IE19990007A1 - Cytotoxic Lymphocyte Maturation Factor - Google Patents

Abstract

ABSTRACT The present invention relates to a novel subunit of the cytokine protein called Cytotoxic Lymphocyte Maturation Factor (CLMF) which is produced and synthesized by a human B lymphoblastoid cell line. CLMF synergistically induces in the presence of low concentrations of IL-2 the cytolytic activity of Lymphokine Activated Killer (LAK) cells. CLMF is also capable of stimulating T-cell growth. The present invention also relates to cloned genes coding for the novel proteins and derivatives thereof, to recombinant vectors comprising a polynucleotide encoding said proteins, to microorganisms transformed with the said recombinant vectors, to antibodies directed to the said proteins as well as to processes for the preparation of the said proteins, vectors and antibodies.

Description

The present invention relates to the field of cytokines, in particular to those cytokines which synergize with interleukin-2 (IL-2) to activate cytotoxic lymphocytes such as the cytokine Cytotoxic Lymphocyte Maturation Factor (Cnfi‘). The present invention also relates to monoclonal antibodies directed to CUE1 'Cytokine' is one term for a group of protein cell regulators. variously called lymphokines, monokines, interleukins and interferons, which are produced by a wide variety of cells in the body. These cytokines play an important role in many physiological responses, are involved in the pathophysiology of a range of diseases, and have therapeutic potential. They are a heterogeneous group of proteins having the following characteristics in common.

They are low molecular weight (580 kDa) secreted proteins which are often glycosylated; they are involved in immunity and inflammation where they regulate the amplitude and duration of a response: and are usually produced transiently and locally. acting in a paracrine or autocrine, rather than endocrine manner. Cytokines are extremely potent, generally acting at picomolar concentrations: and interact with high affinity cell surface receptors specific for each cytokine or cytokine group. Their cell surface binding ultimately leads to a change in the pattern of cellular RNA and protein synthesis, and to altered cell behavior. Individual cytokines have multiple overlapping cell regulatory actions. ‘it is concomitantly exposed. macaw. 84903 The response of a cell to a given cytokine is dependent upon the local concentration of the cytokine, upon the cell type it is acting on and upon other cell regulators to which The overlapping regulatory actions of these structurally unrelated proteins which bind to different cell surface receptors is at least partially accounted for by the induction of common proteins which can have common response elements in their DNA. interact in a network by: Cytokines inducing each other: second, transmodulating cytokine cell surface receptors and first. third. by synergistic, additive or antagonistic interactions on cell function. [Immunology Today lg: 299 (1989)].

The potential utility of cytokines in the treatment of neoplasia and as immunoenhancing agents has recently been demonstrated in studies using human recombinant interleukin-2 (rIL—2). Natural inter1eukin—2 (IL—2) is a lymphokine which is produced and secreted by T—1ymphocytes.

This glycoprotein molecule is intimately involved in the induction of virtually all immune responses in which T—cells play a role. B cell responses in vitro are also enhanced by the presence of IL-2. IL-2 has also been implicated as a differentiation inducing factor in the control of B and T lymphocyte responses. lymphocytes which are activated by rIL—2 in vivo [J. lmmunol. l39:285-294 (1987)).

The anti-tumor effects of in ameliorating chemotherapy—induced immunosuppression [lmmunol. Lett. _g:3o7—314 (l985)].

However, the clinical use of rIL—2 has been complicated 'by the serious side effects which it may cause [N. Engl. J.

Med. 313:14B5—l492 (1985) and N. Engl. J. Med. 3l6:889—897 combination with rIL—2 in vivo.

Kobayashi et al. (J. Exp. Med. (1989) 170, 827-845) relates to the identification and purification of natural killer cell stimulatory factor (NKSF), a cytokine with multiple biologic effects on human lymphocytes.

Thus, the present invention provides a 35 kDa subunit of a cytokine protein called Cytotoxic Lymphocyte Maturation Factor (CLMF) which is produced and synthesized by cells capable of secreting CLMF. Examples for such cells are mammalian cells particularly human lymphoblastoid cells. In the presence of low concentrations of IL-2 CLMF synergistically induces the cytolytic activity of Lymphokine Activated Killer (LAK) cells. CLMF is also capable of stimulating T-cell growth.

CLMF can be isolated in a substantially pure form by the following steps: Egg 9;, lg; M a) stimulating B lymphoblastoid cells such as NC-37 cells to produce and secrete cytokines into a supernatant liquid; b) collecting the supernatant liquid produced by the stimulated cells; c) separating the supernatant liquid into protein fractions: d) testing each protein fraction for the presence of CLM: e) retaining the protein fractions which are able to stimulate T-cell growth. said fractions containing an active protein which is responsible for the T—cell stimulating activity of the protein fractions: f) isolating said active protein into a substantially pure form, said protein being Cytolytic Lymphocyte Maturation Factor (CLMF).

The CLMF protein obtained in this way is free from other cytokine proteins. The natural CLMF protein is a 75 kilodalton (kDa) heterodimer comprised of two polypeptide subunits, a 40 kDa subunit and a 35 kDa subunit which are bonded together via one or more disulfide bonds. The present invention also provides the nucleotide sequence.of the 35 kDa subunit of the CLMF gene and the amino acid sequence of the 35 kDa subunit of the CLMF protein encoded by the said gene. The present invention relates to a protein which exhibits CLMF activity and contains a biologically active portion of the amino acid sequence of CLMF or which contains an amino acid 3,2 2 sequence of CLMF as well as other amino acids or proteins containing analogous sequences to CLMF or its biologically active fragments which proteins exhibit CLMF activity.

The above process steps c) to f) may be used to purify CLMF from any liquid or fluid which contains CLMF together with other proteins. The present invention relates also to protein fractions having CLMF activity and being capable of stimulating T—cell growth, to a substantially purified active CLMF protein, obtained by the above described process, to the isolated cloned gene encoding the 35 kDa subunit, to vectors containing this gene to host cells transformed with the vector containing the said gene and to CLMF proteins prepared in such a transformed host cell. Furthermore the present invention relates to isolated polyclonal or monoclonal antibodies capable of binding to CLMF.

Monoclonal antibodies prepared against a partially purified preparation of CLMF have been identified and characterized by 1: I251-labelled CLMF, 2: immunodepletion of CLMF bioactivity, 3: blotting of CLMF, 4: its cellular receptor and 5: immunoprecipitation of western 1251-CLMF binding to neutralization of CLMF Twenty hybridomas inhibition of bioactivity. secreting anti—CLMF antibodies were identified. The 125 .

I-labelled CLMF bioactivity as assessed in antibodies were found to immunoprecipitate CLMF and to immunodeplete the T—cell proliferation and LAK cell induction assays. Western blot analysis showed that each antibody binds to the 70 kDa heterodimer and to one of the subunits. Each of the above—mentioned 20 anti—CLMF monoclonal antibodies were specific for CLMF and in particular for the 40 kDa subunit of CLMF. A CLMF receptor binding assay has been developed to evaluate the ability of individual antibodies to inhibit CLMF binding to its cellular receptor.

The assay measures the binding of 125 I-labelled CLMF to FHA activated PBL blast cells in the presence and absence of each antibody. of the 20 antibodies tested, 12 antibodies were found to inhibit greater than 60% of the l—1abe1led CLMF binding to the blast cells. inhibitory antibodies, viz. 7B2 and 4A1, Two neutralize CLMF bioactivity while one non-inhibitory antibody, SE3, neutralize CLMF bioactivity. These data confirm that antibodies which block 125I—labelled CLMF binding to its cellular receptor will neutralize CLM bioactivity as does not assessed by the T-cell proliferation and LAK cell induction assays. The ability of the antibodies specific for the 40 kDa subunit of CLM to neutralize CLMF bioactivity indicates that determinants on the 40 kDa subunit are necessary for binding to the CLM cellular receptor.

The monoclonal anti—CLMF antibodies provide powerful analytical, diagnostic and therapeutic reagents for the immunoaffinity purification of natural and recombinant human CLMF, the development of human CLMF immunoassays, the identification of the active site of the 40 kDa subunit of CLMF and may be used in therapeutic treatments of patients which require selective immunosuppression of cytotoxic T cells, transplantation. such as in Monoclonal antibodies which recognize different epitopes on human CLMF can be used as reagents in a sensitive two-site immunoassay to measure levels of CLMF in biological fluids, cell culture supernatants and human cell extracts.

The monoclonal antibodies against CLMF exhibit a number of utilities including but not limited to: l. Utilizing the monoclonal antibodies as affinity reagents for the purification of natural and recombinant human CLMF; . Utilizing the monoclonal antibodies as reagents to configure enzyme—immunoassays and radioimmunoassays to measure natural and recombinant CLMF in biological fluids, cell culture supernatants, cell extracts and on plasma membranes of human cells and as reagents for a drug screening assay; . Utilizing the monoclonal antibodies as reagents to construct sensitive two—site immunoassays to measure CLMF in biological fluids, cell culture supernatants and human cell €Xtl'.'BCtS2 . Utilizing the monoclonal antibodies as reagents to identify determinants of the 40 kDa subunit which participate in binding to the 35 kDa subunit and which participate in binding to the CLMF cellular receptor; . Utilizing the intact IgG molecules, the Fab fragments or the humanized IgG molecules of the inhibitory monoclonal antibodies as therapeutic drugs for the selective blockade of proliferation and activation of cytotoxic T cells. such as in transplantation.

BRIEF DESCRIPTION OF THE DRAWINGS Figure 1 is a plot of a supernatant solution obtained from cultured NC37 lymphoblastoid cells applied to a Nu—Gel P—SP column showing the protein fraction containing TGF activity being eluted with a salt gradient.

