmid pbAR. The plasmid pbAR contains a TGFβ signal sequence without promoter attached to the mature form of the AR comprising 78 amino acids.

It was constructed by ligating the oligonucleotides bLARN3 and bLARN4 with the 240 base pair Ddel to Xbal fragment from pARSTOP in PEMBL18 digested with EcoRl / Xbal. bLARN3 and bLARN4 are complementary and create a 5'EcoRI overhang and a 3'Ddel overhang and a unique internal Nael site compatible with that close to the carboxyterminal end of the TGF β signal sequence. Ligation destroys the Ddel site and regenerates the correct amino acid valine-valine sequence at the amino terminal end of the mature AR.

Sstll

AGVVKPPQNKTESE PHIL1 5 'GGGAGTAGTTAAGCCGCCCCAAAACAAGACGGAAAGTGA

PHIL2 3 'CGCCCTCATCAATTCGGCGGGGTTTTGTTCTGCCTTTCACT

NTSDKPKRKKKGG PHIL1 5 'AAATACTTCAGATAAACCCAAAAGAAAGAAAAAGGGAGG

PHIL2 3 'TTTATGAAGTCTATTTGGGTTTTCTTTCTTTTTCCCTCC

EcoRI

KNGKNRRNRKKKN PHIL1 5 'caaaaatggaaaaaatcgaagaaacagaaagaagaag PHIL2 3' gtttttaccttttttagcttctttgtctttcttcttcttaa \ ί 111

The plasmid pDCHBARl. The plasmid pDCHBAR1 is a mammalian expression vector (Fig. 19) designed to secrete the mature form of the treated AR containing 78 amino acids. It contains the mature TGF / 3 / AR signal sequence from pBAR activated by the CMV / HIV promoter, which is flanked by a SV40 polyadenylation signal. The vector also contains SV2dhfr in the same transcription orientation. PSVDR / boM was subjected to digestion with Nael and Xbal and the 6 kb vector was separated from the excised OncoM coding sequence. The pbAR was also digested with Nael and Xbal, and the fragment of 260 base pairs encoding the 5 carboxyterminal amino acids of the signal sequence TGF Jb, and the mature AR sequence of 78 amino acids followed by 'a stop codon and EcoRV and Xbal sites. The junctions were checked by sequence analysis.

The plasmid pDCHBPHILE. The plasmid pDCHBPHILE is a mammalian expression vector designed for the secretion of a hybrid AR / EGF protein. This plasmid is also configured to facilitate the future fusion of constructs between the hydrophilic domain of the AR and other growth factors. This constructed product was formed by ligating the vector fragment pDCHBAR1 digested with Sst11 / xbal 6.5 kb, the EcoRl / Xbal fragment of 175 base pairs containing the synthetic human EGF gene and the oligonucleotides PHIL1 and PHIL2. These oligonucleotides are complementary with the 5 'Sst11 and 3' EcoRI extensions and code for the last remainder of the TGFβ signal sequence and the entire hydrophilic domain of the AR. PDCHBAR1 provides the CMV / HIV promoter, except the last amino acid of the signal sequence TGFβ, and the polyadenylation sequence of SV40. The EGF fragment was acquired from Doctor Timothy M. Rose (Oncogen) and 112 includes sites suitable for future manipulation.

10.1.2. Preparation of the vector pTAC.

The plasmid TacPak / EGF was acquired from Doctor Timothy M. Rose (Oncogen) and is composed of the following units: the hybrid trp-lac promoter and the Shine-Delgarno sequence of the Cro gene isolated in the form of a BglOO / BamHI fragment the expression vector pl35-l # itself derived from pDR540 (Pharmacia); the alkaline phosphatase sequence originating from the synthetic oligonucleotides TacPakl and TacPak2 (Doctor Rose, Oncogen), the synthetic EGF sequence supplied from the plasmid pBM22 / PAK / EGF; the transcription termination region; the backbone of pBR322 with the neomycin resistance gene, all from the plasmid pl35-1. The TacPak / EGF plasmid was digested with BamHI, supplemented with 2 bases using dATP and dGTP to create a site compatible with Sali supplemented with 2 bases, and then digested with Pvul. This digestion eliminates the synthetic EGF gene and most of the alkaline phosphatase signal sequence, but leaves the Tac promoter, and the Shine-Delgarno sequences and the ATG initiation codon intact. The 2 kb fragment was gel purified.

