Anti-lymphotoxin antibody, its production and use

Recombinant lymphotoxin

The application in question is a continuation of the USSN

616 503, filed May 31, 1984. The reference in question is closely linked to USSN 608 316, filed on 7 May 1984, entitled Human Lymphotox, and USSN 616 502, filed on 31 May 1984, entitled Anti-Lymphotoxin.

The present application concerns lymphokines. In particular, it concerns lymphotoxin and its derivatives.

Lymphotoxin was first identified as a biological factor with anti-cellular activity against tumor cell lines. The activity identified as lymphotoxin and derived from mitogen-stimulated lymphocytes is linked to a series of cytotoxic activities ranging from inhibition of cell growth in certain tumor cell lines to characteristic cell proliferation of other transformed cells. However, lymphotoxin activity is characterized by little or no anti-cellular activity in the primary cell cultures or normal cell lines studied. This putative differential cellular activity of lymphotoxin led to in vivo studies suggesting that lymphotoxin may have potent antitumor activity.

Lymphotoxin is a term used to describe a group of molecules. Lymphotoxin molecules have been identified as glycoproteins divided into five molecular weight classes, each heterogeneous in charge. The human alpha (molecular weight 70-90,000) and beta (molecular weight 25-50,000) classes appear to predominate in most lymphocyte supernatants. The alpha molecular weight classes can be separated by reservation into at least seven subclasses, while the beta subclass has been separated into two distinct subclasses (G. Granger et al., Mozes et al., 1981, Cellular

Responses to Molecular Modulators ss. 287 - 310). A complex (molecular weight> 200,000) and γ form, molecular weight 10-20,000, of lymphotoxin have also been identified. The different forms and classes of lymphotoxin differ in their stability and expression kinetics in culture. At low ionic strength, they may still aggregate with the complex class. The lower molecular weight classes of lymphotoxins have been reported to be relatively unstable and poorly degrading to cells compared to the higher molecular weight classes (Hiserodt et al., 1976, Cell. Immun. 26: 211; Granger et al. De Week et al.; pp. 279 - 283). Class γ activity has not been extensively studied due to its instability (G. Granger et al., 1978) Cellular Immunology 38: 388-402). The beta class has also been shown to be unstable (Walker et al., J. of Immun. 116/3 ?: 807-815 (March 1976)).

It should be understood that the terminology of lymphokines is not uniform. Today, the names given to cell culture products are largely a function of the cells believed to produce the product as well as the result of the products in biological assays. However, these products remain largely poorly characterized because many studies have been performed with partially purified preparations and because the assays used to characterize the products are not molecularly specific and in any case are subject to considerable variation. The true identity of the various cytotoxic factors remains unknown due to the lack of standard terminology based on clearly identifiable, distinguishing features, such as amino acid sequences or immune epitopes. Examples of other names given to cytotoxic cell culture products include tumor necrosis factor, NK cell cytotoxic factor, hemorrhagic necrosis factor, and macrophage cacotoxin.

The present application, designated USSN 608,316, filed May 7, 1984, and EP 100 641A (published February 15, 1984), describe the amino acid sequences of human lymphotoxin isolated from the human lymphoblast cell line RPMI-1788.

Hayashi et al., EP 132 125A (published January 23, 1985) describe the recovery of protein from a rabbit and the subsequent stimulation of the reticuloendothelial system of the rabbit. The protein was reported to have antitumor activity and had the amino acid sequence Ser-Ala-Ser-Arg-Ala-Leu-Ser-Asp-Lys-Pro-LeuAla-His-Val-Val-Ala-Asn-Pro-Gln-Val-Glu -Gly-Gln-Seu-GlnTrp-Leu.

The present application, generally designated USSN 628,059, filed July 5, 1984, discloses purification and recombinant synthesis of a cytotoxic human polypeptide, the polypeptide has been identified as a tumor necrosis factor, and its N-terminal amino acid sequence is Val-Arg-Ser-Ser-Ser. -Thr-Pro-Ser-Asp-Lys-Pro-Val-AlaHis-Val-Val-Ala-Asn-Pro.

Ohnisni et al. (U.S. Patent 4,481,137) disclose that they have obtained from a BALL-1 cell culture a substance having a molecular weight of 7 to 9,000, called 0Β_χ3 and suppresses tumor cell growth and has an N-terminal Ala-Ala.

According to Toth and Granger, Mol.Immun.

16: 671-679 (1979), the removal of sialic acid from lymphotoxin-containing lymphocyte supernatants and the addition of nacetyl-glucosamine, galactose, lactose, mannose, α-methyl-mannoside or fucose to the supernatants had no effect on in vitro activity. Toth et al. thus concluded that simple sugars played no role in the activity of the lymphotoxin in question. However, Toth et al. also found that saccharides play an important role in the action of other lymphokines and concluded that they cannot rule out the involvement of more complex forms of oligosaccharides in the cytotoxic activity of lymphotoxin.

Later, Proctor, Klostergaard, and Granger (Clinical Research, 1982, 30 (1): 55A) reported that human lymphocytes, when activated with PHA in the presence of tunicamycin (to inhibit the addition of N-linked carbohydrate moieties to lymphotoxin molecules), released lymphotoxin molecules. According to these authors, immunochemical studies indicated that although the carbohydrate moiety of lymphotoxin was not required for lymphotoxin transport or release by activated lymphocyte into the supernatant, carbohydrate was required for efficient dissociation of the target cell because the carbohydrate moiety corresponded to the lymphocyte moiety.

Other literature that should be examined in this context includes publications, Evans, Cancer Immunol. Immunother. 12: 181-190 (1982); Lee et al., Cell. Immun. 48: 166-181 (1979); De Week et al. publishers, (1980) Biochemical Characterization of Lymphokines ss. 279 - 312; Khan et al. published (June 30, 1982) Human Lymphokines ss. 459 - 477; Aggarwal et al., Presentation at the 3rd International Lymphokine Workshop in Haverford, PA., August 1-5, 1982; Ranson et al., Canser Research 43: 5222-5227 (November 1983); Kull et al., J. of Immun. 126 (4): 1279-1283 (April 1981); J. Sawada, et al., Jpn. J. Epx. Med. 46: 263-267 (1976); G. Granger et al., Cell. Immunol. 38: 388-402 (1978); J. Rundell et al., Immunopharmacology 3: 9-18 (1981);

G. Granger et al., J. Lymphokine Res. 1: 45-49 (1982); N. Ruddle et al., Lymphokine Res. 2: 23-31 (1983);

M. Mitsuhashi et al., UK Patent Application 2,106,117;

H. Enomoto, European Patent Application 87,087A; B.

Williamson et al., PNAS USA 80: 5397-5141 (1983) and S. Wright et al., J. Immunol. 126: 1516-1521 (1981).

Lymphotoxin (or substances identified as lymphotoxin) obtained to date from lymphocyte culture is present in low concentrations, in the order of 0.05 to 2x10 units / l, in supernatants of RPMI-1788 cells or primary lymphocytes. The amounts collected often vary considerably, and primary lymphocytes are expensive. There is a need for an economical method for preparing lymphotoxin (Yamamoto et al., J. of Biological Response Modifiers 3: / 1/76 - 87/1984 /).

Previous methods also failed to produce a lymphotoxin that is homogeneous in amino acid sequence, which is an important feature of drug utility. Lymphotoxin recovered from cell line cultures has heterogeneity with respect to the amino terminus, probably due to proteolytic activity (see USSN 608,316, supra). Cultures of primary lymphocytes obtained, e.g., from overgrown pharyngeal or peripheral blood of the pharynx, must necessarily, for economic reasons, contain cells from many donors. However, the products of these cells will reflect the genetic variation prevalent among donors, so that the resulting lymphotoxin may in fact be a mixture of allelic species. Apparently, the proportions and identities of such alleles will be unknown from batch to batch. There is a need for a method for producing a lymphotoxin that is uniform in its amino acid sequence.

Previous methods are also limited to the preparation of lymphotoxin, wherein the primary amino acid sequences of the product correspond to the amino acid sequences of naturally occurring lymphotoxin. Substitution, deletion or insertion of various amino acids in these sequences would require extensive and expensive chemical modifications, if any. Methods are required to effect modifications to the amino acid sequences of lymphotoxin.

Although data on the antitumor effects and apparent therapeutic value of lymphotoxin activity have been reported in the literature since 1968, lymphotoxin has not been studied in extensive clinical trials and has not been commercialized due to the heterogeneous nature of lymphotoxin available from previous methods. Methods are needed for the economical production of sufficient amounts of lymphotoxin for clinical trials.

The rabbit antiserum capable of neutralizing the cytolytic activity of various cytotoxins has been described in the literature, including those identified as lymphotoxin (Yamamoto et al., Cell. Immun. 38: 403-416 (1978); Gately et al., Cell. Immun. 27:82). - 93 (1976), Hiserodt et al., J. of Immun. 119 (2): 374-380 (1977), Zacharchuk et al., PNAS USA 80: 6341-6355 (October 1983); Ruddle et al., Lymphokine Research 2 (1) 23-31 (1983); Mannel et al., Infection and Immunity 33 (1): 156-164 (1981); Wallac et al. E. De Maeyer et al. published by The Biology of the Interferon System ss. 293-302 (published September 1983) and Stone-Wolff et al., J. Exp. Med 159: 828-843 (March 1984). Because this antiserum is polyclonal, it contains a variety of antibodies to the lymphotoxin immunogen. One or more of these antibodies acts to neutralize lymphotoxin activity. Furthermore, the data in the literature are generally unclear as to the molecular identity of the substance responsible for the lymphotoxin activity used as the immunogen. Diagnosis and immunoaffinity purification methods require a monospecific antibody directed against a clearly and unambiguously identified lymphotoxin molecule. It is an object of the present invention to provide such an antibody.

It is a further object to provide economical methods of synthesis for a form of lymphotoxin having a structure such that substantially all primary lymphotoxin molecules have the same primary amino acid sequence.

Another object of the invention is to provide predetermined modifications in the amino acid sequence of the lymphotoxin form, in particular amino acid deletions, insertions, substitutions or combinations thereof.

It has been an object of the invention to achieve successful recombinant expression of a protein having lymphotoxin activity. The species of lymphotoxin, which is described herein in terms of both natural and 20 alternative amino acid sequences of its activity, is hereinafter referred to as lymphotoxin. Surprisingly, it has been found that DNA encoding lymphotoxin cannot tolerate small amounts of lymphotoxin expressed in homologous cells, and uncertainty has been found as to the time at which messenger RNA encoding lymphotoxin 25 appears in homologous cells. It is also surprising that the biologically active lymphotoxin is expressed in recombinant cells that do not glycosylate the lymphotoxin (or are not expected to do so in the same way as homologous cells), and the lymphotoxin thus expressed recovered has a substantially uniform amino acid sequence without enzymatic hydrolysis. DNA encoding lymphotoxin is expressed in cell cultures in copy numbers in excess of 0.1 to 1x10^x units / liter 35 cultures from which the cells have been disrupted.

The lymphotoxin expressed by the recombinant host cell depends on the DNA or precursors used to encode the lymphotoxin and the host cell selected. In this context, the nucleic acid sequences used for lymphotoxin synthesis are novel. They are characterized by nucleotide sequences that differ from the original or naturally occurring sequence in one or more of the following ways: DNA has no introns; in the case of human lymphotoxin, the intron present is located between nucleotides 284 and 285 (Figure 2a); DNA does not contain free nucleic acid that encodes other proteins of the organism from which the DNA is derived; the nucleic acid encoding the lymphotoxin is coupled to the vector; and / or the nucleic acid may hybridize to the nucleic acid encoding the lymphotoxin, provided that such hybridizing nucleic acid does not have the nucleotide sequence of naturally occurring lymphotoxin-encoding DNA or RNA.