Figure 2 is a plot of the material containing TGF activity obtained from the separation shown in Figure l as it was being eluted with a salt gradient through a Blue-B-Agarose Column.

Figure 3 shows the plot of the material containing TGF activity obtained from the separation shown in Figure 2 as it was being eluted with a Nacl gradient through a Mono Q column.

Figure 4 shows a SDS-polyacrylamide gel electrophoresis (SDS—PAGE) analysis of the fractions 30 to 45. 48 and SO obtained from the step illustrated in Figure 3. The numbers on the left side. 44 and 68, refer to the apparent molecular weight of standard proteins of 44 and 68 kDa in i.e. lane 8.

Figure 5 shows the elution profile through a vydac Diphenyl column of fraction 38 from the Mono Q Chromatography separation (reversed-phase HPLC) shown in Figure 3.

Figure 6 shows SDS—PAGE analysis of protein purity of the protein fractions 85-90 recovered from the separation process depicted in Figure 5.

Figure 7 shows a SDS—PAGE analysis of fractions 87 and 88 from the reversed-phase HPLC separation under non—reducing (lane A: without B—mercaptoethanol) and reducing (lane B; in the presence of B—mercaptoethanol) conditions showing the 75,000 molecular weight CLMF The remaining lanes in the gel shown in this Figure contain separated into two subunits of 40 kDa and 35 kDa. standard proteins comprising the 44 and 68 kDa marker protein.

Figure 8 shows the elution pattern of the proteins from the supernatant solution from NC—37 cells applied to a Nu- Gel P-SP column and eluted with a salt gradient. :. so ea Figure 9 is a B1ue—B—Agarose column salt gradient elution profile of the active fractions obtained from the Nu-Gel P—SP column elution shown in Figure 8.

Figure 10 is a Mono-Q column salt gradient elution profile of the active fractions obtained from the elution shown in Figure 9.

Figure 11 is the elution pattern through a Vydac Diphenyl column of active fractions 39 and 40 obtained from the Mono Q Chromatography shown in Figure 10.

Figure 12 shows a SDS—PAGE analysis under reducing conditions of the active fractions obtained from the separation process shown in Figure 11.

Figure 13 is a schematic diagram depicting the separation of the 40 kDa subunit from the 35 kDa subunit of the CLMF cytokine.

Figure 14 is a schematic diagram depicting the determination of the amino acid composition, the N—terminal sequencing. the proteolytic digestion and the complete sequencing of the 40 kDa subunit of the CLMF cytokine.

Figure 15 shows a separation of the tryptic peptides of the digested 40 kDa subunit of the CLMF cytokine.

Figure 16 shows a separation of the proteolytic peptides of the Staphylococcus aureus V8 protease digested 40 kDa subunit CLMF.

Figure 17 is a chart which summarizes the information on the protein structure obtained from the analysis of the proteolytic peptides of the 40 kDa subunit of CLMF. The following abbreviations and symbols are used: lEssn@%Z _ 10 _ N—t — N—termina1 sequencing on intact protein Tr - tryptic peptides from map HP2383 numbered by "fraction number V8 — V8 protease peptides from map HP24l2 numbered by fraction number - indicates probable glycosylation site; boxes indicate potential sites Figure 1B shows the SDS—PAGE analysis of Fraction 39 from the Mono Q FPLC elution profile shown in Figure 3. Lane A: standardproteins without B—mercaptoethano1; lane B: Fraction 39 without B—mercaptoethano1; lane C: Fraction 39 with B-mercaptoethanol: lane D: Standard proteins with B—mercaptoethanol.

Figure 19 relates to the purification of the 35 kDa subunit by reversed-phase HPLC and depicts the elution pattern through a Vydac C-18 column of fraction 39 of the Mono Q chromatography which was reduced in 5% B—mercapto- ethanol.

Figure 20 shows a SDS-PAGE gel analysis under non-reducing conditions of the fractions which were fluorescamine positive from the Vydac C-18 column elution profile shown in Figure 19. S: = protein—standard; F:‘= flow—through: numbers refer to the fraction number.

Figure 21 depicts the elution pattern of a tryptic digest of fractions 36 and 37 of the Mono Q Chromatography through a YMC ODS column.

Figure 22 shows the stained PVDF membrane with the smeared bands comprising the CNBr cleaved CLMF before (Fig. 22B) and after (Fig. 22A) excising the regions of about 29, 25, 14, 12, and 9 kDa. respectively. The regiones contain the CNBr fragments having the following sequences: !E9$u@@7 _ 11 _ I (P?)—P—K—N-L—Q—L-K—P-L—K—N—?—V-(Q?)- (New sequence from 40 kDa protein) ?-Q—K—A-(R?)—Q—T—L—E-F—Y—P—?—T~ (New sequence starting at residue no. 30 of 35 kDa protein) III V—V—L—T—?—D-T—P—E-E—D-G-I—T— (starts at residue no. 24 of 40 kDa protein) IV V—D-A-V—(H?)—K—L—K—Y—E—?—Y—T—?—?—F-F-I~ (Starts at residue no. 190 of 40 kDa protein) note: it is assumed or known that the above sequences are preceeded by a Met residue.

Figure 23 shows a reverse—phase HPLC separation of the peptide fragments obtained by cleaving CLMF with CNBr.

Figure 24 shows an SDS—PAGE of pure CLMF and "free" unassociated 40 kDa subunit of CLMF purified by affinity chromatography using the monoclonal antibody 7B2 covalently attached to an agarose resin. Lane A: molecular weight marker proteins; lane B: starting material; lane C: flow- through: lane D: acid eluate: lane E: potassium thiocyanate eluate. ‘ Figure 25 a, b, c and d show the DNA sequence and the deduced amino acid sequence of the 40 kDa subunit of human CLMF. ' Figure 26 a, b and c show the CDNA sequence and the deduced amino acid sequence of the 35 kDa subunit of CLMF Figure 27 depicts the inhibition of CLMF bioactivity by serum from rats immunized with CLMF and from non-immunized rats (control). _ 12 _ Figure 28 shows a SDS-PAGE analysis of immunoprecipitates of I—CLMF by monoclonal antibodie 4A1 (lane 1). 4Dl (lane 2), 8E3 (lane 3) and 9C8 (lane by a control antibody (lane 5), by immune rat serum (la and 8) and by normal rat serum (lanes 7 and 9). On the side the molecular weight in kDa is indicated.

Figure 29 shows the immunodepletion of CLMF bioactivity s 4). nes 6 left (TGF activity) by monoclonal anti—CLMF antibodies (a-CLMF).

Figure 30 shows the immunodepletion of CLMF bioactivity (LAK induction activity) by monoclonal anti—CLMF antibo (a—CLMF).

Figure 31 shows a Western blot analysis of the reactivity of the monoclonal antibodies (mAbs) 7B2, 4A1 8E3, 6A3, 9F5 and 2A3 and of rat polyclonal anti—CLMF antibodies (RS1) with the CLMF 75 kDa heterodimer. NBS: normal rat serum.

Figure 32 shows a Western blot analysis of the reactivity of monoclonal and rat polyclonal anti—CLMF antibodies with the CLMF 40 kDa subunit. In lanes 1 to the following mAbs were used: 4A1, 4D1. 7B2, 7A1, 2A3, 8B4, 8A2, 8E3, 1B8, 4A6, 6A2, 8C4. 9F5. 6A3, 9C8, 8A1 a E7, respectively. In lane 19 a control antibody, in.l a fusion rat serum and in lane 21 a normal rat serum was used. l25I—CLMF to peripheral blood lymphocyte (PBL) Figure 33 FHA-activated shows the binding of lymphoblasts.

Figure 34 PHA—activated PBL blast cells by rat anti-CLMf serum.

SI~CLMF binding to the cells in the presence of the indicated data are expressed as amount (% bound) of dies lCl. nd . . . . 125 . . shows the lnhlbltlon of I—CLMF binding to [59 9h ©i..'Z - 13 _ concentrations of serum when compared to the total specific binding in the absence of serum.

Figure 35 shows the inhibition of the binding of l2SI—CLMF to PHA-activated PBL blast cells by monoclonal antibody supernatants. The data are expressed as % inhibition of the binding of 1251-CLM to the cells in the presence of a 1:1 dilution of supernatant when compared to the total specific binding in the absence of antibody supernatant.

Figure 36 shows the inhibition of the binding of 5I—CLMT to PHA—activated PBL blast cells by various concentrations of purified monoclonal antibodies. The data are expressed as the amount (% cpm bound) of l25I—CLMF bound to the cells in the presence of the indicated concentrations of antibody when compared to the total specific binding in the absence of antibody.

Figure 37 shows a Western blot analysis of the reactivity of a rabbit polyclonal anti—CLMF antibody with the 75 kDa CLMF (nonreduced) and with the 35 kDa CLMF subunit (reduced). The antibody was prepared against a synthetic peptide fragment of the 35 kDa CLMF subunit. Lanes 1 to 5 are without B—mercaptoethanol; lanes 6 to 10 with B—mercaptoethanol.

F‘ D! 23 I'D I-’ O\OCD\lO\U"ol>l.-J|\3l4 IE99n@®7 CLMF 3 ul CLMF 6 ul CLMF Blank Blank ul 1 ul 3 ul 6 ul prestained molecular weight standards CLMF CLMF CLMF ul prestained molecular weight standards om so To The CLMF biological activity of all of the proteins of the present invention including the fragments and analogues may be determined by using a standard T—cel1 growth factor assay.

In accordance with the present invention, natural CLMF is obtained in pure form. The amino acid sequences of the kDa subunit and the 40 kDa subunit of the CLMF protein is depicted in Figures 25 and 26.