10.1.3. Preparation of the modified amphiregulin cDNA fragment.

The plasmid pARSTOP was digested with SalI, supplemented with 2 bases using TTP and dCTP to create a site compatible with a BamHI extension supplemented by 2 bases (dATP, dGTP), then digested with Ddel and BglI. This latter digestion was necessary to remove DNA migrating simultaneously out of the desired fragment of 242 base pairs which codes for the short form of mature AR beginning at 113 at VKPP and ending at CGEK, which follows an introduced stop codon synthetically. The 243 base pair fragment was gel purified.

10.1.4. Preparation of synthetic oligonucleotides a ALKPAR1 and ALKPAR2.

The complementary synthetic oligonucleotides ALKPAR1 and ALKPAR2 were constructed with a 5 'Pvul overhang compatible with a partially completed Pvul / BamHI fragment of TacPak / EGF and a 31 Ddel extension compatible with the partially completed DdE1 / SalI fragment of pARSTOP. They were synthesized on an apparatus for synthesis of Applied Biosystems oligonucleotides and purified on an acrylamide gel. Phosphates were added to the oligonucleotides using T4 kinase and equimolar amounts were renatured by slow cooling after thermal denaturation.

Pvul

IALALLPLL ALKPARl 5 'CGCCCTCGCACTTCTCCCACTGCTG

ALKPAR2 3 'TAGCGGGAGCGTGAAGAGGGTGACGAC

Ddel FTPVTKAVV ALKPARl 5 'TTCACTCCAGTGACAAAAGCTGTAG 3' ALKPAR2 3 'AAGTGAGGTCACTGTTTTCGACATCAAT 5' 10.1.5. Ligation and isolation of pTacAPARl.

The partially completed AR Ddel to Sali fragment of 243 base pairs, the partially completed Pvul / BamHI vector fragment of 2.8 kb of TacPak / EGF and the renatured and kinase-treated oligonucleotides ALKPAR1 + 2 were assembled using DNA ligase, then transformed "into competent E. coli JM109 cells and then selected on LB / neomycin plates. The correct construct was confirmed by digestion with restriction enzymes and DNA sequencing. pTacAPAR sequence is given in Fig. 20.

10.1.6. Preparation of the hybrid gene fragment hydrophilic domain of AR / EGF.

v The plasmid pDCHBPHILE was put to digest with

Sstll and Xbal to generate the fragment of 286 base pairs which codes for the last 2 residues of the signal sequence TGF fi> (AG), 37 residues of the hydrophilic domain of AR (VVKP ... RKKK) and the synthetic sequence of 53 remains of the mature human EGF sequence (NSDS ... WELR). This 286 base pair fragment was gel purified.

10.1.7. Preparation of the oligonucleotides APAREGF1 and APAREGFl.

The complementary synthetic olinucleotides APAREGFl and APAREGF2 are designed with a 5 'overhang Pvul compatible with the Pvul / Xbal fragment of pTacAPARl and with a 31 Sstll extension compatible with the Sstll / Xbal fragment of pDCHBPHILE. The oligonucleotides were synthesized on an apparatus for the synthesis of Applied Biosystems oligonucleotides and purified on an acrylamide gel. Phosphates were added to the oligonucleotides using T4 kinase and equimolar amounts were renatured by slow cooling after thermal denaturation.