Nucleic acid mutants encoding lymphotoxin are the product of recombinant manipulations. Non-functional mutations in the lymphotoxin-corresponding 5 'nucleic acid, read or unread, are provided to increase expression levels in selected hosts, e.g., by reducing the likelihood of stem and helical RNA constructs in the 5' regions of the nucleic acid, or by replacing naturally occurring nucleic acid with preferred codons.

Mutations in nucleic acids that are expressed rather than non-functional make it possible to produce lymphotoxin species that have the amino acid sequence of the original lymphotoxin or a variant thereof with amino acid sequences that differ from the original lymphotoxin. The lymphotoxin mutant is recovered as such or treated by the host cell to obtain the desired lymphotoxin species.

These nucleic acids or nucleic acids that subsequently hybridize, or fragments thereof, are labeled and used in hybridization assays to identify or determine the genetic material encoding the lymphotoxin.

In lymphotoxin synthesis methods, DNA encoding lymphotoxin is ligated into a vector, the vector is used to transform host cells, the host cells are cultured, and the lymphotoxin is recovered from the culture. This general method is used to synthesize a lymphotoxin having the amino acid sequence of the original lymphotoxin, or to construct new variants of the lymphotoxin, depending on the vector construct and the host cell selected for transformation. Lymphotoxin species that can be synthesized in the present case include lymphotoxin in which leucyl is at the amino terminus, lymphotoxin in which histidyl is at the amino terminus, prelymotoxin, and amino acid variants of lymphotoxin, including (a) fusion proteins with a peptide and a heterologous protein or polypeptide and / or carboxyl-terminal amino acids, (b) lymphotoxin fragments, in particular prelymotoxin fragments, wherein any amino acid between -34 and +23 is the amino terminus of the amino terminus of the fragment, (c) lymphotoxin mutants in which one or more amino acid residues have been substituted, inserted or deleted, (d) derivatives in which methionyl or a modified methionyl (such as formylmethionine) other closed methionyl species) is at the amino terminus, and / or (e) non-glycosylated or variable glycosylated species of all of the above.

If a mammalian cell is transformed with a nucleic acid encoded by a lymphotoxin operably linked to a eukaryotic secretory sequence (the original, including an isotantant derived from an isobody), or if the nucleic acid encoding the lymphotoxin is operably linked to a vector, the nucleic acid encoding the lymphotoxin is operably linked to a vector. ) the host is transformed with the vector and cultured, then lymphotoxin species, which usually do not contain methionine at the amino terminus, are recovered from the culture.

If the DNA encoding the lymphotoxin is operably linked to a vector without a secretory introduction sequence and then used to transform a host cell, the synthesized lymphotoxin species are generally substituted with an amino-terminated methionyl or a modified methionyl residue such as formylmethionine.

The present invention relates to a method for the production of lymphotoxins by recombinant DNA technology and to nucleic acids and vectors useful therefor.

It is described how a mutation in nucleic acid encoding a lymphotoxin in vitro results in the expression of lymphotoxin variants that have not been available to date. First, an N-terminally methionyl- or modified-methionyl-containing lymphotoxin is expressed in host cells transformed with a nucleic acid encoding a lymphotoxin that is directly expressed, i.e., operably linked to a secretory preamble sequence.

Second, location-specific, predetermined, or random mutations have been used in vitro to effect deletions, substitutions, and / or insertions into a nucleic acid encoding a lymphotoxin. Lymphotoxin fusions are accomplished in this manner. Lymphotoxin derivatives obtained by expression of a nucleic acid mutation have modified properties.

Finally, unglycosylated or variable glycosylated lymphotoxins are provided as novel lymphotoxin species. Non-glycosylated lymphotoxin is prepared by prokaryotic expression encoding lymphotoxin. Varying glycosylated symphotoxin species are the product of recombinant culture in transformed higher eukaryotic, usually mammalian, cells.

The lymphotoxin prepared according to the present invention is purified from culture supernatants or lysed cells by immunoaffinity adsorption using an insoluble lymphotoxin neutralizing antibody. This antibody, most efficiently produced in monoclonal cell culture, is obtained in mice by immunization with lymphotoxin adsorbed on aluminum sulfate.

For therapeutic use, the lymphotoxin of the present invention is combined with a physiologically innocuous stabilizer and excipient and is prepared in a sterile dosage form, for example, by lyophilization in vials or storage as stabilized aqueous preparations. Alternatively, the lymphotoxin is incorporated into the polymer matrix for attachment to the tumor or tumor site where the tumors have been removed, thereby providing a slow release of the lymphotoxin as a local, high concentration gradient.

The compositions are administered in therapeutically effective doses by implantation, injection or infusion to animals, especially human patients, with malignancies.

Figure 1a depicts the DNA sequence and the amino acid sequence encoding its lymphotoxin fragment.

Figure Ib illustrates the structure of the synthetic DNA encoding the fragment shown in Figure 1a.

Figure 2a shows the complete amino acid sequence of pre-lymphotoxin, the DNA encoding it, and the 5 'and 3' flanking, untranslated regions.

Figure 2b illustrates a method for constructing an expression vector for methionyl-leucyl-amino-terminated lymphotoxin and its amino-terminated methionyl derivatives.

Figure 3 shows a method for constructing an expression vector for methionyl-histidyl-amino-terminated lymphotoxin.

Figure 4 shows the amino acid sequence of human, mouse and bovine lymphotoxin as well as common mammalian lymphotoxin residues.

Figures 5a and 5b show the structure of a plasmid encoding a fusion of a lymphotoxin and a bacterial signal sequence.

Lymphotoxin is defined for purposes of the present application as a biologically active polypeptide having a region that illustrates the amino acid sequence of lymphotoxin shown in Figure 2a with at least a portion of substantial structural amino acid homology. Biological activity is defined as selective, cytotoxic activity as defined below, immunological cross-reactivity with cytotoxic lymphotoxin, or the ability to compete with cytotoxic lymphotoxin for lymphotoxin cell surface receptors. In the latter two examples, the lymphotoxin need not be cytotoxic per se. Immunologically cross-reactive mutants are useful as immunogens for the production of anti-lymphotoxin in animals, e.g., for the preparation of reagents for immunoassays, while non-cytotoxic, competitive mutants have utility as labeled reagents for competitive immunoassays of the type, biologically active.

Selective cytotoxic activity is defined as the selective lysis of tumor cells or the inhibition of growth in vivo or in vitro, under the same conditions as normal cells. Destruction of tumor cells by lysing the cells in vitro or by necrosis in vivo is a preferred endpoint in the assay, although cell growth and proliferative activity have also been used satisfactorily.

Suitable assays for detecting anti-cellular lymphotoxin activities are described in B. Aggarwal, et al., 1984, J. Biol. Chem. 259 (1),

686 691 and E. Carswell, et al., 1975, Proc. Natl.

Acad. Sci. USA 72, 3666-3670.

In the context of the present application, the specific activity of lymphotoxin is defined as the lysis of the target cell rather than the inhibition of cell growth. The lymphotoxin unit is defined as the amount of lymphotoxin required for 50% lysis of target cells pre-transplanted into each chamber as described in Example 1. However, other methods for determining cytotoxic activity are accepted.

Substantial structural homology generally means that more than 60%, and usually about 70%, of the amino acid residues in a polypeptide are the same or conserving substitutions (corresponding residue) in the sequence shown in Figure 2a.

The polypeptide sequence of the lymphotoxin need not be completely homologous to the sequence shown in Figure 2a. Only a portion of it should be homologous to any portion of the sequence shown in Figure 2a, provided that the candidate has the required biological activity. In general, homology should be illustrated for about 20 to 100 amino acid residues, however, it is understood that random gaps may be used to maximize homology.

Less homology is required for polypeptides within the definition if the region homologous to the sequence shown in Figure 2a is not one of the key regions of lymphotoxin, e.g., regions important for cytotoxic activity. The key regions of the sequence shown in Figure 2a are believed to be approximately residues 162-171,

- 83 and 127 - 148.

Lymphotoxin is specifically defined to exclude human necrosis factor or its naturally occurring animal analogs (D. Pennica et al., Nature 312: 20/27 December, 1984, pp. 724-729 and B. Aggarwal et al., J. Biol Chem. 260/4 /: 2345 2354/1985?).

Structurally similar refers to the predominant properties of amino acid side chains, such as basicity, neutrality or acidity, hydrophilicity or hydrophobicity, or the absence of a steric hindrance. Replacing a structurally similar amino acid with another amino acid is generally known as a conservative substitution.

An important factor in establishing the identity of a polypeptide, such as lymphotoxin, is the ability of the antiserum, which is capable of substantially neutralizing homogeneous lymphoblast (i.e., naturally occurring) lymphotoxin, to substantially neutralize the cytolytic activity of the polypeptide in question. However, it is to be understood that immunological identity and cytotoxic identity may not co-exist. The antibody to lymphotoxin of Figure 2a cannot bind to a protein candidate because the neutralizing antibody happens to be directed to a lymphotoxin that is only adjacent to an area that is critical for lymphotoxin activity but acts as a neutralizing antibody to sterically inhibit lymphotoxin. A protein candidate mutated in this indifferent region can no longer bind a neutralizing antibody, but it would still be a lumphotoxin in the form of substantial homology and biological activity.

Lymphotoxin obtained from cultures of lymphoblastic cell lines has been determined to have the following properties: molecular weight 20,000 or 25,000, depending on the degree of glycosylation and N-terminal heterogeneity;

glycosylation in Asn + 62 (Figure 2a); tendency to aggregate, especially tendency to organize into multimers; isoelectric point about 5.8; pH unstable (loss of cytolytic activity greater than> 50% when stored for 24 hours in ammonium bicarbonate buffer at a concentration of 10 ug / ml at a pH below 5 or greater than about 10); substantial losses of activity upon incubation in aqueous solution for 5 minutes at 80 ° C. Two species of lymphoblast-lymphotoxin molecules have been identified. The 25,000 dalton lymphoblast lymphotoxin species has an amino-terminal leucine residue. Polypeptides with a primary amino acid sequence of 25,000 daltons are called leucylamino-terminated lymphotoxins. Lymphoblast lymphotoxins of the 20,000 dalton (da) type are characterized by an amino-terminated histidine, and the corresponding sequences are termed histidyl-amino-terminated lymphotoxin. It is important to note that these properties describe the original as well as the wild-type human lymphotoxin type obtained from lymphoblast cell cultures. While the lymphotoxin defined herein includes the original, glycosylated lymphotoxin, other related cytotoxic polypeptides are within the scope of the definition. For example, glycosylation normally associated with an animal lymphotoxin may be modified by expression in a heterologous, eukaryotic recombinant host cell to provide a modified lymphotoxin that is not of molecular weight in size or has the same isoelectric point as human lymphoblast. Lymphotoxin, which is completely non-glycosylated, is produced in recombinant bacterial culture and has the same molecular weight, isoelectric point, and other properties as modified lymphotoxin. In addition, post-translational production of pre-lymphotoxin from first animal species, in a cell line derived from second animal species, may result in a different amino-terminal residue than is usually the case. Similarly, the mutation provided herein, for example, makes it possible to alter the amino acid sequence and N-terminus of a lymphotoxin, thereby altering pH stability, isoelectric point, and the like.