Thus, the present invention relates to a protein having Cytotoxic Lymphocyte Maturation Factor (CLMF) activity in a substantially pure form, such as the CLMF protein per se, or to a 35 kDa subunit of the said protein which exhibits CLMF activity if combined with the 40 kDa subunit and comprises at least a part of the amino acid sequence of the natural form of CLMF.

The present invention also relates to cloned genes coding for 35 kDa subunit of CLMF and to isolated polynucleotides encoding a subunit as defined above, which polynucleotide contains a sequence corresponding to the CDNA encoding 35 kDa subunit of CLMF, to recombinant vectors comprising a polynucleotide encoding a 35 kDa subunit of the CLMF protein, to microorganisms transformed with the said recombinant vectors, to antibodies directed to the said subunits as well as to processes for the preparation of the said subunits, govectors and antibodies. lE99®@@? _ 15 - The practice of the present invention will employ, unless otherwise indicated. conventional techniques of molecular biology, microbiology, recombinant DNA and immunology, which are within the skills of an artisan in the field. Such techniques are explained fully in the literature. See e.g., Maniatis, Fitsch & Sambrook.

MOLECULAR CLONING; A LABORATORY MANUAL (1982): DNA CLONING, VOLUMES I AND II (D.N Glover ed., 1985); OLIGONUCLEOTIDE SYNTHESIS (M.J. Gait ed., 1984): NUCLEIC ACID HYBRIDIZATION (B.D. Hames & S.J. Higgins eds., 1984); TRANSCRIPTION AND TRANSLATION (B.D. Harnes & S.J. Higgins eds., l9B4); ANIMAL CELL CULTURE (R.I. Freshney ed., 1986); IMMOBILIZED CELLS AND ENZYMES (IRL Press, 1986): B. Perbal, A PRACTICAL GUIDE TO MOLECULAR CLONING (1984); the series, METHODS IN ENZYMOLOGY (Academic Press, Inc.): GENE TRANSFER VECTORS FOR MAMMALIAN CELLS (J.H. Miller and M.P. Calos eds.. 1987. Cold Spring Harbor Laboratory). Methods in Enzymology Vol. 1 559 W and Vol. 155 (Wu and Grossman, and Wu, eds.. respectively); IMMUNOCHEMICAL METHODS IN CELL AND MOLECULAR BIOLOGY (Mayer and Walker. eds.. 1987, Academic Press. London), Scopes, PROTEIN PURIFICATION: PRINCIPLES AND PRACTICE, second Edition (1987. Springer—Verlag, N.Y.), and HANDBOOK OF EXPERIMENTAL IMUNOLOGY, VOLUMES I—IV (D.M. Weir and C.C.

Blackwell eds., 1986).

The DNA sequences and DNA molecules of the present invention may be expressed using a wide variety of host/vector combinations. For example, useful vectors may consist of segments of chromosomal, non—chromosomal and synthetic DNA sequences{ Examples of such vectors are viral vectors, such as the various known derivatives of SV40, bacterial vectors, such as plasmids from E. coli including pCR1, pBR322, pMB9 and RP4. phage DNAs. such as the numerous derivatives of phagex. M13 and other filamentous single-stranded DNA phages, as well as vectors useful in yeasts. such as the 2n plasmid, vectors useful in eukaryotic cells more preferably vectors useful in animal cells, such as those containing SV40. adenovirus and/or retrovirus derived DNA sequences. Useful vectors may be also derived from combinations of plasmids and phage DNA’s, such as plasmids which have been modified to comprise phage DNA or other derivatives thereof.

Expression vectors which may be used for the preparation of recombinant 35 kDa CLMF subunits are characterized by comprising at least one expression control sequence which is operatively linked to the 35 kDa CLMF subunit DNA sequence inserted in the vector in order to control and to regulate the expression of the cloned 35 kDa CLMF subunit DNA sequence. Examples of useful expression control sequences are the lac system, tac system. the the trc system, major operator and promoter the trp system, regions of phage X, the control region of fd coat protein, e.g., the promoters of yeast acid the glycolytic promoters of yeast, 3-phosphoglycerate kinase, the promoter for . _ 19 _ phosphatase, e.g., Pho 5, u—mating factors, the promoters of the yeast and promoters derived from polyoma virus, retrovirus, adenovirus, and simian virus, e.g., the early and late promoters or SV40, and other sequences known to control the expression of genes of prokaryotic or eukaryotic cells and of their viruses as well as combinations of the said promoter/operator sequences.

Among such useful expression vectors are known vectors that enable the expression of the cloned CLMF—related DNA sequences in eukaryotic hosts, cells [e.g.. P. J. such as in animal and human Southern and P. Berg. J. Mol. Appl.

Genet. ;: 327-41 (1982): S. Subramani et al.. Mol. Cell.

Biol. lz 854-64 (1981); R. J. Kaufmann and P. A. Sharp, Mol.

Cell. Biol. igg: 601-64 (1982): S. I. Scahill et al., "Expression and Characterization of The Product of A Human Immune Interferon DNA Gene in Chinese Hamster Ovary Cells“, Proc. Natl. Acad. U.S.A. QQ: 4654~S9 (1983): G. Urlaub and L. A. Chasin, Proc. Natl. Acad. Sci. USA 11: 4216-20 (l989)].

Sci.

Furthermore, within each specific expression vector, various sites may be selected for insertion of the CLMF—related DNA sequences of the present invention. These sites are usually designated by the restriction endonuclease which cut them. They are well recognized by those of skill in the art. It is, of course to be understood that an expression vector useful in this invention need not have a restriction endonuclease site for insertion of the chosen DNA fragment. Instead, the vector could be joined to the fragment by alternative means. The site chosen in the expression vector for the insertion of a selected DNA fragment and the operative linking of the DNA fragment to an expression control sequence is determined by a variety of factors, such as the number of sites susceptible to a particular restriction enzyme, the location of start and stop codons relative to the vector sequence and the desired IEQW »WZ _ 19 _ selection method for the host transformed with the recombinant vector. The choice of a vector and an insertion site for a DNA sequence is determined by a balance of these factors, not all selections being equally effective for a given case.

The host cell used for the expression of the CLMF—related DNA sequence known hosts. Examples for eukaryotic cells. A large may be selected from a variety of such hosts are prokaryotic or number of such hosts are available from various depositories such as the American Type Culture Collection (ATCC) or the Deutsche Sammlung fur Mikroorganismen (DSM). Examples for prokaryotic cellular hosts are bacterial strains such as E.co1i, B.subtilis and others. Preferred hosts are mammalian cells such as the SV4O transformed African Green monkey kidney cell line COS.

Not all host/expression vector combinations function with equal efficiency in expressing a given DNA sequence.

However, a particular selection of a host/expression vector combination may be made by those of skill in the art after due consideration of the principles set forth herein without departing from the scope of this invention. For example, the selection should be based onia balancing of a number of factors, These include, for example, compatibility of the host and vector. susceptibility of the protein to proteolytic degradation by host cell enzymes, possible contamination of the protein to be expressed by host cell proteins difficult to remove during purification, toxicity of the proteins encoded by the DNA sequence to the host, ease of recovery of the desired protein, expression characteristics of the DNA sequence and the expression control sequence operatively linked to them, biosafety. costs and the folding, form or any other necessary post—expression modifications of the desired protein.

The host organisms which contain the expression vector comprising the 35 kDa CLMF subunit DNA are usually grown up under conditions which are optimal for the growth of the host organism. Towards the end of the exponential growth. when the increase in the number of cells per unit time decreases, the expression of the CLMF subunit is induced, i.e. the DNA coding for the subunit is transcribed and the transcribed mRNA is translated. The induction can be effected by adding an inducer or a derepressor to the growth medium or by altering a physical parameter, e.g. by a temperature change.

The CLMF subunit produced in the host organism can be secreted by the cell by special transport mechanisms or can be isolated by breaking open the cell. The cell can be broken open by mechanical means [Charm et al.. Meth. Enzmol. ggz 476-556 (1971)], by enzymatic treatment (e.g. lysozyme treatment) or by chemical means (e.g. detergent treatment, urea or guanidine~HC1 treatment, etc.) or by a combination thereof.

In eukaryotes, polypeptides which are secreted from the cell are synthesized in the form of a precursor molecule.

The mature polypeptide results by cleaving off the so—called signal peptide. As prokaryotic host organisms are not capable of cleaving eukaryotic signal peptides from precursor molecules, eukaryotic polypeptides must be expressed directly in their mature form in prokaryotic host organisms. The translation start signal AUG, which corresponds to the codon ATG on the level of the DNA. causes that all polypeptides are synthesized in a proharyotic host organism with a methionine residue at the N-terminus. In certain cases, depending on the expression system used and possibly depending on the polypeptide to be expressed this N—terminal methionine residue is cleaved off.

The 35 kDa CLMF subunit produced by fermentation of the prokaryotic and eukaryotic hosts transformed with the DNA sequences of this lisenoll _2]__ invention can then be purified to essential homogeneity by known methods such as. for example. by centrifugation at different velocities, by precipitation with ammonium sulphate, by dialysis (at normal pressure or at reduced pressure), by preparative isoelectric focusing, by preparative gel electrophoresis or by various chromatographic methods such as gel filtration, high performance liquid chromatography (HPLC), ion exchange chromatography, reverse phase chromatography and affinity chromatography (e.g. on Sepharose” Blue CL-6B or on carrier-bound monoclonal antibodies directed against CLMF).

The purified CLMF subunit of the present invention can be employed for the preparation of LAK cell and T cell activator and antitumor compositions and in methods for stimulating LAK cell. T—cells or Natural Killer Cells.