115

Pvul

IALALLPLL APAREGF1 5 * CGCCCTCGCACTTCTCCCACTGC

APAREGF2 3 'TAGCGGGAGCGTGAAGAGGGTGACG

Sstll

F T P V T K

APAREGFl 5 ’TGTTCACTCCAGTGACACC 31 APAREGF2 3 'ACAAGTGAGGTCACTGTGGGG 5' 10.1.8. Ligation and isolation of pTACAPHILE.

The 286 base pair Sstll / Xbal fragment coding for the hydrophilic domain of the AR united to the mature EGF sequence, the 2.8 kb pTacAPARl PvUl / Xbal vector fragment and the renatured and treated oligonucleotides APAREGFl + 2. kinase were assembled using DNA ligase, transformed into E. JM109 cells. competent and selected coli on LB / neomycin plates. The correct construct was confirmed by digestion with restriction enzymes and DNA sequencing. The nucleotide sequence of pTacAPHILE is given in FIG. 21. The intended translation porduit should be cut in the Ala-Gly dipeptide immediately after the alkaline phosphatase signal sequence, thus adding an additional residue (glycine) to the hydrophilic domain of the AR. Nucleotide remains that differ from the normal human sequence are written in bold type (Fig. 20).

10.1.9. Purification of the recombinant amphiregulin.

Plasmids pTacAPAR1 and pTacAPHILE were transformed into E. coli JM109 cells. Competent coli which have been grown to confluence in 10 ml of LB medium at 37 ° C to an A ^ qq of 0.7. The cultures were then subjected to induction using 100 micrograms of iPTG and growth was allowed to continue for 24 to 72 hours. The cultures were centrifuged twice at 5000 rpm, each time for 15 minutes, the pellet was collected and the purification was continued with the supernatant. The samples were concentrated 10 times in an Amicon ultrafiltration apparatus with YM5 membranes (cut at a molecular weight of 5000). The concentrate was diluted with 5 volumes of MilliQ water and rediluted to one tenth of the original culture volume. Glacial acetic acid was added to 1M, then the samples were stored at 4 ° C for 2 to 24 hours. The samples were centrifuged at 19,000 rpm for 20 minutes at 4 ° C in Oakridge tubes in the SS34 rotor, then the pellet was extracted with 20 ml IM acetic acid and pooled supernatants. The clarified supernatants were then subjected to dialysis against 0.1M acetic acid for 2 days, then lyophilized and stored at -20 ° C.

By immunoblotting and evaluation of the growth inhibitory activity (GIA), the AR protein was assayed in the cell pellets and crude supernatants brought to dryness. The cell pellet was passed from 50 microliters of confluent culture or from 100 to 200 microliters of supernatant brought to dryness on a polyacrylamide gel containing 16% tricin, then the staining was carried out with solid green and the transfer carried out on nitrocellulose. Western blots were performed as described in section 7.1.6. above.

10.2. Results and discussion.

The pellet of transformants carrying pTacAPAR1 revealed abundant amounts of an immunoreactive protein migrating to 10 kD, as well as a certain amount of degradation products and probable dimers. The supernatants contained approximately 117,100 to 200 ng per ml of AR. secreted immunoreactive. The measurement of the growth-inhibiting activity of the supernatants indicated approximately 150 ng per ml of AR for the control cell line A431-A3, the comparison being made with standards of purified native AR. Large amounts of immunoreactive protein were produced in these systems and remained mainly contained in cells. The assay on the periplasmic preparations and the supernatants revealed a secretion of active proteins after 2 to 3 days of growth.

The plasmid pTacAPHILE constitutes a convenient means for fixing the hydrophilic domain of the AR on the N-terminal end of any cloned gene, this region could confer modified binding characteristics on the normal ligand receptor, serve as a localization sequence nuclear, bind to DNA or allow transport through lipid membranes or the like which are normally impermeable to the attached factor.