Figure 2a shows the amino acid sequence for human lymphotoxin that has been translated. Note that this sequence contains 34 residues of a presequence that is believed to be deleted in human cells, normally modifying the transcription product to which translation is performed (in this context, together with its mutants, pre-lymphotoxin), resulting in a leucylamino-terminal species. The species containing histidyl as the amino terminus is homologous to the species containing leucyl as the amino terminus, except that the first 23 amino acids of the leucylamino-terminal species are missing. All three species, e.g., pre-lymphotoxin, leucylamino-terminal lymphotoxin, and histidylamino-terminated lymphotoxin, as well as their methionyl, modified methionyl, mutant, and glycosylated forms, are included in the scope of lymphotoxin. Non-glycosylated leucyl- and histidyl-amino-terminal species have lower molecular weights than described above for homologous species derived from lymphoblast cells.

Pre-lymphotoxin is a species of lymphotoxin that falls within the definition above. It is characterized by a signal (or introductory) polypeptide at the amino terminus of the molecule. Generally speaking, the original signal polypeptide of lymphotoxin is proteolytically cleaved from lymphotoxin as part of a secretory event in which the protein is secreted from the cell. The signal peptide may be of microbial or mammalian origin (including the original, 34-residue sequence), but is preferably a signal homologous to the host cell. Some signal17 lymphotoxin fusions are not recognized or processed by the host cell into N-terminal, methionyl-free lymphotoxin. Associations that contain microbial signals have utility, for example, as lymphotoxin immunogens.

It should be noted that the expression of possible cytotoxic activity means that the lymphotoxin contains polypeptides that can be converted, e.g., by enzymatic hydrolysis, from a zymogen-like, inactive state to a polypeptide fragment having the desired biological activity. The term "possible cytotoxic activity in vitro or in vivo" is intended to encompass non-cytotoxic polypeptides that can be converted, e.g., by enzymatic hydrolysis, from an inactive, zymogen-like state to a polypeptide fragment having detectable biological activity. Typically, inactive precursors will be fusion proteins in which the lymphotoxin is peptide-linked at its carboxyl terminus to another protein or polypeptide. The sequence flanking or adjacent to this peptide bond is selected to be sensitive to proteolytic hydrolysis to release lymphotoxin, either in vivo or as part of a method of preparation, in vitro. Typical linker sequences are lys-lys or arg-lys. The non-lymphotoxin component, such as pro-lymphotoxin, is preferably a homologous protein so as to reduce the immunogenicity of the fusion. The homologous protein should be harmless and should not bind to cell surfaces. The lymphotoxin thus obtained will then have a determinable required cytotoxic activity.

Although lymphotoxin is commonly referred to as human lymphotoxin, lymphotoxin derived from a mouse, pig, horse, or ox is included in the definition of lymphotoxin18 insofar as it meets the standards described above for homologous regions and biological activity. For example, bovine and mouse lymphotoxins have been found to be highly homologous (approximately 80%) to human lymphotoxin. Lymphotoxin is not species-specific, e.g., human lymphotoxin is active against mouse tumors and neoplastic cell lines. Therefore, lymphotoxin from one species can be used to treat another species.

Lymphotoxin also includes multimeric forms. The lymphotoxin spontaneously aggregates into multimers, usually dimers or higher multimers. Multimers are cytotoxic and accordingly suitable for use in therapy in vivo. Lymphotoxin is expressed in recombinant hosts as a monomer. However, lymphotoxin does not tend to form multimers spontaneously afterwards. Homogeneous multimers or mixtures of different multimers are therapeutically useful.

Lymphotoxin variants include predetermined or targeted, e.g., site-specific, mutations in the molecule or fragments thereof shown in Figure 2a. Lymphotoxin variants are defined as polypeptides that otherwise have the properties defined for lymphotoxin, except that they are characterized by an amino acid sequence that differs from the sequence shown in Figure 2a, either in terms of residue deletion, substitution, or insertion. The non-human lymphotoxins as well as the human lymphotoxin alleles described herein are considered variant lymphotoxins because they are site-directed mutants that do not have any naturally occurring duplicates. The object of the mutation is to construct DNA which encodes a lymphotoxin as defined above, but which has properties that modify the biological activity of the naturally occurring lymphotoxin or facilitate the production of the lymphotoxin. For example, the lysine +89 codon undergoes a mutation to express a histidine residue in place of the lysine residue. Histidine +89 is no longer hydrolyzed by trypsin (which usually cleaves proteins at the arg-X or lys-X bond). Protease resistance is expected to give the mutant a longer biological half-life than that of a lymphotoxin having the sequence of Fi Gure 2a (or a fragment thereof). Other lysine and arginine residues of lymphotoxin may mutate to histidine, such as lysine +28, lysine +19, or arginine +15.

As discussed above, certain regions of the lymphotoxin molecule have substantial homology to a similarly active protein, termed tumor necrosis factor. Amino acid residues in these substantially homologous as well as immediately flanking regions are preferred for mutations aimed at identifying lymphotoxin Mutants with Aberrant Biological or cytotoxic activity. Such Mutants are prepared by methods known per se and screened for the desired Biological activity, for example, increased cytotoxicity, the particular tumor being treated, or, in the case of a species of lymphotoxin for immunization of animals, the ability to elicit a Stronger immune response. Examples of such lymphotoxin variants are as follows: Ala = 168 has undergone a mutation to a branched chain amino acid (vai, Ile or Leu); a hydrophobic amino acid (eg ., Phe, vai, Ile or Leu) is inserted by insertion between thr + 163 and val + 164; tyrosine has replaced thr + 163; lysine has replaced ser + 82; ser + 42 is replaced by isoleucine, leucine, phenylalanine, valine or histidine; lys + 84 is replaced by glutamine, tryptophan, serine or histidine; ser + 82 has been removed; A hydrophobic di- or tri-peptide is attached to Leu + 171; thr + 163 is replaced by aspartic acid or lysine; ala-lys is inserted between Glu + 127 and pro + 128 by insertion; ser + 70 has been replaced by lysine or glycine; thr + 69 is replaced by tyrosine; lys + 28 is replaced by arginine or histidine; his + 32 is replaced by arginine or lysine; asp + 36 has been replaced by proline, serine, threonine, tyrosine or glycamic acid; ser + 38 is replaced by tyrosine, methionine or glutamic acid; ser + 61 is replaced by threonine, tyrosine, histidine or lysine; Gly + 124 is replaced by aspartic acid, serine or tyrosine; his + 135 is replaced by arginine, lysine, tyrosine, tryptophan or proline; thr + 142 is replaced by aspartic acid; and gln + 146 is replaced with lysine or threonine.

A highly desirable group of mutants are mutants from which the methionine residues +20, +120 and +133 of human lymphotoxin residues have been removed, or preferably substituted with corresponding residues present in lymphotoxins of other species, such as those described elsewhere in the present application. . For example, met + 20, +120 and +133 are substituted with threonine, serine and valine, respectively. These are the corresponding residues of bovine lymphotoxin. Substitution is accomplished as described in Example 9, except that met + 133 undergoes mutation vals by the next step of mutation using M13 phage according to methods known per se. This animal hybrid lymphotoxin DNA mutant is used in place of the DNA comprising the leucyl amino terminus of Example 7, and is expressed as a fusion. Following known methods, cyanogen bromide is used to cleave the STII signal from the hybrid lymphotoxin, and a mature lysphotoxin containing the leysyl amino terminus is recovered.

Other useful lymphotoxin variants are those in which residues from tumor necrosis factor replace the corresponding lymphotoxin residues to produce hybrids, tumor necrosis factor, and lymphotoxin variants. A typical example is to replace the first 27 residues of leucylamino-terminal lymphotoxin with the first 8, 11, or 10 residues of a mature tumor necrosis factor (e.g., val-arg-ser-ser-ser-arg-thr-pro-ser-asp- ). This transformation is more likely to be one in which the methionyl has been N-terminally removed by direct expression in E. coli.

As long as the mutation site is predetermined, it is unnecessary that the mutation itself be predetermined. For example, to optimize the efficacy of the histidine +89 lymphoxin mutant, a random mutation in the lysine +89 codon is made, and the expressed lymphotoxin mutants are screened for the optimal combination of cytotoxic activity and protease resistance.

The lymphotoxin may also contain insertions, usually about 1-10 amino acid residues, or deletions of about 1-30 residues. Substitutions, deletions, insertions, or any subcombination may be combined to obtain the final structure. Insertions include amino carboxyl-terminated linkages, e.g., a hydrobobic extension added to the carboxyl terminus. Preferably, however, only a substitution mutation is performed. Obviously, mutations in the DNA to be encoded should not be placed outside the reading frame, and preferably no complementary regions should be formed that could produce a secondary mRNA construct. E. coli extracts transformed with vectors containing DNA encoding lymphotoxin mutants lacking at least 16 carboxy-terminal amino acids or lacking the first 33 amino-terminal residues of the leucylamino-terminal lymphotoxin are not cytotoxic ale new. However, the reasons for the lack of activity are not known and may be any of the reasons given in Example 1 below.

Not all mutations in the DNA encoding lymphotoxin appear in the final product of the recombinant cell culture. For example, a major class of DNA substitution mutants are those DNAs in which the secretion signal shown in Figure 2a has been replaced with a different secretion signal, either by deletions 34, within the residue signal, or by substitutions that exchange most or all of the original signals for a signal more likely to be recognized by the intended host. For example, when constructing a prokaryotic expression vector, the secretion signal of Figure 2a is deleted for bacterial alkaline phosphatase or heat-resistant enterotoxin II signals, and the signal of Figure 2a is replaced by yeast invertase, alpha-factor or acid phosphatase signal. However, this does not mean that hosts other than human cell lines do not recognize a human secretion signal, when the host recognizes a secretion signal, the fusion protein comprising ly mphotoxin and a signal is usually cleaved at the signal-lymphotoxin-Peptide bond in the same event leading to lymphocyte. Thus, even if the DNA Mutant is used to Transform a host, the resulting lymphotoxin product may be either fused or the original lymphotoxin, depending on the efficiency of the host cell in the Fusion event.

Another major class of DNA mutants that do not manifest as lymphotoxin variants are nucleotide Substitutions made to increase expression, primarily by avoiding Stem-loop structures in the transcribed mRNA (see pending patent application

USSN 303,687, which is incorporated herein by reference) or to provide codons that can be more easily transcribed by a selected host, eg, well-known E. coli preference codons for E. coli expression.

The Nucleic Acid Mutant is prepared by methods known per se (A. Hui et al., 1984, the EMBO Journal 3 (3): 623-629; J. Adelman et al., DNA 2 (3): 183-193; UK Patent Application 2,130,219A G. Winter et al., 1982, Nature 299: 756-758, and R. Wallace et al., 1981, Nucleic Acids Research 9 (15): 3647-3656). These methods include the M13 phage mutation, the synthesis of the lymphotoxin mutant gene as described in Example 1 et seq., Or other known methods.

A nucleic acid encoding a lymphotoxin is any DNA or RNA that encodes a polypeptide that falls within the scope of a lymphotoxin as defined herein, whether or not its nucleotide sequences correspond to naturally occurring sequences. In addition, the invention encompasses a nucleic acid that can hybridize, at least under less stringent conditions, to a nucleic acid encoding a lymphotoxin, although the hybridizing nucleic acid does not encode a protein that otherwise meets the requirements specified for a lymphotoxin. An example of the latter would be a probe which, due to the shortness of the polypeptide it encodes, is unable to express a biologically active lymphotoxin. Nucleic acid encoding lymphotoxin or may additionally hybridize is prepared by organic synthesis, essentially as described in Example 1, or obtained from naturally occurring sources by examining gene or cDNA libraries as described in the examples.