The 35 kDa CLMF subunit of the present invention can also be The information from this analysis may be used to predict and produce fragments or peptides, including synthetic peptides, having the activity of CLMF. analyzed to determine the active sites for CLMF activity.

Among the known techniques for determining such active sites are X-ray crystallography. nuclear magnetic resonance, circular dichroism, UV spectroscopy and site specific mutagenesis. Accordingly. the fragments obtained in this way may be employed in methods for stimulating T—ce1ls or LAK cells.

The CLMF subunits prepared in accordance with this invention or pharmaceutical compositions comprising the 35 kDa CLMF subunit may be administered to warm blooded mammals for the clinical uses indicated above. The administration may be by any conventional modes of administration of agents which exhibit antitumor activity auch as by intralesional or parenteral application either intravenously, subcutaneously or intramuscularly. obviously, the required dosage will vary ‘Egg Q‘% Q @j? _ 22 _ with the particular condition being treated, the severity of the condition, the duration of the treatment and the method for administration. A suitable dosage form for pharmaceuti- cal use may be obtained from sterile filtered, lyophilized protein reconstituted prior to use in a conventional manner.

It is also within the skill of the artisan in the field to prepare pharmaceutical compositions comprising 35 kDa CLMF subunit of the present invention by mixing the said CLMF subunit with compatible pharmaceutically acceptable carrier materials such as buffers. stabilizers, bacteriostats and other excipients and additives conventionally employed in pharmaceutical parenteral dosage forms. The present invention also relates to such pharmaceutical compositions.

The preferred form of administration depends on the intended mode of administration and therapeutic application. The pharmaceutical compositions comprising a CLMF protein or peptide derivative of the present invention also will preferably include conventional pharmaceutically acceptable carriers and may include other medicinal agents (e.g. interleukin-2), carriers, adjuvants, excipients. etc., e.g., human serum albumin or plasma preparations.

Preferably, the compositions of the invention are in the form of a unit dose and will usually be administered one or more times a day. The unit dose is preferably packed in 1 ml vials containing an effective amount of the 35 kDa CLMF subunit and if desired of interleukin-2 in lyophilized form. The vials containing the CLMF subunit and if desired the interleukin-2 are preferably packed in a container together with written instructions describing the correct use of the pharmaceutical composition. The present invention relates also to such a unit dose packed in a container, preferably together with a separate unit dose of interleukin—2, most preferably together with the appropriate instructions. Furthermore the present invention relates to a process for the preparation of the said unit dose. “ . ‘.2- gieeefi _ 23 _ In order that our invention herein described may be more fully understood, the following examples are set forth. should be understood that these examples are for illustrative purposes only and should not be construed as limiting this invention in any way to the specific embodiments recited therein. It has to be noted that the specific product names and suppliers mentioned below are not meant to be mandatory. The person skilled in the art is in a position to select alternative products from other suppliers.

EXAMPLE PURIFICATION AND CHARACTERIZATION OF CYTOTOXIC LYMPHOCYTE MATURATION FACTOR (CLMF) Production of Supernatant Liquid Containing CLMF.

Human NC-37 B lymphoblastoid cells (ATCC CCL 214.

American Type Culture Collection. Rockville, MD) were used for production of CLMF. These cells were maintained by serial passage in RPMI 1640 medium supplemented with 5% heat—inactivated (56°C, 30 min.) fetal bovine serum, 2 mM L-glutamine, 100 units/ml penicillin. and 100 ug/ml streptomycin (all cell culture media were from GIBCO Laboratories, Grand Island, NY).

Higher producer sublines of NC-37 cells were derived by limiting dilution cloning in liquid microcultures. Each well of three Costar 3596 microplates (Costar Co., Cambridge, MA) received 100 ul of a cell suspension containing five NC-37 cells/ml. The medium used for the cloning was a lzl mixture of fresh passage medium and filtered, conditioned medium from stock cultures of the parent NC-37 cells. One week and two weeks after culture initiation each of the microcultures was fed with 50 ul of the l:l mix of fresh and conditioned medium. Between 3 and weeks after culture initiation the contents of wells V99; 4:} containing clones of NC-37 cells were harvested and passed into larger cultures. when the number of cells in a given subline exceeded 1.4 X lO6. one million cells were stimulated to produce CLMF in 1 ml cultures containing 3 ng/ml phorbol 12—myristate l3—acetate (PMA) (Sigma Chemical Co., St. Louis, MO) and 100 ng/ml calcium ionophore A23187 (sigma). Supernatants were harvested from the cultures after 2 days, dialyzed against about 50 volumes of Dulbecco's phosphate buffered saline (Gibco) using e.g. SPECTROPOR® #1 tubing (Fisher Scientific) overnight with one change of buffer and then for 4 hours against 50 volumes of RPMI 1640 medium with 50 ug/ml of gentamicin (both from Gibco) and tested for CLMF by means of the T cell growth factor assay (see below).

Three sublines, NC—37.89, NC—37.98, and NC—37.l02, were identified which routinely produced CLME at titers 1 the titers produced by the parental NC—37 cell line. times Since cells from these three sublines produced CLMF at similar titers (3 800 units/ml), culture supernatants derived from the three sublines were pooled for use as starting material for the purification of CLMF.

Bulk production of CLMF was carried out in roller bottle cultures on a roller apparatus set at about 38 rpms (Wheaton Cell Production Roller Apparatus Model II, Wheaton Instruments. Millville, NJ). Cell suspensions were prepared containing 1-1.5 x 106 NC-37.69, NC-37.98 or NC—37.lO2 cells/ml in RPMI 1640 medium supplemented with 1% Nutridoma—SP (Boehringer Mannheim Biochemicals, IN), 100 units/ml ng/ml PMA and 20-25 Two hundred fifty to three hundred fifty ml aliquots of the cell suspensions were added Indianapolis. 2 mM L-glutamine, penicillin. 100 ug/ml streptomycin, ng/ml calcium ionophore A23l87. to Falcon 3027 tissue culture roller bottles (Becton Dickinson, Lincoln Park, NJ) which had been gassed with a mixture of 5% CO2, 95% air. The roller bottles were then "V-’-Yogngli V i aE990@“7 _ 25 _ capped tightly and incubated at 37°C with continuous rolling for three days. At the end of this time, supernatants were harvested. the culture EDTA and phenylmethylsulfonyl fluoride (both from Boehringer Mannheim) were added to the culture supernatants at final concentrations of 1 mM and 0.1 mM, respectively, to retard proteolytic degradation. supernatants were stored at 4°C.

Lympokine Activated Killer (LAK) Cell Induction (LCI) Assay.

Culture supernatants and chromatographic fractions were tested for their ability to synergize with r1L—2 to induce the generation of cytolytic LAK cells as follows. Human peripheral blood mononuclear cells (PBMC) were isolated by the following method. Blood from normal volunteer donors was drawn into syringes containing sufficient sterile preservative—free heparin (Sigma) to give a final The blood was diluted 1:1 with Hanks‘ balanced salt solution (HBSS) concentration of approximately 5 units/ml. without calcium or magnesium (GIBCO). The diluted blood was then layered over 15 ml aliquots of Ficoll/sodium diatrizoate solution (Lymphocyte Separation Medium, Organon Teknika Corp., Durham, NC) in 50 ml Falcon 2098 centrifuge tubes. The tubes were centrifuged for 30 minutes at room temperature at 500 x g. Following centrifugation, the cells floating on the Ficoll/sodium diatrizoate layer were collected and diluted by mixing with 3 2 volumes of HBSS without calcium or magnesium. The resulting cell suspension was then layered over 15 ml aliquots of 20% sucrose (Fisher) in RPMI 1640 medium with 1% human AB serum (Irvine Scientific, Santa Ana, CA) in Falcon 2098 centrifuge tubes.

The tubes were centrifuged for 10 minutes at room temperature at 500 x g, and the supernatant fluids were discarded. The cell pellets were resuspended in 5 ml of HESS without calcium or magnesium, repelleted by centrifugation. and finally resuspended in the appropriate QEQWW7 culture medium. Accessory cells were removed from the PBMC by treatment with 5 mM L—g1utamic acid dimethyl ester (Sigma) using the same conditions as described by Thiele et al. J. Immunol. ;;l:2282—2290 (1983) for accessory cell depletion by L—leucine methyl ester except that the glutamic acid ester was substituted for the leucine ester.

The accessory cell-depleted PBMC were further fractionated by centrifugation on a discontinuous Percoll density gradient (Pharmacia, Piscataway, NJ) as described by Wong et al.. Cell Immunol. l;l:39—54 (1988). cells recovered Mononuclear from the 38, 41, 45. and 58% Percoll layers used as a source of LAK cell precursors in cells recovered from the Percoll gradient were pooled and The were washed and the assay. suspended in tissue culture medium (TCM) composed of a 1:1 mixture of RPMI 1640 and Dulbecco's modified Eagle‘s medium, supplemented with 0.1 mM nonessential amino acids. 60 pg/ml arginine Hcl, 10 mM HEPES buffer, 2 mM L—glutamine. 100 units/ml penicillin, 100 x 1o'5 NJ), and 5% human AB serum (Irvine ug/ml streptomycin (all available from GIBCO), M 2—mercaptoethanol (Fisher Scientific, Fair l mg/ml dextrose (Fisher), Scientific, CA). These cells were incubated in ~we1l tissue culture plates (costar, Cambridge, MA) in cells/culture) to which 10-4 M hydrocortisone sodium succinate (Sigma) was added to Lawn, Santa Ana. l ml cultures (7.5 x 10 minimize endogenous cytokine production. some cultures also received human rIL—2 (supplied by Hoffmann-La Roche, Nutley.