11. Deposit of microorganisms.

The following microorganisms have been deposited with the Agricultural Research Culture Collection, Northern Regional Research Center (NRRL) and have been given the following access numbers:

Microorganimes Plasmid Access number

Escherichia coli HB101 pARl

Escherichia coli HB101 pARH12

Escherichia coli HB101 pARHö

Escherichia coli JM109 pTacAPARl

Escherichia coli JM109 pTacAPHILE

The invention is not limited to the deposited cell lines 118 or to the described embodiments which are susceptible of numerous variants and modifications without departing from the scope of the invention.

It should also be noted that all of the amino acid and base pair residue numbers and all dimensions given for nucleotides and peptides are approximate and of descriptive utility.

Claims (39)

119 1. Protein, characterized in that it has an amino acid sequence comprising î 1 10 20 VVKPPQNKTE SENTSDKPKR K K K. G G K 30 40 50 N G K N RRNRKKKNPC NAEFQNFCIH 60 70 GECKYIEHLE AVTCKCQQEY FGERCG 78. K 2. Protein according to claim 1, characterized in that it has a molecular weight of approximately 8500 to 25,000 daltons. 3. Protein according to any one of claims 1 and 2, characterized in that it has a pi of approximately 7.6 to 8.0. 4. Protein according to claim 1, characterized in that it is glycosylated. 5. Protein according to claim 1, characterized in that it is non-glycosylated. 6. Protein, characterized in that it has an amino acid sequence comprising î 1 10 20 SVRVEQVVKP PQNKTESENT SDKPK 30 40 50 RK K KG GKNGKNRRNR KKKNPCNAEF 60 70 QNFCIHGECK YIEHLEAVTC KCQQE 120 80 84 Y F G E R CGEK 7. Protein according to claim 6, characterized in that it has a molecular weight of approximately 9100 to 25,000 daltons. 8. Protein according to any one of claims 6 and 7, characterized in that it has a pi of approximately 7.6 to 8.0. 9. Protein according to claim 6, characterized in that it is glycosylated. 10. Protein according to claim 6, characterized in that it is non-glycosylated. 11. Protein according to any one of claims 1 to 5 or peptide or fragment thereof, characterized in that it inhibits the growth of human cancer cell lines of epithelial origin. 12. Protein according to claim 11, characterized in that the human cancer cell line of epithelial origin comprises cells of the squamous cell carcinoma of the vulva A431 or of the mammary adenocarcinoma HTB132. 13. Protein according to any one of claims 1 to 5 or peptide or fragment thereof, characterized in that it stimulates the growth of fibroblasts of the human foreskin cultivated in vitro. 14. Protein according to claim 13, characterized in that the fibroblasts of the human foreskin constitute the Sakamoto or Goodwin cell lines. 15. A protein according to any one of claims 1 to 5 or a peptide or fragment thereof, which comprises a bifunctional regulator of cell growth which inhibits the growth of human cancer cell lines of epithelial origin and stimulates growth human foreskin fibroblasts cultured in vitro. 16. Protein according to claim 15, characterized in that the human cancer cell line of epithelial origin comprises cells of the squamous cell carcinoma of the vulva A431 or breast adenocarcinoma HTB132. 17. Protein according to any one of claims 15 and 16, characterized in that the fibroblasts of the human foreskin constitute the Sakamoto or Goodwin cell lines. 18. Protein according to any one of claims 11 to 17 or peptide or fragment thereof, characterized in that the inhibition or stimulation of cell growth is determined by the following test: (a) wells microtiter each containing 3.5 x 10 ^ cells in 50 microliters of test medium comprising DMEM supplemented with 5% fetal heat inactivated bovine serum (FBS), penicillin / streptomycin and glutamine are incubated for 3 hours; (b) 50 microliters of the test medium containing amphiregulin are introduced into each experimental well, while 50 microliters of the test medium free of amphiregulin are introduced into each control well and are incubated at 37 ° C. for 2 to 3 days; (c) 100 microliters