The lymphotoxin of the present invention is prepared by a method which generally necessarily involves transformation of the host with a vector comprising a nucleic acid encoding the desired lymphotoxin. Vekto24 has a replicable DNA structure. Vectors are used herein to expand DNA or to express DNA encoding lymphotoxin. An expression vector is a DNA construct in which a DNA sequence encoding a lymphotoxin is operably linked to a suitable regulatory sequence capable of effecting the expression of the lymphotoxin in a suitable host. Such control sequences include a transcriptional regulatory region (promoter), an optional operator sequence for transcriptional control, a sequence encoding an appropriate ribosomal binding site for mRNA, and sequences that regulate transcriptional and translational termination.

The vector may be a plasmid, a virus, or an insertable DNA fragment (e.g., recombinantly insertable gene material). Once the vector has been transformed into a suitable host, the vector replicates and functions independently of the host's genetic material, or may, in some cases, integrate into the genetic material itself. In the present application, the plasmid and the vector are sometimes used without distinction, although the plasmid is currently the most commonly used form of the vector. However, all other forms of vectors that are equivalent in function and that are or will be known in the research are suitable for use in this context.

Suitable vectors include a replicon and control sequences derived from species compatible with the intended expression host. Transformed host cells are cells transformed or transfected with lymphotoxin vectors constructed using recombinant DNA technology. Transformed host cells usually express lymphotoxin. The expressed lymphotoxin remains inside the cell or is secreted into the periplasmic space or culture supernatant, depending on the host cell selected.

DNA regions are operably linked when they are functionally related. For example, DNA corresponding to a presequence or secretory introductory sequence is operably linked to DNA corresponding to a polypeptide if it is expressed as a preprotein involved in the secretion of the polypeptide; a promoter is operably linked to a coding sequence if it regulates transcription of the sequence; or a ribosome binding site is operably linked to a coding sequence if it is located so as to permit translation. Generally speaking, operably linked means contiguous and, in the case of a secretion sequence, contiguous and a reading step.

Suitable host cells are prokaryotes, yeast cells or higher eukaryotic cells. Prokaryotes include gram-negative or gram-positive organisms, for example, E. coli or Bacilli. Higher eukaryotic cells include engineered, mammalian cell lines as described below. The preferred host cell is the phage-resistant E. coli strain W3110 (ATCC 27 325) described in the examples, although other prokaryotes, such as E. coli B, E. coli X1776 (ATCC 31,537), E. coli 294 (ATCC 31,446), Pseudomonas species or Serratia marcescens are suitable.

Prokaryotic host-vector systems are preferred for lymphotoxin expression. Suitable microbial vectors are abundantly available. Generally speaking, a microbial vector contains an origin of replication recognized by the intended host, a promoter that functions in the host, and a phenotypic selection gene, e.g., a gene encoding proteins that confer antibiotic resistance or complement the auxotrophic requirement. Similar constructs are prepared for other hosts E. coli is typically transformed using pBR322, a plasmid derived from a

From E. coli (Bolivar, et al., 1977, Gene 2:95).

pBR322 contains genes for ampicillin and tetracycline resistance and thus provides easy means to identify transformed cells.

Expression vectors must contain a promoter that is recognized by the host organism but does not need to be recognized by the cloning vectors. The promoter is generally homologous to the intended host. Promoters most commonly used in recombinant DNA construction include β-lactamase (penicillinase) as well as lactose promoter systems (Chang et al., 1978, Nature, 275: 615; and Goeddel et al., 1979, Nature 281: 544), tryptophan ( trp) promoter system (Goeddel et al., 1980, Nucleic Acids Res. 8: 4057 and EPO App. Pub No. 36,776) and the tac promoter / H. De Boer et al., Proc. Nat'l. Acad. Sci. USA: 80: 21-25 (19832). Although these are the most commonly used, other known microbial promoters are suitable. Details of their respective nucleotide sequences have been published, allowing those skilled in the art to link them to plasmid vectors encoding lymphotoxin (Siebenlist et al., 1980, Cell 20: 269) as well as DNA encoding lymphotoxin. The presently preferred vector is a pBR322 derivative containing the E. coli alkaline phosphatase promoter with the Shine-Dalgarno sequence. The promoter and the Shine-Dalgarno sequence are operably linked to the DNA encoding the lymphotoxin, e.g., they are located so as to promote transcription of the lymphotoxin mRNA from the DNA.

In addition to prokaryotes, eukaryotic microbes, such as yeast cultures, are transformed with vectors of photomoxin. Saccharomyces cerevisiae, or common baker's yeast, is most commonly used among lower eukaryotic host microorganisms, although many other strains are commonly available. Yeast vectors generally contain an origin of replication from a 2 micron yeast plasmid or an independently replicable sequence (ARS), a promoter, DNA encoding lymphotoxin (including human pre-lymphotoxin in particular), sequences for polyadenylation, and transcription termination and selection genes. A gene suitable for lymphotoxin expression in yeast is YRp7, (Stinchcomb et al., 1979, Nature, 282:39; Kingsman et al., 1979, Gene, 7:141; Tschemper et al., 1980,

Gene, 10: 157). This plasmid already contains the trp1 gene, which provides a selection marker for a mutant strain of yeast lacking the ability to grow in tryptophan, for example for ATCC No. 44076 or PEP4-1 (Jones, 1977, Genetics, 85:12): presence of trp1 damage in yeast host cell genetic factors then provides an effective environment for detecting transformation in the absence of tryptophan.

Suitable promoter sequences in yeast vectors include promoters for metallothionein, 3-phosphogyly20 serate kinase (Hitzeman et al., 1980, J. Biol.

Chem., 255: 2073) or other glycolytic enzymes (Hess et al., 1968, J. Adv. Enzyme Reg., 7: 149; and Holland et al., 1978, Biochemistry, 17: 4900), such as enolase, glyceraldehyde-3-phosphate25 for dehydrogenase, hexokinase, pyruvate decarboxylase, phosphofructokinase, glucose-6-phosphate isomerase, 3-phosphoglycerate mutase, pyruvate kinase, phosphoserase isosera, triosophosphate isomerase Vectors and promoters suitable for yeast expression are further described in R. Hitzeman et al. in, et al.

EPO Pub, No. 73,657.

Other promoters with the additional advantage of growth factor-regulated transcription are the promoter regions for alcohol holide dehydrogenase 2, isocytochrome C, acid phosphatase, degrading enzymes that are involved in nitrogen metabolism, and the aforementioned metallothionein and glyceraldehyde. responsible for the utilization of maltose and galactose. In constructing suitable expression plasmids, the terminal sequences that bind to these genes are also ligated into the expression vector 3 of the lymphotoxin coding sequences to provide polyadenylation and resolution of the mRNA.

In addition to microorganisms, cell cultures derived from multicellular organisms can also be used as hosts. However, this is not advantageous because so far excellent results have been obtained with lymphotoxin-expressing microbes. However, in principle, any higher eukaryotic cell culture derived from either a vertebrate or inverttebrate culture is suitable. However, there has been the greatest interest in vertebrate cells, and proliferation of vertebrate cells in cultures (tissue culture) has become a routine method in recent years / Tissue Culture, Academic Press, Kruse and Patterson, 1973 //. Examples of useful host cell lines include VERO and HeLas cells cell lines (CHO) and WI38, BHK, COS-7 and MDCK cell lines. Expression vectors for such cells usually include (if necessary) an origin of replication located above the gene to be expressed, together with a ribosome binding site, an RNA binding site (if intron-containing gene DNA is used), a polyadenylation site, and a transcriptional terminal sequence.

Transcriptional and translational control sequences located in expression vectors for use in transforming vertebrate cells are often obtained from viral sources. For example, commonly used promoters are derived from the polyoma, Adenovirus

2 and preferably Simian virus 40 (SV40). Early and late promoters are particularly useful because both are obtained from the virus as a fragment that also contains an SV40-derived origin of replication (Fiers et al., 1978, Nature; 273: 113). Smaller or larger SV40 fragments may also be used, provided that they contain a sequence of approximately 250 bps extending from the Hind β site to the Bgl I site located at the viral origin of replication. Furthermore, it is also possible and often also desirable to utilize promoter, control and / or signal sequences from the human gene that are normally associated with lymphotoxin, provided that such control sequences are compatible with the host cell system.

The origin of replication may be provided either by constructing a vector that includes an exogenous origin, for example, it may be derived from SV40 or another viral source (e.g., Polyoma, Adenovirus, VSV, or BPV), or provided by a host cell chromosomal replication mechanism. If the vector is fused to the chromosome of the parental cell20, the latter is often sufficient. Lymphotoxin is often prepared without amino-terminated methionyl, by transforming higher eukaryotic cells with human pre-lymphotoxin DNA.

When selecting a preferred mammalian host cell for transfection with vectors comprising both lymphotoxin and dihydrofolate reductase (DHFR) encoding DNA sequences, it is appropriate to select the host according to the DHFR protein used. If a wild-type DHFR protein is used, it is preferable to select a host30 cell that lacks DHFR and thus allows

Use of the DHFR coding sequence as a marker for successful transfection in a selective medium lacking hypoxanthine, glycine, and thymidine. In this case, a suitable host cell is a Chinese hamster ovary35 cell line (CHO) lacking DHFR activity, prepared and propagated as described by Urlaub and Chai, 1980, Proc. Natl. Acad. Sci. (USA) 77: 4216.

On the other hand, if the DNA encoding the DHFR protein with low binding affinity for metorexate (MTX) is used as a control sequence, it is not necessary to use DFR-resistant cells. Because the DHFR mutant is resistant to MTX, MTX-containing medium can be used as a means of selection, provided that the host cells themselves are sensitive to MTX. Most eukaryotic cells that can absorb MTX appear to be sensitive to methotrexane. One such useful cell line is the CHO line, CHOK1 (ATCC No. CCL 61).

Transformed host cells are cells that have been transformed or transfected with lymphotoxin vectors constructed using recombinant DNA technology. Transformed host cells usually express lymphotoxin. The expressed lymphotoxin usually remains to be stored intracellularly.

Lymphotoxin is recovered from recombinant cultures of non-secreting cells by lysing the cells and removing particulate matter by centrifugation and the like. Lymphotoxin-secreting cells are separated from the culture supernatant by centrifugation. The contaminated lymphotoxin solution is then purified by the methods referred to above or by the immunoaffinity method described in Example 4 below. The lymphotoxin is purified to be suitable for pharmacological use and placed in suitable dosage forms, e.g., bottles or syringes. Preferably, the lymphotoxin is lyophilized for long storage periods or may be placed in aqueous solution with stabilizers and excipients, for example isotonic saline, and administered to patients as described by B. Aggarwal et al., European Patent Application 100641.

Lymphotoxin compositions are administered to tumor-bearing animals. Administration is performed according to known methods, e.g., intravenous, intraperitoneal, subcutaneous, intramuscular, intramuscular, infusions or injections of sterile lymphotoxin solutions, or using the sustained release systems described below. Lymphotoxin is administered intramuscularly, e.g., by direct injection into solid tumors. In the case of metastatic tumors, such as leukemia, administration is preferably carried out intravenously or into the lymphatic system. Abdominal tumors, such as ovarian cancer, are preferably treated by intraperitoneal infusion, using peritoneal dialysis equipment, and solutions compatible with the peritoneum. Usually, however, the lymphotoxin is given as a continuous infusion, although a bolus injection is accepted.