Inc., NJ) at a final concentration of 5 units/ml and/or supernatants to be assayed for CLMF activity. All cultures were incubated for 3-4 days at 37°C in a humidified atmosphere of 5% CO2. 95% air.

At the end of this incubation. the contents of each culture were harvested, and the cells were pelleted by centrifugation and resuspended in 0.5 ml of fresh TCM. one tenth ml aliquots of these cell suspensions were mixed with _ 27 - lCr—labelled K562 or Raji cells (both cell lines may be obtained from the ATCC) and tested for .1 ml aliquots of release was calculated as [(g — g)/(100 — 3)] X 100, where III) is the percentage of Cr released from target cells were assayed in quadruplicate for lytic activity.

LAK Cell Induction Microassay. The microassay for measuring synergy between rIL—2 and CLMF-containing solutions in the induction of human LAK cells was similar to the LAK cell induction assay described above but with the following modifications. Human peripheral blood mononuclear cells which had been depleted of accessory cells and fractionated by Percoll gradient centrifugation as described above were added to the wells of Costar 3596 microplates (5 x 104 cells/well). some of the wells also received rIL-2 (S units/ml final concentration) and/or purified CLMF or immunodepleted CLMF—containing solutions. . -4 contained 10 All cultures M hydrocortisone sodium succinate (Sigma) and were brought to a total volume of 0.1 ml by addition of TCM with 5% human AB serum. 3 days at 37°C.

The cultures were incubated for after which 0.1 ml of 5lCr—labelled K562 cells (5 x 104 cells/ml in TCM with 5% human AB serum) were added to each well. overnight at 37°C.

The cultures were then incubated Following this, the cultures were centrifuged for 5 minutes at 500 x g, and the supernatant solutions were harvested by use of a Skatron supernatant collection system (Skatron, Sterling, VA). The amount of Cr released into each supernatant solution was measured R99“ QQJ _ 23 - with a gamma counter (Packard, Downer's Grove, IL), 1 Cr release was calculated as described All samples were assayed in quadruplicate. and the % specific above. gytolytic T Lymphocyte (CTL) Generation Assay.

Methods used for generating and measuring the lytic activity of human CTL have been described in detail by Gately et al. in J. Immunol. 136: 1274-1282 (1986) and by Wong et al. in Cell. Immunol. 111: 39-54 (1988). Human peripheral blood mononuclear cells were isolated from the blood of normal volunteer donors. depleted of accessory cells by treatment with L—glutamic acid dimethyl ester. and fractioned by Percoll gradient centrifugation as described above.

High density lymphocytes recovered from the interface between the 45% and 58% Percoll layers were used as responder lymphocytes in mixed lymphocyte—tumor cultures (MLTC). CTL were generated in MLTC in 24-well tissue culture plates (costar #3424) by incubation of Percoll gradient—derived high density lymphocytes (7.5 x 105 culture) together with 1 x 105 uv—irradiated melanoma cells e.g. HT144 (obtainable from ATCC) or with 5 x 104 gamma—irradiated melanoma cells e.g. HT144 in TCM with 5% human AB serum (1.2 ml/culture). For uv—irradiation, HTl44 cells were suspended at a density of 1-1.5 x 106 cells/ml in Hanks’ balanced salt solution without phenol red (GIBCO) containing 1% human AB serum. One ml aliquots of the cell suspension were added to 35 x 10 mm plastic tissue culture dishes (Falcon #3001), (960 uw/cmz for 5 min) by use of a 254 nm uv light (model UVG—54 MINERALIGHT® lamp, Ultra—violet Products.

Inc., San Gabriel, CA).‘ For gamma irradiation. HT144 cells were suspended at a density of 1-5 x 106 cells/ml in TCM with 5% human AB serum and irradiated (10,000 rad) by use of a cesium source irradiator (model 143, J.L. Shepherd and Associates, CA). Uv— or gamma-irradiated HT144 were centrifuged and resuspended in TCM with 5% human and the cells were then irradiated San Fernando, IHWW -29..

AB serum at the desired cell density for addition to the MLTC. In addition to lymphocytes and melanoma cells, some MLTC received human rIL-2 and/or purified human CLMF at the concentrations indicated. Hydrocortisone sodium succinate (Sigma) was added to the MLTC at a final concentration of -4 M (cultures containing uv-irradiated melanoma cells) or 1o’5 M (cultures containing gamma-irradiated melanoma cells) to supress endogenous cytokine production (8. Gillis et al.. J. Immunol. 12;: 1624-1631 (1979)] and to reduce the generation of nonspecific LAK cells in the cultures [L.M.

Muul and M.K. Gately, J. Immunol. 112: 1202-1207 (1984)].

The cultures were incubated at 37°C in a humidified for 6 days. At the end of replicate cultures were pooled, atmosphere of 5% CO2 lymphocytes from in air this time. centrifuged. resuspended in 1.2 ml TCM containing 5% human AB serum, and tested for their ability to lyse HTl44 melanoma cells, and, as a specificity control, K562 erythroleukemia cells (obtainable from ATCC) in overnight 510: release assays.

Slcr sodium chromate as described by Gately et al. [JNCI gg: 1245-1254 (l982)]. Likewise. mediated lysis of Melanoma cells and K562 cells were labeled with measurement of lympocyte- SIC:-labeled melanoma cells was performed in a manner identical to that described by Gately et al. (ibid.) for quantitating lysis of glioma target cells. 5lCr-labeled K562 cells, 0.1 ml aliquots of lymphocyte suspensions were mixed with 25 ul aliquots of SlCr—labeled K562 (2 x 105 cells/ml in TCM with 5% human AB serum) in the wells of costar 3696 For assaying the lysis of "half—area“ microtest plates. After overnight incubation at °C, the plates were centrifuged for 5 min at 1400 x g, and 50 ul of culture medium was aspirated from each well. amount of The lcr in each sample was measured with a gamma counter (Packard), and the % specific calculated as described above.

Cr release was All assays were performed in quadruplicate, and values in the table (see below) represent lhsaafifi? the means 1 1 S.E.M. of replicate samples.

T cell growth factor (TGF) assay.

The ability of culture supernatants and chromatographic fractions to stimulate the proliferation of PHA—activated human T lymphoblasts was measured as follows. Human PBMC were isolated by centrifugation over discontinuous Ficoll and sucrose gradients as described above for the LCI assay.

The PBMC (S x 105 cells/ml) were cultured at 37°C in TCM containing 0.1% phytohemagglutinin—P (PHA~P) (Difco Laboratories, Detroit, MI). split lrl with fresh TCM, After 3 days, the cultures were and human rIL—2 was added to each culture to give a final concentration of 50 units/ml. The cultures were then incubated for an additional 1 to 2 days, at which time the cells were harvested, washed, resuspended in TCM at 4 x 105 and cells/ml. To this cell suspension was added heat-inactivated goat anti-human rIL—2 antiserum (final dilution: l/200) to block any potential ILinduced cell proliferation in the assay. This antiserum may be prepared using methods well—known in the MA. The antiserum used was shown to cause 50% neutralization of 2 art or may be obtained from Genzyme Co., Boston, units/ml rIL-2 at a serum dilution of l/20,000.

Fifty ul aliquots of the cell suspension containing anti-IL-2 antiserum were mixed with 50 ul aliquots of serial dilutions of culture supernatants or chromatographic fractions in the wells of Costa: 3596 microplates. The cultures were incubated for 1 day at 37°C in a humidified atmosphere of 5% CO2 in air, and 50 ul of 3H-thymidine (New England Nuclear. Boston, MA), 10 uci/ml in TCM, were then added to each well. incubated overnight.

The cultures were further Subsequently, the culture contents were harvested onto glass fiber filters by means of a cell harvester (Cambridge Technology Inc., Cambridge, MA), and . . . . .

H-thymidine incorporation into cellular DNA was measured liggg by liquid scintillation counting. in triplicate.

All samples were assayed ln purifying CLMF it was necessary to define units of activity in order to construct chromatographic elution profiles and to calculate the percent recovery of activity and the specific activity of the purified material. To do this, a partially purified preparation of human cytokines produced by coculturing PHA-activated human PBMC with NC—37 cells was used as a standard. The preparation was assigned an arbitrary titer of 2000 units/ml. Several dilutions of this preparation were included in each TGF or LAK induction assay. The results obtained for the standard preparation were used to construct a dose—response curve from which could be interpolated units/ml of activity in each unknown sample at the dilution tested. Multiplication of this value by the dilution factor yielded the activity of the original sample expressed in units/ml.

For antibody neutralization studies, the TGF assay was modified as follows. Twenty—five ul aliquots of CLMF—containing medium were mixed with 50 ul aliquots of serial dilutions of antiserum or antibody solutions in the wells of COSTAR 3596® microplates. incubated for 30 minutes at 37°C, The mixtures were and 25 ul aliquots of a suspension of FHA-activated lymphoblasts (8 x lO5/ml in TCM plus 1:100 anti—rIL-2) were then added to each well.

The cultures were further incubated, pulsed with 3H—thymidine. harvested. and analyzed for 3H—thymidine incorporation as described above.

Natural killer (NK) cell activation assay.

Purified CLMF was tested for its ability to activate NK cells when added alone or in combination with IIL-2 as follows. Human PBMC were isolated by centrifugation over discontinuous Ficoll and sucrose gradients as described above and were suspended in RPMI 1640 medium supplemented with 10% heat—inactivated fetal bovine serum, penicillin, 100 ug/ml streptomycin, units/ml and 2vmM L—glutamine.