of a solution comprising ^ 25j_ ^ 0 ^ 0_2 · _125_j _ ^^ SOXyUrj [(jine (i-udR) are introduced into each well and incubated at 37 ° C for 4 to 6 hours; (d ) the medium is eliminated and each well is washed once with PBS; (e) 200 microliters of methanol are introduced into each well and incubated for 10 minutes at room temperature, then eliminated? '(f) 200 microliters of an IM sodium hydroxide solution is introduced into each well and incubated at 37 ° C for 30 minutes, and (g) the IM sodium hydroxide solution is collected and subjected to counting in a gamma counter, in which the presence shots of radioactivity is a measure of the incorporation of • ^^^ I-IUdR and cell proliferation, while the absence of shots is indicative of the inhibition of cell proliferation. 19. A protein according to any one of claims 6 to 10 or peptide or fragment thereof, characterized in that it inhibits the growth of human cancer cell lines of epithelial origin. 20. Protein according to claim 19, characterized in that the human cancer cell line of epithelial origin comprises cells of the squamous cell carcinoma of the vulva A431 or of the mammary adenocarcinoma HTB132. 21. Protein according to any one of claims 6 to 10 or peptide or fragment thereof, characterized in that it stimulates the growth of fibroblasts of the human foreskin cultivated in vitro. 22. Protein according to claim 21, characterized in that the fibroblasts of the human foreskin constitute the Sakamoto or Goodwin cell lines. 23. A protein according to any one of claims 6 to 10 or a peptide or fragment thereof, which comprises a bifunctional regulator of cell growth iiiii 123 which inhibits the growth of human cancer cell lines of epithelial origin and stimulates the growth of human foreskin fibroblasts cultured in vitro. 24. Protein according to claim 23, characterized in that the human cancer cell line of epithelial origin comprises cells of the squamous cell carcinoma of the vulva A431 or of the mammary adenocarcinoma HTB132. 25. Protein according to any one of claims 23 and 24, characterized in that the fibroblasts of the human foreskin constitute the Sakamoto or Goodwin cell lines. 26. Protein according to any one of claims 19 to 25 or peptide or fragment thereof, characterized in that the inhibition or stimulation of cell growth is determined by the following test: (a) wells microtiter each containing 3.5 x 104 cells in 50 microliters of test medium comprising DMEM supplemented with 5% fetal heat inactivated bovine serum (FBS), penicillin / streptomycin and glutamine are incubated for 3 hours; (b) 50 microliters of the test medium containing amphiregulin are introduced into each experimental well, while 50 microliters of the test medium free of amphiregulin are introduced into each control well and are incubated at 37 ° C. for 2 at 3 days; (c) 100 microliters of a solution comprising 125]; _ i0a0_2, -125_x_ (j4SOXyUri (jine (1-UdR) are introduced into each well and incubated at 37 ° C for 4 to 6 hours; (d) the medium is eliminated and each well is washed once with PBS; (e) 200 microliters of methanol are introduced into each well and incubated for 10 minutes at room temperature, then eliminated; (f) 200 microliters of a solution of IM sodium hydroxide are introduced into each well and incubated at 37 ° C for 30 minutes, and (g) the IM sodium hydroxide solution is collected and counted in a gamma counter, in which the presence of radioactivity shots is a measurement of the incorporation of • * - ^ I-IUdR and of cell proliferation, while the absence of the strokes is indicative of the inhibition of cell proliferation. 27. Amphiregulin, a bifunctional growth-regulating factor, obtained by the method comprising; (a) culturing MCF-7 cells in the presence of 12-0-tetradecanoylphorbol-13 13-acetate (TPA); (b) collecting the conditioned medium from MCF-7 cells, and (c) isolating the amphiregulin from the conditioned medium. 