If desired, the lymphotoxin is administered from an implanted, slow-release preparation. Examples of suitable systems for proteins having the molecular weight of a lymphotoxin dimer or trimer include copolymers of L-glutamic acid and γ-L-glutamate (U. Sidman et al., 1983, Biopolymers 22 (1): 547-556), poly- (2- hydroxyethyl methacrylate) (R. Langer et al.,

1981, J. Biomed. Mater. Res. 15: 167-277 and R. Langer,

1982, Chem. Tech. 12: 98-105) or ethylene vinyl acetate (R. Langer et al., Id.). Preparations containing lymphotoxin are implanted at the sites where the tumors have been removed. Alternatively, the lymphotoxin is encapsulated in semipermeable microcapsules or liposomes for tumor injection. This method of administration is particularly suitable in the case of tumors which cannot be resected, for example for the treatment of brain tumors.

The amount of lymphotoxin to be administered depends, for example, on the route of administration, the tumor in question and the condition of the patient. It is necessary for the attending physician to determine the dose titer and modify the route of administration as necessary to obtain optimal cytotoxic activity against the target tumor, such as by tumor biopsy or diagnostic assays for presumptive cancer markers such as carcinoembryonic antigen. In general, doses of recombinant lymphotoxin in mice of about 50 to 200 ug / kg body weight / day, when administered intravenously, have been found to be essentially non-toxic and effective in vivo. Apparently the dosing instructions vary from animal to animal.

According to the present invention, there is provided a method of obtaining a lymphotoxin neutralizing antibody. A neutralizing antibody is defined as an antibody capable of immunologically binding a lymphotoxin as defined in the present invention by substantially reducing lymphotoxin activity in cytostatic or cytolytic assays for lymphotoxin activity, such as the mouse L929 assay described below. The fact that an antibody is capable of neutralizing lymphotoxin activity does not mean that the antibody must bind directly to the active or receptor-binding site of the lymphotoxin. An antibody may appreciably neutralize lymphotoxin activity if it sterically binds to a region adjacent to a critical site, e.g., adjacent to a conformation and not necessarily adjacent to an amino acid sequence.

In an attempt to prepare a neutralizing, monoclonal antibody against lymphotoxin, it seemed difficult to immunize mice in such a way that a neutralizing antibody to lymphotoxin is generated or raised in animals. Immunization with lymphoblast-derived lymphotoxin and lymphotoxin cross-linked with glu33 tardehyde did not produce any detectable amount of neutralizing antibody in the serum of the immunized mouse, even though the mouse was not treated with a neutralizing anti-lymphotoxin antibody. However, the lymphotoxin-alum (alumina hydroxide or alumina, ΑΙ, ^ Ο ^ · 31 ^ 0) adsorption complex generates a neutralizing antibody even in animals that have failed to develop activity prior to immunization with the alum complex. The preparation of alum and its use in the preparation of antisera is published in C. Williams, et al., 1967, Methods in Immunology and Immunochemistry I, p.

197 - 229.

A fusion of neutralizing antibody-producing animal-derived spleen cells with mouse myeloma cells has been performed. On average, about 50 to 100 clones have had to be cloned to identify a single clone that synthesizes a neutralizing antibody. The method of screening clones for the desired activity is routine and is within the skill of the art and can be easily replicated with little effort.

Serum, plasma or IgG fractions from an immunized animal, as well as immunoglobulins secreted from hybridoma cells developed from spleen and lymphocytes of immunized animals, satisfy the use according to the invention. In a preferred embodiment of the patent, the neutralizing antibody is obtained substantially free of other anti-lymphotoxin antibody present in the hybridoma culture.

The neutralizing antibody is immobilized by adsorbing it to surfaces, for example plastic, such as polystyrene, or by covalently binding to binders, such as cyanogen bromide-activated Sepharose.

It is then used for immunological assays or immunoaffinity purification. Since the antibody is a neutralizing antibody, it is most convenient to adsorb it and express the biologically active lymphotoxin or fragments thereof. The antibody is particularly useful in immunoradiometric (sandwich) immunoassays in combination with a non-neutralizing anti-lymphotoxin, a monoclonal antibody, or a polyclonal antiserum containing a non-neutralizing anti-lymphotoxin. The immunological assay is performed using either a neutralizing or non-neutralizing antibody as a labeled component, which labeling is effective with a detectable agent, such as fluorescent, chemiluminescent, or radioisotope labeling, according to known methods. For competitive lymphotoxin assays, the lymphotoxin is labeled in the same manner. Chloramine-T radioiodination is suitable for the preparation of both lymphotoxin and lymphotoxin antibody tracers, or uses the methods described in J. Klostergaard et al., Mol. Immun. 18: 455 (1980).

To simplify the examples, certain common methods are referred to by shorthand expressions.

Plasmids are marked with a lowercase letter p, which is first and / or followed by uppercase letters and / or numbers. The starting plasmids used in the present invention are commercially available, are available to the public without limitation, or can be constructed from such available plasmids according to published methods. Other equivalent plasmids are known in the research and will be apparent to the person performing the work.

Digestion of DNA refers to the cleavage of DNA by an enzyme that only affects certain parts of the DNA. Such enzymes are called restriction enzymes, and the sites to which they are specific are called restriction sites. Partial digestion means incomplete digestion with a restriction enzyme, e.g., the conditions are chosen to result in the cleavage of some, but not all, of the restriction endonuclease-susceptible sites on the DNA substrate. The various restriction enzymes used in this context are commercially available, and their reaction conditions, cofactors, and other requirements specified by the enzyme supplier were used. Restriction enzymes are usually denoted by abbreviations consisting of capital letters followed by other letters and then usually a number which represents the microorganism from which each restriction enzyme is originally derived. Generally, about 1 ug of plasmid or DNA fragment is used with about one unit of enzyme.

<img file="FI93025C_D0001.tif" />

buffers and amounts of substrate suitable for certain restriction enzymes. Incubation times of about 1 hour at 37 ° C are usually used, but may be changed according to the supplier's instructions. After incubation, the protein is removed by extraction with phenol and chloroform, and the digested nucleic acid is recovered from the aqueous fraction by precipitation with ethanol. Restriction by restriction enzyme is rarely followed by hydrolysis of 5 'terminal phosphates by bacterial alkaline phosphatase to prevent cleavage of the DNA fragment. to form a loop, which would prevent the insertion of another DNA fragment into the restriction site. Unless otherwise stated, plasmid digestion is not followed by removal of the 5 'terminal phosphorus. Dephosphorylation methods and reagents are conventional (T. Maniatis et al., 1982, Molecular Cloning pp. 133-134):

Recovering or isolating a particular fragment of DNA from a restriction digest means separating the digest by polyacrylamide gel electrophoresis, identifying the fragment of interest by comparing its mobility to that of marker DNA fragments of known molecular weight, removing the gel fragment containing the desired fragment and separating the gel fragment containing the desired fragment. This method is well known. For example, see R. Lawn et al., 1981, Nucleic Acids Res. 9: 6103-6114, and D. Goeddel et al., 1980, Nucleic Acids Res. 8: 4057.

Southern Analysis is a method by which the presence of DNA sequences in a digest product or a DNA-containing composition thereof is confirmed by hybridization to a known labeled oligonucleotide or DNA fragment. In the context of the present invention, unless otherwise stated, Southern analysis involves the separation of the degradation product on 1% agarose, denturization and transfer to nitrocellulose by the method of E. Southern, E. Southern, 1975, J. Mol. Biol. 98: 503-517, and hybridization as described by T. Maniatis et al., 1978, Cell 15: 687-701.

Transformation means the introduction of DNA into an organism so that the DNA replicates, either as an extrachromosomal material or as part of a chromosomal assembly. Unless otherwise stated, the method used in this context for E. coli transformation is CaCl₂method described by Mandel et al. have published, 1970, J. Mol. Biol. 53: 154.

Binding refers to a method of forming a phosphodiester bond between double-stranded nucleic acid fragments (T. Maniatis et al., Id., P. 146). Unless otherwise stated, binding is performed using known buffers and conditions with 10 units of T4 DNA ligase (binder) per 0.5 μg of approximately equimolar amount of DNA fragments to be bound.

Preparation of DNA from transformants means isolating palmid DNA from microbial culture. Unless otherwise stated, the alkaline / SDS method of Maniatis et al., Id. p. 90.

Oligonucleotides are short, single- or double-stranded polydeoxynucleotides that are chemically synthesized by the method set forth in Example 1 and then purified on polyacrylamide gels.

All citations to the literature are attached as references.

Example 1

Lymphotoxin Sequencing and Purification Human Lymphoblast Cell Line RPMI-1788 (ATCC No.

CCL-156) were grown in 15 liter spinner flasks to a cell density of 4 x 150 cells per milliliter, using serum-free medium (RPMI-1640). Lymphotoxin was induced 10 to 20-fold (500 to 1000 lymphotoxin units per milliliter, determined as described below) above baseline levels by including ng / ml phorbol myristate acetate in serum-free

In RPMI-1640 medium. After 65 hours of culture, the cells were harvested by filtration, and the lymphotoxin activity of the filtrate was absorbed into glass beads of a specified pore size (Electronucleonics) on a column (5 cm x 20 cm), equilibrated with 5 mM phosphate buffer (pH 7.4) and eluted with 50% ethylene glycol in 5 mM phosphate buffer (pH 7.4). All buffers throughout the purification included 0.1 mM phenylmethylsulfonyl fluoride (PMSF), a protease inhibitor, and 1 mN sodium azide to inhibit microbial growth. The eluate from the glass beads contained 84,000 units of lymphotoxin / mg protein.

This was followed by DEAE cellulose chromatography, lentil lectin-Sepharose chromatography, and preparative, original PAGE, as described by B. Aggarwal et al.

1984, J. Biol. Chem. 259 (1): 686-691. The homogeneity of the protein responsible for cytotoxic activity was determined by SDS-PAGE, reverse phase HPLC Lichrosorb

On an RP-18 column as well as by amino-terminal sequencing.

This lymphotoxin preparation contained more than% by weight of leucylamino-terminated lymphotoxin with an approximate molecular weight of 25,000 as determined by SDS-PADE. The theoretical molecular weight of the N-terminal leucyl-type protein component is 18,664 daltons; the remainder, approximately 6,500 daltons, was associated with the glycosyl side chain at Asn + 62 and possibly other O-mediated sugar residues. The tissue culture supernatant contained putative multimers of this type (60,000 DA as determined by TSK-HPLC or

000 Da Sephadex G-100 chromatography).

The remaining 5% of the lymphotoxin mixture was of the N-terminal histidyl type with a molecular weight of about 20,000. Both types have approximately the same cytolytic activity, at least within the ranges inherent in the assay methods described below for lysed mouse fibroblast cells.

Treatment of intact lymphotoxin molecules with trypsin produced only a few fragments. Histidylamino-terminated lymphotoxin was cleaved into two fragments between amino acids 89 and 90, whereas trypsin treatment of leucylamino-terminated lymphotoxin yielded four fragments cleaved at positions 15 and 16, 19 and 20, and 89 and 90, respectively.

Microsequencing by the Edman digestion technique yielded sequence information for the intact molecule as well as fragments obtained by trypsin treatment.

Additional sequence information was provided by lymphotoxin fragments obtained by digestion with carboxypeptidase P as well as chymotrypsin, extraction with acetic acid, and cleavage by cyanogen bromide. Almost the entire sequence of human lymphotoxin was determined by this method. 156 contiguous residues were determined from the Amino39 terminal. From this sequencing information, it was clear that the difference between the two lymphotoxin species was based on the location of the amino-terminal residue in leucylamino5-terminal species, which was not observed in histidyl-amino-terminal species. The carboxy-terminated sequence on the other side of the first three residues was found to be difficult to determine due to the hydrophobic nature of certain peptide bonds in this region as well as the residues.