The PBMC were incubated overnight at 37°C in 1 ml cultures (5 x 106 cells/culture) together with rIL-2 and/or purified CLMF at various concentrations. After 18-20 hours, the contents of the cultures were harvested and centrifuged, and the cells were resuspended in the same medium used for the overnight cultures. The cytolytic activity of the cultured PBMC was then assessed in 51 described above.

Cr release assays as Concentration of cell supernatant solutions Stored, frozen crude human CLMF supernatant solutions totaling 60 liters prepared from several batches NC—37 cells were pooled and concentrated 30-fold Pellicon Cassette System (30,000 NMWL PTTKOO005: Corp.. Bedford, MA). After concentrating to the volume of approximately 1.9 liters, performed with 10 mM MES.

NaOH. of induced using the Millipore desired a buffer exchange was pH adjusted to 6.0 with 10 N The concentrate was centrifuged at 10,000 x g for minutes at 4°C and the precipitate discarded.

Ion—Exchange Chromatography on NuGel P—SP Column The concentrated supernatant solution was applied at a flow rate of 120 ml/hr to a Nu-Gel P-SP (Separation Industries, Metuchen, NJ) column (5 x 5 cm), l0mM MES, pH 6.0. absorbance monitoring at 280 nm was obtained. equilibrated in The column was washed until baseline Absorbed proteins were then eluted with a 500 ml salt gradient from to 0.5 M Nacl/l0 mM MES, pH 6.0 at a flow rate of 2 ml/min 1). Aliquots of fractions were assayed for T cell growth factor (TGF) activity. Fractions containing TGF activity were pooled and dialyzed (Spectra/Por 7, scientific) against 50 volumes 20 mM Tris/Hcl, (Fig.

Fisher pH 7.5 in gggewm _ 33 _ order to reduce the salt concentration of the preparation by 50-fold.

Qye—Affinity Chromatography on Blue B-Agarose Column The dialyzed sample was centrifuged at 10,000 X g for 10 minutes at 4°C and the precipitate discarded. The supernatant solution was applied at a flow rate of 20 ml/hr to a Blue B-Agarose (Amicon, Danvers, MA) column (2.5 x lo cm) equilibrated in 20 mM Tris/Hcl. pH 7.5. The column was washed with this same buffer until baseline absorbance monitoring at 280 nm was obtained. Absorbed proteins were then eluted with a 500 ml salt gradient from 0 to 0.5 M Nacl/20 mM Tris/HC1, (Fig. 2). activity. pH 7.5 at a flow rate of 15 ml/hr Aliquots of fractions were assayed for TGF Fractions containing TGF activity were pooled and dialyzed (Spectra/Por 7. Fisher Scientific) against 100 volumes 20 mM Tris/Hcl, pH 7.5 in order to reduce the salt concentration of the preparation by 100—fold.

Ion—Exchange Chromatography on Mono Q Chromatography The dialyzed sample was filtered through a 0.45 um cellulose acetate filter (Nalgene Co., Rochester, NY) and the filtrate applied at a flow rate of 60 ml/hr to a Mono Q HR 5/5 (Pharmacia LKB Biotechnology, Inc., Piscataway, NJ) column (5 X 50mm) equilibrated in 20mM Tris/HCl. pH 7.5.

The column was washed with this same buffer until baseline absorbance monitoring at 280 nm was obtained. Absorbed proteins were then eluted with a 1 hr linear salt gradient from O to 0.25 M Nacl/20 mM Tris/HCl. of 60 ml/hr (Fig. 3). pH 7.5 at a flow rate Aliquots of fractions were assayed for TGF activity and protein purity was assessed without reduction by SDS-PAGE [Laemmli, (1970)) using 12% slab gels.

[Morrissey, Anal. Biochem. protein (Fig. 4).

Nature (London) gg1:6BO—685 Gels were silver stained ;;1:307—3l0 (1981)) to visualize Fractions 36 and 37 were of greater than _ 34 _ % purity and revealed a major band at 75.000 weight. molecular Fractions 38 through 41 containing TGF activity, revealed the 75 kDa protein by SDS—PAGE with major contaminants at 55.000 and 40,000 molecular weight.

Therefore, to eliminate these contaminating proteins, fraction 38 of the previous Mono Q chromatography was diluted 1:1 vol/vol with 8 M urea and pumped onto a Vydac diphenyl column using a reversed-phase HPLC enrichment technique. The column was then washed with 5 ml of 0.1% trifluoroacetic acid. Elution of the proteins was accomplished with a gradient of 0-70% acetonitrile over 7 hrs in 0.l% trifluoroacetic acid (Fig. 5). Aliquots of fractions were assayed for TGF activity. Protein purity of the fractions containing TGF activity was assessed by sDS—PAGE under non—reducing conditions using a 10% slab gel. The to visualize protein (Fig. 6). Fractions 86 through 90 were of greater than 95% purity and revealed protein of 75,000 molecular weight. gel was silver stained Fractions 87 and 88 were pooled and aliquots were analyzed by SDS-PAGE under reducing (in the presence of B—mercaptoethano1) and non-reducing conditions (in the absence of B—mercapto- ethanol). Under the reducing conditions, the 75,000 molecular weight CLMF was separated into two subunits of ,000 and 35.000 daltons (Fig. 7). Thus it was concluded that CLMF is a 75 kDa heterodimer composed of disulfide- —bonded 40 kDa and 35 kDa subunits.

The overall purification of CLMF that was achieved is shown in Table 1. The protein content of the Mono Q— and Vydac diphenyl-purified material was calculated on the basis sea x ~.m Dm...~.Q ZDH. gong... oHo.o aoo.c noa x m~.m moa x v>.m H.H Hxzmzamo L .

A NVAIQM comavwum nw nofl x m.a mov.o Hao.o pea x mv.m ooa x oo.a m o ocoz “U v n! m! rm comuomub mum uoa x m.a m>o.o mno.o sea a v.o wofi x ov.u H 0 ozoz nofi x m.H Ha v~.o sea a v.~ cod x HH.m me mmo.mo<»m«o:fim oofi x m.~ no o>.o sea x o.~ mofl x oo.~ om amnm Hmoaz mumuuzmuzou _ vofi x m.m ommm mm.~ cog x o.m noa x ~m.H ova.H cmuwpmmumuuaa L mucmamzummaw az az oz mofl x u.~ moa x mm.~ ooo.ow dfiwu cmfioom .as\:. Amev Aasxaav Any AHE\:v Afisv >aM>fluo< cmmpoum cwmuoum mums: »g«>flyo< ofiiummm Hmuou. vofioom Amuse _wUHOO& ®E:HO> H mqm swam _ 35 - of amino acid analysis. A specific activity of 8.5 x 107 units/mq and 5.2 x 10 units/mg for Mono Q» and vydac dipneny1—purified material respectively, was obtained. fact that the diphenyl-purified protein n specific activity than the Mono Q—purifie The» as a slightly lower d material may be due to inactivation or denaturation of some of the molecules of CLMF in the HPLC elution solvents (i.e 0.1% trifluoroacetic acid). .. acetonitrile in Chemical Characterization The ability to prepare homogeneous CLMF allowed for the first time the determination of the amino acid and a partial sequence analysis of the nature CLMF protein. Between lo and 20 picomoles of Mono—Q—purified CLMF was subjected to hydrolysis, amino acid composition was composition lly occurring and its determined (Table 2). cysteine and tryptophan were not determined (ND).

Quantitation of histidine was not artifact peak.

Proline, possible due to a large associated with Tris, coeluting with His (*).

Between 5 and 30 picomoles of dipheny1—purified CLMF was subjected to hydrolysis with and without pre—treatment with performic acid. Complete amino acid composition was thus obtained (Table 3) with the exception of tryptophan.

Amino-terminal sequence determination was attempted by automated Edman degradation on 100 pmol of the Mono Q—purified CLMF. Data from the first 22 cycles indicated two sequences present, as would be expected from the heterodimeric structure of CLMF. summarized as follows: These results may be ..-———_ | F) ‘< 0 p—- II) c-.—.—..

Amino Agid mol 2 Aspaztic acid or asparaqine 11.8 Tnreonine 7.8 serine 8.4 Glutamic acid or glutamine 14.9 Proline ND Glycine 6.2 Alanine 7.6 Cysteine ND Valine 6.9 Methionine 2.0 Isoleucine 44.6 Léucine 9.0 Tyrosine 3.7 Phenylalanine 4.0 Hiatidine * Lysine 9.3 Arginine 5.4 Trypcophan ND TABLE 3 Amino acid Aspartic acid or ésfiaragine Threonine serine Glutamic acid or glutamine Praline Glycine Alanine Cysteine valine Methionine Isoleucine Leucine Tyrosine Phenylalanine Histidine Lysine Arqinine Tryptophan ‘ DEE]/BE. gggm 0 07’ Reversed—Phase HPLC .___.____.____________ pnic system has been described previously by Stern, A.S. and Lewis. R.V. (1935) in Research Methods in Neurochemistry, Eds. Marks. N. and Rodnignt. R (Plenum, New York) Vol. 5, 153-193 ’ phase HPLC was carried out usinq Vydac C18 or (4.6 x 20mm. The sep/a/ra/tions Group, Hesperia, CA). Proteins were e gradient in 0.1% TFA.

Pgotein gnalvsis lured with an acetonitrile instrument which column reaction with fluorescamine for detection and Stein, 5. (1986) in Methods of Microcnaracterization (shively, J.E., Ed.). pp.

Humana Press. Clifton, NJ].

Biosystems Inc.

CA) Ifiewick. w.J., J.

.. Hood, L.E., and 256:7990~7997 (1981)).

TUTQI P D5 To clarify this concentrate after thawing, the preparation was centrifuged and the precipitate discarded.