28, - Nucleotide sequence coding for the precursor of amphiregulin comprising the nucleotide sequence coding, in substance as illustrated: in FIG. 16, approximately from the remainder of nucleotide No. 1 up to approximately the remainder of nucleotide No. 1243. 29. -Ampergulin precursor comprising the amino acid sequence, in substance as illustrated in FIG. 16, approximately of the remainder of amino acid no. 1 up to approximately the remainder of amino acid no. 252. 30, - Process for producing amphiregulin, characterized in that it comprises: 125 (a) culture of an Escherichia coli cell containing a nucleotide sequence coding for amphiregulin under the control of a second nucleotide sequence which regulates the expression of the gene so that a peptide or a protein having the activity of amphiregulin is produced by the Escherichia coli cell, and (b) isolating amphiregulin from the culture. 31. The method of claim 30, wherein the nucleotide sequence coding for amphiregulin comprises the nucleotide sequence in substance as illustrated in FIG. 16, from nucleotide No. 1 to nucleotide No. 1243. 32. The method of claim 30, wherein the second nucleotide sequence which regulates expression of the gene comprises a Tac promoter. 33. Process for producing amphiregulin, characterized in that it comprises: (a) the culture of the strain JM 109 of Escherichia coli transformed by the plasmid pTacAPAR1 as deposited at NRRL with the access number, and (b ) isolation of amphiregulin from the culture, 34. Process for producing amphiregulin, characterized in that it comprises: r. (a) the culture of the JM 109 strain of Escherichia coli transformed by the plasmid pTacAPHILE as deposited at NRRL with the access number, and (b) the isolation of amphiregulin from the culture. 35. Genetically engineered cell, characterized in that it contains a nucleotide sequence coding for amphiregulin under the control of a second nucleotide sequence which regulates the expression of the gene so that the cell produces amphiregulin . 36, - cell engineered by genetic engineering according to claim 35, characterized in that the nucleotide sequence coding for amphiregulin comprises the nucleotide sequence in substance as illustrated in FIG. 16, from nucleotide No. 528 to nucleotide No. 761. 37, - Cell manipulated by genetic engineering according to any one of claims 35 and 36, characterized in that it is a cell of Escherichia coli. 38, - Cell manipulated by genetic engineering according to any one of claims 35 and 36, characterized in that the second nucleotide sequence which controls the expression of the gene comprises a Tac promoter. 39, - Plasmid pAR1, as contained in an Escherichia coli cell line carrying the plasmid deposited at NRRL under the access number 40, - Plasmid pARH12, as contained in an Escherichia coli cell line carrying the deposited plasmid to NRRL under access number 41, - Plasmid pARH6, as contained in an Escherichia coli cell line carrying the plasmid deposited at NRRL under access number 42, - Plasmid pTacAPARl, as contained in a cell line of Escherichia coli carrying the plasmid deposited at NRRL under access number 43, - Plasmid pTacAPHILE, as contained in a cell line of Escherichia coli carrying the plasmid deposited at NRRL under access number 44, - Antibody, characterized in that it recognizes an epitope of amphiregulin. 45 - An antibody according to claim 44, 127 / characterized in that the antibody comprises the amino acid sequence s DTYSGKREPFSGDHSADGFE. V r 46. Antibody according to claim 44, characterized in that the antibody comprises the amino acid sequence s SSSEPSSGADYDYSEEYDNE. 47. Antibody according to claim 44, characterized in that the antibody comprises the amino acid sequence: VKPPQNKTESENTSDKPKRKKKG. 48. Antibody according to claim 44, characterized in that the antibody comprises the amino acid sequence: EAVTCKCQQEYFGERCGEK. 49. Antibody according to claim 44, characterized in that the antibody comprises the amino acid sequence s VQLRRQYVRKYEGEAEERKK. 50. Genomic nucleotide sequence, characterized in that it codes for the amphiregulin gene containing the nucleotide sequence, in substance as illustrated in FIG. 17.