A synthetic gene was designed that would encode a protein sequence to the extent defined by microsequencing. The constructed gene contained a common E. coli codon strand, which means that rarely used E. coli codons were used in the sequence. Preferred human codons act as surrogates where nothing

The codon band of E. coli was not visible. This band was chosen to aid expression in E. coli and also so that the synthetic gene would be useful as a probe for identifying a naturally occurring DNA sequence from human cDNA or gene libraries. Unique restriction sites XbaI, BamHI, HindIII, and BglII were constructed in the sequence to facilitate the construction of fragments and to allow future manipulation of the gene.

the original oligomer designed for the synthetic lymphotoxin gene was synthesized by the solid phase phosphite method of M. Matteucci et al., 1981, J. Amer. Chem. Soc. 103: 3185-390 and S. Beaucage et al., 1981, 1981, Tet. Letters 22: 1859-18662. The size of these oligomers ranges from 16 bases to 20 bases and is shown in Figure 1a. The overlapping portions between the oligomers were six bases in length and were built to be unique.

The entire gene was assembled as shown in Figure Ib.

The gene was constructed in three distinct parts.

The first, segment A, was 117 base pairs long and represented the amino-terminus40 of the leucyl-amino-terminal species at its end 5¹-coding area. Segment B represented the DNA encoding the middle portion of the lymphotoxin molecule and was 145 bp in length. Segment C, which was 217 base pairs in length, was believed to encode all amino acid residues except 16 at the carboxy terminus of lymphotoxin. The oligomers required to synthesize each segment were purified by electrophoresis and then pooled. The relatively small size chosen for each oligomer (i.e., 16 to 20 bases) was chosen to reduce errors in the synthesis.

Each group of oligomers was phosphorylated in a reaction containing 20 mM Tris-HCl (pH 7.5), 10 mM MgCl 2, 20 mM dithiothreitol, 0.5 mM ATP, and 15 units of T4 polynucleotide kinase in a volume of 50 μl; the reaction contained approximately 50 picomoles of each oligomer. After 30 minutes at 37 ° C, the reaction mixture was heated to 65 ° C to destroy kinase activity and then allowed to slowly cool to 20 ° C over a period of about one hour. The phosphorylated oligomers were then ligated by the addition of 10 units of T4 DNA ligase and the reaction was allowed to proceed for two hours at 20 ° C. The DNA ligase was heat inactivated, and then the bound oligomers were extracted for 3 hours at 37 ° C with restriction endonucleases that recognized the constructed terminal sites (e.g., XbaI and BamHI for segment A). Fragments corresponding to each segment were isolated by electrophoresis with 7% polyacrylamide. Fragments with the correct mobility were identified for each segment by staining with ethinium bromide as well as electroelution from the gel. pFIFtrp69 (D. Goeddel et al., 1980, Nature 287: 411-416 or Crea et al., European Patent Application 48048970) was digested with XbaI and BamHI, and a large vector fragment was isolated by 6% polyacrylamide gel electrophoresis. About 50 ng of segment A was ligated to the pFIFtrp69 fragment. Similarly, segment b was bound to BamHI and HindIII digested pBR322 and segment C was ligated to HindIII and BglII digested pLeIFA-125-1 (D. Geoddel et al., 1980,

Nuc. Acids Res. 8: 4057-4073). The binding reaction mixture was transformed into E. coli ATCC 31446, and the resulting recombinant plasmids were characterized by restriction endonuclease analysis and the Maxam and Gilbert chemical digestion method for DNA sequencing. Five of the six segments of the A clone contained the designed sequence. Four segments of plasmid B and four segments of plasmid C were isolated, and all of these inserts had the correct sequences. Each segment was isolated by digestion with restriction endonucleases that recognized the terminal sites and then ligated into the plasmid vector pFIFtrp69 treated with XBaI and BglII. The resulting recombinant plasmid, pLTXB1, was characterized by sequencing an inserted XbaI-BglII fragment containing the sequence shown in Figure 1a.

To determine whether a synthetic gene would indeed produce a biologically active lymphotoxin, transformants of E. coli pLTXB1 were grown in minimal medium under conditions that abolish the repressor effect of the trp promoter and allow expression of the synthetic lymphotoxin gene. Cultures were grown to an optical density of 1.0 at 550 nanometers and harvested by centrifugation. The cell pellet was slurried to one tenth of a volume, and then the cells were disrupted by sonication.

Lymphotoxin activity was determined by a modified cell lysis method published by B. Spofford in 1974, J. Immunol. 112: 2111. Briefly, mouse L-929 fibroblast cells were grown in microtiter plates in the presence of actinomycin D. After 12 to 18 hours, 0.125 ml of serially diluted sample to be determined for lymphotoxin was added to each chamber. 18 After 1 hour, the plates were washed and lymphotoxin-induced cell lysis was observed as adhesion to the plates by staining the plates with a 1% solution of crystal violet in methanol: water (1: 4 v / v :). The intensity of the staining was observed both visually and spectrophotometrically at 450 nm absorbance and 570 nm transmission, using Dynatech spectrophotometry. Cells inoculated with culture medium alone were assigned a lysis of 0%, whereas those with 3M guanidine hydrochloride gave an endpoint of 100% lysis. One unit of lymphotoxin is determined as the amount required to lyse 50% of the 12,000 cells inoculated into each chamber. It should be noted that other methods for determining cytotoxic activity may also be used. See, for example, B. Aggarwal et al., In Thymic Hormones and Lymphokines, 1983, published by A. Goldstein, Spring Symposium an Health Sceinces, George Washington Univ. Medical Center (the A549 cell line, considered to be part of this material, is available from the ATCC as CLL185 ). Cell lysis products did not show detectable cytolytic activity in the mouse cell assays described above. Control degradation products from cultures expressing γ-interferon contained γ-interferon activity. This result suggests that the synthetic gene did not encode active lymphotoxin. There were often possible explanations for this. For example: (1) E. coli degraded lymphotoxin, (2) the lymphotoxin gene was not transcribed in E. coli, (3) the lymphotoxin message was not translated into E. coli. coli, (4) the protein did not have a suitable sequence due to a protein sequencing error, or (5) the carboxy-terminated sequence of 16 residues or part of it was in fact necessary for activity or the appropriate configuration of the lymphotoxin molecule.

Example 2

Method for obtaining cDNA encoding lymphotoxin RNA was isolated from a non-adherent cell fraction of human peripheral blood lymphocytes after 48 hours after phorbol myristate acetate (10 ng / ml), staphylococcal enterotoxin B (1 μg / ml) and init al. S. Berger et al., 1979, Biochemistry 18: 5143-549). This culture produced 400 units of lymphotoxin activity / ml supernatant. acetate. mRNA was concentrated by adsorption on immobilized oligo-dT, eluted, and cDNA was prepared by reverse transcription (P. Gray et al., 1982, Nature 295: 503-508): Reverse transcriptase was used to prepare a cDNA copy of the messenger RNA to a standard. methods, the second strand was prepared (also by standard methods) by Klenow treatment, and the cDNA was treated with S1 nuclease to remove the hairpin loop. To insert this cDNA into a vector, the ends were ligated to an adapter or linker to form 5 'and 3' restriction enzyme sites, or, preferably, cohesive terminals for a predetermined restriction enzyme site. The oligonucleotide 5ΉΟAATTCATGCGTTCTTACAG was used for this purpose

GTACGCAAGAATGTC-P 5 '. The oligonucleotide was ligated into the cDNA, the cDNA was re-isolated by polyacrylamide gel electrophoresis. Xgt11, which is publicly available (or substantially equivalent to ygt11, available from the ATCC), was extracted with EcoRI to recover a linear fragment (M. Wickens et al., 1978, J. Biol. Chem. 253: 2483-2495). The combined reverse transcription product and the ygt10 digestion product were ligated together, and the fusion product was used. coli C-600 transfection or other host-sensitive γ-phage infection. An estimated 10,000 recombinant phages were inoculated into a 15 cm plate and screened by an inaccurate spot hybridization method (T. Maniatis et al., 1978, Cell 15: 687-701 and P. Gray et al., PNAS 80: 5842-5846) using P a labeled probe prepared from segment A shown in Figure 1a by the method of J. Taylor et al., 1976, Biochem. Biophys. Acta 442: 324-330 using calf thymus DNA primers (PL Biochemicals). Double nirocellulose filters were hybridized by the low stringency method, using probe at 5 x 10 beats per minute, in 20% formamide. The filters were washed twice with 0.3 M sodium chloride, 0.03 M sodium citrate and 0.1% sodium dodecyl sulfonate (SDS) at 37 ° C.

Two phages were hybridized with the probe and purified by plaque purifield. Purified phage were hybridized to both the segment A probe and the segment B probe. The cDNA inserts of the two hybridizing phages, yLT1 and yLT2, were subcloned into M13mp8 and sequenced by the dideoxy chain termination method (A. Smith, 1980, Methods in Enzymology 65: 560-580). The insert in LT2 was only 600 bp and did not contain the entire 3 'coding region for lymphotoxin. The insert in yLT1 contained the entire coding region for leucylamino-terminated lymphotoxin as well as 650 bp 3¹untranslated region (which contains a polyadenylation concordance signal) and codons for 18 amino acids that are amino-terminal to the leucyl terminus. Since this does not form the entire lymphotoxin coding region, a P-labeled probe was prepared from the yLT1 cDNA insert and used to screen an additional 25,000 recombinant ygt10 phage with high precision (see T. Huynh et al., 1984, in Practical Approaches in Biochemistry IRL Press, Oxford). Twelve hybridizing phages were isolated, and the sequence of the longest insert from yLT11 is shown in Figure 2a. The longest, open reading frame was subjected to translation initiation, from the first observed ATG. The numbers above each line indicate the position of the amino acid and the numbers below each line indicate the position of the nucleotide. The leucyl residue, labeled 1, refers to the first sequenced residue of the leucyl amino-terminal lymphotoxin (Figure 1a), and is probably the first amino-terminal residue of the mature type of lymphotoxin. The first 34 residues represent the signal sequence. Residues 156-171 were not detectable by lymphotoxin protein sequencing, but instead could be read from the nucleotide sequence.

Example 3

Construction of a synthetic gene / wild-type cDNA expression vector hybrid for leucylamino-terminal lymphotoxin

This structure is shown in Figure 2b, pLTXB1 (containing the inactive synthetic gene) was partially digested with EcoRI and PstI, and a 685 bp fragment containing DNA encoding the 125 N-terminal residue of lymphotoxin was recovered. A partial PstI digest was formed due to the addition of a PstI site at residue 10 (Figure 1a). 301 The bp fragment containing the DNA encoding the 51 C-terminal amino acid of lymphotoxin was isolated by digesting the yLT1 subcloned cDNA with EcoRI and PstI (these sites are shown in Figure 2a at nucleotide positions 554 and 855). These fragments were isolated by electrophoresis with 5% polyacrylamide and electroelution. The fragments were ligated into pBR322 extracted with EcoRI and dephosphorylated with bacterial alkaline phosphatase. 30 to reduce background transformants. The resulting expression plasmid, pLTtrp1, was characterized for appropriate orientation and sequence by restriction endonuclease digestion and DNA sequencing. The leucylamino-terminated lymphotoxin was expressed by transforming E. coli

31446 pLTrp1 and culturing the transformants in medium containing tetracycline at 37 ° C for 4-6 hours until an optical density of 1.0 was reached. Cell lysis products contained cytotoxic activity. The leucylamino terminus of the expressed lymphotoxin type was found to be substituted by a closed methionyl residue. The product of this synthesis is believed to be of the formylmethionyl rather than the methionyl type.