The supernatant solution was applied to a Nu—Ge1 P—SP column and protein was eluted with a 8). salt gradient (Fig. and the active order to reduce the by 50-fold. This after centrifugation to remove particulates, applied to a Blue-B—Agarose column. a salt gradient (Fig. 9).

Peak TGF activity was determined fractions were pooled and dialyzed in salt concentration of the preparation material. was Protein was eluted with Peak TGF activity was determined and the active fractions were pooled and dialyzed in order to reduce the salt concentration of the preparation by loo-fold. This material. after filtration, was applied to a Mono Q column. Protein was eluted with a salt gradient (Fig. 10). Aliquots of fractions were assayed for TGF activity.

Fractions 39 and 40 of the previous Mono Q chromato- graphy were pooled and diluted 1:1 vol/vol with 8M urea and pumped onto a Vydac diphenyl column using an enrichment technique. The column was then washed with 5 ml of 0.1% trifluoroacetic acid.

Elution of the proteins was accomplished with a gradient of 0-70% acetonitrile over 7 hrs in 0.1% trifluoroacetic acid (Fig. 11). fractions were assayed for TGF activity.

Aliquots of Protein purity of the fractions containing TGF activity was assessed by SDS—PAGE under reducing conditions, B-mercaptoethanol (Fig. 12). Fractions 94 through 97 contained the 40,000 dalton subunit >90% pure. i.e. in the presence of Determination of the amino—terminal sequences of the subunits of CLMF Amino terminal sequence determination was attempted by automated Edman degradation on 20 pmol of the diphenyl— purified 40,000 dalton subunit. summarized as follows: The results may be Cycle 1 2 3 4 5 6 7 Amino ——————— -- Acid I W E L K K D Cycle 8 9 10 ll 12 13 14 Amino — — Acid V Y V V E L D Cycle 15 16 17 la 19 20 21 Amino Acid W Y P D A P G Cycle 22 23 Amino Acid E M ...____.._-_-.._._-.__..__._... with regard to the sequence analysis of 75,000 dalton CLMF and the sequence analysis of the 40,000 dalton subunit of CLMF, one can deduce the amino terminal sequence of the ,000 dalton subunit of CLMF. of the 35,000 dalton subunit and can be summarized as follows: The amino terminal sequences the 40,000 dalton subunit !E990@©7 ,000 dalton subunit: NH —?-?—Leu—Pro-Val—Ala—Thr(?)-Pro—Asp—Pro—G1y_ Met—Phe—Pro-?-Leu—His—His—Ser(?)—G1n— ,000 dalton subunit: NH —Ile—Trp—G1u—Leu—Lys-Lys—Asp—Va1—Tyr—Val-Val—Glu 2 23 Leu—Asp-Trp-Tyr-Pro—Asp—A1a-Pro—Gly-Glu—Met— where ? represents an undetermined or Determination of internal amino acid sequence segments of the 40 kDa subunit of CLMF CLMF was purified as described above. The 40,000 dalton subunit was separated and purified from the 35,000 dalton subunit by the method described by Matsudaira [J. Biol.

Chem. gggz 10035-10038 (1987)). Fifty micrograms of CLMF (in 500 ul of 20 mM Tris, pH 7.5; 0.15 M Nacl) was diluted with 200 ul of a 2 x concentrate of sample buffer [Laemmli, Nature ggz: 680-685 (1970)). containing 12% polyacrylamide and electrophoresed according to Laemmli (supra). After electrophoresis, the gels were soaked in transfer buffer (10 mM 3—cyc1ohexylamino—l—pro— panesulfonic acid, 10% methanol, pH 11.0) for 5 minutes to reduce the amount of Tris and glycine. During this time, a polyvinylidene difluoride (PVDF) membrane (Immobilon; Millipore; Bedford. MA) was rinsed with 100% methanol and stored in transfer buffer. The gel, backed with two sheets "best—guessed" residue. min at 0.5 Amps in transfer buffer. The PVDF membrane was washed in deionized H20 for 5 minutes. The edge of the then destained in 50% methanol, 10% acetic acid for 5-10 minutes at room temperature. The 40,000 dalton stained band was then matched to the corresponding region of the subunit was cut from the By this method, was identified as the the 40,000 dalton protein 40,000 subunit of CLMF. and collected in an Eppendorf They were then immersed in 300 ul of a 2% polyvinyl- pyrrolidone (40,000 dalton) solution in methanol. minutes. tube. the quenching mixture was diluted with an volume of distilled water and further minutes. incubated for 5-10 then discarded and the membrane pieces were washed four times with 300 The supernatant solution was Two containing 2 ug of The sample was shaken and incubated for The supernatant solution was then trypsin was added. 4 hours at 37°C. transferred into a second Eppendorf tube and the membrane pieces were further washed once with 100 ul of 88% gtgeeooi _ 44 _ (vol/vol) deionized formic acid and three times with 100 ul of water. All washing solutions were added to the mixture in the second Eppendorf tube. The digest were digestion resultant peptides contained in the pooled separated by narrow bore HPLC . NJ) TABLE 4 Amino Acid Residue No.

Aspartic acid or asparagine 27.9. (28) Threonine 20.7 (23) Serine 24.6 (34) Glutamic acid or glutamine 44.6 (35) Proline ND (14) Glycine 16.3 (15) Alanine 16.2 (14) Cysteine ND (10) Valine 20.9 (23) Methionine 2.5 (2) Isoleucine 10.3 (12) Leucine 22.9 (22) Tyrosine 12.9 (12) Phenylalanine 9.9 (9) Histidine 5.2 (S) Lysine 24.5 (26) Arginine 12.5 (12) Tryptophan ND (10) yore: The results represent the mean of two analyses The Figures The subunit tryptic peptide map of the digested 40,000 dalton is shown in Figure 15. linear gradient of acetonitrile.

Peptides were eluted with a The peaks which were the intact 40,000 dalton subunit The N-re:mina1 hexapeptide (fraction no. 60) was recovered in high yield.

The carboxy—terminal peptide (fraction no. 72) was recovered the predicted C—terminal peptide although the last two amino acids were not positively confirmed by sequencing. that Cys and Ser residues when they occur at the end This is probably due to the fact are not detected well, especially of a peptide. Four potential Asn—linked carbohydrate sites may be predicted from the CDNA sequence. Two peptides containing two of these sites were when peptide 196-208 (fraction no. 70) was no peak was detected at residue 200 indicating that this Asn (predicted by the CDNA) is indeed glycosylated. Peptide 103-108 (fraction no. at residue 103. Therefore, sequenced. sequenced.

S2) yielded Ash this site is not glycosylated. sequence analysis [Hewick et al., J. (l98l)] of fraction no. 55 was detected at the position corresponding to residue 148. be a Cys residue which is normally analysis unless it is modified.

Chem. 256: 7990 no. The site is predicted to not detected by sequence procedure outline). However. the blotted 40,000 dalton subunit was fragmented with the proteolytic enzyme, Staphylococcus aureus V8 protease (Endoproteinase G1u—C, Boehringer Mannheim, Indianapolis, IN). Membrane pieces were digested for 6 hours at 37°C with 20 ug of V8. The peptides were extracted with 88% (vol/vol) formic acid and separated on a Phase Separations column (2 x 150 mm, C8 83, Queensferry, England, UK) (Figure 16). with a linear gradient of acetonitrile.

Peptides were eluted The peaks which were sequenced are numbered according to their fraction number. The amino acid sequence of these peptides is shown in Table 6.

TABLE 6 V8 (G1u—C) 40kDa peptides off PVDF fraction no. residue no. N-terminal sequence Three major peaks of peptide (fraction nos. 47, 54 and 57) containing four peptides were sequenced. All four peptides were from the amino—termina1 region of the 40 kDa subunit indicating that the N—terminus of the protein is most susceptible to V8—digestion.

Figure 17 summarizes the protein structural determination of the 40,000 dalton subunit of.CLMF.

Direct determination of the amino-terminal sequence of the .000 dalton subunit of CLMF SDS—PAGE analysis of the Mono Q fraction 39 (see Fig. 3) under reducing (in the presence of B-mercaptoethanol) and non»reducing (in the absence of B-mercaptoethanol) conditions (Fig. 18) demonstrated that the 40,000 dalton molecular weight "contaminant" is "free" 40,000 dalton CLMF subunit (i.e. unassociated with the 35,000 dalton subunit).

The evidence which points to this deduction is that without reduction (lane B. Fig. 18) mainly 75,000 dalton CLMF is present with some 40,000 dalton protein. After reduction (lane C, Fig. 18), the 75,000 dalton CLMF is gone yielding the 35,000 dalton subunit and an enriched 40,000 dalton band.

Fraction 39 of the previous Mono Q chromatography was reduced in 5% B—mercaptoethanol in the presence of 4 M urea and heated for 5 minute at 95°C. The sample was pumped onto a Vydac C-18 column using an enrichment technique and the column was then washed with 5 ml of 0.1% trifluoroacetic acid. Elution of the proteins was accomplished with a gradient of 0—70% acetonitrile over 5 hrs in 0.1% trifluoroacetic acid (Fig. 19). Protein purity of the fractions which were fluorescamine positive was assessed by SDS—PAGE under non—reducing conditions using a 10% slab gel.

). The gel was silver stained to visualize protein (Fig.

Fractions 112 through 117 revealed a diffuse band at ,000 molecular weight which was greater than 95% pure.

The 40,000 dalton subunit and any other proteins present in fraction 39 remained bound to the C-18 column. These proteins (including the 40,000 dalton subunit) were finally eluted with a solution of 42% formic acid/40% l—propanol.

The ability to prepare homogeneous 35,000 subunit allowed for the determination of the amino acid composition and partial sequence analysis of the lower molecular weight subunit of the CLMF protein. Approximately 1 ug of 35 kDa E9®®@@7 -49 ._ subunit was subjected to hydrolysis, and its amino acid composition was determined (Table 7). Proline, cysteine and tryptophan were not determined (ND).

TABLE 7 Amino Acid M01 % Aspartic acid or asparagine 10.9 Threonine 6.7 Serine 8.3 Glutamic acid or glutamine 14.9 Proline ND Glycine 6.1‘ Alanine 7.7 Cysteine ND Valine 6.3 Methionine .9 Isoleucine 4.5 Leucine 10.9 Tyrosine 3.2 Phenylalanine 4.4 Histidine 2.3 Lysine 5.6 Arglnine 5.5 Tryptophan ND Amino—termina1 sequence determination was attempted by automated Edman degradation on 100 pmol of the C-18 purified kDa subunit. Data from the first 20 cycles confirmed the sequence obtained by deduction as described above.

Furthermore, the second amino acid was obtained in addition to amino acids 21 through 26. summarized as follows: These results may be

Claims (14)

Claims:

1 (CLMF) protein charac terized in that a) the subunit comprises the amino acid sequence Val Gln Gln Gln Pro Th: Se: Asa Len Pro His Lys Ile Ala Se: Arg His Se: Ala Asp Len Gln Glu Gln Asn A:g His Len Glu Th: Len Cys Arg Se: Arg Gln Lys MET Se: Lys Ty: Asp Ala Asa Lys MET Val Val Lys val Len Ala Th: Th: Ala Len Len Glu Lys Asn Phe Azg Ala Ty: IIe Set, Azg and that (b) if combined with comprising the amino Ile Ty: Glu Asp Glu Gly Ala Se: Len Ala Asp T:p Lys Pro Gly Gly Gin LEU P:o P:o Glu Val Gly Se: Len Se; Gly Len Ile Cys Th: Gly Se: Val Asp His Th: Len Glu Th: Se: Ala Gln Se: Phe Trp Asp Len Arq T:p Len Se: Ala Glu Se: Th: Atg Len Ty: Se: Ala ASH Th: Asp Th: Len Pro Asp Pro Ala Ty: Asp Len Glu Th: Len IIe MET Gln Lys Glu Lys Arg Phe Lys Th: Th: Ala Val Len Ile Glu Phe MET Len Lys Se: Phe MET Asp Ile Pro IIe Th: Asn Len Glu Gln MET Se: IIe Val Ty: Pro Atg Asp‘ Len Glu Lys Ile SEC Len CYS Asp Atg the second subunit of acid sequence Lys Gly Ile Lys Val MET Trp Asp Glu ,Th: Th: Th: Len Ty: Val Th: Th: His Lys Val Val Len Ile Lys Len Ty: Cys Len His Gln Asn Th: Lys Glu Se: Lys Ala Ile Se: Glu Cys Asp Lys Lys Ty: Len Gly Gly Se: Se: Asp Gln Atg Pco Val Glu Asp Atq Gln Asp Gly Val Pro Lys Asn Gly Cys Phe IIe Gln Len Len Th: the Val Len Asp Glu Gly Glu Pro Gly Th: Val Asp Ala MET Se: Cys Th: Glu Se: Leu Lys Phe Ala Glu Len Se: Phe Asn Th: Se: Se: CYS Se: Th: Len Leu Glu His Ty: A subunit of the Cytotoxic Lymphocyte Maturation Factor Pro MET Se: Th: Cys Len Se: MET Asp Asn Pro Ala Leu CLMF protein Glu Th: Gln Val Gly Asp Lys Atg Phe Th: Ash cys Len Cys Se: Lys Glu Gly Asn Phe Se: Cys Lys Pro Asp Asp Se: Glu Val Ile Lys Th: Val Gly Glu Ala Cys Leu Glu Val Len Ala Ile Asn Gln Phe Asp Phe Asn Trp Th: Glu Phe Len Trp Th: Cys Lys Ala Ty: Ala “E9§©®97 Glu-Glu Se: Leu Pro Ile Glu Val MET Val Asp Ala Val His Ly; Leu Lys Ty: Glu Asn Ty: Th: Se: Se: Phe Phe lle Arq Asp Ile Ile Lys Pro Asp Pro Pro Lys Asn Len Gln Leu Lys Pro Leu Lys Ash Se: Acg Gln Val Glu Val Se: Tcp Glu Ty: Rro Asp Th: Tcp Se: Th: Pro His Se: Ty: Phe Se: Len Th: Phe Cys Val Gln Val Gln GIY Lys Se: Lys M9 Glu Lys Lys Asp Acg Val Phe ‘Tht Asp Lys.Th: Se: Ala Th: Val Ila Cys Arg Lys Asa A13 se; 119 59; Val Arg Ala Gln Asp Atg Ty: Ty: se: Se: Se: Trp Se: Glu Ala Se: Val Pro Cys Se: Trp the combined CLMF protein is active in a T cell growth factor assay.

2. The subunit of claim 1, wherein the combined CLMF protein displays a specific activity of at least 5.2 x 107 Units/mg when determined in a T cell growth factor assay, and when combined with the protein as defined in claim l(b). L1.)

The subunit of claim l or 2 comprising the amino acid sequence Arg Asn Len Pro Val Ala Th: Pro Asp Pro Gly MET Phe Pro Cys Leu His His Se: Gln Asn Leu Leu Arg A1a v51 53; A5“ MET Len Gln Lys Ala Aug Gln Th: Leu Glu Phe Ty; Pzo Cys Tn; sec a1u Glu IIe Asp His Glu Asp IIe Th: Lys Asp Lys Tn: sex Tn; va1 Glu Ala cys Leu pm Leu Glu Leu Th: Lys Asa on set Cys Len Asn Se: Arg Glu Th: Se: Phe IIe Th: Asn Gly Se: Cys Leu Ala Sen Arc Lys Th: Se: Phe MET MET Ala Leu Cys Leu Se: Se: Ile Ty: Glu Asp Len Lys MET Ty: Gln Val Glu Phe Lys Th: MET Asn Ala Lys Len Len MET Asp Pro Lys Acg c1n 119 she Len Asp Glu Asn MET Leu Ala Val Ile Asp Glu Len MET Gln Ala Leu Asn Phe Ash Se: Glu Th: Val Pro Gln Lys Se: se; Len Glu Glu Pro Asp Phe TYC LYS Th! LYS lle Lys Leu Cys lle Leu Leu His Ala Phe Arg IIe Rig R18 Val Th: Ile Asp Aug Val Th: Se: Ty: Leu Asn Ala Se: / iE99@@“7

4. A polynucleotide encoding a subunit as claimed in any of claims 1 to 3.

5. A polynucleotide encoding a subunit as claimed in any one of claims 1 to 4 which polynucleotide comprises the nucleotide sequence RTG CTC GRC AGG TTT AAA ACC ACT GCC CTC AGG CTG. TCC TGC RCA

6. a subunit as claimed in any one of claim all or

7. comprising a polynucleotide any one of claims 1 to 3 or all or TGT GAC CCA GCC TAC CAT RAG RAT CTG GAG CAG ATG TCC ATA GTG A recombinant vector comprisin CCA CAC GGA GTC CCT AAA AAT GGG TGC TTC ATC CAG CTT CTT GCC CTC ATG AGC TGC RCC GAG ACT CTT RAG TTT GCC GAR CTT ACG AGC CGC RGT TTC RAC ACT AGC RGT TGC AGT ACC CTR CTG GAR CAT TAT A microorganism of claim 5.

8. as claimed in any on A polycl AGC TTC CCA RTG TCT RCA TGC CTC AGT RTG GRT RAT CCG GCT CTG CTC GCC TGC CTC GAA GTG CTA GCC ATT CAA TTC GAT TTC RAT CTT AAC CAC RTT GCC TCC AGE GAR ATG AGT TAT BTT TCC. GTG CTC CAC GCC cizvr TC 1: MA GAC CTT CTC GAG CGG GCT CCC TCC ASA CAT TTA GAG ACC TTC CTG GCA RCT ACT GCA parts of the polynucleotide of claim 5. RCC GTG CAR CAR GAR CCA RCC TCT RTG GTT GTG GTG CTG GCC ARC ACT GAT TTG TCT TTT RTG GAT RTT CCA ATC ACT CTC ACT CTC CTR RTC GAA TTC ATG TAC CCT CAT CAA RAG ATT transformed with a recombinant vector CTC CCA CTG GAR RCA TTA ETA ATG CAG GAG AAA CTC GAC g a polynucleotide encoding s 1 to 3 or comprising encoding a subunit as claimed in parts of the polynucleotide e of claims 1 to 3. onal or monoclonal antibody directed to a s ubunit 1E99@@@7

9. A process for producing a subunit according to any one of claims 1 to 3 which process comprises culturing a microorganism transformed with a recombinant vector comprising a polynucleotide encoding the said subunit in a culture medium under conditions permitting the expression of the encoded subunit.

10. A process for producing a subunit according to any one of claims 1 to 3 which process comprises (a) preparing sub-unit peptides of the said subunit by conventional peptide synthesis methods; and (b) coupling the sub—unit peptides under conditions favouring the formation of peptide bonds.

11. A process for producing the CLMF protein which comprises a process as claimed in any one of claims 9 to 10.

12. A pharmaceutical composition comprising a subunit as claimed in any one of claims 1 to 3 and a pharmaceutically acceptable diluent, adjuvant or carrier.

13. Use of a subunit as claimed in any one of claims 1 to 3 for the manufacture of a medicament for antitumor therapy.

14. Use of a subunit as claimed in any one of claims 1 to 3 for the preparation of a CLMF protein.