Example 4

Lymphotoxin Immunoaffinity Purification An anti-lymphotoxin-secreting mouse monoclonal cell line (Example 8) was grown in a mouse and purified from ascites fluid by ion exchange chromatography. The anion exchange eluate was coupled to cyanogen bromide activated Sepharose® at a concentration of 2 mg / ml resin. The 20 mL column was sequentially equilibrated with TBS (containing 0.05 M Tris-HCl, pH 7.0, 0.15 M sodium chloride, and 2 mM EDTA); then with elution buffer (containing 0.1 M acetic acid; pH 4.5, 150 mM sodium chloride); and finally with TBS. from sound-disrupted E transformed with pLTtrpl. coli cells, the precipitate obtained with 40% ammonium sulfate impregnation (previously clarified by centrifugation) was slurried in 0.1 M Tris-HCl, pH 7.4, and 5 mM EDTA and loaded onto a column at a rate of one column per hour. After thorough washing with TBS containing 0.05% Tween 20, the specifically bound material was eluted with elution buffer immediately adjusted to pH 7.8 with 0.1 volume of 1 M Tris-HCl, pH 8. 5, and stored at 4 ° C. The specific activity of this purified lymphotoxin was 2-10x10 units / mg, as determined by the mouse L-929 method described above.

The eluate contained most of the activity loaded on the column. The majority of the total protein in the eluate migrated as a single band under both reducing and non-reducing conditions, SDS-polyacrylamide47 gel electrophoresis. The mobility of this zone corresponds to a molecular weight of about 18,000, which is consistent with a predicted molecular weight value of 18,664 for non-glycosylated, leucylamino-terminated lymphotoxin, based on the deduced amino acid sequence. To further characterize its biological activity, purified lymphotoxin was tested for cytolytic activity in vitro as well as antitumor activity in vivo.

Example 5

Biological activity of recombinant lymphotoxin in vivo

Lymphotoxin from recombinant and lymphoblasts was examined in vivo by a tumor death assay. MethA (a) sarcoma was grown for 7-10 days in sensitive mice (BALB / C x C57B1 / 6fl or CB6fl), and tumors were then immediately injected with the lymphotoxin of Example 4, lymphoblast lymphotoxin (prepared and purified as described above), or control samples. 20 - After 24 hours, the mice were sacrificed, the tumors were removed and histologically evaluated for the extent of necrosis. As shown in Table 1, both recombinant and lymphoblastic lymphotoxin caused significant necrosis of MethA (a) sarcoma in vivo. Control samples did not induce necrosis of MethA (a) sarcomas.

table 1

MethA (a) necrosis induced by recombinant and wild-type lymphotoxin

Number of mice

Evaluation of Särcoma necrosis

<td colspan="2">Reading</td><td> + + +</td><td> + +</td><td> +</td><td> ---</td>

<td></td><td>Buffer 1 control Lymphoblast lymphotoxin,</td><td></td><td></td><td></td><td> 3</td>

<td> 10</td><td>25,000 units Lymfoblastilymfotoksiini,</td><td> 4</td><td></td><td></td><td></td>

<td></td><td>10,000 units Rekombinanttilymfotoksiini</td><td> 4</td><td></td><td></td><td></td>

<td></td><td>200,000 units Rekombinanttilymfotok-</td><td> 14</td><td> 2</td><td> 2</td><td></td>

<td> 15</td><td>bus, 25,000 units Rekombinanttilymfotok-</td><td> 3</td><td></td><td></td><td> 1</td>

<td></td><td>bus, 10,000 units</td><td> 3</td><td> -</td><td> 1</td><td> -</td>

<td></td><td>Bumper 2 control</td><td> -</td><td> -</td><td> -</td><td> 9</td>

<td rowspan="2"> 20</td><td colspan="2">Lymfoblastilymfotoksiini</td><td colspan="2">injected</td><td>buffer</td>

<td>1 dissolved (0.01 M Tris-</td><td>HCl,</td><td>0.05 M</td><td>&lt;nh₄&gt;₂</td><td>HCO₃, pH 8.0)</td>

and recombinant lymphotoxin dissolved in buffer 2 (0.15 M NaCl, 0.1 M sodium acetate and 0.1 M Tris-HCl, pH 7.8).

The absence of carbohydrate from recombinant lymphotoxin does not appear to affect biological activity, as the activity of lymphotoxin produced by recombinant culture (2 to 10 x 10 6 units / mg) is approximately the same as that reported for lymphoblast lymphotoxin (4 x 10 units / mg).

The activity of the recombinant lymphotoxin also expressed thermal lability similar to that of wild-type lymphotoxin, e.g., inactivation in aqueous solution, after heating at 80 ° C for one hour.

Example 6

Expression vector for constructing methionyl-histidyl-amino-terminated lymphotoxin

The structure of the plasmid that directs the expression of the methionyl-histidyl-amino-terminal lymphotoxin in E. coli is shown in Figure 3. The synthetic oligonucleotide was inserted into the expression plasmid to encode a methionine start codon at the histidyl-amino-terminated lymphotoxin of the histidyl amino-terminal lymphotoxin. in Figure 2a). This was done by isolating the 4630bp vector fragment from pLTtrpl by XbaI and Claluuto, preparative 1% agarose gel electrophoresis, and electroelution. Similarly, a 570 bp BamHI-ClaI fragment containing most of the lymphotoxin coding sequence was isolated from pLTtrp1. Two synthetic oligonucleotides were synthesized by the methods previously described and mixed with oligonucleotides 6, 7, 52 and 53 of Figure 1a. Approximately 50 picomoles of each oligonucleotide was treated with polynucleotide kinase, as described in the example. The oligonucleotides were annealed and then ligated with a mixture of a 570 bp BamHI-ClaI fragment and a 4630 bp XbaI-ClaI vector fragment. The binding mixture was transformed into E. coli ATCC 31446, and recombinants were selected based on tetracycline resistance. Plasmid p20KLT was recovered from a transformant. Plasmid p20KLT was characterized by restriction enzyme and DNA sequence analysis.

Example 7

Preparation of a cytotoxic lymphotoxin fusion variant

A plasmid containing DNA encoding a combination of lymphotoxin with a bacterial protein was constructed by cloning a sequence encoding a bacterial signal sequence located adjacent to the structural gene for lymphotoxin. The gene sequence for E. coli thermostable enterotoxin II (STII) has been characterized (RN Picken et al., 1983, Infection and Immunity 42: 269-275) and encodes a 23 amino acid signal sequence that directs STII secretion into E. coli. coli periplasmic space.

Plasmid pWM501 (Picken et al., 1983, Infection and Immunity 42/1 /: 269-275) contains the thermostable enterotoxin (STII) gene. A portion of the DNA encoding the STII gene was recovered from pWM501 using the following steps. pWM501 was digested with RsaI, and a 550 bp DNA fragment was isolated. This gene fragment was ligated into phage M13mp8 (J. Messing et al., In the Third Cleveland Symposium on Macromolecules: Recombinant DNA, published by A. Walton, Elsevier, Amsterdam / 1981 / s. 143 - 153), which had previously been broken down by Smal. Bound DNA was used to transform E. coli JM101, a commercially available strain for use with M13 phage. The result was the formation of clear spots. The double-stranded M13mp8 STII Rsa derivative was isolated from E. coli JM101 infected with this phage using standard methods (cited in J. Messing et al.). Using the M13mp8 subcloning method just described, the sets of different restriction endonucleases provided by the phage bind about 550 base pairs of fragments containing the STII leader. The M13mp8 STII RSA derivative is then digested with EcoRI and PstI and a DNA fragment slightly larger than the 550 bp DNA fragment is isolated.

The EcoRI-PstI fragment was subcloned into pBR322.

This was done by digesting pBR322 with EcoRI and PstI and isolating the vector. The isolated vector was ligated to the EcoRIPst1 DNA fragment. This DNA was used in E. coli ATCC

31446, and tetracycline-resistant colonies were selected. The plasmid was isolated from the resistant

E. coli colony and was designated the pSTII moiety.

The pSTII portion was digested with MnII and BamHI, and a 180 bp fragment was isolated containing the STII ShineDalgarno sequence, the STII signal sequence, and the mature

The first 30 codons of the STII gene. The 180 bp DNA fragment was ligated into a plasmid containing the trp promoter. One such plasmid, pHGH207-1, has been described previously (H. de Boer et al., 1982, in Promoters: Structure and Function, published by R. Rodreguez et al. Chamberlin, Praeqer Jul. New York, NY, pp. 462- 481). In this example, a derivative of this plasmid, pHGH207-1 *, was used in which the EcoRI site 5 'to the trp promoter had been converted to EcoRI * by supplementation with DNA polymerase I (DNA pol I) and joining the flank ends by binding to the trp promoter. the plasmid was digested with DNA pol I and kaii wedge with four dNTPs to complete the bulging sequence. The DNA preparation was then digested with BamHI and the vector-containing fragment was isolated. This vector fragment was then ligated to a DNA fragment containing a 180 bp STII signal isolated above. The binding20 mixture was used to transform E. coli ATCC 31446 to confer ampicillin resistance. A plasmid designated STII derivative was isolated from an ampicillin-resistant sequence.

M13 phage containing STII coding sequences was first constructed by binding a 180 bp XbaBamHI fragment of the pSTII derivative to XbaI and BamHI extracted with M13mp10. The resulting phage DNA, pSTIIshuttle, was characterized by restriction endonuclease analysis and nucleotide sequencing. The LT coding sequences were then ligated into this vector by binding a 700 bp fragment of the HpaI-EcoRI site of pLTtrp1 to the replicable (RF, double-stranded) DNA of the pSTII shuttle digested from the Sma-EcoRI site; The Smal and Hpal sites are both flat and connected (resulting in the disappearance of both sites). The resulting phage DNA, M13-STII-LT, was characterized and then used for mutation as follows: primer 5'p CAAATGCCTATGCACTGCCAGGCGTAGG was treated with kinase and mixed with template (M13-STII-LT) in ligase buffer and DNA digested from XbaI-EcoRI in the presence of JP (to promote the production of DNA As described by Adelman et al., 1983, DNA 2: 183-193); the mixture was heated to 95 ° C and then allowed to cool slowly at room temperature for 30 minutes, and then placed on ice for 30 minutes. All four deoxynucleotide triphosphates were then added together with ATP, T4 DNA ligase, and a large fragment of E. coli DNA polymerase I (Klenow). The mixture was incubated for one hour at 14 ° C and then used to transfect the required E. coli JM101, a commercially available strain, or another host of M13 phage. Properly mutated phage were identified by hybridization screening using a p-radiolabeled primer as a probe. The resulting phage ST-LT-mut was characterized by DNA sequence analysis. A replicable form of DNA was prepared from this phage and used to isolate a 761 bp XbaIcoRI fragment containing DNA for the STII signal sequence adjacent to the coding sequence for the leucylamino-terminated lymphotoxin. This DNA was ligated with XbaI-BamHI digested p20KLT (large 4285 bp vector fragment) and the 375 bp EcoRI-BamHI fragment of pBR322. The resulting plasmid, pST18LT, was characterized by restriction mapping and DNA sequencing. A similar structure encoding the amino-terminal attachment of the STII signal to the histidine residue of histidylamino-terminated lymphotoxin was prepared. Plasmids pSTLT18 and pSTLT16 are formed. Using restriction enzyme analysis and dideoxy sequencing, they were confirmed to encode STII fusion. E. coli transformed with plasmids pSTL18 or pSTLT16 synthesized STII signal sequence fusions with leucylamino-terminal and histidylamino-terminal lymphotoxin, which was determined by gel electrophoresis to be consistent with the calculated molecular weight. E. coli cell lysis products containing these fusion proteins had cytolytic activity.

Example 8

A method for producing a mouse monoclonal antibody

Purified lymphoblast lymphotoxin obtained in Example 1 was dialyzed against phosphate buffered saline (PBS). The dialysate contained 200 ug of lymphotoxin per milliliter. Glutaraldehyde was added to the dialysate to a concentration of 70 mM glutaraldehyde, the mixture was incubated for 2 hours at room temperature, glutaraldehyde was added to a total concentration of 140 mM, incubation was continued for another 6 hours, and then the mixture was dialyzed against PBS. 50 yug cross-linked lymphotoxin (hereafter referred to as polylymphotoxin) and 0.5 ml of Freund's adjuvant were injected subcutaneously into mice (strain BALB / c). After one week, mice were booster immunized with 50 μg of polylymphotoxin and 0.5 ml of Freund's incomplete adjuvant, half intramuscularly and half intraperitoneally.

Serum was collected after 7 days and assayed for antilymphotoxin activity by ELISA.

The ELISA assay was performed as follows: A buffered solution of purified lymphotoxin was placed in microtiter chambers and each chamber was allowed to cover with approximately 100 ng of lymphotoxin. The unadsorbed lymphotoxin solution was aspirated out of the chambers. 50 the sample was combined with 100 μl of appropriately diluted test / μl PBS containing mg / ml bovine serum albumin (PBS-BSA buffer) and added to each chamber, incubated for 2 hours at room temperature, washed with PBS containing

0.05% Tween 20, 100 μl of horseradish peroxidase-labeled goat anti-mouse IgG in PBS-BSA buffer li93025 was placed in each chamber and incubated for one hour. Each chamber was washed with PBS containing 0.05% Tween 20 and citrate phosphate buffer, pH 5 containing 0.1 mg O-phenylenediamine / ml (substrate solution), and 30% aqueous H 2 O 2 (ratio 4 / A 30% v / v list of H 2 O 2 in 10 mL of substrate solution) was added to each chamber. The chamber was incubated for 30 minutes, the reaction was stopped with 50 μl of 2.5 M sulfuric acid, the adsorbence was measured at 492 nm. Chambers with an adsorbance greater than 1 OD were considered anti-lymphotoxin-positive.

Experimental samples were also determined for their ability to neutralize the cytolytic activity of lymphotoxin in an experiment with mouse L929. Serum was collected from immunized animals, or hybridoma supernatants were dissolved in the required RPMI-1640 medium containing 10% fetal bovine serum and approximately 100 units of lymphotoxin per milliliter and plated in microtiter chambers containing the same number of cultured L929 cells. In the control, all cells were lysed. A neutralizing antibody was detected as a failure of lymphotoxin to lyse L929 cells.

Animals immunized with glutaraldehyde-polymerized lymphotoxin generated antibodies that were active in the ELISA assay, but no serum neutralizing activity was observed.

A suspension was prepared containing 100 photoxins and 1 ml of 1.64% w / v / ug lymaluminium hydroxide / 1Al (OH)₃suspension and used to immunize the same mouse. The mouse was injected intramuscularly

100 with the suspension.

and intraperitoneally 400 [mu] l of the same

After one week, the mouse was injected intravenously with 10 ug of unpolymerized and tymphoblast lymphotoxin 100

An experiment performed three days later adsorbed / ul in PBS.

1/80 dilution of animal serum, indicated the presence of a lymphotoxin neutralizing antibody.

The spleen was removed from this animal. 3 x 10 spleen7 cells were fused with 5 x 10 mouse myeloma cells and inoculated into mitotiter chambers containing

With HAT medium and about 3,000 peritoneal macrophages / microtiter chamber, according to the method of S. Fazekas De St. Groth, 1980, J. Immunol. Meth. 35: 1 - 21. Hybridoma, from chambers containing 10 supernatants positive in the above ELISA assay, grown in ml of DMEM medium containing 20% ​​fetal calf serum, 10% NCTC-135 medium, - cx 10 M beta-mercaptoethanol and HAT, were plated in microtiter chambers, statistically averaging one cell per chamber, and then cultured in a volume of 1 or 5 ml of the same medium. The supernatants were then assayed for neutralizing antibody. In the ELISA assay, approximately 2% of positive hybridoma cells20 obtained from aluminum hydroxide immunization synthesized a neutralizing antibody. From this group of hybridoma cells, a lymphotoxin antibody with an optionally high affinity was selected.

'25 Example 9

Mutation site-specific lymphotoxin mutation

The procedure of Example 3 was strictly followed in this example, except that segment 6 of the synthetic oligonucleotide-30 was modified to obtain the sequence

5¹CTCAACTCTGCACCCA3 'and its complementary strand (segment 53) were modified so that its sequence became

3'AGACGTGGGTCGTCGT5 '.

The modified oligonucleotides were annealed to the remaining oligonucleotides and ligated into the expression vector as described in Example 6. This vector contains a 2 bp substitution that changed the lysine + 28 codon from lysine to histidine. The histidine mutant is expressed in E. coli ATCC 31446 transformation.

Other locally determined mutants were prepared in a similar manner, preferably by codon selection, so that no EcoRI restriction site was added, which would require partial EcoRI restriction digestion in the pLTXB1 digestion required in Example 3. Mutations 10 should also not add new XbaI or BamHI sites. to fragment B or HindIII or BglII sites to fragment C. Partial digests otherwise require a suitably mounted pLTXB1 mutant; complete digestion would produce a deletion mutant rather than the substitution mutant targeted in this case.

Example 10

Identification of gene DNA encoding mouse and bovine lymphotoxin; amino acid sequence of mouse and bovine lymphotoxin

Mouse and bovine lymphotoxin genes were isolated from gene Ä libraries. The human lymphotoxin cDNA fragment (PvuII-EcoRI, 600bp) was radiolabeled by P-nick translation, and a mouse gene DNA λ library (M600 strain gene DNA in ÄCharon4A, T. Manitias et al., Molecular Cloning) was used as a probe. , p. 31, 1982) and, independently, for screening a bovine gene DNA library (EP 88622A).

Hybridization was performed with low accuracy in 20% formamide (Gray and Goeddel PNAS USA 80: 5842-5846 (1983)), and the filters were washed twice with 0.3M aqueous sodium chloride, 0.03M sodium citrate, and 0.1% SDS: Several phages hybridized to the human lymphotoxin probe were subjected to plaque purification (T. Maniatis et al., Cell

15: 687-701 (1978), was digested with restriction endonuclease 35 Ia and analyzed by Southern hybridization. The mouse

The 3500 bp EcoRI fragment of DNA and the 2200 bp EcoRI fragment of bovine DNA were hybridized with a human lymphotoxin probe. These DNA fragments were subcloned into plasmid pBR322 and then sequenced by the dideoxy chain termination method (AJH Smith Methods in Enzymology 65: 560-580 (1980)). Figure 4 shows the deduced protein sequence of mouse and bovine lymphotoxin compared to human lymphotoxin.

Example 11

Lymphotoxin expression in yeast under the control of the ADH promoter

Plasmid pLTrpl is digested with XbaI to open the plasmid from the XbaI site, just proximal to the lymphotoxin start codon. The XbaI cohesive ends are flattened with the Klenow fragment of E. coli DNA polymerase II with four dNTPs. EcoRI adapter

OH

5' ^UAATTCCCGGG 3 ' _rGGGCCC-P 5 '

The 3 'OH is ligated to a flat plasmid fragment, the protruding 5' hydroxyl termini are phosphorylated using a polynucleotide kinase, the binding mixture is used to transform E. coli ATCC 31446, and restriction analysis identifies plasmid pLTtrplR1, which contains an additional EcoR1 site proximal to the primer. Plasmid pLTtrplR1 is isolated, digested with EcoRI, and the lymphotoxin DNA-containing fragment SP is recovered.

Plasmid pFRPn (EP 60 057A) is digested with EcoRI, treated with alkaline phosphatase to prevent ring remodeling, ligated to the SP lymphotoxin fragment using T4 DNA ligase, and then the binding mixture is used to transform E. coli ATCC 31446. Ampicillin-resistant colonies produce two sets of plasmids with the SP insert oriented in the opposite direction to that determined by restriction analysis on agarose electrophoresis gels. Plasmids are purified by E. colitrans and are used to transform a yeast with a trp1 mutation (e.g. yeast strain RH218, unrestricted ATCC Accession No. 44076) trp⁺phenotype. Plasmids oriented so that the start codon of segment SP is located adjacent to the alcohol dehydrogenase promoter fragment are found and used to transform yeast for lymphotoxin expression. Lymphotoxin is recovered from yeast transformant extracts. The stability of plasmids in large-scale fermentations can be improved by using an expression plasmid that contains a 2 micron start of replication instead of the beginning of pFRPn chromosomal replication (ars 1) and is compatible with the host strain (J. Beggs, 1978, Nature 275: 104-109).

Example 12

Lymphotoxin expression in mammalian cells XLT11 (Example 2) is digested with EcoRI and a lymphotoxin-containing DNA fragment (reverse transcription product) is recovered. Plasmid pEHER (EP 117 060A) is digested with EcoRI, treated with calf intestinal alkaline phosphatase and ligated into an EcoRI-linked reverse transcriptase product of XLT11. The resulting plasmids are grown in E. coli ATCC 31446 (EP 117 060A) and labeled pEHERLT I and pEHERLT II. They contained lymphotoxin DNA in the opposite orientation to that determined by restriction analysis on polyacrylamide gels. These plasmids are used to transfect and select CHO DHFR-DUX-B11, CHO 1 and Ltk cells.

Tissue culture cells are transfected by mixing / μg pEHERLT I or pEHERLT II prepared above with 10 μg rat carrier DNA in a volume of 250 μl of 0.25 M CaCl 2, followed by the addition of pisa59 in portions of 250 HEPES-buffered saline (280 mM MaCl, 1.5 mM Na 2 O, 0.5 mM HEPES, pH 7.1). After 30 minutes at room temperature, the solution is added to tissue culture cells grown in 60 mm plastic tissue culture dishes. CHO 1, CHO DHFR-DUX-B11 and Ltk cells are used. The plates contain 3 ml of medium suitable for the host cell.

A suitable medium for CHO 1 and CHO DHFR-DUX-B11 cells is Ham F-12 medium (Gibco) supplemented with% calf serum, 100 μl / ml penicillin,

100 yug / ml streptomycin and 2 μM L-glutamine. Ltk^-The appropriate medium for the cell line is Dulbecco's modified Eagle's medium (DMEM) supplemented as described above.

After 3-16 hours, the medium is removed and the cells are washed with 20% glycerol prepared in phosphate buffered saline. Fresh medium is added to each plate and the plates are incubated for an additional 2 days.

Selection of transfected cells is performed by treatment of cells with trypsin after 2 days of culture (including treatment of cells

0.5 mg / ml sterile trypsin containing

0.2 mg / ml EDTA) and by adding about 3 x 10 $ cells to 10 mm tissue culture plates with selective medium, for dhfr cells, the medium is a formulation of medium (F-12 GIBCO) lacking glycine, hypoxanthine and thyrine ( GHT medium): DHFR⁺for cells, methotrexate (100-1000 mM) is added to normal medium. Controls are run using plasmid-free transfection conditions and plasmid pFD-11 (EP 117 060A) containing normal DHFR. Colonies derived from cells picked and expressing the DHFR plasmid are visible within 1-2 weeks. Transformants expressing mature lymphotoxin are identified.